planning under pressure, third edition (urban and regional planning series)

Planning Under Pressure

DEDICATION

to MARI (Friend) and JUDITH (Hickling)

whose tolerance, inspiration and support have been so important to us over allthe years while we have been working on successive editions of this book.

They too have had to plan under pressure in managing the work/life balancesof their own creative careers, and in meeting the ever-changing demands ofour growing families while we have so often been absent or preoccupied.

John FriendAllen Hickling

Planning Under PressureThe Strategic Choice Approach

Third edition

John FriendSheffield, UK

Allen HicklingWarwickshire, UK

Plus a new chapter containing invited contributionsfrom 21 users

AMSTERDAM • BOSTON • HEIDELBERG • LONDON • NEW YORK • OXFORDPARIS • SAN DIEGO • SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO

Elsevier Butterworth-HeinemannLinacre House, Jordan Hill, Oxford OX2 8DP30 Corporate Drive, Burlington, MA 01803

First published 1987Second edition 1997Third edition 2005

Copyright© 2005, John Friend and Allen Hickling. All rights reserved

The right of John Friend and Allen Hickling to be identified as the authors of this workhas been asserted in accordance with the Copyright, Designs andPatents Act 1988

No part of this publication may be reproduced in any material form (includingphotocopying or storing in any medium by electronic means and whetheror not transiently or incidentally to some other use of this publication) withoutthe written permission of the copyright holder except in accordance with theprovisions of the Copyright, Designs and Patents Act 1988 or under the terms ofa licence issued by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road,London, England W1T 4LP. Applications for the copyright holder’s writtenpermission to reproduce any part of this publication should be addressedto the publisher.

Permissions may be sought directly from Elsevier’s Science and Technology RightsDepartment in Oxford, UK: phone: �+44� (0) 1865 843830; fax: �+44� (0) 1865 853333;e-mail: permissions @ elsevier.co.uk. You may also complete your request on-line via theElsevier homepage (http://www.elsevier.com), by selecting ‘Customer Support’and then ‘Obtaining Permissions’.

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

Library of Congress Cataloguing in Publication Data

A catalogue record for this book is available from the Library of Congress

ISBN 0 7506 63731

For information on all Elsevier Butterworth-Heinemann publicationsvisit our website at http://books.elsevier.com

Typeset by Integra Software Services Pvt. Ltd, Pondicherry, Indiawww.integra-india.comPrinted and bound in Great Britain

Working together to grow libraries in developing countries

www.elsevier.com | www.bookaid.org | www.sabre.org

Contents

Authors’ preface to the third edition vii

Foreword to the third edition ix

Authors’ preface to the second edition xi

Authors’ preface to the first edition xv

Foreword to the first edition xix

Aquick access guide xxi

1 Foundations 1

2 Working into problems 23

3 Working towards decisions 43

4 Orientations 67

5 Skills in shaping 89

6 Skills in designing 119

7 Skills in comparing 147

8 Skills in choosing 179

9 Practicalities 211

10 The electronic resource 245

11 Extensions in process management 257

12 Invention, transformation and interpretation 273

13 Learning from othersBy 21 invited contributors: see overleaf for names 295

14 The developmental challenge 361

Access to further information 369

Bibliography 373

Index 379

v

List of Contributors to Chapter 13

13.1 Jonathan Rosenhead, London School of Economics, UK 298

13.2 Knut Strömberg, Chalmers University, Sweden

Jaan-Henrik Kain, Chalmers University, Sweden 303

13.3 Frans Evers, Environmental Policy Director, The Netherlands 308

13.4 Richard Ormerod, Warwick Business School, UK 312

13.5 Arnold van der Valk, Wageningen University, The Netherlands

Gerrit Jan Carsjens, Wageningen University, The Netherlands 315

13.6 Dave Friend, Software Design Consultant, West Yorkshire, UK 319

13.7 Alessandro Giangrande, Università Roma Tre, Italy

Elena Mortola, Università Roma Tre, Italy 322

13.8 Rebecca Herron, University of Lincoln, UK

Dennis Finlayson, International Development Consultant, UK 327

13.9 Richard Harris, Independent Process Consultant, UK 331

13.10 Rob Angell, Environmental Policy Consultant and Facilitator, UK 336

13.11 Mike Cushman, London School of Economics, UK

Alberto Franco, Warwick Business School, Coventry, UK 340

13.12 Elisenda Vila, Planning Consultant, Venezuela

Ana Maria Benaiges, Planning Consultant, Venezuela 345

13.13 Brendan Hickling, Independent Facilitator and Mediator, UK 349

13.14 Jackie Phahlamohlaka, Rural Education Pioneer, South Africa 353

13.15 Leny Bregman, Environmental Project Manager, The Netherlands 357

vi

Authors’ preface to the third edition

In 1996, when we drafted our authors’ preface to the second edition of this book, we wouldhave been very surprised by any suggestion that we might find ourselves again invited to cometogether to produce a third edition in the early years of the twenty-first century.We are, of course, delighted that interest in the strategic choice approach and its applications

has continued to grow, spreading to new parts of the world and new generations as fresh areasof application continue to emerge. So we welcome the opportunity to add some new content,and to review the way in which we had presented our concepts and methods in earlier editions.Both of us have now passed normal retiring age, and have started to limit our commitmentsaccordingly. So the question arose: were we ourselves necessarily the ones who should bewriting about these new developments? Or should we now invite others to join us in presentingthe learning points arising from their recent experiences, and speculating about what furtherdevelopments the next few decades will bring?

CONTRIBUTIONS FROM OTHERS

Finally we agreed to invite several of our more recent associates, and also some whomwe haveknown for many years, to make short contributions to a new chapter, entitled Learning from

Others. Fifteen contributions in all are brought together in our new Chapter 13. They presentthe views of twenty-one authors and co-authors, about half of them from our own country andhalf from other parts of the world. Among the key points that come through are the following:

• It is both possible and worthwhile to be inventive in combining the methods presented inthis book with complementary methods of interactive working from other toolboxes – and toreflect deeply on the outcomes so that lessons from these experiences can be widely shared;

• The guidelines to the management of extensive projects offered in Chapter 11 offer newand practical solutions to the development of agreed positions in important fields of nationaland indeed trans-national policy. This has been demonstrated not least in the important andcomplex domain of environmental policy, which different stakeholders tend to approach fromopposing and sometimes deeply entrenched positions;

• In parallel, the potential has now been widely demonstrated for introducing the principlesand methods of the Strategic Choice Approach quickly and informally into more informal andlocalised arenas of decision-making such as community development, urban regenerationand rural capacity-building, where manifold external policy influences can impinge in intricateand unpredictable ways;

• It is a matter for celebration rather than concern that several people who have found theStrategic Choice Approach helpful in their own worlds of decision-making have introducedinterpretations and transformations of their own, to increase its acceptability in the variousdecision-making cultures in which they work.

vii

Authors’ preface to the third edition

OTHER CHANGES

To pave the way for the new multi-authored Chapter 13, we have made some adaptationsto the coverage of the two newer Chapters 10 and 11 that we added in our second edition.Chapter 10 reflects the experiences of John Friend in developing software to support strategicchoice since the first edition appeared, while Chapter 11 reflects Allen Hickling’s experiences inadapting the approach to themanagement of extensive projects. Also, we have inserted a linkingchapter – Chapter 12 – in which we take stock of the variety of changes in presentation andterminology that have been introduced by other people – and sometimes by ourselves. Eachchapter from 11 onwards begins with a synopsis in bold typeface, to highlight its contributionto this edition.After the introduction of the twenty-one new voices in Chapter 13, we return in our closing

Chapter 14 to look afresh at the developmental challenge that was posed in the final chapter ofour second edition – recognising that the world around us continues to change in ways to whichwe and our successors must learn to fashion responses in practical yet creative ways.In general, the additions in this new edition have been more concerned with developments

in practice than in theory. Co-incidentally, however, the international journal Planning Theory

is publishing a special issue (Volume 3, No. 3, November 2004; Mandelbaum, ed.) just as ournew edition appears, in which four sets of invited contributors from Europe and North Americapresent appreciations of the contributions of our work to the development of planning theory.All these papers emphasise our uniquely close link with planning practice, and the issue endswith an invited response from John Friend.

A COMPANION WEBSITE

Advances in technology, since our second edition appeared, now offer a new solution to thefamiliar challenge of keeping the content of any book such as this one up to date. The companionwebsite that our publishers have made available for Planning under Pressure has been designednot only for use by students, but also as a forum inwhich a broader dialoguewith users, research-ers and consultants can be allowed to develop in a flexible, spontaneous and sustainable way.We shall have to see how this new channel develops; all we can do now is to do our best to giveit a flying start. The web reference is http://books.elsevier.com/companions/0750663731.

A NEW FOREWORD

We could think of nobody more appropriate to invite to write a short foreword for this thirdedition than Arnold de Jong, our long-standing Dutch associate who has worked alongside us onmany assignments in Europe and who, through his extensive facilitation practice, has demon-strated repeatedly how the skills and methods presented in this book can generate additionalconfidence and wider support for important development decisions at local, regional, nationaland international levels. Our hope is that his example will be followed by many talented andinventive younger people as our new twenty-first century unfolds.


April 2004

viii

Foreword to the third edition

Throughout my 45-year working life, which spanned successive careers as an agricultural engin-eer, a corporate information manager, a senior local politician and a decision process consult-ant, I have had an ever-increasing concern for interactive participation processes. In spite ofwidespread criticism that they are not effective, it has always been my belief – backed by myexperience – that, skilfully managed and facilitated, the opposite is true.Over my last two decades as a consultant, facilitator and trainer, Strategic Choice has played

a central and indeed a guiding role. So I can confirm from long experience that not only doesit offer a fresh and relevant approach to complex management and planning tasks in theory; italso delivers in practice. It does so at many levels from that of enhancing democratic action inlocal communities to that of consensual policy development in the European Union. It has noparallel in building agreement between seemingly opposed stakeholder groups across nationaland disciplinary frontiers, and across those between the government, business and voluntarysectors. In the mid-1970s, I was becoming drawn into the intricacies of local politics from myposition as Alderman of the municipality of Arnhem in The Netherlands, with the portfolio ofTown Planning and Urban Renewal. Issues of participation and democratic process in planningwere then coming strongly to the fore, and we were looking for new ideas to help us turn theminto reality. In 1976, I led a visiting team from Arnhem on a visit to the offices of the Institute forOperational Research in Coventry, having heard of their innovatory approach and the influenceit was starting to have on public planning in the UK.This first meeting with John Friend and his colleagues of the ‘IOR School’ was to mark a

turning point in my professional career. Then, from 1980, through the 1980s and 1990s, I had theprivilege of working closely with Allen Hickling. We became immersed, jointly and separately, ina succession of demanding projects to tackle daunting issues of environmental, economic andsocial policy, in each case working interactively with as many of the stakeholders as possible.Evidence of what has now been achieved will be found scattered throughout the pages of thisbook. In particular it will be found in the newChapter 13which brings together contributions frompeople in many countries who have recently been extending the frontiers of Strategic Choice innew and promising directions. From my twenty-first century retirement home in Athens, I lookback on more than two decades in which the philosophy and methods of the Strategic Choiceapproach have provided the central thread of my consulting career, helping to turn the ideals ofdemocratic planning into reality.Planning under Pressure is not a theoretical study. From my 20 years working as a facilitator

using the concepts of strategic choice, I can witness to the practical value of this book. The factthat it now appears in a third edition, with many new contributions, holds much promise for thedecades ahead.

Arnold de JongDecision Process Consultant and Facilitator

April 2004

ix

Authors’ preface to the second edition

NEW DIRECTIONS

Our original intention had been to keep our preface to this new edition short. For the first editionincluded an invited foreword and a lengthy preface, both of which we wished to retain; and, inthe reader’s interest, we wished to avoid a surfeit of introductory material. Yet, once we cameto draft the additional chapters for this edition, we recognised that there were many things wewished to say which fitted better here in a new preface than in the body of the book.Over the decade since our first edition went to press, both of us have been developing the

ideas presented in Planning under Pressure in new directions, while continuing to work asindependent consultants with different clienteles. So, inevitably, there have been divergencesin the directions in which the two of us have been moving. Yet, whenever we meet to compareexperiences, we discover intriguing opportunities for synthesis and it is these that have guidedus in drafting the additional chapters for this edition.In brief, we have agreed that we need make only minor changes in the first nine chapters.

However, it was clear to us that the final Chapter 10 of our first edition, concerned with horizonsfor future development as we saw them in the mid-1980s, had now become quite out of date.What we have therefore done is to replace that chapter by three new chapters which, takentogether, reflect our experiences over the last decade, and the consequent shifts in our view ofopportunities for future development.The newChapter 10 – The electronic resource – discusseswhatwe have learnt from our recent

experiences in developing computer software for strategic choice, and in introducing this as anadditional resource both for decision-makers and for teachers and students of management andplanning.Then the new Chapter 11 – Extensions in process management – reviews a complementary

set of experiences in adapting the participatory style of planning discussed in Chapter 9 to evermore demanding challenges; challenges in which the groups involved have tended to becomelarger and more diverse, in terms of both their cultural backgrounds and the interests whichthey represent. Here, the concern has been to develop practical ways of helping people to workcreatively with each other, as much as with the complexities of the issues that they face.In the new Chapter 12 – The developmental challenge – we stand back and review the wider

implications of these two directions of development – one leading towards smaller-scale explor-ations using more tightly structured problem-centred methods, the other towards larger-scaleinteractions usingmore loosely structured people-focusedmethods. Howwidely, we ask, shouldthe boundaries of the strategic choice approach, as presented in this book, now be drawn? Andhow does our approach relate to other recent developments in decision support and participatoryplanning throughout the world?

xi


OUR CONTRASTING EXPERIENCES

Chapter 10 is based largely on the experiences of John Friend over the last decade, whileChapter 11 is based largely on those of Allen Hickling. It does not seem appropriate for us todwell on issues of historical development in those two chapters themselves, so it is importantthat we should say at least a little here in this preface about the principal influences at work.For John Friend, the development of computer software for strategic choice has been a major

programmatic concern since this was identified as a priority in the final chapter of our firstedition. Meanwhile, he has continued to be involved – more intermittently than Allen Hickling –in various facilitation and consultancy projects in which the computer has played little or no role.The decision to start developing software for strategic choice was triggered in 1987 when JohnFriend found that he shared this interest with his former Tavistock colleague John Stringer, thenliving not far away, and they worked together on this in the early stages. The development workhas subsequently been sustained through close collaboration between John Friend and his sonDave, working for him in a software development role.Originally, the software – which has been named Strategic Adviser, or STRAD for short –

was seen as primarily a means of making the philosophy of the strategic choice approach moreaccessible to individuals and to small informal meetings, where the organisational arrangementsinvolved in setting up aworkshopwith expert facilitation cannot readily be justified. However, thesoftware development project has also turned out to havewider implications for the developmentof the strategic choice approach.On the one hand, it has led to some significant extensions and refinements of the specific

problem structuring methods introduced in this book; on the other hand, it has helped to spreadawareness of the strategic choice approachwithin other domains of policy inwhichwe ourselveshad had little direct facilitation experience at the time when our first edition appeared – not least,within the domains of industry and commerce, as opposed to the world of public policy.While Allen Hickling’s work as a facilitator has also taken him into a diverse range of contexts,

themain thrust of his work has been the development and co-ordination of major programmes ofwork in important fields of environmental policy, in local, national and – increasingly – internationalsettings. These programmes had their roots in his earlier work in facilitating policy-making inthe Netherlands in such areas as management of toxic wastes, estuarial pollution and transportand storage of hazardous petrochemical feedstocks – work which is reflected in several of theillustrations from practice that appear in Chapters 5, 6, 7 and 8.Since the publication of our first edition, this programmatic interest has led him first into

process management roles in the shaping of the first National Environmental Policy Plan forthe Netherlands, then into an ambitious programme of projects in which representatives ofgovernment, industry and community interests from different nations of the European Unionhave come together to harmonise their policies for various types of environmentally sensitivewaste. These experiences have in turn led to work with a similar orientation in the Baltic statesand other eastern European countries.Meanwhile, in Britain, a developing relationship with the Environment Council – a national

organisation concerned with promotion of environmental awareness and co-operation – hasdrawn Hickling into facilitation and mediation roles in the management of more acute issues ofenvironmental conflict resolution, in which the stakeholders often set out from deeply opposedpositions. In the process, he has worked alongside other consultants with expertise in suchfields as negotiation, mediation and consensus-building.

xii


EXPANDING NETWORKS

Over the last decade, the range of people who have become exposed to the principles andmethods of the strategic choice approach has been expanding in several directions. This hascome about not only through direct collaboration with one or other of us in project work but alsothrough publications – and significantly through the publication of the first edition of this book.We find it particularly encouraging when we come across consultants, planners, managers andacademics with whom we, as authors, have had little or no personal contact, yet who havedeveloped enough interest in the methods, and enough confidence in their ability to use them,to start applying them to situations in their own contexts without any reference to ourselves.The result is that, in our extended Guide to further reading,1 we are able to point to a numberof published applications of strategic choice methods by other people working in such variedfields as third world development, local community action and information systems strategy incommercial enterprises.Although some of their accounts are tantalisingly brief, others are more extensive. Taken

together, they indicate not only that strategic choice methods have now been successfully usedin a growing range of contexts – but also, significantly, thatmany users have sought to explore thescope for combining themwith other approaches to participatory planning or problem structuringthat have already demonstrated their value in their respective fields.In discussing this trend towards synthesis with thework of others, our new concluding chapter

speculates about the potential for recognising a new generic field of developmental decision sci-ence, differing in its orientation from the broad field of systems sciencewith its various schools –but potentially of comparable significance for coping with complexity in human affairs. Thisleads to a review of future prospects and opportunities in terms of eight interlocking themes –research, methodology, facilitation, communication, technology, sponsorship, application andeducation. Among the opportunities we review here is that of developing computer softwarefor use in what is sometimes called a ‘groupware environment’ – with use of the fast-expandingcapabilities of the Internet to overcome constraints of space and time. Although much of theinitial momentum here has so far been technology-driven, the scope is now becoming clear forintroducing a more process-oriented approach.Our concluding review leads us – as in the concluding section of our first edition – to the

educational challenge, now seen as part of a broader developmental challenge. How can learningof the type presented in our book be made as accessible as possible to new generations ofdecision-makers, so that they can build on it in their own ways? Here, both of us can draw onrecent training experiences of our own – in the case of Hickling, in running facilitation courses forthe Environment Council and other clients, and in the case of Friend in contributing to coursesin the management of sustainable development, in his capacity as honorary professor with theCentre for Development Planning Studies at the University of Sheffield.We both recognise the scale of the longer-term challenge for facilitation skills are not easy to

develop in students whose life experience is so far limited; yet academic staff will only be able tohelp develop such skills to the extent that they can draw upon a reservoir of first-hand facilitationexperience in their own cultural context. We are aware of scattered successes in meeting thiskind of challenge in many parts of the world. One of our wider aspirations is to help in building abroader cross-cultural momentum in this direction – not only through this book but also throughan intensifying web of collaborative projects covering research, development and exchange ofexperiences on a global scale.

1 Now transferred to the companion Planning under Pressure website.

xiii


The authors’ preface to our first edition – reprinted here – contains a long list of acknowledge-ments to people from whom we had drawn inspiration at that time. We could now extend thatlist considerably. However, we shall refrain from doing so here, because many of the newernames will be found in our revised Guide to further reading or in our expanded section on Pointsof contact.2

If the last 10 years provide any guide, we can look forward to a continuing expansion anddiversification of these global networks, through electronic and other channels, into the start ofthe new millennium and well beyond.


1996

2 Both now on the companion Planning under Pressure website.

xiv

Authors’ preface to the first edition

HISTORICAL PERSPECTIVE

Twelve years have passed between the conception and completion of this book. For a book con-cerned with Planning under Pressure, that may seem a surprisingly long time. But the 12 yearsof gestation have seen much in the way of development both in the practice and the theoryof the general approach to planning about which we write. Over this period, we have foundourselves collaborating with users in many kinds of organisations, public and private, throughoutthe world. So several thousands of managers, planners and policy-makers have now becomeexposed to the strategic choice approach; and there are hundreds of these who have playeda part alongside us in its development. There have been many interim publications too; someof them reporting on particular research and application projects, others emanating from train-ing programmes designed to introduce the essentials of the approach to prospective users inparticular countries.The origins of the approach – and of this book – are to be found in thework of IOR – the Institute

for Operational Research. IOR was formed in 1963, as a unit of the Tavistock Institute of HumanRelations in London, as the result of a joint initiative on the part of the Councils of the TavistockInstitute and the national Operational Research Society. Its dual aims were to extend the realmof application of operational research towards broader policy issues, and to build stronger linksbetween OR and the social sciences. These are aims that have remained to the fore throughmany subsequent organisational changes, with the impetus now being maintained through anextensive network of individuals and groups in several parts of the world.The first book on the strategic choice approach was published in 1969 (Friend and Jessop,

1969/77), followed tragically soon by the death of its co-author Neil Jessop, the first directorof IOR. The first experimental applications of the approach to practical planning problems wereconducted in 1970, in collaboration with six teams of British local government officers; andthe first training courses for managers were launched in 1971 at a Coventry hotel. Many othercolleagues from the IOR Coventry and London offices had become involved in these early exper-iences. Then, in 1973 we were also joined in Coventry by Ray Bunker – the contributor of ourforeword – who was revisiting the county of his birth on sabbatical leave from the University ofSydney in Australia. As a professional planner, he readily agreed to make it his task to visit asmany as possible of the planners and other professionals who had taken part in our experimentin application 3 years earlier, in order to discover what influence, if any, the experience had hadon the organisations and individuals taking part.At that stage, it appeared that the impact on individuals had been generally more substantial

than that on organisations. But the extent of that impact was both variable and elusive; and ifdissemination was to proceed further, an obvious next aim was for us to produce a readable,accessible ‘how to do it’ guide to the approach. So, the understanding at the end of 1973 was

xv


that the three of us would work on this task together. Soon, however, the time came for Bunkerto return to Australia – sailing by the long sea route, with prospects of plenty of writing time onthe voyage. Then time scales became extended and, gradually, the idea of Bunker remaininga co-author became a less practical one – to be replaced by the idea of his providing someintroductory remarks from his varied experiences as planning practitioner, consultant, teacherand researcher. Meanwhile, another prospective co-author had emerged: Alan Sutton, an IORcolleague who became closely involved with the two of us in developing training programmesin Canada, and subsequently in a major government-financed project in Britain to apply strategicchoice methods to the exploration of policy alternatives in County Council Structure Plans. Butthen Sutton too receded as a prospective co-author when, in 1977, he moved to a new base inWestern Canada; so, the responsibilities of authorship reverted to the two of us.

SHIFTING PRESSURES

Work on this book continued during the later 1970s – but in a sporadic way as we were bothworking under high pressure on IOR consulting and research projects for clients in Britain andoverseas. For one of us, Hickling, the thrust continued to be on practical applications of thestrategic choice approach; but for the other, Friend, it shifted towards research on the organisa-tional and inter-organisational dimensions of complex planning processes, bringing a contrastingperspective to our training activities and our continuing work on the manuscript of this book.The late 1970s were, for both of us, a difficult time in terms of continuity of our project work.

In 1980, Hickling set up as an independent consultant; Friend continued to work part-time atthe Institute while also taking up an Honorary Senior Visiting Fellowship at the ManagementCentre of the University of Bradford. Around this time we were both becoming immersed inquite different ventures as well. Hickling, having recently relinquished the management of thevillage stores and post office adjoining his home, launched a company called Endless Games,throughwhich to enter the burgeoningmarket of fantasy role-playing games.Meanwhile, Friend,from a new home location in West Yorkshire, found himself working in partnership with his wifeto set up a countryside interpretive centre, as an initiative in environmental education run onsmall business lines.For both of us, the involvement inwork on planning processes continued; andwith it our efforts

to bring the book to completion. The members of Pergamon’s advisory committee for the Urbanand Regional Planning Series – of which Friendwas a long-standingmember – offered a judiciousblend of encouragement and exhortation, supported by the editorial staff. We met togetherwhenever we could, usually at least once amonth, to progress thewriting work. Thesemeetingstook place in all kinds of locations – not only in offices but in hotels and restaurants, at motorwayservice stations, in airport lounges and railway buffets. We met often at our respective homes;indeed, we have photographs of our working one sunny day in an English country garden, withflip charts hung among the greenery climbing up the walls of the cottage behind. Meanwhile,our children grew up and started to go their separate ways; and our wives continued to tolerateour joint writing endeavours with surprising good humour, while developing their own careersin their respective fields of creative art.The breakthrough finally came in the latter half of 1985. By this time, we had both disposed

of most of our other entrepreneurial interests and Friend was again working full time from theTavistock Centre in London. Hickling was now fully stretched in some challenging applicationsof the strategic choice approach for Dutch governmental agencies, while Friend was becomingdrawn into running strategic choice workshops in new fields ranging from community healthplanning to information technology strategy within the firm. A high point came during the new

xvi


year break in January 1986, when a conjunction of circumstances allowed a brief reunion, atHickling’s home, of key people who had been associated with the earlier stages of preparationof the book including both Ray Bunker and Alan Sutton, who was now resident again in Europe.This was not only a convivial occasion, it also saw a significant step forward in the consolidationof our ideas about technology, organisation, process and product, as presented in Chapter 4.There were still to be six agonising months of meetings, long telephone calls and redraftingsbefore we were finally able to commit our finished text to the publishers. It was far from easy,but the sense of relief was overwhelming.

COMPLEMENTARITIES

Our different experiences and work patterns over the 12 years, along with different and com-plementary personal skills, have led us to recognise differences and complementarities in ourrespective contributions to the writing process. For much of the time, the main load of draftingand co-ordination has fallen on one of us, Friend, as and when pauses in the pressures to main-tain a continuing flow of project work have allowed. Yet the endeavour has been a joint one,which neither of us could have brought to fruition without the other. For Hickling’s immersionover this period in practical applications of the strategic choice approach provided an all-importantbase of experience against which to judge the realism of the advice we wished to offer and themost practical way of presenting it; and the interdependence of our contributions became moreand more apparent during the final nine-month period of intensive collaboration in the writingprocess.We found during this period that some significant differences had developed between us on

matters of emphasis and terminology; and we had to work long and hard at these before reach-ing agreement on simple, practical ways in which they could be overcome. We recognised toothat there were differences in our styles of presentation – written, verbal, graphical – but weagreed that these stylistic differences could be a source of strength rather than weakness, ifonly we could achieve a creative synthesis between them. Some clues to the complementaritiesbetween our perspectives can be found in our respective biographical notes. Among the manyfacets of his early work experience, Hickling admits to operating as a semi-professional magi-cian; Friend, meanwhile, admits to having graduated in the abstract discipline of mathematicsbefore embarking on his early career in industrial operational research. So, a background ofmagic comes together with a background in logic; a contrast which, at first sight, seems to sumup neatly enough the main differences in our backgrounds and their influence on our respect-ive styles. But the potential for creative collaboration would not have existed had we not beencapable of meeting each other at least half way. For, in the late 1960s, when Friend was strug-gling to adapt his ingrained belief in rationality, quantification and logical rigour to the untidysocial and political realities of decision-making on Coventry City Council, Hickling was taking hispostgraduate degrees at the University of Pennsylvania, to become both Master of Architectureand Master of City Planning. It was here that he became exposed to the influence of Ackoff andothers in the field of operational research; and it was through the convergence of this experi-ence with Friend’s searing experiences of decision-making in Coventry – together with a sharedinclination towards use of graphics in expressing ideas and their relationships – that the basisfor a productive collaboration was forged.The logic/magic tensions surfaced repeatedly when Friend’s writing tended to become

laboured, in the attempt to pin down more formally aspects of the strategic choice approachwhich had hitherto developed in quite an intuitive way. Sometimes, this led to proposed changesin terminology or technique which did not fit well with Hickling’s evolving base of experience

xvii


in the field; so we found we had to make many fine adjustments: a little more magic here, alittle more logic there. In retrospect, the opportunity which the writing task has provided for usto consolidate, review and modify the concepts and methods of the strategic choice approachhas been an important one for both of us. The hope now must be that the fruits of this labourwill be of as much value to our readers, whether they be practitioners, students, teachers orresearchers. The demand for an authoritative, practical guide to the strategic choice approachhas been expressed to us often enough since 1973 and, indeed, earlier.Our hope now is that this volume will succeed in meeting the demand and thereby help in

sustaining the momentum of application, teaching and development of ideas. If the past is anyguide, the practice of strategic choice is likely to continue to evolve in different ways in differentplaces in response to different demands and pressures. So, we hope that this book can providea significant milestone in maintaining the wide-ranging collaborative endeavour on which theadvance of the strategic choice approach has been built over the last 12 years.

ACKNOWLEDGEMENTS

There have been enough references to other people already in this preface for us to have gonesome way towards the important and congenial task of acknowledging the contributions by oth-ers in thiswork. The contribution of the lateNeil Jessopwas seminal and has been acknowledgedmore fully in the second edition of Local Government and Strategic Choice (Friend and Jessop,1969/77). The far-reaching contributions of Raymond Bunker and Alan Sutton over the 12-yearwriting period have already been mentioned in earlier sections. Another major contribution tobe acknowledged is that of Andreas Faludi who, from his Chair in Planning at the University ofAmsterdam, has persistently sought to encourage development of the comparatively neglectedacademic perspectives of the strategic choice approach, and the study of its relatedness to otherbodies of planning theory.Others who have contributed to the development of the approach over the years includemany

past and present members of IOR and Tavistock Institute staff – Eric Trist, Hugh Murray, PaulSpencer, JohnStringer, John Luckman,DonBryant, JamesMorgan,Michael Luck, Chris Yewlett,Ken Carter, Hadley Hunter, John Pollard, Brian Quarterman, Gloria Overton, Martin Elton, DavidMillen, Peter Spink, Michael Floyd and Michael Norris – the last of whom has contributedmany incisive comments in recent years. Other significant collaborators and sources of ideasin Britain and elsewhere have included Russell Ackoff, Felix Wedgwood-Oppenheim, JonathanRosenhead, Peter Fishburn, John Power, Fritz Scharpf, Bill Ogden, Harry Lash, Robin Fried,Arnold de Jong, Luc Wilkin, Fernand Debreyne, Hans Mastop, Angela de Melo, Robert Glass,Nathaniel Lichfield,Morris Hill, Martin Payne, Doug Spencer, DonMiller, ObboHazewinkel, BramBreure, Paul de Jongh, Fernando Galvão, Maximino Losehiavo de Barros, Moacyr Parahyba, KenBowen, Peter Bennett, Colin Eden, Sue Jones, Stephen Cropper, Christine Huxham, Jim Bryantand Tsunekazu Toda. But even to list the academic, international and other affiliations of theseand other individuals would take up too much space – let alone to describe the rich and variednature of their contributions.Contributors to the typing of successive chapter drafts have also been numerous over the

12 years since this book was conceived – but special acknowledgement must go to Betty Fox,for long the key resource person in IOR’s Coventry office; to Ann Jamieson at the TavistockCentre; and to Jayne Moore who, working from her home, was finally able to commit the textto disk in a way that we could scarcely have conceived in the dim and distant days of 1973.

xviii

Foreword to the first edition

When Neil Jessop’s and John Friend’s seminal work Local Government and Strategic Choiceappeared in 1969, I was teaching in the Department of Town and Country Planning at the Uni-versity of Sydney. I was so impressed that I reviewed it for the then Journal of the AustralianInstitute of Planners, and it has had a marked influence on my work since.I had the profit and pleasure of spending a sabbatical year with John Friend, Allen Hickling

and their colleagues in 1973. This book appears many years afterwards, and it is enlightening toturn back to my impressions of the strategic choice approach at that time. These impressionsfollowed not only from extensive discussions with the authors of this present book, but alsofrom a programme of visits to most of the planners, managers and others who had then begunto use the approach. We generally agreed that the process of strategic choice would benefitfrom being presented as more cyclic rather than linear and sequential, should be extended toaddress policy questions of major significance, and could be applied and used in fields of activityother than urban and regional development. This present book shows howmuch developmentalwork has taken place in these directions since the early 1970s.It would be a pity if the comprehensiveness and thoroughness of this book led to the neg-

lect of opportunities to use the strategic choice approach in partial or informal ways. This isparticularly important in working situations where it is difficult to use the approach deliberatelyand deliberatively, because policies and problems have to be shaped and addressed through adiffuse process of negotiation with many different people and groups. One major example ofthis style of working was the joint Commonwealth-States study of soil conservation in Australiawhich I co-ordinated as a Commonwealth public servant in the mid-1970s.This study had become static and rigidly programmatic. It was dominated by the current tech-

nology of soil conservation and by construction of capital works to arrest land degradation withlittle consideration of any national interests or priorities. The study had been in progress for 2 or3 years and, in the circumstances of that time, it was not feasible for me to introduce strategicchoice explicitly to all the various groups of inter-governmental officers who were involved indifferent ways. But the study was able to conclude with a principal recommendation, agreedto by all parties, about the need for mutual commitments to raise the level of soil conserva-tion effort. This, of course, was supported by a series of statements about what that meant interms of substance and priorities. These conclusions were then supported by a series of subsid-iary recommendations which defined the principles of resource allocation needed to support anenhanced soil conservation effort; the organisational requirements of this expanded programme;and the dynamics of its continued development and modification. Inter-governmental relationswere a particularly important part of this operation, and the recommendations were structuredto express these.In effect, the study was changed from one dominated by the heavily structured characteristics

of traditional planning, towards an emphasis on the qualities of the strategic choice approach as

xix

Foreword to the first edition

expressed in this book. This soil conservation study accordingly represents an example of theapproach being introduced informally but effectively to address the crucial aspects of processand organisation in policy-making and programme/project development and operation.In an educational context, the strategic choice approach can be used in quite comprehen-

sive and explicit ways. At the most ambitious level, I believe that it can be used as the majorstructuring element in the design of courses in town and country planning. More pragmatically,I have used it both as a means of illuminating different philosophies of planning, and as a vehiclefor problem-solving in planning exercises. Either it can be taught carefully and comprehensivelyalong the lines shown in this book; or students can be thrown into the deep end in dealing witha planning situation, after only a brief introduction to strategic choice. In a recent exercise ofthis latter kind, at the South Australian Institute of Technology, I defined the broad attributesof a planning situation, divided students into three groups and asked them to develop differentsolutions. Each group acted as a professional planning group advising the local council andtaking the problem through a series of progressive decisions over three or four months. Toencourage their imagination and to save time on the laborious collection of data and information,I asked them to write up the exercise as three different short stories, inventing information alongthe way which was supportive of, and consistent with, the progressive series of decisions ofdifferent kinds. The three answers showed the leading importance, respectively, of cash flow tothe development agency; of an opportunity to accommodate a major metropolitan showgroundfacing relocation; and of the resolution of land use conflicts with adjoining activities. Along theway the students learned a lot about the roles and relationships of decision-makers and decision-takers.The way strategic choice is used in teaching depends on the educational environment. In the

example cited above, the students were in the third year of an undergraduate course: But ideallythe students should be introduced to the concepts of strategic choice in the first year, and theapproach built up throughout the course.Finally, I believe the stage of development of strategic choice in this book is not the final

one. One of its greatest contributions has been to break down the rigidities attending planning,problem-solving and policy-making. I feel that too many people see implementation of plans andpolicies as simply the routine carrying out of decisions. Yet, I am convinced that aspects andinstruments of implementation often need to be shaped, adapted, or accepted as given, rightfrom the beginning of addressing a problem. Otherwise, we will continue to have too manyineffective policies and too many pigeonholed plans.

Raymond Bunker

xx

A quick access guide

How should you start reading Planning under Pressure? Your choice will depend on the natureof your interests; the time at your disposal; and the extent of any prior familiarity you may havewith the strategic choice approach.The purpose of this quick access guide is to help you in making your decisions about selective

reading. It does so first by outlining the principles on which the book is designed, then by makingsome suggestions to help you establish your priorities.

THE FIGURES

One of the first things you will notice on flicking through the pages of this book is the numberand positioning of the illustrations. There are 102 full-page figures, all boldly numbered. Somefigures are professionally drawn, while others are drawn freehand – deliberately so, in order tostress the spontaneous way in which the methods are usually used in practice. Other figuresagain take the form of photographs, illustrating the approach in use in a workshop setting.Above and below most of the figures will be found various key words and symbols. These are

designed to help the reader in making rapid cross-references within the text. Their meaning willbe explained in the next section.

THE STRUCTURE OF THE BOOK

Three of the fourteen chapters – Chapters 1, 4 and 9 – present the main characteristics of thestrategic choice approach at a general level:Chapter 1 describes its Foundations, which are based on first-hand experience of the chal-

lenges facing decision-makers in group situations; Chapter 4 draws out the general Orientationsand shifts of attitude which are central to this approach; Chapter 9 discusses the Practicalitiesof applying the approach in practice, based on cumulative experience over more than 30 years.Both Chapters 4 and 9 are organised with reference to a general view of four complementary

aspects through which any approach to planning can be described and compared to others.These four aspects, expressed through the mnemonic A-TOPP, are as follows:

Approach : Technology

Organisation

Process

Product

In most of the figures of these two chapters, you will see that the aspect of the approachwhich is currently in the foreground is highlighted in the lower left-hand corner.

xxi


The intervening chapters are grouped into two sets:

1. Chapters 2 and 3, which together introduce the basic concepts and methods of the strategicchoice approach;

2. Chapters 5, 6, 7 and 8, which expand on these basic concepts andmethods, emphasising thepractical skills involved in applying them to the complexities of decision-making in practice.

The structure of these six chapters reflects a general view of four complementary modes ofdecision-making. This view is presented in diagrammatic form in Figure 8. The four basic modesare:

Shaping Designing Comparing Choosing

In these six chapters, a small motif of four circles, based on the framework of Figure 8, appearsin the corner of each figure to identify the mode which is currently in the foreground. Throughoutthese chapters the concepts, methods and skills are introduced gradually through the develop-ment of a case example – the South Side story – in which they are applied to a semi-fictitiousyet realistic situation of Planning under Pressure in the public domain.In this third edition, the main sequence of Chapters 1–9 is followed by five further chapters

presenting significant developments since the first edition was published.Chapter 10 describes briefly – with illustrations – progress in developing computer software

as an additional form of support in applying the approach in practice, making it more accessibleto individuals and small informal management groups. Chapter 11 describes some importantextensions in process management, which have helped in adapting the approach to large andchallenging projects.Whereas Chapters 10 and 11 revise chapters that were first presented in our second edition

of 1997, the final three chapters are new to this edition. In Chapter 12 we describe some of themany ways in which other people have interpreted and sometimes transformed the toolbox forstrategic choice to fit different contexts of decision-making. Then Chapter 13 brings togethertwenty-one invited contributors in reporting their experiences and the lessons that they drawfrom them. Finally, Chapter 14 reviews the altered horizons of future development as they appearat the time this edition goes to press.

READING PRIORITIES

The guidelines that follow are organised according to the results you may reasonably expect toobtain from different reading strategies. They are presented broadly in terms of increasing levelsof comprehensiveness.For a first quick appreciation of the approach, we recommend a fast reading of Chapter 1, then

a skim through the figures of the next two chapters – glancing at the definitions of key conceptsthat appear below the figures in Chapters 2 and 3. This can be followed by reading through thereview of the main orientations of the approach, as summarised at the end of Chapter 4.For a grasp of the main principles, concepts and methods, we recommend a fuller reading

of the first four chapters. To reinforce your understanding, you then have the option of tryingout the short exercises which appear at the end of Chapters 2 and 3. If you are interested ina clearer understanding of the challenges of applying the approach in practice, you can thenskim through Chapter 9; and, if you are interested in using the computer as a tool, you can thenbrowse through Chapter 10.

xxii


For a more thorough appreciation of how you can apply the approach, we recommend firsta familiarisation with the concepts introduced in Chapters 2 and 3, then a browse throughChapters 5, 6, 7 and 8, using Chapter 4 as your point of departure. Having developed a feelfrom these chapters of the kinds of skill which are involved in applying the approach in practice,we suggest you now read Chapter 9 – especially the section on selectivity and adaptiveness.Then you can either proceed to Chapter 10 or to Chapters 11 and 12, for a fuller appreciation ofdifferent contexts of application; alternatively you can return to work through Chapters 5–8 inmore depth.For a fuller feel of the realities of using the strategic choice approach in practice, we recom-

mend selective reading of the fifteen contributions from other people that are brought togetherin Chapter 13, together with some of the snapshots of people at work that appear towards theend of Chapters 5, 6, 7, 8 and 12; then a study of the four management checklists relating totechnology, organisation, process and practice which are presented in Figures 83, 85, 86 and 88of Chapter 9, and the further developments in process management described in Chapter 11.

LEARNING BY DOING

As with any other approach to planning, one cannot expect to prepare oneself for all the chal-lenges one may encounter in practice through reading books alone.Our hope is that you will feel encouraged by what you read to try putting the principles,

concepts andmethods of the strategic choice approach into practice in your owndecision-makingand planning, whether you already carry significant responsibilities for decisions that impinge onother people, or whether you a student on the threshold of your career.

THE WEBSITE

Rather than end up with a Guide to Further Reading as in our earlier editions, it seems moreappropriate in this electronic era to transfer most of this information to the companion website tothis book, where it can be continuously revised and extended as further experiences continue toaccumulate, and as the lessons from those experiences continue to be shared. The companionwebsite will be found at http://books.elsevier.com/companions/0750663731.

xxiii

1 Foundations

A PHILOSOPHY OF PLANNING

There are many ways in which it is possibleto approach the challenge of planning in anuncertain world.The approach to be introduced in this chapter

is one in which planning is viewed as a con-tinuous process: a process of choosing stra-tegically through time. This view of planningas a process of strategic choice is, however,not presented as a set of beliefs which thereader is expected to embrace uncritically atthis stage. That would be too much to expect –especially of an introductory chapter, which isintended merely to open the door for the morespecific concepts, methods and guidelines tobe offered in those that follow. People involvedin any kind of planning activity of course buildup their own sets of beliefs about the prac-tice of planning in the course of their work-ing lives: beliefs which they will not wish toset aside lightly. Yet experience in applyingthe approach offered here has shown that itsfundamentals can usually be accepted withoutmuch difficulty by those planners or managerswhose working philosophy draws more ontheir own practice than on taught beliefs. Thisis because, in essence, the approach sets outto do no more than to articulate, as clearlyas possible, the kinds of dilemma that experi-enced decision-makers repeatedly face in thecourse of their work, and the often intuitivejudgements they make in choosing how torespond.In practice, such judgements may some-

times be accompanied by a sense of discom-fort or even guilt. For the decision-makers mayfeel they are departing from certain principlesof rational behaviour which they have been

taught to respect. Indeed, the view of planningas strategic choice is found to offer more of achallenge to such idealised principles of ration-ality than it does to the intuitive judgementsand compromises that seem characteristic ofplanning practice. If this point can be accep-ted, the reader should be able to relax in fol-lowing the ideas put forward in this chapterand view them as offering perspectives thatcan help make sense of current practice –without necessarily demanding any revolution-ary change in familiar ways of working.

THE CRAFT OF CHOOSINGSTRATEGICALLY

It is important to emphasise that the view ofstrategic choice presented here is essentiallyabout choosing in a strategic way rather thanat a strategic level. For the idea of choos-ing at a strategic level implies a prior viewof some hierarchy of levels of importance indecision-making; while the concept of stra-tegic choice that will be developed here ismore about the connectedness of one decisionwith another than about the level of import-ance to be attached to one decision relative toothers.It is not too surprising that these two senses

of the word strategic have tended to fusetogether in common usage. For it is often themore weighty and broader decisions whichare most obviously seen to be linked to otherdecisions, if only because of the range of theirimplications and the long time horizons overwhich their effects are expected to be felt.This, in turn, can lead to a view that any processof strategic decision-making should aspire tobe comprehensive in its vision and long range

1

FIGURE

1Planning Under Pressure: A View of the Realities

FOUNDATIONS

Foundations

in its time horizon, if it is to be worthy of itsname.But such a view of strategic choice can

become a restrictive one in practice; for it is alltoo rarely that such idealistic aspirations can beachieved. The approach to strategic choice tobe built up in this chapter is not only aboutmak-ing decisions at a supposedly strategic level.It goes beyond this in addressing the makingof any decisions in the light of their links toother decisions, whether they be at a broaderpolicy level or a more specific action level;whether they be more immediate or longerterm in their time horizons; and no matter whomay be responsible for them. This concept ofstrategic choice indicates no more than a read-iness to look for patterns of connectednessbetween decisions in a manner that is select-ive and judgemental – it is not intended to con-vey the more idealistic notion that everythingshould be seen as inextricably connected toeverything else.So this view of planning as a process of stra-

tegic choice implies that planning can be seenas amuchmore universal activity than is some-times recognised by those who see it as a spe-cialist function associated with the preparationof particular sorts of plans. At the same time,it allows planning to be seen as a craft, full ofsubtlety and challenge; a craft through whichpeople can develop their capacity to think andact creatively in coping with the complexitiesand uncertainties that beset them in practice.

ORGANISATIONAL CONTEXTSOF STRATEGIC CHOICE

This relatively modest interpretation of theword strategicmeans that the view of planningas strategic choice is one that can be appliednot only to decision-making in formal organisa-tional settings, but to the choices and uncer-tainties which people face in their personal,family and community lives. For example, anyof us might find ourselves involved in a pro-cess of strategic choice in addressing the prob-lem of where and when to go on a holidaynext year, or how to sell an unwanted vehicle,

or how to deal with a difficult request froma relative or friend. Of course, the craft ofchoosing strategically becomes more complic-ated where it involves elements of collectivechoice – of negotiation with others who viewproblems and possibilities in different ways.Indeed, most of the more demanding prob-lems to which the strategic choice approachhas been applied have involved challengesof collective decision-making, either in organ-isational or inter-organisational settings; andthis can have the effect of blurring many ofthe familiar distinctions of task and discip-line around which organisational structures areusually designed. For the skill of choosing stra-tegically through time is one that can becomejust as essential to the manager or executiveas to those in more formal planning roles. Thispoint is illustrated schematically in Figure 1,through which is presented a view of planningunder the practical pressures of organisationallife. It is a view in which an organisation’sarrangements for making plans and those formaking day-to-day decisions, tend to mergetogether into a less clearly bounded processthrough which progress is sustained. This isa process of choosing strategically in copingwith difficult problems, amidst all the complexrealities – or perceptions of reality – whichcontribute to organisational life.The larger and more complex the organisa-

tion, themore it is to be expected that decision-making responsibilities will have becomedifferentiated according to a multitude of oper-ational, managerial or entrepreneurial roles.The more likely it is too that specialised plan-making functions will have been developed inan effort to maintain a co-ordinated, longer-term view isolated from everyday manage-ment pressures. However, no plan-makingactivity will remain valued within an organ-isation unless it can provide support for themore difficult and important of the decisionspeople face; and it is a common experiencethat carefully prepared plans can quickly losetheir relevance under the pressures of day-to-day events. The combined pressures of

3


urgency, competition for resources and turbu-lence in the world outside can soon lead todisenchantment and confusion in the arrange-ments for making plans; while the pressuresof complexity, conflict and overload can leadto vacillation and inconsistency in the mak-ing of day-to-day decisions. To counter theresulting personal and organisational stresses,those responsible for organisational guidancesometimes look towards some over-archingframework of policies or aims. But, in prac-tice, such policy guidelines can often be dif-ficult to agree – especially when working ininter-organisational settings – and their contri-butions towards sorting out the predicamentsof day-to-day management can be disappoint-ingly small.The making of generalised policies is there-

fore given its place in Figure 1; but it is notgiven pride of place. Instead, the emphasis ison the more subtle process of making pro-gress through time by choosing strategically;and on the creative management of multipleuncertainties as a crucial means towards thisend. And progress through time can itself takemany forms. Immediate progress can take theform of intervening, or negotiating with others,as well as taking decisions on matters wheredirect action is possible. Meanwhile, progressin building a base for later decisions can alsotake different forms – not only investigationsbut also clarification of values and cultivationof working relationships with other decision-makers.So the term ‘planning’ will be used in this

book to refer generally to this more looselydefined process of choosing strategically, inwhich the activities of making plans, decisionsand policies can come together in quite subtleand dynamic ways. But with a wide variety ofways of making progress to be considered, theprocess can soon begin to appear as one notso much of planning but of scheming – to intro-duce a term which has a similar literal mean-ing but which carries very different undertonesin its everyday usage. Whereas the notion ofplanning may invoke a sense of idealism and

detachment, the notion of scheming tends tosuggest working for sectional advantage in anoften devious way. So there is a case to bemade that people involved in planning mustlearn to become effective schemers; and fur-thermore that it is possible to exercise schem-ing skills in a responsible way. Those whoare troubled about social responsibility in plan-ning – including both the authors of this book– may wonder whether there must alwaysbe a divide between responsible planners andirresponsible schemers; and, if so, whether itmust always be the latter who will win. Theconcept of responsible scheming need not beconsidered a contradiction in terms. Indeed,it is towards the search for a theory and apractice of responsible scheming that the stra-tegic choice view of planning can be said to beaddressed.It is, however, one thing for an individual to

embrace a philosophy of planning as strategicchoice; and quite another thing for a groupof people working together to share such aphilosophy as an unequivocal foundation fortheir work. Experience has shown that thereare some settings where a sense of sharedphilosophy can indeed emerge – either wherea set of close colleagues have learnt to worktogether as a coherent team, orwhere they dis-cover that a common professional backgroundallows them to proceed on shared assump-tions as to how decisions should be made.Yet those whose work involves cutting acrossorganisational boundaries must expect oftento find themselves working alongside peoplewith whom they do not share a philosophicalbase. So it is important to think of the philo-sophy presented in this chapter as a helpfulframe of reference in making use of the morespecific concepts and methods to be intro-duced in this book, rather than as a necessaryfoundation from which to build.Indeed, it is a common enough experience,

when working with strategic choice concepts,that people of quite diverse backgrounds canmake solid progress towards decisions basedon shared understandings, with little or noexplicit agreement at a more philosophical

4

Foundations

level. Often it is only through the experienceof working together on specific and immediateproblems that they find they are beginning tobreak through some of the philosophical barri-ers which may have inhibited collaboration inthe past.

DILEMMAS OF PRACTICE

The view of strategic choice presented inthis book gained its original impetus from theexperience of a particular research project,which offered unusually extensive opportunit-ies to observe the kinds of organisational pro-cesses indicated in Figure 1.The setting of this research was the muni-

cipal council of amajor English city – Coventry –which, between 1963 and 1967, agreed to actas host to a wide-ranging project on the pro-cesses of policy-making and planning in localgovernment, viewed as a microcosm of gov-ernment as a whole. This seminal researchwas supported by a grant from the NuffieldFoundation, and has been more fully repor-ted elsewhere (Friend and Jessop, 1969/77).Over the 4-year period, the research team wasable to follow a wide range of difficult issuesincluding the review of the city’s first devel-opment plan; the redesign of its urban roadnetwork; the reorganisation of its school sys-tem; the renewal of its housing stock; thefinance of public transport; and the schedul-ing of capital works. The researchers wereable to hold many discussions with the vari-ous politicians, administrators, planners andprofessional experts involved, and to observethe processes of collective decision-making inwhich they came together – not only in thedepartmental offices and the formal meetingsof Council and its committees, but also in thesmoke-filled rooms of the opposing politicalgroups.Through these experiences, some impres-

sions of the persistent dilemmas of decision-making in such complex circumstancesgradually came to the fore. Among the clearestimpressions were:

• that people held different and continuallyshifting views about the shape of the issuesthey faced and, not least, about how closelyor widely the boundaries of their concernshould be drawn;

• that there were persistent pressures forthem to arrive at commitments to action inan incremental or piecemeal way, howevercommitted they might be in theory to theidea of taking a broader, more comprehens-ive view of the issues before them;

• that there was a continuing dilemma ofbalancing urgency against uncertainty indecision-making through time; and

• that there were persistent difficulties in dis-tinguishing the technical from the polit-

ical aspects of the decision process, eventhough the entire organisational structurewas built around the maintenance of distinc-tions of this kind.

These impressions of the practical difficultiesof choosing strategically in organisations facingcomplex problems have been strengthenedand extended bymany other experiences sincethe conclusion of the Coventry project – notonly in the world of local government but inother public sector organisations, in industryand commerce, in voluntary organisations, andin the increasingly wide range of problemsituations where these different domains ofdecision-making tend to converge. On thestrength of this broader experience, a view ispresented in Figure 2 of the five broad dimen-sions in which difficult choices of balance tendto arise in the management of a continuingprocess of strategic choice. There is a choicebetween:

1. a more focused and a more synoptic treat-ment of problem scope;

2. amore simplifying and amore elaboratingtreatment of complexity;

3. a more reactive and a more interactive

treatment of conflict;4. a more reducing and a more accommod-

ating treatment of uncertainty;5. and a more exploratory and a more decis-

ive treatment of progress through time.

5

FIGURE

2Judgements of Balance in Strategic Choice

FOUNDATIONS

Foundations

The practical task of choosing a position ineach of these five dimensions is not one ofmaking a firm and lasting commitment to oneextreme or the other. It is more a task of main-taining an appropriate balance in continuallychanging circumstances, shifting from time totime in one direction or another, according tothe – often intuitive – judgements of thoseinvolved. In the chapters that follow, the pic-ture presented in Figure 2 will be used as apoint of reference in building more structuredframeworks of ideas through which to expandfurther on the view of planning as a process ofstrategic choice. These frameworks will givedeeper significance to the various contrastswhich, at this stage, can only be indicated inoutline terms.In later chapters, fuller interpretations will

be offered of other related aspects of thedilemmas of practice observed in Coventryand elsewhere, which are not brought outso clearly in the comparatively broad set ofbalances presented in Figure 2. In particular,later chapters will have more to say aboutthe issues of urgency and incrementality, andabout the relationship of the political arenato the technical domain. This is a dichotomywhich, in Coventry City Council, could be seenas the fundamental organising principle onwhich the formal structures of accountabilitywere designed; but it is a relationship with farwider implications for decision-making, even incontexts where such distinctions may becomemore blurred.

RESPONDING TO DIFFICULTYIN MAKING DECISIONS

The view of planning as a process of strategicchoice is, above all, a dynamic one. However,in building up a view of the way this processworks, it is useful to begin with a more staticpicture. This picture, which is quite simple yetalso quite general in its application, has asits focus any situation in which one or moredecision-makers are experiencing difficulty inchoosing how they should act in response tosome particular decision problem with which

they are currently concerned. A snapshot viewof such a decision situation is presented inFigure 3. The decision problem itself isdepicted as a cloud to indicate that itsshape will often be in some degree obscure.However, what makes it problematic to thedecision-makers is that they are experiencingsome pressure to arrive at a decision, yet it isnot clear to them what course of action theyshould choose.Where a group of people find themselves

collectively in such a situation, then it is oftenfound that different members of the groupwill advocate different ways of responding;so some degree of conflict of opinion mayemerge. Three types of response which arerepeatedly offered in practice are indicated bythe three different ‘bubbles’ shown emergingfrom the central cloud in Figure 3.Very often, people will see the way out of

their present difficulties in terms of explor-ations of a more or less technical nature.The suggestions offered typically include vari-ous forms of costing or forecasting exercises,surveys, technical analyses, research stud-ies; or, in some circumstances, proposals forinvestment in more ambitious forms of math-ematical or economic modelling. Whatever theform of investigation, however, the purpose isto reduce the difficulties of making decisionsby investing in a process of exploration into par-ticular aspects of the decision-makers’workingenvironment about which it is felt that too littleis currently known. Other people, meanwhile,may see the way out of the difficulty in termsof other, less technical, kinds of explorationdesigned to establish more clearly what policyvalues should guide their choice of action. Typ-ically, they may call for investment in activit-ies designed to clarify goals, objectives, aimsor policy guidelines, whether through formalor informal channels. In some situations, thismay mean simply consulting decision-takerswho bear more direct responsibility for organ-isational policy; in others it could mean deliber-ately seeking fuller involvement in the processby a range of affected interest groups or theirrepresentatives.

7

FIGURE

3Three Types of Uncertainty in Decision-Making

FOUNDATIONS

Foundations

A third response is to seek the way out ofthe difficulty by moves to extend the currentagenda of decision-making concern. Peopleadvocating this response will often argue thatthe decision problem currently in view is onethat cannot realistically be addressed in isola-tion, because it is connected to one or moreother decision problems which lie ahead. Sothe demand here is likely to be for some formof co-ordination, negotiation or planning exer-cise that will allow the current decision prob-lem to be explored alongside others within abroader, more synoptic problem focus. Eachof the three kinds of demands – most typic-ally expressed as demands for more inform-

ation, for clearer objectives and for more

co-ordination – can be regarded as a differentkind of attempt to manage the current stateof uncertainty over what should be done aboutthe current decision situation. Indeed, it is pos-sible to go on to identify three general cat-egories of uncertainty along the lines indicatedbelow, which are distinguished by the differentforms of response that can be made. Thesethree types of uncertainty play an importantpart in the philosophy of planning as a pro-cess of strategic choice; they can be formallydescribed as follows:

1. Uncertainties about the working Environ-ment: UE for short;

2. Uncertainties about guiding Values: UV forshort;

3. Uncertainties about Related decisions: UR

for short.

It is important to stress that the idea of uncer-tainty in strategic choice is normally viewedin relative rather than absolute terms. It istreated as an attribute of particular situationsand people rather than something with anobjective reality of its own. In practice it isoften far from easy for people to agree whichof the three kinds of uncertainty are most cru-cial in a particular decision situation; and, there-fore, how much attention should be given toeach possible form of response. For instance,members of a city planning team, considering

whether to recommend approval of an applic-ation to build a new hotel, might see possibil-ities either for calling for deeper investigationof its traffic implications; or for seeking clearerguidance on the Council’s policies in relation tothis particular kind of development; or for ini-tiating a wider review of tourism possibilitieswithin the city as awhole. Theymight of coursewant to move in all three directions more orless at the same time; however, this is notalways possible where there are pressures tomake a speedy decision. Nor will it necessarilybe desirable to invest resources in all possibleways of responding to uncertainty – especiallyif some of them are expected to be less effect-ive than others, in terms of reducing the feel-ings of uncertainty among the decision-makersinvolved.

MANAGING UNCERTAINTY:A DYNAMIC VIEW

So, in practice, it may be far from easy tojudge how uncertainty is to be managed at anymoment, even in situations where the sourcesof that uncertainty have been clearly identified.To consider further the possible ways

of managing uncertainty through time, itbecomes necessary to move to a moredynamic view. Such a view is presented inFigure 4, which builds on the ‘snapshot’ pic-ture of Figure 3 by introducing the realitythat any form of investigative, policy clarify-ing or co-ordinating initiative must take sometime to carry through. Indeed, explorations insome of these directions may, in practice, takelonger to carry out than others. However, theintended consequence of pursuing any chosenexploratory path is to make the decision situ-ation less difficult to deal with once the out-come of the exploration is known – in otherwords, to lessen the feelings of uncertaintybeing experienced by the decision-makers, andthus to increase the level of confidence withwhich they can act. In practice, however, itwill not often be realistic to expect that thefeelings of uncertainty surrounding a difficult

9

FIGURE

4Opportunities for Managing Uncertainty through Time

FOUNDATIONS

Foundations

decision problem can be made to vanish alto-gether; however much effort may be investedin exploratory activities. In terms of the symbol-ism used here, the process can be pictured asone whereby the original cloud becomes smal-ler in its dimensions and, by implication, lessobscure.Sometimes, of course, feelings of uncer-

tainty may be reduced through time withoutany conscious action on the part of thedecision-makers. Expected events may or maynot unfold; trendsmay becomemore apparent;the intentions of other partiesmay be revealed;policy positions may become more clearcut. In general, however, uncertainty can onlybe reduced at a cost – whether this be merelythe cost of delay when there may be urgentissues to be settled, or whether it also includesmore direct costs in terms of money, skills orother scarce resources. So themanagement ofuncertainty through time is rarely simple in thetypes of judgement it entails. It is the raisingof these judgements to a more conscious levelthat is one of the most distinctive characteris-tics of the strategic choice approach.

INTERCONNECTED AGENDASOF DECISION-MAKING

Of the three exploratory routes indicated inFigure 4, it is the co-ordinative (UR) routewhich is of most far reaching significance indeveloping the idea of planning as strategicchoice. The demand to move in this directionarises when there is a sense that the presentagenda of decision-making is too restricted –that the decision problem currently in view issignificantly influenced by uncertainties to dowith intended actions in other fields of choice.Such a concern for a wider view will oftenlead to an extension in the time frame aswell, because the pressures for decision maybe less immediate in some of these relatedareas. The concern for co-ordination may alsoshift the process in the direction of some formof liaison or joint working with other sectionsor departments, and sometimes, also, with

other decision-makers quite outside the organ-isational framework within which the currentproblem is being addressed.The concern for co-ordination in dealing with

related fields of choice does not, however,inevitably mean transcending organisationalboundaries in thisway. At amoremodest level,it may simply be amatter of the same decision-maker recognising that an issue to be dealtwith today should be considered in relation tosome other issue to be dealt with nextweek. Inthe case of the hotel development mentionedearlier, for instance, it could be that a proposalto develop an indoor leisure centre is known tobe pending on a neighbouring site, suggestingthat either proposal could affect the other.In general, the pursuit of the co-ordinative

(UR) route implies forging a relationshipbetween one decision process or planning pro-cess and others, in the manner indicated inFigure 5. The dynamic viewhere is taken a stepfurther than in Figure 4, by showing the fullerimplications of a shift from a more limited toa broader decision focus. The investment in‘more co-ordination’ can be seen as shiftingthe focus, temporarily at least, from the originaldecision problem to a broader and more com-plex problem within which it is contained.

INCREMENTAL PROGRESSIN PLANNING

One of the most important points about thisshift to a broader problem focus is that it doesnot automatically mean that those involvedshould be aiming for early decision in respectof all the related choices now brought intoview. It is perfectly possible that the shift to abroader focus will help to reduce uncertaintyin the original decision problem and so enablefirm commitment to be agreed, without lead-ing to simultaneous commitments in any or allof the other related areas. Indeed, the issue ofbalance between exploratory and decisive pro-gress has already been highlighted (Figure 2) asone of themain areas of judgement in strategicchoice; and it is a balance of critical importancein managing uncertainty through time.

11

FIGURE

5Extending the Problem Focus

FOUNDATIONS

Foundations

The broader the focus of the problem withinthe larger cloud in Figure 5, the more it islikely to be thought of not simply as a decisionproblem but as a planning problem becauseit contains elements of both immediate andlonger-term decision. But the distinction is notso much an absolute one as one of degree.This point is made in Figure 5 by showingthe three kinds of uncertainty surfacing againat the broader level of the more complexplanning problem. Indeed, if uncertainties oftype UR again appear important at this level,this may trigger off concerns to move to aneven broader level of concern, and to begin toexplore the shape of an even larger and moreobscure cloud. But this process of continuallyenlarging the scope of the problem will alwayshave its limits in practice; and, if useful pointersto action are to emerge, then the focus of con-cern must be kept within manageable bounds.It is not hard to see how planning pro-

cedures conceived with ambitions towardscomprehensiveness can develop their owninternal momentum. Such tendencies can befound in corporate planning procedures forthe guidance of large and diffuse commer-cial enterprises, and also in exercises in theproduction of land-use plans or economic plan-ning frameworks, through which public agen-cies endeavour to set a context for the actionsof other parties. The danger is always that suchactivities will become separated from othermanagement processes and so cease to exer-cise any real influence on the more immediatedecisions they were designed to inform. Thisrisk of disengagement between arrangementsfor planning and for management has alreadybeen suggested in the keynote diagram(Figure 1); it is a risk that can be confronteddirectly from the perspective of planning as aprocess of strategic choice.

HUMAN SETTINGS FORDECISION-MAKING

The shift from a ‘snapshot’ view of decision-making (Figure 3) to a more dynamic, multi-level picture (Figures 4 and 5) implies that the

imagery of the cloud should itself be conceivedin more realistic, multi-dimensional terms. Toextend the metaphor, clouds in reality are notflat: they have length, depth and breadth; theiredges may be blurred; they progress acrossthe sky, changing shape as time passes; theydissolve, they merge, they break up; and, in sodoing, they assume new and often unpredict-able forms.With such a picture in mind, it is possible

to look more closely at some different kindsof human context for decision, as a steptowards a closer examination of the processesof thought and communication which go on‘within the cloud’, Figure 6 begins by lookingat an organisational context of the most simpleand restricted kind – in a pre-computer officeenvironment: An individual sits on a chair (sym-bolising a defined organisational role), with suc-cessive matters for decision arriving in an ‘in’tray on a table (symbolising an agenda). Thematters are dealt with in sequence, agreedrules are applied, and decisions are then trans-ferred to the ‘out’ tray one at a time.If the rules are unambiguous in their bearing

on the issue currently being dealt with, thenthe cloud representing the thought process ofthe decision-maker is a small one and quicklyevaporates to be replaced by the next. Indeed,the symbolism of the cloud can be replaced bythe more mechanistic image of the black box –and the decision-maker at the table is at risk ofbeing superseded by an electronic counterpart.Of course, the cloud may sometimes becomelarger, when a more complex case arrives. Thedecision-maker now experiences uncertaintyand, as in the case of the public official dealingwith the application to build a new hotel, thisuncertainty may be in part due to awarenessof links to other related cases – as symbolisedperhaps by some matters marked for furtherattention in a ‘pending’ tray.Figure 6 then demonstrates another context

of sequential decision-making, by contrastingthe situation of a single decision-maker sittingat his or her small table with that of a collectivedecision-making body – a committee or man-agement board – grouped around a larger table.

13

FIGURE

6Two Contexts for Sequential Decision-Making

Foundations

Such a group will often have a pre-circulatedagenda, presenting an ordered list of issuesto discuss and where possible resolve, corres-ponding to the ‘in’ tray of the single decision-maker. Among the occupants of the rolessymbolised by the chairs around the largertable, there will usually be someone in a ‘chair-ing’ role, responsible for ensuring that the busi-ness is dealt with in an orderly and expeditiousway. In place of the ‘out’ tray, there will usuallybe a running record of decisions kept by a com-mittee secretary or clerk; while the occupantsof at least some of the other chairs around thetable will sometimes be recognised as havingdifferent representative or expert roles to play.The decision-making process is now not

purely one of cogitation within an individual’shead; it embraces processes of commu-nication, verbal and nonverbal, among themembers of the group. For the observer ofthe process, the elongated cloud above thelarge table in Figure 6 takes on additional sub-stance, in that it becomes possible to followthe dynamics of information sharing, negoti-ation and – if decisions are to be reached –compromise between conflicting views. Forinstance, if the issue of permission to build anew hotel has been brought up on the plan-ning committee’s agenda, there may be avariety of financial, aesthetic, engineering andcommercial considerations to be exposed andshared. Further, there may be various con-flicts of interest to be managed; for instance,there could be conflicts between the commit-tee’s responsibilities to the local communityand the relationships of some members withthe developer, who could perhaps be a well-known and influential local figure.However, many decision processes in prac-

tice fail to conform to either of these tidy,sequential models. If the issues are complexand their boundaries unclear, then organisa-tional responsibilities too are likely to be dif-fuse and probably confused. Communicationsmay take place not just around tables buton the telephone, in corridors, in small back-rooms. The inputs and outputs can no longerbe seen as falling into any clear sequence, and

the image of the single cloud may have tobe replaced by one of several separate cloudswhich continually come together, drift apart,coalesce or disappear. For instance, the hoteldeveloper, in making his or her own invest-ment decisions, may have a series of meetingswith planners and other public officials, as wellas finance houses, landowners and other com-mercial interests. The developer as well as thecommittee members will have uncertainties tomanage in some or all of the three categoriesof UE, UV and UR; and the extent to which thedifferent planning processes can be linkedmaybegin to raise a host of difficult administrative,political, ethical and legal issues.

MODES OF DECISION-MAKING

In developing further the view of planning as aprocess of strategic choice, it is helpful to seethe process within any ‘cloud’ as continuallyshifting between different and complement-ary ‘modes’ of decision-making activity. In thesimple situation of sequential decision-making,where the nature of the problem inputs and theexpected decision outputs is well defined, thismovement can be seen in terms of only twocomplementary modes: the one concernedwith designing possible courses of action, andthe other with comparing them in the lightof some view of what their consequencesmight be. This relatively simple view is indic-ated in Figure 7. The process may not in prac-tice be strictly linear because a comparison ofthe consequences of any pair of alternatives– for instance a straight ‘yes’ or ‘no’ to theapplication to build a hotel – may reveal thateither response could have undesirable con-sequences and so trigger off a search for someother compromise solution. So it becomesnecessary to allow for the possibility of a feed-back loop returning from the comparing to thedesigning mode. So, in Figure 7 the single‘cloud’ is shown as tending to change shapeinto two smaller clouds which may still notbe clearly separable in practice, insofar as theinterplay between designing and comparingmay become rapid and difficult to trace.

15

FIGURE

7A Process of Simple Choice

FOUNDATIONS

Foundations

This picture hasmuch in commonwith other,more orthodox models of decision-making pro-cesses, which tend to present stages or activ-ities in logical sequence, having a beginningand an end, while allowing for elements offeedback or recycling in between. However,the more diffuse, continuous kind of processwhich is characteristic of the making of com-plex decisions in practice involves coping withmultiple problem inputs and multiple decisionoutputs, with no clear sequential relationshipsbetween the two. To represent this kind ofprocess, it is necessary to move to a rathermore elaborate picture of the process withinthe cloud, introducing two additional modes asshown in Figure 8.The two further modes of decision-making

activity which make their appearance inFigure 8 are both of a more subtle and polit-ical kind. One of these is concerned with theshaping of problems; a mode within whichjudgements about the possible connectionsbetween one field of choice and another canhave a crucial role to play. The other, referredto as the choosingmode, is concernedwith theformation of proposed commitments to actionprogressively through time. Here it has to bekept in mind that the more complex the shapeof the problem, thewider the choices that haveto be faced. There will be choices, not onlyabout what courses of action are preferred,but also about what degree of commitmentis appropriate at this stage; which decisionsshould be deferred until later; and what explor-ations could be set in train in response to dif-ferent types of uncertainty. So instead of twopartially overlapping foci within the cloud, therenow appear four, with a variety of possible dir-ections of movement between one mode andanother. Themost orthodox progression mightappear to be from shaping problems, throughdesigning possibilities, to comparing their con-sequences and then on to a final choosing ofactions. However, such a progression is likelyto be neither straightforward nor realistic, inso-far as the process is to be seen as a continuousand incremental one, with no clear beginningand no single end. For the choice of actions

to deal with some parts of the problem situ-ation will leave other choices open for thefuture, creating opportunities for future reshap-ing of problems as unexpected events occurand new connections begin to appear.

CHALLENGES TO MANAGEMENTAND PLANNING NORMS

Already, the ideas presented here can beseen to pose some direct challenges to long-established management and planning norms:norms which have indeed been under sus-tained challenge from other sources, yetremain extremely persistent in the design offormal management and planning procedures– often for reasons of organisational stabilityand accountability which cannot be lightly cri-ticised. Among the more deeply establishednorms in any management system are thoseof linearity, objectivity, certainty and compre-

hensiveness. These can be summarised asfollows:

• Aim for linearity – ‘Tackle one thing at a time’;• Aim for objectivity – ‘Avoid personal or sec-tional bias’;

• Aim for certainty – ‘Establish the full facts ofthe situation’;

• Aim for comprehensiveness – ‘Don’t dothings by halves’.

Such norms may be adequate for the function-ary sitting at a desk, working to highly con-strained terms of reference. However, evenhere the system of rules can rarely be exhaust-ive in representing the situations that couldarise; so feelings of uncertainty about how toact will sometimes surface and with them willarise difficulties in conforming to the norms oflinearity and objectivity in their pristine forms.When a shift is made from decision-makingto plan-making, the same four norms tend toshow a remarkable persistence, even thoughthe language may change. Yet the experi-ence of working on difficult and complex plan-ning problems is that the norms of linearity,objectivity, certainty and comprehensivenesskeep on breaking down. So, in this book, they

17

FIGURE

8A Process of Strategic Choice

FOUNDATIONS

Foundations

will be replaced by less simple prescriptions ofthe following form:

• Don’t aim for linearity – learn to work withcyclicity;

• Don’t aim for objectivity – learn to work withsubjectivity;

• Don’t aim for certainty – learn to work withuncertainty;

• Don’t aim for comprehensiveness – learn towork with selectivity.

These alternative prescriptions may appear tobe less straightforward to interpret in practicethan the more familiar norms. But experiencedemonstrates that they offer a much moreeffective guide for people in attempting tochoose strategically in practice. What is more,once they are stated and developedmore fully,they can help people cope constructively withany sense of residual guilt they may feel infailing to apply simple management and plan-ning norms when they encounter problems ofa more complex kind.

IMPLICATIONS FOR ATECHNOLOGY OF STRATEGICCHOICE

There are many forms of management andplanning technique which have been devisedto help people deal with difficult decision prob-lems. Indeed, systematic methods of design-ing courses of action, and comparing theirlikely consequences, have reached a consid-erable level of sophistication in some profes-sional fields. For instance, systematicmethodshave been developed for assessing invest-ment proposals in the light of predictions ofnot only their economic but also their socialand environmental implications, while thereare various computer-aided methods whichcan help generate a range of alternativeswithin some of the better understood fieldsof technological design. Meanwhile, mathem-atical programming techniques can allow ana-lysts to conduct a systematic search for bettersolutions within a complex, multi-dimensionalfield, provided certain stringent assumptions

about the structure of the problem canbe met.As yet, however, there has been much less

investment in the development of techniquesto support the two modes of decision-makingwhich appear in the upper part of Figure 8 –even though these two modes take on specialsignificance in confronting decision problemsof a less clearly structured kind, where itbecomes necessary to cope with multipleinputs and outputs in a highly flexible, cyclicand, necessarily, subjective way.Just as a distinction can be drawn between

the two lower, more technical, modes inFigure 8 and the upper, more political modes,so can another kind of distinction be drawnbetween the two modes to the left of the dia-gram and the twomodes to the right. Whereasthe former two modes are primarily addressedtowards the task of opening up the field ofchoice facing the decision-makers, the lattertwo modes can be seen as addressed towardsthe complementary task of narrowing that fielddown again in order to work towards agree-ment on action.This distinction will be used as a basis for the

organisation of the next two chapters. Thesewill introduce and illustrate a series of basicconcepts and techniques which have beendeveloped, tested and modified progressivelythrough repeated application to a range ofapplied planning problems. Together, theseconcepts and techniques can be seen as con-stituting an appropriate technology for stra-tegic choice: appropriate in the sense that itis not intended as an advanced technology foruse primarily by the expert. Rather the tech-nology is designed to support the interactivework of groups of people who have differentperspectives to contribute to a problem; whoface quite daunting challenges in communicat-ing with each other; yet who may appreciatethe importance of working quickly and inform-ally under the pressures of day-to-day events.So, the concepts to be introduced in the nexttwo chapters will only occasionally be worthapplying with a high level of analytical soph-istication by specialists in a backroom setting.

19

FIGURE

9A Structure for Subsequent Chapters

10. the electronic resource

11. extensions in process management

12. invention, transformation and interpretation

13. learning from others

14. the developmental challenge

STRUCTURE

Foundations

Indeed, themore complex and unclear becomethe issues and their relationships, the moreproblematic become the more political modesof shaping problems and choosing actions, andthe more vital it becomes that any technologyof strategic choice be capable of use in a flex-ible and relatively non-technical way.

IMPLICATIONS FOR CHAPTERSTRUCTURE

The emphasis in the chapters that follow willtherefore be on quite simple and transpar-

ent concepts and techniques – most of theminvolving graphical forms of representation.These are intended to aid the processes ofcommunication between people whose per-spectives, attitudes and experiences may dif-fer, asmuch as to help individuals in structuringtheir own personal thought processes. Work-ing on these principles, Figure 9 presents a pre-view of the way in which the content of thenine chapters that follow will relate to the fourmodes of strategic choice which were distin-guished in Figure 8. This picture can be usedin conjunction with the ‘Quick access guide’at the beginning, by the reader who wishes to

refer at any moment to the principles of struc-ture on which this book has been designed.Chapter 2, which is concerned with con-

cepts and techniques for working into com-plex problems, will begin with the shapingmode, and will introduce some simple con-cepts which can help in structuring areasof choice and the interconnections betweenthem. It will then move down to the design-ing mode, to introduce some further ideasto help in organising views about the optionsavailable and the patterns of compatibility orincompatibility between them. Chapter 3 isconcernedwith the complementary process ofworking towards decisions; it will begin withsome concepts intended to help in comparingthe foreseeable consequences of alternativecourses of action, taking uncertainty expli-citly into account. It then moves on to intro-duce further concepts addressed to the explicitmanagement of uncertainty and the choice ofincremental actions through time, drawing onthe UE/UV/UR framework which has alreadybeen introduced. Together, the two chaptersprovide a foundation for the discussion, inlater chapters, of the many different ways inwhich the basic concepts and methods can bebrought into play in practice.

21

2 Working into problems

INTRODUCTION

The aim of this chapter is to introduce a setof basic concepts and methods which, takentogether, offer a means of helping people tostructure complex decision problems in termsof interrelated elements of choice. These con-cepts and techniques are addressed in particu-lar to the work of the shaping and designingmodes as set out on the left-hand side of thegeneral process diagram (Figure 8). Thismeansthat they are concerned both with the shapingof problems and with the designing or formula-tion of possible courses of action in responseto those problems. A further set of conceptsand techniques for comparing those possiblecourses of action and for choosing betweenthem will then be described in Chapter 3, tocomplete this introduction to a basic ‘technol-ogy’ for strategic choice.The basic concepts to be introduced in this

chapter will include those of the decisionarea; the option within a decision area; andthe decision scheme, consisting of a set ofmutually consistent options drawn from a setof interconnected decision areas. These andother concepts will be introduced in turn, illus-trated by example and more formally defined.Taken together, they offer a quite general andflexible basis for the formulation of complexdecision problems; and, more specifically, forthe use of an analytical method known asAnalysis of Interconnected Decision Areas –AIDA for short. The essentials of the AIDAmethod will be explained in this chapter, butsome of themore important variations on it willbe deferred for discussion in later chapters.The approach to structuring of complex prob-

lems to be introduced here is quite simple

in its essence. Yet, because it involves try-ing to express complex realities in simple andcomprehensible terms, it can demand subtleand shrewd judgements of those participatingin the process; judgements of a kind whichare often made intuitively by individuals, yetare rarely exposed to argument in the nor-mal course of debate. The nature of thesejudgementswill becomemore apparent as thischapter unfolds.For the purpose of introducing the basic

ideas as simply and clearly as possible, thischapter will begin to develop a case exampleto be known as the South Side Story. It isa story of a group of decision-makers facedwith a set of linked investment and locationaldecisions which impinge on a residential com-munity of around 7000 people, living on thefringe of a larger urban area long dominatedby heavy industry, but now in a state of eco-nomic and environmental decline.1 The storyis one which will be developed gradually inthis and later chapters, as further conceptsare introduced, and the skills and judgementsinvolved in applying them in practice are dis-cussed. For purposes of exposition, the variousconcepts and techniques will at this stage beintroduced sequentially and in an orderly way.However, it has to be kept in mind throughoutthat in practice the process of strategic choiceis normally much more flexible and adapt-ive. The process can shift rapidly from onemode to another, with a continued readinessto ‘recycle’ through earlier stages of analysis

1 The South Side Problem as presented here is closely –though by no means exactly – modelled on one of thefirst successful experiences in applying strategic choicemethods to urban development problems.

23

FIGURE

10

SOUTH SIDE

EXAMPLE

The Concept of a Decision Area

A DECISION AREA is an opportunity for choice in which two or more different courses of action

can be considered.

Working into Problems

as new insights emerge and the level of under-standing grows.

THE CONCEPT OF THE DECISIONAREA

The concept of a decision area provides themost fundamental element in the approachto problem structuring to be described in thischapter.In essence, this concept offers no more than

a means of describing and labelling any prob-lem situation where people see an opportun-ity to choose between different courses ofaction. To begin with a simple example, you,the reader, may even now be thinking abouta choice as to whether to read the remainderof this chapter or to skip ahead to the next.Or a person lying in bed may be conscious –even if only dimly – of a choice about whatto do when a bedside alarm sounds: whetherto get up, to ignore it, or to silence it bywhatever means may be available. On a lesspersonal note, a bank manager may encountera decision area when judging what level ofinterest to charge a particular client for a loan;while the local authority planners mentioned inthe last chapter found themselves faced witha decision area as to whether or not to approvethe proposal for a new hotel. Meanwhile, thedeveloper in question might face a range ofdecision problems, to do with timing, choicebetween alternative locations, scale, design,financial backing and other important commer-cial matters. These could either be expressedas a single, rather complex decision area or –as would be more usual when using strategicchoice methods – as a set of different decisionareas, the mutual relationships of which wouldhave to be explored.Implicit in each of these situations is

an opportunity for decision-makers, whetheralone or in association with others, to act inat least two alternative ways.2 Also implicit ineach situation is some sense of pressure or

2 Purists sometimes point out that the word ‘alternative’applies logically to an ‘either/or’ situation, so it is not

concern to arrive at a commitment to somepreferred course of action amongst thosebelieved to be available even though it is onlyto be expected that some decision areas willcarry a greater sense of urgency to act than oth-ers. It is this sense of pressure to act that cre-ates a decision problem for those concerned– it being useful to distinguish the idea of adecision problem from that of other types ofproblem or puzzle which may be picked up andworked on casually as a diversion by anyonelooking for interesting ways of passing theirtime.To illustrate the concept of the decision

area through the example of the South Sidestory, it is now necessary to set the scene alittle more fully, by describing how the pres-sures to act have arisen in this case. At thisstage, the context will be taken as one inwhich South Side comes within the adminis-trative boundaries of a large urban municip-ality – the city of Dockport, serving arounda quarter of a million people in all. SouthSide itself is an old-established neighbourhoodwith a strong sense of community among theresidents – though they have been steadilydeclining in number as older housing has beencleared and local employment opportunitieshave dwindled.This population base is likely to be reduced

further over the next few years by the impend-ing closure of the local steelworks, which hasbeen a source of many jobs – but also ofsevere local air pollution that has lessenedthe attraction of South Side as a residentialarea. Althoughmany of the older houseswhichremain in South Side are scheduled to bedemolished in the next 2 or 3 years, otherscould have a prolonged life if designated forimprovement, with financial aid from govern-mental grant aid programmes.Suddenly, however, a new note of urgency

has arisen for the municipality in considering

strictly correct to talk of a set of more than two differentcourses of action as ‘alternatives’. However, this neednot be treated as a serious source of difficulty here, as inthe chapters that follow the comparison of alternativeswill usually be treated as essentially a pair-wise process.

25

FIGURE

11

SOUTH SIDE

EXAMPLE

The Concept of a Decision Link

A DECISION LINK is a relationship between two decision areas expressing a belief that it could make

a difference to consider them jointly instead of separately.


what to do about South Side – and, in particular,in addressing the problem of how far to investin its continued viability as a residential com-munity. For a proposal has just been publishedby a transportation agency to route a newarterial highway carrying industrial and othertraffic directly through the neighbourhood. Asmight be expected, this heightens the senseof anxiety about the future among the localresidents and traders. In response, the muni-cipality calls for an early report on the problemfrom a specially formed internal working partyof planners, engineers, accountants, and legaland valuation experts. From their initial discus-sions, it will be supposed that a list of sevenpotentially important decision areas emerge,as indicated in Figure 10.In this list, it will be noticed that each decision

area has not only been described with somecare, but also given a brief label for purposes offuture reference. The question marks – thoughthey can be treated as optional in practice – arehere added as a reminder that each decisionarea is supposed to represent an opportunity

for choice rather than any particular outcome ofthe decision process. This is a point that is alsostressed in the more formal definition accom-panying Figure 10.As the example shows, any list of decision

areas can be quite diverse in the types ofopportunity for choice which it contains. Thefirst decision area in the list concerns choiceof alignment for a road; the second concernschoice of location for a proposed local facility;two others concern choice of land use for par-ticular sites; while the last in the list concernsa choice of timing.Therecanalsobevariationsbetweendecision

areas in the level of generality at which theyare expressed. Whereas the fifth decision areaconcerns choice of policy stance in relation toinvestment in South Side as a whole, thethird concerns a much more specific choice ofaction in relation to one particular street. It isone of the inherent strengths of the decisionarea concept that it allows different types andlevels of choice to be considered togetherwithin a common analytical framework.

LINKS BETWEEN DECISION AREAS

As soon as a set of opportunities for choice hasbeen formulated as a list of decision areas –even if only in a tentative way – it will usuallybegin to become apparent that some of themat least can be viewed as interconnected, inthe sense that there is a case for consideringthem jointly rather than attempting to come todecisions taking each of them one at a time.For example, an appreciation of the local

geography of South Side might make it appar-ent that it could be unwise to consider thechoice of use for the central site without anyreference to the choice of location for thefuture local shopping centre. This implies abelief that the choicesmade if the two decisionareas were considered together could differ

from those that might emerge if each werelooked at in isolation, on separate ‘agendas’of decision-making. This might be the caseif the choice of particular uses for the cent-ral site made some conceivable locations forthe shopping centre physically impossible, orvice versa; it could also be the case if cer-tain choices in either decision area seemedlikely to make some choices in the other lessattractive in terms of costs, implications forlocal residents or other consequences withwhich the working party could be concerned.More obviously, there could be an interconnec-tion between the two decision areas if one ofthe possible locations for the shopping centrewere the central site itself.However, not every pair of decision areas

in a list is likely to be directly linked in any ofthese ways. For instance, there may be no dir-ect reason for supposing that the choice as towhether or not to improve West Street cannotbe arrived at independently from the choice oflocation for the shopping centre, or the choiceof use for the central site.Using labelled circles to represent the

decision areas and connecting lines to repre-sent the presence or absence of direct linksbetween them, a picture of mutual relation-ships can be built up graphically as shownin Figure 11. Here the connecting line, or

27

FIGURE

12

SOUTH SIDE

EXAMPLE

The Concept of a Decision Graph

A DECISION GRAPH is a diagrammatic representation of a set of decision areas and the relationships

between them expressed as decision links.


decision link, represents nomore than awork-ing assumption that, at least at the presentstage of understanding of the problem, itmakes sense to look into the mutual rela-tionships between the two decision areasCENT’L SITE? and SHOP LOCN? Meanwhile,the absence of a link between WEST ST? andeither of the other decision areas representsa working assumption that this choice can bedealt with independently from the other two.The idea of a decision link forms a second

‘core’ concept in the strategic choice vocabu-lary, and is more formally defined in a state-ment appearing below Figure 11. The termdecision link is usually abbreviated simply tolink for working purposes, because it is usu-ally clear from the context that the word isbeing used in this special sense. It is import-ant to note at this stage that the concept of adecision link implies no particular view aboutthe sequence in which the linked decisionsshould be taken, or about possible causal rela-tionships. People who may be versed in otherapproaches to the mapping of relationshipsamong decisions or systems are sometimestempted to add arrowheads to decision links,to suggest directions of influence or prece-dence between one choice and another. How-ever, it has not been found helpful to introducesuch conventions into the approach being dis-cussed here, which is purely concerned withthe logic of mutual connectedness betweenone decision area and another.Another caution concerns the tendency to

interpret the concepts of decision area anddecision link too narrowly in terms of famil-iar forms of relationships. In dealing with landuse or locational problems in particular, thereis a tendency for planners to focus on spatiallydefined decision areas and to look for links interms of geographical adjacency or similar rela-tionships. But decision areas and relationshipsbetween them can be used to reflect all kindsof non-spatial considerations aswell, as the fur-ther unfolding of the South Side storywill makeclear. For this reason, no physical map of SouthSide is introduced at this stage: but any readerwho may find it helpful to form some view of

the geographical layout of South Side may liketo glance briefly at the sketch map at the endof this chapter (Figure 18).

THE DECISION GRAPH AS AREPRESENTATION OF PROBLEMSTRUCTURE

In any situation where a complex problem canbe expressed in terms of a set of severaldecision areas, some but by no means all ofwhich may be directly connected by decisionlinks, it is possible to use the graphical con-nections introduced in Figure 11 to build up awider view of the structure of that problem inthe form of what is called a decision graph.3

A decision graph is, in effect, no more thana two-dimensional ‘map’ showing a set ofdecision areas and a set of links which con-nect some pairs but not others. Figure 12 givesan example of a decision graph for the set ofseven decision areas so far identified withinthe South Side decision problem. This indic-ates that there are eight decision links in all, outof the total of 21 which would be theoreticallypossible if each of the seven had been directlylinked to each of the other six.This particular decision graph reflects an

agreed view, among people who can be sup-posed to have specific knowledge of the real-ities of the South Side situation, that somedecision areas are more directly interconnec-ted than others. For instance, it shows sev-eral links connecting the four decision areasconcerned with ROAD LINE?, SHOP LOCN?,CENT’L SITE? and DIST LIFE? – even thoughROAD LINE? and DIST LIFE? are only linkedindirectly through the other two. The WESTST? and GRIFF SCHL? decision areas, in con-trast, are comparative outliers, each beingdirectly connected to only one of the otherdecision areas. Indeed, had it not been for thesudden introduction of a choice to be madeabout the road line, as a result of a highway

3 In some earlier writings this was referred to as astrategy graph.

29

FIGURE

13

SOUTH SIDE

EXAMPLE

The Concept of a Problem Focus

A PROBLEM FOCUS is any subset of the decision areas in a decision graph which is selected for

closer examination.


planning exercise over a wider area, the WESTST? decision area would have been com-pletely disconnected from the SHOP LOC’N?and CENT’L SITE? decision areas, as wasearlier suggested (Figure 11). This is a relativelysimple example of a common occurrence notonly in urban planning but in other fields, wherethe introduction of new decision areas canintroduce additional complexity into a hithertomuch simpler pattern of decision links.Where a decision graph includes a larger

number of decision areas than the example ofFigure 12, it may be quite important to exploreways of rearranging it to bring out its under-lying structure more clearly. It is important tonote that there is no set rule to guide the pos-itioning of each decision area on the graph: themap is a topological one, themeaning of whichwould not be changed if the relative positionsof the decision areas were altered. The essen-tial information conveyed by Figure 12 wouldbe exactly the same if the WEST ST? decisionarea was shifted from the left to the right ofthe picture. However, itmight then showsomeawkward crossovers between decision links,unless other decision areas were to be repos-itioned at the same time. The value of anydecision graph lies essentially in the picture itpresents about the structure of relationshipsbetween elements of a complex problem; apicture which can be modified through timeand challenged wherever there is disagree-ment between participants in the process. Inthis way the participants can proceed, throughas many iterations as need be, towards deeperexamination of possibilities for action and theirconsequences, either within the graph as awhole or within selected decision areas orclusters within its structure.

THE CHOICE OF A PROBLEMFOCUS

The possibility of focusing on a selected clusterof decision areas within a larger decision graphoffers an important field of judgement in a pro-cess of strategic choice. Indeed, it can mark acritical point of transition from the work of the

shaping mode to that of the designing mode –especially where the decision graph is so largeand complex in its structure that it is difficult tothink of designing possible ways forward whilekeeping the full set of interconnected decisionareas in view.Any focus for the examination of possible

ways forward which has been deliberatelyselected to include some but not all thedecision areas in a decision graph can bereferred to as a problem focus. It is import-ant to note here that the scope of this prob-lem focus can be changed, repeatedly if sodesired, as work on the problem proceeds.There are many different considerations thatcan be taken into account in choosing a prob-lem focus, and these will be discussed morefully in Chapter 5. At this stage it is the generalconcept of the problem focus that is important,as ameans ofmanaging the transition from theshaping to the designing mode.In Figure 13, one possible problem focus

has been selected for the South Side prob-lem, by exercising a degree of selectivitywithinthe decision graph of Figure 12. The compara-tively isolated WEST ST?, GAS SITE? andGRIFF SCHL? decision areas have here beenexcluded, and a boundary indicating the limitsof the resulting problem focus has been drawnaround the other four. In the case of a decisiongraph of comparatively simple structure suchas this, the step of choosing to concentrateon these four may seem a somewhat obvi-ous one to take, working on the basis of thestructural information alone (Figure 12). But itis not hard to see that the judgement couldbecomemore difficult if thereweremanymoredecision areas and connections to consider,and if it was thought important to take intoaccount other reasons for focusing – such asthe relative urgency and importance of differ-ent decision areas.Even in this simple example there are several

other choices of focus that could have beenmade. It could have been decided to focusonly on the triangular cluster of ROAD LINE?,CENT’L SITE? and SHOP LOC’N? decisionareas, on the grounds that each has at least

31

FIGURE

14

SOUTH SIDE

EXAMPLE

The Concept of a Decision Option

A (DECISION) OPTION is any one of the mutually exclusive courses of action that can be

considered within a decision area.


three decision links joining it to other decisionareas, whereas DIST LIFE? has only two. Itcould also have been decided deliberately tokeep all seven decision areas within the prob-lem focus, or even to restrict it to only onedecision area in the first instance – SHOPLOC’N? being one possible candidate becauseof its pivotal position on the graph. The nar-rower the focus, the less work there will beto do in the designing mode, especially if thealternatives within the decision area or areasconcerned can be considered clear cut. So themore rapid can be the progress forward intothe comparing mode. However, this is not tosuggest that the case for choosing a broaderfocus will not re-emerge later, once the uncer-tainties involved in working within the narrowfocus have taken clearer shape.

OPTIONS WITHIN DECISION AREAS

Despite all the information about the structureof a decision problem that may be containedwithin a decision graph, or even within a par-ticular problem focus within a decision graph,this form of problem representation does noth-ing in itself to indicate what range of possibleactions is likely to be open to the decision-makers. To make progress in this direction, itis necessary to move into the more technicaldomain of the designing mode, which takes itsplace in the bottom left-hand corner of the gen-eral process diagram (Figure 8). This is wherethe analytical method of Analysis of Intercon-

nected Decision Areas (AIDA) begins to havean important part to play.4

The term decision option – usually referredto in practice simply as an option – will beintroduced at this point to describe any onecourse of action within a decision area, out ofwhatever range of possibilities may be seen as

4 The AIDAmethodwas first developed in the course of aseminal IOR/Tavistock Institute project on communica-tions in the building industry, conducted in parallel withthe Coventry local government study. See in particularLuckman (1967).

available. In the South Side case it will be sup-posed, for the sake of example, that the mem-bers of the local working party are able to agreethat the range of choice in each of the sevendecision areas can be represented by a set oftwo or more possible options, as indicated inFigure 14.In practice, of course, it may sometimes be

necessary to list more options than indicatedin this example if a fully representative pictureof the range of possibilities within a decisionarea is to be presented. Indeed, there maybe a good deal of debate about these optionsamong participants who may have differentappreciations of the problems before them.During such a debate, different perceptionscould well emerge not only about the numberof options in each decision area, but also aboutthe terms in which they should be expressed.For instance, it could be asked why the set ofoptions for public investment in the continuedlife of South Side as a residential district shouldbe limited to the range of 10-year, 20-year and40-year horizons. Why not 5 years, or 15, or100? And why express the range of possib-ilities in terms of time horizons at all? Could notthe alternative policies perhaps be expressedin some broader, more flexible but still mean-ingful way, such as a choice of short-, medium-and long-term strategies?It is also important to check that the options

within a decision area are mutually exclusive.For instance, if industrial use of the centralsite did not necessarily rule out the possibilityof housing or open space on part of the site,then mixed use options might have to be intro-duced; or perhaps the decision area itself couldbe reformulated in some way, to enable theoptions available to be expressed in a differ-ent form.Questions aboutwhether the optionswithin a decision area can be considered bothrepresentative and mutually exclusive can bewell worth discussing if they help to focus crit-ical attention on what is meant by the decisionarea in question and to suggest possibilities forreformulation of the problem in more realisticways.

33

FIGURE

15

SOUTH SIDE

EXAMPLE

The Concept of an Option Bar

An OPTION BAR is a representation of an assumption that two options from different decision

areas are incompatible with each other.


Once the set of options within a decisionarea is agreed to be adequate as a base for fur-ther analysis, it can be a useful practical step togive them short labels, as in Figure 14. So longas these labels do not suppress too much ofthe information contained in their full descrip-tions, they can save a great deal of time andspace when it comes to examining combina-tions of options from different decision areas.

COMPATIBILITY OF OPTIONS ININTERCONNECTED DECISIONAREAS

Once options have been identified, the ques-tion arises of what possibilities for choice areto be found, not merely within each decisionarea taken separately, but within linked pairs orsets of decision areaswithin the selected prob-lem focus. It therefore becomes necessary tointroduce assumptions about how far optionsfrom different decision areas can be combined.For instance, the ROAD LINE? decision areain South Side contains two options while theSHOP LOCN? decision area contains three;if each option in the first decision area couldbe freely combined with each option in thesecond, this would give a total of 2× 3 = 6possible combinations from which to choose.In practice, however, the range of possibilit-ies may bemore restricted, because of variouskinds of constraint which may be encounteredin trying to combine particular options in onedecision area with particular options in otherdecision areas. For example, a knowledge oflocal geography in South Side might make itsensible to assume that the choice of thesouthern road line would rule out the choice ofboth theMain Street and the King Square shop-ping locations, because either would meanthat the majority of residents would be cutoff from their neighbourhood shopping centre.Such a combination could be seen as violat-ing what might be recognised as an importantdesign principle – if not altogether destroyingthe centre’s economic viability. A similar checkon other combinations of options from thesetwo decision areas, followed by a check of

options in the CENT’L SITE? decision areaagainst those in each of the previous decisionareas, might generate further assumptions onincompatibilities as shown in Figure 15.Such a table is sometimes known as a com-

patibility matrix. When other decision areas areadded, it can be extended further in a stepwisefashion to form a triangular array, until eachpair of the decision areas within the presentproblem focus is covered. Each relationship ofincompatibility between a pair of options fromdifferent decision areas, as indicated by a crossin any particular cell of the table, is known as anoption bar. In the South Side example, thereare altogether three option bars connectingthe ROAD LINE? and SHOP LOCN? decisionareas, which rule out three of the six conceiv-able combinations of options; then there area further three option bars ruling out certainfurther combinations once the CENT’L SITE?decision area is added.As in the identification of options within a

decision area, it is quite normal for differentparticipants to hold different views as to whichcombinations of options are feasible andwhichare not. Again, such differences can be usedas a point of departure in working towardsa clearer shared view of the structure of thespecific decision problem which the decision-makers face.

BUILDING UP AN OPTION GRAPH

Where there are many decision areas andoptions to consider, and so the number ofpossible combinations is large, it can becomea correspondingly laborious matter to checkeach option in each decision area for compat-ibility with each option in every other decisionarea in a systematic way. This itself is onegood argument for choosing a limited problemfocus of no more than four or five decisionareas within a complex decision graph. How-ever, in building up a series of two-way tablessuch as those in Figure 15, it is usually foundthat crosses representing option bars appearin only a minority of the cells. This can greatly

35

FIGURE

16

SOUTH SIDE

EXAMPLE

The Concept of an Option Graph

An OPTION GRAPH is a diagrammatic representation of the compatibilities and incompatibilities

of options within a problem focus.


simplify the analysis of which combinations arepossible and which are not.The same kind of information can be built

up by graphical methods, through an extensionof the kinds of conventions used to developthe decision graph. This involves constructingwhat is known as an option graph, in whichdecision areas are represented by circles, asin the decision graph, but the set of optionsavailable within each decision area is spe-cified within each circle as in the exampleof Figure 16. This allows the structure ofrelationships between specific options to berepresented by drawing in connecting linesbetween those pairs of options from differentdecision areas where option bars have beenidentified.It can avoid clutter in the option graph to

use abbreviated labels for the options withineach decision area and also to write the nameof the decision area itself outside, rather thaninside, the circle. The pattern of option bars canthen be built up gradually, scanning the wholegraph for possible incompatibilities rather thanworking logically through the combinationsone at a time – and concentrating on pairs ofdecision areas which are directly connectedthrough decision links. This kind of approachcan provide a more open means of building upa picture of incompatibilities than the matrixapproach – especially where there are severalparticipants with different kinds of insight tooffer.Although the picture of Figure 16 may look

quite complex and hard to interpret at a glance,it is worth noting that it includes only 10 optionbars in all, as compared to the much largernumber of combinations of pairs of optionsfrom different decision areas – in this example,35 – which remain feasible. This observa-tion helps to explain the convention of usingconnecting lines in an option graph to rep-resent incompatible combinations rather thancompatible ones. When first encountered,this convention can be found surprising andcounter-intuitive. Nevertheless, a little exper-ience soon shows that it is normally farmore economical to use links between options

to represent the few incompatible pairs ofoptions than the many compatible pairs. Notonly does this make the picture of criss-crossing lines less impenetrable to the eye;more importantly, it allows new option barsto be introduced gradually as new reasonsfor incompatibility suggest themselves, andmakes it much easier to keep track of the logicwithin the option graph.It is only to be expected that the pattern of

option bars in an option graph will bear someresemblance to the pattern of decision linkswithin the corresponding part of the decisiongraph – if only because one obvious way inwhich a pair of decision areas can be inter-connected is through some restriction on theextent to which options within them can becombined. However, the correspondence willnot necessarily be precise: for instance, a pairof decision areas may be seen as linked onthe decision graph not because there are anycombinations of options which are incompat-ible, but because there are some combinationswhich appear to bring particular advantages ordisadvantages in terms of costs or other con-sequences. So decision links do not necessar-ily imply option bars. Nor is it inconceivable thatoption bars will be identified to connect pairsof options from decision areas which were notthought to have been linked when the decisiongraphwas first drawn; it is always possible thatdeeper reflection will bring insights into theproblem structure which were not apparent atfirst sight.

GENERATING FEASIBLE DECISIONSCHEMES

At this point the core concept of a decision

scheme will be introduced to describe anycombination of options, one drawn from eachof the decision areas within a problem focus,which is feasible in the sense that it does notviolate any of the option bars included in thecurrent formulation of the decision problem.Even where the number of option bars in anoption graph is quite limited, it can be far fromeasy to see what range of possible decision

37

FIGURE

17

SOUTH SIDE

EXAMPLE

The Concept of a Decision Scheme

A DECISION SCHEME is any feasible combination of options containing one from each of the decision

areas within a problem focus.


schemes is available simply by looking at theoption graph itself. Instead, it is necessary toembark on a logical procedure for testing thefeasibility of different combinations of optionsfrom the decision areas, considered not justtwo at a time but in sets of three or more takentogether.This means first arranging the decision areas

within the current problem focus according tosome chosen sequence, and then proceedinglogically through that sequence in the man-ner illustrated in Figure 17. Here the variousfeasible combinations of options from the fourselected decision areas for South Side are builtup by proceeding through a systematic branch-ing process. The combinations are presentedin the form of a tree-like display somewhat sim-ilar in format to the ‘decision tree’ used in clas-sical decision analysis. However, the aim hereis not to provide a framework for analysing con-tingencies and probabilities, but simply to dis-play all available decision schemes for furtherexamination. To emphasise this difference, thealternative phrase option tree will be adoptedhere.In Figure 17, for example, the path through

the tree which combines the NORTH roadline with the GAS shop location is eliminatedat an early branching point, because of anoption bar linking these two options directly.At a later stage, the NORTH-KING-IND routeis terminated because there is an option barbetween the KING and IND options, whilefurther on again, the NORTH-MAIN-IND-20YRroute is ruled out because of an incompat-ibility between a 20-year district life and theMain Street shop location. So, in progressingalong each branch of the tree, it is necessaryto check the compatibility of each new optionnot only with the option in the immediatelypreceding decision area, but also with all theothers further back. In this example, indeed,the entire branch which begins NORTH-MAIN-OPEN is eliminated, not because of incom-patibilities involving any pair of these threeoptions, but because of two different typesof option bar which are encountered once thefinal DIST LIFE? decision area is added.

In this example, the systematic developmentof the tree shows that there are, in all, ninefeasible decision schemes, each of which hasbeen given an alphabetic label for referencepurposes. Sometimes there will be only a fewfeasible decision schemes, or even none at all;on other occasions, there may be so many thatthey become difficult to compare at all withoutsome further filtering process.The process of working systematically

through the branches of a tree can provideimportant opportunities for learning. It can, ofcourse, become time-consuming if the numberof decision areas and options is much largerthan in this example. In such circumstances,computer methods can sometimes be helpful,both as a check on the logic of the process andameans of testing rapidly the effect of differentassumptions on options and option bars. Thereare various alternative ways of setting out thekind of information contained in Figure 17. Inthis example, the closed branches are includedand terminated by a cross to help demonstratethe logic of the process; they could however,have been omitted, so allowing the range ofavailable choices to be presented in a morecompressed form. If desired, the set of feasibleschemes could have been presented as astraight list rather than in the form of a tree; theadvantageof thebranching format isessentiallyin the structural information it conveys. Also,thearrangementof thesetofdecisionschemescould be changed, either by working throughthe decision areas in a different sequence orby introducing comparative considerations tobring to the fore those schemes which mightbe considered more desirable. But this latterpossibilitymeansmoving on to the perspectiveof the comparing mode, and will be left forfurther consideration in Chapter 3.

ANALYSING INTERCONNECTEDDECISION AREAS: CONCLUDINGREVIEW

This chapter has introduced a set of basicconcepts which provide a foundation for thegeneralmethod of problem structuring that has

39


become known as Analysis of InterconnectedDecision Areas, or AIDA for short. The mostfundamental concepts of the AIDA method –those of the decision area, the option and theoption bar – together form the basic elementsrequired as input to build an option graph asa representation of the structure of choiceswithin a problem or part of a problem; and,from this, to find out what range of possible‘solutions’ or decision schemes is available.Among the other ideas introduced in the earliersections of this chapter were those of thedecision link, the decision graph and the prob-lem focus – all of these being intended to helppeople in debating the overall shape of theproblem before the more specific AIDA meth-ods are brought into play.The AIDA method of problem structuring

does, of course, have its limitations in repres-enting complexity of certain kinds; and theselimitations will be discussed further in laterchapters. In particular, it is not always easyto adapt the method to decision problemswhich are most naturally expressed in termsof adjustments to the levels of a set of moreor less continuous control variables. However,decision problems can only be seen in termsof this kind of ‘control model’ in comparativelystable operating contexts, where the overallshape of the problem can be seen as more orless invariable through time. In such a case, itmay be possible to usemore sophisticated andspecialised forms of analysis concerned withthe systemic relationships among the decisionvariables particular to that operational set-ting. Nevertheless, even such ‘well-structured’problems can often be embedded in widerproblemsettingswith amore volatile structure,to which a more open-ended approach to prob-lem structuring, of the kind described in thischapter, can usefully be applied. For the stra-tegic choice approach has no claims to be a‘systems approach’ in the commonly acceptedsense: rather, it is a process approach in whichthe elements are choices which are normallysupposed to be of a transient nature, and therelationships between elements are not there-fore expected to assume any systemic form.

Relating the concepts and methods intro-duced in this chapter to the five basic dimen-sions of balance in strategic choice (Figure 2),the first of them – to do with the treatmentof scope – has begun to be addressed bythe concepts introduced to guide the workof the shaping mode: the decision area, thedecision link, the decision graph and the prob-lem focus. The second dimension of balance –to do with the treatment of complexity – hasbegun to be addressed by the general con-cepts of the decision option, the option bar,the option graph and the decision scheme,introduced to guide the work of the designingmode. But there is more to be discussed aboutthe treatment of scope and of complexity instrategic choice in later chapters. Meanwhile,the treatment of conflict, uncertainty and pro-gress – while touched upon in some placesin this chapter and Chapter 1 – has scarcelybegun to be discussed in terms of basic con-cepts and techniques. It is the introduction ofsuch concepts and techniques that will be thepurpose of Chapter 3.

SOME EXERCISES

At this stage, some readers may be glad of anopportunity to test their ability to make use ofthe concepts introduced in this chapter. To thisend, the chapter concludes with a short set ofexercises,allof thembasedonsimplevariationsin the formulationof thedecisionproblemfacedin the South Side story as described so far.

1. If every pair of decision areas in Figure 12which is not shown as directly interconnec-ted were to be connected up, how manyadditional decision links would there be?

2. Suppose in Figure 12 that two additionaldecision links were added, one to connectROAD LINE? to DIST LIFE? and the otherto connect WEST ST? to SHOP LOC’N?,could the positions of the various decisionareas be altered to make the structure ofthe decision graph clearer, in particular byavoiding any crossovers between the con-necting links?

40


3. How many different ways can you seeof choosing a problem focus within thedecision graph asmodified inQuestion 2, soas to include three fully interlinked decisionareas?

4. Suppose the 40YR option for DIST LIFE?were to be eliminated as no longer avail-able for some reason, how many of thenine feasible decision schemes in Figure 17would this remove?

5. If an extra option bar were to be addedto Figure 16, to rule out the combina-tion of the 40YR option in the DIST LIFE?decision area with the OPEN option in theCENT’L SITE? decision area, how manyof the nine feasible decision schemes inFigure 17 would have to be ruled out as notfeasible?

6. Which of the decision schemes inFigure 17 – beyond those eliminated byQuestion 5 above – would be cut out ifoption bars were to be added betweenthe NORTH option in the ROAD LINE?decision area and both the 10YR and20YR options in the DIST LIFE? decisionareas?

7. How many additional decision schemeswould be added to the list of nine inFigure 17 if the option bar ruling out theSOUTH option in the ROAD LINE? decisionarea in combination with the IND option inthe CENT’L SITE? decision area were to beremoved?Answers to the above questions will be

found on the Planning under Pressure compan-ion website.Finally, we offer twomore open-ended ques-

tions which, unlike those above, have no oneanswer. First, examine Figure 18,which showsthe spatial relationships between the locationsand alignments assumed in the formulation ofdecision areas and options in this chapter.8. Do the spatial relationships shown in this

map lead you to question the reasoningbehind any of the decision links shown inFigure 12, or to suggest that any new onesshould be included?

9. Do these same spatial relationships sug-gest that there could be a case for question-ing the arguments behind any of the optionbars shown in Figure 16 or for adding anyfurther option bars?

41

FIGURE

18

SOUTH SIDE

EXAMPLE

South Side: Some Local Orientation

BACKGROUND

3 Working towards decisions

INTRODUCTION

The aim of this chapter will be to comple-ment Chapter 2 by introducing a further setof basic concepts and methods, designed toguide work within the comparing and choosingmodes. Taken together, the various core con-cepts introduced in the two chapters will forma skeleton for an appropriate technology ofstrategic choice: a skeleton which will be builtupon further in Chapters 5–8, where a rangeof variations on these concepts and methodswill be discussed. The transition from the lastchapter to this one involves a shift of focus; ashift from a concern with designing possiblecourses of action to a concernwith discriminat-ing among those possibilities in order to makeprogress towards decisions. As in Chapter 2,the intention will be to introduce a limited setof core concepts in as simple and basic a formas possible. However, they are conceptswhichaddress evaluative issues more directly thanthe concepts introduced to guide the work ofthe shaping and designing nodes; and theseevaluative issues can become quite subtle andcomplex.The first three concepts to be introduced

are those of the comparison area, the relativeassessment and the advantage comparison.All of these terms have their counterparts ineveryday use; but they are expressed here inlanguage which is designed to encourage amore explicit consideration of uncertainty thanis found in some other approaches to evalu-ation. These concepts apply generally to anysituation where there are different courses ofaction to be compared, whether there be onlytwo alternatives to consider or a much widerrange of possibilities to be scanned. Also, the

concepts are designed to apply whether ornot there are numerical or other acceptablescales of measurement in view – recognisingthat consequences can often be subtle and far-reaching in their impact, so that an attempt toreduce assessments to a single unambiguousscale may not always be an appropriate way ofdealingwith the complexities and uncertaintiesencountered in practice.The fourth concept to be introduced – that

of the working shortlist – does however meanintroducing more simplified scales of assess-ment to narrow down the range of possiblealternatives. This is especially useful whenworking with several linked decision areas,generating a wide range of feasible combin-ations of options to be compared. So it is atthis point of the chapter that the basic conceptsabout comparing come together with thoseabout designing feasible decision schemesthrough AIDA, which were introduced towardsthe end of Chapter 2. In the second half ofChapter 3, four further basic concepts willbe introduced, designed to guide the work ofthe choosing mode: these are the conceptsof the uncertainty area, the exploratory option,the action scheme and the commitment pack-age. It is in introducing these later conceptsthat the emphasis on management of uncer-tainty through time will come directly to thefore, in the spirit of the view of planning as acontinuous process of strategic choice whichwas presented in Chapter 1.

FORMULATING A SET OFCOMPARISON AREAS

The task of comparing any pair of alternativecourses of action necessarily involves forming

43

FIGURE

19

SOUTH SIDE

EXAMPLE

The Concept of a Comparison Area

A COMPARISON AREA is a description of any field of concern in which it is desired to compare the

consequences of alternative courses of action.

Working towards Decisions

some view of what the consequences, effectsor implications might be if either course ofaction were to be pursued rather than theother. It is possible to conceive of circum-stances in which it may be sufficient toconsider implications in relation to only onedominant area of concern. For example, a per-son choosing a painting to hang on a bedroomwall might be content to compare and choosesolely on the basis of personal aesthetic judge-ment – at least in circumstances where thatperson lived alone and the set of alternativepaintings available were all offered at the sameprice, with the same physical dimensions. Or adeveloper comparing competitive tenders fora building project might conceivably be con-tent to compare on the grounds of quoted costalone – but only in the somewhat artificial cir-cumstances that all tenders met the brief in anidentical way and no other comparative infor-mation was available with which to discrim-inate between contractors in terms of theirreputations for reliability or quality of work.These two examples indicate how rarely in

practice it is realistic to restrict attention to asingle dimension of comparison taken on itsown. Indeed, the more far-reaching the impli-cations of a decision problem, the wider theset of participants likely to become involved.So the less practicable it can become to reducetheir concerns to a single basis for compari-son, whether this be expressed in monetary orother terms.The approach to comparison to be developed

here can therefore be described as essentiallya multi-criteria approach – to adopt a phrasewhich has now become fashionable in rela-tion to more specific mathematical methodsof comparison. The concept of a criterion isof course familiar enough, not only to decisionscientists but also to many practising plan-ners and managers. However, the word is onewhich can convey subtly different meaningsto different people – for instance, it conveysto many people an expectation of a definedscale of measurement, even though dictionarydefinitions tend to suggest the idea of com-parison with some pre-set standard or norm.

For this reason, the concept of criterion willhere be replaced by a more general concept ofa comparison area, which will be more care-fully defined within the context of the strategicchoice approach.In essence, a comparison area can be seen

as simply a description of some area of concernto the participants in a decision process, withinwhich they may wish to consider what theconsequences of alternative courses of actionmight be. Figure 19 illustrates this conceptby presenting descriptions of four differentcomparison areas which could be seen asimportant in addressing the decisions facingthe South Side Working Party, in the planningsituation that was outlined in Chapter 2. Asin the earlier listing of decision areas, each ofthese comparison areas is specified in termsof both a brief label, for quick reference, anda fuller and more careful description of whatit embraces. By convention, the brief label isfollowed by a colon (:), so as to distinguish acomparison area from a decision area, the labelof which is followed by a question mark (?).The fuller description can serve an importantpractical purpose, as a means of checking thatthe same comparison area is being interpretedin a similar way by different participants; andit can of course always be modified later asunderstanding grows.It is often helpful to include, within the

fuller description of a comparison area, notonly an indication of the types of effects orconsequences that it covers, but also someindication of their incidence in terms of com-munity sectors or interest groups, or perhapsover different time horizons. This can be espe-cially significant where the participants seethemselves as accountable to more than one‘constituency’ of affected interests. Indeed,different participants in a decision processwill sometimes be recognised as represent-ing different sectors or groups. Questions ofincidence and perceived equity can sometimesbecome quite crucial to the politics of compar-ing and choosing; and they can indeed emergeas major sources of uncertainty in the UV cat-egory, the management of which can become

45

FIGURE

20

SOUTH SIDE

EXAMPLE

The Concept of a Relative Assessment

A RELATIVE ASSESSMENT is a name for any statement about the consequences within a comparison

area of pursuing one course of action instead of another.


critical to the guidance of the overall decisionprocess. In order to keep the number of com-parison areas manageable, they can if desiredbe formulated so as to bring together sev-eral different elements under a single moregeneral heading. For instance, in Figure 19,the familiar heading of CAPITAL: is used tobring together expenditures on both construc-tion works and property acquisition. In othersituations, however, there might also be com-pensating capital receipts to consider, from thedisposal of surplus land or buildings which,according to accounting convention, might beconsidered either within the CAPITAL: or theINCOME: comparison area.Sometimes, also, different comparison areas

can be combined. For example, it might beagreed that for working purposes the CAP-ITAL: and INCOME: comparison areas inFigure 19 should be combined into a broadercomparison area simply called FINANCE:.These examples merely serve to demonstratethe general point that there may be muchscope for choice in the way a set of compari-son areas is formulated. Indeed, where thereare several participants, it can be valuable toencourage open debate over this choice, lead-ing to elaboration or simplification of the setof comparison areas as work proceeds and thelevel of shared understanding grows.

ASSESSING CONSEQUENCESWITHIN COMPARISON AREAS

Once a set of comparison areas has beenchosen, it can be put to use as a frameworkfor comparing alternative courses of action inthe light of people’s assessments of what theirdiffering consequencesmight be. The idea of arelative assessmentwill be treated as anothercore concept in the strategic choice vocabu-lary; but it will often be abbreviated to thesingle word assessment so long as the contextis clear. The idea of a relative assessment isintended to cover not only any consequencesor implications of a direct and immediately fore-seeable kind, but also any consequences or

implications which may be more gradual, indir-ect and hard to pin down in any tangible way.There are two important qualities to bear

in mind in making a prior assessment ofthe consequences of some proposed futurecourse of action, as opposed to a retro-spective assessment of the consequencesof some course of action already carriedthrough. Firstly, any assessment of future con-sequences will always be to some degreeconjectural, in that it will involve elements ofspeculation or guesswork as to what might fol-low if that course of action were to be setin train. Secondly, such an assessment willbe essentially comparative in the sense that,whether explicitly or otherwise, it involves con-trasting the consequences of pursuing thatcourse of action with those that might followfrom choosing some other course instead. Thispoint applies even if that other course were totake a passive stance and aim to preserve thestatus quo – which is a common baseline formany kinds of assessment in practice.Figure 20 presents some examples of rela-

tive assessments within each of the fourcomparison areas for South Side. Two dif-ferent examples of relative assessments arepresented here, both from within the rangeof nine possible decision schemes which wasdeveloped earlier (Figure 17). First, Scheme Bis assessed relative to Scheme A – the dif-ference between these two alternatives beingonly in the choice of option for use of the Cen-tral Site. Then, Scheme H is assessed also inrelation to the same baseline of Scheme A,recognising that the comparison in this case islikely to be rather less straightforward becauseSchemes H and A differ in the options selectedin each of the four decision areas.It will be noticed in this example that the

forms in which the relative assessments arepresented differ from one comparison area toanother. Even though the capital and incomeassessments are both expressed in monetaryunits, the capital assessments are expressedas lump sums and the income assessmentsas annual flows, in keeping with familiaraccounting conventions. Differences in jobs

47

FIGURE

21

SOUTH SIDE

EXAMPLE

The Concept of an Advantage Comparison

An ADVANTAGE COMPARISON is a name for any statement of belief about the balance of advantage

between alternatives in one or more comparison areas.


are also expressed in numerical terms, butthis time expressed in non-monetary units ofthe net number of local jobs created. How-ever, the assessments of the consequencesfor the South Side residents are expressedhere purely in terms of words – illustrating avery common situation in practice where thereis no accepted numerical scale to which torefer.The example of Figure 20 also illustrates

some of the different ways in which feelingsof uncertainty can be expressed. In one case –the assessment of capital outlay for Scheme Brelative to Scheme A – a single estimate only ispresented; but the word ‘about’ is inserted toconvey the information that there is felt to be atleast some uncertainty over the extent of thedifference. In other places, a range of figures ispresented; this conveys additional informationabout the degree of uncertainty experienced– which may well differ between one rela-tive assessment and another. By moving toa more elaborate format of presentation thanthat in Figure 20, it would be possible to go intothese feelings of uncertainty in more depth,distinguishing between different contributoryfactors, spelling out underlying assumptionsand indicating contingencies which could havea significant effect on the levels of assess-ment presented. These possibilities will be dis-cussed further in Chapter 7; for the time being,it is enough to stress that there is a wide fieldof choice in the level of elaboration or simplific-ation employed in presenting relative assess-ments in practice.

JUDGING COMPARATIVEADVANTAGE BETWEENALTERNATIVES

Because the set of comparison areas isdesigned to reflect fields of direct concernto decision-makers, any statement that onealternative differs from another within a par-ticular comparison area will usually convey asense of positive or negative value in the cur-rent decision situation; it will be seen as eithergood or bad, nice or nasty. For example, from

the comparison of Scheme B with Scheme Ain Figure 20, it will almost certainly count as anadvantage to B that it should incur about 250thousand monetary units less than A in capitaloutlay, but a disadvantage that it should yieldless income. Again, it is likely to be considereda disadvantage to Scheme B that it shouldcreate fewer jobs, but an advantage that itshould generate more confidence among theresidents of South Side.Sometimes, there may be some conflict

of opinion as to whether a relative assess-ment should be viewed in a positive or anegative light; and often, there will be somedoubt as to whether advantages in some com-parison areas should be seen as outweigh-ing disadvantages in others. Such a state ofdoubt appears to surround the comparison ofSchemes A and B for South Side, becausethe overall balance of advantage across thefour comparison areas is by no means clear.However, Scheme H appears to offer noadvantages compared to A in any of the fourcomparison areas, if the same sense of posi-tive and negative values is applied – unlessperhaps further investigation of the uncertaintyabout the CAPITAL: assessment could revealthat H has indeed an advantage over A in thisone comparison area.Various methods of economic analysis have

been developed which allow assessmentsin different comparison areas to be broughttogether by being expressed in commensurateterms. For instance, annual flows of incomecan be converted to capital equivalents byforms of discounted cash flow analysis whichreflect market rates of return. Some econo-mists have also developed methods for com-puting monetary values for other quantitativeindicators, such as numbers of jobs created.But such conversions can have the effect ofsuppressing underlying uncertainties of valuejudgement which, from a strategic choiceperspective, may be important to expose todebate.It is more in keeping with the philosophy

of strategic choice to turn to an openlyjudgemental scale of comparison, in which

49


uncertainties of value judgement can beexposed alongside any other uncertainties thathave arisen in assessing the nature or mag-nitude of the consequences in the various com-parison areas. Such an approach is illustratedin Figure 21, which introduces a non-numericalscale of advantage comparison as a basis fortranslating relative assessmentswithin diversecomparison areas into a common frame-work. The adjectives ‘negligible’, ‘marginal’,‘significant’, ‘considerable’ and ‘extreme’ areintended to represent an ascending scale ofadvantage to the decision-makers in eitherdirection. However, the way in which thesewords should be interpreted – and indeed thejudgement as to what the relative widths ofthe various bands of the scale should be – canbe left open to the discretion of the users in theparticular organisational and political context inwhich they are working.For example, Figure 21 interprets the assess-

ment that Scheme B will involve about250 000 monetary units less capital outlaythan Scheme A as representing somewherebetween a significant and a considerableadvantage to B in relation to the decision situ-ation currently faced in South Side. In anothercontext, the judgement made might be quitedifferent: for example a central governmentsetting a national budget might well regardsuch a difference as negligible, while a smallbusiness, or an individual managing a familybudget, would probably consider it extreme.In each row of Figure 21, the convention isadopted of representing the range of uncer-tainty over where the advantage lies by a pairof arrowheads, with the current ‘best guess’marked in between – frequently, but not neces-sarily, positioned at the mid-point of the range.As the example suggests, this range of uncer-tainty can vary considerably from one compar-ison area to another.It is important to recognise that the

range is intended to embrace uncertaintiesencountered both in assessing alternativeswithin each separate comparison area andalso in judging how these assessments shouldbe transferred to the common advantage

comparison scale. For instance, in the SouthSide case, there may be considerable uncer-tainty not only over how large the differencein local jobs created might be, but also overhow much weight should be attached to anysuch difference in policy terms. Turning to theimpact on South Side residents, uncertaintyarisespartly because there isnoclearnumericalyardstick for assessment of different levels ofconfidence in the future of the area, and partlybecause some decision-makers may place ahigher policy value on residents’ confidencethan others.The value of a common judgemental scale,

however crude, is that it provides a frameworkwithinwhich assessments in different compar-ison areas can be balanced and merged. Forexample, taking the four ‘best guess’ pointsin Figure 21, it can be argued that a consid-erable advantage to Scheme B in terms ofcapital should outweigh a (merely) significantadvantage to A in terms of income; and thatthe advantage of B to local residents shouldroughly balance out the advantage to A interms of jobs, leaving a slight overall advant-age to B when all four comparison areas areviewed together.However, this overall balance of advantage

can become more difficult to judge whennotice is taken of the uncertainties that sur-round the various placings on the advantagecomparison scale. In the illustration presentedin Figure 21, the combined effect of theseuncertainties is to make it by nomeans clear towhich alternative the overall balance of advant-age for the decision-makers will lie. Someapproaches to themore careful analysis of howsuch uncertainties affect the balance of advant-age will be discussed in Chapter 7. However,the most important point about the particularformat of advantage comparison illustratedhere is that it allows many different sourcesof uncertainty to be brought together in acommon perspective: a perspective which isdesigned to reflect the political realities of thesituation within which the alternatives in ques-tion have to be compared. This kind of advant-age comparison between specific alternatives

50


will therefore form an important point of refer-ence when it comes to considering methodsof working in the choosing mode.

RESTRICTING THE FOCUS FORCOMPARISONS AMONG DECISIONSCHEMES

Where there are only a few alternative coursesof action to consider, it may not be difficultto compare each with every other, using thesame kind of methodical approach to pair-wisecomparison which was illustrated in Figures 20and 21. This is usually feasible enough whencomparing a set of three or four options withina single decision area, or when the focus ofcomparison is limited to only a few feasibledecision schemes.However, this pair-wise approach to

comparison can become much more time-consuming where there are many possiblecombinations of options available. Forexample, there are 36 possible pair compari-sons that might be made among the ninedecision schemes for South Side (Figure 17),because each of the nine can be comparedwith each of the eight others – the resultingnumber of comparisons being reduced from72 to 36 when it is remembered that pair-wisecomparison is a two-way process. If thenumber of possible schemes was doubledto 18, the number of possible comparisonswould increase more than four-fold, to 153. Ingeneral, the longer a list of schemes grows,themore essential it becomes to choose somemore manageable set of schemes within thelist – in everyday terms, a shortlist – beforeattempting to compare alternatives morethoroughly in a pair-wise manner. The termworking shortlist will be added to the basicstrategic choice vocabulary to describe anyshortlist formed for such a purpose. In effect,it serves the same kind of simplifying purposewhen people are working towards decisionsas does the idea of problem focus when theyare working into complex problems.One way in which a long list of decision

schemes can be reduced is to focus only on

those which come within acceptable limits interms of one or more chosen dimensions ofevaluation which the decision-makers see asof particular importance. So itmight be decidedto place a constraint on the maximum level ofcapital cost – if capital is regarded as a scarceresource – or the minimum level of incometo be generated by a scheme. Or it mightbe agreed, in the South Side case, to excludeany schemes involving a net loss rather than again of local job opportunities compared to thestatus quo.Introducing such a constraint normally

means resorting to a simplified scale ofassessment, on which each scheme can berepresented by a single point, with all inform-ation about uncertainty set aside for this pur-pose. This then allows the set of schemesto be rearranged, or ranked, in an unambigu-ous order of preference so far as that particu-lar scale is concerned, allowing all schemesabove or below the agreed threshold to beset aside. Often the chosen scale will be anumerical one, but this is not essential: forinstance, a non-numerical scale with sevenpoints labelled ‘very high’, ‘high’, ‘fairly high’,‘medium’, ‘fairly low’, ‘low’ and ‘very low’provides quite an acceptable scale for rank-ing purposes, because it is quite clear whichassessment comes before which other in thesequence.A simplified scale for purposes of ranking

or shortlisting can be defined either within asingle comparison area, or to span more thanone comparison area where there is somecommon scale through which they can belinked. For example, a monetary scale canbe used to span both capital and recurrentcosts, provided there is some agreed conven-tion for converting running costs into capitalequivalents or vice versa. In general, how-ever, the use of any such composite scaleas a shortlisting device can mean a consider-able sacrifice of information in the interests ofsimplification. Indeed, any approach to short-listing can mean sacrificing much informationabout uncertainty which could be importantlater on. For instance, potentially significant

51

FIGURE

22

SOUTH SIDE

EXAMPLE

The Concept of a Working Shortlist

A WORKING SHORTLIST is any subset of a set of decision schemes in which it is intended to compare

alternatives more closely.


information about South Side was presentedearlier (Figure 20) through the use of quali-fying words such as ‘probably’ and throughusing lower and upper limits in place of single-point estimates. Any shortlisting process canalso mean loss of much information aboutthe multi-dimensional nature of consequencesboth within and between the chosen compari-son areas. Such losses of information can bequite justifiable as a means of focusing theprocess of comparison where there are manyschemes to consider; but only so long as thepossibility is not forgotten of reintroducing thatinformation at a later stage in the decisionprocess.The use of simplified scales in forming a

working shortlist is illustrated in Figure 22. Thisexample introduces two contrasting scalesfor the purpose of choosing a more limitedfocus for comparison within the set of nineschemes already generated for South Side.In this example, the scale of expected cap-ital cost is expressed in standard monetaryunits, defined as in Figure 20. The assess-ment for each of the nine schemes has herebeen built up by a straightforward addition ofcapital estimates for each option, each meas-ured against the same baseline which sup-poses there to be no capital investment at all.An alternative approach would, of course, havebeen to make a separate estimate for each ofthe nine schemes seen as a composite entity.Indeed, this might have been judged prefer-able if there were believed to be significantcapital savings or costs associated with par-ticular combinations of options from differentdecision areas.The second scale, relating to confidence

among local residents, is developed in a differ-ent way. The lowest score is awarded to thescheme or schemes judged to have the mostnegative effect on residents’ confidence – inthis case Schemes H and I – while the highestscore is awarded to the scheme or schemesjudged to have the most positive effect – inthis case Scheme E. Then intermediate num-bers of points are awarded to each of the otherpossible schemes. Of course, these scores

too could have been built up cumulatively byoptions, in the same way as the numericalscale of capital cost – but it has been supposedhere that the scoring of each scheme viewedas a whole offers a more realistic alternativein the particular circumstances of South Side.Figure 22 shows not only the scores of thenine schemes on each of the two scales, butalso the two sets of rankings obtained in thisway,with tied rankings in some cases. Inspect-ing this information, it appears that Scheme Bshows particular promise, having a high rank-ing in terms of both indices. In choosing aworking shortlist for closer comparison, otherapparently promising schemes might also beincluded, such as Scheme A and Scheme E –which is the most promising in terms of resi-dents’ confidence, even if the least promisingin terms of capital outlay. Another candidatecould be Scheme C, which is moderatelywell placed in both respects. But a fullercomparison of B with A, C or E might throwup disadvantages in other comparison areas,whichmight suggest that the focus of compari-son should be shifted yet again. Furthermore,some schemes which are apparently poorlyplaced according to the scales of Figure 22might well be brought back into considerationlater because of the types of advantageexcluded from the shortlisting process at thisstage.

EXPLORING AREAS OFUNCERTAINTY

All the time comparisons of alternatives arebeing made, the participants will usually alsobe subject to practical pressures, political andadministrative, to move in the direction ofdecisions: or, in terms of the general processmodel (Figure 8), to shift upwards from thecomparing into the choosing mode. At suchtimes, attention will shift from the comparisonof alternatives under uncertainty to the con-scious management of that uncertainty froma decision-making perspective. So it is at thispoint that the concept of an uncertainty area

will be introduced. This will be regarded as

53

FIGURE

23

SOUTH SIDE

EXAMPLE

The Concept of an Uncertainty Area

An UNCERTAINTY AREA is a description of any source of uncertainty which is causing difficulty in the

consideration of a decision problem.


another core concept in the strategic choicevocabulary; similar, in some respects, to theconcepts of decision area and comparison areaas already introduced. All three concepts canbe seen as expressing areas of concern to theparticipants in a planning process, and as offer-ing wide scope for discretion and judgement inthe way they are formulated. Indeed, circum-stances often arise where what was at firstexpressed as an area of concern of one typemay with advantage be reformulated in termsof any one of the other types.An uncertainty area can be formulated at any

moment in a process of strategic choicewheredoubts arise over the choice of assumptions onwhich the designing or comparing of alterna-tives should proceed. Such assumptions canbe of many kinds, but they can be groupedbroadly according to the three categories ofuncertainty that were introduced in generalterms in Chapter 1:UE forUncertainties abouttheworkingEnvironment;UV forUncertaintiesabout guidingValues; andUR forUncertaintiesaboutRelated decisions (Figure 3).In the UE direction, participants in a process

of strategic choice may experience personaldoubts, or may differ among themselves, asto the assumptions they should make aboutexternal circumstances or trends. In the UVdirection, they may experience doubts or dis-agreements as to the values that shouldinfluence them, especially when they are seek-ing to compare alternatives across differentcomparison areas which reflect the concernsof diverse interest groups. In the UR direc-tion, they may have difficulties agreeing whatassumptions to make about the choices thatare expected to be made in future in otherdecision areas outside the current scope of theproblem on which they are working: decisionareas over which they might conceivably havesome influence, even if that influence may bequite limited or indirect.In Figure 23, the concept of the uncertainty

area is illustrated through seven exampleswhich reflect various doubts and disagree-ments seen as relevant by members of theSouth Side Working Party, at a moment

when they are trying to judge the balance ofadvantage between the two decision schemeslabelled A and B. This list of uncertainty areasmight be part of a considerably longer list builtup at successive stages of the process. Here,however, those listed are all seen as hav-ing relevance to the particular task of judgingwhether the overall balance of advantage liesto A or B. This is a judgement that can beapproached by focusing on the informationabout uncertainty that was presented in theadvantage comparison analysis of Figure 21 –though that analysis on its own gives noinformation about what the main sources ofuncertainty are.As in the listing of decision areas, and indeed

also of comparison areas, the convention ishere adopted of giving each uncertainty areaboth a fuller description and a briefer labelfor quick reference. The convention of pla-cing the question mark before rather than afterthe label is intended merely as a means ofdistinguishing uncertainty areas from decisionareas. Indeed, it allows either kind of area to bequickly transformed into the other whenever itseems sensible to do so.The list in Figure 23 begins by identifying five

uncertainty areas which are seen as betweenthem creating difficulty in judging the balanceof advantage between Schemes A and B inthe JOBS: and RESIDENTS: comparison areas.These two comparison areas are examinedfirst because they appear (Figure 21) to holdthe widest range of doubt when all the fourassessments are translated into terms of theadvantage comparison scale.Looking first at the JOBS: assessment,

the range of 100–200 local jobs fewerfor Scheme B relative to Scheme A mayreflect in the main a feeling of uncertaintyabout the attractiveness of the Central Siteto employment-intensive industries – whichcan be expressed as an uncertainty areaof type UE. However, once the attemptis made to translate this assessment intoterms of comparative advantage, a substan-tial degree of doubt may also arise as towhat weight the participants should give

55

FIGURE

24

SOUTH SIDE

EXAMPLE

The Concept of an Exploratory Option

An EXPLORATORY OPTION is any course of action conceived as a means of reducing current feelings

of uncertainty within some uncertainty area.


to the creation of new jobs in this local-ity, relative to other consequences of thechoice before them; so this is a doubt thatcan be expressed as an uncertainty area oftype UV.Turning to the RESIDENTS: comparison,

Figure 23 again shows a substantial area ofuncertainty of type UV, concerned with thepolicy value that the municipality should attachto meeting the concerns of local residents.However, the main uncertainties in assess-ing the actual impact on local residents ofchoosing Scheme B rather than A can in thiscase be seen as of type UR rather than UE,because they are to do with two related areasof decision outside the current problem focus.One of these is concerned with the futureuse of a nearby site, which is expected tobe vacated soon by a major steel-making cor-poration. Some residents of South Side havebeen campaigning for redevelopment of thissite for light industry rather than other pur-poses; if this case is conceded, it is thoughttheir concern about the use of the CentralSite will be less acute than if the site is putto any alternative use. The other uncertaintyarea relates to the recommendations of themunicipality’s own housing planning team overdensity of any future housing development onthe Central Site itself. If the density is highenough and the cost low enough to makehomes on this site available to local people,then it is judged that residents’ confidencein the future of their community would beincreased.Finally, the list of Figure 23 includes two

further uncertainty areas which impinge onthe INCOME: and CAPITAL: assessments – inone case impinging on them both. However, aglance back to Figure 21 suggests that theseuncertainty areas will not be so critical to theoverall comparison of Schemes A and B asthose affecting the JOBS: and RESIDENTS:assessments. So no attempt need be made atthis stage to look for further uncertainty areaswhich affect these two financial dimensions ofthe overall evaluation frame.

IDENTIFYING EXPLORATORYOPTIONS

Once a list of uncertainty areas has beendeveloped, the question arises of what, if any-thing, should be done about them. For eachuncertainty area that seems to be relevant toa particular comparison of alternatives, there isalways the possibility of accepting the currentlevel of uncertainty over what assumptions toadopt, and looking for ways in which it can beaccommodated. However, there may also bea possibility of initiating some kind of explora-tory action which offers a hope that any cur-rent feelings of uncertainty can be significantlyreduced before decisions have to be made.Any course of action designed to alter the

current state of doubt within an uncertaintyarea can be called an exploratory option. Theidea of an exploratory option is of central signifi-cance to the work of the choosing mode; soit will be considered as another core conceptin the strategic choice vocabulary. In effect,the identification of any exploratory optionsthat could be adopted in response to a parti-cular uncertainty area can be seen as extendingthe problem formulation by introducing an add-itional decision area; a decision area concernedwith whether or not any investment should bemade in the reduction of current levels of doubtwithin the uncertainty area concerned.This point is illustrated in Figure 24, which

gives examples of exploratory options relat-ing to each of the first three uncertainty areasfrom the list in Figure 23 – supposing these tobe judged the most important at the presentmoment in the decision process. In each case,one possible exploratory option is identifiedand is compared with the ‘null option’ of tak-ing no exploratory action in relation to theuncertainty area in question. The nature ofany exploratory action to be considered willtend to vary with the type of uncertainty areaaddressed. As originally suggested (Figure 3),uncertainty areas of type UE will suggestpossibilities for exploration in the form ofresearch, surveys, analytical work or forecast-ing exercises. Uncertainty areas of type UV

57

FIGURE

25

SOUTH SIDE

EXAMPLE

The Concept of an Action Scheme

An ACTION SCHEME is a course of action involving commitments within only some of the more urgent

decision areas within a problem focus.


will suggest possibilities for exploration in theform of policy soundings or exercises designedto clarify objectives or goals; while uncertaintyareas of type UR will suggest possibilities forexploration in the form of liaison, negotiation,joint planning or co-ordinating initiatives to dealwith relationships between the decisions cur-rently in view and others relating to differentagendas and decision-making powers.Sometimes, it is not initially clear which of

the categories UE, UV or UR describes a par-ticular uncertainty area best. The step of con-sidering possible exploratory optionsmay itselfhelp to make this classification clearer. How-ever, there may be cases where two or morequite different kinds of exploratory optionsseem to be available. Thismay sometimes leadto the possibility of splitting the uncertaintyarea into two or more separate elements,which will be discussed further in Chapter 8.In assessing the consequences that might

flow from pursuing any exploratory option,there are in general three dimensions ofevaluation which will be of importance tothe decision-makers. Firstly, they will be con-cerned with assessing any changes in the levelof confidence with which decisions can bemade, arising from a reduction in the stateof uncertainty surrounding key assumptions.Secondly, they will be concerned with assess-ing any resources which may be used upin pursuing that exploratory option; such anassessment might cover not only such tan-gible resources as money and scarce skills,but also less tangible resources such as per-sonal energy and goodwill, which may also bein short supply. Lastly, they will be concernedwith questions of the delay involved in pursu-ing any exploratory option; this may be a crit-ical consideration where there are pressuresfor early action to be taken into account.These three dimensions of evaluation can

be viewed as different kinds of comparisonarea within which relative assessments of theimpacts of different exploratory options can bemade – the aim being to judge whether or notto invest any exploratory effort towards the

reduction of doubts within any of the uncer-tainty areas currently in view. For example, theinformation presented in Figure 24 suggeststhat the proposed informal soundings amongpolicy-makers, directed towards the UV-typeuncertainty area ?VALJOB, could yield betterreturns in terms of confidence, for a lessercost, than the commissioning of a market sur-vey directed towards the UE-type uncertaintyarea ?SITEJOBS – and furthermore that theycould do so with less serious consequences interms of delay to the decision-making process.

CHOICES IN THE TIMING OFDECISIONS

Because exploratory actions invariably take atleast some time to carry through – rangingperhaps from a few minutes in the case of aquick telephone call, to several years in thecase of a major research study – the con-sideration of how to manage uncertainty alsomeans confronting difficult choices about thetiming of commitments within a continuousplanning process. For there is little point intaking actions to improve the confidence withwhich decisions can be made unless thosedecisions are deferred until the outcomes ofthose explorations are known.However, if the problem involves consider-

ation of choices in several interrelated decisionareas, it does not necessarily follow thatcommitments to decisive action must bedeferred in all those decision areas until anyagreed exploratory actions have been followedthrough. There may be some decision areasin which the participants’ preferred course ofaction will be little affected, if at all, by theoutcomes of these explorations; and, further-more, there may be some decision areas inwhich the external pressures for early commit-ment are much more intense than in others.So, in practice, a choice arises in relation toeach decision area within the current problemfocus, not merely as to which option should bepursued, but also as to whether commitmentin that decision area should be made now,or should be deferred until some later time.

59


This means that the extent of commitment

within the present problem focus has itselfbecome amatter of choice for the participants.So at this point they must strike a balancebetween a more exploratory and a moredecisive approach to progress through time(Figure 2). This is a type of judgement whichis often confronted in practice by decision-makers, even though it has as yet receivedcomparatively little attention in the develop-ment of planning methods.As a means of dealing with this element

of choice in the timing of decisions, the coreconcept of an action schemewill nowbe intro-duced.1 This term will be used to describe anycourse of proposed action in which commit-ment to a specific option is indicated in one ormore of a set of decision areas, while commit-ment in others is explicitly deferred until later inthe process. An illustration of the way in whichthis concept can be applied to the South Sideproblem situation appears in Figure 25.In this illustration, five possible action

schemes for South Side are compared. Eachis based on a different combination of optionsin the ROAD LINE? and DIST LIFE? decisionareas – it being supposed that these arethe areas of choice in which pressure forcommitment is currently most intense. Thepatterns of subsequent choice associatedwith each action scheme are displayed by astraightforward rearrangement of the branch-ing sequence through which the original set ofnine decision schemes was developed. So thesame set of nine decision schemes appears asin the original array (Figure 17), but this timein a slightly modified order because the DISTLIFE? decision area has been brought furtherforward.Of the five action schemes, there are two

which, in effect, leave no flexibility of futurechoice at all, because of the structure ofoption bars. The choice of either of theseaction schemes implies commitment in eachof the remaining three decision areas as

1 I The term ‘action set’ was used in the same sense inearlier writings on strategic choice.

well; however, each of the other three actionschemes leaves open a future choice ofbetween two and three of the original ninedecision schemes.This illustrates the important general point

that action schemes which embody the samelevel of intended commitment in the morepressing decision areas can differ from eachother in terms of the flexibility of futurechoice which is retained. Such flexibility maybe of considerable practical value to the par-ticipants, especially where they are beset withdoubts about what course of action will bethe most advantageous in the longer term.One simple, yet practical, means of comparingaction schemes for flexibility involves the useof what is called a robustness index; a ‘robust’action being seen as one which is preferable toothers in that it leaves open a wider range ofacceptable paths for the future. The underlyingassumption is that themore paths are left openwhich meet some specified level of acceptab-ility, as viewed under current circumstances,the better equipped the decision-makers willbe to respond to unexpected circumstances,should they arise in future.This idea of robustness is not a difficult one in

its essentials, but its interpretation in practicecan sometimes call for a considerable degreeof care. This is firstly because the threshold ofacceptability may not be easy to define and,secondly, because flexibility in some decisionareasmay be valuedmore highly than flexibilityin others. These points can be illustrated withreference to Figure 25, where the five actionschemes are compared in terms of two differ-ent indices of flexibility. The first is based ona straightforward count of all schemes still leftopen, while the second is based on a countof only those schemes which pass above aparticular threshold of acceptability in terms ofresidents’ confidence (Figure 22). Both theseindices assume that flexibility in the SHOPLOCN? and CENT’L SITE? decision areas arevalued equally highly; an assumption whichcould, however, be modified by giving themdifferent weightings if that was felt to be morerealistic.

60


DEVELOPING A COMMITMENTPACKAGE

To conclude this discussion of work withinthe choosing mode, a final core concept willbe introduced, bringing together the variousdimensions of the choices people face in con-sidering how to move forward at any momentin a continuous planning process. This isthe concept of the commitment package –conceived as a balanced assemblage of pro-posed steps forward which may embrace aset of proposed immediate actions; a set ofexplorations to deal with important areas ofuncertainty; and a set of understandings aboutthe ways in which any deferred choices shouldbe addressed. Any commitment package will,therefore, contain a considered set of propos-als as to how to move progressively towardscommitment through time: a set of proposalswhich can be compared against other altern-ative sets of proposals and, if agreed, can beadopted as a basis for incremental progress ina continuous planning process.2

Figure 26 presents an example of one pos-sible commitment package that might be con-sidered as a basis for making progress in theSouth Side problem situation. It takes intoaccount not only the analysis so far presentedin this chapter, but also the political realities ofthe situation in which the South Side WorkingParty is currently operating. The basic formatthat is used here – some variations and exten-sions of which will be discussed in Chapter 8 –involves dividing the overall tableau into rowscorresponding to the various decision areas,and also into columns corresponding to the dif-ferent types of proposal that could be includedin relation to each.

2 In recent practice, the term progress package is fre-quently substituted for commitment package, as itemphasises the incremental nature of the output andleaves open the possibility that, in some circumstances,the plan of action it representsmay bemore of a recom-mendation to others than a firm commitment on thepart of those who have put it together. Some of theimplications of this will be discussed in later chapters.

In the format of Figure 26, the headingof ‘Immediate Decisions’ is used to registernot only the set of substantive actions towhich commitment is now proposed – in otherwords, the proposed action scheme – but alsoany explorations which it is proposed shouldnow be set in train to create an improved basisfor future choice. Such proposals for explor-ation reflect a deliberate choice of exploratoryoptions to deal with some but not necessar-ily all of the more important uncertainty areaswhich have been identified. So the set ofimmediate decisions is in effect conceived asan incremental step forward, not just in relationto the substantive decision problem, but alsoin relation to the associated problem of how tomanage uncertainty through time.Under the second main heading of ‘Future

Decision Space’, the format of Figure 26 allowsnot only for an indication of which optionsin which decision areas are to remain open,but also for any further proposals relating tothe arrangements by which future decisionsshould be made. In relation to these deferredchoices, proposals may be made as to thefuture time horizons at which decisions shouldbe made – proposals which may reflect notonly the time scales involved in undertakingrelevant exploratory actions but also any pro-cedural considerations arising from the particu-lar administrative and political context in whichthe participants are working. In South Side,for example, the proposed informal soundingsand liaison activities, designed to cope withuncertainty in deciding a use for the CentralSite, might be expected to take no more thana month, but it might be two months beforethe next scheduled meeting of the decision-making group atwhich a formal decision on thismatter could be taken.The practicalities of the organisational con-

text in which decisions are to be taken canalso play an important part in the second sub-division of future decision space, concernedwith what is here called contingency plan-

ning. This heading recognises that it will oftenbe important to prepare in some way forparticular contingencies of a foreseeable kind,

61

FIGURE

26

SOUTH SIDE

EXAMPLE

The Concept of a Commitment Package

A COMMITMENT PACKAGE is a combination of actions, explorations and arrangements for future

choice designed as a means of making progress in a planning process.


which could have a crucial effect on futuredecisions – including possibilities that particularassumptions which seem to offer a firm basisfor proceeding in present circumstances mightbe overturned by subsequent events.In considering what proposals to enter in the

last two columns, it may become necessaryto probe certain basic assumptions relating tothe terms in which decision areas were formu-lated for purposes of analysis. For example, inSouth Side it could be important to realise thateven a firm commitment now to the northernroad line may be more in the nature of arecommendation to an autonomous highwayauthority than a matter to be decided on theauthority of the municipality of Dockport alone;so some preparationmight be advisable for thecontingency of that recommendation beingturned down. Also, it might be important torecognise that the option of declaring a 10-yearcommitment to the continued life of SouthSide as a residential community does not implyan irrevocable commitment to the demise ofthat community 10 years hence; there mightbe the possibility of an extension in that lifeat some later time, given certain conditionswhich might be anticipated now, even if onlyin general terms.It will be noticed that the example of an

action scheme incorporated in Figure 26 dif-fers from the examples compared earlier(Figure 25) in that it proposes current com-mitment in three rather than two decisionareas. The scope for future choice within theproblem focus is in this case reduced to thetwo Schemes A and B; so the proposals forexploratory action in this commitment pack-age are focused on the comparison of thesetwo schemes, drawing on the earlier analysisof uncertainty areas and exploratory options(Figures 23 and 24).A final point illustrated by Figure 26 is that

the design of a commitment package offersa point at which aspects of a decision prob-lem thatwere deliberately omitted from a prob-lem focus or a working shortlist can be broughtback into view. In the South Side case, itwill be noticed that the West Street decision

area has now resurfaced, even though it hadbeen deliberately left out of the problem focuschosen for most of the designing and com-paring work. Despite the judgement that thisdecision area can be treated in an isolatedway, the same judgements about immedi-ate or deferred choice arise in this case; andthe same general ideas about management ofuncertainty apply.

CHOOSING INCREMENTALLYUNDER UNCERTAINTY

Summarising at this point, this chapter hasfollowed the same principles as Chapter 2in presenting a set of core concepts andmethods – this time designed to help whenworking in the comparing and choosingmodes. As before, the aim has been to intro-duce these concepts andmethods at as simplea level as possible, So the chapter began byintroducing some core concepts about com-paring alternatives, starting with the conceptof the comparison area and then introducingthose of the relative assessment, the advant-age comparison and the working shortlist,before turning to other concepts to assist workin the choosing mode. These later conceptswere those of the uncertainty area, the explor-atory option, the action scheme and the com-mitment package, the last of these offeringa general framework within which to buildcommitments incrementally through time.As may already be appreciated, these ideas

cannot always be applied in practice in such astraightforward way as in the South Side caseexample as presented here. But the difficultiesencountered in practice lie not so much in theconcepts themselves as in the realities of thesituations to which they are applied; and thesedifficulties remain to be addressed whatevermethods ofworking itmay be decided to apply.In principle, there is nothing in the con-

cepts introducedherewhichshouldmake themincompatible with more formalised methodsof technical, economic or environmental evalu-ation of the kind sometimes employed on prob-lems which are seen as of enough importance

63


to justify the investment involved. Thekeypointemerging from the strategic choice philosophyis that there are always choices open to parti-cipants in the way in which the processes ofcomparing and choosing are to be addressed.Referring to the five dimensions of balance thatwere presented in Figure 2, choices of scopearise in agreeing the range of comparison areasto be included; choices in the treatment ofcomplexity arise in judging how much elabor-ation or simplification there should be in theassessment of consequences; choices in thetreatment of uncertainty arise in judging howfar to invest in exploratory actions; and choicesin the treatment of progress arise in the designof a commitment package. Choices in the treat-ment of conflict arise too, in deciding how farpeople should work on these issues in an inter-active, as opposed to a reactive, way; and thisraises issues which will be discussed furtherin Chapter 4.It is important to emphasise that these vari-

ous balances need not remain unchangingthrough time within a continuously evolvingprocess of strategic choice. It is found in prac-tice that this means approaching the work ofthe comparing and choosing modes in a spiritof dynamic comparison, within which atten-tion is continually fluctuating between com-parison of many alternatives at quite a roughand ready level, taking little explicit account ofuncertainty, and deeper comparison of a selec-ted few alternatives within which issues ofmanaging uncertainty can bemore consciouslyaddressed. The broader level of comparisonleads to restricted shortlists of alternatives tobe examined more closely at the deeper level:while the deeper level in turn affords oppor-tunities to examine critically the simplifyingassumptions introduced in any such shortlist-ing process.In Chapters 7 and 8, the skills appropriate

to management of such a process of dynamiccomparison, and the conscious managementof uncertainty within that process, will beexamined more closely in the light of work-ing experience in a range of different contexts.In particular, the concept of the commitment

package will be developed more fully, with ref-erence to organisational contexts where theparticipants cannot always be considered asa cohesive team, and the politics of incre-mental action can therefore become subtle andcomplex. For simplicity of presentation in thischapter, the participants in a process of stra-tegic choice have been assumed to share acommon framework of accountability and acommon view of the task they face. Yet themore difficult and far-reaching become theproblems to be addressed, the wider andmorediffuse the interests on which they are likelyto impinge; and the more unrealistic it maybecome to think of those involved in the pro-cess as having the same level of internal cohe-sion as has been assumed in the case of theSouth SideWorking Party. Therefore, themoreit is to be expected that the choices to bemadeabout the treatment of conflict will involveskilled judgements as to how the politics oforganisational and inter-organisational relation-ships should be addressed.

SOME EXERCISES

This chapter – like Chapter 2 – will concludewith a few simple exercises based on vari-ations of the South Side story. These aredesigned to help the reader begin to develop aworking familiarity with the basic concepts andmethods introduced here, somost of the ques-tions are of a quite specific and sharply focusedform. However, some of them do call for ele-ments of inventiveness as well as interpreta-tion, in that the information about South Sidepresented in the text and the diagrams will notalways be enough to give the required answer.Thismerelymeans that the readerwill be facedwith some realistic problems of managinguncertainty in addressing even some of thesimpler of the exercises which appear below.

1. Working solely on the information pre-sented in Figure 19, which of the four com-parison areas might be broken down intomore narrow comparison areas coveringdifferent sub-categories of consequenceswithin the general headings as defined?

64


2. Working from information in Figure 20,what set of relative assessments couldyou make for Scheme H if you usedScheme B rather than Scheme A as abaseline? (Where you are not sure how toexpress this assessment in any compari-son area, try to find some way of express-ing the uncertainty you feel.)

3. Suppose that the relative assessments forincome and jobs were at the same levelsas shown in Figure 20, but in favour ofScheme B rather than A, could there thenbe any grounds for hesitation in express-ing an overall preference between A andB? If so, what would these grounds be?

4. Turning from Figure 20 to Figure 21, couldthe value judgements implied in convert-ing the four relative assessments to a com-mon advantage comparison scale lead youto modify your response to Question 3?

5. Does the advantage comparison ofFigure 21 suggest that either scheme islikely to be preferable to the other in termsof a strictly financial evaluation coveringboth capital and income assessments?

6. Suppose in Figure 22 that the expectedcapital cost index were+ 200 k rather than+ 400 k for the King Street shop loca-tion, and 1200 k rather than 1350 k for thesouth road line, how might this modifyyour choice of working shortlist?

7. Can you infer anything from Figure 22about the way in which residents’ con-fidence is expected to be influenced bychoice of the location for the shoppingcentre?

8. Would you judge that any of the threeexploratory options compared in Figure 24

has a clear advantage over any one of theothers as a response to the difficulty ofjudging comparative advantage betweenSchemes B and A? If so, on what assump-tions does your judgement depend?

9. In Figure 25, suppose the most urgentdecision areas are now DIST LIFE? andCENT’L SITE? Which combination ofoptions in these two decision areas leavesopen the widest range of choice ofschemes, first in total and second whenthe constraint of at least RRR on the resi-dents’ confidence index is applied?

10. Can you modify the commitment pack-age of Figure 26 to reflect the proposalthat choice should be left open in theSHOP LOC’N? as well as the CENT’LSITE? decision areas, as in the first actionscheme listed in Figure 25? Referring ifyou wish to the sketch map of Figure 18,can you make a guess as to the sorts ofexplorations that might be recommendedto prepare for this deferred decision?

11. In Figure 27 overleaf, some further inform-ation is introduced about the numbers ofpeople, dwellings and traders affected bydecisions on the South Side problem. Inwhat ways might this extra informationhelp you in answering any of the ques-tions posed above? What further typesof information might you look for if youwere a member of the South Side Work-ing Party, assuming time is short and pres-sures of decision are intense?

Answers and comments relating to thesequestions will be found on the Planning under

Pressure companion website.

65

FIGURE

27

SOUTH SIDE

EXAMPLE

South Side: Some Considerations of Scale

BACKGROUND

4 Orientations

BUILDING ON THE FOUNDATIONS

Taken together, the first three chapterspresent a set of foundations for the approachto planning under pressure which has becomeknown as the strategic choice approach.Chapter 1 presented a view of the realitiesthat people face when attempting to choosestrategically under the manifold pressures oforganisational life. It was argued that the key toa realistic approach lay in the idea of consciousmanagement of uncertainty through time; anidea which was seen as posing a fundamentalchallenge to the familiar, if sometimes impli-cit, norms of linearity, objectivity, certainty andcomprehensiveness which have traditionallybeen valued in the design of formal manage-ment and planning systems. In Chapters 2and 3, this approach was extended by intro-ducing a set of core concepts which could beseen as laying foundations for a technology ofstrategic choice, with reference to a view ofprocess in which four complementary modesof working – shaping, designing, comparingand choosing – were seen as interrelated in adynamic way.The aim of this chapter will be to pause and

take stock of what has been covered so far.It will do this within a general framework forreviewing alternative approaches to decision-making and planning, in which the considera-tions of organisation and of product are givendue weight alongside those of technology andprocess which were the main focus of con-cern in the introductory chapters. This willprovide a broad base from which more prag-matic advice on the use of the strategic choiceapproach will be developed in Chapters 5–9.

So, in essence, this chapter offers a philo-sophical bridge between the three introductorychapters and the five that follow.

REVIEW OF THE CORE CONCEPTS

Before addressing this main task, it will be use-ful briefly to take an overall view of the set ofsixteen core concepts which have been intro-duced in the last two chapters. Figure 28 bringsthese concepts together and shows how theyrelate to the four complementary modes ofshaping, designing, comparing and choosingpresented earlier (Figure 8).Because the essence of the process of stra-

tegic choice is that work should proceed in anadaptive and exploratory fashion rather thanin a rigid sequence, there is no fixed orderin which the sixteen core concepts should beapplied in practice. The particular sequence inwhich they were introduced in Chapters 2 and3 was chosen merely for ease of exposition,recognising that some linear sequence has tobe followed when introducing any set of ideasfor the first time.Starting with the shaping mode, a com-

plex and possibly ill-defined decision prob-lem can be structured in terms of a set ofdecision areas, some pairs of which are dir-ectly interconnected through decision links.This leads to a picture of a network of rela-tionships among decision areas, referred to asa decision graph, within which a choice ofproblem focus can be made.Turning to the designing mode, a repres-

entative set of decision options is identi-fied within each decision area, and option

bars are specified wherever it is assumed thatpairs of options drawn from different decision

67

FIGURE

28A Vocabulary of Strategic Choice

ORIENTATIONS

Orientations

areas are mutually incompatible. The result-ing option graph can be seen as a form ofmap, indicating which combinations of pairs ofoptions are available within the current prob-lem focus. The corresponding set of feasibledecision schemes can now be developed byarranging the decision areas in some chosensequence and then moving systematicallythrough all possible branches in the resultingtree-like pattern.Moving on to the comparing mode, com-

parison areas are formulated for comparingalternative options or schemes, and withinthese relative assessments are made,representing judgements as to what theirconsequences might be. Such assessmentscan be balanced against each other throughadvantage comparisons between pairsof selected alternatives, allowing valuejudgements and feelings of uncertaintyto be expressed in a common decision-centred framework. Wherever there are manyschemes available, the use of simplifiednumerical or other scales within some com-parison areas becomes important as a meansof choosing a limited working shortlist.In the work of the choosing mode, sources

of doubt are expressed in the form of uncer-tainty areas, leading towards the identifica-tion of exploratory options whereby someof these doubts might be addressed. Timingconsiderations are now introduced, leading tothe consideration of action schemes whichmay contain proposals for early action in some,but not necessarily all, decision areas. Suchproposals can be brought together with explor-ations in response to uncertainty, and arrange-ments relating to deferred decisions, withinthe framework of a commitment package,designed as an incremental step in a continu-ing process of decision-making through time.In distilling the essentials of a technology

of strategic choice into this basic vocabularyof sixteen core concepts, the guiding principlehas been to avoid specialised and esotericwords. Instead, words in common usage arelinked together in pairs in such a way as to

give thema heightenedmeaning. Formal defin-itions have already been recorded underneaththe appropriate illustrations (Figures 10–17 and19–26) – so the figure references attached tothe sixteen core concepts in Figure 28 canbe used as a quick cross-reference wheneverclarification is required.

TECHNOLOGY, ORGANISATION,PROCESS AND PRODUCT

In any approach to the challenge of addressingcomplex decision problems, people and theirjudgements become involved. In general, thewider the scope of the problem, the wider islikely to be the range of participants and ofpossible relationships between them. So themore it is to be expected that questions ofappropriate organisation will come to the fore.Furthermore, if the process is to be effectivefor the participants and those to whom theyare accountable, judgements will arise as towhat the most appropriate forms of its productmight be. These judgements too are likely toincrease in complexity as the scope of a prob-lem widens.The four headings of Technology, Organ-

isation, Process and Product togetherprovide a useful framework through whichto summarise at this stage the general ori-entations of the strategic choice approach.This will be done by contrasting them withthe emphases of more traditional approachesto management and planning; approacheswhich still have a persistent influence onthinking in both commercial and govern-mental organisations. The mnemonic A-TOPP

(Approach – Technology, Organisation, Pro-cess and Product) will be used here in referringto this framework for comparison, with theoverall concept on an Approach seen ascovering all the other four elements.1 It is

1 This framework evolved from one first developed byHickling; it was later applied to the comparative analysisof approaches in a study of methodologies of regionalplanning by a joint Birmingham University/TavistockInstitute research team (Hart, Hickling, Norris andSkelcher, 1980)

69

FIGURE

29Orientations and Guidelines in Strategic Choice

Orientations

not only the strategic choice approach thatcan be examined in terms of the four ele-ments of technology, organisation, processand product; so too can any other approach,whether it is one which has been consciouslydesignedwith reference to particular principlesor beliefs – such as the norms of linearity,objectivity, certainty and comprehensivenessdescribed in Chapter 1 – or whether it hasevolved pragmatically in response to particularlocal circumstances and constraints.

ORIENTATIONS OF THE STRATEGICCHOICE APPROACH

The four elements within the A-TOPP frame-work – Technology, Organisation, Process andProduct – are represented in Figure 29 as thefour vertices of a tetrahedron. The solid struc-ture of the tetrahedron itself is intended to rep-resent the idea of an overall approach viewed ina relatively complete, multi-dimensional way.The picture as a whole is intended to conveythe message that each aspect of the approachcan be selected in turn as the focus of atten-tion – yet none of them can ever be viewedentirely in isolation from the other three. In Fig-ure 29, the tetrahedron has been positioned sothat the process vertex appears on top. But thewhole structure stands on the base formed bythe other three vertices, and the tetrahedroncould just as easily be tipped over to bring anyother vertex to the top, resting firmly on thebase provided by the other three.In Figure 29, the emphases which are char-

acteristic of the strategic choice approach ineach of the four ‘corners’ of the A-TOPP frame-work are first expressed in terms of generalstatements of orientation. This is to bring outthe essential contrasts withmore conventionalapproaches which tend to be well entrenchedin decision-making arrangements set up to dealwith comparatively clearly structured, boundedproblems. There is no intention here to sug-gest that these well-tried methods should beabandoned altogether; it is more a case ofadvocating a shift of emphasis, so that a moreappropriate balance can be struck.

In terms of technology, the strategic choiceapproach involves a shift away from a relianceon the kinds of expert techniques of solution-finding and evaluation which are characteristicof more routine decision-making arrange-ments. Rather, the orientation is towards whatcan be described as an open technology

intended to be freely accessible to participantswho have differing and complementary con-tributions to make. In situations where thevery shape of a problem may be obscure and,indeed, a matter of possible controversy, therelevance of what may be called ‘black box’technologies becomes considerably reduced.It is no longer enough to produce outputsof advice through the application of well-establishedmethods, the nature ofwhich neednot be disclosed because they are consideredso well validated by past exercises of a similarkind. Such ‘black box’ methods may still havetheir place; but it will be a place of limited sig-nificance in the wider process. For in problemsof complex and ill-defined shape, it is likely thatmany different types of choice will becomeinter-meshed – for instance, choices about fin-ance, about timing, about location, about tech-nological or marketing matters, about peopleand their roles, and about contractual or pro-cedural considerations. Such matters are, inturn, likely to cut across the preserves ofmany different experts; and it is only throughan emphasis on open technology – a techno-logy to support communication and interactionacross these boundaries – that anymomentumof joint working on complex problems can besustained.In terms of organisation, an orientation

towards interactive participation is indicatedas being characteristic of the strategic choiceapproach. This is in contrast to the emphasis onindividual working on assigned tasks which ischaracteristic of organisational arrangementsfor working on more clearly structured, recur-rent problems. There is, of course, a directlink between the emphasis on participativeforms of organisation and that on open tech-nology. Both stress the idea of people work-ing together in an exploratory way, so as to

71


transcend established boundaries of respon-sibility and specialist expertise. Again, the shiftof emphasis is not intended to suggest thatthe clear assignment of tasks to specific indi-viduals or organisational units no longer has itsplace. For this latter emphasis remains appro-priate in situations where predictable typesof issues have to be processed in a con-sistent and efficient way. So the questionis not whether such organisational arrange-ments should be replaced, but how they canbe complemented in circumstances wheremore challenging situations of strategic choicearise.In terms of process, the shift of emphasis is

from reliance on routine procedures for deal-ing with issues which fall into clearly recog-nisable categories, towards an acceptance thatpeople should be engaged in a learning pro-

cess about issues which no one person canclaim to understand in full. Again, the emphasison a learning process has a clear alignmentwith the emphasis on open technology and oninteractive participation. For, in responding tocomplex decision problems, people must beprepared to learn from each other, recognisingthat there may be a variety of complementarysources of insight and experience upon whichto draw. Again the emphasis on a learning pro-cess is not intended to suggest that routineprocedures no longer have a part to play; it issimply that they become less appropriate tothe more complex challenges of a process ofstrategic choice.Finally, in terms of products, the emphasis is

shifted from the conventional idea of problem-

solving, towards an orientation to incremental

progress through time. Where problems areof a bounded nature, it is reasonable to lookfor equally clear and definitive ‘solutions’. Butthere will always be some instances where itis not such a straightforward matter to decidewhat should be done. Action must sometimesbe postponed, referred elsewhere, or enteredinto in a partial or qualified way. In a processof strategic choice, this kind of progressivecommitment becomes more the rule than theexception.

INTERPRETING THE ORIENTATIONSINTO MORE SPECIFIC GUIDELINES

None of the four phrases used to describethe general orientations of the strategic choiceapproach – open technology, interactive par-ticipation, learning process, incremental pro-gress – can be considered as exclusive to thestrategic choice approach as presented in thisbook. Other people – practitioners as well asconsultants and scholars – have argued forthe adoption of these or similar orientations inresponding to complex problems. So, as gen-eral prescriptions, such statements now havequite a familiar ring. In particular, the virtues ofparticipative forms of organisation have beenwidely proclaimed, both in industry and thepublic sector, while the idea of public plan-ning as a learning process has now almostpassed into the realmof accepted conventionalwisdom.However, it has proved far from easy for

people to give effect to such changes of ori-entation in practice. They can all too easilyremain as pious exhortations, which evokea cynical response from those who considerthemselves to be realists in the art of decision-making under the practical pressures of organ-isational life. So the challenge is to find waysof interpreting the four broad orientationsinto more concrete operational terms. Thisis a challenge which is rarely faced in rela-tion to technology, organisation, process andproduct considered in combination; however,it is the challenge to which the strategic choiceapproach is explicitly addressed.In Figure 29, the first step in this direc-

tion is taken by interpreting each statementof preferred orientation into a more specificoperational guideline which reflects moredirectly the philosophy of planning as a pro-cess of strategic choice. So the orientationtowards an open technology is interpreted interms of a focus on decisions; the orientationtowards interactive participation is interpretedin terms of an emphasis on lateral connections;the orientation towards a learning process isinterpreted in terms of cyclic continuity; and

72

Orientations

the orientation towards incremental progressis interpreted in terms of a guideline of stra-tegic products.Each of these guidelines calls for further

explanation. This is done in the next four sec-tions of this chapter, drawing on particularaspects of the philosophy of strategic choiceas already presented. However, the value ofthese guidelines does not rest merely on theirphilosophical content. They are intended togive the statements of orientation amore prac-tical slant; and they serve to draw attention topractical issues that arise in the managementof the strategic choice approach, as applied tocomplex planning situations. So the next foursections will not only offer more precise mean-ings for the operational guidelines offered hereunder the four headings of technology, organ-isation, process and product; they will alsoindicate the nature of the practical manage-ment choices which they imply.

GUIDANCE IN TECHNOLOGICALCHOICE

If any technology of strategic choice is to betruly open, then it must offer a means ofbridging the differences in perception that canexist between participants who view theworldfrom different professional, organisational orcultural perspectives. The adoption of a focus

on decisions offers one pragmatic means ofdrawing the attention of participants towardsmatters which, individually or collectively, theyfeel to be important in their current planningtask. It is only to be expected that they will fre-quently see the decisions before them in verydifferent terms. They may disagree about theimportance of one decision relative to another,or they may have differing views about thelevel of specificity or generality at which mat-ters for decision should be expressed. Indeed,they can have different views about manyaspects of decision-making; but such differ-ences can provide a constructive focus fordebate with appropriate guidance – and thushelp the participants in moving towards amorerealistic sharing of views.

The centrality of the decision perspective isreflected in the focus on a current decisionproblem which was originally adopted as astarting point in developing a view of plan-ning as a process of consciousmanagement ofuncertainty through time (Figure 3). In the sub-sequent diagrams of Chapter 1, the symbolicshape of the ‘cloud’ representing this decisionproblem was variously expanded, pulled apartand examined from a range of contrasting per-spectives. Then, in Chapter 2, the decision per-spective was given more concrete expressionby adopting the decision area as the funda-mental unit of analysis, accepting that it couldbe used to represent choices about many dif-ferent kinds of things – investment, location,alignment of roads, to mention only a few ofthe types of choice which were introduced inthe telling of the South Side story. It is by theclearer articulation of such choices and theirrelationships that the form of the cloud repres-enting a current decision problem can bemadeless obscure. So the adoption of a decision per-spective offers one important means – nowwidely tested and adopted in practice – bywhich the orientation towards an open tech-nology can be given practical expression, andthe transparency of complex planning prob-lems thereby increased.It has to be recognised, however, that the

focus on decisions is not the only possibleemphasis that can be adopted in introducingan open technology. Other writers and practi-tioners have embraced with equal convictionthe idea of a systems perspective as a key to amore effective approach to complexity in plan-ning. However, in its more traditional forms,the adoption of a systems approach can be cri-ticised as focusing attention on some forms ofstructural relationships at the expense of oth-ers, thus working against the spirit of an opentechnology. Such a comment would apply, forexample, to the modelling of corporate plan-ning problems in terms of systems of financialrelationships only, or the modelling of cities interms of systems of land-use/transport inter-actions alone.

73

FIGURE

30Technological Choice: Settings and Equipment

Orientations

Meanwhile, some advocates of systemsthinking have moved towards what is some-times called a soft systems approach. Thisrecognises that complex phenomena can beexamined from many different perspectives,and that much can be learned by subject-ing different views of the purposes, bound-aries and inter-relationships of systems tostructured analysis and debate.2 So the adop-tion of a ‘soft’ systems approach offers analternative means of giving expression to theemphasis of an open technology. It does, how-ever, focus attention on relationships of a com-paratively stable, enduring form, as opposedto the ever-changing relationships betweendecisions – expressed at various levels andcarrying varying degrees of urgency – whichprovide the primary focus for the strategicchoice approach. The approach developed inthis book is, above all, intended as an instru-ment to help people plan under the pressuresof the continuously evolving circumstancesthat they face; which is not to say they cannotalso benefit from standing back occasionally toreach out for a broader, more systemic view.In pursuing the particular expression of an

open technology presented here, it is import-ant to see the scope of an appropriate techno-logy for strategic choice as embracing not onlya set of basic concepts and analytical methods,but also the choice of the physical settings inwhich people work, the equipment they useand the media through which they commu-nicate with each other. In the more conven-tional settings of sequential decision-making(Figure 6) – the clerk sitting at a desk, thecommittee sitting around a long table – suchaspects of technology are well established byconvention and therefore rarely considered ina conscious way. Background information ispresented verbally or in documentary form;while working on a problem, pens and papermay be used by individuals so that notes can betaken and action points recorded. In the case of

2 See in particular Checkland (1981) and Ackoff (1974).

the individual decision-maker, dealing with rel-atively bounded and recurrent issues, the elec-tronic work station has brought changes in thisfamiliar pattern. On the other hand, most col-lective decision processes continue to rely onthe simplest of technological equipment; andthere are good arguments for this, in so faras full rein can be given to people’s capacit-ies to interact with each other through verbaldiscourse and debate – freely augmentedby those nonverbal signals through which awider range of reactions and emotions can beexpressed.However, there are alternatives available to

the technology conventionally used to sup-port the processes of communication and ofinteraction; and some of these have beenfound to give fuller expression to the orienta-tion towards an open technology of strategicchoice. One alternative setting which has beenfound to be particularly effective is reflectedin the main sketch of Figure 30. This showsa group of people relating to each other notby sitting facing each other from fixed pos-itions around a table, but in an alternative set-ting which allows them to communicate andinteract in a more flexible way. The basic tech-nology in this kind of setting consists of:

• large sheets of paper that can be arrangedaround the walls of a room, which has to bespacious enough to enable people to movearound freely;

• a simple, non-permanent means of stickingpaper to walls;

• a liberal supply of wedge-tipped pens of sev-eral contrasting colours.

These materials are not, in themselves, verysophisticated; but many decision-makers areunaccustomed to their use, other than in rela-tively formal presentations. Used together,they have been found to play an important partin encouraging interactive working in groups.More specifically, they can encourage flexib-ility in use of the various graphical methodsfor expression of decision problems and theirimplications which have been introduced in

75


earlier chapters. It has to be recognised, how-ever, that successful group interaction is notautomatically assured merely by creating anappropriate setting, and providing appropriateequipment and tools. It is always possible thatone or two members of a group will tend todominate the process, with others reacting bybecoming alienated or withdrawn. Therefore,among the variations in strategic choice tech-nology that will be offered in the next fivechapters, some suggestions will be offered forhelping members of a group to become morefully involved. For example, this can be doneby creating opportunities for people to draw upindividual lists of decision areas or uncertaintyareas, before attempting to merge these into ashared picture on larger sheets of paper rangedaround the walls of the room.Figure 30 includes a reminder that it is not

exclusively when they are working in groupsthat people can draw on the basic conceptsand methods of the strategic choice approach.As several experienced decision-makers havefound, ideas such as the UE/UV/UR classifica-tion of uncertainty can also be helpful to themwhen they are reflecting alone on difficult prob-lems – as in the symbolic armchair – or whenengaged in informal discussions with one ortwo colleagueswho share an awareness of thebasic language and ideas. It is in such individualand small group settings too that the computercan most readily play a role in augmenting thetechnology of communication and interaction;for, as will be demonstrated in Chapter 10, itcan then be used as a facilitation tool in guid-ing decision-makers through the various pathswhich are open to them in an exploratory yetdisciplined way.However, within the overall technology of

the strategic choice approach, it is importantto regard the computer as no more than anoptional resource. Even at the time this thirdedition goes to press, very many of the recor-ded applications of the approach have madelittle or no use of computer methods. Yetas the technologies of human-machine inter-action and telecommunications continue their

rapid advance, new opportunities will inevit-ably continue to arise for harnessing electronicresources to the support of strategic decisionprocesses; not least, where those processesinvolve collaboration or negotiation throughglobal communication networks.The state of progress so far in developing

software for strategic choice will be reviewedin Chapter 10, along with the longer-termopportunities for development. What has tobe emphasised at this point is that computermethods should in no way be regarded as anessential element in the technology of the stra-tegic choice approach.In pursuing the emphasis on an open tech-

nology, the general rule is to aim continuallyfor simplification rather than elaboration, sothat high levels of effective communicationand interaction can be sustained. However, itis important to remember that there are alwayschoices of balance to bemade in this and otherdimensions of the approach (Figure 2). It is onlyrealistic that the appropriate point of balance atany moment should be judged in the light ofthe participants’ local appreciation of their ownpresent problem situation, and the pressureswithin which they must work.

GUIDANCE IN ORGANISATIONALCHOICE

Turning at this point to the organisationaldimension of the A-TOPP framework, therecommended orientation towards interactiveparticipation is by no means as straightfor-ward to interpret in practice as might at firstappear. For the wider the scope of the prob-lem being addressed, the broader the rangeof organisational and other interests likely tohave a stake in that problem. Then the harder itbecomes to reconcile the desire that all theseinterests should be adequately represented

with the desire to encourage interactive work-ing. For the larger and more diverse the group,the more difficult it becomes for all to feelinvolved, and for progress to be sustained.In addressing the choice of organisational

arrangements to deal with complex planning

76

Orientations

problems, the conventional approach is tofocus attention on the design of comparativelystable, clearly structured frameworks of hier-archical guidance. These usually incorpor-ate elements of conscious forecasting anddirection-setting as well as managerial control,and are shaped to fit closely the establishedmanagement structure of the corporation con-cerned. The intention is – at least in theory – forhistorical inertias to be questioned; for peopleto come together to work across departmentalboundaries on selectedmajor issues; and oftenalso to involve representatives of other signi-ficant interest groups, such as employees, res-idents or public service clients.However, the basic management struc-

tures of most corporate organisations canstill be seen as designed primarily as ameans of responding efficiently to certainrecurrent categories of decision problems,for which clear policy guidelines and rulesof delegated responsibility can be defined.Where problems are of a more fluid, indistinctshape, experience shows that formal plan-ning arrangements which reflect the estab-lished corporate structure can quickly breakdown. It is then that more informal, adaptiveapproaches take over – with the impetus usu-ally coming from those points in the man-agement system where the pressures fordecision are most intense. It is in the faceof repeated pressures of this kind that organ-isational arrangements for planning based onthe hierarchical control structure begin tolose credibility, because their contribution tothe making of key management decisionsbecomes increasingly hard to demonstrate.So, to express the decision-centred

philosophy of the strategic choice approach,an alternative guideline to the design oforganisational arrangements is required. Asan operational guideline, intended to reflectthe orientation towards interactive participa-tion as against individual working, the ideaof a focus on lateral connections will nowbe introduced. The significance of this termin relation to a decision perspective is mademore clear in Figure 31, which is an extension

of the view of three contrasting responsesto uncertainty in decision-making as firstpresented in Chapter 1 (Figure 3).Figure 31 starts from the situation of an indi-

vidual or working group currently concernedwith some particular decision or planning prob-lem – whether it be complex or more limitedin scope – and experiencing some degree ofdifficulty in deciding what should be done.Different aspects of this current difficulty canbe analysed in relation to the three basic cat-egories of uncertainty – UE, UV and UR –and each of these types of uncertainty canbe seen as having organisational implicationsof a somewhat different kind. In each direc-tion, too, there may be various levels of invest-ment to consider, reaching out increasinglyfar beyond the particular organisational set-ting in which the problem is currently beingaddressed. The possibilities may range fromthe level of quick and informal contacts withpeople who are relatively close in organisa-tional terms, to the level of more deliberate,and possibly formal, approaches to others whoare organisationally more remote.When working outwards in the UE direction,

the concern is essentially to acquire additionalinformation relevant to the decision-makers’working environment. Some kinds of informa-tion can be acquired easily enough by meansof informal enquiries from close colleagues;others may involve somewhat more formalapproaches across departmental or corporateboundaries, while others again may mean ini-tiating substantial survey, assessment or fore-casting exercises, perhaps involving formalcontractual arrangements with specialist con-sultants.Turning next to the UR direction, there may

again be several possible levels of investmentin outward linkage to consider. However, thepeople approached now appear not merely inthe role of knowledge providers, but also in therole of decision-makers in related fields. So anycommunication across boundaries will tend totake on a rather different flavour, involvingmore explicit elements of negotiation in so faras each party may be in a position to help the

77

FIGURE

31Organisational Choice: Levels of Response to Uncertainty

Orientations

other in reducing some of the uncertaintiesthat they face. The organisational options to beconsidered in the UR direction can range fromoccasional bilateral exchanges – whether bytelephone or other means – tomore interactiveworking arrangements calling for substantialtime commitments by the parties concerned.Turning finally to the UV direction, the most

familiar and straightforward means of seek-ing to reduce uncertainty about guiding valuesinvolves referring upwards within a hierarchyto a superordinate point of authority. Beyondthis, there may be a possibility of referringfor guidance to a meeting of some collect-ive body – a Board, Council or Committee –in which formal authority resides; acceptingthat at this level there may be many compet-itive pressures on its agenda and only limiteddiscussion time available. Beyond this again,there may be opportunities to launch specialconsultative exercises to probe the value pos-itions of interest groups which have a majorstake in the present decision problem,whetherby use of existing representative channels, byconvening openmeetings, or by some combin-ation of the two.The successive levels of response shown in

Figure 31 are intended to give no more than abroad indication of the opportunities for work-ing across organisational boundaries thatmightemerge at any moment in a planning process.Somemore specific examples were presentedin Chapter 3 (Figures 24 and 26), and furtherillustrations will appear in Chapter 8. Some-times it will be appropriate to refer upwardsor downwards within a hierarchy, especiallyas a means of addressing significant areasof uncertainty of the UV type. However, themain orientation in strategic choice is essen-tially towards building connections in a lateral

way, working outwards from a specific prob-lem focus rather than downwards from a pointof central authority.Often, there will be choices to be faced as

to whether particular connections should beactivated through formal or informal channels.In particular, there is always the possibility that

conscious investment in building close work-ing relationships with particular individuals willprovide a foundation for quick and informalconsultation later, when particular issues orcircumstances arise. In managing a processof strategic choice, it is generally the choicesabout lateral working in the UR direction thatpose the most subtle challenges; for, as wasearlier indicated (Figure 5), it is in this direc-tion that the boundaries of the decision prob-lem itself become enlarged. So, the questionnowemerges as to how far the range of peopleinvolved in interactive working can be exten-ded without the momentum of progress beinglost.Taking these considerations together, the

significant areas of management choice inorganisational terms can be seen as extendingbeyond the formal structuring of arrangementsfor internal co-ordination, as the hierarchicalcommand perspective of planning organisa-tion might suggest. For important choicesmust also be made about arrangements forbilateral linking and multi-lateral grouping, inresponse to particular problems as they arise.The choices about linking relate to the informalcultivation and activation of networks of inter-personal working relationships by specific par-ticipants in the process. On the other hand, thechoices about grouping relate to judgementsas towhen and how to initiate interactive groupworkingwith others, or to enlarge the boundar-ies of an existing interactive process. It shouldnot be forgotten too that choices may arise asto when and how to disband a groupwhich hasoutlived its usefulness, or otherwise to modifythe lattice of arrangements through which aprocess of interactive working is pursued.In confronting such choices, it has to be

recognised that invitations to participate ininteractive group processes will not always bewelcomed without reservation. Resistancesmay be encountered because of fears aboutloss of autonomy, or straightforward timepressures, or conflicts of interests, or vari-ous other factors of a broadly political kind.Factors such as these can contribute tomakingthe emphasis on interactive participation more

79

FIGURE

32Process Choice: Opportunities for Movement Between Modes

Orientations

difficult to interpret in practice than it mightappear. The point to be underlined is again thatthere are always choices of balance to be con-sidered. Not least, there is always a balanceto be struck between the extremes of react-ive and interactive working (Figure 2). This isa balance which must be expected to changethrough time in an adaptive, evolutionary way –and which can only realistically be judged bythose directly involved in the process.

GUIDANCE IN PROCESS CHOICE

To talk of a process of strategic choice as beingessentially a learning process may seem to beno more than to repeat a statement which,applied to planning, has nowbecome sowidelyaccepted as to have become little more than acliché. But this orientation, like the correspond-ing orientations towards open technology andinteractive participation, has proved to be verydifficult to put into practice in a satisfactoryway. This is because the question of what itis most important to learn in relation to com-plex problems can be by no means easy toanswerwith any confidence in advance. Often,it is only after interactive work has begun onsuch a problem that the participants can beginto form any kind of informed, coherent view astowhat themost serious obstacles to progressare, and therefore what course their learningprocess should take.Most of the more conventional approaches

to planning and complex decision-making haveadvocated a process based on clearly spe-cified sequences of steps or stages, followingthe norm of linearity which is widely acceptedin the design of procedures for dealing withmore recurrent, clearly structured decisionproblems. Following this principle, specifictimetables are laid down for consecutivestages of the planning process, covering suchactivities as survey, formulation of objectives,design of alternatives, evaluation and plan pre-paration. In those situations where the aimis to add a more purposive direction-seekingemphasis to an existing system of corporatemanagement and control, such steps are then

fitted into a periodic – often annual – plan-ning cycle. Arrangements of this kind havethe virtue of apparent predictability; this canbe important in organisational settings wheresubstantial human, financial or other resourcesare being committed to the planning task,creating a demand that their use be subjectto due procedures of accountability. However,the risk is that the requirement to adhereto a pre-ordained sequence of activities willconstrain people to concentrate their effort onlearning the wrong things at the wrong times,so far as the more important pressures fordecision are concerned. Instances frequentlyarise, both in business and in public plan-ning, where lengthy and expensive exercisesin survey, forecasting or evaluation are carriedout to a pre-arranged timetable, yet turn outto yield results of little relevance to the morecrucial problems of the day.So, in describing the process emphasis of

the strategic choice approach, the generalorientation towards a learning process willbe interpreted more specifically in terms ofan operational guideline of cyclic continuity.In this, the conventional representation of aprocess in terms of a linear sequence of stagesis replaced by amore flexible set of possibilitiesfor movement between one mode of decision-making and another. It may be useful for somepurposes to follow an agreed sequence of pro-gression to ensure that all modes are con-sidered in a methodical way; but to maintainthe emphasis on a learning process in a com-plex and evolving problem situation, decision-makers must be prepared to adopt a muchmore flexible approach.Often the shape of a problem is only dimly

perceived, and there are many partially formedviewpoints to be exchanged. Then the parti-cipants have to be ready at any time to movebackwards tomodes in which substantial workhas already been done, so as to challenge andmodify past assumptions. Equally, they mustbe ready to skip forward to modes where itmight seem premature to work in any depth,according to a more conventional sequentialview. Indeed, the very concept of a decision

81


area implies a readiness to look ahead to whatis to be decided at various points in the future,effectively skipping forward from the shapingto the choosing mode.In Figure 32, the guideline of cyclic continuity

is expressed more specifically in terms of thetypes of process management choice which itimplies. Starting from the view of a processof strategic choice as involving a continuinginterplay among four basic modes of activity,there are two types of choice that can bemadein considering how to direct the learning pro-cess. First, there are choices about when andhow to switch out of the primary mode ofthe moment and into any of the others, for aspell of deliberate working within that othermode. Then there are those more frequent,often informal, choices that peoplemake aboutwhen and how to loop briefly outwards fromthe primary mode of the moment in the direc-tion of anothermode, to deal with some limitedsource of difficulty, before returning to pick upagain the momentum of work already estab-lished in the primary mode.Much of the advice to be offered in the

next four chapters relate to forms of switchingand looping judgement that people can con-sider when working in each of the four modes.Therefore, each of the next four chapterswill end with a diagram which elaborates therelevant ‘corner’ of Figure 32 by summarisingthe types of switching and looping judge-ments that can be made when working in thatparticular mode.The broader significance of the guideline

of cyclic continuity is that, when peopleare consciously following the strategic choiceapproach, the conventional linear processof agenda control, as regulated through anappointed chairperson, no longer need apply.Instead, it becomes replaced by a more subtle,adaptive process of route finding with manypossible paths among which to choose. In thisprocess, both switching and looping judge-ments have to be made continually. Briefloops out of the primary mode of the momenttowards another mode are often made sub-consciously; however, in group working it

can be a significant contribution to the learn-ing process if they can be more consciouslyrecognised and discussed. Switching judge-ments, on the other hand, have a wider signi-ficance in the management of the process asa whole. So it is all the more important thatthey should be openly debated within a group,if a conscious emphasis on an interactive learn-ing process is to be maintained. More specificadvice on how to manage such switches ofmode in the course of interactive group work-ing will be given in Chapter 9, reflecting accu-mulated practical experience as to how best tosustain the emphasis on relevant learning at alltimes.The significance given to switching and loop-

ing judgements here does not mean thatmore conventional scheduling decisions canbe ignored in the management of a processof strategic choice. Each participant in the pro-cess has inevitably to make at least someadvance arrangements to organise his or herworking life, inwhich therewill usually be otherrelatively fixed commitments to fulfil. So peri-ods of interactive working will often have tobe scheduled in advance, even if the organisa-tion of work within such a period is kept asflexible as possible by following the precept ofadaptive route-finding in place of that of linearagenda control.In practice, problems can arise in recon-

ciling the demands of cyclic continuity withexpectations of linear procedure, not only inthe advance scheduling of work but also inaccounting to other people retrospectively forthe way in which any conclusions or recom-mendations have been reached. To argue thata particular set of proposals has been arrivedat as a result of a cyclic, adaptive, learning pro-cess may be to do no more than describe therealities that underlie any process of workingat complex problems. But it is only natural thatthose not directly involved should wish to seesome kind of rational argument by which theconclusions offered can be justified, withoutbecoming drawn into the full subtleties of thelearning process. The requirement for somedegree of ‘post-rationalisation’ can be a vital

82

Orientations

one in some circumstances, and it will be thesubject of further discussion in Chapter 9.

GUIDANCE IN PRODUCT CHOICE

The shift in orientation of the strategic choiceapproach in relation to the products of the pro-cess is expressed (Figure 29) in terms of a shiftaway from an emphasis on definitive resolutionof problems at specified end points of the pro-cess, and towards a more subtle emphasis onmaking incremental progress through time.Where problems can be treated as simple anddiscrete, it may make sense to deal with themin accordance with some linear schedule oragenda, with the aim of settling each itemof business in full before moving on to thenext.When people encounter issueswhich aremore interrelated and complex, the tendency isoften to translate this desire for complete solu-tions into a concern with the production of sub-stantive plans at as comprehensive a level aspossible. The conventionalwisdom is that suchplans should be implemented in an equallycomplete manner through more specific oper-ational decisions, budgets and programmes,after they have been steered through therequired procedures of authorisationwithin theorganisation.However, it is a common experience for

people to encounter very considerable diffi-culties in attempting to conform to suchprinciples in practice. Indeed, the more com-prehensively they may seek to view the scopeof a strategy or plan in terms of the range ofsubstantive issues it should cover, the morethese difficulties tend to proliferate. It is in suchcircumstances that the alternative orientationtowards incremental progress comes into itsown.Within the strategic choice approach, the

operational guideline which is offered as amore specific interpretation of this idea isthat of generating a balance between differentkinds of strategic products through time. Thespirit of this idea is captured in the conceptof the commitment package, as introducedtowards the end of Chapter 3 (Figure 26).

The structure of the commitment packageis designed to help people in thinking andtalking about what level of balance betweencommitment and flexibility is appropriate at anymoment of a continuous process. This entailstheir weighing up the various urgencies anduncertainties impinging on the choices withwhich they are currently concerned, and alsoconsidering how foundations for the makingof future decisions can be built. Such consid-erations are of only marginal value in a con-ventional problem-solving approach, but are ofprofound importance in a process of strategicchoice where the need for continuity is expli-citly recognised.The concept of the commitment package

provides a particularly clear expression of theprinciple of treating commitment as a variable

rather than as something to be conceived inall-or-nothing terms. Yet a commitment pack-age is essentially a product of the work of thechoosing mode; and it is important to recog-nise that there are also products of work in thevarious other modes of strategic choice whichcan make a direct contribution to progress ina broader sense. A wider view of the varioustypes of strategic product which can contrib-ute to the momentum of progress in strategicchoice is offered in Figure 33.Figure 33 presents a two-dimensional frame-

work for identifying different types of strategicproduct, first in terms of a distinction betweenproducts of substance and those relating to theprocess; and, secondly, in terms of a distinc-tion between visible products and those of amore invisible or intangible form. The essenceof the distinction between substance and pro-cess can be illustrated by referring again to thecommitment package framework (Figure 26),the aim of which is, essentially, to present theproducts of work in the choosing mode in aclearly visible form.Within a commitment package, the products

which relate most directly to the substance

of the decision problem are those appearingin the first column, concerned with imme-diate actions. It has to be recognised that

83

FIGURE

33Product Choice: A Classification of Strategic Products

Orientations

these actions may take various forms, includ-ing statements of policy and agreement onresponsibilities, budgets and deadlines, aswellas the commitment to here-and-now actionsunder the direct control of those involved in theprocess. There is also a substantive contentwithin the right-hand half of the commitmentpackage framework – that relating to the futuredecision space. For this includes informationabout decisions which have been deferred orare seen as contingent on future events.The existence of process elements within

a commitment package is implicit in its verynature as an incremental step in a continuingprocess. Those parts of a commitment pack-age which are most directly process relatedinclude the arrangements for managing uncer-tainty, along with any proposed arrangementsfor scheduling; for future decision-making pro-cedures; and for linking with other people witha part to play in the wider process.There are also other kinds of visible products

which are not directly represented in the com-mitment package framework. These includeany documented evidence of progress madethrough work done in the shaping, designingand comparing modes. Work done in thesemodes can lead to clearer expressions of theparticipants’ shared views about:

• thenatureof thedecisionproblemsthey face;• the range of options or schemes available;• the consequences of different alternatives;• the uncertainties that make it difficult toexpress preferences between them.

It is common in planning for such kindsof information to be assembled as part ofa reasoned justification of recommendedcourses of action. There are variousconventional forms – written, diagrammatic,numerical – in which such information aboutproblems, opportunities, criteria, constraintsand sources of uncertainty can be assembledand documented whenever importantmoments of formal commitment to policiesand actions arise. But the various conceptsand conventions illustrated in Chapters 2 and3 are specifically intended to help people build

up a visible record of progress in all modesin a more continuous way. This is of partic-ular importance when work is proceedinginteractively in groups. There is still a task ofinterpreting visible products expressed in suchways into forms which are intelligible to thosenot directly involved; but there are practicalways of addressing this task, and these willbe discussed further in Chapter 9.Important as these various visible products

may be, it is vital also to recognise that anyprocess of strategic choice generates invis-

ible products which can have a profound influ-ence in the longer term. First, in terms of thesubstance of problems, people can expect toextend their perceptions in ways that cannotbe fully reflected in any documented evidenceof progress that is produced. Such exten-ded perceptions may embrace all aspects ofsubstantive progress already discussed at themore visible level, including problems, oppor-tunities, effects and uncertainties. Such per-sonal shifts in perception will generally be inthe direction of increased sharing of viewsamong those involved in interactive working;or increasing intersubjectivity, to use a termwhich others have found useful in relationto group processes (Eden, Jones, Sims andSmithin, 1981). Moves in this direction canbe of considerable significance beyond theboundaries of the group involved in the inter-active working process; for they can affectthe ways in which individuals seek to influ-ence the views and behaviours of colleagues,superiors and associates within their widerworking environment.Also, more realistic perceptions of the

external political realities surrounding a prob-lem can be an important invisible product of aprocess of strategic choice. However, such aproduct must be seen as relevant as much tofuture process as to the substance of partic-ular problems; so this leads naturally into thediscussion of the final quadrant of the productclassification of Figure 33.

85

FIGURE

34Technology, Organisation, Process and Product: An Interim Review

ORIENTATIONS

Orientations

Among the important invisible productswhich relate more to process than to sub-stance is fuller understanding of:

• other people’s values;• their ways of working;• the pressures and constraints acting onthem.

In most cases, this is backed up by extensionsto the communications networks of individualsinvolved in the interactive working.In interactive working, it is important to con-

siderhowfarthefull rangeofvisibleandinvisibleproducts can be expressed in more accessibleforms, so that they can be better understoodat a collective level. This raises important man-agement choices in the areas of recording andinterpretation. Recording is simply a questionof keeping track of progress, as it is made.Various methods for recording progress onthe substance of problems – and the morevisible aspects of progress on the processside – have already been illustrated through thetelling of the South Side story in Chapters 2and 3; but recording of invisible progress interms of process is often less straightforward.However, conscious efforts to document shiftsin understanding of others, and in communi-cation networks, can be important if theseaspects of progress are not to become lost.It is important to recognise that any con-

scious documentation of progress requires aprocess of interpretation as well as record-ing. This must involve the translation of jar-gon used in group working – whether verbalor graphical – if progress is to be reviewedin terms accessible to others rather than intermswhich are intelligible merely to those dir-ectly involved. This is important enough in thedocumentation of visible products, but is evenmore essential in attempting to document pro-gress at the more invisible level. It is especiallyimportant, from a longer-termperspective, thatpeople should be helped to recognise invisibleproducts through explicit review of the direc-tions in which progress is – or is not – beingmade. Some practical advice on these aspectsof the management of strategic products willbe presented in Chapter 9.

ORIENTATIONS OF STRATEGICCHOICE: A GENERAL REVIEW

The various points made under the four head-ings of Technology, Organisation, Process andProduct are summarised in Figure 34. This isintended to form a point of reference in follow-ing the more concrete guidance to be offeredin the next five chapters. The first two columnsdo little more than present in a more compactform the contrasts in orientation which werehighlighted earlier (Figure 29), where the struc-ture of a tetrahedron was introduced to conveythe essential unity of the A-TOPP framework.So, in general terms, the strategic choice

approach is interpreted in terms of shiftsof emphasis from conventional ideas abouttechnology, organisation, process and producttowards the four preferred orientations of opentechnology, interactive participation, learningprocess and incremental progress. These com-paratively general, exhortative statements arethen translated in the second column ofFigure 34 into the four operational guidelinesof focus on decisions, lateral connections,cyclic continuity and strategic products. Allthese can be seen as linked to the more gen-eral guideline of the conscious management

of uncertainty, which distinguishes the stra-tegic choice approach as a practical meansof responding to complex problems. Each ofthese four guidelines can be contrasted withmore conventional guidelines for respondingto complexity which, in practice, are often dis-appointing in their effects. These more con-ventional guidelines – systems perspective,hierarchical command, prescribed stages andsubstantive plans have all been used as pointsof reference in discussing the strategic choicealternatives in the four preceding sections. Sothese too are brought within the comparativeframework in the second column of Figure 34,linked to amore general guideline of consistentpolicy at the level of overall approach.The third column of Figure 34 extends

this framework to indicate the main areas ofmanagement choice which it is important tobear in mind when attempting to follow the

87


recommended operational guidelines in prac-tice. Again, these areas ofmanagement choicehave already been introduced in the precedingsections of this chapter; but it will be helpful togather them together here.

• In terms of technology, it is important to keepin view choices about ways of interactingand facilitating as well as about methods ofanalysis as such.

• In organisational terms, the importantchoices concern ways of linking and

grouping across sectional or organisationalboundaries, as well as about the design ofmore formal co-ordinating structures.

• In process terms, there are choices ofswitching and looping to be consideredwhile working interactively, as well as themore familiar choices about scheduling ofmeetings and deadlines.

• In terms of products, there are choicesto be made about ways of recording

and interpreting progress, as well as themore familiar choices about production offormal reports which emphasise substantiveactions and policies.

All these aspects of management choice canbe seen as describing different facets of amore general task of managing the strategicchoice approach. This means maintaining aspirit of selectivity and adaptiveness in plan-ning and decision-making, as contrasted witha more familiar emphasis on the regulationof more simple decision-making tasks. Yetanother level of contrast is added by the fourthand final column of Figure 34, which intro-duces a set of contrasting emphases in termsof the continuous monitoring of the qualityof a process of strategic choice. In brief, themessage is that judgements about the use oflimited resources have to be approached notjust in terms of some clearly defined criterionof optimality but more in terms of a broaderconcept of effectiveness. In the successiverows of the framework, this concept is seen asembracing such concepts as sharing, synergy,

understanding and confidence, in addition tomore conventional indicators of success.These evaluative issues will be discussed

more fully in Chapter 9. They remain relevantwhether the concern is to make quick pro-spective judgements about what to do nextin the course of an interactive process, or toreview progress in a more reflective vein. Theimportant point is that it is not necessary to dis-card completely such traditional concerns asaccuracy and rigour in technology; efficiencyand control in organisation; productivity andtimekeeping in process; or accountability andcompleteness in product. It ismerely that othercrucial considerations have to be explicitly con-sidered as well. Indeed, this point applies to allthe contrasts offered in Figure 34; to advocatea shift in emphasis from the top left section ofeach square towards the bottom right is not tobe taken as implying that these more conven-tional emphases do not have their place. It hasto be recognised that all organisations requirea base of stability and predictability on whichtheir more adaptive activities can be built; andthe concern here is that this should be effect-ively complemented rather than undermined.The concern with identifying areas of choice

and evaluation which emerges in the third andfourth columns of Figure 34 may suggest thatsome of the core strategic choice conceptssuch as those of the decision area and the com-parison area could be applied in structuring themanagement of the process of strategic choiceitself. There is no reason in principle why thisshould not be possible. However, any explicituse of strategic choice methods for this pur-pose must not distract the participants fromthe real problems of substance that they face.So it is suggested that Figure 34 be treated asbackground rather than foreground in readingthe chapters that follow. Working on the well-established principle that there is nothing sopractical as a good theory, this framework canbe referred to as andwhen required, as a broadphilosophical context within which all the prac-tical guidance offered in the next five chaptersshould make sense.

88

5 Skills in shaping

INTRODUCTION

This chapter offers some practical guidance onapproaches to the shaping of complex decisionproblems. It does so by building on the basicconcepts of the shaping mode which werepresented in the first half of Chapter 2 – theconcepts of the decision area, the decisionlink, the decision graph and the problemfocus. Each of these ideas is quite simplein its essence. Within the overall technologyof strategic choice, they together provide afoundation for the more technical activities ofgenerating options and exploring their mutualcompatibility which distinguish the work of thedesigning mode.The importance of work in the shaping mode

arises from all the difficult and subtle judge-ments that can be involved in expressing prac-tical problems in terms of decision areas andlinkages between them, and then in agreeinga problem focus on which to work within theresulting decision graph. This can be an espe-cially challenging task in situations where thenature and scope of the choices to be madeare far from clear cut and where there aremany participants in the process, each witha different understanding of what is at stake.Experience has repeatedly shown that it can beworthwhile spending a substantial amount ofgroup time working to achieve more satisfac-tory formulations of a decision graph. This maymean changing the ‘map’ of decision areasand their connecting links several times, as thelevel of shared understanding grows. Indeed,such a process of recycling is of value not onlywhen a group of people is still in the earlystages of working together. It can also be wellworthwhile for the group to return and carry

out further work in the shapingmode after theyhave consciously moved ahead and investedeffort in the work of the designing, comparingand choosing modes. This point is inherent inthe guideline of cyclic continuity in the process,which was emphasised in Chapter 4 and gaverise to the various switching and looping oppor-tunities indicated in Figure 30. Taking a longer-time perspective, the idea of the commitmentpackage demonstrates how incremental pro-gress towards commitment will usually meandeferring choice within some of the decisionareas within a decision graph. Any such defer-ments will, in turn, create opportunities for thereshaping of decision areas and their relation-ships at future points in time – when otherquite new areas of choice may possibly havecome to the fore through the unfolding ofexternal events.In Chapter 2, it was suggested that the

choice of problem focus within a widerdecision graph marked an important point oftransition between the work of the shapingand the designing modes. There are, how-ever, many different considerations which canbe taken into account in this process of ‘cut-ting a problem down to size’. Some of theseconsiderations – urgency, degree of import-ance, level of interconnectedness, degree ofinfluence on the part of a particular set ofdecision-makers – may be much more import-ant in some decision contexts than in others.Indeed, their importance can vary with the pas-sage of time, even within the same decisioncontext. So the whole process of problem-shaping can be seen as one in which a con-tinuing tensionmust be expected between thedesire to expand the boundaries of a prob-lem to encompass all conceivable elements of

89

FIGURE

35

SOUTH SIDE

EXAMPLE

Building up a Tentative List of Decision Areas

The building up of an unstructured list of decision areas such as this can be a useful step when a group

is just starting work. The list can be extended, altered or restructured as understanding grows, with

some entries transferred to lists of comparison areas or uncertainty areas. Individuals can be invited to

list their own suggestions either at the start or later in the process.

Skills in Shaping

choice, and the desire to keep its dimensionsmore manageable so as to secure a sense ofprogress towards action.

APPROACHES TO BUILDING UPA SET OF RELEVANT DECISIONAREAS

Given some agreement to carry out work in adifficult but as yet unstructured problem field,the task of beginning to build up a set ofdecision areas to represent the choices withinthat field is one that can be organised eitherat an individual or a group level. It is broadlyakin to the process of ‘agenda building’ whichis, literally, no more than the construction ofa list of things to be done, or matters to beaddressed. However, in strategic choice, thetask is not simply to arrange these things to bedone into some appropriate order so that theycan be tackled sequentially. This is becausesome of them at least are likely to be interre-lated, in a way which implies that they shouldperhaps be grouped or clustered so that theycan be examined together.Wherever an individual or group is working

to a reasonably well-defined remit, as in theproblem of looking into the implications of theproposed new road through South Side, thenit may not be too difficult to decide where theprocess of building up a list of decision areasshould begin. For example, among the sevendecision areas listed in Figure 10, it might bethat the first four at least would suggest them-selves fairly readily to anyone familiar with thelocal planning context. Yet in other circum-stances, where the remit may be broader orlooser, and the elements of choice less clearor less urgent, difficulties are to be expected inagreeing the way in which even themost basicelements of choice should be expressed.Such difficulties may, of course, arise even

where an individual is working alone; butwhere people are working as a group they canbe expected to surface rapidly in the form ofdebates and disagreements between one per-son and another. Usually, the different mem-bers of the group will bring different kinds of

perception to bear, rooted in different kinds ofexperience, different bases of professional ortechnical knowledge and different responsibi-lities to people outside the group, apart fromany deeper differences in ways of construingthe world around them. Such differences canbe valuable in so far as they can help the groupdevelop a more rounded appreciation of thechoices which they face.In South Side, for example, it might be

expected that any engineers in the groupmightsee the relevant choices in terms of roadalignments, construction techniques and othertechnical matters to dowith elevations, accessroads, surface drainage arrangements or bidsfor resources on capital programmes. Urbanplanners might see them in terms of opportun-ities to specify intended uses or other controlsrelating to particular areas of land; economistsmight see them in terms of investment levels,amortisation of fixed assets, balance betweenprivate and public expenditures, and incomes.Then again lawyers, administrators, projectco-ordinators, politicians and any other par-ticipants in the process might all have theirsubtly different perspectives on the range andcontent of the choices to be considered.In a group setting, one good way to start

the process of building up a set of decisionareas is simply to begin with a large sheet ofblank paper on the wall, on which an initiallist of issues or matters of concern is recor-ded by one participant in such a way as to beclearly visible to all the others. Figure 35 givesan example of the kind of tentative list thatmight be built up by the South Side WorkingParty at the outset of its work, before any kindof sorting or grouping of decision areas hasbeen attempted. What matters at this stage isprimarily to encourage the involvement of asmany participants as possible in the listing ofan initial set of elements, however imperfect itmay seem. The spirit called for is one of organ-ised brainstorming, in which ideas put forwardby one participant can trigger off new direc-tions of thought by others.An alternative way of starting is to ask each

participant to spend a few minutes writing

91

FIGURE

36

SOUTH SIDE

EXAMPLE

Sorting Decision Areas by Categories and Levels

It can be useful to rearrange decision areas by fields or levels — or both — whenever an unstructured

list seems to be becoming too long or unwieldy. This can help in redefining or clustering similar decision

areas; in considering the possibility of an multi-level problem structure; and in choosing a problem focus

where there are many decision areas.

Skills in Shaping

down a list of a few decision areas he or shebelieves to be important, before the interact-ive group process begins – or at least before ithas built up a momentum of its own. If eachindividual uses a separate sheet of paper orcard to record each suggested decision area –and if these sheets are not too large thenthe results can be shared by spreading all thesheets prepared by the participants out on atable top or, perhaps, merely on a patch ofclear floor space in the centre of the room.Wall space can, of course, be used for dis-playing these sheets of paper, provided thereis a simple, non-permanent means of stickingpaper to the wall.This kind of approach can be helpful in gen-

erating an initial sense of involvement by allparticipants in a group. Of course, it is alsolikely to lead to a degree of redundancy in theinformation generated, because some of themore obvious decision areas are likely to bethought of independently by different people.But this kind of redundancy may be no badthing at an early stage in a group process, bothbecause any area of choice which is identi-fied by many people can be given more weighton that account, and because the elements ofa problem may be perceived in subtly differ-ent ways by different participants. Such differ-ences can then provide a focus for constructivedebate when it comes to the point of formulat-ing decision areas in more precise terms.In this kind of activity, it is only to be expected

that doubts will arise as to whether some sug-gested elements of a problem should be con-sidered as decision areas at all. Some areas ofchoice may seem to lie largely outside the dir-ect sphere of influence of the participants, inwhich case they might perhaps seem betterexpressedasuncertaintyareas.Other ‘decisionareas’ offered may seem more in the natureof statements or choices to do with general-ised aims, criteria or desiderata, in which casethere would be a possibility of reserving themto be introduced later as comparison areas.As will be seen later, comparison areas anduncertainty areas can always be transferred to

separate sheets of paper on the wall, for fur-ther consideration when the process movesinto other modes. However, there is no reasonwhy they should not be included in a singleunstructured list of problem elements in thefirst instance, with a view to further rearrange-ment when the group decides to move ahead.

SORTING DECISION AREAS BYCATEGORIES AND LEVELS

Where the problem to be tackled has no clearboundaries, and especially where many partici-pants are involved, it is not uncommon forthe process of generating decision areas todevelop a momentum of its own. Each newaddition to the list can then prompt furthersuggestions; so that the process begins toappear a never-ending one. But once the list ofdecision areas reaches around 15 or 20 – as inthe example of Figure 35 – then questions willbegin to arise as to whether it is useful to con-tinue any further in this way. Participants maythenwell begin to askwhether, before thinkingof any further additions, those decision areasalready on the list should not be sorted orrearranged in some way, to identify possibleoverlaps and gaps in the problem formulationas it has shaped up so far.One obvious way in which to begin to

restructure a list of decision areas is to intro-duce some framework of categories of choicewhich are felt appropriate to the presentdecision setting. In a setting such as that ofthe South Side Working Party, for example, itmight be felt that the decision areas in the listcould be rearranged into broad fields of choiceto do with transport, with land use, with fin-ance and, perhaps, with other familiar func-tional areas such as housing, education, legalor personnel matters. In an industrial manage-ment context, the key functional areas mightwell be different, with categories such as pro-duction, marketing, purchasing, maintenanceand product development forming the mostnatural frame of reference for the participants.The differentiation of levels of choice offers

another obvious means of sorting decision

93


areas, especially where either the nature of theremit or the diversity of the participants leadsto the generation of a list of decision areassome of which are expressed in more gener-alised terms than others. For example, in thelist in Figure 35, the decision area about ‘cut-backs in municipal spending’ could be seen asimpinging not just on South Side but on thepopulation of themunicipality as awhole; whilethat on investment in the continued life of theSouth Side district could be seen as clearlybroader than that to do with the comparativelylocal West Street improvement issue.Figure 36 illustrates one approach to the sort-

ing of decision areas both according to cat-egory and to level of generality at the sametime. Each decision area is here positioned,vertically, in relation to a spectrum rangingfrom the most specific operational choice tothe most generalised choice of policy orienta-tion. It is also located, horizontally, accordingto some set of functional categories which,for this purpose, need not be very preciselydefined. These two dimensions can bemarkedout either on a large sheet of paper on awall, or on a sufficiently large area of flooror table space, depending on the preferencesof the group and the facilities of the room inwhich they are working. The use of floor ortable space, with decision areas representedbymovable cards, sheets of paper or discs, hasthe advantage that their positions can be shift-ed around easily, until the participants agreethat the resulting layout represents a goodenough ‘map’ as a basis for further shapingwork. Of course, the same flexibility can beachieved on awall, provided the technical prob-lems of adhesion and easy repositioning canbe dealt with – for example, by use of semi-adhesive ‘Post-It’ type notes.Themain purposeof this sorting andposition-

ing procedure is to build a broad picture of theoverall balance of the set of decision areas sofar generated; a picture within which it is pos-sible tosearchforbothoverlapsandgaps.Theremay, for example, be very good reasons why,in Figure 36, there should be several comparat-ively specific transport decision areas,whereas

theonly financial decision areas seemtoappearat the more generalised level: but at least itbecomes possible to debate whether there areoverlaps that should be eliminated or import-ant gaps to be filled. Overlaps are, of course,especially likely to arise where some or all ofthe decision areas have been generated byindividuals independently, rather than as partof a group process. Often, overlaps will pointtowards opportunities for more careful refor-mulation of decision areas, using the kind ofguidelines to be discussed in the next section.Where decision areas seem to be poorly

balanced between levels, a useful way ofproceeding is to review whether any of thecomparatively specific decision areas can bethought of in terms of choices of means

towards more general ends, or whether someof the more general ones can be thought ofin terms of choice of ends, suggesting morespecific choices of means. For example, thegeneral CUTBACKS? decision area might, onreflection, suggest more specific choices todo with the relative severity with which anyparticular departmental budget or capital pro-gramme was to be cut back in a comingannual review cycle; while the specific issueof closure of the Griffin Road School might, onreflection, suggest broader areas of decision todo with educational reorganisation policy in awider area.This kind of ends/means investigation is

likely, in practice, to cut across whatever div-isions were agreed as a basis for organisingdecision areas into functional categories. Forinstance, choices about the closure of theGriffin Road School might be seen as lead-ing towards broader policy choices beyond thefield of education – especially if possible non-educational uses were being kept in view forthe buildings or site, with implications for otherpolicies and budgets.Experience tends to indicate that categories

are often simplest to distinguish in the middlerange of the general type/specificity spectrum,while tending to converge both at the broadpolicy level, and in some cases at the level ofspecific local impact. So the vertical divisions in

94

Skills in Shaping

Figure 36 can be thought of as rather like linesof longitude on a globe, which gradually con-verge as they are extended upwards or down-wards towards the polar regions where theymeet.As in the construction of an initial tentative

list, one outcome of this process of sorting bylevels and categories may be a growing recog-nition that some decision areas may be bet-ter expressed either as comparison areas oras uncertainty areas, so may be best set asidefor later consideration in this light. Also, as inthe preliminary listing process, it is importantnot to spend too long in re-sorting decisionareas according to categories and levels; thisis because of the connections across levelsand categories which are of more funda-mental importance to the work of the shapingmode.

REFORMULATING DECISIONAREAS

For a group to establish momentum, it is goodpractice to formulate decision areas looselyand informally in the first instance, withoutworrying about the finer points of how theyshould be worded or the specific nature of thechoices which may be available within each.However, this can mean bypassing issues ofdefinition which can be of much importance inpractical terms; and it can be well worthwhilereturning to consider such issues more care-fully once the participants in the process feelsatisfied that they have built up a good enoughset of decision areas to form the basis of aninitial decision graph. This kind of pause forreformulation of individual decision areas willbe particularly important where there appearto be overlaps between somewhat similardecision areas; overlaps which could be over-come by more careful formulation of what ismeant, so that the decision areas can eitherbe merged or else more clearly distinguishedfrom each other.There are two kinds of difficulty which

often arise in the initial listing of decisionareas, especially among people who have

not previously been accustomed to think-ing explicitly in decision area terms. Oneis that decision areas are confused withoptions within decision areas: for instance,the decision area described as ‘West StreetImprovement’ could be read as the specificproposal to improve rather than the area ofchoice as to whether to improve or not. It isthis kind of scope for misunderstanding thatmakes it useful to maintain the discipline ofwriting a question mark at the end of boththe shorter description and the shorter label ofevery decision area.The other source of confusion encountered

in practice is a tendency to express decisionareas in terms of choice of preferred state,rather than choice of preferred action. Forinstance, depending on the powers at the dis-posal of the decision-makers, it may be onething to consider whether the vacant centralsite in South Side should be in residential orindustrial use at some future time; yet quite adifferent thing to consider which zoning shouldbe indicated on a local plan to be prepared andpublished as a guide to future developmentdecisions. So ‘choice of zoning’ may be a morerealistic phrasing than ‘choice of use’, whenconsidered in relation to the levels of influencewhich the public decision-makers in South Sidecan actually bring to bear. The choice of actionwithin their control may relate, strictly speak-ing, to the action of setting down a policy state-ment in print, accompanied by the action ofshading a map in one tone rather than another;this act may have some influence over the waythe land is used, but perhaps by no meansthe only influence once market pressures andother policy influences come to bear.In some contexts, refinements in the expres-

sion of a decision area may not matter verymuch: but in others they may be crucial.In essence, the attempt to express decisionareas in more careful and realistic termsinvolves looping outwards from the shapingmode towards the designing mode. It meansthinking about options within decision areas, inorder to arrive at a clearer view of the natureof the decision areas themselves. Figure 37

95

FIGURE

37

SOUTH SIDE

EXAMPLE

Refining the Expression of Decision Areas

It is worth giving some thought to the more careful formulation of particular decision areas if they appear

to overlap with other decision areas, or if they appear at any time to be emerging as part of a problem

focus for the design and comparison of alternatives. The operational meaning of decision areas can

become especially crucial when it comes to the design of possible commitment packages.

Skills in Shaping

illustrates this point by offering some morecareful formulations for five of the seventeendecision areas for South Side which were tent-atively listed in Figure 35. These same fivedecision areas will also be found in the set ofseven decision areas that were first presentedto illustrate the general concept of a decisionarea (Figure 10). For that purpose, they weredefined with more care than in the tentativelisting of Figure 35, but with less precisionthan in Figure 37. This illustrates the generalpoint that there is often considerable scope fordebate and judgement in theway any particulardecision area is expressed.The closer formulation of decision areas can

be seen not just as a means of reaching aclearer understanding of problems, but as acheck against wishful thinking – in particular,against the illusion that the participants are in aposition to choose between end states, in situ-ations where in reality they may have only limi-ted opportunities for intervention or influencein relation to the decisions of others. So thiskind of probing can lead to subtle distinctionsbetween decision areas in terms of decision-making agencies and roles. In a situation suchas that of South Side, for instance, therecould be an important distinction between pro-fessional officers, who are expected to offerrecommendations on particular matters, anda politically accountable council or committee,which has the power and the responsibility toaccept, modify or reject these recommenda-tions as a basis for executive action. Such dis-tinctions can become crucial when turning tothework of the choosingmode, andwhen con-sidering the design of commitment packagesin particular. This means that any attempt toformulate decision areas more clearly in actionterms can be seen as a form of looping out-wards in the direction of the choosing mode.Of course, too much reflection on such sub-

tleties may inhibit progress in the constructionof a decision graph. So again it is important toavoid becoming too deeply immersed in thiskind of activity. However, in moving ahead itis important to record any progress which ismade in themore exact formulation of decision

areas – as in theexample of Figure 37 – so that itis available for reference if andwhen required.

INTRODUCING LINKS BETWEENDECISION AREAS

As was emphasised in Chapter 2, the drawingof a decision link between any pair of decisionareas represents no more than a workingassumption that it could make a difference ifthe choices available within those two decisionareas were to be considered together ratherthan independently.As in the generating of the decision areas

themselves, it is useful to treat the introduc-tion of decision links and the resulting build-upof a decision graph as a creative group pro-cess, in which one person begins by suggest-ing one or two obvious linkages, stimulatingother participants to join in and suggest others.Where the number of decision areas is large,it can be advisable to defer this process untilthey have been arranged in terms of categor-ies and levels, as in the example of Figure 36.However, it is more usual in practice to startfrom a set of decision areas which has notbeen arranged in this way, but can be exten-ded and rearranged freely as the group interac-tion gathers momentum. Whichever approachis used, it is important to follow a selectiveapproach to the introduction of decision links,omitting any which seem less significant thanothers. This is because a decision graph canbecome hard to understand if it shows virtu-ally every decision area as joined up to everyother. Indeed, as soon as the pattern becomesas complex as this, it can quickly lose any valueit might have had as a guide to the understand-ing of problem structure within the group.Once the decision graph begins to take

shape, a useful way to proceed is to look ateach decision area in turn and ask which ofthe other decision areas – if any – are likelyto affect it to a significant degree. It is only tobe expected that doubts will arise from timeto time over whether the connection betweenone decision area and another is clear enoughor important enough to be shown as a link on

97

FIGURE

38

SOUTH SIDE

EXAMPLE

Building up a Picture of Decision Links

It is important to begin introducing links between decision areas at a comparatively early stage of work

in the shaping mode, recognizing that many of the decision links will be only tentative at this stage.

The pattern of decision links can if preferred be generated by using non-graphical methods, and the

example from practice presented in Figure 45 gives an example of an alternative matrix format.

Skills in Shaping

a decision graph; and it is only to be expectedthat participants will disagree with each otherover judgements of this kind. Where this hap-pens, progress can be sustained by adding abroken line to the graph, to indicate an uncer-tain or debatable decision link, as shown in theexample of Figure 38. Sometimes, indeed, itmay be worth expressing more than one levelof doubt about the status of decision links,for instance by distinguishing between brokenlines and more tentative dotted lines whichidentify links of even more dubious status.As more and more links are added to a

decision graph, it usually becomes apparentthat it would be helpful to redraw it with someof the decision areasmoved across to differentpositions, so as to avoid or reduce the confu-sion caused by criss-crossing lines and thus tobring out the underlying structure of relation-ships more clearly.If the aim were merely to represent the

inherent structure of relationships within thegraph – its topological structure to use math-ematical language – there would be no needto spend time altering the positions of decisionareas to avoid inelegant features of graphicdesign. However, because the purpose of thedecision graph is primarily to facilitate com-munication within a group, the attempt to‘clean up’ its representation can be well worth-while in practice. A confused and untidy pic-ture with many bent or intersecting lines canobscure aspects of problem structure which,with a little rearrangement, could be broughtout much more clearly; and this in turn canhelp members of a group to identify possibleclusters of decision areas which they mightwish to explore more closely within an agreedproblem focus. So it is worth at this point intro-ducing some simple guidelines for rearrange-ment and clustering of decision areas within adecision graph.

REARRANGING A DECISIONGRAPH

There are various working rules, some of themsimple and obvious enough,which can be used

as an aid to the reshaping of a complex decisiongraph. First, it is usually not hard to pick out anydecision areas which have no more than onedecision link connecting them to the rest of thegraph, as in the case ofWEST ST? in Figure 38.Sometimes, indeed, there may be one or twocompletely isolated decision areas which wereoriginally seen as important when drawing upa preliminary list, but then turned out to haveno links to other decision areas at all. Theremay also be ‘chains’ of decision areas – suchas that connecting GRIFF SCH? and GAS SITE?to SHOP LOC’N? in Figure 38 – which haveonly limited connections with the rest of theproblem structure. So a useful preliminary stepin the restructuring of a graph is to identifysuch comparatively isolated parts and set themaside, as a step towards exploring more care-fully the configuration of the decision areasthat remain. Typically, there will be several ‘tri-ads’ of three fully interlinked decision areasin a decision graph of any size. In Figure 38,for example, there are seven such triads, ifonly solid links are counted, and several moreif broken links are counted as well. It will beless common to find totally connected clustersof four or more decision areas. However, inFigure 38 it is possible, with perseverance, topick out three such clusters by eye, if broken aswell as solid lines are counted. All three, it maybe noted, incorporate the same ROAD LINE?decision area, which emphasises its import-ance to the overall problem structure.Figure 39 presents the result of one attempt

to redraw the graph so as to bring out some ofthese structural featuresmore clearly. Throughthe use of visual judgement and graphical dex-terity alone, some of the decision areas havebeen shifted in position, mainly to reduce thenumber of criss-crossing lines, while prevent-ing the more isolated decision areas from clut-tering up the middle of the picture. Also, toidentify closely linked clusters more clearly,boundary lines have been drawn around twoclusters of four decision areas and one of five.Not all of these clusters include totally inter-connected sets of decision areas, and there isno need to be rigid in this respect. What the

99

FIGURE

39

SOUTH SIDE

EXAMPLE

Rearranging a Decision Graph

Where the structure of a decision graph is difficult to make out visually because there are too many

decision links connecting decision area which are far apart on the graph, it can be important to give time

to the repositioning of at least some decision areas to bring out the underlying structure more clearly. It

is useful to start by identifying more isolated decision areas and closely connected clusters.

Skills in Shaping

clustering in Figure 39 does reveal is that thereare three closely connected clusters which inthis case connect up in a chain: an observationwhich offers an important aid in comprehend-ing the overall structure of this particular graph.It is not easy, nor perhaps too helpful, to

try to codify more precisely the kind of rulesto be used in this kind of structuring process.This is because any attempt to codify createsa need to agree more precisely what statusa broken link should be given as comparedwith a solid link, and this may be to place toomuchmeaning on what is intended as nomorethan a loose and informal convention for mak-ing assumptions about relatedness betweendecision areas. Yet it is always possible todevelop and apply more precise codes of rulesfor the limited range of situations where thiscould be a worthwhile aid to the work of theshaping mode. For instance, decision links canalways be formulated as yes/no or zero/oneentries in a two-way table or matrix of mutualrelationships between pairs of decision areas:and question marks or other symbols can thenbe introduced to represent doubtful links ifdesired. Then that matrix can be rearrangedaccording to some set of logical rules which,in turn, may indicate how the decision graphcould, with advantage, be redrawn.Resort to such methods is most likely to be

worthwhile in circumstances where a graph isso complex that visual inspection offers fewclues to its underlying structure, so that anyattempt to disentangle it without resort to acomputer is likely to involve judgements ofan arbitrary and therefore misleading kind. Inpractice, experienced users of strategic choicemethods usually stop well short of this levelof elaboration in the size of a graph – not leastbecause input of structural information to acomputer can mean slowing down the pace ofinteractive working within a group.

HIGHLIGHTING MORE SIGNIFICANTDECISION AREAS

A decision graph may have to be rearrangedmore than once to bring out its structure as

clearly as possible, and to draw attention toclosely linked clusters without other decisionareas getting in the way. However, this kindof structural information on its own cannotbe expected to bring out all those character-istics of a problem situation which might besignificant in choosing a more restricted prob-lem focus or foci within which to switch tothe work of the designing mode. In particular,the participants will often be aware that somedecision areas are more important to themthan others, in the sense that they representchoices where there is more at stake. Also,there may be some decision areas which maycarry particular urgency even if they are notso significant in themselves – and it couldbe unwise to exclude these from any morerestricted problem focuswhich the participantsmight choose.It is always possible, however, to add

information on importance, urgency or otherconsiderations to a decision graph, by thesimple device of introducing further symbolsto highlight those decision areas which arebelieved to call for particular attention on anysuch grounds. Figure 40 illustrates how thiscan be done by simply adding to the graphvarious kinds of symbol to distinguish decisionareas which the participants see as signifi-cant to them on different grounds. Here circlesare marked in one way to indicate particularurgency of decision, and in another way toindicate particular importance in terms of theirconsequences. An asterisk is used to indicatedecision areas within which there are thoughtto be comparatively clear alternatives, becausethese offer a prospect of straightforward pro-gresswhen turning attention to thework of thedesigning mode. A different convention againis used in Figure 40 to highlight those individualdecision areas which have a particularly highlevel of connectedness to other decision areas.The symbols chosen in this illustration are of akind to allow them to be used in conjunctionwith each other, if necessary. So, for example,the CUTBACKS? decision area is identified asboth urgent and important in its consequences;the ROAD LINE? decision area is identified as

101

FIGURE

40

SOUTH SIDE

EXAMPLE

Highlighting More Significant Decision Areas

Where there are several possible problem foci within which work in other modes might be carried out,

it becomes useful to use colour or other coding to highlight key decision areas on grounds of urgency,

importance of consequences and any other such characteristics additional to those which are to be

found in the structure of the decision graph itself.

Skills in Shaping

both urgent and highly connected; the DISTLIFE? decision area,meanwhile, is identified asimportant in its consequences and also highlyconnected, if not thought to be so urgent atthis stage.The identification of highly connected or

‘nodal’ decision areas is of course one aspectof this process that can be carried out by work-ing from the structural information containedin the decision graph itself. In the graph ofFigure 40, it is not too hard to pick out ROADLINE? and DIST LIFE? as the two most inter-connected decision areas, as each has eightlinks to others, if the more doubtful links areincluded.However, judgements about urgency and

level of importance have to be made on thebasis of information which comes not fromthe graph itself but from people’s percep-tions of the particular pressures and respons-ibilities that they face. It is, therefore, onlyto be expected that there will be differencesof opinion as to relative urgency and import-ance where the participants are working as agroup. There is no reason why different cod-ings should not be used to pick out import-ance to different sets of interests; and whenworking with coloured wedge-tip pens, con-trasting colours can always be used to build aricher picture than can be presented in a blackand white diagram, such as that of Figure 40.For instance, one colour could be used to pickout decision areas of particular importance tolocal people in South Side, and another to pickout those of broader importance to the muni-cipality as a whole, as a means of ensuringthat neither perspective is lost sight of when itcomes to choice of an agreed problem focus.It is very much in the spirit of the strategic

choice approach that differences of perspect-ive should be debated openly within the group,while in the process of adding information onurgency and impact to the decision graph. Inthe course of such a discussion, specific argu-ments may emerge as to why one participantsees a decision area as urgent which anothersees as not so urgent; or again, in the case ofanother decision area, it could emerge that one

participant has foreseen that the choice couldhave serious importance for some sector ofthe community with which other participantswere not so closely concerned. So the high-lighting of more critical decision areas on thegrounds either of urgency or importance canbring valuable benefits in terms of increasedmutual understanding – involving as it doeselements of anticipation of timing and evalu-ative considerations that the participants maywish to explore in more depth when it comesto future work in other modes.

SUPERIMPOSING ORGANISATIONALRESPONSIBILITIES ON A DECISIONGRAPH

A further type of consideration, which can beused to differentiate between decision areaswithin a decision graph, relates to the degreeof influence or control which the participantsexpect to be able to exercise. In South Side,for instance, themembers of theworking partymight see themselves as collectively having adirect influence on the decision as to wherethe shopping centre for South Side should belocated and, possibly, also on the choice ofdesignated use for the central site. However,they might recognise that their influence wasmore limited when it came to recommenda-tions on the line of the road; and their influ-ence might be even more marginal in relationto the high-level corporate budgeting decisionswhich are implied by the CUTBACKS? decisionarea. Therefore, they might wish to discusswhether CUTBACKS? would be better refor-mulated as an uncertainty area instead of adecision area at this stage.One means of reflecting judgements about

relative influence or control is simply to des-ignate another symbol with which to distin-guish individual decision areas on a decisiongraph, along the lines of those used to indi-cate urgency and importance in Figure 40.This approach is indeed often used in practicalapplications of the strategic choice approach.However, in many situations it is useful to

103

FIGURE

41

SOUTH SIDE

EXAMPLE

Superimposing Organisational Boundaries on a Decision Graph

It is often important in practice to distinguish between different areas of organisational or departmental

responsibility in the manner indicated above, using different coloured boundaries where possible. This

can help in considering who should participate in the work of other modes, and in what ways linkages

to the work of other parties might be maintained.

Skills in Shaping

go further than this by distinguishing severaldifferent zones of organisational responsibilityand influence, and to do this by superimposinga new set of boundary lines on the decisiongraph as a whole.For example, Figure 41 once more takes

the overall decision graph arrived at for theSouth Side problem and superimposes threenew sets of boundaries, to do with trans-port decisions; housing decisions; and financialdecisions – each of these relating to a zoneof organisational responsibility which is recog-nised in this particular decision setting. Suchorganisational boundarieswill sometimes over-lap; and both the patterns of overlap and thepatterns of decision links across the boundar-ies can then reveal much about the issues ofinter-departmental or perhaps inter-corporateco-ordination that would arise if any particularset of decision areas were chosen as a pro-spective problem focus.The kind of information used to develop

the graph in this way will often be similar tothat used to sort a preliminary list of decisionareas into functional categories, as illustratedin Figure 36. So the results of any earlierwork in sorting decision areas by categoriescan be drawn on in the process of superim-posing a map of organisational responsibilitiesonto a decision graph. However, it is import-ant to stress that the decision graph is intend-ed primarily as a picture of problem structurerather than organisational responsibilities. Itsvalue in organisational terms is that it can beexpected to highlight issues of co-ordinationwhich could have remained hidden, had theprocess of problem formulation been carriedthrough in a more traditional way.Of course, there is a danger that the introduc-

tion onto the same decision graph of inform-ation about organisational boundaries, on topof all the other information it contains, willpresent the participants with too rich and com-plex a picture for them to comprehend in itsentirety. This iswhy Figure 41 –while highlight-ing particular decision areas as in Figure 40 –omits the cluster boundaries which appearedearlier in Figure 39. However, when a decision

graph is being built up on a large sheet of paperon a wall, the risk of confusion can again bemuch reduced through the use of contrastingcolours to distinguish different sets of bound-aries. For colour distinctions have a muchclearer impact than distinctions among linescontaining various patterns of dots, dashes andcrosses, as used in Figure 41. Furthermore,they take less time to draw, which can be asignificant factor where there is a momentumof group interaction to be sustained.

CONSIDERATIONS IN CHOICEOF A PROBLEM FOCUS

The work done in adding all this furtherinformation to a decision graph can offer abasis either for further rearrangement andreformulation or – once the participants aresatisfied that they have a good enough rep-resentation of the problems before them – formoving on to look more closely at choiceswithin selected parts of the graph. It is only ifthe graph as a whole contains no more thanfour or, at most, five decision areas that agroup – or for that matter an individual – islikely to find it useful to move into the work ofthe designing mode without limiting the prob-lem focus in some way. To attempt to explorecombinations of options within a set of morethan five decision areas is to risk moving intoa depth of analysis which can be very time-consuming and, as experience shows, onlyrarely productive in terms of the progress towhich it leads.Indeed, it is advisable to err on the side of too

narrow a problem focus in the early stages ofanalytical work, by choosing a set of no morethan three linked decision areas – perhaps two,or even only one. Alternatives can then becarefully developed and compared within thatfocus before attempting to broaden the prob-lem focus out again, possibly by introducingonly one additional decision area at a time.In the original example used to illustrate

the concept of problem focus – Figure 13 ofChapter 2 – the choice of focus was a simpleone. For there were only seven decision areas

105

FIGURE

42

SOUTH SIDE

EXAMPLE

Selecting Multiple Problem Foci

It becomes useful to distinguish more than one problem focus within a decision graph in any situation

where it is possible to identify areas for more specialised analysis; where it is desired to define tasks

for sub-groups of a large working group; or where there is a concern to schedule a succession of linked

analytical task over a period of time.

Skills in Shaping

in all, and three of those clearly occupied com-paratively outlying positions. But, when a setof participants is presented with a more com-plex picture, such as that of Figure 41, whereshould the process of choosing one or morefoci for more intensive analysis begin? Thisjudgement must inevitably depend on the con-text, as it is only the participants themselveswho are in a position to judge what weightto give to decision areas they have marked asurgent compared to those they have markedas important or easier to influence; and it isonly they who will be able to appraise theadvantages to be obtained through workingat clusters of issues which cut across depart-mental or corporate boundaries. Furthermore,it is only theywho canweigh these advantagesagainst any possible political difficulties theyforesee, drawing on their own knowledge ofthe history ofworking relationships – good, bador indifferent – between the individuals, pro-fessions or organisational units concerned.In some situations, there is a case for adopt-

ing not just one problem focus but two ormore,on which different groups of people can worksimultaneously. This possibility is illustrated inFigure 42. Here it is supposed that it has beenagreed in South Side to adopt two overlappingproblem foci as a basis for deeper investiga-tion. One of these contains four of the fivelinked decisions within the transport sphere ofresponsibility, so is likely to raise comparativelyfew organisational difficulties. Within this par-ticular problem focus it is quite possible thatthe work of the designing mode can be treatedas a purely technical exercise, though the workof comparing and choosing may still involvechallenges of a wider political nature.The second problem focus – which the

readermay recognise as corresponding exactlyto the set of four decision areas selected asthe original problem focus in Chapter 2 – over-laps with the first focus in that it contains theROAD LINE? decision area. It includes oneurgent decision area and two of high import-ance, along with one other decision area whichis connected to all these three. Three of thefour decision areas are marked as containing

reasonably clear alternatives; but the fourth,DIST LIFE?, is a broader policy-level decisionarea which presents more difficulties in thisrespect. This problem focus, in contrast to theother, does not lie entirelywithin awell-definedfield of organisational responsibility; so it mightwell be agreed that the South Side WorkingParty, or some sub-group of those concerned,would afford the most appropriate setting forthe next stages of the work.It will be noticed in Figure 42 that some

urgent and important decision areas are stillomitted from either focus. So questionswill, ofcourse, arise as to what should be done aboutthem – especially themore urgent ones. It maybe agreed that some urgent decision areassuch as PROPACQ? can best be exploredfurther on their own at this stage, in effectforming a ‘focus of one’ in terms of the prob-lem focus concept. And it may be that otherdecision areas such as HOUS CLEAR? andCUTBACKS? raise issues which can hardly betackled without taking various wider organisa-tional relationships into account. So it might beagreed that any deeper consideration of theseissues should be set aside by the membersof the South Side Working Party at this stage,accepting that they might wish to return tothem in later work.

SUMMARY: PROCESSJUDGEMENTS IN THE SHAPINGMODE

There is not much which is inherently difficultin the concepts of the decision area and thedecision link, which have remained as the basicbuilding blocks of the approach to shaping ofproblems throughout this chapter. Where thechallenge lies is in applying these simple con-cepts to the representation of complex prob-lems, in such a way as to capture as muchas possible of the understanding of theseproblems which is to be found among theparticipants; to bring out differences of viewwherever these are significant; and to helpthem find ways of moving forward towards anagreed collective view.

107

FIGURE

43Process Choices when Working in the Shaping Mode

Skills in Shaping

Because the concepts and tools used whenworking in thismode are simple in themselves,it is possible to take on board more complex-ity in the dimensions of the decision prob-lems being looked at than would be the casewith more sophisticated methods. So it can bequite realistic in practice to build up a decisiongraph containing 20 or even more decisionareas, within which more limited foci for fur-ther analysis can be picked out, using somecombination of the methods illustrated in thischapter. This is not, however, to say that itwill invariably be useful to build up a graph ofthat size or complexity. If it is readily agreedthat a problem can be represented in terms ofonly four or five important areas of choice –or indeed only one or two – there is no pointin striving to make its representation morecomplex.When working in the shaping mode, one of

the few opportunities there is for moving togreater elaboration in analytical methods liesin the use of matrix conventions for represen-ting the structure of a decision graph. Suchmethods were briefly touched on earlier inthe discussion of ways of rearranging a com-plex decision graph and identifying clustersof closely linked decision areas. This kind ofapproach can normally be useful in a situationwhere a decision graph contains such a com-plex pattern of links that it is unclear from visualinspection alone how it could be rearranged tobring out its inherent structure more clearly;and where there are, therefore, dangers thathuman bias will play a significant part in itsinterpretation. Figure 91 in Chapter 10 illus-trates the format in which a decision graph isdisplayed when using the STRAD 2 software,and the accompanying text discusses brieflysome of the ways in which the software canhelp in redrawing the graph and in suggest-ing possible problem foci. The value of thesoftware is that it can enable changes to bemade more freely than when working on flipcharts. However, the role of the computer isonly to make recommendations on possiblefoci; the final judgement must remain one forthe people concerned.

Here, as in other aspects of the strategicchoice technology, the essential point is thatdepth of analysis can be treated as a variable tobe controlled. Yet, experience shows that thecreativity of the process of shaping problems isbest sustained by keeping the level and meth-ods of analysis as simple as possible, espe-cially when working in groups. This is becauseit keeps the process open and transparent, andprovides flexibility for changes in formulationas the work proceeds. For instance, it maybe only after a comparatively complex graphhas been gradually built up on the wall, by col-lective effort, that there is a shared basis fordiscussing what new decision areas might beadded, and whether existing ones should bemerged or expressed in different ways. Theobvious next step will then be to start againon a clean sheet of paper, keeping the earlierversion as a secondary record of the processrather than throwing it away. It is quite com-mon for a group to go through many differentformulations of a decision graph, before theycan agree it offers a good enough basis formoving ahead. But such recycling of effort willonly be creative if there is a shared sense thatthose involved are continually making progressin terms of mutual understanding of the under-lying problem structure.Figure 43 summarises the main points of

this chapter, with reference to the concepts ofswitching and looping as explained in Chapter 4(Figure 32). The types of judgements to bemadewithin the shapingmode include not onlychoices of appropriate depth of analysis whenworkingwithin thatmode, but choices of whenand how to move on into any of the othermodes.Where some sub-set of linked decisionareas has emerged as a useful problem focus,then the obvious next step is to move into thedesigning mode and, whether through AIDAor other methods, to search for feasible com-binations of options among the decision areasconcerned. But it is always possible that atten-tion will be quickly drawn to just one crucialdecision area, where the options are alreadyclear cut; in that case, it may be possible to

109


move directly to an examination of the con-sequences of those optionswithin the compar-ing mode. Then, if there is little doubt whichoption is to be preferred, it may be possibleto move almost straight away into the choos-ing mode, where a commitment to act on thatparticular option can be made. This describesthe kind of fast route through a decision pro-cess which is often followed by an individualworking on hunch or experience; but it is lesscharacteristic of groups formed to work on dif-ficult decision problems, the consequences ofwhich for other interests may be profound.If two or more different problem foci have

been identified within a large decision graph,then the process can, of course, be carriedforward by agreeing to move from the shapingmode into more than one of these othermodes, more or less simultaneously – perhapsby sub-dividing the working group for thispurpose, provided that the case for somecontinuing linkage between their activitiesis kept in view. Apart from indicating thevarious possibilities for switching out of theshaping mode, Figure 43 also shows threeshorter loops which represent brief excursionsin the direction of the other three modes,while continuing to work primarily within theshaping mode. For instance, any attempt toformulate decision areas more precisely, interms of the kinds of alternative actions likelyto be available, involves a loop in the directionof identifying options – and therefore towardsthe task of the designing mode. Any attemptto balance a set of decision areas accordingto some framework of categories or levels,as in Figure 36, means looping in the directionof the comparing mode, because the broaderlevels of choice offer a potential evaluativeframework. The highlighting of especiallyimportant decision areas within a decisiongraph can also be seen as a form of looping

in the direction of the comparing mode.Meanwhile, the highlighting of more urgentdecision areas on the decision graph tendsto involve looping out towards the choosingmode; so does the superimposing of differentorganisational domains where responsibilitiesfor action lie. This brief review of relationshipswith other modes only serves to emphasisethe underlying unity of the process of strategicchoice. It is important to stress the value ofmaking these choices of movement betweenmodes in a flexible, evolutionary way, ratherthan feeling constrained to conform to anymore rigid and prescriptive sequence ofoperations. Some practical guidelines to helpin managing this wider adaptive process willbe offered in Chapter 9.

ILLUSTRATIONS FROM PRACTICE

On the three double pages that follow, thereader will find a set of three contrasting illus-trations from practice which have been chosento supplement the points about skills in shapingwhich have already been made in this chapterwith reference to the South Side problem.The illustrations are taken from applications

in very different contexts – in Brazil, in theNetherlands and in Russia. The first includesphotographs of people in action, to give someimpression of theway inwhich decision graphscan be built up and reshaped through interac-tive group working. The second demonstrateshow simple forms of ends-means mappingcan be used to generate decision areas andto explore relationships between levels ofdecision. The third illustration describes howa focus was selected in a strategy workshopaddressing issues of metropolitan food dis-tribution and retailing, making use of bothPost-It stickers and computer methods inpriority-setting.

110

FIGURE

44Illustrations from Practice – Shaping 1

Theme: Group Interaction in Problem Structuring using Decision Graphs to find a Working Focus.

Problem: Policy Formulation in a Comprehensive Development Plan for an Offshore Island.

Context: Five-Day Workshop with a Local Inter-Disciplinary Planning Team, Recife, Brazil, 1984.

Skills in Shaping

COMMENTARY ON FIGURE 44

These photographs were taken during thesecond day of a five-day workshop held inNovember 1984 in Recife, in the tropicalnortheast of Brazil. The task was to applystrategic choice methods to the explorationof alternative land use, economic, socialand infrastructure policies for the offshoreisland of Itamaraca in the Recife MetropolitanPlanning Region. This island, with an area ofabout 60 square kilometres and a permanentpopulation of around 8000, was connectedto the mainland by a road bridge and hadbecome subject to intense pressures fordevelopment, which had to be reconciled withconcerns about the conservation of the naturalenvironment and of the islanders’ traditionalway of life.The group which participated in the work-

shop included members of an interdisciplinaryplanning team from FIDEM, the metropolitanplanning agency, two of whom were attachedto the island’s local municipal council. The pro-cess was facilitated by two Brazilians whowere already familiar with the strategic choiceapproach, with John Friend as visiting con-sultant. The workshop was jointly sponsoredby FIDEM, the Federal University of Pern-ambuco, the State Transportation Departmentand the British Council. The two decisiongraphs on the wall in the top left photographare shown in detail to the right. The first wasdeveloped by focusing on a single import-ant decision area selected from an initial set

of about 60; this concerned the choice ofroad access to the undeveloped north of theisland. The group then identified eight otherdecision areas which were directly linked tothis, and some cross-links were inserted (2).On the second decision graph (4), this set ofnine decision areas was rearranged to bringout clusters more clearly, and different col-our codings and symbols were used to high-light decision areas of special significance ongrounds of urgency and importance (the cod-ings being listed on the wall on a sheet belowthe graph itself). Of particular interest in thiscase was the use of different colours to pickout not only decision areas of metropolitanimportance, but also those of particular localimportance to the islanders themselves. Thisdistinction was made in response to a con-cern among some members of the group thatthe analysis might ignore issues about whichthe residents themselves were concerned.There was a great deal of lively discussionover the choice of a problem focus for fur-ther analysis, as shown in the second groupphotograph (3) where the second decisiongraph has just been replaced by a redrawn ver-sion showing the pattern of links rather moreclearly.It was later agreed to explore and compare

alternatives within a focus of three decisionareas chosen with both metropolitan import-ance and importance to the islanders in view.A later stage in the comparison of alternativeswithin this focus is illustrated in Figure 69 ofChapter 7.

113

FIGURE


Theme: Use of Ends-Means Mapping to identify Decision Areas which are then analyzed to find a

Problem Focus.

Problem: Disposal of Accumulated Polluted Silts dredged from the Rhine Delta.

Context: Inter-Ministerial Government Working Group, Den Haag, The Neterlands, 1985.

Skills in Shaping


These charts were made during one of themiddle sessions in a series of exploratorywork-shops held in Den Haag, The Netherlands, inthe periodMay–August, 1988. The project wasset up to explore the extent of the problem, andthe feasibility of inter-ministerial working in thiscase. It also formed part of an action-researchproject aimed at developing ways of managinguncertainty in environmental decision-making.The problem centred on recent findings aboutheavy metals in the sediments lying underHolland’s extensive water surfaces. Particularurgency created a focus on the Rhine estuary.Here a deep passage has to be kept clear tothe harbour of Rotterdam. Unfortunately thesludge is replaced by the river currents almostas fast as it is removed. Thus to stop dredgingis no solution – and there is an immediate needfor a decision as towhere to place the removedsludge. Sources of the pollution are known toextend far up the Rhine and its tributaries sothere are international aspects to be taken intoaccount also.The working group was formed from vari-

ous directorates of the Rijkswaterstaat (traffic,waterways and water supply) and the Min-istry of VROM (housing, physical planning andenvironment), with Allen Hickling as ProcessConsultant.The ends/meansmapping (1) provided a way

of helping the group to focus at several levelson what could be done to manage the situ-ation. Thus, starting with ‘clean underwatersoils’ (in the rectangle), the first means to

be identified included prevention, cleaning andisolation of the toxic sludge. Ends (to the rightin the diagram) were very broad, including theelimination of bad effects on flora and fauna.The next level down (to the left in the diagram)brings out possibilities with respect to thetechnology for cleaning, policy over emissions,approvals – and, of course, the internationaldimension. These lead to more specific issuesof law-making, subsidies, publicity, transporta-tion and so on.The list of decision areas (2)was one of those

developed out of the ends-means activity. Thetwo first words of each line demonstrate thediscipline of using ‘choice of’ (‘Keuze van’ inDutch) as the opening phrase of all descrip-tions of decision areas. In this case they aremostly associated with publicity, claims fordamages, various initiatives, and (at the bot-tom) one which is crossed out because it wasthought to be too all-embracing and in need ofworking out more detail.This was done in the decision graph (3) next

to the list of decision areas. The messinessof the chart is typical of work done when thelearning process is going quickly. The analysisto find a working focus was by use of the basicdecision-making criteria of urgency, impact,controllability and connectedness. Two candi-dates for such a focus have been identified – acluster of three decision areas at the top of thechart, and a set of two at the bottom. In fact theupper one was chosen, together with the veryimportant central decision area concerning dif-ferent ways of integrating the organisationsinvolved.

115

FIGURE


1

2

Theme: Listing of Issues, drawing of a decision graph and choice of a problem focus.

Problem: Development of food distribution and retailing systems for a major metropolitan area.

Context: A project in St Petersburg, Russia, conducted by Danish consultants for a European Com-

mission technical assistance programme.

Skills in Shaping


These photographs come from a two-dayworkshop in which John Friend was invited toact as facilitator to five members of a projectteam from the Danish Technological Instituteshortly before the final reporting stage of atwelve-month contract for the European Com-mission TACIS programme (Technical Assist-ance to the Commonwealth of IndependentStates) to help in developing food distributionand retailing policies for the metropolitan areaof St Petersburg in Russia.After an initial listing of 27 issues covering

three flipcharts – only one of which is shownhere (1) – ten of these issueswere classified as

decision areas, and brief descriptions of thesewere transcribed to adhesive ‘Post-It’ notes.Thesewere transferred to another flipchart andrearranged to bring out their interrelationships,using connecting lines to represent decisionlinks and outer rings to represent importanceand urgency (2). One of the team membershad received some introductory training in theuse of the STRAD software. She was there-fore able to enter the resulting decision graphinto the computer so as to generate a recom-mendation for a reduced set of decision areaswhich might form the core of a proposed prob-lem focus. This recommendation was thendebated critically before a problem focus offour decision areas was selected.

117

6 Skills in designing

INTRODUCTION

In Chapter 2, the set of concepts for shapingcomplex problems led into the introductionof a further set of concepts for designingfeasible courses of action within any selec-ted problem focus. These further conceptstogether provided a set of foundations for adesign method – Analysis of InterconnectedDecision Areas, or AIDA – which could beapplied quite generally to any problem thatcould be expressed in terms of multiple areasof choice, however diverse these might be.This chapter will build on those foundations

by discussing the challenges which can arisein applying the conventions of AIDA to untidydecision problems, and the types of judge-ment which are called for in meeting thesechallenges in practice. The chapter will drawon working experience in applying strategicchoice methods to problems of many differentkinds. Although a number of elaborations ofthe basic concepts and techniqueswill be intro-duced, it is not intended to suggest that theseforms of elaboration are always either neces-sary or desirable: it is simply that they are avail-able to the skilled user of the strategic choiceapproach,wherever they can be of help in over-coming practical difficulties as they arise.The principal advantage of the AIDA method

of problem formulation is that it can allowmany different kinds and levels of choice tobe expressed in the same general language.So it offers a means of opening up the pro-cess of problem structuring to participantswhomay see the elements of a decision situationfrom quite different professional or represent-ative perspectives. Its strength is that it has noinherent bias towards any particular frame of

reference, whether financial, locational, engin-eering, marketing, administrative, political – or,indeed, anything else.This high level of generality is, however only

achieved at a cost. For it may be necessary toaccept quite drastic simplifications when striv-ing to represent complex design choices interms of once-and-for-all decision areas, witha finite range of discrete options within each.There are, of course, other design methodswhich offer the possibility of overcoming theseconstraints – but usually thesemeanmoving toa more specialised frame of reference, so as toview the problem primarily – for example – asone of engineering design; or of land-use plan-ning; or of financial budgeting; or of schedulinga set of interconnected activities through time.More specialised approaches to problem

structuring can, of course, have an import-ant place in practice, especially within man-agement contexts where similar kinds ofchoice tend to arise recurrently over time.Choices can then often be expressed in termsof repeated adjustments to continuous con-trol variables, rather than choices betweendiscrete optionswithinmore transient decisionareas. The AIDA approach to problem struc-turing was not developed with these compar-atively well-structured management situationsin view. Nevertheless, situations sometimesarise when it can be used to complement,rather than replace, more refined methods ofproblem structuring within particular special-ist frameworks. Indeed, an initial quick ana-lysis of options and relationships within thesimplified conventions of AIDA can often forma useful prelude to deeper analysis by othermethods. Its particular value can lie in open-ing up the debate about problem structure to

119

FIGURE

47

SOUTH SIDE

EXAMPLE

Testing the Range of Options within a Decision Area

Both when first identifying options within decision areas, and at intervals thereafter, it is useful to question

briefly whether any additional options should be included to make the set of options more representative

of the full range of choice available. This is often done verbally and informally, but a record of arguments

for exclusion or inclusion can provide a useful point of reference for later.

Skills in Designing

a wider set of participants, and affording aclearer perspective of how far more special-ist examination of particular parts of the prob-lem may, or may not, be justified when setagainst the pressures for decision which theparticipants currently face.

IDENTIFYING SETS OF OPTIONSWITHIN DECISION AREAS

In generating a set of options within a decisionarea, the most important consideration is thatthey should be mutually exclusive; that thechoice of any one option should foreclose thechoice of any of the others. In an idealised,analytical world, such a set of options should,where possible, also be exhaustive: in otherwords, it should be a complete list of all thepossible courses of action available. But inpractice, this may be quite an unrealistic con-dition to impose, and it is usually necessary tosettle for a set of options which is agreed tobe representative of the full range of choiceavailable, rather than one which covers everyconceivable possibility.For instance, in South Side, it might be pos-

sible to think of other lines for the new roadapart from the northern and southern routeswhich were identified earlier (Figure 14). Itcould equally be possible to conceive furtherlocations for the proposed shopping centre, orother quite different uses for the central site.Turning towards the broader policy decisionon level of public investment in South Side,this was expressed in terms of choice of timehorizon; and the choice of time horizon is onewhich could, of course, be regarded as a con-tinuous variable, to be extended more or lessindefinitely into the future. So the questionarises here of whether the particular horizonsof 10, 20 and 40 years are to be consideredsufficiently representative of the full range ofalternatives available in practice. It may bethat this is readily agreed to be so – perhapsbecause, in the municipality responsible forSouth Side, these represent familiar periodsof amortisation for different kinds of capitalassets. Yet any such assumption can always

be challenged, debated and then modified ifthe participants agree.Figure 47 illustrates the process of question-

ing which is involved in reviewing whether ornot additional options should be included in aprocess which can be conducted either purelywithin an individual’s head, or through inter-active debate within a group. Whatever themethod of search may be, it can be useful inpractice to keep a record of possible optionswhich have been considered but excluded,with reasons for the exclusion noted as inFigure 47. These options can always be rein-troduced at a later stage if a case for doing socan be made.The examples of possible additional options

which appear in Figure 47 – relating to each ofthe four decision areas which were chosen asan initial problem focus (Figure 13) – illustratevarious types of consideration which can arisein judging whether a proposed set of optionsis both representative and mutually exclus-ive. In the case of ROAD LINE?, for instance,the question arises of whether what is some-times called a null option should be included,to reflect the possibility of building no newroad at all. In this case, the null option is sup-posed to be excluded because of the particularbrief given to the South Side Working Party:but of course this brief could, at any stage,be changed and the list of options modifiedaccordingly. As will be seen later, the possibil-ity of null options can arise again,more directly,when working in the choosing mode, whenpossibilities for deferring choice into the futurebegin to come to the fore.Another question that can sometimes arise

in practice is whether or not an option shouldbe includedwhen it is uncertainwhether or notthe conditions will occur which would make itpossible. In Figure 47, this question arises inthe case of the football ground as a possiblelocation for the new shopping centre; and theassumption is recorded that this option shouldbe excluded for now, possibly to be reintro-duced at a later stage. Turning to the reviewof the DIST LIFE? decision area in Figure 47, itwill be noticed that the 40-year option has been

121

FIGURE

48

SOUTH SIDE

EXAMPLE

Reformulating Decision Areas with Composite Options

Occasionally, it is helpful to examine the possibility of breaking a decision area down into two or more

separate but connected decision areas, especially if the number of possible options is large and some

of them appear to be composite in form. Such a reformulation can help in examining whether the set of

alternatives indicated is fully representative, and in exploring structural assumptions.

Skills in Designing

taken to represent the 60-year horizon, on thegrounds that this possible additional option isbelieved to be broadly similar to the 40-yearoption in its consequences, and also in its pat-tern of option bars. Indeed, this latter test castsdoubt on whether it is even necessary to dis-tinguish between the 20- and 40-year options;for a glance back to Figure 16 will show thatthese two options are identical in their pat-terns of option bars within the chosen problemfocus.

REFORMULATING DECISIONAREAS CONTAINING COMPOSITEOPTIONS

Another situation often encountered in prac-tice is that in which doubts arise as to whetherthe initial list of options can realistically be con-sidered as mutually exclusive – or whether,perhaps, some mixture of one option andanother might be possible. For instance, inthe case of the CENT’L SITE? decision areain South Side, the question could arise – as issuggested in Figure 47 – of whether it is pos-sible to combine different uses on the samesite and, if so, which particular combinations ofuse might be possible.If the possibility of such ‘composite’ options

arises, then it can be important to exploremorecarefully what the full range of possibilitiesmight be. For instance, if it were possible tomix housing and industrial uses on the cent-ral site, could this imply an integrated designin which housing and light industry might beclosely woven together throughout the site?Or would physical or other considerations sug-gest that housing should be sited towards oneend and industry towards the other?Would dif-ferent proportional mixes be possible, and howfar could these different possibilities be reflec-ted in the reformulation of options within thedecision area?The answers to such questions might some-

times depend on technicalities such as soil anddrainage conditions, or on structural factorssuch as existing land uses on other adjacent

sites. Figure 48 illustrates one simple way ofdealing with this kind of complexity, whichinvolves simply dividing up the original decisionarea into two more specific decision areas.In this example, the composite options withinthe CENT’L SITE? decision area in South Sidehave been dealt with by defining two simplerdecision areas covering the western and east-ern parts of the site respectively.Whether this is a realistic and a useful way

of reformulating the choices availablewill oftendepend on an understanding of local realit-ies. In this case, it is assumed that the siteis such that a division into two parts alongsome east/west boundary line makes sense.The new formulation embodies an assump-tion – which can always be challenged andreviewed at any time – that industrial use in thewest is incompatible with housing in the east.Whether the opposite combination is possibleis supposed, in this case, to be more doubtful;and this is indicated by a broken line connect-ing the option of housing in the west to theoption of industry in the east. So this exampleintroduces a new convention, a tentative oruncertain option bar, represented by a brokenline connecting the options in question. So,a broken line is used to express a feeling ofuncertainty over the existence of an optionbar – just as it was used to express uncertaintyover existence of a decision link when workingin the shaping mode.The expansion of a decision area into two

or more linked decision areas, in order toarrive at a clearer formulation of compositeoptions, means re-examining any option barswhich might have linked the original decisionarea to others in the wider problem focus.For example, it could be that those bars inFigure 16 which originally excluded the com-bination of industrial use on the central sitewith the option of the southern road line, andalso with the option of the King Street shop-ping location, apply to industrial uses at oneend of the central site but not necessarily at theother.

123

FIGURE

49

SOUTH SIDE

EXAMPLE

Recording Assumptions Underlying Option Bars

The recording of assumptions behind the inclusion of particular option bars can be useful both as a

record and as a means of challenging or justifying their validity in the course of interactive working.

Cross-references to such assumptions can be included either on an option graph or a compatibility

matrix. If some assumptions are deemed weaker than others, these can be removed on a trial basis.

Skills in Designing

RECORDING ASSUMPTIONSBEHIND THE INCLUSION OFOPTION BARS

There are many possible grounds for mak-ing a judgement that some pair of optionsfrom linked decision areas should be viewedas incompatible. Sometimes, the incompatib-ility will arise from the inherent logic of thesituation. For instance, one option within thedecision area about the use of the gasworkssite (Figure 10) might relate to its developmentas the location of the new district shoppingcentre. This option would then be logicallyincompatible with any option in the SHOPLOCN? decision area which involved locatingthe new centre in Main Street, at King Squareor on any other conceivable site apart from thegasworks site itself.Often, however, option bars are used to

rule out combinations which are not so muchincompatible in strict logical terms, but agreedto be undesirable on other grounds; perhaps,because they may violate some agreed policy,or because they are thought likely to incurconsequences which are agreed to be unac-ceptable. For example, there might be somedesign configurations for South Side whichwould present such severe engineering diffi-culties that they could only be overcome atunthinkable expense – for example, by theconstruction of a tunnel or deep cutting for theproposed new arterial highway.The assumptions behind the inclusion of dif-

ferent option barsmay bemany and varied, andit is important to keep track of these assump-tions so that they can be re-examined at anytime. Figure 49 illustrates one simple way inwhich this kind of documentation of assump-tions can be handled. This involves attachinga symbol to each option bar, whether as anaddition to the information on the option graphitself or as a substitute for the simple crossused earlier (Figure 15) to register an incom-patibility in the option compatibility table.The symbols used in this example are simply

letters of the alphabet, with an accompanyingkey to indicate the nature of the judgement

involved in each case. Thus the two option barsmarked ‘a’ are included because these combin-ationswould both involve the unthinkable com-bination of locating a shopping centre whereit would be cut off by an arterial road from itslocal residential catchment area. The combina-tion marked ‘b’ is ruled out because it is judgednot to be viable in economic terms; and the twocombinationsmarked ‘c’ are excluded becausethey would mean permitting housing or indus-trial development on a restricted site unac-ceptably close to a busy main road. The moredoubtful option bar marked ‘d’ reflects a moredebatable assumption; the assumption thatindustrial use of the central site might erodethe attractiveness of shops at King Square tolocal shoppers, who might then be able to taketheir custom to alternative shopping centresfurther afield.While assumption ‘d’ has been picked out as

more debatable than the rest, it may also bepossible to distinguish in other ways betweenthe levels of conviction with which option barsare to be regarded as imposing unacceptableconstraints. This can be important if there areso many option bars as to seriously constrainthe range of possible schemes that can betaken forward for comparison, in which casethere will be an incentive to relax some of theless stringent of the option bars.As a guide to this procedure, it is sometimes

found useful to adopt a rough-and-ready judge-mental scale to differentiate between optionbars according to the relative ease or diffi-culty of breaking the constraints they imply. Ona three-point scale, for example, the highestgrading – perhaps indicated by three crosses –might indicate very strong grounds for anincompatibility judgement; the second grading,with two crosses, could indicate an incompat-ibility which might be broken if there werestrong enough grounds for doing so; while thesingle cross grading could mark a weaker ormore doubtful incompatibility. Such symbolsfor grading the strength of incompatibility couldbe used instead of more specific reference let-ters either in a compatibility matrix or an optiongraph. The procedure for developing a set of

125

FIGURE

50

SOUTH SIDE

EXAMPLE

Introducing Multiple Option Bars

It is only occasionally in practice that it is worth trying to identify multiple option bars, to rule out particular

combinations of options from three or more different decision areas. In practice, they can often be

avoided by reformulation of the decision areas. The device of expressing a multiple option bar as a

dummy decision area is useful mainly when generating schemes by computer.

Skills in Designing

feasible schemes can then be repeated, eitherwith or without the ‘weaker’ option bars, totest what effect different assumptions wouldhave on the range of choice available. It shouldbe emphasised, however, that the grading ofoption bars offers no more than a crude filter-ing device which can occasionally be helpfulwhenworking in the designingmode. It alwayscarries the risk of suppressing deeper evaluat-ive issues, of a kind which can be subjectedto more explicit scrutiny when working in thecomparing mode.

INTRODUCING MULTIPLE OPTIONBARS

One of the basic simplifying assumptionsof the AIDA design method is the assump-tion that the majority of important designconstraints can be represented as simpleoption bars: as relationships of incompatibil-ity between pairs of options drawn from pairsof interconnected decision areas. Indeed, it isthis kind of simplification which makes AIDAsuch a versatile and transparent means ofexploring complex problems. In most problemsituations encountered in practice, design con-straints can indeed be represented in this way,at least for working purposes. Occasionally,however, important design constraints arisewhich can only realistically be represented interms of a more complex form of incompat-ibility, involving some combination of optionsfrom not just two decision areas but three ormore considered together.Where some combination of three or more

options from different decision areas is incom-patible, this can be formulated through anextension to AIDA conventions known as amultiple option bar. For example, in South Sidethere could be subtle economic argumentswhy it should be considered impractical to con-sider a district life as short as 10 years if theparticular combination of housing developmenton the Central Site and a King Square shop-ping location were chosen – even though a10-year life might be quite conceivable in com-bination with either of these options taken on

its own. So, a triple option bar could be formu-lated to connect the particular triad of optionscombining KING in SHOP LOC’N?, HOUS inCENT’L SITE? and 10YR in DIST LIFE?. Onequite simple means of representing such atriple option bar on the option graph is illus-trated in Figure 50. Here, an unlabelled node onthe graph is shown linked to each of the threeoptions in different decision areas which are tobe excluded in that particular combination.It is usually not too difficult to add a limited

number of triple or even four-fold option barsto a decision graph in the manner shown inFigure 50. But, in general, the introduction ofmultiple option bars tends to make the struc-ture of a problem more difficult to compre-hend, and is to be avoided unless it is clearlyof pivotal importance to the present problemfocus. It is, therefore, comparatively rare –though by no means unknown – for multipleoption bars to be introduced during periods ofinteractive working, when the main concern isto develop a picture of problemstructurewhichcan contribute to mutual understanding withinthe group.However, circumstances do arise in practice

when it is important to explore what effectsthe introduction of multiple option bars mighthave on the range of decision schemes avail-able. If the range of choice has already beendisplayed in the form of a tree, it is not too dif-ficult to filter out those branches in which mul-tiple option bars appear: for instance, it is nothard to see that there is only one of the ninedecision schemes for South Side presented inFigure 17 – Scheme C – which would be elim-inated if the KING-HOUS-10YR multiple optionbar were introduced.Multiple option bars can also be introduced

into computer programs for developing therange of feasible decision schemes availablewithin any specified formulation of a prob-lem. Even if the program is designed to dealwith incompatibilities between pairs of optionsalone, the higher-order option bars can behandled through the device of the dummy

127

FIGURE

51

SOUTH SIDE

EXAMPLE

Coalescing Closely Linked Decision Areas

It is usually only worth coalescing two or more decision areas into a single decision area in situations

where the number of options and option bars can be significantly reduced by this step, and where

the resulting set of composite options can be readily visualised and compared. In many cases, it will

be possible to reformulate these composite options at a broader level of generality for subsequent

reference.

Skills in Designing

decision area.1 The dummy decision area isdefined to include asmany options as the num-ber of decision areas which are linked throughthe multiple option bar – three in the case ofthe multiple option bar introduced in Figure 50.Each of the options affected in these otherdecision areas is then excluded by an optionbar linked to a different one of the options inthe dummy decision area, as shown by theset of three additional compatibility matrices inFigure 50. The dummy decision area has theeffect of excluding this particular combinationof three options, but no other combinations oftwo among the three. The three simple optionbars can then be used as computer input inplace of the triple option bar, and the searchfor feasible routes through the tree can pro-ceed in the normal way. The dummy decisionarea can also, if desired, be shown directlyon the option graph; in the example of Fig-ure 50, this would involve replacing the unla-belled node by a full-sized circle to representa decision area labelled DUMMY?, includingoptions labelled 1, 2 and 3, each with its appro-priate connections to an option within one ofthe other decision areas. However, this kindof representation has little value in practice,as – unless the dummy decision area can be solabelled to draw out some deeper significanceto the substantive decision problem – it is likelyto serve more as a distraction than as a sourceof illumination to the participants.

COALESCING DECISION AREAS

Wherever there is a high density of option barsconnecting the various options within any pairor cluster of decision areas, the number offeasible combinations which remain availablewithin that part of the problem focus is likelyto be quite small. There may then be a case forcoalescing two or more of the decision areasconcerned into a single decision area contain-ing a smaller set of composite options. This is,in effect, the converse of the device illustrated

1 Introduced by Hunter in his early work on the develop-ment of an AIDA computer program.

earlier (Figure 48), where a single decision areacontaining composite options was split up intotwo separate decision areas with option barsconnecting particular options within them. Anexample of the device of coalescing decisionareas appears in Figure 51which is again basedon the three closely linked decision areas forSouth Side considered in Figure 50. Here,there is a high density of option bars betweenthe ROAD LINE? and SHOP LOC’N? decisionareas, which has the effect of excluding threeof the six possible combinations. Therefore,these two decision areas can be coalescedinto one, here labelled ROAD/SHOP?. Thisdecision area includes the three compositeoptions of NORTH/MAIN, NORTH/KING andSOUTH/GAS, each with its corresponding pat-tern of option bars linking it to options in theremaining CENT’L SITE? decision area.In general, it is not likely to be worth coales-

cing decision areas where the result wouldbe to create a decision area containing morethan three or, at most, four composite options;as would be the case, for example, if theROAD/SHOP? decision area in Figure 51 wereto be further coalesced with either the CENT’LSITE? or the DIST LIFE? decision area, creatinga set of seven composite options in the firstcase, or six in the second. Nor are compositeoptions of much value once they cease to bereadily intelligible in design terms. In Figure 51,for example, it is not too hard to look at eachof the three options within the ROAD/SHOP?decision area as a coherent physical option forthe broad-scale design of a future South Sideneighbourhood. However, the ability of parti-cipants in the decision process to conceivealternatives in this way could easily be lost ifany additional dimensions of choice were to beintroduced as well.As soon as a first attempt has been made to

identify options and option bars within a selec-ted problem focus, various opportunities maybe identified both for expanding some decisionareas and for coalescing others. Changes ineither direction will depend on the judgementof participants as to which aspects of their

129

FIGURE

52

SOUTH SIDE

EXAMPLE

Estimating the Number of Feasible Decision Schemes

This kind of procedure for obtaining a rough estimate of the range of feasible schemes is rarely used

explicitly in interactive working. It can however provide a useful quick aid to judgement when deciding

whether or not to extend an option tree by adding further branches — especially when a tree has grown

to a point where it is beginning to become difficult to comprehend the overall range of choice it contains.

Skills in Designing

problems are of greater or of lesser import-ance, as well as on more analytical considera-tions to do with the number of options within adecision area and the densities of option barswhich connect particular pairs. In general, theaim will usually be to arrive at a design formu-lation which the participants feel to be well bal-anced in relation to their current perceptionsof the choices before them. In this respect,they are likely to find that their work within thedesigningmode yields changes in their percep-tions of the decision areas and their links, andso contributes to a sense of progress in thework of the shaping mode.Once a problem has been expressed in

terms of a pattern of options and optionbars connecting some set of decision areas,it becomes possible to work out the num-ber of feasible decision schemes by arrangingthe decision areas in some chosen sequenceand then working systematically through thebranches of the resulting tree of possibilities, inthemanner illustrated in Figure 17. However, itis sometimes useful to make a rough estimateof the number of schemes before embarkingon this systematic process; and there is a ruleof thumb which is sometimes useful in thisrespect.2

The basis of the rule is simply to work out aseries of fractions, one for each pair of decisionareas, representing the proportion of optioncombinations which remain feasible after anyoption bars are taken into account. Then thetotal number of combinations which would beavailable in the option graph, if there were nooption bars, is multiplied by each of these frac-tions in turn. The outcome is an estimate –usually quite a close one – of the number ofschemes that could be available if all the com-binations were to be worked through in full.Figure 52 illustrates how this process works

for the four decision areas in the original

2 This was first suggested by John Luckman and con-sequently became known as ‘Luckman’s Lemma’; alemma in mathematics meaning no more than anunproven rule.

South Side option graph of Figure 16. The pro-cess of estimation is built up in a stepwisefashion, starting with only two decision areasand adding one more at each stage. At thefirst stage, the number ‘3’ simply reflects theinformation that only three of the six combin-ations in the first small table are still avail-able once the three option bars are taken intoaccount.When the third decision area is added,the number of combinations available in thefirst two decision areas – three in all – is firstmultiplied by the number of options in thisnew decision area, which is also three; the res-ult is then multiplied by the two fractions 4/6and 8/9 representing the proportions of feas-ible combinations in the two additional optioncompatibility tables. This gives an estimate forthe number of schemes now available for thethree decision areas taken together.The process of estimation indicates that

approximately 5.3 out of the conceivable 18combinations so far are likely to be feasible;this compares with the exact number of six asrevealed by the tree of Figure 17. Figure 52then takes the estimation procedure a step fur-ther, showing that the estimate remains quiteclose – 9.5 compared to 9 – when the fourthdecision area is added. It is, of course, quitepossible to work directly from the option graphwhen building up the estimates, if the informa-tion is not already set out in the formof compat-ibility tables.Wherever there are no option barsbetween a pair of decision areas, the appropri-ate ‘multiplier’ fraction is 1; a point which cangreatly simplify the computations involved.This kind of estimating procedure can be car-

ried out by hand, as in Figure 52, or it can bebuilt into an interactive computer program; itcan be especially useful as an aid to decidingwhether or not to add an additional decisionarea to an option tree in which the range offeasible decision schemes so far has alreadybeen estimated or enumerated. The estima-tion process can also help in making quickassessments of the consequences of addingor removing option bars; if an estimate showsthat a problem is so constrained as to be likely

131

FIGURE

53

SOUTH SIDE

EXAMPLE

Altering the Sequence of Decision Areas in a Tree

At any moment when the range of possible decision schemes is set out in the form of a tree, it can

be helpful to try changing the sequence of decision areas to explore different structural features. In

particular, this makes it possible to focus on more urgent, important or controllable decision areas by

bringing them further forward, or to develop combinations on either side of a pivotal decision area.

Skills in Designing

to leave open a very small number of feas-ible schemes, then this may suggest that cer-tain of the more doubtful option bars shouldbe removed in order to open up a wider rangeof choice. Where there is a large number ofdecision areas within a problem focus, theprocedure can be used to estimate the likelynumber of schemes to be found either in theformulation as awhole or in particular branchesof the tree. In other words, it is a flexible toolwhich is by no means essential to the strategicchoice ‘toolbox’ – but which can sometimessave time, and alter the course which furtheranalysis may take.

ALTERING THE SEQUENCE OFDECISION AREAS IN A TREE

Figure 17 illustrated one kind of standardformat in which it is possible to display the setof decision schemes that is feasible under anychosen set of assumptions about options andoption bars. This format indicates those pointswhere option bars have been encountered,by means of the device of displaying ‘dead’branches wherever they arise in the devel-opment of the option tree. However, variousalternative forms of presentation are possible,including a format which simply suppressesinformation on dead branches, to make theresulting picture more compact (this simpli-fied form was adopted in Figure 22). Anotherpossibility is simply to list the set of feasibleschemes in straight tabular form, or to list themusing repeat signs wherever the same optionin a decision area recurs on successive lines.This gives a presentation which is logicallysimilar to that of the tree, but avoids any use ofgraphics; so it offers a useful alternative formatin the production of formal documents.A more fundamental choice that arises in

developing a set of feasible decision schemesconcerns the choice of sequence in whichthe decision areas should be considered. Bybringing different decision areas to the fore,different structural features can be displayed.This point is illustrated in Figure 53, which

begins by comparing just two of the twenty-four possible sequences available for arrangingthe four decision areas in the South Side prob-lem focus. The first sequence is that originallyused in Figure 17, with the information aboutdead branches omitted. It draws attention tothe information that the majority of feasibleschemes involve choice of the northern roadline and that, of these, five out of seven involvechoice of King Square as location for the shop-ping centre.The second sequence illustrated in Figure 53

brings the DIST LIFE? choice to the fore – arearrangement which might be decided uponin the light of the broad policy significance ofthis particular choice. To save space, optionsin the final decision area are this time listedhorizontally rather than vertically. This altern-ative sequence demonstrates that the threebranches relating to the 10-, 20- and 40-yearhorizons, all allow equal numbers of possib-ilities in the remaining decision areas. Fur-thermore, the combinations available underthe 20- and 40-year life policies are otherwiseidentical – a point which might indeed havebeen spotted from the structure of the optiongraph, but is demonstrated more clearly in thestructure of the option tree.The third example in Figure 53 illustrates

another kind of format, in which an option treeis developed in two directions, both to the leftand the right of a selected ‘pivotal’ decisionarea. This can be particularly useful in circum-stances where a decision graph can be parti-tioned into two parts which are only connectedthrough a single decision area or decision link.In this example, just one decision area is intro-duced to the left of the pivotal ROAD LINE?decision area – the WEST ST? decision area,which was linked in the decision graph of Fig-ure 12 to the ROAD LINE? decision area aloneand, for that reason, was omitted from theoriginal problem focus (Figure 13). It is sup-posed here that there is just one option barwhich involves the WEST ST? decision area –an option bar which rules out the YES option incombination with the southern road line. Thispiece of information can be introduced quite

133

FIGURE

54

SOUTH SIDE

EXAMPLE

Exploring Consistency Between Levels of Choice

Distinctions between broader and more specific levels of choice offer a valuable means of dealing with

complexity where the number of decision areas is large. When exploring choices over many interrelated

fields of policy, it can occasionally be helpful to introduce a hierarchy of three or even more levels of

generality, using feasible combinations at the broader levels as a framework for evaluation at other

levels.

Skills in Designing

simply when the new decision area is added tothe left of the tree, adjacent to the only otherdecision area to which it is directly related.The wider potential that may sometimes

exist for developing a tree separately on eitherside of a pivotal decision area can be appre-ciated by glancing back at the larger decisiongraph of Figure 42, where the ROAD LINE?decision area forms the sole point of overlapbetween two potential problem foci. If the pat-tern of option bars in the transport cluster ofdecision areas was an intricate one, then therange of possible decision schemes within thetransport focus could be explored by develop-ing another option tree branching out to the leftof the pivotal ROAD LINE? decision area.Mention has already been made of the

potential of computer methods in developingsets of feasible decision schemes. Computermethods can of course make short work ofthe multiple basic ‘logic-crunching’ operationsthat are involved, aswill be discussed further inChapter 10. When there are only three or fourdecision areas in focus, it is not too arduousto go through the procedure without computersupport. However, experience has shown thatit is easy for humans to make errors in this pro-cess, so a computer can always provide a use-ful check. It is of course technically feasible forthe computer to handle quite large numbers ofinterconnected decision areas, but it is anothermatter to present the results in a form that canbe readily absorbed by the participants. Wherethe power of the computer can be especiallyuseful is in the realm of sensitivity analysis,for it can quickly be tested whether differentassumptions about the inclusion of any doubt-ful options or option bars have any signific-ant impact on the range of decision schemesavailable.

EXPLORING CONSISTENCYBETWEEN LEVELS OF CHOICE

There is another variation on the procedure fordeveloping feasible combinations of optionsfrom several decision areas, which has beenfound particularly useful wherever there are

some decision areas expressed at a higherlevel of generality than others. This is some-times the case where a set of people arereviewing choices of broad policy orientationat the same time as more operational choiceswhich might be consistent with some policypositions but not with others. In such cir-cumstances, the options or combinations atthe more operational level can be evaluatednot merely according to their anticipated con-sequences in certain agreed comparison areas,but also according to their consistencywith dif-ferent policy orientations which could be adop-ted at a broader level. An example will helpto illustrate how this analysis of consistencybetween two levels of decision can work inpractice. Figure 54 shows an example of aconsistency matrix (or table) relating to SouthSide, in which four of the seven decision areasfrom the original problem formulation – includ-ing theWESTST? but excluding theDIST LIFE?decision area – have been regarded as formingtogether the more operational level of choice.For purposes of exploring relationships

between levels the DIST LIFE? decision areahas been reformulated here as a broaderpolicy-level choice of investment horizon withonly two options – short and long. To thischoice is added another choice at the broaderpolicy level concerned with emphasis on pub-lic versus private investment – reflecting per-haps a well-recognised ideological differenceon the municipal council. There are then fourcombinations of emphasis at the policy level,which have been expressed as four columnsin the matrix of Figure 54. In this case, it isjudged that there is an element of doubt overwhether a private investment emphasis canbe considered compatible with a longer-terminvestment horizon in view of prevailing eco-nomic circumstances. So there is one possibleoption bar within the broader level of choice,which may possibly eliminate one of the fourcolumns in its entirety.The main cells of the matrix are then com-

pleted by entering a cross wherever an optionwithin a decision area at the more operational

135


level is judged inconsistent with a policy orient-ation represented by a combination of optionsat the broader level. As always, there can bescope for argument as towhether or not partic-ular entries in the table should be included, andany such argument can yield useful productsin terms of learning and mutual adjustmentamong the participants. Any incompatibilitiesbetween options among decision areas at themore operational level can also be displayedwithin this kind of format; and the triangulargrid that appears to the left of the main mat-rix of Figure 54 offers one compact means ofadding information of this kind.The consequences of any set of assump-

tions about incompatibilities can then beworked out and displayed in terms of a seriesof separate option trees covering the moreoperational decision areas, one correspond-ing to each broader choice of orientation.So it is possible to explore what range ofchoice is available for any one of the broaderpolicy orientations, and to present the res-ults in the normal branching form. It is oftenfound that a particular choice of orientation willpre-determine the choice of option in somedecision areas, while leaving open a rangeof alternatives in others. In Figure 54, forexample, a choice of a shorter-term publicinvestment emphasis restricts choice of ROADLINE? to the northern option, and choice ofshopping location toMain Street, while leavingthe WEST ST? choice completely open.Where there are many decision areas it can

be useful to draw out such structural featuresby varying the sequence of decision areasbetween one branch and the next, to bring any‘no choice’ decision areas to the front and toset back any ‘free choice’ decision areas tothe end of the sequence; this will sometimesleave a ‘conditional choice’ zone in between,the structure of which can be examined moreclosely. This kind of representation is some-times referred to as a decision stream; it can beespecially useful in displaying structural char-acteristics of problems which include a largenumber of decision areas, and then seekingreactions from policy-makers before going on

to search for compromise schemeswhichmaysometimes be consistent with more than onepolicy orientation.This kind of analysis of consistency between

levels of choice, in effect takes the processof strategic choice in the direction of thecomparing mode. For it introduces a formof evaluation which stresses the assessmentof alternatives not so much directly in termsof their consequences, as in terms of their per-ceived consistency with different political ori-entations or aims. In effect, the comparisonareas described in Chapter 3 are replaced bythe broader-level decision areas – illustratinghow one kind of strategic choice concept canbe replaced by another in a flexible, adaptiveway. This therefore provides a fitting point atwhich to conclude this chapter and to reviewthe various points made about skills of workingwithin the designing mode.

SUMMARY: PROCESSJUDGEMENTS IN THEDESIGNING MODE

The various points made in this chapter havetaken the process of strategic choice into asomewhat more technical domain than thosemade in Chapter 5 – reflecting a general shiftof emphasis in moving from work in the shap-ing mode to work in the designing mode. Itis worth first re-emphasising the general pointthat work in the designing mode may some-times be very important, but at other timesmuch less significant. In the extreme case,there will be little designing work to be doneat moments where the problem focus hasbeen restricted to a single decision area, andthe options within that decision area are nottoo hard to identify. However, even in thesecircumstances, it may be difficult to choosebetween these options because of uncertain-ties of type UR relating to possible futuredecisions in related fields; so sooner or laterother decision areas may be brought backwithin the problem focus, and the range of pos-sible combinations may begin to broaden outagain.

136

Skills in Designing

When working in the designing mode, asin each of the other modes, the choice ofappropriate level of analysis is always a mat-ter of judgement. In general, there is a casefor inclining towards simplification in the firstinstance, and only then moving in the directionof further elaboration if or when there seemsjustification for doing so. In the work of thedesigning mode, simplicity generally meansworking within a limited focus of no more thanthree, four or five decision areas, usually withbetween two and four options in each. At thislevel, the number of option bars will usually bemanageable and the process of developing aset of feasible decision schemes will not betoo complex.The simpler the structure, the more practic-

able it will be to develop the set of schemesin a way which keeps all the participantsinvolved. But when participants are working inthe designingmode, as opposed to the shapingmode, there ismore likely to be a case for hand-ling at least some of the complexity in otherways. In particular, a computer can be usedin checking combinations in terms of feasibil-ity; in rapidly changing the sequence of a treeto display different structural information; andin testing the consequences of removing oradding option bars, or options, in terms of theoverall range of choice available. But the com-puter can become a distraction in an interact-ive group situation; and its benefits in termsof speed and flexibility may be of little valueif the momentum of group working is therebydisrupted.Figure 55 shows the various directions of

movement out of the designing mode towardsother modes. One frequent direction of move-ment will be towards the comparing mode,to develop a basis of comparison among thevarious decision schemes that are available.But, as was explained in Chapter 3, there aredifferent levels of comparison from which tochoose. If there are only two or three schemesavailable in all, then it becomes possible tocompare them all in considerable depth, takinguncertainty explicitly into account; but if thereare many schemes, it may be more practicable

to compare them first at a much broader level,to form amore limitedworking shortlist as illus-trated in Figure 22.Alternative directions of exit from the design-

ing mode take the process into the choosingand the shaping modes respectively. Exit dir-ectly into the choosing mode will be possiblewhere the participants agree that one courseof action in a particular decision area is clearlysuperior to the others. For example, the parti-cipants in the South Side exercise might be soconvinced by the advantages of the northernroad line, after realising that it presents a farwider choice in other decision areas than thesouthern alternative, that they agree to make arecommendation for the northern line straightaway, without conducting any further analysis.And there are also many circumstances inwhich an exit from the designing to the shap-ing mode will make sense. In particular, a veryrestricted range of alternatives within a prob-lem focus may suggest that this focus mighteasily be extended, whereas a very wide rangemay suggest it should be narrowed down.Also, as indicated in Figure 55, there are vari-

ous ways in which the guidance offered inthis chapter can involve more modest process‘loops’ in the direction of the other modes.A loop towards the shaping mode is impliedboth in the coalescence of decision areas andin the breaking down of decision areas withcomposite options into two or more linkeddecision areas. A loop towards the comparingmode is implied in the step of separating outdifferent levels of choice for the developmentof consistency matrices. It is also implied inthe device of annotating option bars to indic-ate different kinds of arguments for introdu-cing assumptions of incompatibility betweenoptions; arguments which in many cases willbe grounded in value judgements of a debat-able kind. Finally, the resequencing of a set ofdecision areas to bring more urgent decisionareas to the fore implies a form of loopingtowards the choosing mode, where the pres-sures for action begin to dictate the directionin which the process moves.

137

FIGURE

55Process Choices when Working in the Designing Mode

Skills in Designing


Following the precedent set in Chapter 5, a setof three diverse illustrations from practice ispresented on the pages that follow. They havebeen selected to illustrate some of the moreimportant practical points that arise in relationto the work of the designing mode,The first illustration, based on extracts from

documents published in connection with aland-use policy plan for an English rural area,demonstrates one way in which information

about options and theirmutual consistency canbe presented clearly for public discussion. Thesecond illustration, from an application in theNetherlands, shows how members of a groupcan work together in developing a set of feas-ible schemes from the information containedin an option graph. The third illustration is takenfrom a group exercise in organisational choice,and anticipates some of the points to be madein Chapter 9 about systematic documentationof the progressmade through interactivework-ing on flip charts around the walls of a room.

139

FIGURE


Theme: Presentation for Public Discussion of Policy Options, Incompatibilities and Choices within

Orientations.

Problem: Preparation of a Structure Plan as a Policy Framework for a Rural Area.

Context: Planning Team for Rutland with Various Community Groups, County of Leicestershire,

England, 1975–76.

Skills in Designing


These three extracts are taken from docu-ments published by the County Council ofLeicestershire in central England as part of theprocess of preparing a first official StructurePlan for Rutland. Rutland is a predominantlyrural district of some 30000 inhabitants whichat that time (1975/76) had recently lost its inde-pendent County status (since recovered).The table at the top is one of a series of six

similar tables presented for public discussionin November 1975, as part of a consultationdocument entitled ‘Let’s talk about Rutland’.This particular table focuses on employmentpolicy for Rutland, and lists in the first columna set of five possible options for employment,treated as a single decision area. Each row ofthe table contains notes on the incompatibilityof that particular employment option with spe-cific options in other decision areas concernedrespectively with population change, housing,movement, recreation, settlement patternand finance. For presentation purposes, theindividual options in these other decision areaswere not shown in separate sub-divisions ofthe columns as in the standard format of optioncompatibility tables. This kind of simplificationis most desirable in published documents. Itwill be noticed that the table does howeverbriefly indicate the arguments on which partic-ular option bars are based, so that the underly-ing assumptions can be challenged by readers.

Below the table appears an excerpt from asubsequent page of the same document (2),arguing for the exclusion of certain optionson the strength of the foregoing analysis –among them the fifth of the employment policyoptions, excluded on the grounds of its excess-ive financial implications. After a descriptionof the challenge of policy design and evalu-ation that remains (paragraphs 9.11 and 9.12),the report then goes on to list out the mainoptions that remain in the various decisionareas (only those for the first two decisionareas, concerned with population and employ-ment, being included in this extract).The final illustration opposite is taken from

a technical document written to present theanalysis onwhich the submitted Structure Planwas based. The five ‘solution streams’ hadbeen developed to indicate the range of choiceavailable in the six main decision areas (exclud-ing the financial decision area) for each offive broad policy ‘orientations’ or emphaseswhich could be clearly envisaged by theCountyand District Councillors. It was indicated bythe planners that the set of policies finallyselected might be based on a compromisebetween orientations rather than the selec-tion of any one orientation in its pure form.The final decision was to go for a set ofpolicies based on a position part way betweenthe conservation and the socially directedorientations.

141

FIGURE


Theme: Deriving Alternative Schemes from an Option Graph and forming an Option Tree.

Problem: Implementation Policy for the Law against Nuisances, in an Environmental Policy Plan for a

Municipality.

Context: Action-Research with the Project Group, Municipality of Emmen, The Netherlands, 1982.

Skills in Designing


These photographs were taken at a work-ing session of the project group during thesecond month of a 1-year project to producea Municipal Environmental Policy Plan. Thiswork was undertaken by the Municipality ofEmmen in the Province of Drenthe in thenorth of the Netherlands as part of an actionresearch project sponsored by the CentralGovernment.The project group was drawn from vari-

ous parts of the organisation and containeda mixture of administrators, planners, engin-eers and other professionals. The alder-man responsible for the plan was oftenpresent for parts of such sessions. AllenHickling and Arnold de Jong, who wereconducting the research, acted as processconsultants.A high level of unemployment in this part

of Holland, and around Emmen in particu-lar, had led to the environment receivingvery low priority in the recent past. Thishad been especially the case with respect toindustry and commerce. Thus the need foran Environmental Policy Plan was a matterof debate. Environmentalists argued it to beessential if not too late, while others con-sidered it not only unnecessary, but evenharmful.

The focus of the work in this sequence ofphotographs is the implementation of controlsof various kinds – legal, planning and so on(1). It represents a good example of how dif-ferent levels of choice can be combined inan option graph – although this may be onlyof value early in a project. Here there aredecisions about whether to produce detailedlocal plans for the industrial areas; whetherto be involved in zoning and the setting ofnorms; and the thoroughness of the imple-mentation plan to be made for the law aboutnuisances. At the same time there were moregeneral considerations about the attitude tomanagement of controls and the relaxation ofby-laws.The process of deriving the decision

schemes on the wall is important (2). In thisinstance there were three members of thegroup directly involved, and this helped to keeptrack of the logic. The option tree itself (3)is typical in that an unexpected number ofschemes had to be accommodated, leadingto a very asymmetrical layout. This does noharm to the learning process and was correc-ted in the official loose-leaf record of the meet-ing. Here also the reasoning and assumptionsunderlying the option bars were recorded. Inthis case recycling occurred soon after, and thewhole problem was broken out into its con-stituent levels for further analysis.

143

FIGURE


Theme: Documentation using a Photo-Record of Progress in Designing Alternative Decision Schemes.

Problem: Organisational Choice in Developing a Programme of a Study Group of a National Society.

Context: Process of O.R. Study Group, Operational Research Society, Birmingham, England, 1985.

Skills in Designing


The four illustrations opposite show how evenrelatively quickly drawn, exploratory flip chartsproduced in the course of a working ses-sion can be incorporated in a documentaryrecord to help the participants recall after-wards something of the spirit of the pro-cess in which they were involved. The flipcharts are part of a set of 17 generatedduring a one-day workshop held at the Uni-versity of Aston in Birmingham by a studygroup of the Operational Research Societyof Great Britain. This study group, called theProcess of OR Study Group (PORSG), hadbeen formed several years earlier by a setof members of the Society with an interestnot just in the analytical techniques of oper-ational research but in the wider process ofcarrying out operational research projects inorganisations.Nine members of the study group took part

in the workshop, the task of which was to con-sider possible directions for the future of thegroup after a period in which its momentumseemed to have been slipping. Allen Hicklinghad agreed to act as facilitator to the work-shop, of which John Friend was also a mem-ber. Because only a few hours were available,and because not all those present were famil-iar with the strategic choice approach, AllenHickling did not involve the group as a wholein building up the record of progress on thewall; instead he combined the facilitator andrecorder roles, attempting to capture the con-tributions of others on flip charts in a rapid,free-flowing way.

Quite early in the workshop, the initial con-cerns expressed by participants were struc-tured loosely in the form of an ends/meansdiagram rather similar to that shown in Fig-ure 45; from this, an initial set of 17 decisionareas was generated. The two wall charts onthe right hand side reflect attempts to exam-ine some of the broader decision areas moreclosely. It will be noticed that some questionsarose in discussion about whether the labelsof particular options should be changed, andwhether or not options could be treated asmutually exclusive.The more ‘finished’ option graph in the top

left corner includes a set of three key decisionareas concerned with the future of the group;the form of organisation, and the questionof expansion of its membership – a furtherdecision area on frequency of meetings havingbeen first included but then set aside. The ana-lysis of this option graph indicated a choice of14 decision schemes. A debate on preferenceswas centred on four comparison areas, someof which were treated as constraints; thesewere concerned with commitment, relevanceto others, availability of enough willing peopleand effect on OR education.The upshot of the meeting – corresponding

broadly to scheme 10a in the tree – was theformation of a committee to work on a pro-motional strategy with a six-month deadline inview. The committee then went on to devisea programme of joint meetings with regionalgroups of the Society, thus opening up thedebate about the practical process of conduct-ing operational research projects to the widermembership of the Society.

145

7 Skills in comparing

INTRODUCTION

In this chapter the aim will be to explore,more fully than was possible in Chapter 3,the subtle judgements that arise when work-ing in the comparing mode. The relevant coreconcepts were introduced in Figures 19–22;they are those of the comparison area, therelative assessment, the advantage compar-

ison and theworking shortlist. These conceptsare all very general ones, which can be adap-ted to guide the work of the comparing modeat a variety of levels, ranging from the mostrough and ready appraisal of ‘pros’ and ‘cons’of some proposed action, to the most elabor-ate exercises in quantification and predictivemodelling.However, experience so far in applying stra-

tegic choice methods to practical decisionproblems supports the view that it is only veryrarely that themore elaborate forms of compar-ison exercise can be justified. In the first place,it will usually take an exceptionally weightydecision to justify the time and resourcescalled for in collating and analysing largeamounts of data: more fundamentally, how-ever, the strategic choice philosophy drawsattention to the question of whether extens-ive investment in analysis will produce a goodenough return in terms of the level of confid-ence with which decisions can be taken, oncethe full range of uncertainties bearing on thatdecision has been considered.So experience indicates that it usually pays to

concentrate on quite simple, crude approachesto the comparison of alternatives. Yet it isimportant to stress that there is always scopefor choice in the ways alternatives are to be

compared. There are choices to be made inthe formulation of comparison areas; choicesin methods of relative assessment; choices inways of presenting broader advantage com-parisons; and choices in the extent to whichsimplified numerical or other indices are intro-duced to narrow down the full range of altern-atives to a more limited working shortlist. Theguidance offered in this chapter is intended tohelp users in addressing these choices of eval-uative method and style in a more consciousand effective way.It is important to recognise that the work of

comparing alternatives can become very dif-ficult in practice, however simple or elabor-ate the chosen methods of evaluation may be.The difficulties encountered can be politicalas much as technical, especially where thereare many different constituencies of interestwhich might be affected – some perhaps dir-ectly represented in the decision process andothers not. Furthermore, the procedural set-ting itself may be far from simple, and maychange as the problem focus shifts. For thesereasons, the guidance in this chapter will beshaped according to the principle of dynamiccomparison, which has already been touchedupon in the concluding section of Chapter 3.This principle captures the evaluative implica-tions of the wider view of planning as a pro-cess of choosing strategically through time: aprocess in which there may be a successionof ‘evaluative moments’ as the focus of com-parison changes or as new information comesto light, whether through unforeseen circum-stances or through explorations deliberatelyset in train.

147

FIGURE

59

SOUTH SIDE

EXAMPLE

Developing a Set of Relevant Comparison Areas

In building up a set of comparison areas which are relevant to a specific problem focus, it can be

useful to start by examining each of the decision areas in turn, identifying the more important areas of

difference between options as indicated here. When working interactively, this kind of activity is usually

carried out verbally, and can be combined with the use of various brainstorming methods.

Skills in Comparing

IDENTIFYING APPROPRIATECOMPARISON AREAS

In general, a set of comparison areas shouldbe chosen to be as relevant as possible tothe scope of the decision problem currentlyin view, and also to the political domainwithin which decisions are to be made. Withproblems of any complexity, there is alwaysa danger that these considerations takentogether will draw the participants in the dir-ection of quite an elaborate evaluation frame,with a longer list of comparison areas thancan be conveniently managed when work-ing under practical pressures. So, as with thedefinition of decision areas, it can be worthdevoting some time and effort to the formu-lation of an appropriate yet manageable set ofcomparison areas. Yet again, as with decisionareas, people’s views on comparison areasmay change as a group process builds upmomentum and the level of shared under-standing grows; so the formulation of com-parison areas is better viewed as a continuingmatter of concern than as something whichshould take up a great deal of time in the earlystages of an interactive process. Where a cur-rent problem focus is expressed in terms ofa set of linked decision areas and a choice ofoptionswithin each, one practical way in whichto begin generating comparison areas is totake each decision area in turn – perhaps start-ing with those which are believed to be mostimportant – and then to consider what kindsof consequences or effects are thought to bemost relevant to the comparison of optionswithin each. One way of organising such a pro-cess is illustrated in Figure 59.In this example, the first question concerns

the consequences of choosing between thenorthern and the southern road line throughSouth Side. It might be immediately apparentthat these options are likely to differ in termsof capital outlay and also in terms of impacton local residents. Further discussion mightthen suggest that there is another importantconsequence to be considered, relating to theflexibility to pursue further local development

options in the future. So the considerationof this one decision area has indicated a setof three different comparison areas straightaway. These are listed in Figure 59 using briefand tentative definitions. Such definitions canalways be extended later, in the manner illus-trated earlier in Figure 19, and can be modi-fied, if need be, as the work proceeds. Someof these comparison areasmay, of course, alsobe relevant to other decision areas as well;this is indicated in Figure 59 by checking offsubsequent columns in the table where this isthought to be the case, or where there is somedoubt over this point.Turning to the second decision area – that

on location of the shopping centre – an addi-tional set of comparison areas may emerge asimportant. These again are listed in a tentat-ive way, using the options within the decisionarea as one source of guidance in reviewingwhat kinds of comparison areas are likely tobe significant. As further decision areas areadded, the list is extended further – though itbecomes more and more likely that some atleast of the relevant comparison areaswill havealready been generated at an earlier stage. Inthe example of Figure 59, it will be seen thata set of 11 tentative comparison areas hasbeen generated once the four decision areaswithin the chosen problem focus have beenconsidered in turn.But a list of comparison areas generated in

this way is unlikely to provide a well-balancedbase for comparing alternatives, until some fur-ther work has been done on the structure ofthe list. For a start, the list may well be toolong for working purposes; furthermore, it maycontain a certain amount of repetition or over-lap, raising doubts about the level of general-ity or specificity with which comparison areasshould be expressed. For example, in consider-ing the ROAD LINE? decision area, it was atfirst considered good enough to talk simplyof ‘impact on residents’ without distinguishingone group of residents from another. Furtherdown the list, in comparing options for DISTLIFE?, it was however felt to be appropriateto formulate a rather more specific comparison

149

FIGURE

60

SOUTH SIDE

EXAMPLE

Building a Balanced Evaluation Framework

Even where a set of comparison areas has been generated initially with reference to a particular

problem focus, it can be useful to check the list for balance and coverage by cross-reference to broader

evaluation frameworks. Different perspectives can emerge by referring to frameworks expressed in terms

of organisational purposes and of political interests, so use of both types of framework can be worthwhile.

Skills in Comparing

area concerned with the level of confidence ofSouth Side residents in the future of the area.If the West Street decision area were to beadded to the list, then a further question couldarise of whether impacts on this more com-pact group of residents should be consideredseparately from those on the larger South Sidecommunity.In many practical problem situations, ques-

tions will arise about discrimination betweendifferent affected groups, between differentkinds of impact on the same group and,possibly, between different time horizons ofimpact as well. In the South Side situation,it might indeed be very important that theworking party should debate what differentcategories of impact on residents there mightbe: whether, for instance, it could be import-ant to distinguish the severe and immediateconsequences for those few residents facingthe possibility of physical demolition of theirhomes from the less severe but continuingnuisance to others from high levels of trafficvibration and noise.

DEVELOPING A BALANCEDEVALUATION FRAME

Participants in a decision process do not nor-mally work entirely from scratch when build-ing up a list of comparison areas relevant toeach new decision problem that comes up.Usually, each member of a group begins withsome preconceptions about a relevant frameof evaluation; a frame which may either haveevolved as a purely intuitive response to feltpolitical pressures, or may have been articu-lated more systematically as an expression offormal organisational policies or aims.One of the more explicit forms that a frame

of evaluation can take is a written statement ofguiding objectives of the type often developedby those responsible for the management andcontrol of large corporate organisations. Sucha statement may sometimes be structured ina hierarchical form, with expressions of broadaspirations at one level and expressions ofmore specific policy intentions at another. In

other situations the elements in the frame-work may be expressed in less aspirationaland more pragmatic terms; perhaps in termsof broad problem areas to be addressed, suchas poverty or unemployment, rather than interms of idealistic aims to be achieved. Usu-ally, however, a statement expressed in termsof amelioration of problems can be rewritten interms of achievement of goals, or vice versa;the underlying sense of purpose may be quitesimilar in either case.Underlying such statements of purpose,

however, may be more subtle politicalpressures, relating to the structure of account-ability within which the participants are oper-ating. Sometimes, this sense of accountabilitywill differ significantly from one participantto another: one member of the South SideWorking Party, for example, might feel a spe-cial responsibility for the interests of SouthSide residents, and another for the interestsof unemployed school leavers throughout thewider municipality. Such variations are likelyto be influenced by some mix of personalcommitments, representational roles and theprinciples of accountability on which themembership of the working group has beenselected.Figure 60 illustrates how it is possible to use

more than one type of evaluation frame as apoint of reference in the search to developa more structured set of comparison areasfor a process of strategic choice. In the firstlist, based on a more purposive approach, aset of broad headings and more specific sub-headings is used to develop a framework ofpolicy appropriate to the South Side problem;in the second list, a different sets of headingsand sub-headings is used to identify relevantinterests or ‘stakeholder’ groups, reflecting aconcern to be responsive to varied politicalpressures.Either list would probably have to be exten-

ded quite considerably before it could beconsidered to be at all complete in its ownterms. Nevertheless, these two examples ofevaluation frames are both reasonably typicalof their kind. The first is typical of the kind

151


of hierarchically ordered statement of object-ives which is sometimes formally adopted asa set of central guidelines for the work of acorporate organisation;while the second is typ-ical of the range of interests to which a setof publicly accountable decision-makers mightsee themselves as having to respond. If thesetting were one of a commercial rather thana political organisation, the headings might,of course, be different – for example, theinterests to be consideredmight include share-holders, employees, customers, suppliers –but the organising principles would be muchthe same.Either type of list could, of course, be used

directly as a set of comparison areas withinwhich to evaluate alternative courses of action.However, in addressing any specific decisionproblem it will often be more helpful to gen-erate comparison areas directly in the firstinstance, as illustrated in Figure 59. Then thetentative comparison areas generated in thisway can be matched against whatever morestructured lists of policy concerns and interestsmay be available, to help generate insights intoways in which the set of comparison areasto be used might be reformulated in a morelogical and consistent way.Although the two types of framework illus-

trated in Figure 60 are organised on differentprinciples, there will often be more similaritiesbetween them than might at first appear. Forexample, a concern with employment appearsin both lists, expressed in different ways, asdoes a concern with municipal expenditurelevels. The value of this kind of cross-referencebetween contrasting frames of reference isthat it illuminates the scope for choices in theway comparison areas are expressed. Amongthem are choices between different levels ofspecificity or generality, and between levels ofaggregation or disaggregation which might beconsidered appropriate to a particular decisionprocess and a particular ‘evaluative moment’within that process.One particular opportunity for evaluative

choice concerns the possibility of reformu-lating some comparison areas as decision

areas, relating to choices of policy orienta-tion at a quite general level. This opens upthe opportunity for an alternative approach toevaluation which was introduced in Chapter 6(Figure 54), based on the exploration of con-

sistency between different levels of choice.This can be particularly helpful where there isuncertainty or disagreement about choices ofappropriate policy stance; for example, lack ofagreement on whether corporate expenditurecommitments should be held to their presentlevel, allowed to increase to some degree,or perhaps deliberately reduced. Such mat-ters can often be the subject of long-standingpolitical controversy, which cannot always beaddressed too explicitly in a process of stra-tegic choice – butwhich even so can be accom-modated in the analytical work.

APPROACHES TO THECOMPARISON OF ALTERNATIVESIN TERMS OF FLEXIBILITY

Whatever the range of policy concerns orinterests that may be relevant to a decisionproblem, there is one particular aspect of com-parison that is of fundamental importance tothe process of choosing strategically throughtime. This is the comparison of alternativecourses of short-term action in terms of theflexibility of future choice which they allow.One approach to this aspect of comparisonwas illustrated in Figure 25, where five pos-sible action schemes for South Sidewere com-pared in terms of the range of longer-termdecision schemes left open by each. This rangewas first presented in terms of a straight countof schemes; but then a robustness index wasused to indicate how many schemes in eachcase met a specified threshold of acceptab-ility, reflecting the concerns of one of themore important of the various interest groupsaffected in this case.In general, flexibility of future choice tends to

be valued positively by decision-makers. Thisis because the more courses of action are

152

Skills in Comparing

left open for the future, the greater in gen-eral will be the prospects of successful adapta-tion to whatever changes in circumstances thefuture may bring.1 However, it may not alwaysbe realistic to treat flexibility of future choiceas one simple comparison area, because thevalue of such flexibility may differ betweenone area of future choice and another, andalso between one set of interests and another.Indeed, flexibility to one party may sometimesmean restrictions of opportunities to others.In South Side, for example, the opportunitiesfor residents to improve their homes, or toexercise other kinds of choices in their owndomestic lives, might become more restrictedif certain options for development on nearbysites were not foreclosed.Where many alternatives are being com-

pared in broad terms, it may be adequateto treat flexibility or some such term as asingle broadly defined comparison area, as inthe list of Figure 59. At other times, how-ever, it may be important to relate this aspectof comparison more carefully to some evalu-ation framework which identifies a range ofaffected interests or policy concerns, as in thetwo examples in Figure 60. This finer level offlexibility analysis becomes practicable in anysituation where opportunities for future choicehave been structured in terms of decisionareas and options, so that flexibility in somedecision areas can be given greater weightthan flexibility in others. For example, in thesituation of Figure 25, the first of the five short-term action schemes could be regarded withparticular favour as it is the only one to leaveopen the opportunity of developing the centralsite for industry – an aspect of flexibility whichcould be viewed as of particular importance tothe local unemployed.Another possibility – which will be discussed

more fully in Chapter 8 – is that the preserva-tion of options in some decision areasmight be

1 This can be seen a essentially an expression of thecybernetic law of requisite variety, first enunciated byRoss Ashby in his ‘Introduction to Cybernetics’ (Ashby,1956). At its simplest, the law states that ‘it takes vari-ety to control variety’.

seen as of particular significance as a meansof responding to certain specified contingen-cies that might arise in future. For example, theopportunity to zone the central site in SouthSide for industry might be seen as of particu-lar significance if the contingency of closure ofthe local steelworks were to occur.Circumstances sometimes arise in practice

where another related aspect of flexibility hasto be considered; the flexibility not just tochoose between alternative courses of actionin future but to alter courses of current action atsome point in time after an initial commitmenthas been made. Where decision areas andoptions have been rigorously defined in termsof immediate action commitments of an irre-vocable nature, this possibility should, strictlyspeaking, not arise. However, when people areworking on problems under pressure, this kindof rigour is not always appropriate. This is apoint that is of particular significance in relationto choices of broad policy orientation, whichmay be publicly adopted at one moment butcan be modified or reversed later, should cir-cumstances change.One means of approaching this question of

flexibility to modify a policy position at somefuture time is illustrated in Figure 61. Thisexample builds on the comparison of SouthSide policy orientations presented earlier(Figure 54). Figure 61 starts by presenting thecourses of possible action at an operationallevel which are left open by each of thethree possible policy orientations – omittingthe short-term private orientation because noschemes at all at the operational level werefound to be consistent with it.Also, it was indicated in Figure 54 that a

long-term private investment orientation wasbelieved to be of doubtful feasibility. In suchcircumstances, oneworking assumptionmightbe that such an orientation could not be adop-ted in the short term, yet that the choiceof a short-term public investment orientationnow could leave open the possibility of alater switch to a long-term private orientation,should a sufficiently interested and influentialprivate investor appear. The full logic of this

153

FIGURE

61

SOUTH SIDE

EXAMPLE

Expressing Interchangeability in Policy Choice

Where decision areas relate to choices of policy stance or other positions which allow future modification,

it can sometimes be useful to compare short-term alternatives in terms of the possibility of changing to

other positions at some future time. Such comparisons are usually made informally and interactively,

but this example of a more formal comparison illustrates the general principles involved.

Skills in Comparing

situation is represented in Figure 61, wherePOLICY NOW? and POLICY LATER? are for-mulated as separate areas of choice, in placeof the composite decision area which was ori-ginally labelled INVESTMENTORIENTATION?.The choice now of an orientation towardsshort-term public investment can then be com-pared to the choice now of a long-term publicinvestment orientation – or, indeed, any otheralternative – in terms of the additional flexibilityit offers in terms of its changeability to a differ-ent policy position in future. However, it shouldbe noted that, in this example, the flexibilityavailable in the ‘Level 2’ decision areas may belimited if some of them are of high urgency,calling for commitment at or before the timethe initial policy is agreed. This kind of difficultywill be considered further in Chapter 8.

CHOICE OF METHOD OFASSESSMENT WITHIN ACOMPARISON AREA

Just as there are differentways inwhich altern-ative courses of action can be compared interms of their relative flexibility, so more gen-erally there are different ways in which altern-atives can be compared within any other kindof comparison area. This means that there arealways judgements to be made as to howany kind of comparative information should bepresented to help people make comparisonsbetween specific alternatives – not leastwhereinformation from varied sources has to becompressed into a compact and intelligibleform. Among the practical choices to be facedare choices relating to the use of differentforms of language for communication such aswords, figures and graphs, as well as choicesabout theways inwhich feelings of uncertaintyshould be expressed. Some of the differencesof approach that can be considered in boththese directions were considered in Chapter 3(Figures 20 and 21).Underlying such choices of presentation,

however, may be deeper choices about theactual techniques or procedures to be used inmaking assessments of alternatives, and the

sources of information or judgement on whichthose procedures should draw. Such choicesof assessment procedure can sometimes havean important bearing on the levels of effortand time to be devoted to different comparisonareas, and also on the levels of confidencewithwhich assessments can be expressed.Some procedures of assessment are highly

intuitive, while others involve precisely definedsequences of calculation. For example, to forman assessment of the relative capital outlaysinvolved in different development options forthe central site in South Side, there couldbe a choice of either asking a single finan-cial or engineering expert for a quick ‘off thecuff’ guess, based on accumulated personalexperience; or calling for a more painstakingcalculation taking several weeks, in which sev-eral different experts might be asked to playsome part. However simplified or elaborate theprocedure may be, it is important to rememberthat the range of underlying sources of uncer-tainty will be the same – the choice being oneas to whether or not effort should be investedin exploring them in a conscious way.Whenever a particular procedure of assess-

ment is thought to be critical to the decisionscurrently in view, there is always the oppor-tunity to pause and explore its structure inmore depth, in a search for a fuller understand-ing of the pattern of elements and operationson which it is built. One systematic meansof conducting this kind of exploration is illus-trated in Figure 62. For illustrative purposes,this example is restricted to one comparativelytangible aspect of the assessment proceduresin South Side – that of the relative assessmentof capital outlays for two of the developmentoptions for the central site.The breakdown of elements and operations

in Figure 62 begins by taking two of the appar-ently straightforward numerical assessmentsof capital cost which were first presented inChapter 3 (Figure 22): the estimates of 450 kand 200 k entered against the industrial andhousing options for use of the central site.The baseline for each of the assessments in

Figure 62 is assumed to be the ‘null option’

155

FIGURE

62

SOUTH SIDE

EXAMPLE

Uncovering the Elements in an Assessment Procedure

This kind of systematic exploration of the elements contributing to an assessment procedure is not

usually worth carrying out unless and until the focus for comparison has been narrowed down to a

few alternatives which differ critically in terms of assessments in particular comparison areas, making it

important to explore key sources of uncertainty and alternative ways in which they might be managed.

Skills in Comparing

of leaving the site in its present derelictstate – an option which may be purely hypo-thetical in so far as it is not considered a real-istic alternative within the present South Sideproblem formulation but, nevertheless, mightoffer awell-defined starting point for the applic-ation of standard cost assessment techniques.In Figure 62, the assessments of capital out-

lay for the industrial and housing options areboth shown broken down in a stepwise wayinto various contributory elements of assess-ment. Some of these take the form of othermore basic estimates of a quantitative kind –for instance, estimates of the usable site areaand the unit cost of site servicing per hectare –while others take the form of non-numericalstatements of assumptions which are stillworth recording explicitly, especially wherethey remain open to challenge. For example,one engineer might assume that engineeringdifficulties on the central site were ‘not excep-tional’; yet another might cast doubt on thisassumption in the light of a somewhat differentappraisal of drainage problems or geologicalconditions. Again, some participants mightwish to challenge an underlying policy assump-tion that industrial sites should be providedwith a comprehensive range of infrastructureservices before being advertised for rentalor sale, whereas only minimal infrastructureshould be provided if housing developmentwere being considered.The breakdown of assessments into con-

tributory operations and elements is a proced-ure that could, in theory, be pursued almostindefinitely, exposing more and more hiddenassumptions all the time. However, this depthof investigation will not normally be justi-fied unless it is suspected that it will exposenew areas of uncertainty which might havea critical effect on work within the choos-ing mode. Under practical time and resourcepressures, it is more usual to trust the judge-ment of the experts; however, the opportun-ity is always there to ask probing questionsabout the assumptions underlying any expertassessments, and Figure 62 illustrates onegeneral procedure that can be used to probe

systematically the range of assumptions onwhich particular assessments rest.2

Of course, the nature of the assumptionsand the underlying procedure of assessmentmay be quite different in different types ofcomparison area. For instance, any assess-ments of annual incomes for the two alternat-ives in Figure 62 could depend on judgementsabout the year-by-year build up of incomesover some fixed period of future years, withmounting levels of uncertainty as the time hori-zon extends. And, wherever there is no obvi-ous numerical unit of assessment on whichto rely – as in the assessment of the impactany course of action might have on exist-ing residents of South Side – then the break-down of contributory assumptions is likelyto be a less straightforward matter. But thesame principles apply; and the method of step-wise investigation of assumptions illustratedin Figure 62 remains valid even when mostor all of the contributory assumptions have tobe explored by interrogating an expert whoseassessments are based purely on personalexperience and informed judgement.

EXPRESSING FEELINGS OFUNCERTAINTY WHEN ASSESSINGALTERNATIVES

Because any assessment of the effects ofpursuing a course of action involves at leastsome elements of conjecture, anybody whois asked to contribute towards the process ofassessment must expect to encounter feel-ings of uncertainty – whether these feelingsare addressed consciously or at a more uncon-scious level. Referring again to the exampleof an assessment procedure considered inFigure 62, it might be possible for a local plan-ner to feel very confident about quoting a figureof 4 ha for the full area of the central site,

2 Other examples relating to actual planning studies car-ried out by Coventry City Council were presented inLocal Government and Strategic Choice (Friend andJessop, 1969/77, pp. 69–95).

157

FIGURE

63

SOUTH SIDE

EXAMPLE

Eliciting Limits of Surprise

This kind of stepwise questioning process is usually carried out verbally, and can be introduced briefly

and informally at any stage of interactive working where it is proving difficult to arrive at an assessment

either on a numerical or a more judgemental scale. It can be used either to overcome unwillingness to

offer any assessment at all, or to probe the level of uncertainty surrounding a point assessment.

Skills in Comparing

accepting that details of boundary demarca-tion could make a marginal difference. Yet theextent of the proposed buffer zone – assumedto be wider in the case of the housing thanthe industrial option, so as to provide a higherlevel of insulation from traffic noise – might beseen as a rather more debatable matter. Andthe assessment of site servicing costs per hec-tare might be expected to involve higher levelsof uncertainty again – aswas indeed suggestedin Figure 62 by the fuller breakdown of this ele-ment into four different contributory elements,not all of them of a readily quantifiable form.People who are asked to make assessments

under uncertainty – even if they be professionalexperts in their field – do not always behavein a similar or consistent way. One expert, forinstance, might prefer to quote a single estim-ate of 15 k per hectare for industrial servicingcosts, as if it were quite incontrovertible, treat-ing any feelings of uncertainty as a purely per-sonal concern. However, another expert mightprofess such a high state of uncertainty as tobe reluctant to offer any figure at all – perhapsfor fear of being called to account, should anyestimate offered later be falsified by events.To overcome such feelings of reluctance inquoting figures – or conversely, to get a senseof the level of uncertainty which surroundsa single apparently confident estimate – it isoften worth adopting a simple questioning pro-cedure which has become known as the sur-

prise limit method. This method is illustratedwith reference to South Side in Figure 63.The surprise limit method can be applied

to any element within an assessment proced-ure which calls for judgements along somenumerical or equivalent scale. It involves ask-ing a series of questions of a person whoholds information in that field – the knowledge

source, in expert systems language – as towhat levels on that scale would cause themsurprise, starting with extreme levels at eitherend and graduallyworking inwards until a rangeof ‘non-surprising’ possibilities remains.Figure 63 demonstrates two different levels

at which this approach can be applied to thecomparison of capital outlays for Schemes A

and B in the South Side problem situation.In the first set of questions, it is supposedthat an expert – in this case perhaps a civilengineer – starts from a position of reluctanceto give any estimate of the capital cost perhectare of servicing the central site for indus-trial development. So the interrogator beginsby taking what is likely to be an unrealisticallylow level – in this case 5 k per hectare – andasking whether it would cause the expert anysurprise if the servicing cost per hectare wereas low as this. If the expert says ‘yes’ then thequestion is repeated with successively higherlevels until a level which no longer causes sur-prise is reached. The same kind of questioncan also be asked, starting fromwhat is initiallyjudged to be an unrealistically high level andworking downwards.Usually, the questioner works from the two

ends more or less alternately, accepting thatthere will be a tendency to ‘overshoot’ thelimits of surprise from time to time. The pro-cess is, therefore, one of gradual narrowingdown from both ends until a view is arrivedat, with the range of tenable assumptions inbetween. Such a procedure, of working bysuccessive approximation towards a feasiblerange, can usually persuade even the mostreluctant expert to give some expression to hisor her limits of surprise, even when startingfrom a position of refusal to make any estimateat all. In the opposite situation,where an expertrefuses to deviate from a single point on thescale, it becomes possible to test the limits ofsurprise by working outwards from that pointrather than inwards from the extremes. In prac-tice, the procedure is usually conducted purelythrough verbal questions and answers; it ismainly for illustrative purposes that Figure 63sets out the successive steps in written form.The second illustration in Figure 63 shows

how the surprise limit method can be appliedeven when the scale is a non-numerical one.The scale here is the generalised one of degreeof comparative advantagewhich was first illus-trated in Chapter 3 (Figure 21). The process ofasking surprise limit questions in relation to thisnon-numerical scale is essentially the same as

159

FIGURE

64

SOUTH SIDE

EXAMPLE

Combining Advantage Judgements Across Comparison Areas

It is occasionally worth calibrating an advantage comparison scale numerically as in this illustration,

either to provide a firmer basis for arriving at judgements of advantage across different comparison

areas for a particularly important pair of alternatives, or to carry out periodic checks for the types of bias

that can arise in combining judgements of ranges in a more informal and judgemental way.

Skills in Comparing

before; this example demonstrates how relat-ive assessments can be arrived at even withincomparison areas where there is no basis forcomparing alternatives other than in terms ofsome intuitive sense of level of advantage –marginal, significant or whatever – within thespecific setting in which decisions are to bemade.There is, of course, no reason why surprise

limit questions should not be addressed tomore than one expert. Indeed, the possibilitythat there may be several different knowledgesources opens the way to the use of system-atic approaches such as the Delphi method(Dalkey, 1969) for pooling the judgements ofseveral individuals with differing kinds or levelsof expertise to contribute. Such experiments,if used as a background for discussion ratherthan a substitute for it, can expose all kindsof hidden differences in the assumptions ofdifferent individuals. One exercise in which asurprise limit approach was used to exploredifferences in perception between colleaguesin the same team – an administrator, a plan-ner and an engineer – has been fully reportedelsewhere (Friend, Power and Yewlett, 1974,pp. 140–158).

COMBINING ADVANTAGEJUDGEMENTS ACROSS DIFFERENTCOMPARISON AREAS

It can be difficult enough at times to arriveat judgements of the balance of advantagebetween alternatives within any single com-parison area: it can be even more difficult tojudge the overall balance of advantage acrossa set of dissimilar comparison areas. This wasillustrated by the comparison in Chapter 3(Figure 21) of two alternative schemes forSouth Side across the four comparison areas ofCAPITAL:, INCOME:, JOBS: and RESIDENTS:,each represented by a range of points on acommon advantage comparison scale. It is atsuch moments of evaluation across compar-ison areas that major feelings of uncertaintyabout value considerations tend to come mostdirectly to the fore, and have to be taken into

account, alongside whatever other feelings ofuncertainty may have been encountered inmaking assessments within the separate com-parison areas taken one at a time.It is never an easy matter to bring diverse

sources of uncertainty within a common ana-lytical framework. But it is necessary to doso if it is intended to compare alternativeapproaches to the management of uncertaintywithin the choosing mode; and it was as a stepin this direction that a broad-based methodof advantage comparison under uncertaintywas introduced in Chapter 3. The judgementwas presented (Figure 21) that Scheme Bhad the advantage over Scheme A in termsof capital outlay and probably also in termsof impact on residents; but that Scheme Ahad the advantage in terms of income andjobs. But the levels of advantage were subjectto much uncertainty – uncertainty deriving inpart from the process of assessment withineach comparison area and, in part, from thevalue judgements involved in conversion to thecommon advantage comparison scale. Theseuncertainties made it all the more difficult inthat example (Figure 21) to judge where theoverall balance of advantage lay when all fourcomparison areas were taken into account.This kind of balancing process is one which

people continually have to undertake in prac-tice, if only in an intuitive way. However, oncea comparative advantage scale has been usedto record an explicit picture of judgements andassumptions, it becomes possible to supple-ment that intuition – and to carry out checks onits reliability – by resort to rather more system-atic methods. One way of doing this is illus-trated in Figure 64. In Figure 64, a numericalscale of advantage assessment has been intro-duced as ameans of calibrating the distinctionsbetween bands on the original non-numericalscale, and thus as a basis for combining judge-ments across the different comparison areas.The calibration of the scale is a matter of con-venience: the particular numerical scale usedin Figure 64 treats the point of no advantageeither way as zero, and takes the boundarybetween the ‘significant’ and ‘considerable’

161


bands as 10 points in either direction – one dir-ection being treated as negative and the otherpositive so as to keep the arithmetic straight.Such a scale allows the mid-point of

the range for each comparison area to beexpressed as either a negative or a positivenumber. These numbers can then be addedtogether to arrive at a rough estimate of themid-point of the range of belief about overallcomparative advantage, for any set of compar-ison areas taken in combination. In the illus-tration of Figure 64, there is a total of 20‘advantage points’ in favour of B to be coun-ted by adding the mid-points of the CAPITAL:and RESIDENTS: ranges, against 19 points infavour of A obtained by adding the mid-pointsof the INCOME: and JOBS: assessments. Thenet effect is a marginal single point of netadvantage in favour of B when all four compar-isons are combined.But then there is the influence of uncertainty

to be considered, which is considerable in thiscase. A rough rule of thumb, which is goodenough for most practical purposes, is that therange of belief for a combined advantage com-parison will be a little wider than the range ofbelief for the most uncertain of the separateadvantage comparisons within the individualcomparison areas; if there are two comparisonareas with an equally wide span of uncertainty,then a span about half as wide again as eitherof them can be assumed. In the illustration ofFigure 64, it will be noticed that the range forboth RESIDENTS: and JOBS: extends about 10points on either side of the mid-point, whereasthe range for both CAPITAL: and INCOME:assessments is significantly less. So the ruleof thumb suggests that a range of about 15points on either side of the mid-point will beroughly right as an expression of the total rangeof uncertainty when all comparison areas arecombined. As shown in Figure 64, this meansthere could be a significant or even consider-able advantage to either alternative when allsources of uncertainty are taken into account;so there is a case for some serious thoughtabout how the overall range of uncertainty

could be managed, before a preference in thedirection of either alternative is expressed.This kind of procedure for combining advant-

age comparisons can also be carried outin a step-by-step way, introducing one newcomparison area at a time. This possibilityis illustrated in Figure 64 by the introductionof FLEXIBILITY: as an additional comparisonarea; the range of positions on the scale beingbased, in this case, purely on a quick intu-itive judgement about the relative flexibilityof A and B. In the event, Figure 64 showsthat this addition makes only a marginal dif-ference to the mid-point and range arrived atfor the other four comparison areas combined.With practice, it becomes possible to use anon-numerical advantage comparison scale ina quick and informal way, either as a groupactivity or by asking people to work individu-ally on pre-prepared sheets, then comparingand collating the individual judgements whichthey make. It usually takes a little experienceto arrive at consistent interpretations of the dif-ferent levels of significance within a particu-lar working situation; for people have to learnto use the successive bands of a comparativeadvantage scale in a broadly similar way.For example, in a particular context, it might

seem to make sense to rate one alternative ashaving a ‘considerable’ advantage over anotherin each of two comparison areas, but littlesense to regard their combined advantage as‘extreme’. In that case, it might be decided toextend the width of the ‘considerable’ band onthe scale until such inconsistencies tended todisappear. Also, methods based on statisticaltheory can provide a check on any more intu-itive rules of thumb by which ranges of uncer-tainty are combined.3 However, in interactive

3 If a set of variables are independent and can beassumed to follow the normal (Gaussian) probabilitydistribution, then the standard deviation of their sumcan be taken as the square root of the sum of thesquares of the separate standard deviations. So if theranges shown in Figure 64 are assumed to representtwo standard deviations on either side of the mean,the corresponding distance for the first four comparisonareas combined would be the square root of �42+52+

102+102�, which is about equal to 16.

162

Skills in Comparing

group working, such checks are mainly of useon an occasional basis, as a means of buildingconfidence that intuition is not generating res-ults which are at too much variance with logic.

APPLYING CONSTRAINTS INTHE CHOICE OF A WORKINGSHORTLIST

Where there is a large number of decisionschemes to be compared, it will usually bequite unrealistic to subject more than a fewof them to the kind of carefully structured paircomparison process under uncertainty whichhas just been discussed; so the need to choosea more restricted working shortlist becomesacute. One approach to the choice of such aworking shortlist was illustrated in Chapter 3(Figure 22), where all schemes which wereestimated to come above a specified thresholdof capital cost were set aside, as were allthosewhich did not meet a specifiedminimumlevel on the non-numerical residents’ confid-ence index.The application of such minimum or max-

imum constraints on selected indices ofassessment is a useful, if obvious, meansof arriving at a working shortlist where thereis a very large number of possible decisionschemes to be considered. This is especiallyso where the options within the individualdecision areas can be assessed separately onnumerical or equivalent scales, and assess-ments for at least some of the combinationsof options can be made simply by adding theoption assessments together. In practice amonetary index is often used as a constraint,because there are many situations in which itis politically appropriate to apply some upperlimit to the overall cost of a decision scheme.However, in other situations a constraint mightbe placed on the minimum number of jobscreated, or the maximum number of resid-ents to be displaced, or the minimum ratio ofannual return to investment. In general, it will,of course, only be worth introducing a con-straint on any scale if it corresponds to a con-

cern which is important in the particular con-text in which the participants are working.When applying a constraint to a large and

complex set of decision schemes, it is notalways necessary to work through all possibleschemes in full. One means of simplifyingthe procedure is illustrated in Figure 65. Here,assessments of capital cost are indicated forthe set of possible decision schemes for SouthSide, these being presented in the form of atree as in Chapter 3 (Figure 22) – but with someadjustments introduced to illustrate particularpoints. First, the WEST ST? decision area hasbeen added, which increases the total rangeof schemes. Secondly, the options within theSHOP LOC’N? and DIST LIFE? decision areasare assessed jointly rather than separately –because it is supposed now that there are cer-tain costs associated with choosing options inparticular combinations in these two decisionareas – and the sequence of decision areashas been modified so as to bring these twodecision areas together. Thirdly, the capitalcost assessments in each column have beenrecalibrated so that the alternative with thelowest capital cost is taken as zero and theexpected capital cost of every other alternativeis assessed as a positive figure relative to thisminimum amount. Such recalibration can beuseful where it is desired to apply a constraintof expected relative cost – but of course inmany situations cost limits may be conceivedin more absolute terms.Because only positive cost figures can now

appear, it is possible to set a constraint on themaximum cumulative figure which is accept-able at any branching point in the tree. InFigure 65, a capital cost limit of 500 kmonetaryunits has been introduced. This has the effectof terminating the branches emanating fromchoice of the southern road line quite early on,and also closing some of the other branchesbefore the full set of feasible schemes hasbeen developed. In this example, only fourfeasible schemes remain which do not violateeither the option bars or the cumulative capitalcost constraint of 500 k. These four schemeshave been given the labels P, Q, R and S for

163

FIGURE

65

SOUTH SIDE

EXAMPLE

Applying Constraints in Generating Decision Schemes

Wherever options or combinations can be assessed in terms of a simplified scale, it becomes possible

to cut down the range of schemes for closer comparison by introducing, on a trial basis, some constraint

on the maximum or minimum acceptable level. It can save time and effort to close off entire branches

at an early stage, and recalibration to make all relative assessments positive can facilitate this.

Skills in Comparing

working purposes, and together they offer onebasis for a manageable working shortlist. Ifdesired, the working shortlist can be furtherreduced by grouping together schemes whichare rather similar in terms of their options, thenusing one of them to represent the othermem-bers of that group at this stage in the compar-ison process. In the example of Figure 65, thejudgement has been made that Scheme R canrepresent Scheme Q within a more restrictedshortlist of only three schemes, because thetwo schemes differ only in terms of the WestStreet option and R is the less costly of thetwo.Often, it can be difficult to judge, in prac-

tice, at what level a constraint should be set,if the purpose is to develop a working short-list which is neither too large nor too limitedto serve as a base for more careful compar-ison of alternatives. The level of the constraintcan, of course, always be raised or lowered inretrospect, if the level first chosen does nothave the desired effect. Where computer sup-port is available, then such adjustments canbe carried out quickly; indeed, if desired, thelevel of constraint can be varied automatic-ally in accordance with specified rules. Onepossibility is to specify in advance the num-ber of schemes required in the working short-list, leaving the computer to adjust the levelof the constraint accordingly. Another possib-ility is to specify the maximum level of differ-ence to be considered between the least-costscheme and any other scheme to be includedin the list recognising that, as in the exampleof Figure 65, the level of cost for the least-costscheme cannot be known at the outset if thereis a possibility that the combination of all thezero-cost options will be excluded by optionbars.

SHORTLISTING ACROSS MULTIPLECOMPARISON AREAS

The application of constraints in relationto numerical or similar indices can providean effective way of reducing the range ofschemes where this is very large. But this

reduction can carry a cost, because it canmeanignoring for the time being any other compar-ison areas which cannot be treated in this way;and also because potentially vital informationabout uncertainty has to be temporarily setaside. So, it is often wise to view the use ofconstraints as only a crude filtering phase in theformation of a shortlist, leading to an interme-diate list of schemes to which other methodsof shortlisting can subsequently be applied.As was also illustrated earlier (in Figure 22)

the ranking of alternatives according to orderof preference in different comparison areasoffers another useful reference point in select-ing promising alternatives for closer examina-tion. Simple rankings can be deduced quicklywherever decision schemes can be comparedin terms of numerical or equivalent indices,with tied rankingswherever the indices for twoor more alternatives are the same; and it is notdifficult to scan two or more comparison areassimultaneously, once the information has beenreduced to this form. But a set of alternat-ives can always be arranged in a tentativerank order in other comparison areas, whereassessments cannot be quantified – so longas there is some basis for judgement whichallows one scheme to be rated, however hes-itantly, as more desirable than, less desirablethan, or roughly similar to any other.Figure 66 develops this point by compar-

ing the nine schemes, A–I of Figure 22, interms of their rankings on all four of thecomparison areas which were originally intro-duced. The CAPITAL: and RESIDENTS: rank-ings are as earlier indicated in Figure 22; butthe INCOME: and JOBS: rankings have beenadded on the basis of quick intuitive judge-ments about orders of preference within eachof these two additional comparison areas.On the basis of a set of rankings such as

that in Figure 66, it may be possible to pick outone or more schemes which are dominated byparticular others, in the sense that any domin-ated scheme is inferior to, or at least no betterthan, the other in each of the rankings in thedifferent comparison areas. So, in South Side,there are four schemes – C, F, H and I – each of

165

FIGURE

66

SOUTH SIDE

EXAMPLE

Shortlisting Across Multiple Comparison Areas

The ranking of alternatives taking several comparison areas separately offers a simple but useful aid

to comparison where the set is neither too large to be readily handled in this way — in which case

constraints can be introduced — nor so small as to make this step unnecessary. The setting aside of

dominated schemes and the picking out of those ranking highly on several comparison areas can then

aid the shortlisting process.

Skills in Comparing

which is dominated by Scheme A in terms ofthis set of four comparison areas; and amongthem, Scheme F is dominated by Scheme Bas well. So it could be judged appropriate toexclude these four schemes from any short-list selected as a basis for closer evaluation,accepting the risk that the excluded schemesmight score quite highly in other comparisonareas not considered at this stage, and alsothe risk that the rankings might change if majorareas of uncertainty were to be investigatedand new information uncovered.As an alternative means of displaying the

same information, it can often be useful torearrange a set of rankings by reference tothe principle of rank ordering, as shown in thesecond listing of Figure 66. Here the set ofalternatives is arranged in four different ordersof preference, each based on their rank order inone of the four comparison areas; so schemeswhich rate highly in terms of two or more com-parison areas will tend to rise towards the topin the corresponding columns. It can be seen inthis instance that SchemeA –which comes topin two of the columns – rates quite well also inthe other two columns, while Scheme E – themost attractive from the residents’ viewpoint –scores well enough in terms of income andjobs but is the least favourable in terms of cap-ital cost. So, the comparison of A with E couldbe a promising one to explore more closelyfrom the point of view of exposing underly-ing value issues; thus these two alternativesmight well be taken forward to a committeeof elected representatives for debate, perhapsafter further assessment of their relative con-sequences has been carried out.Further inspection of Figure 66 might sug-

gest also that Scheme B be carried forwardfor closer comparison with A and E, in view ofits high ranking in terms of both the residentsand capital assessments – despite its low rank-ing in each of the other two comparison areas.Conceivably, too, Scheme D could be carriedforward as a possible compromise alternative,making a quitemanageableworking shortlist offour schemes in all – A, E, B and D – to be usedas a basis for closer pair comparisons taking

uncertainty more explicitly into account. Andthis in turn could provide a basis from which toswitch into the work of the choosing mode.

FURTHER EXTENSIONS INCOMPARING METHODS

In Chapter 10 there will be found a brief dis-cussion of the use of computer software tosupport the work of the comparing mode, andexamples of three of the main window lay-outs used to support this mode in STRAD 2for Windows will be presented in Figure 92.In the design of this software, the opportun-ity has been taken to make significant exten-sions in the methods of comparison presentedin this chapter, although these remain basedon the four basic concepts of the comparisonarea, the relative assessment, the advantagecomparison and the working shortlist. Theseextensions are possible because, providedassessments of option differences have beenentered for all combinations of decision areasand comparison areaswhere these differencesare judged to be significant, the computer canaggregate this information rapidly. The usercan then introduce any adjustmentswhichmaybe judged necessary to take account of expec-ted effects of combining options from linkeddecision areas in particular ways.The computer can carry out this process

of aggregating differences not only acrossdecision areas but also across comparisonareas, provided some assumptions have beenentered about their relative weightings ortrade-off rates. A particular feature of theSTRAD software is the way it enables intuit-ive assumptions of these weighting factors tobe entered at an early stage of the comparingwork, and subsequently adjusted as and whenrequired.

SUMMARY: PROCESS JUDGEMENTSIN THE COMPARING MODE

This chapter has discussed various choicesof method and emphasis that are continually

167

FIGURE

67Process Choices when Working in the Comparing Mode

Skills in Comparing

encountered in practice when working in thecomparing mode. They are choices whichinvolve repeated judgements about the bal-ance between simplification and elaboration inevaluation method; a balance which, as arguedat the beginning of the chapter, is more use-fully conceived in dynamic than in static terms.To maintain an appropriate balance throughtime can involve alternating between roughand ready comparisons among many altern-atives, and closer comparisons of a selec-ted few; however, this balance is one thathas to be judged not only in the light ofthe application of broad guidelines such asthose offered in this chapter, but also in thelight of political and administrative considera-tions which can vary from context to context.These will include the perceived importanceof different decisions; the range of interestsaffected; and the nature of their represent-ation, direct or indirect, in the processesby which comparisons and choices are tobe made.Underlying many of the evaluative choices

to be made are questions of response touncertainty – feelings of uncertainty being,in practice, inseparable from the necessaryprocesses of conjecture about what the con-sequences of alternative courses of actionmight be. The concepts of relative assessmentand of advantage comparison developed in thischapter have provided away of coming to gripswith these feelings of uncertainty, at least sofar as their influence on comparative judge-ments is concerned. But the closer analysis ofperceived areas of uncertainty, and the consid-eration of what might be done about them, isa matter that will be deferred for discussion inChapter 8, as it is more germane to the workof the choosing than the comparing mode.Figure 67 summarises the points made in

this chapter, in terms of the various typesof looping and switching judgements whichcentre on the comparing mode. Within thecomparing mode itself, the idea of dynamiccomparison, involving alternation betweencrude sifting of many alternatives and closerevaluations of a selected few, offers a key

to the process judgements that have to bemade. Considering first the brief loops that canbe made out of the comparing mode towardsother modes, Figure 67 indicates a loop in thedirection of the choosing mode in situationswhere some rough and ready assessment offlexibility of future choice seems important; aloop in the direction of the shapingmodewhendeveloping a set of comparison areas relevantto a particular problem focus; and a loop in thedirection of the designing modewhenever, forcurrent working purposes, it is felt useful toreduce the range of alternatives by imposingadditional constraints – or, for that matter, toextend the range by removing constraints pre-viously assumed.Turning to the question of less transitory

switching into other modes, the normal pro-gression in a conventional sequential processof decision-making would be from the compar-ing into the choosing mode – not necessarilyto make a definitive choice among the altern-atives that have been compared, but at leastin order to make incremental progress in thatdirection. However, where a process is guidedby the strategic choice philosophy, other direc-tions of progress are possible. A move into theshaping mode will often be appropriate wherethe balance of advantage between alternativesacross different comparison areas seems soproblematic that choices of policy orientationshould be broughtmore explicitly into the prob-lem focus itself. At a more technical level, aswitch back into the designing mode makessense whenever it seems that a reformula-tion of options and option bars could lead to aclearer expression of the set of alternatives tobe compared.


There now follows a set of three illustrationsfrom practice illustrating some further practicalpoints about the work of the comparing mode.The first of these, from a planning exer-

cise for North Holland, illustrates the mergingof contributions from different members of agroup in the building up of an initial set of

169


comparison areas, and the collation of opinionsabout their relative importance. The secondillustration, from northeast Brazil, shows a vari-ant of the advantage comparison method inwhich a list of uncertainty areas is built up asdifficulties are encountered in arriving at groupjudgements. The third illustration presents an

example of the successful use of advant-age comparison methods with the manage-ment committee of a housing co-operative ina disadvantaged former mining community inSouth Yorkshire, England, leading to a broaderappreciation of the positions of some signific-ant external stakeholders.

170

FIGURE

68Illustrations from Practice – Comparing 1

Theme: Developing Key Comparison Areas for Exploring the Differences between Decision Schemes.

Problem: A Provincial Development Plan for the Disposal of Household Waste.

Context: Inter-organisational Planning Team, Province of Noord Holland, The Netherlands, 1986.

Skills in Comparing


These flip charts were drawn up during thefinal day of the first of a series of four work-shops, held to bring together the thinking ofplanners from the province of Noord Holland,the Central Government and the Municipal-ity of Amsterdam on the problem of disposalof solid waste. The workshops were held inHeemstede, the Netherlands during Januaryand February 1986, with Allen Hickling act-ing as process consultant to a team of twofacilitators from the Directorate General for theEnvironment.A number of alternative decision schemes

had already been formulated, and these wereset out in a list. A set of comparison areas(‘criteria’) was then composed in order todifferentiate among these. The first step inthis procedure involved the use of cardson which participants were encouraged towrite their suggestions. All these were stuckup on the wall and then sorted accordingto three categories: physical (‘fysiek’); social(‘soc. maatsch.’); and economic (‘fin. econ.’)(1). Some cards naturally fell between categor-ies, and were therefore positioned so as tooverlap sector boundaries; others were foundto be duplicates (‘dubbelle’).From this, a straightforward list was drawn

up, in no particular order, defining each cri-terion more clearly. Stickers were issued tothe participants, who were asked to placethem next to the comparison areas which they

thoughtmost significant in relation to the rangeof alternatives under consideration. The areaof greatest concern emerged as practical feas-ibility (‘realiseerbaarheid’), followed by envir-onmental damage (‘milieubelasting’), cost ofdisposal to local citizens (‘verwijderingskos-ten’, ‘burger’) and flexibility of waste disposalmethod (‘flexibiliteit techn. verwerking’) (2).Assessments were then made by con-

sensus in the group using very simple scales.A scale of + and − signs was used for theenvironment; the more plus signs the better.The Dutch symbol for Guilders (f) was usedfor cost; the more symbols, the higher thecost. Spots were used for flexibility; the morespots, the more flexible. The note by the sideheaded ‘PM’ ismerely an elaborated definition,describing the system of separating the wasteat source.This work in the comparing mode was not

taken further on that day, because it was feltto be more important to explore further theuncertainties and assumptionswhich had beenthrown up; these were listed elsewhere onthe wall. As is quite usual with work carriedout in the early stages of a strategic choiceworkshop, this particular part of the analysiswas not resumed later. By the time the groupwas ready to move back to work in the com-paring mode, the formulation of the altern-atives and comparison areas had changedagain, reflecting the further understanding ofthe problem that had been gained in themeantime.

173

FIGURE


Theme: Generating a List of Uncertainty Areas during an Advantage Comparison of Shortlisted

Decision Schemes.

Problem: Policy Formulation in a Comprehensive Development Plan for an Offshore Island.

Context: Five-Day Workshop with a Local Inter-Disciplinary Planning Team, Recife, Brazil, 1984.

Skills in Comparing


These two photographs were taken in Recife,Brazil, during the course of the workshopalready described in Figure 44, concerned withpolicies for the offshore island of Itamaraca.At the moment when the first photograph

was taken, attention had become focused ona comparison between two sharply contras-ted decision schemes, numbered 2 and 5,which had been picked out of a set of six feas-ible combinations of options from three linkeddecision areas. These covered respectively thechoice of economic base for the island; thearea chosen for urbanisation; and the accessroute to the undeveloped north. Scheme 2involved retention of a traditional economicbase, centred on farming and sea fishing, witha brake on any urban development beyondexisting commitments; Scheme 5 involved anemphasis on tourist development with addi-tional allocation of land.Initially, five comparison areas were chosen,

concerned with fulfilment of metropolitanaims; with levels of income for the islanders(‘renda’); with preservation of their way of life(‘vida’); with conservation of the natural envir-onment (‘ambiente’); and with demands ona limited water supply (‘agua’). A sixth com-parison area was added later, concerned withinternal transport for the islanders; it is inter-esting (Figure 44) that some of these compar-ison areas had earlier been viewed as decisionareas.In the topphotograph, amember of thegroup

has just filled in the top row of an advant-age comparison chart, using a nine-point scale

labelled (e c s m n m s c e). This is essentiallysimilar to that introduced in Figure 21 – thewords extreme, considerable, significant, mar-ginal and negligible all having close equivalentsin Portuguese. On the first row in the photo-graph, excluded parts of the scale have beenblocked out. So there was thought to be nopossibility that the balance of advantage onthemetropolitan benefit comparison area couldbe in favour of Scheme 2 to any degree, oras much as considerable or extreme in favourof Scheme 5. For the group to arrive at thisjudgement, they had to confront many areas ofuncertainty; and a tentative list of uncertaintyareas isbeingstarteduponanotherchartunder-neath.The threeuncertaintyareasencounteredin debating metropolitan benefit are to do withwhich groups might benefit; with the actionsof another agency; and with assessing how farinvestors would be attracted.The second photograph shows the same

two wall charts at a later moment when allrows of the advantage comparison table hadbeen completed and the list of uncertaintyareas extended accordingly. The various uncer-tainty areas have now been classified (the Por-tuguese equivalents of UE, UV and UR beingIA, IV and IR, respectively). Also appearing inthe second photograph is a second advantagecomparison chart, in which Schemes 2 and 3are compared. In this case, the process wascarried out initially by each individual complet-ing a smaller version of the chart. Onememberthen collated the results and presented themto the group in the form of the mode and rangeof the frequency distribution for each row, asshown here.

175

FIGURE


2

1

3

Theme: Comparing the advantages of a ‘crash programme’ versus a slow but steady approach to

takeover of housing maintenance.

Problem: Strategy for Tenants’ Co-operative in taking over responsibility for maintenance of houses on

their estate having recently become owners.

Context: Support from the Community Operational Research Unit to the Thurnscoe Tenants Housing

Co-operative in South Yorkshire.

Skills in Comparing


This illustration from practice comes from thefirst in a series of workshops carried out in1988 with the management committee ofthe Thurnscoe Tenants Housing Co-operative,serving a former mining community in SouthYorkshire, England. Themembers of this grouphad just taken over the ownership of theirestate of about 300 houses from British Coal,previously the National Coal Board. Havingfought a successful campaign to become own-ers of the estate – with loan finance froma large mortgage company and professionalsupport from a regional housing association –the tenants (1) were now faced with takingover the very different responsibilities of estatemanagers, involving decisions onmatters suchas rent levels, maintenance and sales that dir-ectly impinged on their relationships with theirneighbours and friends.This first one-day workshop was facilitated

by John Friend, acting as a consultant to theCommunity Operational Research Unit thathad recently been established at Northern Col-lege near Barnsley, working with Peter Long ofthe Operational Research department of Shef-field City Polytechnic – now Sheffield HallamUniversity. The focus of the workshop was onthe choice of arrangements for housing main-tenance. It quickly emerged that one of themost critical early choices was whether theco-operative should plan for a gradual transitionto setting up its own organisation for medium-scale maintenance work over a 3-year period,as had been originally envisaged, or whetherit should commit itself to a faster ‘crash’ pro-gramme leading to a takeover of this workwithin about 12 months – as most members ofthe committee preferred.The main picture (2) shows an advant-

age comparison that was made between thescheme for a crash programme and a more‘slow but steady’ alternative – defined in terms

of options for doing the work, for administra-tion and for priority setting. Quick assessmentswere first made in terms of the six criteriaof control; commitment to the work; cost tothe co-operative; local job creation; speed ofresponse; and quality. Stickers were used toindicate the participants’ view on the likelybalance of advantage (spots) and the rangeof uncertainty (arrows) on each row. Therewere agreed to be significant sources of uncer-tainty on several criteria, and thesewere addedto an accompanying list of uncertainty areas(3) on another flip chart. Overall, however,the advantage on each of the six criteria wasjudged to lie with the crash programme.This prompted the facilitator to ask why, if

all the advantages seemed likely to lie on oneside of the balance sheet, a decision shouldnot be made straight away. Might there beany obstacles to making it happen? Mightthere not be other interests that had not yetbeen thought of that would suggest otheradvantages to the slow, but steady, alternat-ive? Prompted by these questions, two addi-tional comparison areas were added, wherethe balance of advantage was thought likelyto lie in the opposite direction: these wereconcerned with the load on committee timeand the attitude of the loan provider. The parti-cipants thenwent on to review the actions thatthey might take to deal with those uncertain-ties that appeared to be most critical, and todevelop an action plan to address these – lead-ing to a decision point on whether to appointa maintenance manager in a few months’time.This episode demonstrates that a facilitator

may sometimes find it necessary to play a‘devil’s advocate’ role in questioning whetherthere are any biases among workshop parti-cipants in the range of comparison areas thatthey initially select as significant, especiallywhen all the advantages seem to lie to onealternative.

177

8 Skills in choosing

INTRODUCTION

The work of the choosing mode calls for quitedifferent kinds of judgements to the work ofthe comparing mode. In essence, they arejudgements to do with the management ofuncertainty and the development of commit-ment through time; for this is the mode inwhich the time dimension comes to the fore,and the pressures for urgent action have tobe balanced against any concerns that thedecisions faced may be too difficult to addressbefore further explorations have been carriedout.The pressure for a switch towards the choos-

ing mode can frequently be observed in thecourse of a conventional committee decisionprocess. Typically, some participants start tobecome impatient, look at their watches andsay ‘isn’t it time we reached a decision andmoved on to the next business?’, or ‘isn’t ittime we brought the meeting to an end?’.In practice, however, the issues involved inmaking progress through time become mostcomplex and challenging where the problemunder consideration is itself complex in struc-ture, embracing many interrelated elementsto which different urgencies and uncertaintiesapply.Even though the challenges of making pro-

gress through time are readily recognisablefrom personal experience, they have receivedmuch less attention in the literature of plan-ning and management methods than the moretechnical challenges of evaluation. This maybe because judgements to do with choosingthrough time are much less easy to separatefrom the particular organisational or political

context in which they arise than are judge-ments to do with comparison of specific alter-natives, which scholars have generally beenable to treat in amore detached, analytical way.So in practice, the shift from the comparing

to the choosing mode means a shift from thetechnical domain towards the political arena.But one political arena can differ sharply fromanother, in terms of the configurations ofconflict, competition, consensus and coalitionwhich influence the ways in which decisionsare made; and such differences can reflect notonly variations in the ways in which the par-ticipants themselves have shaped their owninternal working relationships through time,but also deeper, underlying differences in theiraccountability to others with a stake in thedecisions to be made. Even where progressis being sought towards action commitmentsby a set of participants who come togetherrepeatedly in the same group setting – as hasbeen supposed to be the case with the SouthSide Working Party – the members may dif-fer from each other in terms of their externalaccountability. Indeed, in practice, the wish toinclude representation of all the most relevantdepartments, agencies or interest groups isoften one of the explicit principles on whichinvolvement in a decision process is designed.So, in addressing complex issues, the work-

ing group in which all members share exactlythe same accountability tends to be more theexception than the rule. Not only problems butalso personal responsibilitiesmay be perceiveddifferently by different members of a group;and this in turnmeans that any guidance on thework of the choosing mode cannot always beaddressed to all members of a working groupas if they formed a single coherent team. There

179

FIGURE

71

SOUTH SIDE

EXAMPLE

Building up a Working List of Uncertainty Areas

It is useful to build up a rough-and-ready listing of uncertainty areas progressively in interactive working,

as new areas of uncertainty can surface when working in any mode. The use of the UE/UV/UR typology

helps in drawing attention to any under-represented types as well as borderline cases. Notes on

origination are here included for illustrative purposes and would be omitted in practice.

Skills in Choosing

is no shortage of useful advice that can beoffered in this chapter, at a generalised level,building on the four core concepts of the uncer-tainty area, the exploratory option, the action

scheme and the commitment package.1 Butreaders should not be surprised when ques-tions of whose uncertainty, whose exploration,whose action and whose commitment keepbreaking through. Yet even in contextswhere itis unrealistic to expect that views about thesematterswill be freely shared, there remains thepossibility that individuals will be able to makegood use of the concepts andmethods offeredhere as a guide to their own personal contribu-tions to a decision process.

BUILDING A WORKING LIST OFUNCERTAINTY AREAS

The concept of the uncertainty area can beused to represent any area where alternat-ive assumptions can be made about matterswhich are of some importance to decision-making. Such assumptions may relate to vari-ous aspects of theworking environmentwithinwhich people are trying to make decisions;they may relate to aspects of the policy val-

ues to which they are expected to pay heed;or they may relate to other related decisions

where commitments have not, as yet, beenmade. The labels UE, UV and UR, which wereintroduced in Chapter 1 to differentiate thesethree types of uncertainty, serve not only asa means of classifying particular uncertaintyareas as they are identified, but also as a usefulreminder of the broad scope of the uncertaintyarea concept. Such a reminder can be espe-cially appropriate to participants of an analyticalbent, who are often pre-disposed to view themanagement of uncertainty primarily in termsof prediction or survey exercises using estab-lished statistical techniques.In Chapter 3, the concept of the uncer-

tainty area was first illustrated at a point in

1 As will be explained later, this is now often referred toby the alternative phrase progress package.

the South Side story where the focus of com-parison had been narrowed down to a par-ticular pair of alternative schemes, A and B,for which the overall balance of advantageacross comparison areas was far from clearcut. However, it is only to be expected thatpeople will experience personal feelings ofuncertainty at many moments in a decision-making process, and that these feelings willchange continually as they are exposed towhatother participants have to say. People may feeluncertain as to where the boundaries of theirproblem should be drawn; theymay feel uncer-tain as to whether or not particular decisionareas should be seen as interconnected; theymay feel uncertain as to whether particularoptions or combinations of options should beconsidered feasible; and they may feel uncer-tain as to the terms in which particular compar-ison areas should be formulated. Then, whenit comes to comparing specific alternatives,they are bound to experience feelings of uncer-tainty both in attempting to assess their con-sequences within particular comparison areasand in judging where the overall balance ofadvantage lies; and it is at such moments thatit becomes most important to find ways ofviewing all areas of uncertainty together assources of difficulty in choosing – some ofwhich may be more significant than others –using the methods which will be discussed inthis chapter.Because feelings of uncertainty may surface

at virtually any moment in a process of stra-tegic choice, it is often useful in group work toset aside a sheet of paper on the wall wherea cumulative list of uncertainty areas can bedeveloped, recognising that such a list canalways be restructured, and the items within itreformulated, at some later time. An exampleof a relatively unstructured list built up gradu-ally in this way is presented in Figure 71. Thisillustration follows broadly the development ofthe South Side story, as recounted in earlierchapters, so the listing of uncertainty areas fol-lows the broad sequence of shaping, designingand comparing in a more or less linear way –after starting by registering one uncertainty

181


area, to do with air pollution, that is in thiscase supposed to have been so clearly recog-nised among the participants that it could beplaced on record even before the group activityof shaping problems began. In practice, how-ever, the more work is carried out within agroup, the more the group is likely to switchfreely between one mode and another; so thelonger a list of uncertainty areas becomes,the more mixed it will tend to become inthe modes of activity from which successiveuncertainty areas are drawn.In a group process, the recording of each

uncertainty area on the list implies a work-ing assumption that the feeling of uncertaintyit contains is shared among different mem-bers of the group. Sometimes, of course, onemember may express a feeling of uncertaintywhich is promptly dissolved because of somepiece of information which another memberof the group can supply, from a position ofgreater knowledge or authority in some partic-ular field. Sometimes, too, there may be areasof uncertainty which remain largely latent untilexposed within the group. For example, onemember of the South Side Working Party mayfeel quite confident in assuming that the foot-ball groundwill be available for redevelopment,only to discover in discussion that anotherparticipant is equally confident that it will not;and they may then discover that neither isin a position to refute the other’s assump-tion on the grounds of superior knowledge.Whereas neither participant had feelings ofuncertainty on this point before the group inter-action started, both share a feeling of uncer-tainty after some communication has takenplace; so a new uncertainty area is recorded,and the decision process has become themorerealistic as a result.Because a working list of uncertainty areas

may be subject to considerable restructuringbefore it is put to operational use, it is usu-ally not worth while assigning brief labels tothe uncertainty areas at the time they are firstrecorded, in the manner that was illustratedin Figure 23. It can, however, be helpful toregister from the outset whether each new

uncertainty area seems to be of type UE, UVor UR, even if there remains some doubt overwhich classification is most appropriate. Forthe attempt to classify uncertainty areas in thisway can provide a valuable check against anytendency to bias towards or against any oneof the three basic types – a bias towards con-cern with uncertainties of type UE being acommon experience among participants of amore analytical, apolitical cast of mind. Oftenthere is a tendency in an initial list to includecomparatively few uncertainty areas of typeUR – most of them referring to decisions overwhich the participants feel they have relat-ively little control. This is natural enough at astage of the process before a clear problemfocus has emerged, because the tendency willhave been to include most of the significantareas of choice within the decision graph itself.However, as soon as particular decision areasbegin to be excluded from the problem focus,they become potential candidates for the listof uncertainty areas of type UR.Finally, it is often helpful, as in Figure 71,

to include brief notes against all the variousuncertainty areas on the list, referring to themode or moment in the process where eachof them first surfaced. The value of this can beappreciated by noticing that some of the mar-ginal notes recorded against particular uncer-tainty areas on Figure 71 – though not all ofthem – can help in making reference back toearlier chapters which covered other stages inthe development of the South Side story.

PUTTING UNCERTAINTY AREASINTO A CLEARER DECISIONPERSPECTIVE

Each uncertainty area on a working list, suchas that of Figure 71, will have been recordedbecause at some moment in the process itappeared to have at least some relevance tothe decision problems being addressed. How-ever, the longer a list grows, the more it willusually become apparent that some of theuncertainty areas are more relevant than oth-ers – and that some of the less prominent ones

182

Skills in Choosing

might perhaps be set aside, or amalgamatedwith others to which they may seem closelyrelated. So there will always be opportunitiesfor the use of judgement and creativity in refor-mulating particular uncertainty areas, and inrestructuring the overall list: opportunities ofa similar kind to those already addressed inChapter 5 in relation to the reformulation ofdecision areas.Although a long and unstructured list of

uncertainty areas can be tidied up at any time,where obvious opportunities for so doing canbe seen, it is in practice usually worth savingany serious efforts towards restructuring untila moment arrives when there are pressuresto explore preferences within a quite restric-ted working shortlist – perhaps a set of threeor four promising schemes or, in many cases,only two promising alternatives – within whichit is difficult to see where the overall balanceof advantage ties.The argument for leaving more detailed for-

mulation and investigation of uncertainty areasuntil such a moment stems from the obser-vation that in practice the relative prominenceof different uncertainty areas can change quitedramatically as the decision perspective shifts.For example, some of the uncertainty areaslisted in Chapter 3 (Figure 23) as relevant tothe choice between Schemes A and B couldwell be overshadowed by others if some thirdscheme – such as Scheme E – were broughtinto the comparison as well. One way of pick-ing out those uncertainty areas which are mostrelevant to a particular decision perspective issimply to use asterisks or some other sym-bol to pick out those in a longer list whichare agreed to be the most important in thatparticular light. However, Figure 72 illustratesone alternative form of presentation, which issometimes found useful because it reflects theemphasis on creative use of graphics in com-munications, which is characteristic of the stra-tegic choice approach.Figure 72 is an example of an uncertainty

graph, in which the more relevant uncertaintyareas are represented by labelled circles, usingrather similar conventions to those of the

decision graph. But there is an important dif-ference, in that the positioning of uncertaintyareas on the uncertainty graph is judged in rela-tion to a particular decision perspective thathas been selected in advance. In the centreof the graph is indicated the current focusof comparison, in this case the set of threeschemes, A, B and E, which was selected inFigure 22. Different directions of movementoutwards from this central focus then corres-pond to the three basic categories of uncer-tainty, UE, UV and UR, occupying broadly thesame sectors of the graph as in the generalintroductory diagram of Figure 3. However,the graph leaves the boundaries between thethree sectors undefined, so that any borderlinecases can be plotted in intermediate positions.In addition, themore relevant uncertainty areasare positioned closest to the centre of thegraph; these being the ones which are judgedto bear most closely on the difficulty experi-enced in making choices among the particu-lar alternatives currently in view. As an aid tothis aspect of positioning, two or three con-centric rings can be drawn around the centreof the graph. When working with wall chartsand coloured pens, these rings are best drawnin an unobtrusive colour, such as yellow, sothat they do not get in the way of the other,more specific, information which the graph isintended to convey.Working from an unstructured list such as

that of Figure 71, it is usually best to begin byscanning it to pick out the most relevant uncer-tainty areas first. These can then be positionedwithin the innermost ring, working methodic-ally outwards to add others of lesser relevance.Another approach is to begin by placing onthe graph those uncertainty areas that are bestunderstood and use these as points of ref-erence in positioning others. It is not usuallynecessary to overload an uncertainty graph byattempting to locate within it all uncertaintyareas, however insignificant relative to others;and it is rarely in practice worth going bey-ond a set of eight or nine uncertainty areas

183

FIGURE

72

SOUTH SIDE

EXAMPLE

Placing Uncertainty Areas in a Decision Perspective

This kind of radial uncertainty graph offers a useful way of putting the more important uncertainty areas

into a decision perspective, and bringing together judgements about type, relevance and tractability.

It is usual to start by plotting the more clearly understood uncertainty areas, and position others with

reference to these, reassessing the first few when this process is complete.

Skills in Choosing

which are judged to be of particular relevancein relation to the current focus of comparison.Although the classification of uncertainty areasas UE, UV or UR may have been noted on theoriginal working list, the act of plotting themon the uncertainty graph can provide a usefulopportunity for second thoughts and for fur-ther debate on any doubtful cases. Usually, atleast some of the more relevant uncertaintyareas will have come to the fore in the com-paring of particular schemes, for example, thethree uncertainty areas ?VALJOB, ?SITEJOBSand ?STEELSITE all emerged from the originalcomparison of Schemes A and B in Chapter 3(Figure 23). However, other uncertainty areasmay have surfacedwhile comparing other pairsof alternatives, or even in the work of the shap-ing and designing modes – for instance one ortwo of themmight reflect doubts as towhetherparticular options should be considered feas-ible or whether particular options bars shouldbe assumed.Various additional kinds of information can

be added to the basic information conveyedby an uncertainty graph about the type andrelevance of different uncertainty areas. Buttoo much elaboration can confuse rather thaninform; so the only additional information intro-duced in Figure 72 is the use of a serrated ringto pick out particular uncertainty areas whichare judged to be more intractable than others.The message is that there is thought to bevery little that could be done to reduce the cur-rent state of doubt in each such uncertaintyarea, whereas in the case of other uncertaintyareas it is possible to conceive of at least someform of exploratory action whereby currentfeelings of uncertainty might be reduced. Inthe case of these more tractable uncertaintyareas, any decision to carry out this explorat-ory action will then have the intended effectof pushing that uncertainty area some distancefurther outwards from the centre of the graph;however, in the case of a more intractableuncertainty area, this possibility either does notexist, or is assumed to be realisable only at anunacceptable cost.

RESTRUCTURING COMPOSITEUNCERTAINTY AREAS

Once attention has been focused on a fewof the most relevant uncertainty areas –whether through use of an uncertainty graphor simply through picking out the most import-ant uncertainty areas on a list – it is only tobe expected that doubts will begin to arise asto whether they have been formulated clearlyenough. In some cases, closer investigationwill suggest that a particular uncertainty area iscomposite in form, and could with advantagebe broken down into two or more separate ele-ments for which different kinds of exploratoryaction would be appropriate. Some of theseelements might then be found to be morerelevant than others to the present focus ofcomparison; furthermore, the elements mightoccupy quite different parts of the uncertaintygraph in terms of the UE/UV/UR classifica-tion. Therefore, the reformulation of compositeuncertainty areas in terms of their more signi-ficant elements can sometimes lead to quitea radical restructuring of the content of theuncertainty graph, and a reappraisal of the pic-ture of opportunities for managing uncertaintywhich it conveys.It is possible to explore such possibilities

for restructuring uncertainty areas by workingdirectly from the uncertainty graph. But thegraph itself can become overloaded and con-fused if too much information of this kind isadded. Figure 73 illustrates a way in whichthe elements of composite uncertainty areascan be explored on a separate sheet, with aview to possibly modifying the graph at a laterstage. The first example of a composite uncer-tainty area relates to the policy value of jobcreation; this is an uncertainty area occupyingquite a central position on the graph of Fig-ure 72, which was classified as of type UV.In Figure 73, this value uncertainty is brokendown into two different value elements, oneof them relating to the general priority given bythe municipal council to job creation as againstother policy aims; and the other relating tothe more specific question of whether there is

185

FIGURE

73

SOUTH SIDE

EXAMPLE

Reformulating Composite Uncertainty Areas

This splitting down of composite uncertainty areas into more specific elements is often carried out

directly on the uncertainty graph, reassessing type and relevance in the process. Judgements about

the relevance of different elements can be seen as a type of informal sensitivity analysis, and the more

explicit form of sensitivity analysis illustrated here is used comparatively rarely.

Skills in Choosing

to be discrimination in favour of economicallydeprived neighbourhoods, such as South Side,in the attraction of new jobs to different partsof the municipal area. The meaning of eachelement is made more explicit in Figure 73 byspelling out, in parenthesis, a set of two ormore alternative assumptions which can beregarded as representative of the current rangeof doubt. Such a set of alternative assump-tions is exactly analogous to the set of optionsused to represent the range of choice withina decision area; indeed, within an uncertaintyarea, similar problems of howbest to representa complex or open-ended range of possibilitiesmay sometimes have to be faced.In this instance, it is possible that the par-

ticipants will be experiencing doubts overwhether the general priority attached by mem-bers of the municipal council to job creationhas shifted upwards or downwards since itslast written policy statement on this issuewas agreed. But it may be judged that thiselement of uncertainty is not so importantas the element of uncertainty relating to pos-itive discrimination is in favour of particularlocalities, about which there might recentlyhave been much political controversy. There-fore, Figure 73 shows the latter element ratedas of medium-to-high relevance and so worthsubjecting to further scrutiny straight away;while the former – rated as of only low-to-medium relevance – can perhaps be setaside for the time being. The latter elementis assigned a brief label – ?JOBDISC – witha view to this replacing the composite uncer-tainty area ?VALJOB on the uncertainty graph.As a further step in analysing this more care-fully defined uncertainty area, a note ismade ofthe kinds of exploratory option which could beconsidered to reduce its relevance further. Inthis example, it is indicated that two alternativelevels of policy soundings might be consideredas possible exploratory options – the first ofthem more formal and the second comparat-ively quick and informal.Turning to the second of the uncertainty

areas in Figure 73 – labelled ?EASTWL – its ori-ginal description in terms of ‘scale and timing’

of the proposed Eastwell shopping complexindicates that it also is composite in form. Itis supposed here that further discussion of itscontent leads to the judgement that it is uncer-tainty over scalewhich ismuch themore prom-inent element. In the case of ?SERVCOST, itis supposed in Figure 73 that the analysis ofelements can be taken further with the helpof some more explicit sensitivity analysis, ofa kind that only becomes possible in circum-stances where assessments can be made onthe basis of well-defined sequences of calcu-lations or logical steps.The particular procedure of sensitivity ana-

lysis used in Figure 73 can be appreciatedmore fully by referring back to Figure 62. Thisindicated a sequence of operations involved inestimating the costs of providing site servicesfor either industrial or housing development.Some of these steps involved calculations ofa straightforward arithmetical kind, but othersdepended on expert judgements – and thelogical structure of these will often be muchless transparent.In the particular example of sensitivity ana-

lysis included in Figure 73, it has been sup-posed that it is the cost of servicing the site forindustrial use rather than housing use which isthe source of most of the uncertainty; and therelative contributions to this uncertainty of twocontributory elements are explored. The differ-ence in usable site area for the two alternativeuses was estimated as 0.5 ha (Figure 62) –the difference between 2.5 ha for housing and3.0 ha for industry. In Figure 73 it is supposedthat the range of uncertainty over this estim-ate has been assessed, through a process ofsurprise limit analysis, as extending fromamin-imum of 0.3 ha to a maximum of 0.7 ha. In thesameway, a range from 10 to 20 k/ha has beenassessed for the cost per hectare of servicingthe site for industry (see Figure 63); the equi-valent cost of servicing for housing being takenas a more predictable 8 k/ha.Repeating the sequence of calculations in

Figure 62 with theminimum andmaximum fig-ures substituted for the original estimates, first

187

FIGURE

74

SOUTH SIDE

EXAMPLE

Comparing Alternative Responses to Uncertainty

It is only in the case of highly relevant and relatively intractable uncertainty areas that it is likely to

be worth comparing different exploratory options in as much depth as illustrated here. Nevertheless,

the general principles apply to any judgement about alternative responses to uncertainty, including

judgements about whether to take no current action but prepare contingency plans.

Skills in Choosing

for the loss of site area and then for the ser-vicing cost per hectare, Figure 73 shows thatthe former element causes very little variationin the estimated servicing cost for the site,while the latter element contributes consider-ably more.2

In pursuing this example further, it will benoted (from Figure 62) that this estimatedsite servicing cost per hectare can itself bebroken down into four more specific elementsof assumption – some expressed in numer-ical form and others not. So the process ofbreaking down elements of uncertainty can, inthis case, be extended further. In Figure 73this point is illustrated by indicating the twomore important sub-elements of the uncer-tainty over industrial site servicing cost – themore significant of them in this case beingof an intractable policy nature. This illustratesthe general point that analysis of elementswithin an uncertainty area can lead to a revi-sion of views about its classification within theUE/UV/UR framework.

ASSESSING EXPLORATORYOPTIONS WITHIN CRITICALUNCERTAINTY AREAS

Once attention has been focused on a few cru-cial uncertainty areas, attention can be turnedto the question of what might be done aboutthem. This involves examining more closelyany exploratory options that are seen as real-istic – whether they have already been iden-tified or whether they only come to mindat this stage – and weighing up what theconsequences of following these exploratoryoptions might be.In Chapter 3, one approach was illustrated

(Figure 24) by which the implications of follow-ing any particular exploratory option could be

2 This is a contrast which can be quantified if statist-ical methods are used to calculate – on conventionalassumptions of independence and normal distributionalform – that the latter element explains some 97% ofthe combined variance.

compared with the consequences of not pur-suing it. In this, the three comparison areas ofconfidence, resources and decision delaywereused to represent the three most importantdimensions of evaluation that normally arisewhen making judgements about how uncer-tainty should be managed. In practice, it is onlyrarely that it is worth while evaluating explor-atory options in this explicit way. But the prin-ciples involved are crucial to the managementof uncertainty in strategic choice; so Figure 74illustrates how the comparison of exploratoryoptions can be taken a step further in relation toany especially critical uncertainty areas wherethis deeper level of analysis may be justified.Because the proposed scale of the Eastwell

district shopping complex has emerged as anespecially relevant uncertainty area (Figure 74),the implications of taking different exploratoryactions in response are reviewed in Figure 74.The first exploratory option to be consideredis that of undertaking a full and systematicanalysis of alternative combinations of districtandmore local shopping centres in the broadereastern sector of the municipality that includesboth Eastwell and South Side. But, in addition,a more modest exploratory option is also con-sidered, which might go at least some waytowards reducing the level of doubt within thissame uncertainty area; this is the option ofengaging in some informal liaison with lead-ing members of a specialist team of plannerswhich the municipality has set up to look atshopping provision within its overall admin-istrative area. This second option may meanrelying on the specialist team’s own expertjudgement of the consequences of adoptingdifferent shopping patterns, rather than on anymore rigorous methods of survey and analysis.Nevertheless, it could be well worth consid-ering as a more modest – but possibly moreeffective – way of reducing the level of doubtin this same uncertainty area.Figure 74 records the judgement that the

first exploratory option – labelled ANALYSIS –is expected to lead to more confidencein judging the overall balance of advantage

189


between Schemes E and B – this pair com-parison being a more appropriate one to con-sider than A:B within the working shortlist(A:B:E) because A and B do not differ in termsof options within the SHOP LOC’N? decisionarea. But the more modest LIAISON option isalso judged likely to lead to ‘somewhat’ moreconfidence in relation to the samepair compari-son – and to do so at considerably less costin terms of both demands on resources anddecision delay. The call on resources for theANALYSIS option is here assessed in termsof a rough estimate of the number of plannerdays required – although of course it could haveother dimensions such as the use of computertime and the cost of origination of data, pos-sibly involving the recruitment of interviewersthrough agency or other channels.To weigh up whether it is likely to be worth

pursuing the full ANALYSIS option as againstthe more modest LIAISON option, it becomesessential to take the urgencies and resourcepressures of the current decision situationinto account. In practice, this will usually bedone intuitively. But intuitive appreciations ofurgency may differ from one participant toanother; so it can sometimes be a useful aid tocommunication, in the case of especially crit-ical uncertainty areas, to introduce an advant-age comparison framework as an aid to thiskind of judgement. This possibility is demon-strated in the lower part of Figure 74.In this case, the advantage comparison indic-

ates a view that the additional gain in con-fidence from adopting the ANALYSIS ratherthan the simpler LIAISON response to theuncertainty about the scale of the Eastwellshopping centre should be placed in themarginal-to-significant range. Yet Figure 74shows that this benefit is likely to be out-weighed by the considerable additional invest-ment of resources required, along with theevenmore serious implications – in this particu-lar decision situation – of a four-month delaybefore the decision in question can be made.This example merely makes explicit the kind ofjudgements that are being made all the time inan intuitive way, when people have to decide

how far to invest in any kind of exploratoryaction to improve the confidence with whichimportant choices can be made.This particular example concerns an uncer-

tainty area of type UR; but the judgementof whether or not to invest in exploratoryaction can be expressed in a similar waywhether an uncertainty area of type UE, UVor UR is involved – as was indeed demon-strated earlier in Chapter 3 (Figure 24). Quickand informed judgements of this kind canbe debated with reference to the comparat-ive information about the relevance of differ-ent uncertainty areas which is contained in anuncertainty graph such as that of Figure 72.As a general rule, it is rarely worth makingmajor investments in exploratory action direc-ted towards uncertainty areas of lower rele-vance, so long as uncertainty areas of higherrelevance remain.

RELATING EXPLORATORYACTIONS TO THE TIMING OFDECISIONS

As was discussed in Chapter 3, the consider-ation of what to do about particular uncertaintyareas can bring concerns about the timing ofchoices in different decision areas directly tothe fore. For any exploratory action invariablytakes some time to carry through – whetherthis be measured in minutes, hours, days,weeks, months or even years – arid can thusimply delays of a more or less serious extent inthe taking of those decisions they are designedto inform.The opportunities for taking immediate

actions in some decision areas while defer-ring choice in others were approached inChapter 3 through the core concept of theaction scheme, chosen to cover some but notnecessarily all of the decision areas within aproblem focus. An example was then presen-ted (Figure 25) of how action schemes couldbe compared in terms of the relative flexibil-

ity of choice left open, and various possibleapproaches to the comparison of flexibilitywere discussed in Chapter 7.

190

Skills in Choosing

There are, therefore, two types of timingjudgement which have to be considered inpractice. On the one hand there is the judge-ment of how far to invest in pursuing explorat-ory actions which imply delays in at least somedecision areas; on the other hand there is thejudgement of how far to make commitmentsto action in some decision areas in advanceof others. This means that any consideration ofthe time dimension can make the structure ofinterconnected choices for participants moresubtle and complex; indeed, this is a reality thathas to be faced whether or not the concepts ofstrategic choice are being applied in an explicitway.In dealing with these choices, a useful

guideline is to focus first on any decisionareas where considerations of both urgencyand uncertainty arise. This is illustrated in Fig-ure 75, which takes as its point of departurethe set of nine possible decision schemes firstdeveloped in Chapter 2 (Figure 17) for the fourdecision areas in the original South Side prob-lem focus. In Figure 75, these schemes arearranged (as in Figure 25) with the two moreurgent decision areas – ROAD LINE? And DISTLIFE? – brought to the fore; then, focusingmore closely on the comparison of schemeswithin the more limited working shortlist ofA, B and E, Figure 75 identifies the points ofgreatest difficulty in the decision process, bymeans of question marks positioned at therelevant branching points of the tree.It can now be seen that the DIST LIFE?

decision area is the more urgent of thetwo decision areas in which there are differ-ences between the shortlisted schemes A,B and E; and it has already been discovered(Figure 74) that the ?EASTWLSC uncertaintyarea has an important bearing on the choiceof routes at this branch point of the tree. So itcould be especially important to explore howto deal with this particular uncertainty area ifthe urgencies of the problem situation are tobe addressed. So, in the second picture of Fig-ure 75, the first part of the option tree is shownexpanded to introduce, as additional branch-ing points, a choice of two exploratory options

within the ?EASTWLSC uncertainty area. Thefirst is the ‘null option’ of taking no explorat-ory action at all, and the second is the quickoption of a brief liaison exercise with the shopsplanning team – it being here supposed thatthe more costly and time-consuming option offuller analysis has been ruled out after the kindof advantage comparison exercise illustratedearlier (Figure 74).In Figure 75, the assumption is made that

the DIST LIFE? decision should definitely bedeferred if the LIAISON exploratory option isto be followed through. It is also supposedthere may possibly be an argument for defer-ring choice of DIST LIFE?, as an alternative toearly commitment to any specific option, if it isdecided to take the ‘no explorations’ route. Forsuch a deferement could make sense if therewere any likelihood that some event mightoccur which could clarify the choice withinDIST LIFE? without any explicit exploratoryaction being taken. For example, it might beknown that there was a meeting of a SouthSide community forum scheduled in a week’stime to debate this very issue.The introduction of an option to defer choice

within the DIST LIFE? decision area has theeffect of subtly altering the meaning of thatdecision area; for it now represents a choice asto what should be done now about the life ofthe district, rather than what should be done inany more timeless sense. However, the mainpoint about the expanded option tree in Fig-ure 75 is that it displays five possible com-binations of current choices and exploratoryoptions that can be compared with each otheras a basis for incremental progress; and thesefive paths can quickly be compared in termsof how many of the nine decision schemesremain available in each case. In Figure 75, theschemes remaining open are listed for eachpath, drawing attention in particular to the avail-ability of the three shortlisted Schemes A, Band E.In general, this analysis indicates that those

courses of action involving immediate commit-ment in the DIST LIFE? decision area carry a

191

FIGURE

75

SOUTH SIDE

EXAMPLE

Weighing Uncertainty against Urgency in Decision-Making

In considering how far to defer decisions in strategic choice, it is useful to focus on any decision areas

which may be urgent yet subject to major uncertainties which could be reduced through exploratory

action. The judgement of whether to decide now or defer while explorations are carried through is usually

made informally in constructing a commitment package, but the principle is as illustrated here.

Skills in Choosing

risk of subsequent regret, through the fore-closing of particular decision schemes whichmight have been found to be advantageoushad the explorations been followed through.For example, commitment now to a 40-yearlife for South Side might lead to regret if thework of planning team of the specialist shopswere to result in a proposal for a particularsector-wide shopping pattern which made ituneconomic to develop a new South Side localshopping centre on the King Square site; whilecommitment now to a 10-year life might leadto regret if geological difficulties on the centralsite were later found to be so severe that nouses other than open space could realisticallybe considered there.In general, the choice of whether to opt

for commitment or deferment in an urgentdecision area involves some process, howeverintuitive, of weighing the risks of early com-mitment against the negative consequences ofdelay – including any political or professionalpenalties that might be incurred. In the particu-lar circumstances of Figure 75, the defermentbeing considered is only a couple of weeks, somight well be considered worth while; unless,perhaps, there were an imminent meeting ofthe municipal council, or a closely fought localelection, to introduce an added note of urgencyover the commitment to district life, evenwithin this otherwise insignificant time scale.

ACCOMMODATING UNCERTAINTYBY PREPARING FOR ALTERNATIVECONTINGENCIES

Various approaches have now been discussedto the comparison of alternative courses ofimmediate action, in terms of the flexibility offuture choicewhich they allow. At the simplestlevel, an intuitive judgement can be made thatone alternative is likely to leave open moreopportunities for future choice than another,without any analysis of what these opportunit-ies are.3 At another level, the patterns of future

3 This was the kind of approach adopted when ‘flexibil-ity’ was introduced as an additional comparison area inFigure 64.

choice left open can be analysed and presen-ted for visual comparison, or the comparisoncan be simplified by introducing some form ofrobustness index (Figure 25).However, it is possible to go further in cir-

cumstances where there has been some ana-lysis of the relative importance of differentuncertainty areas, and where some attempthas been made to represent the more import-ant of these in terms of the range of alternativeassumptions that could be held. Where this isso, then it is possible to bring those alternativeassumptions more directly into the compari-son of current actions, and to explore how farsome such actions could have advantages overothers in their capacity to respond to particulareventualities that could arise. One way of tak-ing the analysis in this direction is illustrated inFigure 76. In Figure 76, two alternative actionschemes for South Side are compared, eachof which involves commitment to the north-ern road line and also to a particular optionin the DIST LIFE? decision area. Referringto Figure 75, each such course implies thatno exploratory action is being taken in rela-tion to the important ?EASTWLSC uncertaintyarea. So the feelings of uncertainty over thescale of the proposed Eastwell shopping com-plex remains unchanged; and this uncertaintyarea has already been represented in termsof a choice of three possible assumptionslabelled AS IN PLAN, LARGER and SMALLER(Figure 76).According to which of these assumptions is

held, either the feasibility or the relative attract-iveness of different courses of future actionmay be affected. In Figure 76, the judgementis made that the uncertainty about the scalefor the Eastwell shopping complex is of suchdirect relevance to the choice of shoppinglocation for South Side, that the King Squaredevelopment must be ruled out on grounds ofeconomic viability if the Eastwell complex is tobe significantly larger than proposed in the cur-rent plan. Turning to the second of the actionschemes in Figure 76 – Action Scheme III –this appears to allow no alternative to theKing Square site; but it does afford protection

193

FIGURE

76

SOUTH SIDE

EXAMPLE

Accommodating Uncertainty in the Future Decision Space

Uncertainty areas in which the choice of assumptions can have a critical effect on the choice of path

through a tree may become an important focus for contingency planning. This consideration is usually

introduced only when attention is turned to the design of a commitment package, but the extension of

the tree to show contingencies as extra branching points can be a useful aid to communication.

Skills in Choosing

against another contingency – the perhapsremote contingency of abnormal geologicalconditions being discovered on the central site,precluding the possibility of any use exceptthat of open space.However, another point in favour of Action

Scheme I is that it leaves open the optionof industrial use on the central site; providedgeological conditions make this option feas-ible, Figure 76 shows that its attractivenessas a means of creating alternative local jobscould be enhanced, should the contingencyof early closure of the steelworks arise. So,the consideration of flexibility to accommodatedifferent kinds of uncertainty can becomemore complex, the more contingenciesare explored. Sometimes this kind of analysiscan lead towards a searching re-examinationof earlier assumptions. For example, theassumption that there is no possibility of usingthe central site for open space if a 10-yeardistrict life were chosen might now seem arather unnecessary constraint. Referring backto the structure of the problem as reflectedin the option graph (Figure 16), the reasoningbehind this particular option bar could wellnow be challenged. The result could be anagreement that the option bar was no longernecessary, so removing an apparent advant-age towards Action Scheme III on groundsof capacity to accommodate geologicaluncertainty.Once the possibility is considered of introdu-

cing someof themore crucial uncertainty areasinto the structure of possible paths through theproblem situation, as illustrated in Figure 76,then it becomes possible to move in a num-ber of different directions of further analysis,depending on whether these crucial uncer-tainty areas are of type UE, UV or UR. Inthe case of a UE-type uncertainty area – suchas that over geological conditions in SouthSide – it may be important to ask whetherthe contingencies reflected in the alternat-ive assumptions are of high or low probab-ility: and questions then arise of how far itis worth safeguarding or insuring against spe-cific risks. If these questions are of sufficient

importance, then the analysis can be taken inthe direction of the classical form of decisionanalysis in which numerical probabilities areassigned to different contingencies treatedas alternate branching points in a decisiontree.4

Any important uncertainty areas of type UVcan, if desired, be absorbed into the ana-lysis of decision areas expressed at a broaderpolicy level – and such a step will open upthe possibility of introducing the methodsalready illustrated for analysing relationshipsbetween levels of choice (Figure 54). Again,any crucial uncertainty areas of type UR canbe absorbed into the problem structure as add-itional decision areas, thus in effect extend-ing the problem focus. However, where thesenew decision areas relate to choices which areentirely or partially under the control of otherparties – and, where there are elements of con-flict or competition in relationships with thoseother parties – then the decision situation mayhave to be considered as more like one ofa game in which moves, counter-moves andpoints of potential stability might have to beexplored: ‘If they did this, we could retaliate bydoing that’. This points to another direction ofanalysis, based on the ideas about themanage-ment of conflict and collaboration discussed byauthors such as Howard (1989) and Bennett,Cropper and Huxham (1989).

MOVING TOWARDS THE DESIGNOF POSSIBLE COMMITMENTPACKAGES

This chapter has, so far, covered various kindsof judgement which are relevant to the designof a commitment package, conceived as a pro-posed incremental step towards commitmentthrough time. The four basic sections of acommitment package were first illustrated inFigure 26: a set of immediate decisions cov-ering both actions and explorations to reduce

4 See, for example, Raiffa, 1968.

195

FIGURE

77

SOUTH SIDE

EXAMPLE

Building an Appropriate Commitment Package

The basic framework of a commitment package offers a framework for discussion in considering how

to act which can be extended in a number of different ways, according to context. When working with

large sheets of paper it is useful to add marginal notes against immediate decisions specifying agreed

organisational responsibilities, resource commitments and deadlines.

Skills in Choosing

uncertainty; then a set of proposed arrange-ments for deferred choices and contingency

planning within a future decision space. In anysettingwhere decisions call for formal commit-ment of organisational resources, there will bemoments when it is important to give carefulattention to the design of alternative commit-ment packages. Yet it is important to emphas-ise that at other times, when there is a desireto move ahead informally, or to respond toparticular urgencies of the moment, a commit-ment package may be much more skeletal inits content, and will not necessarily be recor-ded in written form.In practice, important points of judgement

can arise not only in developing the content ofa commitment package, but also in judging theformat in which it should be presented. Thisapplies in particular to the rows as opposed tothe columns of the framework. In the simpleexample presented in Chapter 3 (Figure 26),the problem structure was brought out by des-ignating a separate row for each decision area.This presented no difficulty in that particularcase, because each of the exploratory actionsproposed to address uncertainty could be dir-ectly linked to a particular area of deferredchoice. In practice, however, the number ofdecision areas to be brought together withinthe compass of a commitment package cansometimes be larger, and the relationshipsbetween explorations and deferred choicesless straightforward. In these circumstances, itcan be more practicable to group two or moredecision areas together in the same row, per-haps relating to some designated sphere ofresponsibility. This way of presenting a com-mitment package is illustrated in Figure 77.This example of a commitment package

presents one coherent set of proposals whichis, in effect, an expanded version of the firstof the two alternative action schemes com-pared in Figure 76. The rows of the frame-work are organised into three broad spheresof responsibility, concerned respectively withlocal South Side matters, with transportissues and with economic development in the

municipality as a whole. Only the first ofthese is within the direct sphere of respons-ibility of the South Side Working Party; andeven then there may be procedures of formalauthorisation or endorsement of proposals tobe followed, requiring some consideration ofcontingency planning if any recommendationsshould be rejected or opposed. In the trans-port and economic development spheres, allthe working party may be able to do is to makerepresentations to other parties, and to usewhatever influence or leverage they can exertto follow those representations through – per-haps again with some thought to possible con-tingency actions, should the proposed repres-entations or negotiations fail.In Figure 77, asterisks are used to indicate

the principal set of interconnected decisionareas around which the commitment packagehas been built, as a reminder that it may be dif-ficult to consider changing course in any one ofthese, without considering what implicationsthere might be for the others. However, theremay be more scope for considering variationsin other elements of the package; and the num-ber of elements shown in this particular illus-tration, though quite large, is only a limitedselection from those that could have beenincluded if all the decision areas and uncer-tainty areas discussed in the course of thelast few chapters were to be included. Forexample, extra spheres of housing or finan-cial responsibility could have been included toreflect these other organisational interests inthe South Side problem (Figure 41). Variousother decision areas from this wider decisiongraph might then have been considered, inthe shaping of either proposed actions ordeferred choices. Further kinds of technicalexploration could also have been suggested.For example, explorations could have been pro-posed into the feasibility of doubtful options,such as the possibility (examined in Figure 48)of combining some industrial with some hous-ing development of the central site. Furtherexplorations into policy values might also per-haps have been recommended: for example,

197

FIGURE

78

SOUTH SIDE

EXAMPLE

Helping Decision-Takers to Make Progress

With ingenuity, it is often possible to present action proposals to decision-takers in forms which do not

appear to deviate significantly from familiar forms of report, yet which reflect the underlying structure of

a strategic choice analysis, as in this illustration. Use of diagrams should be viewed with caution and

contingency plans may not be indicated explicitly where sensitive negotiations may arise.

Skills in Choosing

into the issue of discrimination in job creationin favour of low-income neighbourhoods.So, however broad in scope it might appear,

any commitment package is designed to beselective rather than comprehensive in itscontent, reflecting the various resource con-straints, urgencies and priorities of the specificdecision setting within which it is shaped. Forit is intended as an incremental step in a con-tinuing process of commitment, rather thanas a conclusive response to the full range ofproblems that has been identified. Even so,there are many additional subtleties that canbe introduced informally into a decision pro-cess while a commitment package is beingdesigned, extending its scope beyond that ofthe analysis as so far pursued. For example,Figure 77 introduces for the first time the pos-sibility that the uncertainty about the steel-works closure might not be so immune tomunicipal influence as earlier assumed, oncethe idea of a joint initiative to attract smallbusiness enterprises to the site has been con-ceived.So, within the commitment package frame-

work, a considerable amount of richness andcomplexity can be encompassed, even wherea conscious effort is made to present its con-tent in a form which is intelligible to peoplewho may not have been involved in the ana-lytical work which has gone into its present-ation. Cross-referencing between elements inthe package itself can be indicated by dis-creet annotations; in Figure 77, for example,row and column codes are used – spar-ingly – for cross-reference both within thecommitment package and also in the briefstatement of comparative information whichappears below. The intention is that the formatselected for the commitment package shouldoffer a basis for structured debate aboutchoices and assumptions; for closer compari-son of alternatives, either within particularelements or on a broader scale; for modifica-tion to reflect new insights or representations;and – when the moment is ripe – forprogress through commitment to decisiveactions.

HELPING DECISION-TAKERS TOCHOOSE

Even a carefully organised and selective com-mitment package such as that of Figure 77can present an overload of information forthose decision-takers who may be called uponto authorise decisive actions. This problem ofpotential overload is of course not specific tothe strategic choice approach; it applies equallyto any form of decision technology, whereveractions have to be endorsed by busy peoplewith multiple roles and responsibilities, whomay not have the time to keep themselvesclosely acquainted with the progress of anyanalytical work.However, if decision-takers are to exercise

their right to choose responsibly in other thana token and ritualistic way, it can sometimesbe important that they be presented with morethan one commitment package from whichto choose. Judgements must then be madeabout how many alternative packages shouldbe presented, how much back-up informationshould be offered in support; and what thebalance should be between different forms ofcommunication – text, graphics, numerical tab-ulations, verbal presentation – as opposed todocuments circulated in advance. These arejudgements which will depend on the particu-lar decision context, and the way in whichits opportunities and constraints are under-stood by those who carry the responsibility ofpresenting proposals to decision-takers.Figure 78 gives one example of the way in

which the members of the South Side Work-ing Party might decide to present their initialrecommendations to the policy committee towhich they report. It is here assumed thatthe latter group, with many pressures on itsagenda, is accustomed to acting mainly on thebasis ofwritten documentation of a condensedsummary form – backed up, if need be, byfuller information introduced in the course ofdiscussion.Figure 78 therefore takes the form of a

one-page summary report, presented in fairlyconventional written form, offering guidelines

199


which can be followed, questioned or chal-lenged at the discretion of the decision-takers –in this case, the members of the municipalpolicy committee, meeting as a group withthe authority to commit resources to whatevercourse of action they may agree. To keep thereport brief and readable, it is restricted in itsfocus to two main alternatives expressed interms of contrasting policy orientations – oneof these being a condensed version of the com-mitment package of Figure 77, and the othera contrasting package based on a 40-year hori-zon for the life of South Side as a residentialarea. In the interests of brevity, all but themostcrucial characteristics of the two packages areomitted from the report; for the main aim ofthe summary is to indicate what are the maindifferences in structure between the two.It is likely that the main structural assump-

tions behind the alternative proposals will havebeen put to quite stringent testing well beforethis reporting stage; but it is important that theyshould be checked again carefully before thereport is finally submitted. For example, if therewere any serious residual doubts as towhetherthe 40-year life might, after all, allow the MainStreet shopping centre to be retained, or as towhether industry on the central site might, insome circumstances, be compatible with theKing Square shopping location, then the wholestructure and emphasis of the report could besubject to challenge. The proposals might theneither have to be redesigned in their entirety– or at least revised to allow this element ofuncertainty to be acknowledged and an appro-priate response prepared.In any case, it will be only prudent that

the working party members should preparethemselves carefully for the sorts of debatethat could conceivably develop at the meet-ing, on the basis of their prior knowledgeof the responsibilities of committee mem-bers, their interests, their power relationshipsand the idiosyncrasies or prejudices of par-ticular members. As a first step, they couldbe well advised to prepare more detaileddescriptions of each of the alternative com-mitment packages – perhaps using wall charts

or other visual aids to present a broad pictureof actions, explorations and future choices ofthe kind illustrated in Figure 77. They mightalso agree to come to the meeting armedwith this kind of picture not only for the twocontrasting packages highlighted in the report,but also for a ‘compromise’ 20-year package,to demonstrate how marginal the differenceswere between this and the 40-year alternative.Indeed, they might agree that it was a goodmove to prepare a further commitment pack-age based on the southern road line, to indicatehow limited a future it offers to South Side andwhat a strong case the Council has in recom-mending the northern alternative to the trans-portation authority.Knowing the political inclinations of the com-

mittee members – which might cover quitea broad spectrum – the working party mem-bersmight also feel it was important to preparecarefully the case for an emphasis on publicrather than private development, in terms ofthe constraints presented by the local situationin South Side. They might, however, anticipatethat the first contingency planning elementof the package in Figure 77 offers a possiblebasis for helping the committee to convergeon an agreed compromise view. They mightalso anticipate some surprise from committeemembers that this package is thought to offerlocal residents better short-term economic pro-spects than one based on a longer-term invest-ment horizon. The explanation could be thata longer-term commitment to the future ofSouth Side was expected to lead to risinghouse prices followed by gradual disintegra-tion of the existing deep-rooted communitystructure; so they might have to be preparedfor some challenging debate in the committeeabout the subtle issues of policy values thiscould raise.The example of a commitment package

given in Figure 77 is of a form whichassumes that the decision-takers to whom it isaddressed have significant influence not onlyin the sphere of responsibility of the Work-ing Party itself, but also in their relationshipswith others. But this may not always be the

200

Skills in Choosing

case; and a decision situation involving nego-tiation between two or more autonomous orsemi-autonomous parties may call for a dif-ferent approach – especially if the parties arepotentially in conflict, in which case it might notbe thought tactically wise that all elements ofthe commitment package should be declared.For example, if the working party did not feelthat the Council was in a position powerfulenough to ensure that its recommendationof the northern road line would be accepted bythe transportation authority, then it might haveto rethink the whole commitment package interms of a more subtle negotiating stance.To take another example, it might be thoughtmore tactful – and more likely to yield results –if the idea of a joint small business initiative onthe steelworks site were seeded informally inthe mind of the Municipal Enterprise Officer,rather than suddenly presented to him or her ina background document during a formal com-mittee meeting. So, the overt content of writ-ten proposals and background documents canbe as much a matter of choice as any otheraspect of the decision-making process.To round off this particular episode in the

South Side story, it will be supposed thatsomeparticular commitment package is accep-ted by the committee as a basis for action –whether this be Package IA, as presented inFigure 77; or some variant of it which hasattracted support in the course of discussions;or possibly some quite different package, per-haps embodying a higher level of commitment;or perhaps a lower level. For it is the committeethat has the final responsibility and authority todecide.Whatever the outcome of the meeting, the

use of strategic choice concepts and methodsdoes not mean that the problems of the SouthSide community are now ‘solved’ in any final orcomprehensive way, any more than it is everrealistic to expect in a case as complex as this.But progress has beenmade; and it is progressboth in the direction of action and of learning.For both the decision-takers and their advisersshould have gained valuable insights into notonly the problem but also the process; insights

whichwill help to lay foundations for more con-fident decisions when further chapters cometo be written in the continuing South Sidestory.In Chapter 10, an illustration will be presen-

ted of the way in which the format of thecommitment package has been adapted inthe design of computer software to supportthe strategic choice approach. While the basicgrid format of Figures 26 and 77 is closely fol-lowed in the design of the software, the phraseprogress package is adopted in this context asa generic description as opposed to the phrasecommitment package.The aim of this is to encourage users to

develop alternative packages, which can bestored in different data files, before proceedingto the further step of adopting a preferred pro-gress package as an agreed basis for action.The STRAD software offers options of adopt-ing any progress package either as a basis forcommitment, in a situation where those gen-erating the package have sufficient authorityto do so, or as a basis for recommendation toothers, or as a strategic option to be comparedwith other strategic options by thosewho carrythe primary responsibility to decide.

PROCESS JUDGEMENTS IN THECHOOSING MODE

Of course, not every moment of commitmentin a continuous planning process need beapproached as formally or with as much pre-paration as this last episode in the South Sidestory might suggest. Sometimes, a commit-ment package might contain no more thana single action commitment on some aspectof the problem under consideration; and thisaction may fall within the personal sphere ofresponsibility of a single decision-maker. Per-haps this action may be accompanied by adecision to embark on some form of explorat-ory action to provide a clearer basis for otherchoices that have been deferred. Perhaps too –though not of necessity – that small step in

201

FIGURE

79Process Choices when Working in the Choosing Mode

Skills in Choosing

the process will include some agreed arrange-ments for the making of those other choicesat a later time.But the latest episode of the South Side

story does illustrate some of the subtle con-siderations of dialogue between advisers anddecision-takers that can arisewhenever amoreformal moment of commitment arrives. Andthe story has at least touched on some ofthe complexities that can arise where thereis more than one authoritative decision-takingbody, and where different parties may be con-tinually pulling in conflicting directions. In thesecircumstances, progress may well depend onquite complex negotiations across organisa-tional boundaries – raising issues which will bediscussed further in the next two chapters.This chapter has shown how the work of the

choosing mode can draw on prior work in anyor all of the other modes – shaping, designingand comparing – sometimes in a very subtleway. Also, as summarised in Figure 79, thework of the choosing mode itself presentsmany opportunities either for looping brieflyin the direction of these other modes, or forswitching more deliberately to work for sometime in a different mode, if it is judged thatthe moment for making commitments has notyet arrived. In particular, once a commitmentpackage is adopted as a basis for immedi-ate decision, there will remain the deferredchoices to be dealt with at some later time.When this time arrives, all the various oppor-tunities for shaping, designing, comparing andchoosing will arise once again – perhaps indifferent forms, influenced by events and

reappraisals over the intervening period. Forthe process of strategic choice is above alla continuing one which cannot be isolatedfrom the processes of change within the widerenvironment in which it is set. The content ofFigure 79will not be discussed in detail, as wasthe content of the equivalent figures at the endof the last three chapters. Rather, it is left asan exercise for the reader to reflect on the vari-ous switching and looping opportunities whichit indicates. Indeed, the reader is encouragedto reflect on these not only in relation to thedevelopment of the South Side story in earlieractions, but also in relation to personal experi-ence in choosing strategically in his or her ownworking life.


This chapter again concludes with a set ofthree illustrations from practice selected toillustrate further points about the work of thechoosing mode.The first illustration shows how uncertainty

areas can be mapped and sorted quickly andinformally in a group setting. The secondillustration shows the use of a commitmentpackage framework in presenting the propos-als from a national policy development project,showing the wide range of actions and explor-ations involved in putting them into effect.The third illustration shows how uncertaintyareas were explored and a first trial commit-ment package was developed on flip charts ina workshop in Venezuela to explore options forresponse to a major flood disaster.

203

FIGURE

80Illustrations from Practice – Choosing 1

Theme: Putting Uncertainty Areas into Perspective through the Use of an Uncertainty Graph.

Project: Development of an Integrated Policy Framework for the Management of Solid Waste.

Context: Workshops with an Inter-Organisational Project Team, Province of Limburg, The Netherlands,

1986.

Skills in Choosing


These charts record work done by a projectgroup at the end of the first day of a series ofthree two-day workshops, designed to get amore substantial planning project under way.The group consisted mainly of Provincial Gov-ernment employees, and was charged withpreparing a policy plan to integrate the man-agement of different kinds of solid waste.In the first instance, uncertainty areas were

listed as they occurred to members of thegroup, drawing on their everyday experienceand on understanding gained through earlieranalysis (2). They ranged from the quite spe-cific (quantity, type and source of solid waste)to the relatively general (basis for recyclingpolicy).When this process slowed, the analysisproper was started, although the list was leftopen for additions if there were any.Firstly, each uncertainty area was classified

according to type, as can be seen down theright-hand side of the list (‘OG’ = UE in Dutch,‘OW’ = UV, ‘OR’ = UR). As usually happens, anumber of the uncertainty areas were seen asfalling between categories and so given dualclassifications.The second stage of the analysis then iden-

tified views about the difficulty or otherwiseof reducing uncertainty in the various areas.A pictorial distinction between ‘diamonds’(hard to reduce) and ‘jellies’ (easy to reduce)was adopted as a starting point. As might beexpected, this quickly developed into a more

subtle range which is shown just below theuncertainty graph (3).The third stage of the analysis involved

assessment of relevance to the plan. For this,use was made of stickers. Each member ofthe group was given four stickers and asked toplace them next to the four uncertainty areaswhich they considered most relevant to thepolicy plan. The results appear down the left-hand side of the list. The three boxed uncer-tainty areas emerged as more relevant thanthe others. They included the uncertainty oversupport for a policy of recycling; over level ofcommitment tomaking the planwork; and overpossibilities for combining with other fields ofpolicy.All this information was then transferred to

the uncertainty graph (3), which has as itscentre the focus: ‘Beleidskeuzen in PAP II’(policy choices in the second Provincial WasteManagement Plan). This was further analysedin line with principles outlined by means ofa simplified 2× 2 taxonomy (1). As a resultthe decision was taken to reduce uncertaintyarea number 6 about the level of support fora policy of recycling. However, because theproject was still at an early stage, it was pos-sible to give some more thought to ways oftackling the other central uncertainties whichseemed more difficult to reduce. Had theprocess been nearer the end, they would prob-ably have been tackled by making assump-tions, perhaps backed up by some contingencyplanning.

205

FIGURE


Theme: Finalisation of a Policy Statement using a Commitment Package Framework.

Project: National Policy on the Handling and Distribution of Liquid Petroleum Gas (LPG).

Context: A Multi-Disciplinary Inter-Ministry Working Party, Den Haag, The Netherlands, 1983–84.

Skills in Choosing

COMMENTRY ON FIGURE 81

This illustration is taken from a published casestudy (van de Graaf, 1985; see also Hickling,2001) of the application of strategic choicemethods to the development of a new nationalpolicy in the Netherlands for the landing, trans-portation, storage and distribution of LiquidPetroleum Gas (LPG). The author of the pub-lished paper was a civil servant in the DutchMinistry of Economic Affairs who played aleading role in the project alongside represent-atives of other Ministries, with Allen Hicklingand Arnold de Jong in the role of processconsultants.The situation was one in which the basic

assumptions of the previously adopted policyhad become eroded, because it had notproved possible to concentrate the landingof this economically attractive but hazardoushydrocarbon – of international importance as afuel and feedstock – at a single installation atthe mouth of the Rhine, with distribution bypipeline. For 2 years, the search for an alter-native policy had exercised the attention of civilservants from theMinistry of EconomicAffairs,the Ministry of Transport and the two Direct-orates of the Environment Ministry concernedwith land-use planning and environmental pro-tection. Progress had been blocked by thewell-known tendency to defend departmentalpositions. The explicit use of amore interactiveapproach based on strategic choice methodsallowed rapid progress to be made towardsan agreed set of policies. However, the parti-cipants found it useful to switch back from timeto time to a more adversarial mode of work-

ing, so as to appreciate more clearly the policyand other constraints they would have to con-tend with outside the context of the workinggroup.The working document reproduced here (1)

was one of several in a set which were neededto cover all aspects of the problem. The prin-ted framework adopted here is one that canbe used large (A1 size) on the wall or small(A4 size) in a loose-leaf file. Notice the explicitaccommodation of responsibility (‘wie’ – who),deadlines (‘wanneer’ – when) and budgets(‘fl.’ – guilders (money)). The decision-taker islisted on the left (‘secretariaat’), as is the orient-ation (‘grote stroom’). Otherwise it is a stand-ard commitment package framework.The condensed extract from the final

recommendations (2) is presented using thefour headings of the commitment packageframework – as was the actual set of propos-als. These were formally adopted in February1985 by the second chamber of the Dutchnational parliament. It will be noticed thatthe explorations are primarily of a technicaland economic nature, reflecting concern withuncertainty areas of the UE type. However, thesignificance of uncertainty areas of type UR isclearly indicated in some of the proposals lis-ted under the contingency plan heading – theother decision-makers identified ranging frominternational regulatory bodies to local agen-cies concerned with the siting and safety offilling stations. Uncertainties of type UV in theform of the conflicts between economy, safetyand other criteria were accommodated in thedesign of the inter-ministry working process.

207

FIGURE


Theme: Developing proposed actions in response to key uncertainty areas, so as to build an initial

progress package.

Project: Post-disaster reconstruction in the Province of Vargas, Venezuela.

Context: Second session of inter-agency workshop convened by Ministry of Planning and Development,

Caracas, Venezuela, March 2000.

Skills in Choosing


These three photographs illustrate the processadopted to make rapid progress towards a firsttrial progress package during the second dayof a workshop convened in March 2000 bythe ViceMinister of Planning andDevelopmentin the Venezuelan government. The purposeof the workshop was to explore options forresponding to the challenges of reconstruc-tion within the coastal State of Vargas, and inthe adjoining States that together comprise themetropolitan region of Caracas, in the wakeof the disastrous mudslides that had inund-ated the coastal communities of Vargas inDecember the previous year.Representatives of several government

departments, emergency services, technicalsupport agencies and planning consultanciesinvolved in the reconstruction programmewere invited, at short notice, to participate intwo exploratory workshop sessions to be heldon consecutive afternoons.During the first session, some twenty

decision areas were generated in the full groupand prioritised, by the use of stickers, fordeeper investigation. There appeared to befew direct links between the most criticaldecision areas, which varied from the relat-ively technical – to do with the design ofearly warning systems for floods – to widerquestions of the future economic role of theregion. There was energetic debate over theissue of how far ecological as well as engin-eering considerations should influence thedesign of works to protect for the upper rivervalleys.Five decision areas were carried forward

to the second session, when five sub-groupswere formed and invited to explore onedecision area each in more depth. Each sub-group was asked to concentrate on a pairof realistic yet contrasting options and todraw up a pair comparison of these on a flipchart, taking all relevant comparison areas intoaccount. They were asked to indicate ranges

of uncertainty for their assessments on eachcomparison area, and to arrive at a group judge-ment on whether or not their overall prefer-ence between the alternatives was clear (1).They were then asked to indicate the mainsources of uncertainty on another chart, show-ing any exploratory actions they would recom-mend to deal with these, before presenting theresults of their work to the full group (2).While they were presenting these conclu-

sions, the main proposals were being loggedusing ‘Post-It’ slips on a pair of flipchartsformatted in the form of a progress pack-age – or commitment package – grid. At thisstage of the workshop, it was not surpris-ing that there were fewer entries in the firstcolumn, concerned with recommendations fordecisions now, than in the second column,concerned with explorations to inform futuredecisions. It was stressed by the facilitatorsthat most entries in a package generated bysuch a workshop would be expected to takethe form of recommendations to the respon-sible decision-taking authorities rather thanfirm commitments, as most of the participantsin the workshop had been specialised advisersrather than managers with authority to committheir agencies.The workshop was facilitated by a team

in which John Friend, as visiting consultant,was accompanied by Ana Maria Benaiges asrecorder and occasional co-facilitator, using theSTRAD software as a recording device. Oth-ers in the facilitation team were José Madrid,Head of the Ministry’s Strategic Analysis Unit,and John Foley of the Central University’s Plan-ning School. Issues of Spanish/English trans-lation were handled by this team in a flexibleand adaptive way, with little evident disruptionin the momentum of progress. The followingmonth, John Friend returned to Caracas withJonathan Rosenhead, and two further work-shop sessions were held with senior repres-entatives of key executive agencies. This time,Elisenda Vila and a Cuban consultant acted asco-facilitators.

209

9 Practicalities

TAKING STOCK

The last four chapters have completed asecond and fuller cycle around the four basicmodes of planning seen as a process ofstrategic choice. In contrast to the briefer,introductory tour made in Chapters 2 and 3,the emphasis has been on appropriate waysof drawing on the core concepts and basicmethods of the strategic choice approach,when adapting to different working situations.Although certain elaborations of these con-cepts and methods have been introduced, it isimportant to stress again that moves towardselaboration should not necessarily be seenas moves towards ‘better’ use of the stra-tegic choice approach. Simplification is usuallya virtue when working under pressure; and instriking a balance between simplification andelaboration the most useful working rule is tofavour simplification except where there areconvincing arguments to the contrary.At various points in Chapters 5–8, consider-

ation was given to some of the organisationalaspects of the strategic choice approach; tothe switching of the process between modes;and to the forms that themore visible productsmight take. Also, those aspects of techno-logy concerned with effective interaction andfacilitation in a group setting were consideredmore closely than in the earlier introductorytour of core concepts and methods. But theexposition of the last four chapters has stillfollowed the structure of the four modes in asequential way. It has to be recognised that,the more effective a working group becomes,the more the process will tend to movefreely and adaptively between modes through

a succession of rapid switching and loopingjudgements. So the offering of operationaladvice on the management of the strategicchoice approach cannot be organised accord-ing to the structure of the four modes alone.There is another level of practical advice whichrelates generally to the management of theapproach as a whole. It is this kind of advicewith which the present chapter is concerned.The starting point for the advice to be offered

here will be the general picture of the ori-entations of strategic choice presented at theend of Chapter 4 (Figure 34). That diagramused the framework of technology, organisa-tion, process and product to compare the stra-tegic choice approach with more conventionalapproaches that could be regarded as quiteappropriate in dealing with simpler problems.In successive columns, contrasts were drawnfirst at the level of general orientations, thenat the level of operational guidelines, then atthe levels of emphases in management choiceand in evaluation. This chapter will interpretthese last two aspects in still more concreteterms, reflecting the body of experience inthe application of strategic choice ideas thathave been built up through working with manydifferent people on a wide range of practicalproblems over a considerable period of years.First, this body of practical experience will bebriefly reviewed. Then, the main lessons to bedrawn from itwill be presented; first in terms ofthe emphases on selectivity and adaptivenesswhich distinguish the overall approach, andthen in terms of the four headings of techno-logy, organisation, process and product, con-sidering each in turn.

211


THE BASE OF EXPERIENCE

The advice that will be offered in this chapteris rooted primarily in the experiences of theauthors, working with many other people inmany different project settings, and the pic-ture that has gradually developed through thisexperience of what constitutes ‘good practice’in working with the strategic choice approach.This view of the current state of the art isone which is unlikely to stand still in the yearsahead, as further experience accumulates. Forthe range of applications to different fields ofdecision-making can be expected to extendand, with it, both the variety of people andorganisations involved, and the range of furtheradaptations made.However, our current view of good practice

is one which already has quite a broad basein terms of the range of planners, managers,specialists and lay policy-makers who havebecome exposed to the approach and playedsome part in its development. The people withwhom we ourselves have worked have comefrom several different countries and decision-making cultures; and the types of working rela-tionship through which we have collaboratedwith them have also been diverse. So, it will bea useful prelude to the advice of this chapter toindicate briefly what the range of these work-ing relationships has been and, broadly, whathas been learned through each type of experi-ence. For this purpose, three broad headingswill suffice – sponsored experiments, trainingevents and consultancy projects. Each typeof experience has complemented and rein-forced the others: but the order in which theywill be introduced here corresponds broadlyto the historical sequence in which they firstemerged.Much of the early development work on the

strategic choice approach took place throughthe medium of sponsored experiments, inwhich an organisation interested in the possib-ility of promoting newmethods of planning anddecision-making has been prepared to investin a programme of one or more experimentalapplications of the strategic choice approach

to ‘on line’ problems within its general areaof concern. Leading examples included the ini-tial LOGIMP experiment of 1970; the struc-ture plan project of the mid-seventies; theNetherlands project on environmental policyplans in the early eighties; and successive pro-jects in the states of São Paulo and Pernam-buco in Brazil. Each exercise was sponsorednot so much to help with specific decisionproblems, as to test the relevance of the stra-tegic choice approach to a wider class of prob-lems; yet success in helping with specificcases was quite central to this testing process.In each of the first three projects mentioned,between two and six ‘live’ planning problemswere pursued in parallel, with the authorsand other colleagues or associates acting asadvisers to the local teams responsible for pro-ducing recommendations on these selectedproblems. The different teams would all meettogether from time to time, to exchange exper-iences, review difficulties and provide mutualsupport – with representatives of the centralsponsoring body maintaining oversight of theprogramme as a whole.The overall style of working in these

sponsored experiments has been one of actionresearch, designed through negotiation withthe dual aims of discovery and practical ser-vice to users in view. In contrast to the moreclassical action research studies (Clark, 1976),there was the added dimension that the actionresearch team was engaged in collaborationwith people not just from a single organ-isational context, but from a set of two ormore local ‘host’ organisations together witha central sponsoring body, each with its ownexpectations of what the collaboration shouldachieve.Inevitably, each sponsored experiment has

involved elements of training in the strategicchoice approach – partly in the form of ‘onthe job’ training for the members of the localteams, and partly in the form of introduct-ory lectures on philosophy and method. Inaddition, however, from 1971 onwards, free-standing training events have been arrangedunder the auspices of a wide range of other

212

Practicalities

organisations, in various parts of theworld. Theduration of such an event has generally beenshort – ranging from as little as half a day to twoweeks at the most – because those participat-ing have generally been not full time studentsbut decision-makers and planners with manyother pressures on their time. The design for-mula for all but the shortest half-day events – atwo- or three-day duration being most typical –has been for lectures to be interspersed withwork in small groups on a realistic problemexercise. The aim is to move progressivelythrough the shaping, designing, comparing andchoosing modes, as in this book, with as muchattention as practicable within the time avail-able to the various opportunities for recyc-ling that can be explored as understandinggrows and circumstances change. The experi-ences of both sponsored experiments and con-sultancy projects have provided a rich sourceof case material for use in these trainingevents, allowing the various facets of planningunder pressure to be realistically reproduced.At the same time, the wide variety of man-agement contexts and positions representedby those participating in the various trainingevents has considerably enriched the processof feedback from practice through which stra-tegic choice ideas and methods have graduallyevolved.Involvement in direct consultancy projects,

where the primary aim has been to help witha specific planning or decision problem ratherthan to explore the relevance of the approachto some class of problems, has achievedmomentum as the experiential foundations ofstrategic choice have become more secure.In almost all cases where we or others asso-ciated with us have been involved as con-sultants, arrangements for direct collaborationwith people from client organisations havebeen treated as an important part of the projectdesign, and opportunities for interactive work-ing have been stressed.The scale of such projects has varied

from substantial engagements with nationalgovernments on national policy issues, tobriefer assignments for commercial, public and

voluntary organisations. Sometimes, too, briefexercises in mutual consultancy have beendesigned within short-term training events.Typically, an hour or two has been allowed forparticipants to talk about instances of specificproblemswhich are currently concerning themin their own management situations.While the various training and consultancy

experiences have provided opportunities toexperiment with many different styles in man-aging the process of strategic choice, theadvice that follows is based primarily on one‘prototypical’ style that has proved success-ful in many different settings and has beenwidely used both in sponsored experimentsand in consultancy projects. It is a style thatreflects the basic orientations and operationalguidelines of Chapter 4, in that it is builtaround a programme of interactive group ses-sions of limited duration in the kind of workingenvironment illustrated in the upper picture ofFigure 30.Within the group, the methods of strategic

choice are used as an ‘open technology’,starting from some initial state of knowledgeor uncertainty about the problem among theparticipants, that will gradually be revealed.The agenda is treated as flexible within thetime constraints of the working session orsequence of sessions that has been arranged.Little or no backroom preparation is expectedfor the first session – though priorities for back-room work can build up as a shared view ofissues and uncertainties develops within thegroup. Typically, a working session will covertwo or three hours either in a morning or anafternoon,with the same kinds of pressures ontime as in the more traditional committee set-ting depicted in Figure 6 – but with the import-ant exception that there is no expectation ofworking through a preconceived agenda fromstart to finish as the session proceeds.

SELECTIVITY AND ADAPTIVENESS:THE TOOLBOX ANALOGY

It is most important that the strategic choiceapproach is not seen as a mere technique.

213


The straightforward application of techniques,with some well-defined rules as to how andwhen they should be applied, has all the attrac-tion of simplicity and is easily taught in aclassroom setting. But such an approach hassevere limitations, unless the problems beingaddressed are ‘bent’ to fit it – something whichin practice happens rather more often thanit should. The strategic choice approach hasbeen developed in a way designed specific-ally to avoid this difficulty. This means thatthere is no one right way to use the approach.There can be no ‘correct’ sequence in which todo things, and no prescribed combinations toadopt – whether these be viewed in terms oftechnology, organisation, process or product.It is all a matter of judgement as to what isappropriate to the circumstances at that par-ticular time. Every problem is different. Eachmust be tackled on its merits.Use of the strategic choice approach has

been likened to the use of a toolbox. Theanalogy is extremely apt so long as it isremembered that in strategic choice the toolsare many and varied – including a wide rangeof concepts, modes, frameworks, techniques,activities and media.Imagine taking a toolbox to mend a car

engine. Having opened the bonnet and lookedat the engine, it is likely that a start has to bemade by undoing a nut. Reaching into the tool-box, one selects a spanner, takes it out andtries it on the nut. The first time, almost invari-ably, it does not fit. So one puts it back and triesanother. It may be that the nut is still stubborn,in which case one dips down into the toolboxand chooses another spanner, using now thetwo in combination. It may be that this combin-ation fails too. If so, they can be put back andanother combination tried; or perhaps somepenetrating oil is used in addition, thus usingthree items in combination; and so on, until itworks.As progress is made through the job in hand,

different situationswill emerge requiring differ-ent combinations of tools – and so on, until thejob is finished. Also, the next time it is neces-sary to do the same job, it will probably be

appropriate to use different tools, in differentcombinations and sequences. Imagine that thesame car engine requires the same mainten-ance three months later. It may now be easierto shift that nut, which was once so difficultthat several spanners had to be used to shift it;perhaps because it was greased before beingtightened. This time, the tools needed may bedifferent again. Perhaps a pair of pliers will beenough to get it undone – or it may be foundto be only finger tight in any case. Meanwhile,however, other things might have changed; abolt might have become rusted in, or brokenoff, or weakened in some way. So yet moredifferent tools, sequences and combinationsmight have to be tried.This analogy illustrates well the two key

ideas which govern all the operational adviceoffered in the sections which follow. The firstis the idea of selectivity. This follows inevit-ably from the wide variety of choice availablein the use of the strategic choice approach.Indeed, it is clearly implied in the toolbox ana-logy; there aremany toolswhich obviously can-not all be used on a specific task, even if theywere all appropriate – which is most unlikely.Even more inconceivable is the idea that theycould all be used at the same time. Thus it isnecessary to be selective; to choose carefullywhat to do and what not to do – and how, withwhom and when.The second idea is that of adaptiveness. This

follows from the feelings of uncertainty whichare endemic to working with complex issues –uncertainty about what the problem is (and isnot); what the alternative ways of dealing withit may be; how they rate one against the otherand, indeed, what to do about it. The idea thatone can always, unerringly, select a good wayof working in such a context is plainly unreal-istic. It may serve some purpose to try to pre-dict the course that the analysis should take;but to believe that it will be sensible to main-tain that course whatever happens in the pro-cess is similarly far-fetched. So, there can beonly one way of working; and that, paradoxic-ally, is to be ready to work in many ways – in

214

Practicalities

fact, to work adaptively via the explicit accept-ance of the learning process, part of which islearning how to conduct that process. One’scapacity to think about what to do; what not todo; how, when and in which context, is boundto evolve as one’s understanding develops. Itis only sensible, as long as time allows, to givethat evolutionary process full rein.Therefore, every opportunity has to be taken

to promote that evolution – to keep the learningprocess going in a smooth and adaptive way.Referring back to the description of the tool-box analogy, there was an important momentwhen the first spanner selected proved to bethewrong one. Thiswas not bad or inefficient –it was a vital step in the learning process. Beingmistaken in this way is important because somuch is learnt from it. Thus the second span-ner was more likely to be right because thechoice of it was so much better informed thanthe first. This is a classic example of learn-ing from one’s mistakes – a phenomenon wellknown in all cultures of the world, old and new.It leads to the idea of learning by doing. Theonly way to make mistakes is by doing some-thing – so it follows that the more one does,the more mistakes will be made, and the moreone will learn.But this does not mean that selectivity is

unimportant. Selections have to be made allthe time, in order to keep the process going,and there are times when making a mis-take can be costly. So it means that care mustbe taken in that selection; yet that too muchtime must not be wasted on it. Going backto the toolbox analogy again, a lot of thoughtcould be given to the choice of which tool touse at any time. It is easy to hesitate whenthere are somany alternatives – not only whichsize of spanner to use, but also which sortof spanner and, indeed, whether a spannerwas an appropriate tool in the first place. Theimportant thing is not to agonise too long overthe choice ofwhich to use; but to try one, draw-ing as much as possible on past experience,with an open mind and in a spirit of learning.As a rough and ready rule, it is worth bearingin mind that the time spent thinking about it

should not exceed the time it would take justto try one out.Naturally, experience is a vital help in

selectivity. This in itself can lead to the ideaof each user having his or her own individualrepertoire. This is the sub-set of all the avail-able tools which that person tends to usemore freely and frequently than others – aform of behaviour which is systematically rein-forced because it is the most frequently usedtools that tend to stay at the top of the tool-box, thus being the first for consideration thenext time. Tools from lower in the toolboxonly get used when the earlier ones do notwork so well, or when the specific task is verywell defined. Different people’s personal rep-ertoires are likely to have a number of toolsin common, and these may be expected tobe those with the more general application;the equivalent of hammers, pliers and screw-drivers, though even these may be of varioussizes and types. For users of strategic choice,among these common tools would probablybe the basic concepts of AIDA; though evenhere such core concepts as the option bar andthe decision scheme can be represented andused in different ways, as earlier chapters haveshown.So, even the experienced consultant or prac-

titioner in strategic choice will have a personalrepertoire, which will be similar but probablyby no means identical as between individu-als. There will be differences in their personalstyles as well as in their work experiencesin collaborative settings. The two authors, forexample, recognise some clear personal differ-ences in the emphases they give to particularconcepts and methods in their respective rep-ertoires.Most people already have personal tool-

boxes of their own and, indeed, their own rep-ertoires with which they are likely to feel quitecomfortable. What is more, they are probablyreluctant to abandon these just to take upwhatcould seem to be a self-contained box of toolsfor strategic choice, such as that presentedhere. In most cases, this is not necessary.The strategic choice approach is intended to

215


be open and flexible – and most of its toolscan be used in combination with other waysof doing things. Indeed, mention has alreadybeen made of certain other tools of evaluationand design at particular points in the precedingchapters. Further discussion of these relation-ships will be found in Chapters 11 and 12.

CHOOSING TOOLS UNDERPRESSURE

The most distinctive and significant featureof the strategic choice approach is that it ishas been developed directly out of practice,working with planners and decision-makerswho are usually working under pressure. Con-sequently, it is under these conditions that stra-tegic choice is most useful, and it is gener-ally where deeper and more lengthy analysisis relevant that some of the other approachescome into their own.In developing the strategic choice approach,

it has been found helpful to distinguishbetween contrasting ways of handling workunder pressure – especially pressure causedby the shortage of time. These are what isoften called the quick and dirty way, and whatmay be called the fast and effective way. Tak-ing up the toolbox analogy again, they canboth be likened to types of screwdriver. Thequick and dirty way can be likened to the‘Brummigem’ screwdriver, whereas the fastand effective way can be likened to the ‘Yan-kee’ screwdriver. The ‘Brummigem’ (which isa corruption of the name Birmingham) is, infact, a heavy hammer. Although designed forother tasks, such as banging in nails, it can be avery quick way of driving in screws. Of course,there tends to be a loss of effectiveness inthe result, although the appearancemay some-times be satisfactory. The ‘Yankee’ (actuallya trade mark) is a specially designed ‘plunge’type screwdriver which, by means of a helicalthread inside the handle, turns the screw auto-matically as the tool is pushed on. It is veryfast and, indeed, can be very effective. But thetrouble is that it is not suitable for all situations

and, unlike the ‘Brummigem’, it cannot be eas-ily used for anything else. Thus, it can be seenthat there is no clear-cut recipe for the usein practice of the strategic choice approach.Selectivity and adaptiveness are both essen-tial and, by definition, require that choices bemade. And, as was implied in the toolboxanalogy, these choices mean balancing ideasabout effectiveness against ideas about theexpenditure of effort, time and resources –whether the balance is struck in a conscious orsub-conscious way. Here it is worth mention-ing the Sutton Principle, which many users ofstrategic choice have found helpful as a simple,light-hearted way of reminding themselves ofthe issues involved.The Sutton Principle is named after

Willie Sutton, who was an infamous Americanbank robber. One day, when he eventuallybecame available for interview, Willie Suttonwas asked why he robbed banks. To thishe came back with the now immortal reply:‘Because that is where the money is!’ Hisreasoning was very simple. Given a limitedamount of time and effort (in Willie’s casehis own), these resources should be directedwhere the rewards are likely to be greatest.Without examining Willie Sutton’s logic too

deeply, two important lessons can be drawn.The first is that if one cannot deal witheverything adequately then, rather than dealwith all things inadequately, it is better to dealadequately with some things only. The secondis that, in choosing which things to do, the pro-cess should always be steered in the directionwhere it is believed the greatest progress canbe made. So it is also with deciding how toselect tools from a toolbox – always remem-bering that in the strategic choice approachthe toolbox includes modes, concepts, frame-works, media of communication and waysof working, as well as analytical techniques.The costs are generally the demands whichthe use of each tool places on the availableresources and, in the management of a pro-cess of strategic choice, the resources in ques-tion tend to take many forms. They include

216

Practicalities

elapsed time; the skills and experience of rele-vant people; access to relevant information;and access to the capacity to process suchinformation by electronic or other means. Theeffectiveness of alternative ways of proceed-ing has to be weighed against the demandsthey make on resources of all these kinds and,as such resources tend to be limited, theycan become constraining, limiting the rangeof choice. Or sometimes, of course, they canbe bought in from elsewhere. Money has notbeen mentioned yet as a resource because itis rare for the selection of a tool from the tool-box to carry a direct financial cost. Where suchcosts do arise is in trying to make up for short-falls in resource supply. This means giving con-sideration to improvement, support or, in somecases, replacement of the resources alreadyavailable, in such forms as:

• more andmore effective people (e.g. trainingand consultants);

• more and more relevant information (e.g.surveys and research);

• better and/or faster technological support(e.g. computers and software).

However, this is always assuming that there isenough calendar-type time available – and thisis often the most inflexible constraint of all.The effectiveness side of the resource

effectiveness balance is difficult to judge inadvance because it is so intangible and uncer-tain. It is concerned essentially with the poten-tial of the tool; in particular, the potentialprogress it might generate in terms of thosebroad and comparatively subtle dimensions ofevaluation which were indicated in the finalcolumn of Figure 34. Progress in the directionsof sharing, synergy, understanding and confid-ence can be difficult enough to assess in ret-rospect, let alone to judge with any clarity inadvance.So far, this chapter has discussed the prac-

ticalities of working with the strategic choicetoolbox in rather general terms. The aim hasbeen to explain more fully the broad man-agement emphasis on selectivity and adapt-iveness which appeared in the top row of

Figure 34, and also to develop further theidea of effectiveness as an evaluative prin-ciple. To move to a more concrete level inthis discussion of the practicalities of strategicchoice, it will be necessary now to consider theoverall approach in terms of its four compon-ents: technology, organisation, process andproducts.

TECHNOLOGY: MANAGING THEOPEN TECHNOLOGY IN PRACTICE

In Chapter 4, it was argued that the technologyof strategic choice was designed above all asan open technology; as a means of assistingand encouraging effective communication andinteraction within a working group and not justas an aid to backroom analysis by experts. It isbecause of this emphasis on sharing betweenparticipants with diverse perspectives to con-tribute, that the technology of strategic choicecan be considered an appropriate technologyfor group working on issues the very shape ofwhich may be both complex and confused.The emphasis on an open technology

was interpreted in Chapter 4 into an oper-ational guideline of a focus on decisions. Thisguideline finds more specific expression inthe set of core concepts to do with decisionareas, decision options, decision schemes andthe various other linked concepts which wereintroduced in relation to the work of the com-paring and the choosing modes. In addition,an impression was presented of the kind ofinteractive work setting in which the spirit ofan open, decision-centred technology can beachieved in practice (Figure 30). It was arguedthat choices about the use of rooms, furnitureand equipment were just as much a part ofthe management of the strategic choice tech-nology as were matters to do with the applic-ation of specific concepts and techniques. Sothere are various practical matters to do withthese physical aspects of technological choiceto be discussed in this section; after which,some further points will be made about the

217


kinds of social technology that can be helpfulin stimulating productive forms of interactionwithin a working group.Some of the advice to be offered on

effective use of walls, paper and pens mayseem quite trivial and mundane. However,long experience in using the strategic choiceapproach with groups does indicate that thequality of group communication and interactioncan be quite severely impaired by failure topay heed to such considerations, especially inthe early stages of group working, when pat-terns for future interaction are being set. Itis important, firstly, to choose a room whichis not so small in relation to the number ofparticipants that their freedom of movementis restricted, constraining them to sit in fixedpositions in relation to each other, as theywould in the familiar committee setting. Forthe same reason, it is important to avoid cre-ating a situation in which a table becomes thecentrepiece. Indeed, a first step in preparing aroom is usually to push any tables to one side,where they can still be used as required, fordocuments, refreshments, materials or note-taking, but do not create barriers between theparticipants as they move around. It is a matterof comfort and convenience to provide enoughchairs for participants to be able to sit downwhenever they desire. However, they are likelyto spend much of their time standing and mov-ing around when the process of interactiveworking is in full swing; and at such timesthe focus of attention will normally be on theever-changing pictures of decisions and theirrelationships which are building up around thewalls.Fixtures such as blackboards or whiteboards

do not allow this to happen. This is simplybecause the information they contain has to be‘wiped’ once they are full; this is not consist-ent with effective working in a cyclic process,which requires the ability, at anymoment, to goback to a previous point and to pick up whereone left off, Nor do ‘flip chart’ boards on easelsby themselves serve the purpose, because oftheir limited surface area.

It is important in strategic choice to display asmuch of the work as possible at any one time,thus encouraging instantaneous, ad hoc loop-ing to take place. Participants should be ableto compare the focus of the moment with pre-vious work, skipping backwards and forwardsthrough the process as freely as possible.So, walls are important; and the interact-

ive process can be impeded quite seriously ifpeople are working in a room where the wallspace is broken up by too many windows, orprojecting panels, paintings or shelves; whereaccess to the available wall space is restrictedby too much furniture; or where the walls areuneven or richly decorated in a style which isdesigned to impress. This means that conven-tional boardrooms or council chambers do notmake an ideal environment for this kind of inter-active working; for tensions can often arisewith other people who wish to preserve suchrooms as settings formore orthodoxmeetings.Such tensions can usually be avoided by choos-ing to work in a more sparsely furnished andplainly decorated space. To work effectivelyon walls, plentiful supplies of paper of gener-ous dimension are required. Continuous rollsof paper have sometimes been used; but padsof flip chart sheets of international A1 size havebeen found to be more flexible when it comesto rearranging and consolidating information asthe work proceeds.The more the sheets of paper accumulate

around the walls of the room, the more import-ant the practicalities of paper management

become. Sheets of paper can be affixed to thewalls either by masking tape or by a reusableputty-like substance such as ‘Blu-tack’; pinsonly provide an acceptable substitute wherewalls are covered with cork or some sim-ilar soft surface. Blu-tack has been found toprovide high flexibility in removing and repos-itioning sheets of paper quickly and, if properlyhandled, leaves no permanent marks on mosttypes of surface. It is generally quite sufficientto fix each sheet by the top two corners only,placing the Blu-tack a centimetre or two in fromthe corner; each new sheet should be pos-itioned with its upper edge a little above head

218

Practicalities

height, for purposes of writing and of visibil-ity. As wall space at this level becomes morescarce, sheets recording past work can eitherbe moved down to a lower level or overlaid byothers. However, in doing so, it is importantto avoid hanging paper on paper – for the add-itional weight is then likely to pull the originalpaper away from the wall, usually to the dis-traction of all concerned.It is not always possible to keep the papers

on the wall neatly arranged – evenly spacedand vertical. However, well-ordered walls canhelp a group overcome the strangeness ofworking in this way and give a sense of organ-ised purpose. Architectural features, such asjoints in walls, door frames and even wallpaperpatterns can be used for guidance – not onlywith hanging the papers but also with settingup lines, grids and matrices on them. Paperover-printed with a feint grid is also helpful inthis respect. Large sheets of paper with pre-drawn frameworks, such as:

• compatibility and consistency matrices;• concentric circles for uncertainty graphs;• advantage comparison charts;• commitment package frameworks;

not only make the work look more organised,they also promote faster, more effectiveworking. They can be copied quite easily by thesame processes that architects and engineersuse. In this way, one can have a large enoughsupply not to have to worry about runningout.The paper used need not be of high qual-

ity, so long as it is not too porous when putto the test of being written on with colouredmarker pens. The pens themselves should besufficient in number for participants not to haveto compete with each other when they feelmoved to add something to the ‘maps’ of prob-lems that are building up around the walls;and they should be sufficiently varied in hue toallow colour differentiation to be used freely asan aid to communication.Experience has shown that some brands of

marker pens are much more reliable than oth-ers in terms of ink flow. Those using water-

based coloured ink have been found to providethe best service; and those with a wedge-shaped felt tip, 3–5mm wide, allow a varietyof line thicknesses, and clear visibility in mostinteractive environments.Many experiences ingroup working have been impoverished by reli-ance on pens which can only make thin, palelines; which squeak in a jarring way; or whichare prone to dry up at important moments.However, this is not to say that pale coloursare never useful. In particular, a barely visibleyellow makes a good colour with which topre-enter frameworks for guidance in addingmore substantive information; for example, itcan be used to draw up the rows and columnsof a grid or matrix, or the concentric rings ofan uncertainty graph. In drawing up an optiongraph for a selected problem focus, one use-ful ‘trick’ is first to draw a faint yellow ringintersecting each of the decision areas, thento space out the options within each decisionarea at intervals around this ring.There are also more general points of cal-

ligraphy and style which regular users of flipcharts have tried to cultivate – points aboutholding the pen, about drawing freehand linesand circles, about positioning of information ona new sheet of paper in a way that anticipatesthe further information that is still to come.Practice and discussion can lead to gradualimprovements in these directions; but it isimportant above all that concern with the qual-ity of the graphics should not be so pervas-ive that participants who feel they have fewgraphic skills should feel inhibited from takingpart.It is not to be regarded as a sign of failure to

makemistakes, or to revise earlier views; it is avital part of the learning process. So it is import-ant to make corrections quickly and move on,rather than go to pains to erase any signs thata change has been made. Trivial mistakes canbe covered up by self-adhesive labels; but lesstrivial changes or corrections are better left aspart of the record on the wall, in so far as theyrepresent a significant step in the learning pro-cess which people may wish later to recall.A decision graph or similar picture which has

219


become too untidy and confused can alwaysbe redrawn and the new version placed over oralongside the old, treating the original as a sec-ondary record only. Every now and again, too,it may be sensible for the participants to pauseand rearrange the total set of flip charts aroundthe walls, perhaps grouping together those towhich they expect to refer most frequently inthe course of subsequent work.Turning to what has been called the social

technology of strategic choice, there are fur-ther management choices which arise whenworking with a group of participants in whichthere is little or no prior experience of this styleof interactive work. It is appropriate for the per-son or persons acting in a facilitating role totake the lead in the use of pen and paper in theearly stages, if only to demonstrate the generalmethod of working and set a style for others tofollow. Yet, in training courses which includesmall group exercises, it is usually found thatmost people start working on the walls spon-taneously enough, provided they have had atleast a brief demonstration of the art of papermanagement beforehand.How quickly and energetically moves are

made to open up the process of writing on thewalls to wider participation is a matter whichhas to be judged according to the situation andthe sense of how well a group momentum isdeveloping without active steps to promote it.Sometimes it will be easy to encourage peopleto start using the walls, paper and pens in aninteractive way; sometimes, they will be moreinclined at first to sit and talk and leave anywriting to the facilitator, unless they are expli-citly offered a pen and encouraged to use it. Ina brief exercise with inexperienced users withonly an hour or two to run, it may be accept-able that the facilitator continues to play themain recording role, concentrating on captur-ing what other people have to say; but thelonger the time available, the more feasible itbecomes to build up pictures around the wallsin which the draughtsmanship of many parti-cipants has been merged, so that the sense ofcommon ownership can take on a visible realityof its own.

Mention has been made in earlier chaptersof the advantages of punctuating interactivegroup sessions with brief opportunities forindividuals to work on their own. It can helpto ask participants to write down their ownideas about relevant decision areas, uncer-tainty areas or other elements in their ownwords, using cards or small sheets of paper.These individual contributions can then becompared and merged in such a way that eachparticipant feels he or she has participated dir-ectly in building up the shared pictures that aretaking shape around the walls. The use of flooror table space to sort and regroup the variousindividual contributions can itself help to buildup a spirit of interactive working.1

There is a related technique, which can beespecially helpful in bringing invisible productsto a conscious level. It involves the use ofsmall circular-coloured stickers (about 2 cmdia-meter), for the structuring of informal interact-ive evaluation. For example, if it is thought agood idea to review the growth of joint com-mitment in the grouping, each participant canbe asked to place one sticker on a scale fromvery good to very bad which has been drawnon the paper. Small stickers can also be usedfor identifying the most important items in alist – such as comparison areas or uncertaintyareas. The resulting pattern is then used as thebasis of a creative discussion of the subject,often leading to consensus in the group.It can pay to introduce brief opportunities

for individual working from an early stage ina group process, with further such opportun-ities whenever the momentum of interaction

1 Working on paper on the wall is not exclusively a stra-tegic choice style. There are others who adopt a similarstyle of working (Doyle and Straus, 1976); but very fewhave a structured approach to the use of the papersthemselves. One such approach, which has much tooffer, originated in Germany in the early seventies. Itis called Metaplan, and is based on the use of shapedcoloured cards (oval, circular, rectangular) on which theparticipants write, before pinning them onto specialscreens (Schnelle, 1973). There are some close par-allels with the use of cards in strategic choice groupprocesses, which started about the same time.

220

Practicalities

appears to be flagging or dominated by a sub-group. In order to be ready when such oppor-tunities occur, it is recommended that a supplyof cards and stickers be on hand at all times –in addition to the pens and the paper. It isalways possible to improvise by cutting or tear-ing sheets of paper into smaller pieces andmarking themwith coloured blobs; but a supplyof cards and stickers in several sizes, shapesand colours gives the process an air of pro-fessionalism which can be reassuring to someparticipants.In Figure 83, all these practical points about

themanagement of the technology of strategicchoice are drawn together, in a form intendedfor quick reference, bothwhen preparing for aninteractive group session and while reviewingprogress subsequently. The key distinctionsmade are between choice of the physicalsetting, which usually has to be negotiatedin advance; the provision of adequate sup-plies both for interactive participation and indi-vidual working, not forgetting refreshments;and, finally, the requisite techniques and skillsto allow these resources to be used to goodeffect once the group process is under way.

ORGANISATION: MANAGINGINTERACTIVE PARTICIPATIONIN PRACTICE

It is neither practicable nor useful to attemptto draw a firm line to define where the man-agement of technology ends and the manage-ment of organisation begins. There are somefurther points to be made about ways of man-aging the dynamics of a working group, andabout opportunities for forming sub-groupsand subsequently bringing them together.These points will be touched on briefly here,leaving some deeper points about the under-standing of group dynamics to be opened upin later chapters. However, it is most import-ant to recognise that any working group has tobe seen as embedded in a wider organisationalenvironment in which many other people mayhave a part to play. So important practical ques-tions have to be asked about what levels and

forms of participation and of interaction will beappropriate for other people beyond the con-fines of the present working group; and thesequestions too will be dealt with in this section.It is quite easy to think of good reasons for

involvingmany and various people interactivelyin the process of decision-making. But it is notso easy to make a case for the meetings thatthis entails. It is regularly argued that meet-ings are a waste of time; that there are toomany of them; and that the decisions made asa result cannot be proven to be better. This isentirely justified and will continue to be so, aslong as meetings are organised in the conven-tional manner. Therefore, in using the strategicchoice approach, changes have to be made.Already, in the last section, ways in whichthe conduct of meetings can be changed havebeen dealt with at some length. In this sectionthe concern shifts to questions about how themeetings should be set up, especially with ref-erence to who should meet with whom.The form of association will vary from per-

son to person and from time to time. Of allthe people who might be involved, only somemay wish to be; others may be content to beconsulted occasionally, in a reactive rather thanan interactive way. It has to be rememberedthat interactive working, for all its advantages,is demanding in terms of time and energy; andpeople in practice must always find ways ofdistributing their personal resources of timeand energy effectively, depending on the over-all range of responsibilities they carry.The general picture of three kinds of uncer-

tainty was used in Chapter 4 (Figure 31) toindicate some of the main ways of initiat-ing connections with other people, wheneverit was found difficult to deal with a currentdecision problem in the present working con-text – whether it be that of individual work oran interactive working group. However, thereare various types of relationships which canbe distinguished,with different implications forthe grouping and linking decisions that have tobe made. They reflect different roles peoplecan play in relation to an overall decision pro-cess, as contrasted to the roles theymight play

221

FIGURE

83Management Checklist: Technology

Practicalities

in relation to the work of a particular group.It is helpful to think of people in seven proto-typical roles in relation to the decision-makingprocess:

• those who are charged with working on theproject in the sense of performing all thetechnical tasks involved in shaping, design-ing, comparing and choosing (as described inChapters 5–8);

• those who undertake the continuous pro-

cess management of the project, lookingafter the day-to-day co-ordination of all thevarious activities and groupings;

• those who are accountable for the decisionsto be taken in a broadly political way;

• those who are directly responsible for guid-ing the conduct of the decision-making pro-cess, at a managerial or senior professionallevel;

• those to whom periodic reference should bemade because they have roles in other fieldsof decision-making which are instrumental inthis case;

• thosewho fill a representative role in relationto specific interests which may be affectedby the decisions;

• those others who are stakeholders in thesense that they will be directly impacted bythe decisions.

It is important to recognise that, in anygiven decision process, the exact way in whichpeople in such roles should participate is adesign question in its own right. This meansthat some of the roles – though not necessarilyall of them – will be played by groups or semi-formal groupings. There will often be overlap-ping memberships and/or vacancies; and, insome cases, sub-divisions may be recognised.For example, it is common, in public agen-cies, to distinguish between internal and inter-agency reference groups.Nevertheless, in discussing these roles and

relationships, it is useful to take as a referencepoint the task of a particular working group –a group which, in the prototypical decisionsetting being considered in this chapter, may

be constituted as a more or less formal pro-ject group charged, like the South Side Work-ing Party, with working on an issue of someimportance over a period of several months.However, most of the points to be made hereare also relevant to briefer, more informal exer-cises. The typical relationships between thesegroupings, one to the other and all to the work-ing group, are illustrated in Figure 84. Broaddistinctions are superimposed between thosegroupings which have a role to play in thetechnical domain and those with a role morein the political arena. In addition, a distinc-tion is drawn between groupings which areinternal and those which are external to themain locus of responsibility for the decisionsbeing addressed.There are advantages to be gained by keep-

ing the working group as compact as possiblewithout losing the representation of importantinterests, especially with respect to the ease ofcalling the group together often at short notice.Therefore it has been found helpful to considertwo sorts of participant:

• Regular participant : responsible for regularsubstantive input and team member func-tions;

• Ad Hoc participant : responsible for a spe-cific area of specialised substantive input andgeneral team functions when present.

In the typical project situation, the role ofmaintaining continuity is usually handled by asmall sub-set of all the people who form themain working group. The task of this processmanagement grouping, or core group, is to pro-mote the effectiveness of the wider process inwhich members of all other groupings shownin Figure 84may at different times be involved.It is a small group – generally two or threepeople – because it must have the capacityfor extremely flexible behaviour, especially inits arrangements for meetings. It normallyembraces the four specific intra-group roles:

• Leader : responsible for thesubstantivework,controlling the quality and quantity of theanalysis and information input;

223

FIGURE

84Organisational Responsibilities in Strategic Choice

Practicalities

• Co-ordinator : responsible for process ratherthan substance, making sure that the rightpeople are at the right place at the right time;

• Facilitator : responsible for internal grouprelations, guiding interactive sessions andhelping with grouping and linking;

• Recorder : responsible for the visible results,co-ordinating the continuous output from theprocess and its distribution in appropriateforms.

Process management is sometimes seen asa function which can ultimately be handled byone person. When this occurs, that person isoften described as the project manager. Unfor-tunately, neat as this arrangement may seem,all four roles are very difficult to play together.In particular the facilitator role sits uncomfort-ably with the other three. Also, it is often thecase that the co-ordinator and facilitator havelittle direct knowledge of the specific problemsbeing addressed. In such a case it is import-ant to recognise the vital contribution whichthe leader can make. Thus, there are distinctadvantages when such a project managementfunction is given over to a small group. Whileany one person can take on any one role, thereare a limited number of combinations of roleswhich any one person can play. These are thecombinations of:

• leader and recorder (product manager);• co-ordinator and recorder (secretariat);• co-ordinator and facilitator (process man-ager).

Another issue which can arise in the choiceof core group membership concerns the intro-duction of external consultants. While a cap-able process consultant should be able totake on any of these functions or roles, thereare some to which they are better suitedthan others. For example, in many cases thefacilitator role can be better performed bysomeone from outside the immediate organ-isational framework. Basically, there are tworeasons, in addition to the extra manpower,why help of this sort can be useful. The firstis that strategic choice skills may be lacking

and need to be developed in the team; andthe second, where several organisations areinvolved, is that an independent view may berequired. In either case, it is important that theexternal consultant is given a counterpart fromwithin the client group, to provide the detailedknowledge and experience of the problems,the people involved and the culture withinwhich they work.Next to be described will be the set of

people who are formally accountable for thedecisions with which the working group isprimarily concerned. These can be thought ofas the decision-takers in relation to a wider,more diffuse set of decision-makers, if the dis-tinction is made between decision-making asa continuous process and decision-taking assomething that happens at comparatively raremoments when formal authorisation of actionsor policies is required. In the classical manage-ment hierarchy, decisions are taken at mostlevels throughout the process; but there isoften a level of importance at which author-isation by some accountable group, such as aBoard or Council, is required. In the public sec-tor as opposed to commercial organisations,the stakeholders are usually more diffuse, sothe principle of group accountability is rathermore widespread. Sometimes, accountabilityis vested in joint organisations – joint commit-tees or ‘standing conferences’. In general, themore complex the problem, the more complexwill be the politics of accountability; so a work-ing group may have to relate to more than oneaccountable grouping in some cases.There have been a few applications of the

strategic choice approach where decision-takers in such accountable roles have becomeinvolved in interactive working with membersof a working group.2 This can bring powerful

2 Members of the accountable grouping tend to find sucha role shift easiest through involvement in the respon-sible grouping, and there is extensive experience ofthis. However, there have been some examples of thedirect involvement of local politicians in the workinggroup. This occurred during preparation of the StructurePlan for the County of Hereford and Worcester in Eng-land, and also in the preparation of local environmentalpolicy plans in the Netherlands.

225


benefits both in terms of the level of aware-ness with which formal decisions are takenand in terms of the direct input of politicalconsiderations into the work of the workinggroup. However, it has to be recognised thatsenior decision-takers – whether appointed orelected – are often subject to multiple andsevere pressures on their time. Their inputscan be particularly significant in the work of theshaping and choosing modes and, given lim-ited opportunities to involve them in interactiveworking, it is important that their involvementshould be focused in these areas.Often, a link between a working group

and a set of accountable decision-takers ismade by means of a project steering group.Meeting more or less regularly, guidance isprovided on the directions and priorities ofthe work, progress is reviewed and any dif-ficulties discussed, especially so far as thehandling of external relationships is concerned.In multi-accountable decision settings, sucha responsible grouping will often have beendeliberately designed to bring together repres-entatives of different organisations or depart-ments involved in sponsorship of the projectwork. The membership can be relatively small;leading members of the project group maybe directly involved either as members orin a reporting capacity; and there should beopportunities between scheduled meetings towork through informal links with individualmembers.So a great deal of informal negotiation about

difficult or contentious matters can take placethrough the channels offered by the respon-sible grouping, both during and betweenmeet-ings; and there will be opportunities here touse various aspects of the strategic choiceapproach. It is also through this set of rela-tionships that members of the working groupcan retain the sanction of their departmentalheads to continue acting on their behalf, andthat changes in themembership of theworkinggroup can be discussed and approved. So theresponsible grouping offers a channel throughwhich the highly cyclic, adaptive work processof the working group can be adjusted to the

more formally structured procedures and polit-ical realities of the wider organisational envir-onment. It is, therefore, important to allowthe time and resources for this element ofthe wider participative process to work in aneffective way.What is referred to in the prototypical work-

ing context as a reference grouping consistsof delegates or representatives of all thoseother organisations whose co-operation will beimportant if proposals arising from the workof the project group are to be put into effect.Where the project group’s task is to producea relatively generalised ‘plan’ or set of policies,the application of which depends on actionsof many other agencies, it is important thatthere should not only be co-ordination dur-ing some eventual ‘implementation’ stage; arealistic framework for pre-coordination is alsorecommended, by which these other organ-isations or departments can have the oppor-tunity to participate at an appropriate levelthroughout.Even in comparatively simple contexts of

inter-agency or inter-departmental working, anelement of pre-coordination can be importantso that duplication, omission or contradictionin the provision of services can be anticip-ated and avoided. However, when workingon broader problems of strategic choice, pre-coordination can involve more complex issuesof adjustment of values and perceptions. Inlarger-scale exercises of policy planning it issometimes appropriate to invite representat-ives of all related agencies to participate inoccasional seminars. In these, some of thespirit of interactive working can be generated –even though, given time pressures, itmight notbe feasible to introduce the strategic choicemethods in an explicit way.There are two further types of grouping

which it is important not to overlook; bothfall into the external political arena. They arethe representative grouping through which thepolitical interests of specific sections of thepopulation, such as car owners, conserva-tionists or small businesses, are protectedand promoted; and the wider stakeholder

226

Practicalities

grouping of those others directly impacted bythe decisions, such as residents, consumers,or employees. They involve those who areconcerned directly from time to time throughexercises in direct public or employee partici-pation.Conventional attempts at a more general

sort of participation, involving ‘the public atlarge’, tend to end up being conducted withmembers of these groupings. Unfortunately,because such exercises are aimed at a broaderaudience, they most often take the form ofseeking reactions to well-formulated propos-als, which it is too late to influence in anyfundamental way. A more sensitive appraisalof which people in which roles are beingaddressed – and are willing and able to con-tribute – is the key to selective design of par-ticipative frameworks in which a high level ofcreative feedback can be realised.So the way in which all these types of rela-

tionships evolve through time forms part ofthe broad concern withmanagement of organ-

isation in a process of strategic choice. Thisapplies not only in the prototypical projectgroup situation that has been discussed herebut also in other more modest working situ-ations. Sometimes, in such cases, it can bemore realistic to relate to those in other rolesthrough more familiar patterns of bilateral con-tact and negotiation. So this is where pat-terns of linking through personal networks canplay an important part in supplementing anygrouping arrangements of a more formal kind.The extent to which strategic choice ideascan be used to guide such interactions willdepend on the judgement of the user andthe extent to which support has already beenestablished.The larger a working group, the more dif-

ficult it becomes to sustain a spirit of inter-active participation through time. To makeprogress in such circumstances, it is usefulat times to suggest that the group breaksinto smaller sub-groups to pursue comple-mentary aspects of the overall task. Thereare many possible grounds on which such

a sub-division of work can be made. Con-ventionally, work is sub-divided by disciplines,professions or organisational units; but, oncesome shaping of problems and their impli-cations has been recorded on the walls ofthe room, other possibilities become avail-able. The conceptual frameworks of strategicchoice provide a rich source of logical waysfor defining sub-groups. The four differentmodes of work taken individually, or in loopedpairs, provide one useful division. And thereare many other possibilities such as differ-ent problem foci, separate working shortlists,or commitment packages based on alternat-ive orientations. Some such tasks for sub-groups may require considerable time, and soshould be scheduled to be done in the peri-ods between the main interactive sessions ofthe working group. However, there are manyways in which strategic choice structures canbe used to identify briefer tasks for sub-groupswithin a working session. Examples of suchtasks include designing optionswithin decisionareas separately; assessing alternatives byindividual comparison areas; and exploring dif-ferent areas of uncertainty. While such div-isions of labour can be most helpful, it isimportant:

• to use the structure of the analysis as thebasis of the division;

• to set, and keep to, a strict time limit;• to allow plenty of time for the sharing of res-ults afterwards;

• to use the structure of the analysis to inte-grate the various contributions.

Of course, in suggesting which people mightbe assigned to which sub-groups, the skilledfacilitator should be able to take into accountwhich participants work well – or not so well –with each other, and which individuals can belooked to in order to provide stimulus and lead-ership for the more reticent members of thegroup.Then, if it is intended that there should be

not just one working session but a series ofsessions – whether concentrated within a fewsuccessive days or spread out over a longer

227

FIGURE

85Management Checklist: Organisation

Practicalities

period – it is likely that the composition of thegroup will alter at least a little from sessionto session, whether by design or because ofrestricted availability on the part of some par-ticipants. Such changes can, of course, affectthe momentum and patterns of interactionwithin the group. This means that there isa balance to be struck between a desire forcontinuity of membership, with the advantagethat everyone is working from the same basisof shared group experience; and a desire tokeep the membership open, so that new parti-cipants can join and others drop away.This desire to keep the group boundaries

permeable can be especially important whereideas about the shape of the problem are ini-tially unclear. Where this is so, the focus ofattention may shift in unexpected ways, sug-gesting that others should be involved whowere not seen as relevant earlier on. But itis important to keep in mind the point thatextending the boundaries of a working groupis not the only way of extending the processesof communication and interaction: and thisis where a whole range of possible ways ofgrouping and linking can be considered. For anycomplex problemwhich impinges onmany dif-ferent organisational interests, it is only to beexpected that participation will also becomea complex matter, with many people becom-ing associated with the process in one way oranother.Thus, the whole task ofmanaging the organ-

isation for strategic choice can be seen as aprocess of dynamic grouping and linking, inwhich there are many variables. It is not onlya question of which people in which decision-making roles should be active, but of how fre-quently; in which combinations; and in whatrelationship they should be to each other, andthe current decision situation. A summary ofthe considerations which arise in these areasof choice is put forward in Figure 85. This isintended as a guide for those faced with thetask in practice – probably in the core grouprole. The key emphasis in evaluation of suchorganisational choices will be not so much on

efficiency and control, as in a classical hier-archical structure as on the creation of synergywhich can be realised through the orientationtowards interactive participation.

PROCESS: MANAGING THELEARNING PROCESS IN PRACTICE

In the process aspect of the A-TOPP frame-work, the practical choices which people faceare primarily to do with the effective use oftime. Choices about the use of time have tobe faced continuously during an interactiveworking session; and these have already beendescribed at the end of Chapters 5, 6, 7 and 8,in terms of the various switching and loop-ing judgements that have to be made in mov-ing from one mode of work to another. Thissection will offer some further practical adviceon how this switching and looping processcan be managed within an interactive workingsession.However, in a longer-term project, where

there can be a commitment to a series ofsuch working sessions over the project period,there are also important choices to be con-sidered about the management of time in theintervals between one working session andanother. During these periods, there will usu-ally be backroom and investigative work to bedone by at least some of the participants, indi-vidually or in sub-groups, and there will oftenalso be various activities of a consultative orco-ordinative nature involving other people.There are two interrelated tasks which are

often overlooked, but which are of crucial sig-nificance – especially in situations where thereare conflicting interests. These are the need forparticipants to:

• report progress made in the working groupto their home organisation;

• and gather feedback as input to the groupprocess in return.

If these activities are neglected, or given toolittle time, the result can be, at worst, that thework loses all credibility; or, at best, that thereis a withdrawal of the backing that is necessary

229


if the recommendations are going to be carriedout.Some of these tasks may have emerged

directly from the activities of the workinggroup, as ways of dealing with importantareas of uncertainty; while some of them mayhave been explicitly recorded in the format ofa commitment package. But of course, noteverything that happens between sessionsflows from the work done in this interactivesetting. All participants will have competingclaims on their time between sessions, somemore than others. They may have to respondto unexpected events of quite separate origins:some of them may be involved in more thanone interactive process or have pressing out-side deadlines to meet. So there is a fine bal-ance to be struck between adapting to thesepressures and maintaining the momentum ofprogress in the agreed common task.The core group task embraces all the

mundane jobs associated with organisingmeetings: tasks which range from the time-consuming job of getting people together atthe same time and in the same place, to theprovision of appropriate spaces to work inand adequate sustenance for the participants.However, in addition, it involves proper prepar-ation for and steering of the interactiveworkingsessions themselves; and it can take continualeffort to keep those not involved in these ses-sions in touch with what is going on.One of the greatest cultural difficulties

encountered when introducing the strategicchoice approach to organisations is that mostof these tasks are not conventionally con-sidered to be ‘real work’; or at most theyare considered as simple administrative taskswhich can be delegated to clerical staff to bedealt with in a routine way. Some aspects ofthe arrangement of meetings can, of course,be dealt with as a matter of routine; but exper-ience has repeatedly shown that the timeinvolved between meetings on the part ofthose who carry the main core group respons-ibilities should not be underestimated. Manynegotiations about involvement in meetings

and related matters have to be handled inform-ally on the basis of the rapport which is devel-oping through interactive working; and a failureto allow time for this can soon begin to erodethe quality of the process as a whole.Even where there are several members of

the working group who are committed to theproject for much or all of their working time,some element of advance scheduling of meet-ings is essential. For example, the same dayof the week can be scheduled for a numberof weeks in advance, with the understandingthat members of the working group will nor-mally meet for two and a half to three hours inthe morning and a similar period in the after-noon. Where possible, some flexibility shouldbe built in by an understanding that, if neces-sary, the afternoon and/or evening before asession might also be used or perhaps themorning of the next day. In some cases itmay be necessary to use two or more daystogether; while, in other cases, less time maybe necessary, allowing sessions to be can-celled. At intervals, other types of meetingsmay be fitted into this kind of schedule. In par-ticular, allowance can be made for meetingsof the responsible grouping, which involvessenior managers in a steering capacity, and ofthe reference grouping through which work-ing liaison is maintained with other depart-ments and agencies. This flexible scheduling

approach means arranging in advance a seriesof meetings, in which there are some basicand fairly firmexpectations about time commit-ment and participation in meetings, with theunderstanding that either may be extended orcurtailed should circumstances require.Between sessions, the core group should

arrange to meet every few days, althoughsuch meetings need not be very lengthy. Theperiods immediately before group sessions,and also immediately after, are also import-ant times to exchange notes on progress andideas on how the dynamics of group interac-tion can be maintained in the current situation.Laying out the room, setting targets for theday, monitoring progress towards them, ana-lysing the results and recording wall charts are

230

Practicalities

among the vitally important activitieswhich thecore group must undertake during such ses-sions. Without them, progress and directioncan quickly become lost. It is better than noth-ing to get together in the main working rooma few minutes in advance of the others – butthat is really not enough. On the other hand,it should not be necessary to organise suchmeetings formally – they should happen spon-taneously – but it is important that they shouldnot be so squeezed out by other pressures. Itis also important that the core group meetingsshould not be seen as secretive. If other mem-bers of the working group happen to come intothe room while the core group is discussingprocess management issues, it is importantthat they should not feel excluded from the dis-cussions but should be explicitly brought in.It is not recommended that a half-day work-

ing session be formally broken, for instance togo to another room for coffee or tea at someappointed time. For the loss of momentum canbe serious and it is much preferable that cof-fee, tea or other light refreshments be availablein the working room itself. Lunch breaks, onthe other hand, are more essential; they canbe important not only as respite for the parti-cipants and as consultation time for core groupmembers, but also for other informal purposes.In particular, process consultants working forthe first time on a problem, in an organisation ofwhich they have little or no inside knowledge,will often find themselves being briefed dur-ing the lunch hour on aspects of the internalpolitical background which are well enoughknown to most of the participants, yet cannotbe openly discussed during the course of aworking group session. Also, of course, socialinteractions in the evenings andweekends canplay an important part in the cultural acclimat-isation of all concerned, especially those whocome from a distance to an unfamiliar organ-isational setting.The facilitator role is of particular importance

because it is concerned with the group pro-

cess. The role is one with many parts andis sometimes handled flexibly between two

or more people who are experienced in pro-cess management. This sort of team facilit-

ating can be most effective, with the activefacilitator role changing hands from time totime. It allows the benefits of active facilitat-ing to be combined with simultaneous reflec-tion on it, which improves the effectivenessof the review process. In many cases, it canenable process issues to be discussed openlybetweenmembers of the group, with the inputof two or more independent views of what ishappening at any one time.The facilitator has always to be ready to

aid communication between the participants,especially where they may have difficulty insharing their different perceptions, concepts,assumptions and priorities. The task is to pre-vent progress being blocked whenever pos-sible, and to unblock it when it is. It meanskeeping all the participants feeling involvedby creating opportunities for all to maketheir contributions effectively. It includes act-ive guidance of the interactive work of thegroup and, in particular, guidance of the judge-ments associated with switching and loopingbetween modes of work. In making switchingand looping judgements, the main evaluativeemphasis is on seeking always to move in dir-ections where the growth of understanding isexpected to be greatest in relation to the timeand energies of the participants. It is not, ofcourse, easy for anyone, however skilled, tomake judgements in advance as to whethermore will be learned by a shift of mode inone direction rather than another. These are,essentially, judgements about where the rateof potential learning is likely to be greatest; andhere the idea of the learning curve is of value.When starting work in any mode, there is

usually an initial period of quick learning, theduration of which will depend on how muchis already known – and shared among the par-ticipants – at that time. When little is knownabout a problem, it can be some time beforethe rate of learning – the learning curve –begins to level off. But where there is anestablished base of understanding, a tailing-offeffect in learning is soon experienced, so more

231

FIGURE

86Management Checklist: Process

Practicalities

rapid switching and looping are to be expected.A group which starts with little experience instrategic choice is likely to be dependent on aprocess manager or facilitator for judgementsof this kind; but as a group becomes accus-tomed to this way of working, the breadthof participation in active process managementcan increase.What is important is that people should be

conscious of the mode in which they are cur-rently working, and of the opportunities forswitching and looping which exist at any par-ticular moment. For this reason, it is a use-ful guideline that any group which includesmembers who have never worked with stra-tegic choice before should have at least someexperience of work in each mode during thecourse of the first working day. A flip chartshowing the four modes (as in Figure 8) canbe posted on the wall as a point of refer-ence to which such people can refer; and occa-sional reference to the ‘Sutton Principle’ willhelp people to be conscious at any time ofthe opportunities to switch to the mode wherethey think the greatest rewards will be found.But switching judgements can be difficult,

especially if there is a sense of leaving workunfinished when moving into another mode.Part way through a lengthy analysis, the feelingoften develops that there is little more to belearned by going on – so the learning curve isnow becoming flatter – but nonetheless thereis resistance to a switch because the analysis isincomplete. At this stage, a relaxed approach isrecommended and the switch should bemade.The analysis can always be completed later if,in coming back to it through the cyclic process,that seems a helpful thing to do.A somewhat different form of this is experi-

enced when people find themselves labouringto define some aspect of the problem, oftenin terms of a decision area, comparison areaor uncertainty area the meaning of which isambiguous, or the wording of which is diffi-cult to agree. In such situations, the device oftemporary labelling has been found very help-ful in maintaining the pace of learning. The idea

is that a name is given to whatever is unfin-ished; it is put on one side; and then it can bepicked up later if it is still relevant. This devicehas become known as the ‘Rhubarb Principle,3

and around it has developed the informal short-hand of putting unfinished work ‘in the rhubarbsack’. In practice, the rhubarb sack is usuallyan open-ended list of unresolved items built upon a separate flip chart on the wall. This is thenused during review, and as input to agendabuilding exercises for future sessions.To summarise at this point, processmanage-

ment choices are vital to effective progress,though not often recognised adequately in con-ventional practice. It is important that theybe made explicit wherever possible, and thatadequate time be allowed for them. As an aidto this, a summary of relevant process issuesis presented in Figure 86. Some of the mostimportant aspects of process management tobe considered are those associatedwith judge-ments about switching and looping within andbetween group sessions, combined with thepassage of time. In evaluating the use of timethrough process choice, it is important to beguided by considerations of growth of under-standing rather than just good time-keeping orproductivity in terms of substantive plans; andthese are judgements to which all members ofa group can contribute, once they have becomeconscious of the considerations which arise.

PRODUCT: MANAGINGINCREMENTAL PROGRESSIN PRACTICE

In a process of strategic choice, there areboth invisible and visible products. Variousforms can be identified (Figure 33), takingthe substance–process dimension as well as

3 This name arose during early work on the formulationof policy choices for the Avon County Structure Plan,where theworking groupwas struggling with the defin-ition of a decision area about the quality of life. Theimpasse was resolved by one member suggesting thatthey label this decision ‘rhubarb’ for the time being andpressed on to something else.

233


the visible–invisible into account. The manage-ment emphasis is placed on the recording andinterpreting of these products as well as theact of producing them (Figure 34).In any interactive working session, visible

products will be continuously building up onflip charts around the walls of the room – apartfrom any notes that may have been made byindividuals. At the end of a session, the choicearises as to what should be done about theseproducts. At one extreme, they can be thrownaway, while at the other extreme, they canbe left intact as a starting point for the nextsession, provided the space is not required forother purposes. But there is a range of othermore practicable possibilities in between; andthis is where the task of recording becomesimportant.Flip charts can, of course, be saved at the

end of a session; and where this is the inten-tion it is useful that somebody charged withthe recorder role should number them in thesequence they were first written on – and adda date – at the end of the session. But flipcharts are bulky, and to save too many can beto create both a storage problem and potentialinformation overload,with severe difficulties ofaccess and retrieval.One of the most useful and simplest means

of recording is simply to photograph the setof flip charts around the walls of the room atthe end of a session – and possibly also at crit-ical moments during the course of a session,especially at any moment when the informa-tion on the wall is being reorganised, break-ing up a pattern of relatedness between flipcharts which has become familiar to the par-ticipants and which they might later wish torecall. Colour prints are recommended to cap-ture theway inwhich coloured pens have beenused; and use of a digital camera enables theresults to be quickly displayed. It is import-ant that the information in the photo-record

be easily retrievable, and that closely relatedparts of the work are kept together. Mount-ing sets of three or four prints together on A4sheets of paper has been found to serve suchpurposes well, especially if they are suitably

coded, dated, labelled and indexed (Figure 58).It is important that all members of a work-ing group have access to copies of the photo-record. It is sometimes found useful to takesome pictures of people working together aswell – both as a reminder to the participantsof the process, and as an indication to othersof the general style of working within thegroup.Another way of recording is to write up the

key aspects of visible progress in the form ofa loose-leaf record in a ring-back file, whichcan provide a more accessible set of referencepoints for the participants when they recon-vene for another working session. This can ofcourse be supplemented by photographs ofthe flip charts in groups on the wall – in whichcase, some or all of the charts themselvescan be scrapped, or the reverse sides writ-ten on again where the paper is of sufficientquality.Any act of recording – even one of mere pho-

tographic recording – involves some elementof interpretation of information into anotherform – usually a more condensed form appro-priate for later use. If the intended use isreference by the same working group at alater meeting, the extent of the interpretationmay be quite small, because those concernedhave become used to the language of stra-tegic choice – the graphical conventions, thevocabulary, the abbreviated labels by whichthey have learned to communicate with eachother. But it is important to remember thatthis sort of recording can easily become, ineffect, another cycle in the process. In somecases this may not be too serious; in othersit can be interpreted by other participants asan attempt by the recorder to manipulate theresults. It is hardly necessary to point out thatthis could have very negative effects on thegroup process. Where there is a danger of this,and when there is enough time, it is better tomake recording the explicit subject of anothergroup session – perhaps using a photo-recordas a starting point. Another way, if time isshort, is to allow the recorder to go ahead, butalso to allow a reasonable amount of time for

234

Practicalities

questioning and clarification at the start of thenext group session. This can be usefully com-binedwith circulation of the record in a reactivestyle between sessions.It is only to be expected that the members

of a working group will develop their own formof jargon – graphical as well as verbal. Thiscan be extremely useful insofar as it helpsthem talk about problems rapidly and effect-ively among themselves. But such jargon, ofcourse, can be of little value in communicatingwith others and, indeed, it can have the effectof confusing more than it informs. In practice,there is always a temptation for members of aworking group to try to communicate with oth-ers in such terms, presenting decision graphsor lists of alternatives described in terms ofbrief labels which have acquired significanceto them through the process of interactionwithin the group. But such information can-not have the same significance to others notso intimately involved. There is a danger thatfindings and supporting information presen-ted in terms of apparently esoteric jargon willannoy and alienate rather than bring any senseof involvement to those only intermittentlyinvolved – such as members of accountable,reference and representative groupings (Fig-ure 84). So, in reporting progress to others, andin providing a basis for interactive discussionswith those who are only occasionally involved,some more substantial form of interpretationinto plain language is required.A general impression of the progressive

build-up of visible products is presented inFigure 87. This shows the development of aloose-leaf record drawn from the flip chartsrelevant to each mode and, from that, the fur-ther development of documents suitable foruse outside the group. The process of pro-duction is one of incremental documentation.In practice, the loose-leaf record builds upand evolves gradually, providing at any onetime a statement of the latest findings andpossible recommendations coming out of thework. Whenever it is necessary to producea formal statement or recommendation, it isthen a matter of taking the relevant leaves

out of the loose-leaf system; assembling themin an appropriate order; and considering howto present the result. The remainder of theloose-leaf system provides the basis for anyreasoned justification for the choice; a sourceof supplementary information; and a startingpoint for future cycles.It is important inbuildingupa loose-leaf record

that it should reflect the products of work in allfour process modes. It will, of course, be vitalto record thecontent of a commitmentpackagewhere theworkof a sessionhas culminated in aproductof this form;but informationondecisionareas, decision graphs, problem foci, options,decision schemes, assessments, advantagecomparisons, shortlists, uncertainty areas andexploratory options can be vital for future refer-ence as well. On some major strategic choiceprojects, standard pre-printed forms have beenused to record some of these kinds of inform-ation, so that it can be retained in as organisedand immediately accessible a form as possible.What is most important is that the decision-takers should be in a position to question andchallenge the presented argument; and thattheir advisers, when challenged, should beable to retrieve further information from theloose-leaf system wherein the fuller recordsof their work are stored. So the outer rings ofFigure 87 are linked by loops which indicate atwo-way flow of questions and responses. Itis in this way that decision-takers, with manypressures on their time, can become inter-actively involved in the process – and cando so on their own terms, as it is they whoset the questions on which the dialogue isbased.In taking a view of planning as a decision pro-

cess in which uncertainties are managed con-tinuously through time, it is vital to recognisethe importance of process-related products.Nonetheless, it is the recommendations aboutthe substantive problems which will be thefocus of attention when the time comesfor formal decision-taking. Even where theyhave been developed using a commitmentpackage framework, these may differ littlein visible form to recommendations made

235

FIGURE

87Recording and Interpretation in Strategic Choice

Practicalities

without the use of strategic choice methods.4

But it has to be recognised that such recom-mendations may include many partial andinterim actions, to complement those whichare explicitly deferred or made contingent onfuture events. Although what decision-takersare being asked to sanction may be primarilythe products of work in the choosing mode,the provision of some supporting informationon the work of the other modes is essential ifthey are to feel that:

• the advice they are offered is soundly based;• their advisers have explored the shape of theproblem in a realistic way;

• they have looked at a wide enough range ofalternatives;

• they have taken all the most significant con-sequences into account;

• they have attempted to deal with uncertain-ties in an adequate way.

To demonstrate these things can mean adegree of post-rationalisation of the highlycomplex and adaptive processes of switching,looping and recycling through which membersof the working group in reality arrived at thefindings they present; processes which couldseem quite chaotic if any attempt were madeto present them realistically in full.As suggested in Figure 87, it is possible in

strategic choice, as in other ways of working,for a reconstructed argument to follow a linearcourse from shaping, through designing andcomparing, to choosing – though there is noreason why one or two feedback loops in theprocess should not be presented as well, ifthey are sufficiently critical in their implications.However, in strategic choice, such a linear

4 This was particularly evident in the English StructurePlan project. However, sonic of the participating teamsmade use of structured frameworks of ‘policy manage-ment information’, to interpret commitment packageconcepts into forms of presentation which indicatedhow policies might be adapted through time. Assump-tions were clearly stated, together with contingencieswhich would warrant change and constraints whichwould limit that change.

framework is not essential. Free of the con-straint of a retrospective process-based logic,the reasoning can be based instead on thestructure of the problem – with an increasedemphasis on the prospective aspects of thecontinuous decision process.Indeed, this kind of ‘prospective process-

based logic’ is essential in strategic choice,where the decisions are structured over time.Future procedures form an important part ofmost commitment packages. Not only is thereexplicit provision for various forms of explor-ation to reduce uncertainty, but statementsin the ‘future decision space’ are likely tobe couched in terms of procedures to reachdecisions, rather than definitive plans. Suchrecommendations are likely to be needed inorder to:

• keep track of how proposed actions turn out,and review the need for adaptive planning;

• monitor the performance of assumptions sothat critical misjudgements can bemanaged;

• provide for decision-making on a regularbasis, as well as prepare for cases of emer-gency and/or unforeseen circumstances;

• allow for participation by individuals and rele-vant groupings with respect to expected andunexpected future decisions;

• maintain progress in implementation pro-grammes related to time schedules andbudgetary cycles.

In addition, documentation in relation to futureprocesses is likely to include potential sourcesof information, as well as key individuals andorganisations who are particularly relevant tothe case in hand. However, it is often difficultto define procedures exactly – and especiallyuncertain is the behaviour of the individualpeople who will ultimately enable them towork. Therefore, it can be very helpful to havedeveloped a high level of understanding amongthe participants during the decision-makingprocess, so that there is a foundation of sharedperception,mutual respect and trust which canbe built on in the future. This entails not onlyknowledge of other ways of working, but alsoextended networks of communication within

237


which these invisible products can be put toeffective use.Such products, which are continuously

developing in a process of strategic choice,emerge fromwork in all modes and take shapein many ways. They take shape in the mindsof individuals, as peoples’ limited perceptionsof problems, possibilities, implications anduncertainties gradually become replaced byricher perceptions through the sharing of viewswithin an interactive group setting. Indeed, itcan be said that invisible products build upin the ‘hearts and souls’ rather than just theminds of individuals, insofar as they cometo share values as well as perceptions andto develop a sense of shared commitment tocommon directions of action or policy. Thesesomewhat nebulous products can have animmediate value through the production ofbetter integrated and more strongly suppor-ted decisions. But, in most cases, their fullworth is experienced over time – paving theway for more effective working in the longer-term future. However, they are likely to escapeunnoticed unless explicit efforts are made tohelp participants:

• first of all to become conscious of them;• and then to find ways of harnessing them.

Invisible products can be brought to a con-scious level most effectively as they hap-pen; and the facilitator will be trying to dothis. Unfortunately, this is no easy task; so itbecomes important also to try to capture theinvisible products retrospectively. Most com-monly in a strategic choice exercise, this isdone at the end of sessions, through allow-ing opportunities for reflection at that time.A quick recognition of the visible work doneduring the session is followed by evaluationnot only in terms of progress on the wall, butalso in terms of the growth – or lack of it – inunderstanding, consensus and commitment.People’s ideas about matters of process aswell as substance are likely to change; andthese can also have profound implications forthe quality of decision-making in the longer

term. So these also can be usefully included inany end-of-session sharing of views.Taking a more extended view, it is worth-

while to suspend business for slightly longerafter a series of working sessions, so that par-ticipants have an opportunity to consider thecumulative growth of invisible products incre-ments of which may easily slip by withoutcomment during more frequent evaluations.Longer, more formal surveys can be mosteffective at the end of a project.5 Question-naires can be focused on the potential ofinvisible products and ways in which parti-cipants intend to capitalise on them. Retro-spective questions paired with prospectiveones, addressing the same issue in differ-ent words, reveal changes in attitudes. Thesechanges may have been perceived at an indi-vidual level but become much more powerfulwhen they are seen as part of a group develop-ment. Naturally, as with any survey of this kind,feedback of the results to the respondents isessential.But all of this emphasis on process-oriented

products – visible and invisible – is to no availif nothing happens as a result of the work.What is required also is a convergence ofview towards a common understanding aboutwhat has to be done – or not done, as thecase may be. For this sort of joint commit-ment, there must be a degree of sharing ofperceptions about what the problems are; thesolutions which are available; and their likelyconsequences. All of which can build up to awell co-ordinated set of feasible recommend-ations with the backing of all concerned –especially given a framework for structuringthem over time.Where a working group has an inter-

departmental or inter-organisational member-ship, there is a need for a framework which

5 Evaluative questionnaires were used for this at theend of the LOGIMP experiment (Friend, Wedgwood,Oppenheim et al., 1970) and, more recently, in a num-ber of environmental policy projects in the Netherlands.On two occasions, follow-up studies have been carriedout by means of interviews with participants three ormore years after the event, and the findings presentedin internal papers.

238

Practicalities

will enable the actions required of the variousparticipants to be brought together. In the stra-tegic choice approach, this can be achieved viaa mutual commitment package. Each set ofparticipants in a working group will have theirown separate commitment packages, withintheir own spheres of responsibility which maybe expressed in a variety of forms. Parts ofeach such commitment package can howeverbe merged to form a composite statement ofthe way forward or mutual commitment pack-age. But howmuch should be expressed in thisform will be a matter for negotiation, as theremay be various political sensitivities to keep inmind.In such a multi-organisational group, the

range of uncertainties encountered in agree-ing how to move forward at any time mayinclude areas of recognised conflict, whichcannot be resolved at that point. But adecision process, structured within a commit-ment package framework, provides opportun-ities for the management of such conflicts.It allows decisions over which there is con-flict to be deferred, whilst others over whichthere is agreement can be put into action. Inmany cases, when the time comes for thedeferred decision to be taken, the conflict willhave dissolved. Options can become infeasibleover time, while new ones, over which thereis more agreement, can emerge. In any case,even if the conflicts have not disappeared, atleast those involved will by then have a betterchance of understanding their consequences.Another source of conflict arises from uncer-tainty over events which may – or may not– occur. The commitment package includesspace for contingency planning which can beused to handle this sort of uncertainty. Pro-vision can be made to reduce the impact ofadverse circumstances and thus remove thefear which is causing the conflict.Of course, there may be times when it will

be necessary to recognise deep-seated con-flicts of interest, reflected in divergent pres-sures from different members of the group.Such conflicts are not often such as to blockthe opportunities for progress in a group using

strategic choice methods; but where this hap-pens, resort may have to be made to othermeans of resolution, such as external arbitra-tion or mediation. Sometimes, where there isnot one accountable grouping but two ormore,conflicting interests may have to be reconciledin an incremental way, through a process ofgradual mutual adjustment over time.6 Evenhere, the commitment package framework canstill provide a dynamic structure to aid collab-orative negotiation.Figure 88 contains a summary of the range

of products whichmight be considered – struc-tured as in Figure 33, along the dimensions ofsubstance/process and visible/invisible. Natur-ally, it is necessary to be selective in the choiceof products as in the other aspects of strategicchoice. The evaluative emphasis which ismostappropriate to the consideration of products isan emphasis on confidence; the gradual accu-mulation of confidence among participantsthat they are moving towards decisions whichare soundly based, at least so far as a real-istic appraisal of uncertainties and time andresource pressures allow. This is the emphasiswhich was contrasted with the more conven-tional emphases on accountability and com-pleteness of products (Figure 34), importantas these more tangible considerations remain.And the evaluation of growth in confidence isa matter not just for the working group itself,but for all those other participants who may beinvolved in other process roles.

CONTINUITY AND THE INDIVIDUAL

In the prototypical working situation wheresubstantial human and other resources arebeing committed to a major planning task, therate of progress towards a more continuousway of working becomes an important testof the extent to which the philosophy of stra-tegic choice is becoming diffused; especiallyso where there remains the expectation that

6 See, for instance, the case of Droitwich Town Devel-opment – Friend, Power and Yewlett (2001) or Batty(1977).

239

FIGURE

88Management Checklist: Products

Practicalities

themain substantive product will take the formof a fully integrated master plan, delivered bysome appointed deadline. This shift in the dir-ection of continuity has been called the ‘con-tinuisation’ of a planning process (Hickling,1982) and it becomes an important evaluativeemphasis in process terms, to accompany theemphasis on growth of confidence in the moresubstantive products.But the shift towards continuity has its limits,

especially when combined with the emphasison interactive working. People cannot workinteractively all the time and, indeed, practicalways of managing the process and productsare essentially to do with the management ofthose discontinuities that inevitably arise whenpeople can only come together periodically andfor limited lengths of time. Two expressionsthat are often heard from people involved incomplex planning tasks are ‘let’s start fromscratch’ and ‘at the end of the day’. In mostreal-life planning, there is no such thing as‘scratch’ and people never do reach ‘the endof the day’ in that final sense. Yet success-ive working days will come to an end and newworking days will begin – whether the newday is spent working on the same problem oron another. And these are the realities againstwhich moves towards continuity of processand incremental products must be judged.Inevitably, it is at the level of the individual

that the experience of continuity finds itsfullest expression. For individuals exist con-tinuously, in states of consciousness punc-tuated by more or less regular periods ofsleep. While conscious, they make choices ofmany kinds, some of them in a more or lessprogrammed and automatic way. They makechoices about eating, drinking, moving fromplace to place, about small-scale social interac-tions and economic transactions. Indeed, oneof the first examples of a decision area quotedin Chapter 2 was of the choice facing an indi-vidual as to how to respond to an early morningalarm, as the first choice of all at the start ofanother decision-making day.Some personal choices, of course, call for

more deliberation – either through a process

of simple choice (Figure 6), with work concen-trated in the designing and comparingmodes –or through a more subtle process of strategicchoice such as that exercising the individualdepicted earlier (in Figure 30) relaxing in anarmchair and contemplating how some set ofuncertainties should be managed – in whichsituation one possible response could be torise out of the chair and to discuss it withsomeone else.Choices for the individual arise over where

to direct his or her attention at any moment;where to look; what to read; to whom to listen;and, indeed, how to use any of the varioussensory mechanisms with which the body isequipped. These choices about scanning arecontinuously being made, more or less con-sciously, and are accompanied by choices inthe realm of action, or doing, through themotor mechanisms of the body. So any com-pletemodel of individual decision-makingmustinclude some representation of these motorand sensory mechanisms which provide con-nections to the physical world outside.Following this line of argument, any com-

plete representation of a process of stra-tegic choice should include scanning and doingmodes in addition to the four already con-sidered in depth in this book. A more completesix-mode representation appears in Figure 89.The two additional modes are shown asembedded in a world of operations in con-trast to the technical domain of designing andcomparing, and the political arena in whichthe shaping and choosing work are done. Sowhat has been called the process of simplechoice (Figure 6) in effect by-passes the polit-ical arena – unless and until difficulties areencountered which mean moving in the direc-tion of increasing complexity and, therefore,increasing interactions with others whose val-ues and perceptions may conflict.Of course, organisations as well as individu-

als have their ‘worlds of operations’ which tendto provide the strongest strands of continuityand predictability in their work. Individuals whoare employed by organisations will spend partsof their conscious lives making choices within

241

FIGURE

89Strategic Choice as a Continuous Process

PRACTICALITIES

Practicalities

the framework of their employer’s field of oper-ations, and other parts of their lives makingchoices within more personal settings. Manyof these choices will be simple, others morecomplex – possibly involving not just one tech-nical domain or one political arena, but several,which may be interrelated in intricate ways.So the six-mode diagram of Figure 89 providesno more than a conceptual reference point inattempting to follow activities within a set ofprocesses which may, in practice, have a veryrich and complex texture – with threads ofcontinuity provided by the scanning and doingactivities of all the various individuals who maybecome involved.The extended six-mode model of strategic

choice will not be further developed here,but it can be used as a base from which toexplore further what is involved in work in anyof the four main process modes, with refer-ence to the others. In particular, Hickling (1985)has developed a view of evaluation as a five-finger exercise, in which the comparing modeis treated as the palm of the hand, with the five‘fingers’ of designing, shaping, scanning, doingand choosing being used in various sequencesand combinations as the pianist consciouslyworks to extend his or her skills.

KEEPING IT SIMPLE

The emphasis of this chapter has been onpractical guidance in managing a process ofchoosing strategically through time. Atten-tion has been given in turn to the fouraspects of technology, organisation, processand product. These together comprise a coher-ent approach – or merely a ‘way of working’

if these aspects are developed in an informalrather than a conscious way. In the use of thestrategic choice approach, these four aspectshave provided a structure for the designand management of comparatively large-scaleexercises. In some cases, governmental bod-ies have committed substantial human andother resources over a period of months towork on important national, regional or localproblems. It is in such settings that deliber-ate attention has been paid to the finer pointsof technological choice (Figure 83); to the rela-tionships of different groupings (Figures 84and 85); to the scheduling arrangements formeetings (Figure 86); and to the systematicdocumentation and interpretation of products(Figures 87 and 88).Important as these practical points are, they

cannot, of course, be taken into account soconsciously in every application of the strategicchoice approach. Themost important guidelineof all is to keep things simple wherever pos-sible. Problems vary and pressures vary. Ifan opportunity suddenly emerges to draw onstrategic choice ideas in the heat of workingwith a few other people on an urgent prob-lem, the setting andmaterialsmay not be ideal;the working group may have to be treated asgiven and the time constraints may be severe.So the immediate products might have to besimple and direct. But a consciousness thatthere are management choices involved ineach of the four aspects of technology, organ-isation, process and product can provide a valu-able background to the snap judgements thathave to be made, even in such informal work-ing situations – and a framework for reflectiononce the moment of creativity has passed.

243

10 The electronic resource

WHAT ROLE FOR INFORMATIONTECHNOLOGY IN STRATEGICCHOICE?

Should electronic information and communica-tion technologies have a central, and possiblyan expanding, role to play in the future develop-ment of the strategic choice approach (SCA)?This has been a subject of sustained and vigor-ous debate since the early years of the designof the toolbox presented in earlier chapters. Onthe one hand, it is not hard to see that there aresome aspects of the methods presented therethat might be handled more quickly and pain-lessly by electronic means. On the other hand,some users have argued that there shouldbe little or no place for electronic methodsif the emphasis of the approach is to remainon interactive working among the people whoare involved together in any decision process,drawing primarily on the knowledge and valuesthat they carry with them rather than on sharedsources of external data.Yet the question of the role of electron-

ics in strategic choice is one that has to beraised repeatedly, as circumstances change,for the pace of development in informationand communication technologies continues toaccelerate, and at the same time the breadthof access to these technologies continues toextend. Whereas the world of electronic com-munication might once have been seen asthe province of the backroom expert, it hasbeen since gradually extending its reach intothe home, the schoolroom and the meetingplaces of local communities. Younger gener-ations have begun to make casual and intuitiveuse of mobile communication devices which

their elders have had to learn to use in a moreconscious manner, if at all.

PROGRESS IN DESIGN OFSOFTWARE SUPPORT FOR THESTRATEGIC CHOICE APPROACH

Software to support some of the more tech-nologically demanding methods within thestrategic choice approach – specifically, thecombinatorial aspects of the analysis of inter-connected decision areas – has a historyextending back as far as the early seventies,when a program to deal with the combinatorialaspects of AIDA was developed in FORTRANby Hadley Hunter with other colleagues in theInstitute for Operational Research.Then it was at around the time when

the first edition of this book was publishedin 1987, when microcomputers were begin-ning to become widely available at affordableprices, that pilot work began on the designof interactive software to support the widerprocess of working flexibly through the fourmodes of shaping, designing, comparing andchoosing, helping decision-makers to moveincrementally towards commitment over time.Through a small family company formed by

John Friend in the Sheffield Science Park,1

work proceeded on the development and mar-keting of a prototype software package calledSTRAD – short for strategic adviser – designedto guide users interactively in drawing onthe full range of tools as presented in earlierchapters. The design, writing and testing of the

1 See the Planning Under Pressure companion websiteor www.stradspan.com.

245

FIGURE

90Design of the STRAD 2 software

DESIGN PRINCIPLES:

STRAD 2 provides two windows of access to additional software resources for each of the

four strategic choice modes. One window of each pair is primarily concerned with managing

inputs to the work of the mode, while the other window is concerned with managing outputs

from the work of the mode – see examples in Figures 91 and 92.

As a matter of principle, the management of the continuing decision process itself, including

any switching or looping between modes, is treated as a matter for human judgement rather

than machine.

The Electronic Resource

program was carried out by John Friend’s sonDave, who writes about some of his experi-ences in his contribution to Chapter 13.The first commercial version of STRAD

was published in 1991, designed for IBM-compatible microcomputers running under theMS-DOS operating system. Several potentialusers had previously advised that it would havebeen more appropriate for the developmentwork on STRAD to be carried out in the alterna-tive Apple Macintosh environment with itsmore user-friendly graphical interface. It waswith reluctance that this suggestion was setaside; for it would have made it more difficultto test the resulting package with a wide rangeof potential users.However, shortly after the launch of

STRAD 1, Microsoft was to introduce themore flexible graphical user interface of MSWindows 3.1, with enhanced capabilities anda capacity to display more than one windowon screen at a time. This made it possible forusers to view simultaneously the progressthey had made so far in the different modes ofthe strategic choice approach – thus simulatingthe way in which a workshop facilitator canswitch attention flexibly from one flipchart toanother, and from mode to mode, when work-ing with the more familiar materials of paperand felt-tip pens. Initially, the storage limita-tions of early personal computers imposedsevere limitations on this flexibility. However,it was not long before the development ofmore powerful microcomputers was to enablethese constraints to be overcome.STRAD 2 for Windows was first released in

late 1994. Since then, it has been upgradedseveral times to enable it to operate withnewer andmore powerful versions of MSWin-dows. The scale of investment in marketinghas remained modest, yet single user or net-work licences have since been purchased inover 30 countries, by users in a range of mar-ket sectorswhich gradually extended the previ-ous range of application of the Strategic ChoiceApproach. Among the main types of user ofthe software have been managers in business

and the public and voluntary sectors; plan-ners, consultants and policy analysts involvedin the management of both intra- and inter-organisational projects; and teachers, studentsand researchers in leading schools of manage-ment, planning and public policy.

WINDOWS OF ACCESS TO WORKIN THE FOUR MODES OFSTRATEGIC CHOICE

In the early design work on the STRAD soft-ware, the first aimwas to design a linked set ofprogram modules corresponding as closely aspossible to the fourmodes of strategic choice –shaping, designing, comparing and choosing.It was intended that these modules shouldreplicate as closely as possible the graphicalformats introduced in Chapters 2 and 3 of thisbook, and further elaborated in Chapters 5, 6,7 and 8. From this point of departure, oppor-tunities could then be explored to offer usersaccess to additional procedures reflecting theanalytical power of the computerwherever thismight be helpful. For example, the computermight introduce new ways of supporting thechoice of a focus within a complex decisiongraph; or the development of a set of feasibledecision schemes where the range of choiceis constrained by many option bars; or thecomparison of alternative schemes taking mul-tiple comparison areas and diverse sources ofuncertainty into account.In the design of STRAD 2, this logic was pur-

sued further, through the design of eight pro-cesswindows, two for each of the fourmodes,as shown in Figure 90. In general, the purposeof the first window relating to each mode is tohelp users in organizing inputs to the work ofthat mode, while the purpose of the second isto help in organizing outputs.The principle behind this picture is that the

process of sustaining flexible progress throughthe four modes of the approach, from shap-ing of problems to the achievement of incre-mental and continuing progress towards actioncommitments, should continue to be viewedas a human process, whether it entails people

247

FIGURE

91Shaping and Designing with STRAD 2

Designing

Shaping


working individually or collectively.What is rep-resented by the eight associated process win-dows in Figure 90 is merely a set of windowsoffering access to certain supporting resourceswhich the STRAD program can bring them intheir endeavours.The set of illustrations of screen formats

that follows in Figures 91 and 92 illustrateshow, in the design of STRAD 2, it has beenpossible to replicate several of the graphicalformats usedwhenworking on flip charts, withadaptations to the alternative medium of thecomputer screen. Furthermore, they illustratehow, through these adaptations, the softwarecan offer certain additional forms of supportto users of a kind that is difficult to introducewhen working on paper alone. The illustrationsin these two Figures are all based on the SouthSide case that runs throughout this book, soas to facilitate comparisons with the freehanddrawings that accompany earlier chapters.

EXTENDED TOOLS FOR SHAPINGAND DESIGNING

Figure 91 brings together examples of three ofthe main types of window layout available inSTRAD 2 to support the work of the shapingand designing modes.The first stage in working on a problem is to

set up a project file for the problem situationbeing examined, which can then be developedover the course of one or more working ses-sions, and stored at intervals under successiveversions of the file name if desired. The func-tion of the firstwindowof the shapingmodule –the Overview window, not shown here – is toprovide rapid access to current information onall decision areas, uncertainty areas and com-parison areas at any stage of work.The first picture of Figure 91 shows the

format of the other main window of the shap-ing module, the Focus window, which is usedin STRAD 2 to help the user in developing suc-cessive versions of the decision graph. Thedecision graph shown in this example con-tains the same decision areas and links as thatpresented in Figure 40 of Chapter 5. The focus

window enables various operations that cor-respond to those discussed in Chapters 2 and 5to be followed through. Decision areas can bedragged to or from a list of those not currentlyincluded in the graph (‘ex graph’), or they canbe positioned and repositioned freely withinthe graph. Further dragging operations enabledecision links to be inserted or removed. Col-our distinctions are available to emphasiseimportant or urgent decision areas; to distin-guish uncertain from firm decision links; andto indicate differences in the ‘ownership’ ofdecision areas, much as when using flip chartsin a workshop.In addition, STRAD2offers a choice of aids to

the development of a new problem focus thatdoes not depend on visual judgement alone.A toolbar displays various buttons designedto help users in selecting a new focus – forexample by highlighting a suggested ‘triad’ ofthree interlinked decision areas as a startingpoint, based on a user-adjustable points sys-temwhich enables criteria such as importance,urgency and density of links to be taken intoaccount. As a matter of principle, any suchrecommendations are always subject to theoverriding judgment of the user. A strip at thefoot of the window displays miniature picturesof all earlier stages in the development of afocus, enabling any of these earlier graphs tobe called up again as a starting point for furtherwork.The second window of Figure 91 illustrates

the layout of the compatibility window, whichis used in STRAD 2 to organise inputs to thedesigning mode. A grid format similar to thatof Figure 15 is used in preference to the optiongraph format of Figure 16, as this representa-tion is simpler to work with in any situationwhere there are more than a few option bars.One of the option bars is here shown as uncer-tain (?) as in Figure 49.Reasons for inserting option bars can be

registered in a list that is available for future ref-erence; and procedures are available to enableusers to inspect, suspend or cancel any reasonon the list, in order to expand the set of feas-ible schemes. Below this list, an estimate of

249

FIGURE

92Comparing and Choosing with STRAD 2

Comparing

Choosing


the total number of feasible decision schemesis presented, using the formula explained onFigure 52.The final window in Figure 91 presents an

example of the schemes window, throughwhich the set of decision schemes resultingfrom any specified pattern of option bars canbe rapidly displayed – normally in the form of anoption tree as in Figure 17, though an alterna-tive list format is also available. In this window,there are procedures for altering the sequenceof decision areas in the tree; for calling up acount of the appearances of each option; andfor selecting and ‘locking’ a preferred optionin the first one or more decision areas in thetree, working from the left. This is normallyonly done after some relative assessment ofthe schemes has been carried out, and a work-ing shortlist has been developed through theassessment procedures now to be described;this gives effect to the principle of robustnessdiscussed towards the end of Chapter 3.

EXTENDED TOOLS FORCOMPARING AND CHOOSING

In Figure 92, a set of four further illustrationsof STRAD windows is presented, indicatingsome of the main procedures used to sup-port the work of the comparing and choosingmodes. To support the work of the compar-ing mode, STRAD introduces several featuresthat go some way beyond the relatively crudemethods advocated in Chapters 3 and 7. Forthe software makes it possible to carry outquickly various aggregating and sorting oper-ations of a kind that it has been found scarcelypracticable to introduce in a workshop settingif a climate of interactive working is to be sus-tained. The main Assess window, illustratedat the top of Figure 92, provides access to aset of secondary assessment windows, twoexamples of which appear below. This mainwindow displays a grid of all decision areas cur-rently in focus against all comparison areas thathave been selected for current use. Within thisgrid, users can enter their first intuitive judg-ments about the relative levels of impactwithin

particular comparison areas of the choice openwithin each decision area, as a preliminary stepbefore entering their assessments of differ-ences between specific options. Having thusentered their initial judgments about the rela-tive value weightings of comparison areas inan intuitive yet logically defensible way, theycan then focus their attention on those cells ofthe grid in which they expect to find the mostsignificant differences. They can modify thesejudgments later if they wish, by means of anoptional display at the foot of the grid coveringall decision areas combined.Coloured rectangles indicate those cells

where option differences have already beenassessed; open ones indicate those in whichthis is not yet done. The width of any cell isinitially set at zero, as in the case of the cell forROADLINE?with JOBS: in this example. How-ever the cell width can be altered, by meansof a dragging operation, to indicate anythingbetween a marginal to an extreme impact, inanswer to the question posed in the spacebelow the grid.A double click on any cell opens up a second-

ary option assessment window for that cell, asshown in Figure 92. Each option is initially rep-resented by amarker in the centre of the range.These markers can bemoved to new positionsto represent the user’s assessment of optiondifferences. Each option also has a pair of sym-metrically placed ‘range markers’, the distancebetween which can be increased to representperceptions of uncertainty.The scale used for assessment in any com-

parison area can be calibratedwith reference toany specified numerical unit of measurement,wherever this can be defined. Yet STRAD doesnot require the user to make such a calibrationin the case of less tangible comparison areas;for the software applies its own internal metricdefined by the intuitive value judgments thathave been entered in the main Assess win-dow, and there are procedures for automaticadjustment whenever the limits of the workingrange of the window are exceeded.The option assessments from all the filled

cells of the assess window can then be

251


automatically combined – across decisionareas, across comparison areas or both. Thecombined scheme assessment window ofFigure 92 shows the outcome of this process.It displays a set of multi-criteria assessmentsfor all the feasible decision schemes displayedin the tree of Figure 91, after aggregating allthe option assessments entered so far. Theformula applies the weightings to the comparison areas as they were last adjusted in themain assess window, applying a standard sumof squares rule as a means of estimating thecombined range for each scheme.Any of these scheme assessments can be

adjusted at this stage to reflect any furtherknowledge about the effects of combiningoptions from different decision areas in par-ticular ways; and STRAD can quickly sort theschemes to display them in descending orderof preference. Schemes near the top of the sor-ted list can then be selected to form a workingshortlist. Information on the schemes includedin this shortlist will then be carried back to theschemes window, as a basis for the selectionof robust actions in the more urgent decisionareas.The set of shortlisted schemes can also

provide a starting point for closer compari-sons of pairs of promising alternatives usingthe balance window. This window is similarin its layout to the type of advantage com-parison chart shown in Figures 21 and 64,and is therefore not reproduced here. Also notshown in Figure 92 is the format of the uncer-tainty window, which is designed to help in thejudgments about alternative ways of managinguncertainty, using the types of procedure illus-trated in Figures 24, 71, 72 and 73.Figure 92 concludes with an illustration of

the layout of the progress window which isdesigned to bring together proposed outputsfrom the choosing module. This windowdisplays the current state of progress towardsdecisions in the format of a commitment pack-age grid, using a broadly similar format to thoseillustrated in Figures 26 and 77. In STRAD,the less definitive phrase progress package isadopted to describe the general format of the

grid; then it is up to the user whether any suchpackage is subsequently adopted, either asan explicit commitment, or as a recommenda-tion to others; or as a strategic option to besaved and compared with any alternative stra-tegic optionswhichmight have been designed.Clicking operations enable fuller details of anyelements of a progress package to be entered,edited or inspected – including details of anyproposals for implementation (who, when,how). There are print and export proceduresthat then enable the content of a progresspackage – or indeed of any other STRAD win-dow display – to be printed out as hard copythat the user or users can take away as a recordof what has been agreed at this stage in theprocess.

EXPERIENCES IN USING THESOFTWARE

Since the first production version of the STRADsoftware was launched in 1991, and especiallysince this was followed by the first Windows-based version STRAD 2.1 in 1994, there hasbeen enough feedback on its applications byusers who have applied it in management,consulting, project co-ordination, teaching andresearch to provide us with several learningpoints to guide the further development ofelectronic support to users of the strategicchoice approach.It is also important to ask what implications

this additional electronic ‘compartment’ in thestrategic choice toolbox can be expected tohaveonthefuturedevelopmentof theapproachas awhole, and on the future range of its applic-ations. For the provision of software for stra-tegic choice, supportedby tutorial andhelp filesand a comprehensive user’s manual, makesavailable an additional channel for disseminat-ing understanding of the approachmorewidelyon a broad international front.Contacts have been maintained, through

both personal and electronic channels, withmany people in the developing network ofusers of the STRAD software. Also, oppor-tunities have arisen in many countries to

252


demonstrate the capabilities of the software toprospective users, ranging from policy-makersto local community development workers. Ithas been found that one of the most effectiveforms of demonstration is to begin from a livesituation on the current development agendaof those present, then to start building up aproject file based on a direct process of ques-tioning by the facilitator/consultant. The resultis a gradually developing picture of problemstructure jointly constructed by the problemowner(s), the consultant and any others whomay be present. One successful frameworkfor generating such an informal computer-supported discussion is reported in thecontribution by Rebecca Herron and DennisFinlayson in Chapter 13.

SOFTWARE AS AN AID TOPROGRESS IN SMALL INFORMALMEETINGS

One of the more surprising conclusions fromthe experience built up so far in the use of soft-ware as a problem-structuring aid is its poten-tial in supporting the use of strategic choicemethods by clusters of two or three peoplemeeting casually and informally in their ownworking environment – so long as they haveaccess to a desktop or laptop computer witha monitor large enough to enable all to fol-low progress. The contrast here is with themore spacious kind of workshop environmentthat becomes so important as a setting for theuse of flip charts in wider groups. While theworkshop process itself may be highly informaland adaptive, this is counterbalanced by themore formal preparation that may sometimesbe necessary to reserve a suitable workingspace in advance; to requisition supplies of flipcharts and pens; and to reserve dates in thediaries of those invited to participate.There have now been several occasions

when STRAD has been put to good effect infacilitating smaller, more intimate meetings ofa few people grouped around a single work-station. All that is required is that at least oneof those involved should be familiar enough

with the software; and should also have suf-ficient personal facilitation skill, to be able todrive the process forward. As in the case ofa facilitated workshop, the art is in balancingthe concern for steady progress against theconcern to encourage each participant to con-tribute spontaneously as the picture of sharedprogress gradually builds for all to see.At this scale of the small intimate group,

the evidence is that it is quite feasible to sus-tain a creative synthesis between a processof spontaneous group interaction and a pro-cess of flexible interaction between humanand machine. For this is a scale where peoplecan switch their focus casually and repeatedlyfrom the computer screen to the other partici-pants grouped around it and back again. How-ever the evidence is that this kind of informalswitching can become more difficult as thesize of the group increases; the content of thescreen can of course be projected onto a wall,but there can then be a noticeable loss in termsof the interactive dynamics of the group.One effect of the use of the computer is

inevitably to make the developing picture ofproblem structure look more impersonal thanif the picture had been built up freehand on aflip chart. Through the introduction of the com-puter, the shapes become more regular andthe conventions more standardised. There is arisk that this may reduce the sense of owner-ship of the work; yet there can be compensat-ing advantages in terms of focusing attentiononmatters of content rather than presentation.Another consequence of the switch from flip

charts to computer is that the record of pastmistakes, and of the learning points to whichthey have given rise, may become less easyto retrieve. Yet there are facilities within thesoftware to enable these learning points tobe recorded, as systematically as the requiredpace of progress allows. This can be done bya combination of recording text notes on thekey assumptions underlying any new informa-tion entered at each stage; and saving succes-sive versions of the project file at intervals. Inthis way, a cumulative record of progress canbe built up; and revisited later as an audit trail

253


of the way in which steps towards importantdecisions were agreed.

COMPUTER-AIDED CONSULTING

The situation of a consultant using STRADinteractively with an individual client – or with asmall informal client group – is inmost respectslittle different from that of the small computer-supported meeting as already described. Thekey distinction is that – assuming that the con-sultant is acting as ‘process driver’ – he or shewill be concentrating on drawing inputs out ofthe others without being concerned to makemore substantive inputs of personal know-ledge to the decision process. The benefit ofthis style of consulting, as compared to con-sultancy without computer support, is that theclients can see a continually evolving pictureon the screen of the decision situation as builtup from the various information inputs thatthey themselves have provided. This enablesthe clients to make adjustments and correcti-ons at any time, and provides a foundationfrom which a firmer and more confident client-consultant relationship can be built.

STRAD AS AN INTERACTIVELEARNING TOOL

In several countries, staff members of schoolsof management, planning and policy haveexperimented with the interactive use ofSTRAD on both undergraduate and postgradu-ate courses. The challenge is to help studentsin exploring the realities of strategic decision-making in complex fields where the interestsof multiple stakeholders converge and mayconflict. Whether working with case studies orwith ongoing decision situations, the softwarehas proved a versatile means of introducingstudents to the general principles and workingtools of the strategic choice approach. Also,it has provided significant feedback on waysof improving the software both in terms of itshuman interface and its range of functions.

THE CHANGING SCENE:ADVANCES IN TECHNOLOGY,EXTENSIONS IN ACCESS

It would have been difficult for either ourselvesor others to have started work in developingSTRAD as a flexible and interactive tool tosupport ongoing processes of decision-makingearlier than the mid-eighties, when micro-computers were starting to become widelyavailable for individual desktop use. Fromthat time onwards, further waves of techno-logical advance followed in quick succession;advances in hardware capacities, in softwaredesign and in human–machine interfaces, andalso in communication via local and wide areacomputer networks. Among the innovationsof this period were the development of thefirst computer-supported group decision sup-

port systems, in which members of a group ofdecision-makers are equipped with individualterminals through which to contribute to thebuild up of a shared view of problem structure,projected onto a wall under the guidance of afacilitator. It was not long before there followedeven more dramatic advances in opportunitiesfor communication at a distance, through therapid spread of access to the Internet, to e-mailand to other new forms of messaging basedon mobile telecommunication systems.The limited scale of investment so far in

the development of software to support thestrategic choice approach has offered fewopportunities to keep pace with these devel-opments; even though it has helped to indi-cate the wider prospects that these advancesin technology, coupled with extensions inaccess, have been bringing into view.Shortly after John Friend, as initiator of the

STRAD project, reached normal retirement agein 1996, he was to take up a new research anddevelopment challenge through a 3-year part-time contract with the Lincoln School of Man-agement, addressed primarily to the extensionof participatory planning methods in develop-ing countries.Through this challenging new assignment,

the platform for experimenting with the

254


application of strategic choice methods andthe STRAD software was to continue to grow;while, the priority in terms of software devel-opment turned to the search for internationalpartners who could bring both experience inrelevant fields of application and resources ofmore professional software development andmarketing expertise. In early 2000, an ambi-tious project proposal was submitted fromLincoln to a new European Commission pro-gramme designed to stimulate the develop-ment of technologies to support the concept ofan information society. The proposal broughtthe Lincoln School together with partners infour other European countries, with the aim ofdesigning an Internet service to support theengagement of citizens in the development oflocal action plans.This proposal was to be unsuccessful in

the competition for funds in a highly com-petitive field. Yet the relationships with thesepartners – who include three of the newcontributors to Chapter 13 – have continuedto develop. Gradually, negotiations to build analternative base for the further developmentof electronic support to processes of strategicchoice have been moving in new directions.For further news on these developments, thereader is referred to the relevant section of thecompanion website, or to the Stradspan web-site at www.stradspan.com.

IMPLICATIONS OF WIDENINGTECHNOLOGY ACCESSFOR PARTICIPATORYDECISION-MAKING

Contrary to what might have been anticipatedtwo or three decades ago, growing experi-ence with STRAD is pointing to the viewthat experiments in access to flexible decisionsupport software of this kind can be morequickly and successfully designed with peopleinvolved in small-scale settings, such as thoseof local community action or small businessdevelopment, than with people engaged inthe management of larger and more power-ful organizations. For those people are likely

to have already learned to approach their plan-ning and management responsibilities in othermore closely structured ways, designed tofit the particular organisational and budgetarystructures through which corporate planningand control functions in those organisations areexercised.Yet many people who do not work for large

organizations are now becoming familiar withthe principles and conventions of informationand communication technologies through theirday-to-day access to the graphical user inter-face, to the Internet and to cellphone net-works, as access to these technologies hasbecome widely diffused to the home, theschoolroom, the community centre and othermeeting places within the wider civil society.Members of younger generations repeatedlydemonstrate that they can acquire knowledgeof these technological capabilities more rapidlyand intuitively than members of earlier gener-ations. Whether through immersion in interac-tive learning programs or in playing computergames, they are acquiring impressive skills ofa kind that have not been passed down fromearlier generations in the traditional way.In Chapter 13, several recent experiences

are described in introducing strategic choicemethods – sometimes with support from theSTRAD software – to people who have littleor no prior experience of other managementmethods, in a range of experimental settings inAfrica, South America, Italy, the Netherlandsand Lincolnshire in England. These accountsgive a few indications of the potential thatis now emerging for a widely diffused processof experimentation with electronic technolo-gies to support and guide participatory decisionprocesses in such relatively informal settings.Meanwhile, opportunities to support the dif-fusion of lessons from such experiments areemerging through a range of governmental andinter-governmental investment programmesdesigned to support capacity-building for sus-tainable development at grass roots level, bothin developing countries and in the more disad-vantaged regions of the developed world.

255


The continuing development of ever-morepowerful technologies, in combination withparallel extensions in access, means that newopportunities will continue to emerge for sus-tained innovation in forms of support fordecision-makers who must plan under pres-sure, both locally and in broader policy set-tings. Yet it is important that questions of

appropriate balance between electronic andother forms of technology to support respon-sible decision-making should be kept undercontinuing critical review. For deep politicalissues of access and empowerment will per-sist, and will have to be confronted in aresponsible way, however dramatic the futurepace of technological advance may appear.

256

11 Extensions in process management

Since the first edition of this book went to press, opportunities have arisen to applythe general approach presented here to the management of ambitious projects on anincreasingly large scale. Typically, these projects have involved people with many differentallegiances and cultural backgrounds coming together to address important issues in whichthey have a shared concern. The theme of this chapter will be the challenge of adaptingthe strategic choice approach to extensive projects of this kind, and the further adaptationsin technology, organisation, process and products which have been evolving in response.

EXTENSIVE PARTICIPATORYPROJECTS AND THE STRATEGICCHOICE APPROACH

As mentioned in the preface to this edition,many extensive projects – though by nomeansall – have been concerned with issues of envir-onmental policy in a national or an internationalsetting. Some of them have brought togetherparticipants from different countries of west-ern and eastern Europe, building on the suc-cess of the earlier projects in the Netherlandsfrom which several of the illustrations frompractice in earlier chapters are drawn.The emphasis on environmental policy is no

accident – for it is this field of policy which,more than any other, reflects the current con-vergence of two highly significant trends in themodern world of policy formulation, planningand decision-making. These are:

1. A growing appreciation of the complexity of

the issues – It is becoming more and moreclear that the issues we are now facing areoften of global significance on a time scaleinvolving generations far into the future –as expressed in the concept of sustainabledevelopment. Not only are the issues them-selvesmore complex but, at the same time,they are being seen as increasingly interre-lated – to the extent that what could once

be seen as externalities have now to behandled as central concerns.

2. A growing emphasis on collaborative meth-

ods of working – Thewider the set of issuesbeing addressed, the wider the range ofinterests and organisations that will con-sider themselves as having a stake in theprocess – and so the more widely it is nowbecoming recognised that collaborative asopposed to adversarial styles of workingoffer the only constructive opportunities foreffective progress. There is a widespreadunderstanding that it is only through theirinvolvement in the process that stakehold-ers will feel the sense of ownership of theresults necessary for there to be a high levelof commitment to their implementation.

While there is recognised to be still muchscope for improvement, the proliferation ofprojects based on the ideas of consensus-building is increasingly apparent. This has beenespecially evident in projects where the aimhas been to harmonise policies across inter-national boundaries, for example within theEuropean Union.These two trends have led to three types of

changes in the political context within whichextensive projects must be planned; and thesein turn have called for adaptations in the

257


strategic choice approach. The three mainchanges are:

1. Extended numbers of participants – Thewider the set of issues being addressed,the wider the range of interests and organ-isations that will expect to have a stake inthe process – and so the larger the num-ber of people with a claim to be involved.No longer is it sufficient to rely on one ortwo informal workshops, producing quickresults. It has become important that anystakeholder who has the power to makethings happen – or conversely to blockthings happening – should be an active par-ticipant.

2. Extended channels of communication –The wider the issues, the more attenuatedbecome the channels and procedures ofcommunication which link the participantsto their constituents. The more extensivethe range of participating organisations, themore complex will become the requiredprocedures of authorisation, resource com-mitment and control. In particular, thoserepresenting governmental organisationswill have to assure those to whom theyare accountable that they are undertaking afinite project with prescribed forms of out-come.

3. Extended time-frames – An important con-sequence of this is that time scales becomeextended if lasting progress towards agree-ment is to be secured. A process has tobe designed which may often extend over1 or 2 years, with successive meetings ofmany different types, managed together asa series with ample time allowed for main-taining the link between representativesand their constituents.

All these broad trends reflect a growth in thecomplexity of the organisational and politicalcontext within which the search for progress isset. Therefore, they require that greater atten-tion be given to the design of themore political,as opposed to the more technical, aspects of

the work. There have been the following con-sequences for the adaptation of the strategicchoice approach:

• in the overall design of the process, consider-ably greater attention has had to be given tothe design of the work that has to be done in

the political arena, as presented in Figure 89,as against the technical domain;

• an increased emphasis is required on meth-ods designed to help participants to work

with each other, as much as with the com-plexities of the issues they are addressing;

• it becomes important to simplify the analyt-

ical methods used as far as possible, espe-cially where there are differences of cultureand mother tongue among the participants;

• concepts relating to the management of

uncertainty may have to be kept in the back-

ground of thework rather than brought to theforeground in the early stages, as they aresometimes seen as threatening and can bestressful until some level of trust has built up;

• it becomes necessary to reinforce the skills

involved in processmanagement, by introdu-cing other third-party skills from such relatedfields as mediation and alternative disputeresolution;

• it becomes increasingly difficult for one per-son to handle effectively the extended facili-tation role; so a requirement arises for theformation of facilitation teams – often callingfor a planned investment in training in facilita-tion skills;

• it has become necessary to reconcile the

cyclic process with the linear principles of

conventional project planning and to agreea clear series of stages through which pro-gress towards outcomes can be monitored;

• it becomes important to give more explicitattention to the challenge of actively man-

aging the process within an extended organ-

isational context, throughout all phases of aproject.

For all these reasons, the strategic choicetoolbox as presented in earlier chapters hascome to be supplemented, in the manage-ment of extensive projects, by various further

258

Extensions in Process Management

concepts, processes, techniques and meth-ods. Some of these have been borrowedfrom other toolboxes – with adaptations whereappropriate – while others take the form ofmore direct extensions to the concepts andmethods presented in earlier chapters. Themain purpose of this chapter is to introduce theextended principles of process managementwhich have evolved as a result of a decadeof experience in the management of extens-ive projects. A fewmore specific examples willalso be given of the more specific adaptationsthat have been introduced in the technologyand the organisation of the strategic choiceapproach, and in the strategies used to captureboth the visible and the invisible products.

A TWIN-TRACK APPROACH

The argument that any process of strategicchoice involves a subtle interplay between rel-atively technical and relatively political types ofwork was first introduced in the discussion ofthe basic process diagram of Figure 8, and wasdeveloped further in Chapters 4 and 9. This lineof argument culminated in the presentation ofa general model – Figure 89 – indicating anintricateweb of relationships linking a technicaldomain; a political arena and a field of opera-

tions.To reflect the increased complexity and sig-

nificance of the political arena in the circum-stances of an extensive project, an extendedmodel of the process has now evolved, basedon the concept of a twin-track approach. Thismodel is based on a recognition of parallel andinterdependent streams of work for the tech-nical domain and for the political arena.These streams have come to be known

as the technical and the socio-political workstreams. The phrase ‘socio -political’ is usedas a means of reassuring senior civil servantsthat their role is not being seen as intruding onthat of elected politicians. This point can alsobe reflected by talking of a policy stream ratherthan a political stream – and this is the termthat will be adopted in this chapter.

The two streams of work play complement-ary roles in the overall process:

1. the work in the technical stream involvesmanaging the complexity of the issues, soas to develop confidence in the quality ofthe results;

2. the work in the policy stream involvesmanaging the conflicting positions of thoseinvolved, so as to develop commitment tothe results.

It is important for the participants to recog-nise that the distinction is one between typesof work rather than between types of roles.Indeed, those in more technical roles oftenhave a significant part to play in the policy aswell as the technical work streams and it ishelpful to keep them aware of this reality. Like-wise, those in policy roles may make contribu-tions to the technical work streams, althoughpressures on their time will often limit theextent of these.Most conventional planning projects begin

with a series of technical stages leading tothe production of a proposal of some sort,for example, a draft plan. Professional normsoften create an expectation that this be madeas technically perfect as possible, before workbegins in a series of political stages aimed atdeveloping commitment to that proposal.The first diagram of Figure 93 illustrates this

relationship – recognising that, for a short whileat least, the two types of work may run in par-allel, and also that each stage may in practiceinvolve work of a cyclic rather than a strictlylinear nature.It is at the interface between the two

types of work that the all-important processof consensus-building takes place. Conven-tionally, however, the duration of this inter-face is so short that the opportunities forconsensus-building become severely limited.If these opportunities are to be increased, it isimportant that both types of work should startas early as possible in the life of a project. Thismeans extending the length of the interface, asillustrated in the second diagram of Figure 93.

259

FIGURE

93Linking Technical and Political Work


In an extensive project, this can beachieved through structuring the overall pro-ject timetable with reference to an agreedschedule of events, each with an emphasison consensus-building at the policy–technicalinterface, and each usually involving a substan-tial number of people. The progress of theoverall project can then be managed throughcommitments on the part of those involved inboth work streams to the production of draftdocuments in preparation for future events.The policy work and the technical work

in combination can be viewed as coveringthe four modes of work which together formthe basic processmodel of the strategic choiceapproach – shaping and choosing within thepolitical arena, together with designing andcomparing within the technical domain. Tofurther clarify the nature of the work involvedin each mode, it has been found helpful tosub-divide these activities further, as follows:

POLICY WORK

scoping

shapingmode

��

focusing

��

packaging

choosingmode

��

committing

TECHNICAL WORK

generating

shapingmode

��

scheming

��

assessing

choosingmode

��

balancing

Several of the terms introduced above reflectthe terminology of specific strategic choiceconcepts introduced in earlier chapters – for

example, the terminology of the problemfocus, the decision scheme and the commit-ment package. Some correspondences willalso be noticedwith the terms used to describethe principal process windows of the STRADsoftware, as illustrated in Figure 90.The fundamental relationship between the

policywork and the technical work is illustratedin the third diagram of Figure 93. This modelhas become known as the political–technicalU-loop. Recognising that each stage of policywork will rarely in practice follow a linear logic,each policy stage is shown as a cycle, involvinginputs not only from previous cycles but alsofromwhat is described in Figure 89 as the field

of operations. The outputs of each policy cyclewill then provide inputs to later cycles, togetherwith any agreed policies or actions directedtowards the field of operations.In the diagram, each technical cycle is seen

as, in effect, an extension of the middle part ofa political cycle. It begins with inputs from thefirst part of the political cycle, taking the formof agreed directions for the work sometimesexpressed in ‘terms of reference’. It ends withsome form of recommendation or proposal,providing input to the latter part of the polit-ical cycle – often taking the form of a draftdocument.What is essential is that the two streams

of work should not be taken forward in isola-tion; rather, it is important that the interactionbetween them should be carefully structured.The principle of the political–technical U-loop,as illustrated in Figure 93, offers an importantconceptual reference point in doing this. How-ever, if it is to provide practical guidance overthe course of an extensive project, it has firstto be adapted to the gradually changing natureof the work carried out at successive stages.

APPLYING THE CONCEPTSTO PROJECT PLANNING

The design of any major project involves plan-ning some progression of activities extendingfrom its beginning to its intended end date,related in some way to the passage of time.

261

FIGURE

94Phasing Work in an Extensive Project

Adapting the ISCRA model for project management

WORK BITES DOCUMENTS time dimension → → → → → → → → → → → → → → → →

Initial focus; aims; organisation; planning

PROJECT PLANNING I S C R A

Status quo; current action commitments and policies

INFORMATION

BASE I S C R A

Trends; scenarios; problems & opportunities

PROBLEM FOCUS I S C R A

Criteria; constraints; shortlisted strategies

OPTIONS I S C R A

Uncertainties; assumptions; selected strategy

OVERALL STRATEGY I S C R A

Agreed actions; budgets; responsibilities; deadlines

OPERATIONAL INITIATIVES I S C R A

Research; communication; monitoring; adjustment

ON-GOING MANAGEMENT I S C R A


The purpose of formalising this is to enable allthose involved to keep track of what is goingon at any time.The following set of terms indicates the

changing emphasis of both the policy workand the technical work over the course of anyextensive project:

policy work technical work

1 initiation structuring2 orientation generating3 preferencing evaluating4 revision composition5 agreement finalisation

In practice, it is usual for more than one kindof work to be in progress at any particular time,even though the primary emphasis will tend toshift gradually over the course of a project.In order to reflect this shift, it has been found

helpful to view the progression of a projectin terms of a succession of linked U-loops, inwhich the nature of the work to be done atthe policy–technical interface changes, as wellas the nature of the work within each of thepolicy and technical streams. This progressionis reflected in the first diagram of Figure 94.In Figure 94 five sorts of activity are iden-

tified which, based on experience, are suitedto interactive working at the political-technicalinterface. These are summarised in theacronym ISCRA:

I Identifying the IssuesS Setting the StructureC Confirming the CourseR Reviewing the RecommendationsA Approving any Amendments

It will be noticed in Figure 94 that these fiveactivities do not proceed strictly in step withthe five sorts of work in the political arenaand the technical domain. In particular, the twoclosely intertwined technical cycles of gener-ating and evaluating are shown as appearingtogether between Setting the Structure andConfirming the Course.It will also be noticed that the primary inputs

to the early stages (I and S) are shown as

coming from the political arena, whereas theprimary inputs to the later stages (C, R and A)are shown as coming from the technicaldomain. This change in emphasis recognisesthat, in an extensive project where thereare many diverse interests at stake, valuejudgements are required in order to identifywhat are the key issues to be addressed, andthus give direction to the technical work. Thecontrast here is with more conventional mod-els of planning in which technical issues areregarded as dominant in the early stages.However, this relationship does not extend

throughout the process. Towards the end,those in more technical roles will tend tobecome more proactive, as it becomes neces-sary to draw the threads together and preparedocuments for the approval of those who carrypolitical accountability for the project.The most important practical challenge can

then be seen as one of designing, over thecourse of a project, the settings within whichthe work at the policy–technical interface is tobe done. In an extensive project, it has becomeusual for much of this work to be carried outin the course of relatively large events whichare sometimes called strategic discussions orstakeholder dialogue. These discussions aredesigned to bring together a considerable num-ber of people in both policy roles and relativelytechnical roles. Some of the work may, how-ever, be done in smaller sub-groupmeetings ofvarious kinds. At other times – especially at thebeginning or end of a project – it will be appro-priate for all the participants to come togetherfor events of a more celebratory nature.The number of full strategic discussions

scheduled over the course of a project willdepend on the project context. In a projectdesigned to be completed in a few monthsthere may be relatively few, whereas as manyas seven or eight strategic discussions may bescheduled in the case of an international pro-ject extending over 2 years or more.A strategic discussion is usually planned to

extend over at least one full day and usuallytwo days, in the course of which variousmore specific activities – sometimes called

263


‘work bites’ – will be scheduled, often inseries but sometimes in parallel. Many ofthese work bites will involve interactive work-ing, but others may involve a more reactivestyle of work. The overall nature of the workdone in a strategic discussion can therefore bedescribed as inter-reactive in nature.As might be expected, the nature of the joint

work carried out in such meetings is designedto be different from the nature of the work nor-mally carried out in either the political arena orthe technical domain. For there are only certainforms of consensus-building activity that suchmixed groups can be expected to undertake.As is also to be expected, the nature of thiswork will change through time to reflect thechanging emphases of the work within boththe political arena and the technical domain.In an extensive project involving many

participants, it becomes a practical necessitythat most of the primary inputs and outputstake the form of documents of some kind. Atsome stages, these may be relatively formalpapers; at other times, they may be briefer andmore informal. This will depend on the overallscale of the project as well as on the progressthat has so far been achieved.

THE EXTENDED ISCRA MODEL

In planning major projects, it has become com-mon practice to apply the ISCRA model notjust to the project as a whole, conceived as aunitary work programme, but to a set of inter-related work bites, each associated with theproduction of a different type of document-ary output. Each document is developed incre-mentally over the course of the project andbecomes a self-contained report on that part ofthe work – ultimately forming one of the mainsections of the overall project document.A typical set of work bites and their asso-

ciated documents forms the left hand side ofthe lower diagram in Figure 94. The names andtitles used to identify the work bites are notsacrosanct and, indeed, the whole set, whichhas been designed to conform to relatively

familiarmodels of strategic planning, should beadapted to the project in hand.The ISCRA sequence of activities is then

applied to each work bite of the project spe-cific set. This resulting framework is then usedfor the phased drafting and revision of a seriesof documents – one for each of the bites. Thework in each is usually scheduled to begin atdifferent stages of the project, then to run inparallel with the others to phased completiondates, as shown in the right hand part of thelower diagram in Figure 94. In fact, the workrepresented in the first few columns and thelast few columns is usually undertaken react-ively, and only the central columns are used inthis way.It is important to reiterate that this some-

what ‘pure’ generalised model is intended onlyas a framework for the design of a processspecific to the requirements of each project –depending on such variables as the complex-ity of the issues, the number of stakehold-ers and the time available. It is common fordocuments to be combined and/or some ofthe steps brought together or extended intime. For example, the Information and Prob-lem Focus documents together can be seen asa database; the Options and Overall Strategydocuments combine to form a strategy, andthe Operational Initiatives and On-going Man-agement documents could provide the basiccomponents for an action plan (see the illustra-tion from practice in Figure 95).There are two important advantages in the

application of the extended ISCRA model as ageneral framework for the design of extensiveplanning projects.

1. Firstly, it provides a practical means ofreconciling the principle of cyclic continuityas discussed in Chapter 4 with the expect-ation of steady and sustained progresswhich is so vital tomost of the participants –and their sponsors – in any project of thisnature. For example, opportunities will stillremain to review the focus of the pro-ject, and to reformulate the options, evenwhen the project is well advanced, as the

264


participants develop closer mutual under-standing and as they adjust to any importantchanges in their working environment.

2. Secondly, the extended ISCRA modelprovides a framework for the design ofsufficiently compact ‘work bites’ even inrelatively large strategic discussion meet-ings to be sub-divided in such a way as togive each participant a series of meaningfulparts to play, working for much of the timein relatively small sub-groups with definedtypes of outcome which have a recognis-able role in the overall project design.

EXTENSIONS TO THE PROJECTORGANISATION

One significant extension to the principles oforganisation as discussed in Chapter 9, andillustrated in Figure 84, concerns the overallmanagement of a project.In an extensive project, it is especially import-

ant that there should be a recognised project

leader who carries continuous responsibilityfor the overall management of the project, yetwho draws authority from an extended pro-

ject group which includes representation of allthe main interests involved. In a typical largeenvironmental project, these will include vari-ous governmental bodies plus a range of organ-isations representing business and communityinterests.This project group carries the ultimate

responsibility not only for the conduct of theproject work itself, but also for the productionof a management programme or implementa-tion document through which the work will becarried forward into the field of operations, toinfluence future decision-making after the pro-ject comes to an end.In the formation of the project group, it can

be important to go through explicit processesof stakeholder analysis – preferably extendingbeyond the kind of brief brainstorming exer-cise which is often considered adequate ina more limited project setting. Various meth-ods of stakeholder analysis are available (GTZ,1991;Wilcox, 1994) which not only enable lists

of stakeholders to be identified, but also helpin understanding the positions they might beexpected to adopt, the factions or coalitionswhich might be expected to form, and the pos-sible strategies through which they might beapproached when invited to participate in theproject work.It is important too that an extended view be

taken of what are referred to in Figure 84 asthe reference and representative groupings. Inan extensive project, it has been found use-ful to introduce an extended concept of a ref-

erence network, emphasising that the task ofdeveloping this falls on all those who becomedirectly involved in the project work. For thereference network is intended to tap into thepersonal networks of all participants, who areencouraged to suggest the names of individu-als whom they consider should be included inthe first instance. From this basis the networkcan be allowed to grow freely, encouraginganyone who shows an interest to join.Members of this wider network can then

be kept in touch with the course of theproject work through a periodic project news-letter, and encouraged to contribute by provid-ing inputs of information or opinion throughwhatever channels are available. In some cir-cumstances it may be appropriate to invitethem to particular sessions of the projectgroup – for example in the evening betweenthe two days of a strategic discussion. Themembership of this reference network willtend to expand over the course of a projectas more and more people find out that thework is in progress – ultimately evolving into animplementation network focused directly onthe often complex processes of implementingthe agreements reached.Wherever it is important that the project

group should have a larger membership –as is especially the case in international pro-jects – it becomes essential that smaller, moreflexible sub-groups be formed to carry outcertain aspects of the project group’s workat some stages of the process. There is nointention that these groups should take overthe responsibilities of the project group – they

265


are merely intended to enable it to work moreeffectively.It is usual to view such groups as falling

into one or other of two categories – taskgroups, which are primarily responsible formore technical aspects of the work, and inparticular for preparing specific outcomes ofthis for handling within the project group; andguidance groups, which are responsible forpolicy aspects and can act as political soundingboards where necessary.Whereas the task groups will tend to oper-

ate inaproactivestyle,developingproposals forconsideration in the full project group, the guid-ance groups will tend to play a more reactiveor responsive role, acting as a sounding boardwhere necessary – typically, during the finalstages inthepreparationofspecificdocuments.In an extensive project, as in any other, con-

tinuous management through a relatively com-pact core group between one scheduled eventand another is crucial. Typically this group willbe enlarged to include at least one processconsultant, whose role is to advise the coregroup on matters of organisation, communica-tion, process management, network develop-ment and project methodology. These peoplewill usually be centrally involved also in the stra-tegic discussions and many of the sub-groupmeetings, acting as facilitators as and whenrequired.It is also important in some circumstances

that the core group includes a technical sup-

port team, with expert knowledge of the sub-stance of the issues. Its members are likely tobe consultants, though this is not necessarilyso; they will often have a role in assisting anyadministrative staff in compiling the document-ation necessary to keep the process movingforward, based on the directions indicated bythe project group.

EXTENSIONS TO PROCESSMANAGEMENT SKILLS

An important point to recognise, especiallyfor those who are accustomed to the man-agement of more conventional projects, is

that the management of extensive participa-tory projects calls not only for practical pro-ject management skills of the kind which arewidely recognised as vital to the success ofany large and complex project, but also formore subtle skills of process management.The more extensive the project in terms ofthe range of interests involved, the higher willbe the premium on skills oriented towards thebuilding of such invisible products as synergy,mutual understanding and commitment, along-side the incremental growth of more visibleproducts.Furthermore, it is important to recognise

that process management in such a contextinvolves much more than the skill of facilitat-ing progress within a single working group –even though this has come to be regarded asthe key to success in managing a more inten-sive ‘one-off’ strategic choice workshop. Ser-ious attention must also be given to a rangeof complementary skills, to each of which thestrategic choice approach is relevant to a dif-ferent extent.These skills cannot be dealt with in detail

here, but there follow outline descriptionswhich are intended to give readers some ideaof what is involved and of ways in which theirown acquired skills might be extended andimproved:

• Organisational/group development skills –The concept of the extended ‘project organ-isation’ – as expressed in Figures 84 and85, and further developed in this chapter –provides plenty of scope for creative use oforganisational and group design skills. Theconcept of dynamic grouping described inChapter 9 is highly relevant both at this leveland at the level of designing sub-group workwithin a strategic discussion event.

• Process design skills – In the design ofstrategic discussion events on inter-reactiveprinciples, the concept of flexible schedulingdiscussed in Chapter 9 plays an importantrole. The use of ‘flexi-time’ (time put aside foruse when that use becomes more clear) and‘slop’ (allowing more time than is technically

266


necessary for each task) represents a recog-nition of process uncertainty and can providevital opportunities for the facilitation team torespond to the needs of the participants asthey are expressed.

• Network management skills – Recognitionof people on the fringe of the regular pro-ject organisation, who nevertheless have aserious interest in the subject, can offer thebasis for building a broader consensus thanwould otherwise be possible. The challengeis that, while setting up an extended projectnetwork as described earlier is not difficult,without proper maintenance it will just fadeaway. For this reason members of the net-work can become the main target audiencefor regular project newsletters, while innova-tive thinking in process design can provideoccasional opportunities during strategic dis-cussions to obtain input from the wider net-work. Two-way communication within thenetwork can also be enabled by encouragingwritten inputs, using methods ranging fromsimple questionnaires to exercises using theDelphi method (Moore, 1987).

• Effective negotiation/participation skills –These skills are sometimes thought of asuseful mainly at the point of interaction; butit is now recognised that the way in whichany negotiation or interaction is initiated,the amount of effort put into preparation,and the attention paid to follow-up actionscan be just as critical, in the context of awider understanding of process (Hawkinsand Hudson, 1986). The concept of strategicchoice as a continuous learning process, aspresented in earlier chapters, is significant inthis respect.

• Skills in the management of information

flows – The effect of the strategic choiceapproach here is to introduce new perspec-tives on the type of information being com-municated. Conventional approaches focuson what is known, and there is a strongemphasis on facts. Recognition of uncer-tainty, as in strategic choice, leads to anemphasis on what is not known and therange of assumptions which may have to

be made. What is more, the focus changesfrom a concentration on uncertainties of typeUE (i.e. data) to a more balanced approachin which peoples’ values and intentionsbecome equally important.

• Project evaluation skills – The more exten-sive a project, the more important itbecomes to build in explicit methods forkeeping track of progress at each stage. For,the more any personal feelings of either sat-isfaction or dissatisfaction can be shared,the more opportunities will be created forthe process manager to make adjustmentsin the subsequent process, as well as todemonstrate that it is being managed in atruly participatory way. Evaluation exercisesmay be either at a micro scale – involvingquick checks taken during interactive work-shops – or at a macro scale where the sub-ject is the whole project and where it isleading. The latter type of evaluation is usu-ally most relevant some way into the pro-ject, and can be used to switch the focusof the group to finishing the project as wellas to obtain important process-related feed-back. This principle applies to both the visibleand the invisible products of the process. Sofar as the visible products are concerned,the division of the work into manageable‘work bites’ within parallel streams of workcan provide a useful basis for assessing pro-gress. As discussed in Chapter 9, it is notso easy to keep track of the developmentof invisible products: although they are oftenmade a part of open evaluations within theproject group, a more sensitive way is forcore group members to check continuouslywith participants at a personal level.

• Conflict mediation skills – So far the prin-cipal ‘third party’ role discussed has beenthat of facilitator, which may be adopted byan individual or a team. The task is to enableeffective utilisation of the time of peoplewith diverse responsibilities and skills, withrespect to the prevention of open conflictamong the key objectives. However, it is onlyrealistic to expect that unproductive conflictwill occur from time to time and, when it

267


does, it is helpful to have available the medi-ation skills to manage it. The main differencebetween mediation and facilitation is thatthe mediator will spend more time with theconflicting parties separately, helping themcome to terms with the needs of the otherside – especially with respect to the rela-tionship with their constituents. He or sheis also likely to be a little more involved inthe substantive content, and the concept of‘options for mutual gain’ can play a usefulrole. The use of a progress package in resolv-ing conflict is also significant here, as conflictis often based on differing opinions about theassumptions to be made. Use of the idea ofa contingency plan in case of the assump-tion being wrong can provide just the break-through which is needed.

• Project planning skills This can be viewed interms of process design at the scale of theproject as a whole rather than at that of theindividual interactive event. The basic stra-tegic choice concept of cyclic continuity ismost relevant here, and the ISCRA processprovides the framework for putting it intoaction. The more extensive the project, themore important too that process manage-ment should become seen as a team activity;for it can become a heavy load for any indi-vidual to carry alone, and the process can bemuch enriched when each member of a pro-ject management team has the opportunityto draw on the experience and resources ofothers.

CREATING A CONTEXT FORSUCCESS IN EXTENSIVE PROJECTS

Experience has repeatedly shown that the suc-cess of an extensive participatory project doesnot depend on the quality of its project man-agement alone, even when the importance ofgood process management is fully recognisedand applied.The experience of ambitious projects that are

set up to address complex areas of policy –and, in particular, of environmental policy – hasbeen that a project can only be expected to be

fully successful if the conditions for successare in place at the outset. This has proved gen-erally true, but is especially the case where theproject has had trans-national sponsorship andthe aim has been to work towards harmonisa-tion of policies between different nation stateswith different languages and cultures.The effectiveness of such a context has

been found to depend critically on the follow-ing three conditions:

1. a significant set of issues which is agreedto be of sufficient common concern;

2. an experienced project leader with a clearunderstanding of process managementissues, as a focus for the project manage-ment team;

3. an internal sponsor highly placed in theclient organisation, with a commitment toseeing the project conducted in an inter-active participatory way.

While these aspects are indeed critical, andtheir absence is likely to make a projectextremely difficult – if not impossible, there areothers which it may be possible to influenceto some extent. These are listed here in anapproximate descending order of significance:

• a long enough project in terms of the timeavailable for developing the long-term ‘addedvalue’ represented by the invisible products;

• a committed project group which believesin the benefits of the proposed style ofworking, or is at least willing to suspend itsdisbelief;

• an external sponsor who plays a similar sup-porting role as the internal sponsor, but fromoutside the organisation;

• a non-hostile culture in the ‘client’ organisa-tion or organisations, so that the work is notconstantly under threat of being sabotaged;

• an appropriate workplace which is some-what larger than normal, with good light,ventilation and uncluttered walls.

The absence of all these characteristics of anappropriate context is likely to make the suc-cessful conduct of an extensive participatory

268


project impossible. However we have neverexperienced such a serious case. The mostlikely explanation for this is that, in such asituation, this type of project would ever bemooted, let alone get to the stage of seriousconsideration.So it is probable that all is not lost. While it

may be difficult or impossible to create theseconditions if they are not already in place, itmay be possible to influence them to someextent. Therefore, in the initial negotiationsleading up to the launch of a project, it is vital todiscuss explicitly how far such conditions canbe fulfilled.Where any shortfalls exist, it will pay to

review ways in which they might be substi-tuted. If this proves impossible, the practicalquestion to be faced is one of how SCAcan be adapted to address any of the morenegative characteristics that may have beenrecognised. This is where the process man-agement skills described above, such as thoseof organisational development, process designand project evaluation, will play a crucial role.Inevitably there will be much uncertainty to bemanaged, and there is likely to be a case forapplying the concept of the progress packageto the management of the process itself. It willthen be a case of careful negotiation to balancethe needs of all the key stakeholders withoutsacrificing the viability of the approach.The management of an extended project is a

process that offers rich opportunities to manypeople for many kinds of learning. One of themost important forms of learning concerns thehighly complex process of managing the pro-gress of the overall project, linking togetherthe technical and political work by appropri-ate design of interfaces at every stage of theISCRA sequence. The more ambitious the pro-ject, the more important it becomes that thelessons learnt from it be taken on board inthe design of any future projects of similarscope.


Figure 95 on the next page presents an illustra-tion from practice showing interpretations ofthe ISCRA sequence presented in this chapter.As in the case of the illustrations from prac-tice at the end of Chapters 5 to 8, this figureis accompanied by a commentary to explainmore about the context of the European policyprojects concerned.

269

FIGURE

95Illustrations from practice: working with ISCRA



These illustrations from practice have beentaken from the European Commission’s pro-gramme to develop implementation plans forthe Union’s waste management policy. Thereare examples from two of their six projects,each focused on a different priority waste

stream.They are both taken from flip charts that

formed parts of the photo-reports of CoreGroup meetings. They demonstrate clearly themessy process of adaptation that the ’pure’generalised model illustrated in Figure 94 hasto be put through. The circumstances of eachproject can be very different, thus influencingthe processmanagement decisions about howthe model can best be made to work in eachcase.

1. The upper example comes out of the projecton the management of building and demoli-tion waste, which was carried out in Ber-lin and Brussels. The flip charts illustratedwere drawn up during a session which tookplace someway into the project, but not yetas far as half way.The left hand chart shows how the

ISCRA activities have, in some cases, beendoubled up to be handled together in onesession - while, in others, they have beenextended to take place over several ses-sions. This is quite normal, and reflects howdifficult and/or controversial that part of thework is expected to be.Also on the left hand chart it can be

seen that some of the ISCRA activitieshave been programmed to end together sothat the documents can be more stronglycoordinated and released as one. The sev-eral vertical areas which have been outlinednear the middle of the time scale each rep-resent a programme of work for a strategicdiscussion.

The right hand flip chart shows at the topthe simple ISCRA sequence. Indicated on itare theareaswhere theparticipants involvedin the policy work stream are expected tobe more proactive (i.e. in the early stages),and those areas where the technical workwill take over the lead – albeit guided bypolicy input (i.e. the later stages). Below thisis anabstractedversionof the ISCRApatternfromthe lefthandchart.Onthisare indicatedthe areas where strategic discussions areless useful (i.e. the triangles to the left andright), and, by deduction, the areas wherethey aremost useful (i.e. in themiddle).

2. The lower example comes out of the pro-ject on the management of electric andelectronic waste, which was carried out inRome and Brussels. The flip charts illus-trated were drawn up during a sessionwhich took place very early in the projectbefore it had really got started.Here themain concernwas towork out howthe project could be completed in a muchshorter time span than had been the normup till then. This is reflected in the morevertical shape of the ISCRA pattern, whichwas achieved by not only doubling up activ-ities, but ’tripling’ them up in some cases.Another way to finish more quickly was toreduce the work to be done, and the sac-rifice of the on-going management of theplan was considered to be feasible in thiscase. This is reflected in the fact that thereare only six rows as opposed to seven in theother example.

Probably the most important point arisingfrom these examples is the value of the ISCRAmodel as a structure for communication. Itcan serve the project management team wellin their difficult process discussions, whichare necessary when one tries to combine theadvantages of cyclic working with the linearconstraints of most organisations.

271

12 Invention, transformationand interpretation

In our early chapters we took care to present our view of the Strategic Choice ‘toolbox’ asa coherent and balanced whole. Yet we acknowledged from the very first page of our firstchapter that it would be too much for us to expect readers to embrace the strategic choiceapproach (SCA) as a complete new framework for planning or management, without criticalcomparison with what they have learnt from other sources – including, in the case of moremature decision-makers, their own working experiences as managers or consultants.

So it should be no surprise that many people have taken on our ideas in a selectiveway. They have imported features of the strategic choice toolbox to their own practiceswherever it seems to make sense to them to do so, adapting their language to the situationsthey meet where they feel it will better serve their own needs and – often crucially – tothe needs of the stakeholders they are there to help and to serve. In this kind of process,they have introduced differences of many kinds, as we have ourselves.

ADAPTATIONS IN WORKINGMETHODS

As described in Chapter 11, many peoplealongside whom we have worked, especiallyon relatively extensive projects involving inter-active workshop methods, have devised cre-ative ways of combining methods from thestrategic choice toolbox with complementarymethods from other sources – especially insituations where people have not only to learnways of handling the complexities of the issuesbefore them, but also to learn ways of workingconstructively with each other.In each of the five sections that follow,

examples will be given of adaptations to thestrategic choice toolbox that tend to arise atsuccessive stages of project work. These canbe loosely related to the successive stages inthe ISCRA sequence described in Chapter 11;

and also to the strategic choice cycle ofFigure 8.

• engaging diverse participants in identifyingissues;

• distinguishing different levels of choice;• extending the areas of agreement;• agreeing preferred ways forward; and• developing commitment to action proposals.

The chapter ends with a review of thefactors to be considered when customisingthe approach to different situations, followedby a selection of four further illustrations frompractice.

ENGAGING PARTICIPANTSIN IDENTIFYING ISSUES

We have already described, in Chapter 5, ageneral approach to the shaping of problemsthrough building up a shared list of issues of

273

FIGURE

96Setting Levels of Choice

Zone ofstrategicconcern

Choices ofstrategic

orientation

Choices ofstrategic

policy

Zone ofpolicyconcern

Zone ofoperationalconcern

Choices ofoperational

initiative

Choices ofoperational

policy

larg

e

many

dis

tant

long

sm

all

few

near

short

LIF

ET

IME

TIM

E H

OR

IZO

N

GE

OG

RA

PH

IC A

RE

A

TA

RG

ET

GR

OU

P

• The terminology used in this diagram is indicative only. In practice the language should be

adapted so that the stakeholders can relate to it easily.• There can be as many levels as necessary for comfortable working – 2 to 5 is the normal

range, with 3 or 4 being the most common.• This is not a process so much as a framework for analysis and communication. It is best to

start wherever it is most comfortable, and to keep moving up, down or across according to

which is likely to provide the greatest learning experience.

Invention, Transformation and Interpretation

concern – issues which may later become vari-ously reframed as decision areas, comparisonareas or uncertainty areas – or in some casesas assumptions or constraints.When working with diverse participants it

is helpful to start building consensus in thegroup as early as possible in the process. Anopportunity for this lies in the generation ofa first shared list of issues, as this has beenfound to provide a non-threatening means ofenabling participants to begin working interac-tively together.Here it is important that the facilitator should

demonstrate both listening skills and question-ing skills. For example, it is quite commonfor the participants to tend, at first, to formu-late issues in the form of position statements;this can be seriously counterproductive in thesearch for consensus. Here the skilled facili-tator can demonstrate ways in which theseissues can be reformulated in more funda-mental terms such as organisational interestsor personal needs.Over time, this process can have the effect

of making the contributions of different par-ticipants more consistent in their language –with some issues being reframed in termsof broader goals or values, and others interms of more concrete action possibilitiesor contingencies. Some examples of relevantquestions, sometimes known as ‘restructuringheuristics’ (Shakun, 1995), are:

• So why should this be a problem?• What might be the causes of that?• Why might we want to do that?• If that is a goal, how might we get there?

Time spent in this kind of discussion can beimportant in demonstrating to the participantshow their own concerns, which may earlierhave been expressed in quite diverseways andat diverse levels, can be gradually reframedin terms of a shared framework of under-standing. Once in this form, the issues canbe analysed further by clustering similar onestogether and, where there are many, estab-lishing the priorities between them so that a

collective approach can be taken in decidingwhere to begin.Then, more specific concepts of problem-

structuring from the strategic choice toolbox –such as those of the decision area, the decisionoption, the comparison area and the uncer-tainty area – can be introduced gradually andselectively, as a means of enabling people tobuild together on the inputs that they and theother participants have provided.

DIFFERENTIATING LEVELSOF CHOICE

Towards the end of Chapter 6, the notion wasintroduced that it is sometimes useful to for-mulate decision areas at different levels of

generality. This applies particularly in relativelyambitious projects where the participants areresponsible not only for exploring relativelyspecific choices of action, but also broader,more generic policy positions and choices forthe longer term.One of the strengths of the Strategic Choice

Approach is that it enables people to breakaway from more rigid, linear models of plan-ning in which it is supposed that one shouldbegin by setting broad goals and progress-ively work downwards towards a more spe-cific operational level. Yet, where a need isfelt for some kind of differentiation of levels ofchoice, SCA can be adapted to this by offer-ing a framework for checking the consistencyof options in either direction. Only two levelswere suggested in the South Side illustration(see Figure 54): one encapsulating more gen-eral and long-term issues, and the other spe-cific projects and operations. This distinctioncorresponds broadly to the familiar differen-tiation of strategy and tactics in military andbusiness planning.In some extensive projects, it may be help-

ful to distinguish more than two levels inaddressing the complexity of the issues identi-fied. For example, during a 3-year project con-cerned with the development of policy optionsin the preparation of a new generation ofstructure plans that was being introduced in

275

FIGURE

97A Common Grounding Table

Homogenous Sub-Groups

Decision Areas Options A B C D E F

i

ii

iii

i

ii preferences / choices

i

ii

iii

iv

3

2

1

This technique can be used with as few as 10 participants but its real value iswith 30 or 40. There is no real upper limit, but it must be remembered that themore people are involved the more difficult it becomes to find some commonground with which to start building.

The time necessary depends on the number of participants, but it should bepossible to get the main table on the wall in an hour. If it is taking longer thanthat it is likely that the participants are not ready for it yet. The time requiredfor the subsequent ‘give and take’ is much more difficult to judge, and canbe considerable. However it has to go on until agreement is reached, even ifthat agreement is to go back and review the formulation of decision areas andoptions.

The amount of space required also depends to some extent on the number ofparticipants, but also on the scale of the problem. There are no particular spatialrequirements, except that the sub-groups should be able to work privately inthe first round, if possible.

It is most important that the recording is thoroughly carried out at this stagebecause it provides the only record of the consensus agreements.

See also the illustration from practice in Figure 99.

• This type of interaction should only be attempted when the group have had the opportunity to establisha reasonable level of trust through working together – for example in agreeing the decision areas andoptions.

• When the preferences of the homogenous sub-groups have been recorded, and the common groundidentified, the process of extending it begins where the disagreement is weakest.

• If such an extension proves difficult, some form of organised reformulation in the light of the ‘needs ofthe other side’ is most likely necessary.

• See also the illustration from practice in Figure 99.


England alongside local government reorgan-isation (Hickling, 1978), many choices arose todo with the choice of policies either in relationto specific planning topics or of a more fun-damental nature. Because these three zonesof concern – strategic, policy and operational –seemed to overlap, a four-level differentiationwas identified (see Figure 95):These levels were described as:

• choices of strategic orientation;• choices of strategic policy;• choices of operational policy;• choices of operational initiative.

In any other application the number and thedescription of levels can be varied accordingto the range and the complexity of the issuesidentified. In most extensive projects, experi-ence suggests that a four-level split is usuallyappropriate.The definition of the levels of generality

should be agreed, at least provisionally, beforeworking across levels can begin. There arethree scales that recur: space, population andtime, though there can be others that maybear consideration in specific cases. The spa-tial scale refers to the geographic area relevantto the choices being made, which may varyfrom a specific site to a wider region or nationstate. The population scale refers to the targetgroup for the choices being made – probablyhuman, but not necessarily so. This could besimilar to the geographic scale, but is morelikely to vary from an individual or small group,through various sectors, to the total population.The time scale is different because there aretwo aspects that are usually relevant. One is asin the time horizon of the choices being made(when the effects may be experienced), whichmay vary from tomorrow through next year, to20 or 50 years from now. The other is as inthe life expectancy of the choices being made(the length of time before they should be re-examined), which may be similar to the timehorizon, but is most often more short-term.Any set of levels of choice comprises a

framework for problem structuring, and needimply no particular process for working with

them. In a hierarchy of choice, it is oftenassumed that one should start at the top andwork down; yet it has been found equallyrevealing to start at the bottom and work up.In the structure plan project (Hickling, 1978),some teams started working at all levels at thesame time. In keeping with the cyclic processof strategic choice, we would recommendstarting wherever it is most comfortable, andto keep moving up down, or across accordingto which is likely to provide the greatestlearning experience.

FACILITATING EXTENSIONSTO THE COMMON GROUND

Turning to the next stage of ISCRA –confirming concepts – consensuswill graduallydevelop, as the work of a project proceeds, onthe nature of the key choices ahead – in stra-tegic choice language, the key decision areas –and also on the range of realistic optionsavailable within each, although they may notalways be presented as such. An approachcan then be introduced which has come to beknown as common grounding. The purpose ofthis is first to identify and make visible anycommon ground which already exists amongthe various interests represented; then grad-ually to enlarge this. The experience of effect-ive negotiators (Fisher and Ury, 1981) confirmsthat this can be amore productiveway ofwork-ing towards agreement than focusingmore dir-ectly on the differences between the parties.Common grounding is usually used at some

stage well into a project, after a sufficient levelof mutual trust has been built up. In any casethere will have to have been some activityin developing options for mutual gain, lead-ing to some reasonably clear choices to bemade and demonstrating that a minimal levelof consensus already exists.This technique requires that the choices

faced by the participants should be formulatedin the form of decision areas and options as inAIDA, after which homogenous sub-groups ofparticipants (in terms of ‘where they are com-ing from’) are formed to decide which of the

277

FIGURE

98An Elaborated Progress Package

PROCESS

CRITERIA FOR

MANAGEMENT

OF

UNCERTAINTY

NOTES ON THE

PROCESS CRITERIA

What immediate decisions should we take(which may include to do nothing yet; toundertake direct, partial or holding actions;or to make statements of policy)?

R obustness.Flexibility.Changeability.

R obustness is like ‘rugged’.Flexibility refers to keeping futureconnected options open.Changeability is opposed to ‘rigidity’.

What are the crucial assumptions entailed(including successful implementation) which, ifthey turn out to have been wrong, wouldmake the immediate decisions unwise?

Sensitivity. Sensiti vity of the decision to anassumption being wrong is a measureof the degree to which that assumptionis ‘crucial’.

What monitoring system would be sensibleto keep track of how the crucial assumptionsare standing up over time – probably usingindicators?

C ost.Effectiveness.

Cost refers to all types of resource(time; manpower, money, materials).Effectiveness is about the provision ofaccurate indicative information.

What contingency plans will be needed tooffset the negative effects of not having gotthe crucial assumptions right?

Assurance.

Assurance is the level of confidencethat all will not be lost when crucialassumptions are wrong.

Whe re there is no imperative to decide yet,what deferred decisions should be identifiednow (including how long they should bedeferred for)?

Opportunity costConfidence (nowand when the timecomes to decide).

Opportunity cost is what one misses bynot deciding now.Confidence is the increased probabilityof a good decision later.

Which are the key uncertainties (of all threetypes – UE, UV and UR), which will makethese deferred decisions difficult when thetime comes?

Sensitivity.Tractability.

Sensitivity of a deferred decision to anuncertainty is a measure of how ‘key’that uncertainty is.Tractability is the practical ease ordifficulty of reducing an uncertainty.

What research programme would it beappropriate to undertake to reduce the keyuncertainties before taking the deferreddecisions?

Cost.Benefit.

C ost refers to all types of resource(time, manpower, money, materials).Benefit here refers to the expectedreduction of uncertainty.

What performance specifications, to whichdeferred decisions and/or contingency plansmust conform, can be formulated now toinsure a good result?

Consistency.Acceptability.

Consistency relates to statements ofpolicy (see immediate decisions above).Acceptability is for the stakeholders.

• In designing a progress package it will be helpful to first identify: for whom it is being prepared (the decision-taker); what it is aimed to achieve (the strategic orientation); and for how long it should be valid (its expected 'life').• If alternative progress packages are to be put forward the process criteria are likely to play as important a role in comparing them as the substantive comparison areas - but they may have to be explained first.• The practical aspects of the commitment (or progress) package in use are described in Chapter 8 – they are also relevant here.


options they would prefer. The output of thesegroups is then plotted on a simple matrix asshown in Figure 97, and the result then openedfor discussion.Common ground is easily identified where

all the sub-groups choose the same option. Itis also identified where any option is chosenby no sub-group – so there is common groundthat this option is not supported. It is at thisstage that it is quite usual for the participantsto be surprised to find the extent to which theyalready agree.Discussion is then aimed at building on the

common ground, so the focus starts wherethere is least difference in preferences. Typ-ically this will be where only one sub-groupis not in agreement. That sub-group is thenquestioned as follows: ‘Howwould this option,which the others have all chosen, need tobe modified to make it acceptable to you?’Frequently the changes needed are quitesimple: a slight change of wording, some jointmonitoring, or a contingency plan in case of anegative outcome. Themodification is then putto the other sub-groups in the form of: ‘Can youlive with this proposed modification?’ Whenthe changes are genuinely small there is rarelya problem, and once again consensus is iden-tified. Where the changes required are moredifficult to agree, and where the differencesare more evenly split between the sub-groups,there is a clear focus for further work. Here, forexample, a decision area can be set aside untillater, and another one explored. When only afew decision areas remain which require fur-ther work, the experience of working togethersuccessfully will often have had the effect ofenabling options to be reformulated in a cre-ative way so as to resolve the outstandingdifficulties.Where difficulties prove harder to resolve,

one way of proceeding is to invite the partici-pants to join one or other of two competingteams, each with its own appointed ‘captain’.Each team is then set the challenge of con-sidering: ‘What are the needs of the otherside?’, and reformulating its preferred optionas an option for mutual gain.

At this point, the principles of managinguncertainty covered in earlier chapters becomerelevant; for they can help people to probe theconflicting assumptions which often underlieapparent disagreements. Inter-group discus-sion can then help to reduce uncertainty bothabout the different values in operation, andabout the future intentions of relevant parties –leading in many cases to agreement on appro-priate exploratory actions or contingency plans.

FRAMING A WAY FORWARDIN THE FACE OF UNCERTAINTY

Chapter 3 concluded with a frameworkfor managing decisions under uncertaintyunder the name of a commitment package

(Figure 26). This term has worked well in manydifferent situations, but has been found to bemore difficult to use when more people, andthus more different opinions, are involved. Ithas also been found to need elaboration if itis to be seriously helpful in aiding continuity.As mentioned earlier, we now tend to speakmore often about a progress package, whichaddresses some of these issues.The basic principles are unchanged. How-

ever it can often be important to expandthe various elements so that they can beaddressed more explicitly, and interrelated ina more structured way. One format for doingthis is shown in Figure 98. The two main sec-tions remain, but now they are clearly related toeither the decisions to be taken now, or thosethat are to be deferred. The third section ismerely included to clarify some of the qualitycontrol content which was previously hiddenwithin the cells of the table.This more explicit structure allows clarifi-

cation of the role of process criteria for themanagement of uncertainty. Aspects such asrobustness and flexibility can be used moreeffectively, and others, such as tractability, canbe introduced. This is not to say that theyshould not be handled alongside, and in thesameway as, the comparison areas in a regularcomparison table if that is more comfortablefor the participants, as illustrated in Figure 24.

279


However it is more usually a case of exploringdifferent aspects of comparison, and interpret-ation of assessments using substantive com-parison areas in the light of uncertainty. Thiscan be especially productive when the levelof conflict in the group begins to undermineany progress which is being made towardsconsensus.This focus on uncertainty and its manage-

ment in the form of a progress packageis important because it opens the way forthe constructive management of conflict. Notonly does uncertainty cause conflict – mostlythrough the parties making different assump-tions – but the stresses of conflict can causeuncertainty. These are the classic ingredientsof a vicious circle. In an extensive project,if one has the means for effectively man-aging uncertainty, conflicts can be handledconstructively – and their negative symptomsmade to disappear.

SECURING COMMITMENTTO ACTION PROPOSALS

At a stage when proposals for action start tobe developed in some depth – often throughthe work of sub-groups charged with gen-erating these within specific policy areas orfrom the point of view of a specific group ofstakeholders – it becomes important to subjectthese to critical scrutiny, and to encourage con-structive amendments or counter-suggestionsby members of other sub-groups. At this stageof a major project it has been found helpful touse consistently structured action forms witha prescribed sequence of headings such as:

• Description of the action• Responsibilities• Funding and Cost/Budget• Planning (uncertainties, conditions, etc.)• Crucialassumptions/‘Criticalsuccessfactors’

A structure such as this offers a means of rela-tively easy access to what can be a consider-able amount of detailed information, so that itcan be shared among all the members of thewider project group – even in circumstances

where conventionally it might be thought moreappropriate to use such concepts explicitly onlywithin the core group. It also provides a way ofintroducing the set of ideas about the manage-ment of uncertainty and the design of commit-ment (or progress) packages which have beendescribed above, thus releasing the potentialfor managing conflict which that allows. Someillustrations at the use of action forms in prac-tice are presented in Figure 101 at the end ofthis Chapter.One technique which can be used to bring

the proposals of the different sub-groupstogether for display is to create what hasbecome known as an action gallery – enablingmembers of other sub-groups to comment andadd any further suggestions of their own. Thisprocess can take various forms – including thatof the carousel described below – all based onthe need to allow free interaction between theparticipants in order to achieve a high measureof pre-coordination, as described in Chapter 9.The principle is that of a picture gallery, inwhich the action forms are displayed in group-ings which may be defined in various ways –for example, according to those with primaryresponsibility for implementation; according tothe relevant issues; or accroding to who isfunding it.

OPPORTUNITIES FORSOCIO-TECHNICAL DESIGN

At all stages of a project, opportunitiesarise for exercising creativity not only in themethods by which people address the sub-stance of the work, but also in the methodsby which they are helped to work witheach other. So the challenge can be con-ceived as one of socio-technical design (Tristand Murray, 1993). The choices open caninclude: (1) devising completely newmethods;(2) adapting and combining established tech-niques; or (3) making a myriad of spontaneousadjustments as particular challenges presentthemselves.

• The Carousel – An example of the first ofthese categories of socio-technical design

280


involves a way of structuring interactionsbetween members of different sub-groupswhich has become known as the carousel.This has shown itself to be effective at thestagewhere sub-groups have been assignedto differentiated but parallel tasks withinthe context of a strategic discussion event.This approach involves first inviting the sub-groups to work on allotted tasks in differentcorners of a large room, or a suite of adjoiningrooms, using flip charts as ameans of record-ing their progress. Then a schedule is agreedwhereby the members of each sub-groupvisit the territory of each other sub-group inrotation, inspecting the results of their work,commenting spontaneously on what theysee, asking questions on anything whichappears controversial or unclear, and addingtheir own ideas and opinions if they wish.It is essential that one or two members of

each ‘home’ sub-group should stay behindto answer questions, to make notes on thepoints raised and to report back to the othermembers of their own sub-group once theyhave completed their tour and returned tobase.

In the second category of socio-technicaldesign, there are four well-established andreasonably well-known methods which haveproved to be fully consistent with the strategicchoice approach. Either in whole or in part,these form basic tools in the toolboxes ofmanyfacilitators. They are:

• Metaplan (Schnelle, 1973) – This is a com-plete approach in its own right, developedby the Quickborn Group in Germany in theearly seventies; however two techniqueswithin it have a general application and canbe used independently. The first is a form ofbrain-writing in which the participants writetheir ideas on small pieces of paper prior tosticking them on the wall for further ana-lysis – usually a form of clustering. Thesecond is the use of small coloured stickyspots as a means of collecting opinionswithin a group – whether that is a matter ofprioritisation in a list of items or an evaluation.

• Brainstorming (Rickards, 1974; van Gundy,1981) – This is probably the best-knowntechnique, although the word is often usedwith very little understanding of what itactually means. There are many derivativeforms, and it is considered to be the fore-runner of most idea-generation methods.Nevertheless, application of the rules of theclassical form is still the best way of helpinga group to be creative while also maintaininga sense of discipline.

• Nominal Group Technique (Moore, 1987) –There are two aspects of this technique(usually abbreviated to NGT), which haveproved valuable in association with strategicchoice methods. The first is the use ofsub-groups working together in the sameroom, provided that their tasks are kept thesame and quite simple. This allows intensiveinteractive working – even for very shortperiods of a minute or so, without the wasteof time caused by the sub-groups having tochange rooms frequently. The second is asimple sequence starting with the gener-ation of ideas (e.g., issues, options oruncertainties) in small groups, leading to theformation of a mutually understood joint list,then discussion of the relative importanceof the items on the list – often using thesticky spots technique from Metaplan.

• Open Space Technology (Owen, 1985) – Thisis a way of enabling people in large groups towork effectively together. It was developedabout 1985 by an American named HarrisonOwen, to enable people to come together,often in large numbers and usually repres-enting enormous diversity, to pool their ideasfor creative and collaborative action. Theevents generally last from one to three days,and can involve 10–1000 or more people.The secret of Open Space Technology is

that it is based on the passion and responsib-ility of the participants. All that is provided isa structure, a process, and logistic support.Thus after the opening phase, in which theagenda is devised collectively, manage-ment and performance of the process iseffectively given over to the participants

281


themselves. This includes the means ofreporting in a very short space of time.It is a pre-requisite that the central

theme must be of genuine concern to allinvolved. This provides an essential focusfor maintaining the relevance of diversecontributions, and is best formulated withan action orientation

It is in the third area of socio-technical design –that of spontaneous adjustments – that com-binations and variations of these techniquesare often devised. There are many choices tobe made, not only about which techniques touse when, but also about how they are to beintroduced. Questions of who does what withwhom; the sequence and duration of activities;and adjustments to the physical layout of theroom(s) can often be all-important in sustainingthemomentum and spirit of participative work.For example, sub-groups can be formed in

different ways – either to be as mixed as pos-sible in theirmemberships or to bemore homo-geneous in terms of the interests or disciplinestheyrepresent.Thetasks assigned to each sub-groupmay be the same; variations on a theme;or completely different. Also, it is possible todraw on approaches which combine elementsof individual, sub-group and plenary work, notonly to demonstrate how each participant canmake a full contribution from the start, but alsoto start to build on the contributions of others.These are choices which should not be madelightly because the outcomes can vary accord-ing to continually changing circumstances.

CUSTOMISING THE STRATEGICCHOICE APPROACH

The various concepts, processes, techniquesandmethodspresented inthischapterasexten-sions and adaptations to the strategic choiceapproach have been devised in response to theneeds of those involved. They cover, betweenthem, all the aspects of the A-TOPP model −Approach : Technology, Organisation, Processand Products – often touching on two or moreof these in combination. But they cannot be

used all at the same time, and individually theyare often not enough – so they have to be insome way ‘orchestrated’. This is a process inwhich the all-important twin ideas of selectiv-ity and adaptiveness which were introduced inChapter 9 become a matter of even greaterconcern.Although some of the people with whom

we have worked have adopted the approachvirtually in its entirety, others have seen a casefor reshaping it to their own contexts, introdu-cing some of the adaptations described abovealongside any other favoured methods withwhich they personally feel comfortable.For example, civil servants who adapted the

approach in theWaste Directorate of theDutchMinistry of the Environment in the 1980swere encouraged to merge SCA with some oftheir own ideas to produce an approach calledStrategie Afvalstoffen (STRAF). The phrasetranslates simply as Waste Strategy – thoseconcerned also being aware that the STRAFabbreviation also means punishment in Dutch!Not long after this Bram Breure, a civil servantwho was later to become a private consultant,generalised it to Interactieve Beleidsvormingen Implementatie Systeem (IBIS) – meaningInteractive Policy-Making and ImplementationSystem. Another variant, developed with HansKnikkink and presented to the World Bankin Washington, was called the CollaborativeAction Planning Strategy (CAPS). Some morerecent initiatives in interpreting the SCA for par-ticular types of client situation – for instanceas Strategic Action Planning (SAP) and Cross-Organisational Learning Approach (COLA) – aredescribed by contributors to Chapter 13.In most of these cases, the approach as

presented clearly has some differences fromSCA, and their advocates feel the need torecognise this. In any case, their sense of own-ership is increased with the new name thatthey introduce.We welcome such continuing exercises

in interpretation. For we recognise that weourselves in the early days did not find iteasy to settle on the phrase ‘Strategic ChoiceApproach’ to describe our toolkit; and we have

282


since become increasingly aware of how itlends itself to differing interpretations, not all ofwhich reflect the spirit in which it was originallyintroduced.One difficulty we have had to face is that

the English language phrase ‘strategic choice’is itself in common usage, with several subtlydifferent interpretations. We may ourselveschoose to interpret the adjective ‘strategic’ asindicating a (selective) awareness of connec-tions between decision agendas, at whateverlevel of organisational or inter-organisationalchoice. However, we should not be surprisedwhen other people relate the adjective to thelevel of choice to be found in the board-room of the corporate enterprise. We find alsothat the three-word phrase ‘Strategic ChoiceApproach’ is too long to use frequently in dis-cussion, without shortening them either tojust ‘Strategic Choice’ or an acronym such as‘SCA’. Although the use of ‘Strategic Choice’to describe our particular approach tends tobe readily accepted by students, we find thatit can become more problematic when intro-duced into communicationwithmanagerswhowill often have their own long-acquired senseof what that phrase implies. So we eitherbecome drawn back to the acronym SCA, orstart searching for alternative terms.

ADAPTING LANGUAGETO CONTEXT

So what are the basic needs that have to besatisfied by those who wish to interpret ourapproach – or indeed any other such approach?We have mentioned ‘ownership’ from time totime – especially in relation to the products of aproject where its value lies in the commitmentthat comes with it. But the question is morefundamental than that. Yes, having a sense ofownership of the toolbox in use is helpful, andalmost certainly reassuring. However themorefundamental need is to feel confident whenoperating the approach with and on behalf ofothers – often clients of one sort or another.This seems to be a key factor that has

led people to change the names of particular

methods and tools that we have introducedhere– includingsometimessuchbasic termsasdecision area, comparison area and uncertaintyarea. We have always encouraged this sortof ‘customising’ to different situations, recog-nising that it may be less to do with the con-fidence of a facilitator – although it can allowthem some measure of self-expression – thanwith the confidence of their clients. If these cli-entsarepresentedwithanapproachcouched interms with which they are uncomfortable, theprobability of acceptance is likely tobe reduced.If on theother hand, after a process of interpret-ation often involving the clients themselves,more sympathetic terms can be developed, astronger basis for acceptance can be built.For example, there are some situations

where an emphasis on the management ofuncertainty can be resisted as seemingly nega-tive, yet the principle behind it may be morewidely acceptable when replaced by moreaction-directed language. Then we have grad-ually come to replace the phrase commitment

package by the more dynamic term progress

package in most – but not all – contexts. Forthe focus on commitment may be a welcomeone where a group is vested with authority tomake decisions; yet where its members areonly in a position to make recommendations toothers, the emphasis on progressmay becomea much more acceptable alternative.Whenever users of our approach have

expressed interest in taking it on board as facili-tators, we have encouraged them to modifyit as they see fit. The hope has always beenthat, if the modification works, we will findout about it and SCA may evolve accordingly.The development of SCA through this form ofaction research has been a way of life for us,for what is nowmany years. Along the way wehave been helped wittingly, and sometimesunwittingly, by very many others.

INSIGHTS FROM OTHERLANGUAGES

Theexperienceofseeingourdevelopingvocab-ulary of strategic choice translated fromEnglish

283


into other languages – both European and non-European–has itselfbeenan important learningexperienceforus.Asourexperiencehasgrown,we have come to question the value of someof the basic terminology we presented in ourfirst edition – most of it carried forward intothe second. So how far should we introducechanges now, in our third edition, based on ourexperiences in subsequent years?Frequently, we have found ourselves learn-

ing important lessons from translations ofour terminology into other languages, whichsometimes seem to express the underlyingideas better than our native English. A goodexample is the translation of commitment

package as pacote de compromisos in bothSpanish and Portuguese – conveying a senseof ‘promising together’ which becomes dis-torted in English once other less-benign inter-pretations the English word ‘compromise’ areintroduced.

INVITING VIEWPOINTS FROMOTHERS

In the preceding chapters, we have presen-ted the Strategic Choice Approach as viewed

through the eyes and the experiences of thetwo of us who have played central roles indeveloping the ideas, and in introducing it toother people, over more than three decades.In this third edition, we believe that the time

has come for us to introduce the voices ofsome of themany other people with whomwehave worked – especially over the years sinceour second edition was published in 1997 – asthey can bring fresh and varied perspectivesto bear on present practice and its implicationsfor the decision-making of the future.So, in the spirit of learning from the experi-

ences of others, and in thewider spirit of actionresearch that we have always embraced, wehave invited fifteen short contributions fromtwenty-one people with whom at least oneof us has worked closely, and who seem tous to have important things to say. In doingso, we suggested that they tell us aboutwhat they have learnt from their experiencesusing their own preferred terminology, sothat the diversity of interpretations, transform-ations and translations of the ideas and experi-ences we have presented here can be morewidely shared. These contributions are broughttogether in Chapter 13.

284



The final pages of this chapter present fourillustrations from practice of the introduction ofthe methods discussed in this chapter. As inthe case of the illustrations from practice at theend of Chapters 5–8, each of Figures 99–102,starting on page 286, is accompanied by a com-mentary to explain more about the context ofthe project or projects concerned.

285

FIGURE

99Illustrations from Practice: Carousel Techniques in Action

1 Croatia 1997 Exploring organisational futures in post-war Bosnia (Mercy

Corps / Scottish European Aid)

2 Botswana 1989 Planning faster availability of land for housing (Ministry of Land and

Resources)

3 Latvia 1996 Developing the National Environmental Action Programme (Ministry

of Environment and Regional Planning).



The basic carousel technique is described inpage 286, but there are several variations of it.Together they form a suite of techniqueswhichcan be used in various combinations. Theyprovide a means for keeping large amounts ofinformation visible as a basis for structured col-laboration between participants.The Promenade This is a much less struc-

tured variation of the carousel. Different tasksare carried out in different parts of a largeroom – or different rooms which would, how-ever, be ideally inter-connected. For example,there might be several metaplan exercisesgoing on, each concerned with a different sub-ject (various issues, aspects of the work, etc.).The participants move around contributing tothe work as and when they are best able andmotivated to do so.The Picture Gallery This is a way of shar-

ing with each other the work done separately,either individually or in sub-groups. It is con-ducted, as the name implies, rather like an artfair or gallery in a large room with much wallspace or screens (separate rooms can be usedwith some loss of ‘flow’). Each of the individu-als or groups hangs summaries of the work ithas done on the wall – ideally in a form whichis easily accessible to others; sometimes in apre-arranged format. A table, and/or half-circleof chairs, is placed in front of them and thisbecomes the focus of that group’s activities.Participants then walk around learning aboutthe work that has been done. They may sug-gest new ideas, or alternatives to those alreadyon display. The Action Gallery is a special caseof this technique.There is a second phase of this process

which is an optional extra – used speciallywhenworking with a group of professionals. Inthis the groups are requested to remove theirpapers from the wall and to redistribute themaccording to a completely different classifi-cation, such as specific issues, actors, targetgroups or impact areas. The groups are thenre-formed according to the new classification –grouping like specialists together. Their task is

to sort out the new set of ideas or proposalson the wall before them. This can be an idealstarting point for the Trade Fair technique.The Trade Fair This allows free exchange

of ideas between groups who will have setthemselves up as in the Picture Gallery. Par-ticipants seek to improve their proposals bynegotiating with others to eliminate the threeproblems which can occur when groups havebeen working separately:

1. duplication in which two (or more) groupspropose the same thing – often in differentway – a straightforward waste of effort;

2. contradiction in which two (or more)groups propose things – often for differentpurposes – which at best cancel each otherout and at worst lead to seriously negativeresults;

3. omission in which two (or more) groupseach think that the other is going to pro-pose something – leading to neither of themdoing it.

The Guided Tour Sometimes known as the‘travelling plenary’ – this is a form of sharingthe work done in sub-groups, which is halfwaybetween the Picture Gallery and conventionalpresentations. Starting with a room, or rooms,set up as for a Picture Gallery, Promenade orCarousel, the groups are shepherded from onegroup’spositiontothenext.Ateachpositiontheyare given a run-down onwhat is there. It is veryusefulwhentherearemanypaperstobeviewedandmoving themaroundbecomesahassle.The three photographs in Figure 99 are all

of different ways of working with the carou-sel suite of techniques (although photographsare notoriously uninformativewhen being usedto show process). The top one (1) is a reason-ably conventionalCarousel. As is frequently thecase, the rooms are not ideal, but the groupscanbeseenworking intheircorners.Themiddlephotograph (2) shows aPromenade in progresswhere the participants are clearly operating intheir ownstyle andspeed. Thebottomexample(3) is a much more structured variation whichutilises the ‘sticky dot’ prioritisation systemborrowed from the Metaplan technique.

287

FIGURE

100Illustrations from Practice: Building Common Ground



The common grounding technique is illustratedin Figure 100 through photographs of someof a set of flip charts from a European Unionproject concerned with implementation of theEuropean waste management policies – in thiscase for the priority waste stream for healthcare waste, carried out in the mid-1990s.The flip charts to the left show some of

the key decision areas and options, as agreedthrough lengthy dialogue part way through theproject. The columns at the right show the pre-ferred options as identified by five sub-groupswhich, at that stage, were formed to representdifferent, and relatively homogeneous, sets ofinterests such as directorates of the EuropeanCommission, member states, industry, wastemanagement companies, hospital and otherhealth managers and the environment lobby.As often happens in such circumstances, the

participants were surprised to find the extentto which they already agreed: there were afew decision areas where there was completeagreement, as well as some options whichnone of the groups had selected. The processof developing common ground then continuedby focusing on decision areas where there

was a considerable measure of agreement, forexample, where there was only one sub-groupout of line.That sub-group was then asked ‘In what

way would this option, which the other sub-groups have selected, have to be modifiedto make it acceptable to you?’ Often, slightchanges were all that was required. Each ofthe other sub-groupswere then asked the clas-sic consensus-building question ‘Could you livewith that?’ If the answer was positive, then thecommon ground was thereby extended.Where there was difficulty, the principles

of managing uncertainty became helpful inprobing the conflicting assumptions that oftenunderlie apparent disagreements. Inter-groupdiscussion then helped to reduce uncertaintyboth about the different values in operation,and about the future intentions of relevantparties – leading to agreement on appropriateexploratory actions or contingency plans.Where difficulties proved harder to resolve,

only a few decision areas remained whichrequired further work. Here the experience ofworking together successfully had the effectof enabling options to be reformulated in acreative way so as to resolve the outstandingdifficulties.

289

FIGURE

101Illustrations from Practice: Use of Action Forms

Theme: The use of action forms in developing draft and final documents (Action Plan).

Project: 1. Action planning for European waste management: used tyres.

2. Preparation of the National Environmental Action Plan for Latvia.

3. Work document for the Netherlands national environmental policy plan.

Context: Various national and international multi-organisational project groups.



There are three projects illustrated here. Theintention is to reflect different stages of thechoosing process and the way concepts wereadapted to the situation ‘on the ground’ for thespecific client at the time. Each project will bedescribed in turn, in an order which reflectsdifferent stages of development.

1. The first example shows an early attempt toenable a multi-organisational project group,set up in 1993 by of the European Com-mission, to produce an action plan as partof the Implementation Strategy for the ‘Pri-ority Waste Stream’ of Used Tyres. Thiswas towards the end of a 30-month projectwhich had demonstrated all the problemsof being the first in a series.The Project Group had difficulty in adapt-

ing to the new style of working, and therehad been a long settling in period. The facil-itation team were also on a steep learningcurve, as they had previously only experi-enced this kind of working at national level –the increase in scale being not a little chal-lenging. All participants had been asked tofill in formswhichwould then form the basisof the proposed Action Plan element of adraft strategy. Thisworkwas done betweenstrategic meetings and assembled later bya technical support team. As it happened,little more work was done on theseforms and they were submitted withoutthe benefit of sustained analysis andcomparison.

2. The second example is taken from the workof an inter-governmental project group setup in 1996 by the Ministry for the Environ-ment and Regional Development in Latvia.

While the language here is Latvian, theboxes and their content are similar to thosein the first example.At this early stage in developing an envir-onmental strategy in Latvia, it was decidedthat this first attempt would be restrictedto the Environment Ministry. Accordinglythese forms were first filled out by thestaff of the regional offices, then brought toan Action Gallery event (see page 290) inRiga, prior to being put through a processof integration to produce a first draft plan.This example shows a format adopted inthat second stage of development just afterthe Action Gallery, after word-processingfor presentation purposes.

3. The third example is chosen to demon-strate how an action form can end up bybeing presented in a published Governmentdocument – in this case the so-called WorkDocument of the second National Envir-onmental Policy Plan for the Netherlands.It is the work of a project group of theNetherlands Directorate General for theEnvironment in 1994.The document contains 125 actions

grouped in each case according to theprimary actors – described as ‘targetgroups’ in the plan (e.g., Agriculture;Industry; Central Government). It alsoshows the current status of all the actionsagreed in the first National EnvironmentalPolicy Plan. Although the presentation hereis more sophisticated, the various headingsare similar to those used in any typicalaction form. The amount of informationunder each heading is limited according tothe intended readership, but referencesare provided to enable more detailedinformation to be located.

291

FIGURE

102Illustrations from Practice: Managing an Extensive Environmental Project



Figure 101 illustrates the importance of invest-ment in creating and maintaining an appro-priate context. It does so by summarising allthe stages that were involved in the designand management of an important extensiveproject – in this case, involving assistanceto the government of Latvia from the gov-ernments of the Netherlands and Swedenin developing its first National EnvironmentalPolicy Plan.This summary – taken from an official pro-

ject document – begins at the time whenit was agreed to set up the project in thespring of 1994, omitting details of variousearlier exploratory discussions in 1993. How-ever, the political complexity of the project iswell reflected both in the arrangements pre-ceding the series of three workshops heldin the summer and autumn of 1994, and inthe events leading up to the final adoption

of agreed policy proposals. Also, the cross-cultural aspects of the project are reflected inthe planned transition from Dutch to Latvianco-facilitators after the first workshop, suppor-ted by appropriate training events.The success of this projectwas to lead in turn

to another in which the focus is on the NationalEnvironmental Action Programme. This isan ongoing programme which deals withimplementation of the National EnvironmentalPolicy Plan, and is reviewed and updatedregularly.Similar projects have been launched in

other Baltic and central European states,each of them providing important learningopportunities. The experience of each projecthas contributed to the design of further extens-ive projects in response to expressions ofinterest in other countries, relating to otherfields of policy. So the process of developmentgoes on, in pursuit of the ‘holy grail’ of anylearning process – continuous improvement.

293

13 Learning from others

This chapter enables several other people who have been using the Strategic ChoiceApproach in their practice to tell us what they have learnt from their experiences. Thedevelopment of the Strategic Choice Approach has always been based on this style ofdiscovery: action research. As we use the phrase, this means working with practitioners,and together sharing the learning from doing so, then together working out how it can bebest codified for the benefit of others.

LEARNING ABOUT PROCESSCHOICES

We have found it helpful to see all decision-making as a learning process. This is in keepingwith the cyclic nature of our view of decision-making as developed in Chapters 1 and 4. Inthis respect the learning model of Kolb (1984),which was being developed in parallel to ourwork in the early 1970s (Hickling, 1974), offersa valuable point of reference.

Up to this point in this book we havebeen primarily addressing decision-making asit relates to the substantive problem beingworkedon.However inourownaction researchprogramme, focused as it always has been onthemethodology, it has been necessary to con-centrate more on the decisions being takenabout the process itself. For example: Whichmode ofwork shouldwemove to next? Shouldwe use a cost-effectiveness-based compari-son or one based on ends-means? How manydecision areas can we address effectivelyhere? Would it be sensible for us to coalescethese decision areas at this stage? What formof progress package is appropriate in this case?What size sub-groups would be sensible forthis next part of the work? How are we goingto help those stakeholders outside the pro-cess to catch up? Such questions relate closelyto the areas of concern addressed by Argyrisand Schön (1973) in their concept of double

loop learning. Although that is concernedmorewith organisational learning, we see a clear linkbetween their work and ours.

ACTION LEARNING

It seems appropriate here to refer to the ideasof Revans (1983) in his concept of action learn-

ing, which is based on the value of practition-ers learning from each other, and has beena major influence on our decision to add thischapter. The process of Revans, in which smallgroups form to learn together, clearly has a dif-ferent focus to ours; yet we see the aims ofthis chapter as being entirely consistent withthe basic notion.

So we asked people with whom we haveworked before – though not necessarily on theprojects they describe – whether they wouldlike to be involved. Of course there were manymore people we could have asked, but therewere limits in the extent to which we couldadd new material in this third edition, so wehad to be highly selective. Using criteria suchas geographical spread, diversity of problemsaddressed and relevance to the twenty-firstcentury, we produced a shortlist. We aremuchencouraged by the 100 per cent response wereceived.

We challenged our contributors to reflecton their decision-making about process

295


management in practice, so as to identify whatthey had learnt from their experience, whichmight be of value to others using the approach.We asked them to select the most significantlessons they had learned, whether positive ornegative, and to present them using examplesfrom their experience.We asked them to focuson up to three lessons, and to aim for a targetof 1800words or less, including an introductionand biographical notes.

SELECTIVE READING

We recognise that readers of this chapter mayfind it more rewarding to begin by selecting

those contributions that appear most relevantto their interests, in the spirit of our Quickaccess guide at the start of this book. Sowe begin with a listing of all our contributorsand the titles of their contributions. We havearranged these in the sequence of their firstworking contact – ranging from the early 1970sto the early twenty-first century – with eitherof us, the chapter editors.

Each contribution starts with a shortresumé of what is in it, with biographical noteson the contributor(s). So we suggest you pickout first one or two titles which attract you,then turn to the relevant page where you willfind the resumé and biographical notes.

LIST OF CONTRIBUTIONS

13.1 Controversy on the Streets: Stakeholder workshops on a choice of carnival route

Jonathan Rosenhead, Department of Operational Research, London School of Eco-nomics, UK

13.2 Communicative Learning, Democracy and Effectiveness: Facilitating private – public

decision-making in Sweden

Knut Strömberg, Department of Urban Design & Development, Chalmers University,Goteborg, Sweden and Jaan-Henrik Kain, Built Environment & Sustainable Development,Chalmers University, Goteborg, Sweden

13.3 Delving into the Toolboxes: National environmental policy-making and strategic

choice

Frans Evers, former Director of Strategy, Ministry of Environment, The Netherlands

13.4 Less is More: Controlling decision complexity when using strategic choice

Richard Ormerod, Professor Emeritus, Warwick Business School, Coventry, UK

13.5 Feet on the Ground: Engaging planning students with political realities

Arnold van der Valk, Land-Use Planning Chair, Wageningen University, The Netherlandsand Gerrit J. Carsjens, Department of Land-Use Planning, Wageningen University,The Netherlands

13.6 Designing Electronic Support: Empowering decision-makers through software for

strategic choice

Dave Friend, Software Design Consultant, West Yorkshire, UK

13.7 Neighbourhood Renewal in Rome: Combining strategic choice with other design

methods

Alessandro Giangrande, Professor of Planning, Universita Roma Tre, Italyand Elena Mortola, Professor of Urban Design, Universita Roma Tre, Italy

296

Learning from Others

13.8 Fast Forward with Strategic Choice: Mutual consulting for small enterprise

development

Rebecca Herron, Community Operational Research Unit, University of Lincoln, UKand Dennis Finlayson, International Development Consultant, Derbyshire, UK

13.9 The Plutonium Predicament: Managing conflict through strategic action planning

Richard Harris, Independent Process Consultant, East Sussex, UK

13.10 Dealing with the Dumps: Using decision areas and levels in developing national

policy for used tyres

Rob Angell, Environmental Policy Consultant and Facilitator, Somerset, UK

13.11 Cross-organisational Learning: Sharing insights from managing major construc-

tion projects

Mike Cushman, Department of Information Systems, London School of Economics, UKand Alberto Franco, Warwick Business School, University of Warwick, Coventry, UK

13.12 Capacity Building in Venezuela: Using strategic choice methods with policy-

makers and students

Elisenda Vila, Planning Consultant and Professor, Caracas, Venezuelaand Ana Maria Benaiges, Planning Consultant and Professor, Caracas, Venezuela

13.13 Differential Learning: Managing different rates of progress between participants

in extensive projects

Brendan Hickling, Independent Facilitator and Mediator, Warwickshire, UK

13.14 Building Commitment in a Rural Community: Use of commitment packages in

empowerment through information technology

Jackie Phahlamohlaka, Community Education Project Chair, Mpumalanga, South Africa

13.15 Commitment is the Key: Building inter-agency agreement over the future of an

historic estate

Leny Bregman, Environmental Project Manager in Natuurmonumenten, The Netherlands

297

13.1 Controversy on the Streets

Stakeholder workshops on a choice of carnival route

By Jonathan Rosenhead

This contribution describes the application of the strategic choice approach and relatedmethods in debating the future of the Notting Hill Carnival, a massively popular but risk-prone street festival in West London. A team from the London School of Economics under-took two successive projects in 1998 and 2001, focused on the use of problem structuringmethods in the management of complex risk. This involved facilitation of workshops withstakeholder groups whose relationships varied between wariness and antagonism. Thesegroups included local government agencies, emergency services, local residents, transportproviders, arts funders and carnivalists. The author reflects on how far the methods helpedthe stakeholders to make progress.

Jonathan Rosenhead has been on the staff of the London School of Economics since 1967,and was Professor of Operational Research there when these projects took place. His long-standing interests are in problem structuring methods; in supporting the decision-makingprocesses of community groups; in appropriate analytic work in developing countries; andin health service planning and service delivery. His seminal work on robustness analysisin sequential decisions under uncertainty influenced the early development of the strategicchoice approach, which he has taught and applied over many years. He has edited a bookin which strategic choice is presented alongside other related methods for the interactivestructuring of complex problems. (Rosenhead and Mingers, 2001)

THE NOTTING HILL CARNIVAL

Each year the Notting Hill Carnival attractscrowds of up to a million revellers for two daysin August to a relatively small area of WestLondon. The first Carnival took place in 1958 aspart of the response to riots in which violencewas directed against the black immigrant pop-ulation; then the carnival was small and spon-taneous; now it is suffering the problems ofsuccess.This account is of two linked attempts

to assist in resolving some of these prob-lems with the chief stakeholders. They were

remarkable for the number of these stakehold-ers, the high political profile of the process, andthe prevailing level of tension, conflict and tur-bulence. Could SCA and relatedmethods reallyfunction effectively in this environment?The Carnival is a unique cultural street fest-

ival with its roots in the Caribbean, whereit was based on the experience of slavery.The freedom to walk the streets represen-ted symbolically the lifting of that oppression.In Notting Hill the Carnival procession wendsits way through the network of streets ofthat neighbourhood with music and costumebands. There is a wide range of associated

298


activities off the actual route, involving streettrading – especially ethnic foods – and staticsound systems.Carnival has often been controversial. At

times it has been threatened by significantcriminal activity, at others by hostile confront-ations with the police. By the late 1990s thesehad become historical, but the problems ofsuccess were evident – a congested circu-lar route with the threat of injuries or deaththrough crushing; the move from a largely eth-nic minority base to a mass audience attunedto contemporary commercial culture; the pre-carious reliance on sponsors for financial viab-ility; and the tension between the desire fora safe carnival with all its disciplines and thesymbolic celebration of the spirit of misrule.

THE INITIAL WORKSHOPS

My involvement with Carnival arose initiallythrough a research project funded by theEconomic and Social Research Council that Icarried out with Tom Horlick-Jones, also asso-ciated with LSE though based in Cardiff. Thatwork seemed to have run its course when theissue that we had been dealing with flared upagain, and we were again drawn in – this timefunded by the BP Complex Risk Programmeadministered by the Centre for the Analysisof Risk and Regulation at LSE. Both projectswere designed to explore the scope for theuse of problem structuring methods (PSMs)(Rosenhead and Mingers, 2001) in the man-agement of risk. The work of the first projectis described in more detail in Horlick-Jones,Rosenhead, Georgiou, Ravetz and Löfstedt(2001).Tom supplied the expertise in ‘risk’ and also

in ethnographic methods (Hammersley andAtkinson, 1995). These we used during the1997 Carnival cycle to gain insights into theinter-organisational dynamics at work.During the 1998 Carnival cycle we negoti-

ated the participation of the four most signi-ficant organisations involved in Carnival in twolinkedworkshops, somoving on from researchmode into action research mode. The four

organisations were the Metropolitan PoliceService; the Royal Borough of Kensington andChelsea; the Tenant Management Organisa-tion, representing the residents in social hous-ing in the Carnival area; and the Notting HillCarnival Trust. Each was represented by its‘top brass’ – for example the Commander incharge of policing major events in Londonand the Leader of the Borough Council – aswell as supporting personnel. The aim of thegathering was to do some strategic thinkingabout the future directions of the Carnival. Wewere told in advance that we had just the oneday. However at the end of the day, the par-ticipants spontaneously started talking aboutthe arrangements for a follow-up workshop.The methods used in these first workshops

were not restricted to SCA. In fact, being underextreme time pressure, we started with a sim-plified version of Soft Systems Methodology,on which we received helpful advice fromPeter Checkland.1 Our purpose was to bringto the surface the different possible objectivesthat people might have for staging Carnival –important as three of the four organisationswould prefer Carnival not to happen! Start-ing from these possible objectives we elicitedmodified or new actions that might be relevantto achieving them, and the uncertainties asso-ciated with those actions. Following this dis-cussion we prioritised a number of issues forexamination in depth.That was the first workshop. It was in the

second workshop one month later – and witha slightly less senior membership – that weemployed SCA. We had recorded and tran-scribed all the discussion at workshop 1, andused cognitive mapping methods to makesense of it. On this basis we had organisedthe prioritised ‘decision areas’ into what wethought might be a decision graph represent-ing their interconnections; this was now con-firmed by the group. However they selected asubset of three of the decision areas that they

1 For an introduction, see Rosenhead and Mingers(2001).

299


wanted to talk about, none of which were dir-ectly linked to any of the others. The result wasthat each of the areas was discussed separ-ately. Just one of these lent itself to the advant-age comparison method of SCA’s comparingmode. This was funding. Three new options(in addition to the status quo) were identified.Pair-wise comparison of two of them gener-ated a useful discussion. Then we repeatedthe exercise, contrasting the status quo withan option of raising money by moving to anon-circular procession routed to pass throughan arena with chargeable seating. This timethe result was electric. Pooling their expert-ise, experience and judgement, they came torealise that across a range of criteria (safety,disruption, cultural celebration, promotion ofthe Carnival arts) the latter option had strik-ing advantages. In a mood of unprecedentedenthusiasm and amity, the various representat-ives took the idea back to their constituencies.Nothing happened.

A LATER INVITATION

Two years passed. The first ever Mayor ofGreater London, Ken Livingstone, was elec-ted. Two people were murdered at the August2000Carnival. TheMayor took on the issue andset up a Carnival Review Group. The Reviewstaff discovered our report, and the move toa ‘linear’ route was promptly adopted as aReview Group recommendation. In the sum-mer of 2001 Tom and I were appointed asAdvisers to the Group. Mike Cushman becameactively involved later.Our principal responsibility now was to pro-

gress agreement on a new route. At our sug-gestion, a Route Working Party representingstakeholders was convened. Nineteen organ-isations took part, some of them with morethan one representative. With these num-bers we had to rethink our intention of usingproblem structuring methods designed for thedynamics of smaller groups. However ourchoice of approach was informed by our exper-ience with these methods, including SCA.

The stakeholders were certainly diverse:three boroughs plus the Greater LondonAssembly, the Association for London Gov-ernment and the Royal Parks; the CarnivalTrust plus two arts funding bodies, three policeagencies and three emergency services; threetransport organisations; and the Tenant Man-agement Organisation. We had redefined thetask of the working party as that of prepar-ing the ground for the selection of a shortlistof routes, to be agreed on by a full confer-ence. This shortlist would then be evaluated byconsultants who specialised in space-relatedtechnical risk analysis. The Greater LondonAssembly (in effect theMayor) would fund thisanalysis for three routes.The aim of our process design was to stop

the parties, who had very different object-ives, from simply disputing the advantagesof one route over another, or the need for achange of route at all. To develop a space fordiscussion and negotiation, we started withabstraction, getting the participants to identifypossible generic route shapes. They ended upwith seven, from circle, line and horseshoethrough to ‘tuning fork’. Criteria were elicitedand used to reduce the range to the four iden-tified above. So progress was made despite aconcerted initial attempt by some members tochallenge the approach and the authority of thefacilitators.Ahead of the secondmeeting of theWorking

Party, the members generated specific routes,11 in all. A set of criteria relevant to choicebetween specific routes was developed,which in turn were condensed into sevenclusters – effectively meta-criteria. In hetero-geneous break-out groups the routes wererated out of 10 on each criterion.Looking at the results produced an enlight-

ening discussion. Routes that scored highly oncriteria most relevant to the service agenciesscored poorly on those of most concern to theCarnival community, and vice versa – so thevarious parties seemed to gain a better under-standing of the logic of each other’s positions.Furthermore some routes seemed to be dom-inated by others on all criteria. However none

300


of the representatives felt they had authorityto withdraw them.The culmination was the tense route eval-

uation sessions of a two-day conference onOctober 29–30 with over 50 present. Carnival-ists for the first time were there in some num-bers. Documentation on the progress madeby the working party had been pre-circulated,along with a statistical analysis of the routes bythe consultants. Discussion was held in threesyndicate groups, this time as homogeneousas possible to promote frank discussion – butwith the ground rules that they should eachtry to produce a shortlist ‘balanced’ across dif-ferent types of route. The facilitators ran theirthree syndicates differently, according to theirjudgements of their internal dynamics. Eachhowever used methods employing flip chartsand ‘post-it’ notes with an agreed list of evalu-ative criteria as background.The Mayor’s senior adviser took over the

chairing of the final session. That morning hehad told me of his strong doubts that agree-ment would be possible; but I persuaded himto wait and see what themethod could deliver.Only one of the groups managed a balanced

portfolio of routes. But in plenary discussionmovement took place and a consensus beganto appear. A larger than intended shortlistwas agreed unanimously, and the Mayor’sadviser said that money would be found forthe additional risk analysis. There was a palp-able break in the tension, and fraternisationbetween members of all the groups.

POSTSCRIPT

Our formal involvement ended there – mis-sion accomplished. We received warm let-ters of thanks from the Mayor’s adviser. Andthe following year the Carnival route changed,rather timidly, from its previous circle to ahorseshoe formation. However that summaryof subsequent events misses out much con-tent. In effect between October 2001 and the2002 Carnival the two principal London bor-oughsmanaged to regain control of the agendafrom the Mayor. Also, in legal action suppor-ted by the main Carnival funders, the Directorand most trustees of the Carnival Trust werereplaced. This turmoil may in part have milit-ated against more radical route change.

CONCLUDING REFLECTIONS

This was a very rich experience. Most of the lessons relate to multi-organisational engagements.Any conclusions are tentative.One key question is that of whether SCA and other problem structuring methods can function

usefully in an antagonistic environment. The answer in the case of the second engagementseemed to be ‘yes – but only just’. It took some facilitators well-toughened in the university oflife, and flying by the seat of their pants, to get through to a result.Politics was of the essence. It was in the room all the time. It has become obvious retrospect-

ively that we only knew a part of what was going on. Better political intelligence would havehelped us greatly. We now know that, despite our academic status and independent funding,we were, by some parties for at least some of the time, seen as the Mayor’s agents and so notimpartial. In a contested environment, the sponsorship of consultants can be crucial.Carnival is a community-based activity. Carnivalists had day jobs to hold down, and so were

under-represented in most of our process, compared with the boroughs and the agencies. Weshould have been more sensitive to who was not in the workshop.When powerful decision-makers are not in the room, good results at a workshopmay not carry

through to implementation. Clearly reserving time at the workshop to devise a plan of action isa good strategy, but may be impracticable when groups are hostile.

301


Lastly, we found ourselves improvising methods in order to work in larger groups. Recentprogress has been made in developing such methods both within the strategic choice tradition,as discussed in Chapter 11 and by later contributors to this chapter (Sections 13.09 and 13.10),and by others (Bunker and Alben, 1997). In the circumstances of the Notting Hill projects, it isvery doubtful whether the participants would have accepted more elaborate rules of the game.

REFERENCES

BUNKER BB and ALBEN BT (1997) Large Group Interventions: Engaging the Whole System for Rapid Change.Jossey-Bass

HAMMERSLEY M and ATKINSON P (1995) Ethnography: Principles in Practice (Second Edition). RoutledgeHORLICK-JONES T, ROSENHEAD J, GEORGIOU I, RAVETZ J and LÖFSTEDT R (2001) Decision Support for Organisa-

tional Risk Management by Problem Structuring. Health Risk and Society, 3 (2), 141–65ROSENHEAD J and MINGERS J (eds) (2001) Rational Analysis for a Problematic World Revisited: Problem Structuring

Methods for Complexity, Uncertainty and Conflict. Wiley

302

13.2 Communicative Learning, Democracyand Effectiveness

Facilitating private – public decision-making in Sweden

By Knut Strömberg and Jaan-Henrik Kain

Three cases from Sweden, in which public and private interests meet and sometimesconflict, demonstrate the value of moving complex issues from formal arenas for decision-making to informal fora for communication and learning in a wider circle of participants.However, it is vital to link the mutually developed learning experiences back to the formalarenas to inform decision-makers. Lessons from these processes are discussed under threeheadings: communicative learning, effectiveness and democracy.

Jaan-Henrik Kain is conducting research on trans-disciplinary management of multi-facetedknowledge related to urban development processes. He is a CEI Post Doctoral Fellow at theDepartment of Built Environment and Sustainable Development at Chalmers University ofTechnology.

Knut Strömberg has been working with the strategic choice approach (SCA) since the late1970s. He is using urban design as a tool for social, cultural and economic developmentand has introduced a profound social science perspective in postgraduate architecturaleducation at Chalmers University of Technology, where he is professor in Urban Design andDevelopment.

BACKGROUND

The public sector’s role in urban design andplanning has diminished in Sweden as in mostEuropean countries during the last 20 years –a transition from formal government to morecollaborative forms of governance. This shift isby no means simple to manage and is oftencriticised for failing in democratic control andfor resulting in unsatisfactory mechanisms forchange management.The traditional approach to urban plan-

ning, with its formal and sectoral organisa-tions, is not adequate when dealing withnon-standard issues going beyond normal

agendas and standard procedures. Instead,other approaches are needed to cope with theintra-sectoral and multi-actor decision develop-ment needed for new forms of governance.

RESTRUCTURING A MUNICIPALHOUSING MARKET

During the 1980s, Köping, a Swedish municip-ality of 24 000 inhabitants, experienced prob-lems with a growing stock of unlet apartmentsresulting in social and economic segregation.The main explanations were a reduction injobs, a decrease in population and an increase

303


of single-family housing. Many efforts to solvethe problems failed.An analysis showed that planning proced-

ures followed a narrow sectoral approachbased on rigid routines. Although most ofthe municipal departments were involved, nosingle department could solve the problemwithin its own competence. As the situationdeteriorated, different administrative agenciesblamed one another for the deadlock.To tackle these problems, SCA was used

in dialogue between politicians, administrat-ors, planners and stakeholders external to thepolitico-administrative organisation. As a res-ult the setting shifted to an informal forumwith lateral communication and interactiveparticipation.This new way of working induced a spirit

of collaboration and working groups wereset up. By establishing a reference groupwith representatives from different stake-holders and a steering group of politiciansand leading officers, the deadlocked processwas reset within new structures (Khatee andStrömberg, 1993).

DEVELOPMENT OF A REGIONALAGENDA 21

During the 1990s, concerns for sustainabledevelopment came to the fore. To estab-lish common ground for future Local Agenda21 processes, a 3-year Regional Agenda21 project was initiated by four major publicauthorities inWest Sweden, an area of 700 000inhabitants.The project organisation followed standard

operating procedures with a political steeringgroup and linear project plan. However, theproject got stuck already in the goal formula-tion phase and the description of the currentenvironmental situation consumedmost of thetime and money available.Some participants were disappointed and

critical but a new approach was taken whenone of the critics was appointed head of thesubsequent strategy phase. The task was to

make best possible use of the remaining yearof the project to regain enthusiasm and trust.Five strategy groups were formed, of which

the interdisciplinary life in the city group wascomprised of planners and officials but alsoof two researchers from Chalmers University,who suggested that an adapted version of SCAshould be applied (strömberg, 2001).

IMPLEMENTING LOCAL WASTEMANAGEMENT

In recent years, infrastructure systems havebecome more and more complex due toa shift from centralised public provision tomore decentralised market-driven services.Moreover, different infrastructure systemsincreasingly tend to overlap and affect eachother.Within such a context, in 2002 the LOCO-

MOTIVE project was designed to explorepossibilities for integrating operation of alocal waste management system with issues,such as social cohesion, management ofgreen space and local economic developmentin a city district of Göteborg with 14 000inhabitants.A cross-sectoral working group of local

actors from city district and housing manage-ment was set up, while Chalmers providedprocess facilitation. During the studymost SCAmodeswereworked through. One observationwas that the number of theoretically possibledecision schemes in a system comprising fourdifferent waste streams became very large(2916 schemes), and the process of delimitingthe scope of the study thus became an essen-tial issue.Nonetheless, the use of advantage com-

parison turned out to provide powerful sup-port for assessment of this complex system.Through a series of consecutive evaluationsunder diminishing uncertainty, a synthesisingassessment across eight comparison areaswas achieved (Söderberg and Kain, 2002; Kain,2003).

304


LESSONS ABOUTCOMMUNICATIVE LEARNING

How did SCA facilitate problem perception andprovision of transparent information? In Köpingthe SCA process generated a shift in problemperception. Moving the problems away fromstandard agendas and established agenciesdramatically revised participant attitudes. Thisresulted in a complex set of interconnectedproblems – questioning established rules fortaxation, subsidies and structure of decision-making –which engaged political leaders,mostdepartments in the municipality and severalstakeholders outside the municipality. Sub-sequently, incremental decisions by the steer-ing group were developed in collaboration andimplemented during the following years. Theafflicted housing area was redeveloped fromhousing only to mixed use.Also in the 1990s case, work in interdis-

ciplinary constellations revealed complexrelationships between different problemareas – such as car emissions, housing marketsegregation and lack of co-ordinated lead-ership – and their related decision domains.As an example, out of the dialogue betweenthe different representatives a dramaticshift in problem perception occurred: shift-ing the focus from over-fertilisation onlyto the interplay of under-nourishment andover-fertilisation. When the project wasformally concluded, it had developed intoseveral informal learning processes compris-ing numerous organisations and individuals.Although the original problems of the publicconsortia were not solved, agreement wasreached over several more fundamental issuesand inter-dependencies.The stakeholders in the last case repeatedly

claimed that the main advantage of SCA wassimply the opportunity to work in such adiverse group. Invisible products (cf. Figure 88)such as knowledge exchange and processing,were seen as very productive, generating pos-sibilities for future networking. Consequently,the outcome of the learning process was a

close-to-consensus progress package. Learn-ing was so effective that some participantseven claimed that this outcome was identicalto the opinions they held before entering theprocess – an argument refuted by our analysis.Another observation was that the argumentsunderpinning the advantage comparison andthe weighting of different evaluation areasremained somewhat obscure – an issue nowbeing studied in ongoing research on comple-mentary multi-criteria approaches at ChalmersArchitecture.

LESSONS ABOUT EFFICIENCY

How did SCA influence the consumption ofmanpower, time and other resources? The1980s’ Köping process went on for 2 yearsbefore the housing market turned. It is difficultto evaluate what really influenced the happyending, but the general feeling in Köping wasthat the application of SCA made it possibleto break a locked situation. One measure ofthe perceived efficiency is that the municipal-ity later on applied the same approach to otherplanning sectors.From the public consortia’s perspective, the

Agenda 21 work did not reach its anticipatedgoal: a Regional Agenda 21 Strategy. However,the working group participants were satisfiedwith the progress and the learning processes.This was also demonstrated by the extensiveefforts put in on voluntary basis by individu-als, businesses and non-governmental organ-isations – engagements that lasted long afterthe formal project was ended.Yet another measure of process efficiency

is whether participants return throughout aseries of meetings. In the LOCOMOTIVE casealmost all participants were present at allworking meetings. Our interpretation is thatthey judged the time they invested was beingmatched by achieved results. Even so, someparticipants argued that SCA ‘made simplethings complicated’ while others saw the pro-cedure as comparatively swift. These differentviews may be related to differences in prob-lemperception, i.e. how complex is localwaste

305


management? It is thus essential that the levelof complexity be matched by choice of facilit-ative approach.

LESSONS ABOUT DEMOCRACY

How did SCA support transparency, manage-ability and involvement of relevant stakehold-ers? In Köping, decision-making was movedfrom established formal arenas, with demo-cratic framing structures for decision-makingand access, to temporal informal fora emphas-ising learning and communication. Since thesteering group took decisions in a grey zonebetween formal and informal structures, ques-tions about democratic control and influencebecame essential.Another aspect of this problem is illustrated

by the LOCOMOTIVE case where participantswere reluctant to acknowledge that others,such as local inhabitants, could not be fully rep-resented by them. In this way, the processgrew somewhat isolated from local everydaylife. Moreover, in Köpingmost processes werenegatively influenced by hidden power playsamongparticipants resulting frompersonal ant-agonism and other forms of complex humanrelations with little relevance for the issues athand. Although harming the process these hid-den agendas were not discussed openly. Sub-sequently, a fourth category of uncertaintywasintroduced, describing in-group power play:intra-organisational uncertainty.Both the Agenda 21 and the LOCOMOTIVE

cases illustrate the essential problem of

communicating findings of more or lessinformal working groups to others, such aspoliticians active in formal arenas (cf. BrendanHickling’s contribution to this chapter, Sec-tion 13.13). The outcome of the processesmay hence be understood at two levels. Atone level, the invisible results of mutual learn-ing and mutual action-oriented commitmentwere strongly present. In the LOCOMOTIVEcase, the participants were even preparedto embark on co-operative local area man-agement based on their mutual learning.However, at a second level, the informalworking forums of both cases had little link-age to formal decision-making arenas. Visibleoutcomes, such as published reports, werenot accommodated at higher municipal levels.As an example, the LOCOMOTIVE report onlocal waste management received a posit-ive response by the central Recycling Officeof Göteborg and was to be brought to theattention of relevant decision-makers for swiftprocessing.Even so, today – 2 years later – nothing has

evolved within the mainstream waste man-agement system in response to the study.The learning effort thus largely remains unre-cognised by Göteborg mainstream activitiesand bodies, and the suggestions of the localstakeholders have been absorbed into thepolitical system without any further reaction.Fortunately, in relation to the Regional Agenda21, the informal networks have later beeninvolved in problem-solving processes imple-menting some of the findings.


In the development of decision-making approaches within the European Union since theMaastricht treaty, with the emphasis on subsidiarity and the current shift from government togovernance, we observe that SCA has been well ahead of its time.The three cases of SCA in Sweden are typical examples of how the planning context has

developed during the last three decades, from emergency intra-organisational problem-solvingthrough collective strategic knowledge building to implementation of sustainable developmentin a complex context of governance with shared power and no one in charge. In all three casesSCA has been robust and adaptable to ever-changing contexts.

306


Work in political contexts requires special attention to intra-organisational power plays andhidden agendas. The explicit management of such has – at times – been successful in ourexperience. Separating the roles of process leader and project leader has been important for themanagement of complex and sometimes conflict-laden problem structuring.

REFERENCES

KAIN, JAAN-HENRIK (2003) Sociotechnical Knowledge: An Operationalised Approach to Localised Infrastructure Plan-ning and Sustainable Urban Development. Department of Built Environment and Sustainable Development Diss.Göteborg: Chalmers University of Technology.

KHAKEE A and STRÖMBERG K (1993) Applying Future Studies and the Strategic Choice Approach in Urban Planning.In: Journal of the Operational Research Society, vol. 44, No. 3

STRÖMBERG K (2001) Facilitating Collaborative Decision Development in Urban Planning. In: Scandinavian Journal ofArchitectural Research, 2001: 4

SÖDERBERG H, KAIN J-H (2002) Integrating Knowledge or Aggregating Data – Multi-Criteria Approaches for Sustain-able Water and Waste Systems. Proceedings, 3rd International Conference on Decision-Making in Urban and CivilEngineering. London.

307

13.3 Delving into the Toolboxes

National environmental policy-making and strategic choice

By Frans Evers

In this contribution the author draws on his experience as Deputy Director-General forthe Environment in the Dutch government in the early 1980s. The focus is on internalpriority setting in the field of environmental policy, which was very poorly developed atthat time. Techniques and methods from a number of sources were used in combinationsthat varied according to the task in hand. They drew selectively from the SCA tool-box,using in particular AIDA and the levels of choice.

Frans Evers is currently Chairman of the Board of the Sustainability Challenge Foundation;Deputy Chairman for the Independent Commission for Environmental Impact Assessment;Adjunct-Professor at Tias Business School; and Senior Fellow at Globus (an inter-facultyinstitute at Tilburg University). He was Chief Executive of ‘Natuurmonumenten’ (the DutchAssociation for Nature Conservation) from 1996 until 2002; Director-General of the DutchPublic Buildings Agency from 1987 until 1996; and, before 1987, Deputy Director-Generalfor the Environment in the Dutch government (Assistant Deputy Minister in other countries).He has been chairman of several international committees, including the OECD Committeeon Environmental Assessment and Development Assistance.

INTRODUCTION

I bumped into Allen Hickling at Schiphol airportsometime in 1980; and I was sceptical whenJan de Koning, the Dutch consultant who intro-duced us, declared that Allen and his methodscould be the answer to many of my problems.It was the tool-box idea that convinced me

though. Being the son of a carpenter, I knewthat there is no single way to solve a problem.There is always the combination of man, skilland tools. The inventive, creative carpenteruses his tools very differently from the onethat is skilled and experienced, but who livesby tradition. My father used to borrow toolsfrom the blacksmith for jobs that seemed to beimpossible. He made his boss rich.

THE SITUATION

As Director for policy and strategy in theEnvironment Ministry, I was confronted withambitions to introduce environmental impactassessment, environmental policy planning,risk management-policies (e.g. for the ship-ping, landing, distribution and storage of LPG)and, last but not least, priority setting withinnational environmental policies.Since the department was basically organ-

ised by sectors (air pollution, water pollution,radiation, waste, noise, etc.), there was inev-itable competition for the scarce resourcesavailable. But the ministers, up until then, hadrefused to embark on the issue of priorities.There was simply no answer to the question

308


whether, for example, the reduction of air pol-lution had more or less priority than the abate-ment of noise, which, in the densely populatedwestern part of the country, had attractedmajor attention from parliament and scientists.

INTRODUCING THE APPROACH

So suddenly there was a workshop to let a fewof us know about SCA. Several of the parti-cipants recognised that therewere possibilitiesfor different areas of concern, and I decided togive the approach a try in the area of internalpriority setting. Consequently a proposal forthis was launched in spring 1981 with the sug-gestion to experiment with the AIDA method.Now here is part of the content of the letter Isent to the Director-General (my boss) and mycolleague-directors.

In the spring we decided that we can nolonger avoid setting priorities for our nationalenvironmental policies. It was acceptedthat, although the final responsibility formaking these choices lies with the minis-ter, we are not relieved of the duty to bringforward the necessary data for him to doso wisely. We decided to use a methodbased on SCA. This means that we donot start from general policy problems andgoals, as a top-down approach, but that wefocus on the concrete decisions that youexpect to be confronted within the comingyear. I suggested that after you made yourinventory of the fields of decision in yourown directorate, wewould together look forthe relations between the different fieldsof decision across the whole organisation.We would then be able to identify areaswhere decisions should be taken together,because there are certain to be choices thatexclude each other.I also suggested looking for possibilit-

ies and reasons to cluster certain areasof decision, because I expected that thefirst inventory of decisions would result

in decisions of different levels. We iden-tified at first three levels, namely stra-tegic, policy and operational decisions.1

Later we decided in the discussions to adda fourth level, called the normative level ofdecisions.2

This highest level seems to be most rele-vant for a government organisation, whichis related to political goals that are intendedto guide decisions outside our organisation.It is clear that it is not easy to do this lev-elling of decisions right at the beginning onyour own. In practice there has to be someoverlap, and some civil servants hardly havea clue about this normative level.Let me be clear: the whole process of pri-

ority setting is meant to focus on decisionsat the strategic level, but I am quite awareof the fact that limitations at the policy andoperational levels can restrict the room forstrategic choices.

So far the young director in 1981.

WORKING WITH LEVELS OFCHOICE

The letter not only caused a stir in many civilservants’ lives, but it also set us on a difficultand tiresome road of trying to use AIDA to helpus find and keep track of the interconnecteddecision areas. Coincidentally it also forced usto get to grips with the whole idea of strategicplanning. It became clear, through use of thevarious tools from the tool-box, that we hadto embark on a discussion with the ministerand, indeed, the whole organisation, about theimportance of having a strategy for the environ-ment. Otherwise people would continue beingtoo late in dealingwith environmental risks.Wewould continue to be a clean-up organisation,

1 See ‘AIDA and the Levels of Choice in Structure Plans’by Hickling (1978); and the contribution to this chapterby Angell (Section 13.10).

2 This is the same as strategic orientation in the vocabu-lary of Hickling (1978) and Chapter 12.

309


while allowing other ministries and the busi-ness community to continue messing aroundwith the environment.

A CHANGE OF DIRECTION

So we embarked on a fascinating trip on whichit was not clear what the destination wouldbe. All directors produced their ideas about theconnected decision areas within their field ofresponsibility. As an example, my own in itsfirst draft is presented here (Figure 13.1).

Figure 13.1 Major Decision Areas for Policy Development Directorate in 1981.

In a huge effort we connected all of them.The result covered awholewall ofmy room.Ofcourse we got stuck in such an elaborate effortand had to simplify a lot of things. The mostimportant result was the insight that we couldnot continue creating solutions in one areawhile creating new problems in another. Atthat time we were like many of our colleaguesin other countries, shifting waste from the airinto the water and from the water into solidwaste, and we disposed of the solid waste bycreating a soil pollution problem.

310


THE RELEVANT QUESTIONS

We decided to stop this muddling throughand go for a strategic approach. The newminister in 1982, Pieter Winsemius who hada McKinsey background, asked the relevantquestions about strategy and priorities after afew months in office. I doubt we would have

had any idea how to answer them if we hadnot done our prior exercises. In finding therelations across the sectors, in creating anatmosphere where levelling of decisions wasno longer a threat, we laid the foundations forthe guidance position of Dutch environmentalpolicies that lasted until the middle of thenineties.


We found tools in the toolbox of SCA that we liked and were able to use. But we also used othertoolboxes thatwe borrowed or bought,whenwe felt that theywere better for the job. Consensusbuilding using the mutual gains approach was probably the most important. Professor LawrenceSusskind, the relentless promoter of that approach, became another important advisor. Othertools came from project management as taught by Gerd Wijnen.

311

13.4 Less is More

Controlling decision complexity when using strategic choice

By Richard Ormerod

In this contribution the author draws on his experience as a consultant helping SevernTrent, a regional water utility in the UK, to develop a customer service strategy. He drawsattention to three aspects of the application of the SCA:

1. the need to manage down the complexity of the decisions to be taken;2. the option of introducing quantitative scoring into the evaluation; and3. the extension of the progress package to plan a timetable of future decisions.

Richard Ormerod originally became aware of SCA when he read a paper about it in theJournal of the Operational Research Society in the early eighties. He found the categorisationof uncertainty insightful in his then role as a corporate planner in the National Coal Board.However, it was not until the late eighties that he first used SCA as a decision-making aid whenworking as a management consultant in the PA Consulting Group. On entering academia inthe early nineties, as well as teaching SCA and other approaches to postgraduate students,he continued to use SCA in practice as a consultant. He has used SCA on a number ofassignments in the retail, mining, water and electricity industries – usually helping clients todevelop their information systems strategy, but the projects also involved process, technicaland organisational change.

THE SITUATION

Some time before I came on the scene SevernTrent had decided to develop a customer ser-vice strategy to address the one area wherethey performed less well (as measured by theUK water regulator) than other regional watercompanies.A core team had drawn up terms of refer-

ence for half a dozen or so task forces of seniormanagers. Each task force was asked to ana-lyse a particular aspect of customer serviceto develop options and recommendations, andto report back. For example, one task force

had examined the nature and organisation ofthe physical interface with customers and thegeneral public (by letter, telephone and face-to-face). They had addressed issues such aswhether customer services should be organ-ised centrally or locally, and whether call hand-lers should be specialised (billing inquiries only)or deal with several types of query (operationaland billing inquiries). The task forces werereporting back their findings. The core teamwas facedwith awide range of overlapping andconflicting recommendations coming from thedifferent task forces.

312


THE ASSIGNMENT

It was at this point that they asked me to helpthem develop a combined coherent strategyfor the Board to agree. Time was an issue: anexpensive venue had already been booked fora high-profile public launch of the new strategy.One of the points made in Chapter 5 is that it

is advisable to keep the problem focus narrow.I found that this point is key to the success ofan SCA intervention. It is my experience thatthe facilitator must actively manage the com-plexity of the decisions under consideration ifthe process is to avoid getting bogged down.The Severn Trent case illustrates the point.

MANAGING COMPLEXITY

In this case it was not difficult to identify lots ofdecision areas and plenty of options of differingdegrees of importance and urgency. Thereinlay the problem. To be inclusive everythingneeded to be acknowledged and made part ofthe analysis from the start. However, when itcame to shaping the decision space the ques-tion of complexity had to be addressed or theproject would be overwhelmed. For instance,if it is decided to include six decision areas inthe problem focus, each with three options,there are potentially 729 decision schemes tobe evaluated.It would take an enormous number of option

bars to get the number of schemes down to amanageable level. It is far better to restrict thenumber of decision areas in the first instance.I persuaded the team to limit themselvesto four areas for the first pass while theywere getting to understand the approach.They needed the reassurance that the issuesparked for the time being would be broughtback in a later cycle of the process. If eacharea has three options this limitation wouldreduce the maximum number of schemes to81. In the event they chose four options in onearea, two in another and three in the other two(reducing the total number of schemes to 72).Again it pays to struggle to find as many optionbars as possible. In this case after identifying

some option bars we were left with about 20schemes. This was still a lot to understandand evaluate, but it was then manageable.Reducing the number of schemes by limitingthe number of areas included in the problemfocus, constraining the number of options,and searching out all the option bars is crucialto the success of the evaluation phase.The decision to restrict the number of

decision areas to four in the first instancehad been a difficult one. Nevertheless, onreturning to the original map of the decisionareas it proved fairly simple to consider thosedecisions that had been left out. Each decisioncould be considered in the light of the decisionschemes favoured by the evaluation. Someoptions could be barred while others had aparticular affinity with one scheme or another.Others still resulted in variants of particularschemes. It seemed easier for participants tohandle the complexity this way than to gothrough the SCA shaping and evaluation again.The three or four remaining schemes werethen fully evaluated including a comparisonwith the way that other regional water com-panies organised their customer service oper-ations.

QUANTITATIVE SCORING INEVALUATION

The initial evaluation using a few criteria isanother opportunity to manage down the num-ber of schemes, as discussed in Chapter 3.We set the hurdles high so that only a fewschemes remained. When it came to the moredetailed evaluation, the teamwanted to be ableto give different weights to the various evalu-ation criteria. The approach taken was to scoreeach scheme on a scale of 1–5 for each eval-uation criterion. Each criterion was then givena weight. An overall score for each schemewas derived by multiplying scores by weightsand adding up across the criteria. In doing this,something is lost by turning the evaluation intonumbers, but it gave the option of changing theweights to explore the sensitivity to differentassumptions.

313


On another assignment, Sainsbury’s, a UKsupermarket chain, went a stage further andinsisted that all the evaluation scores weretranslated into monetary terms even thoughsome of the criteria were soft in nature. Formany participants this would be a step too far.

EXTENSION OF THE PROGRESSPACKAGE

Finally, when drawing up the progress pack-age the participants decided to go furtherthan distinguishing between decisions nowand decisions later. They divided the decisionslater into time periods depending on when the

results of further explorations would becomeavailable. The result was a critical path networkof explorations. This provided a basis for set-ting time limits on explorations and developinga schedule of decision meetings.Crucially they were able to plan which issues

they could reasonably hope to resolve by thetime of the launch and those which would beincluded in the launch as issues that wouldbe explored and resolved in the future. Withthe launch date rapidly approaching, the real-isation that they could present both decisionstaken and decisions yet to be taken (dependingon further exploration) side by side in the newstrategy was a great relief to all concerned.


In the Severn TrentWater case described above both the introduction of a quantitativemethod ofevaluation and the extension of the progress package using critical path concepts are illustrationsof the more general point that it is often desirable to enhance the use of one approach withothers – as argued in the previous contribution of Frans Evers. At RTZ (an international miningcompany) and PowerGen (a UK electricity utility), soft systems methodology (SSM) (Checkland,1983) was used to generate strategic options while SCA was used to shape and evaluate them.On its own SCA provides little support for creative idea generation, while SSM provides littlesupport for converging on a decision. Combined SCA and SSM can be used to support a strategydevelopment process from idea generation to agreed decisions. It is experiences such as thosedescribed above that have led me to appreciate the strength of the SCA framework. All thesame I do not to follow it too rigidly. Groups of decision-makers have particular requirementsand they appreciate an approach that can be adapted to their circumstances. They like to be incontrol of the process rather than controlled by it. The consultant/facilitator should have someof the possible variants in mind so as to respond appropriately and confidently as the decisiondynamic unfolds and the decision-making requirements become apparent.

REFERENCES

ORMEROD R (1998) Putting Soft OR Methods to Work: Information Systems Strategy Development at Palabora.In: Omega International Journal of Management Sciences 26, 75–98

ORMEROD, R (1999) Putting Soft OR Methods to Work: The Case of the Business Improvement Project at PowerGen.In: European Journal of Operational Research 118, 1–29

PAULEY GS and ORMEROD R (1998) The Evolution of a Performance Measurement Project at RTZ. In: Interfaces28, 94–118

314

13.5 Feet on the Ground

Engaging planning students with political realities

By Arnold van der Valk and Gerrit J. Carsjens

The strategic choice approach stands students in planning and management with bothfeet firmly on the solid ground of day-to-day decision-making. Today planners cannotafford to be called dreamers or ivory tower types. They are supposed to come up withproposals that have a chance of acceptance and of implementation. If not, it reduces thelimited supply of public confidence. That is why teaching staff of Wageningen Universityin the Netherlands includes the philosophy and toolbox of strategic choice in the corecurriculum for landscape architecture and spatial planning. Since undergraduate coursesin planning methodology are open for students in environmental and social sciences, some25 engineers graduating at Wageningen University yearly have been exposed to the basicsof strategic choice since 1999. In the masters curriculum, ‘Planning under Pressure’ is usedas a textbook by planning students.

Arnold van der Valk has been a full professor in land-use planning at Wageningen Universityin the Netherlands since 2002. He took a PhD in planning from Amsterdam Universitywith distinction in 1989. He holds an MSc in human geography and an MSc in spatialplanning. The emphasis in his scientific work is on planning theory and planning history. Hisseminal work so far is Rule and Order; Dutch Planning Doctrine in the Twentieth Century,co-authored by Faludi (1994).

Gerrit Jan Carsjens is a senior staff member of the Land Use Planning Group at WageningenUniversity. His educational work focuses on technical and methodological aspects of land-use planning. This includes working with strategic choice in two strategic planning courses, asdescribed in this contribution. His research focuses at the development of decision-supportmethods using geographical information systems (GIS). Two main projects are: (1) the FAOproject on area-wide integration of crop and livestock activities in Southeast Asia and (2) theintegration of environmental aspects into local planning procedures in The Netherlands.

STRATEGIC CHOICE IN APLANNING CURRICULUM

Strategic choice was introduced by one of theauthors who was appointed professor in land-use planning inWageningen in 1999. He wrotetwo introductory texts to strategic choice in

the Dutch language for undergraduates. TheDutch texts present elementary informationabout a decision-oriented view of planning,an overview of strategic choice technologyand a glossary both in Dutch and in English.The decision-oriented view of planning wasdeveloped by Andreas Faludi and Arnold van

315


der Valk in the 1980s, building on the found-ations laid by the pioneers of strategic choicein the preceding decade (Faludi and van derValk, 1994; Faludi, 2000). This approach setsout to develop a sound scientific foundationfor planning rooted into Popperian social philo-sophy and epistemology. Faludi acknowledgesstrategic choice as the best application of hisinterpretation of rational decision-making inpractice.In the Wageningen planning course, the

foundations of strategic choice are explainedusing contemporary examples taken fromplan-ning practice. Since jargon has proved to bea major obstacle in the process of teachingstrategic choice, staff members have designedrole plays, studios and applications based onreal-time planning experiences. These situ-ations have been carefully selected and simpli-fied for educational purposes.Strategic choice basics are taught in an

introductory planning course in the secondyear of the Bachelor’s Landscape Architec-ture and Spatial Planning (in Dutch, BSc land-schap, planning en ontwerp). This basic coursetakes two weeks of part-time classes andexercises. Students participate in a simu-lated decision-making process about the useof land in a rapidly urbanising region nearthe city of Rotterdam. The case comes witha lot of substantive information about ananticipated harbour, about the claims of hor-ticulturalists, administrative procedures, theinstitutional context, and social and economicstatistics. Above all the students have to learnabout empathy by taking up the role of a stake-holder group or playing the role of a profes-sional. In the role play they are facing the ten-sions inherent in political power play. They geta first impression of facilitating, mediating andorganising participative planning processes.In the third year, SCA is at the heart of two full

grown planning courses that include severalweeks of full time exercises. The courses arebuilt upon the foundations of strategic choiceas explained in Chapter 1 of this book. Thefirst course is an in-depth planning methodo-logy course, as described below. In the second

course the students are asked to produce astrategic spatial plan for the long termof a com-plex region in the Netherlands. This secondcourse will be described in more detail in thenext section.The manual for the planning methodology

course familiarises the students with thefoundations of strategic choice as explainedin Planning under Pressure. The manualexplains the intricacies of choosing strategic-ally, the organisational context, the dilem-mas of decision-making and the responses touncertainty. The emphasis is on the philosophyof strategic choice, perceived as a joint learningprocess. In this phase of the bachelor, studentsin planning and architecture are supposed tobe familiar with a lot of the conventional experttechnology characteristic of routine decision-making.

RESPONSES FROM STUDENTS

In some instances this proved to be anobstacle to the reception of open technology.The good news is that the majority has anopen mind for the social conditions that pressdecision-makers to opt for open technology(see Chapter 4). Group decision techniques areused to sensitise students to an orientationtowards participative interaction. The ultimatechallenge for most students is the empathypart in the simulation of stakeholder particip-ation. The expert roles, included facilitation,group management, quick and dirty researchand chairing negotiations seem to pose muchless of a challenge. An explanation for this kindof conventional expert behaviourmay be foundin the character of the traditional technical train-ing in the first and second year. The emphasisin these 2 years is still very much on tradi-tional learning, i.e. technology for quantitat-ive analysis and cognitive skills for engineersand architects. Traditional engineering com-petences are still kept in high esteem withinWageningen University, formerly known asWageningen Agricultural University.A majority of the students struggle with the

strategic choice philosophy as it conflicts with

316


their perception of planning. This holds forsocial science-oriented students in planningbut has even greater relevance for students inengineering disciplines and architecture. Stu-dents in their primes tend to equate plan-ning with plan making and design of physicaland social constructions. Carefully constructedtraining exercises contribute to a change inperspective from the dominant blueprint plan-ning philosophy to a decision-oriented processapproach.Our experiences with the application of SCA

in academic planning courses show that theproof of the pudding is in the eating. Stu-dents cannot grasp the idea of strategic choiceby reading a textbook alone. It takes practicaltraining exercises to get students acquaintedwith these concepts.The use of the STRAD software is optional

in the curriculum. Since planning methodologyclasses suffer from severe time restrictions,students are warned not to spend too muchtime in playing around with software. Froma software engineering perspective STRADis obviously still in an experimental phase.Staff and students though praise the STRADmanual as a valuable tool in the course. Thesoftware is appreciated by the students forits clarity but too limited in the scope of itsapplications.STRAD classes have been replaced recently

by group decision sessions. Group decisionroom (GDR) sessions are limited in the face ofhigh cost. GDR software has proved extremelyuseful for our students in the shaping mode.Well-prepared group decision sessions save alot of time and produce more high quality con-tent during brainstorming sessions. Integrationof group decision technology with the STRADsoftware is a must from our perspective.

COMBINING COMPLEMENTARYAPPROACHES

The landscape, planning and design curriculumcombines a decision-oriented approach anda design-oriented approach to planning. The

decision-oriented approach focuses on pro-cess management and draws its main sourcesof inspiration from policy sciences. Thisapproach takes day-to-day decisions and theanticipated consequences as a starting point.It focuses on systematically working towardspolicy frames and a commitment package.The emphasis is on the ‘here and now’. Thedesign-oriented approach emphasises the con-struction of spatial plans and is built upontraditional architectural design theory. Thisapproach focuses at the construction of altern-ative futures with the help of creative leapsthen taking consecutive steps backward intime in order to arrive at practicable policyguidelines. It encompasses elements of a tra-ditional survey-analysis-plan approach to plan-ning and makes use of prospective scenariotechniques. These techniques are indispens-able for initiating debate about desirable andprobable futures.The combination of strategic choice tech-

nology and prospective scenario techniqueshas silenced complaints about the neglectof creativity and substance in a decision-oriented approach of planning. For educa-tional purposes it seems highly attractive todevelop sophisticated methods for learning,combining the fruits of both perspectives forplanners, architects and other professionals-in-training. Friend and Hickling have taken upthis challenge under the header ‘practicalities’in the technical domain (cf. Figure 89 – stra-tegic choice as a continuous process).A preference for problem-oriented courses

and design studios is an integral part of theWageningen educational tradition and philo-sophy. Students are challenged to analyseand solve complex problems taken from prac-tice. Interdisciplinary research and stakeholderparticipation are necessary ingredients of thisapproach. In the third year of the bachelorcurriculum, students embark upon a realisticregional planning studio. Learning by doingis combined with systematic application ofprescriptions taken from the planning meth-odology classes. Text books provide themwith additional information about technology

317


for decision-making. Staff members performthe role of advisers on the job. Core dataare taken from a real-time planning processand presented by practicing planners. In thefinal studio, taking several weeks of full timework, students practise a combination of thedecision-oriented approach and the design-oriented approach.The studio reflects an incremental and cyc-

lical planning process. Students are temptedto produce a proposal for a regional spatialplan. At the heart of this plan is a realisticand reasoned commitment package. In theannexes, students produce a detailed reportof the process and the invisible products.Emphasis is laid upon the argumentative partof the work and the role of professionals inplanning. Stakeholder participation is simulatedin non-regular plenary sessions in the begin-ning and at the end of the course. Studentswork under very strict time constraints and arethus confronted with a range of uncertainties.

The design-oriented approach emphasisesthe use of quantitative analyses i.e. applied soilscience, hydrology, landscape ecology, trafficmanagement, land consolidation technology,historical geography, economy, demography,institutional analysis, stakeholder analysis andanalysis of cultural values. Future trends areestablished. Imagery, statistics and maps playa role in visualising alternative futures. Stake-holders and professionals are invited to parti-cipate in a process of mapping consequences.This approach diverts attention from seeminglyunsolvable conflicts and helps in reframingcontested decisions. Playing around with sub-stantive topics takes away some of the stressinvolved in group processes. The design-oriented element is particularly helpful inscanning distant futures, thus providing eye-openers for politicians and one-issue parti-cipants. These results enable students toproduce reasoned lists of decisions and uncer-tainty areas, and discuss priorities.

CONCLUSION

Strategic choice is an indispensable tool for planning education and related fields of professionaltraining. It encompasses a powerful set of ideas and concepts offering help in complex situationsof choice under uncertainty. It helps students to develop the facilitation and mediation skillsso much needed in modern collaborative planning practice. The open technology leaves scopefor add-ons taken from the domain of physical design and planning. These add-ons are usedto help students to produce richer arguments by way of combining substantive and proceduralknowledge. Combining strategic choice with design-oriented approaches helps in bridging thegap between science and political action.

REFERENCES

FALUDI A (2000) The Performance of Spatial Planning. In: Planning Practice and Research, Vol. 15, No. 4, pp. 299–318FALUDI A and VALK AJ van der (1994) Rule and Order; Dutch Planning Doctrine in the Twentieth Century. Dordrecht:

Kluwer Academic Publishers

ADDITIONAL INFORMATION

Details on theWageningen Landscape Architecture and Spatial Planning curriculum can be foundon: http://www.wau.nl/rpv

318

13.6 Designing Electronic Support

Empowering decision-makers through software for strategic choice

By Dave Friend

In this contribution, the architect of the STRAD software for strategic choice draws on hisexperience in designing and testing marketable prototypes of this package over a 9-yearperiod – starting in the late eighties when the diffusion of access to the fast-developingworld of information technology was still in the early stages. He reviews the choices ofdirection he himself has had to make in the development of this software in the face ofchanging circumstances and opportunities, and draws some lessons for the design of futureelectronic support for decision-makers.

Dave Friend is the second son of Mari and John Friend. He was aged 18 when, working withhis father, he started developing computer software for strategic choice. At that time, he hadbecome angry about many things he saw in the world around him, and disenchanted withhis experiences of the educational system; yet, along with many others of his generation, hewas becoming stimulated by the creative potential of the new world of home computing.As a self-taught programmer, he spent 9 years developing successive versions of the STRAD(Strategic Adviser) software, and has since continued to help John Friend in maintainingtechnical support and upgrade services to users. He lives near Bradford in West Yorkshireand now works as spares supervisor for an international microelectronics corporation.

EARLY WORK ON STRAD

It was in 1987 that I first started work onthe development of the STRAD softwaredescribed in Chapter 10, initially on a four-month trial contract. The personal computeron which I started working was then new onthe market but had capacities that would nowseem ridiculously limited; it was an Amstrad1512, with 512 kb of RAM, no hard drive andtwo 360 kb floppy drives. With some guid-ance from my father’s former colleague JohnStringer, I started work using Microsoft Quick-BASIC 3 as a readily available programminglanguage in which a first prototype could bedesigned.

Then, a few years later, the arrival ofmore advanced operating systems for IBM-compatible personal computers, with asmarter graphical user interface, allowed fur-ther aspects of the Strategic Choice Approachto become more easily incorporated into thedesign of STRAD. Users now became ableto draw decision graphs much as they hadpreviously appeared on flipcharts. They couldalso interact with software in a more intuitiveway, negating the need to learn sequencesof keystrokes to navigate around a series ofscreens.It was particularly valuable that the Win-

dows interface allowed the results of workfrom the shaping, designing, comparing and

319


shaping modes to be displayed on screen atthe same time. This enabled users to referback to their earlier work, and to switch frommode to mode, as if they were in a strategicchoice workshop with a developing gallery offlipcharts around the walls.The storage limitations of the desktop com-

puters of that time meant that care always hadto be taken not to overload the screen with toomany different windows displaying the resultsof work so far. Yet, with the growing capacit-ies of personal computers, it was to be only amatter of a few more years before these con-straints were overcome.

CHALLENGES IN SOFTWAREDESIGN

Throughout the early nineties, the develop-ment of STRAD was held back at times by theneed to keep up with successive advances inoperating systems, and the continuing needto ensure bug-free operation on a range ofplatforms.It is easy to concentrate on developing new

features, but in doing so one risks the usabil-ity of the application. There are many ‘what-if’pitfalls to beware of; and I suspect that soft-ware that is designed to give flexible supportto decision-makers in responding to changingsituations will always pose more challenges inthis area than most other types of software.Take this example: A focus of four decision

areas may produce a list of 20 possibleschemes. The user proceeds to compareoptions within each decision area and makesan assessment on the combined data thatappears. Comparing the relative advantages ofthese schemes leads the user to exclude someless advantageous schemes and at somestage to adopt a progress package based onthese assessments.After some discussions with colleagues, the

user realises that another option is availablein one of these decision areas. How is thecomputer to deal with this kind of scenario?Surely the data requires some review? Should

the computer be able to recognise the con-sequences of this otherwise basic action?Does it simply warn the user, or should itattempt to interlace the new option into theexisting choices already made by the user?

SOFTWARE DEVELOPMENT ASA PROCESS OF STRATEGIC CHOICE

I have found that programming the softwarecan itself pose some tricky policy choices andregularly produces areas of uncertainty. Forexample uncertainty is raised by:

– the range of possible scenarios that userswill be able to create under the range ofcomputer platforms available (this could beexpressed in SCA terms as UE);– the effect of a course taken in one area of

development on another area of the program(UR);– policy decisions where it is a matter of

debate whether control should lie with thecomputer or with the user, or indeed whetherthe user should be given an option either todecide or to trust the computer (UV).

KEEPING PACE WITH FURTHERTECHNOLOGICAL CHANGE

In our small-scale programme to develop soft-ware to reflect both the spirit and the prac-tical toolbox of the strategic choice approach,we have had to focus our limited resourceson developing a versatile package to meet theneeds of individual decision-makers and smallinformal groups who can meet around a singleworkstation – a package that could quickly betested in a range of potential markets. Thechoices and uncertainties that I have faced indesigning the first production versions of theSTRAD software can be viewed as a devel-oping story of planning under pressure, in asetting where the technology has been by nomeans standing still.Since I started developing STRAD in the

late eighties, there have been impressive

320


advances in information and communicationtechnologies on a wide front, as discussedmore fully in the later sections of Chapter 10.Among the most important advances in termsof support for decision-making have been notonly the technological changes themselves butalso the rapid growth in access to new chan-nels of electronic communication. The spread

of access to the Internet, e-mail and mobilephone networks has been dramatic in recentyears, and the further advances now on thehorizon will present not only new opportunit-ies, but tough new choices and new uncer-tainties for the designers – and the promoters– of future electronic technology for strategicchoice.

REFLECTIONS

Many lessons have been learnt from this initial small-scale programme of software developmentfor strategic choice, and some of the wider lessons have already been reported in Chapter 10.I have been able to build some, but as yet by no means all, of these lessons into the design oflater editions of the software and its user interface. Meanwhile, some of the lessons have alsobeen absorbed into the wider field of practice on which this third edition builds.Further development of software support for strategic choice need not result in ever more

complicated systems, replacing human interaction, discussion and opinion with rigid and pre-defined computer gimmickry. A more useful way forward is to further develop resources forflexible working around an intuitive interface, which will take the number crunching but not theunderstanding and control away from the decision-makers.Several of the other contributors to this chapter report on ways in which they have so far

been able to make use of the STRAD software, not only in working on decisions but also incapturing lessons from their experiences. Most of the experiences of software support so farhave been at a small and relatively intimate scale of decision-making. So they raise questionsof how to reach out towards influencing the choices of other decision-makers, as shown inFigure 31, rather than of co-ordination from some central policy position. In future, continuedextensions of access to information and communication technologies, as well as advances in thetechnologies themselves, will create many opportunities for empowering the decision-makersof the future. As indicated in the previous contribution to this chapter (Section 13.5), they willhave an expanding role to play in introducing younger people to all the complexities of real-lifeplanning that they must prepare to meet in their future careers.

321

13.7 Neighbourhood Renewal in Rome

Combining strategic choice with other design methods

By Alessandro Giangrande and Elena Mortola

The subject of this contribution is a continuous, participatory and incremental procedurewhich combines three approaches cyclically connected: Visioning, the Strategic ChoiceApproach and A Pattern Language. In the view of the two authors it helps to strengthen SCAas a planning and design procedure. As members of the team that won a national designcompetition, the authors utilised it for developing some planning and design proposalsfor the renewal of Centocelle Vecchia, a sub-urban neighbourhood of Rome that sprangup during the twenties. As a consequence of this success, the municipality of Rome hasgiven the team the task of forming the preliminary plan for the rehabilitation of theneighbourhood. This work is still under way.

Alessandro Giangrande taught Design Methods at the Faculty of Architecture of the Uni-versity of Rome La Sapienza from 1973 to 1992. Since 1993 he has taught EnvironmentalAnalysis and Evaluation at the Faculty of Architecture of University Roma Tre, where he isalso the manager of the Information Technology Laboratory for Sustainable Urban Design.During the last 8 years he has been involved in various programmes of community planningand design: in particular he has collaborated with the Municipality of Rome in the estab-lishment of the Neighbourhood Municipal Laboratories, and with the municipality of Quitoand the San Francisco University in Ecuador in starting up laboratories for sustainable urbandevelopment.

Elena Mortola was professor of Design Methods in the University of Rome La Sapienzabetween 1983 and 1992. Since 1993 she has taught Computer-Aided Architectural Designat the Faculty of Architecture of University Roma Tre. In association with the EuropeanMETA University network, she directs a master’s course on Interactive Sustainable Designand Multimedia. She has been a visiting professor at universities in Scotland, the Netherlands,Ecuador and Mozambique, and has developed methods in the fields of Environmental ImpactAssessment, urban design and CAAD.

OUR ‘IDEAL’ PROCEDURE

In our capacity as teachers in the field ofurban design and regeneration, both of ushave had opportunities to test over a longperiod, with our collaborators and students,the effectiveness of the Strategic Choice

Approach, especially when linked to com-plementary design methods of Visioning andAPattern Language.Wehave applied this com-bination to several consulting and researchprojects, as well as using it extensivelyin our teaching for first degree courses,master’s courses and doctoral theses. In a

322


Visioning1 exercise, all the interested parties –such as local authorities and other public bod-ies, entrepreneurs, professional, cultural andenvironmentalassociationsandneighbourhoodcommittees – are invited to attend a workshopaimed at developing a participatory scenario.A participatory scenario is not a picture of the

future state of a place, but a tool for identify-ing problems and exploring possible actions toenable a community to move towards a prefer-able future. It differs from an expert scenario,where the major focus is on technological andeconomic forecasting.In our version of Visioning, the scenario

takes the appearance of a ‘chronicle from thefuture’. The participants to the workshop inter-act under the guidance of facilitators to writea short story set in the distant future, 15 or 20years later. The participants identify with theprotagonists of the story and describe whatthey see, say or do in order to built up a ‘vis-ion’, that is a representation of a desired future.Scenario making is the first step of a learningprocess that continues through the use of SCA.In SCA the interaction takes place in a work-shop involving both some regular participantsand some ad hoc participants, responsible forspecific areas of specialised substantive input.The learning process develops in an adapt-ive and exploratory fashion, guided by thefour working modes of SCA: shaping, design-ing, comparing and choosing – as shown inFigure 8.The scenario provides the main input to

the workshop. We identify issues as decisionareas, options, uncertainty areas or com-parison areas, updating this picture – forinstance, by opening a new Visioning session –whenever the decision situation changes.In the designing mode the working group

identifies some alternative proposals (options2)

1 See,for instance:AmesSC(1993)AGuidetoCommunityVisioning:Hands-on Information for LocalCommunities,Oregon Vision Project, American Planning Association(OregonChapter),Oregon,Portland.

2 The options are the natural consequence of the factthat the Visioning procedure usually generates different‘visions’ i.e., different solutions to the same problem.

for each decision area. The options that entailtransformations of the physical features of theterritory can be more closely defined with theaid of A Pattern Language 3. This is a languagedeveloped by Christopher Alexander and hisassociates that helps a community to designa set of consistent and effective transform-ations. Each pattern is an ‘archetype’ thatdescribes a problem that occurs over and overagain in different spatial settings, pointing topossible lines of solution that can be adop-ted countless times without ever doing it thesameway twice. Alexander described 253 pat-terns that he was able to arrange, starting withthe largest, for regions and towns, then work-ing down through neighbourhoods, clusters ofbuilding, buildings and rooms, ending finallywith details of constructions. Newpatterns canbe created, if necessary, to supply solutions toany problems not considered in this language.Like SCA, A Pattern Language is an inter-

active and incremental approach. It is possibleto start a process of territorial developmentor change incrementally through selection andaggregation: each pattern interprets the pat-terns at the higher scale and is in turn inter-preted by the patterns at the lower scale. Therules of the language can help a group to defineoptions within decision areas as patterns, orclusters of patterns, that are compatible andsynergetic.This planning/design process is cyclic. It is

not important always to start with Visioning: forsome decision areas, options and sources ofuncertainty can be identified from the existingdecision situation, before developing any scen-ario. The scenario making can also be interrup-ted to allow the participants to use A PatternLanguage as a diagnostic tool to understandthe problems of the area and access some sug-gestions (patterns) that can help to solve themand so on.

3 Alexander C, Ishikawa S and Silverstein M (1977)Oxford University Press, New York: A Pattern Lan-guage. Towns – Buildings – Construction.

323


THE PROCEDURE APPLIED WITHINROME

As members of a team4 that won the secondnational competition in Participatory and Com-municative Planning5, the authors applied thisprocedure to develop some planning anddesign proposals for the renewal of the neigh-bourhood of Centocelle Vecchia in Rome. Theparticipation of the inhabitants was managedby way of a laboratorio (workshop) expresslyset up by the municipality of Rome for thecompetition. Through this laboratorio, publicmeetings were arranged with different groupsincluding local authorities, associations, neigh-bourhood committees and schoolchildren. Asa result, extensive documentation was pro-duced on the problems of the neighbourhoodand the wants of the residents. The membersof the teammade wide use of this documenta-tion, togetherwith other data gathered throughspecial field research, to produce a (simulated)scenario6.From this scenario the team developed 21

decision areas, with levels of importance andurgency and a provisional list of options (2–5 foreach area). Then they built up a decision graphas an overview of the structure of the problem,and identified several foci with the help of theSTRAD software described in Chapter 10.Within each focus, the choice of a preferred

decision scheme was explored using the fol-lowing six comparison areas:

1. Environmental sustainability2. Urban sustainability3. Social and cultural sustainability4. Economic and financial sustainability

4 E. Mortola (team leader), M. Bastiani, G. Cafiero,B. Del Brocco, M. Felici, A. Fortuzzi, A. Giangrande,F. Mecarelli, F. Sartogo, A. Simone and A. Zarfati.

5 The competition had been advertised in 2000 by INU(National Institute of Urbanism of Italy) and WWF(World Wildlife Fund).

6 The rules of the competition prevented the team fromkeeping continuing contact with the inhabitants ofthe neighbourhood, so the team was unable to applythe Visioning procedure to make a ‘true’ participatoryscenario.

5. Expressed desires of the inhabitants6. Degree of innovation in solutions.

The first focus includes four decision areas:

1. What interventions for via Tor de’ Schiavi?2. Where to design new pedestrian routes?3. What traffic improvements in neighbour-

hood?4. Where and how to plant new vegetation?

Starting with the preferred scheme for thisfocus, all options were found impracticable inthe immediate future: so they were put in thelist of deferred choices of the progress pack-age. In the explorations now section, the teamput only the exploratory options that they con-sidered appropriate for reducing critical uncer-tainties. For instance, to realise the preferredoption for pedestrian routes, it was importantto test whether the owners of the courtyardswere willing to transfer to the municipalitythe ownership of those spaces necessary tocomplete the network – perhaps in exchangefor certain benefits. Similarly, the exploratoryoptions for Tor de’ Schiavi were needed to testopposition to the project from residents andtradesmen, including repair garages; and alsothe possibility of diverting the buses presentlyusing that route.Meanwhile, the team proceeded to develop

more detailed designs for the preferred optionsin these two decision areas with the aid ofA Pattern Language. They wished to testwhether the designs, if of sufficient quality,might reduce opposition to the conversion ofvia Tor de’ Schiavi into a pedestrian street; ormight induce the owners of the courtyards togrant the spaces needed to build the pedes-trian routes. These options are illustrated inFigure 13.2. At first the designers chose theprincipal patterns, i.e. the patterns that in theiropinion should play a major role in implement-ing the projects. These patterns were Promen-

ade (for pedonalizz) and Children in the city (forretescuole). Then they used the rules of the

324


Figure 13.2 Options Retescuole and Pedonalizz Designed with the Aid of A Pattern Language (Artwork by

F. Mecarelli).

language to identify both a set of detailed pat-terns7 to complete the principal patterns, andthe structure of their relationships, visualisedin the form of a jigsaw puzzle8.Promenade was used to ‘create a centre for

public life: a place where you can go to seepeople, and to be seen’ and ‘encourage thegradual formation of a promenade at the heartof every community, linking the main activ-ity nodes and placed centrally, so that eachpoint in the community is within 10 minute’walk of it’. To complete the pattern the design-ers made use of Pedestrian street, Shoppingstreet and Street cafe. The project establisheda strong relationship between Children in the

city and Network of paths and cars.

7 The full list of these patterns is the following: Scatteredwork, Activity nodes, Shopping street, Network of

paths and cars, Connected play, Street cafe, Pedestrianstreet, Paths and goals (see Alexander et al., in ibidem).

8 In this diagram the interlocking pieces represent thepatterns that realise the two options (retescuole andpedonalizz): the nearer are the pieces, the strongershould be the connection between the correspondingphysical spaces and functions in the project.

THE REHABILITATION PLAN

The new procedure was later utilised by theteam to draw up a preliminary design for therehabilitation plan – but, this time, with the dir-ect participation of the inhabitants of Cento-celle Vecchia. The working group was nowopened up to the residents of the neighbour-hood and some representatives of the localgovernment. They were invited to participatein an introductory session of the workshop, inwhich the proposals were illustrated in somedetail, explaining the principles of SCA and therules of A Pattern Language.During the sessions that followed, the

decision areas and options changed consid-erably; three new foci were identified andnew uncertainty areas emerged through theinteraction of the working group with localpeople. Many people without planning anddesign expertise expressed their interest in APattern Language, and some of them utilisedit to represent in greater detail their preferredoptions.

325



This experience confirms our view that, at least in the field of urban design, the cyclic procedureworks better than each approach – Visioning, SCA and A Pattern Language – employed as astand-alone procedure. This is because a participatory scenario helps to identify the aspirations ofa whole community, but we need SCA to develop the incremental strategy that helps us to carryit out. Then A Pattern Language enables significant field knowledge to be introduced that SCAalone cannot offer. On the other hand SCA is necessary to establish the connections betweenthe options that entail physical and functional transformations of the territory, and also the non-spatial options such as policy choices in the field of education, social welfare or energy productionthat, generally, A Pattern Language disregards. Also, the introduction of social, environmentaland financial sustainability as comparison areas brings these important future considerations tothe attention of local people.

326

13.8 Fast Forward with Strategic Choice

Mutual consulting for small enterprise development

By Rebecca Herron and Dennis Finlayson

The format of a mutual consulting workshop has recently been found to offer a fastyet effective means of introducing the philosophy and methods of the Strategic ChoiceApproach to a medium-sized group of experienced decision-makers who have been broughttogether from different organisational settings through some kind of training or managementdevelopment programme. For it enables the participants to start interacting with eachother quickly in a structured yet informal way, around examples of the decisions anduncertainties that they currently face in their own management roles. The two authors ofthis contribution draw lessons from the use of this framework when working with ownersof small food businesses in the agricultural county of Lincolnshire in eastern England.

Rebecca Herron joined the University of Lincoln in 2000 to direct the Community OperationalResearch Unit after it moved in 1998 from Northern College in South Yorkshire, where ithad been formed 10 years earlier on the initiative of Jonathan Rosenhead while serving asPresident of the Operational Research Society. Before coming to Lincoln, she worked asplanning and development officer for the City of Nottingham Education Department. Sheholds a doctorate in mathematics and operational research from the University of Leeds.

Dennis Finlayson studied development economics at the University of Southampton thenworked on consulting and training contracts in several Asian, African and Latin Americancountries before coming to the University of Sheffield in 1986 to conduct residential coursesfor planners and senior administrators from the developing world. Here he met John Friendand started introducing SCA and STRAD as vehicles for enabling the participants to worktogether on issues facing them in their own development roles. This programme was trans-ferred to the Lincoln School of Management in 1997.

THE FAST PROJECT

In 2000/2001, the authors served as jointco-ordinators of a development project thatengaged managers from small and medium-sized enterprises in the agricultural countyof Lincolnshire in a programme of learn-ing, reflection and consultancy to strengthentheir capacities to adapt to change. This

project1 was supported by the EuropeanSocial Fund and conducted by a team fromthe University of Lincoln’s Faculty of Busi-ness and Management. It enrolled 15 ownersand senior managers from local food indus-tries in a series of University-led workshops

1 Codenamed FAST, for Fully Acquainted with Systems

Thinking.

327


and site-based consulting visits. Differentmembers of staff introduced complement-ary approaches including Checkland’s SoftSystems methodology;2 Beer’s Viable Sys-tems Diagnosis3 and the Strategic ChoiceApproach.

THE MUTUAL CONSULTINGEXERCISE

After an introductory meeting and overnightstay at a hotel, the programme began witha two-day module to introduce the stra-tegic choice approach as a decision-focusedapproach to managing complexity, to be fol-lowed by further modules that introducedother methods based on systems concepts.Most of the first day was spent in groups

of three, exploring examples of decision prob-lems on each participant’s current agendathrough a mutual consulting design. Thisformat had already been well received bysimilar groups, most of which had broughttogether managers from public sector or vol-untary organisations on a management devel-opment programme.4

Each group started work with a short ses-sion in which each participant was askedto offer one example of a difficult currentdecision problem – in strategic choice terms,a single decision area – that they were pre-pared to offer for discussion. After a lunchbreak, there followed a longer and more struc-tured session in which the members of eachgroup rotated through the three roles of prob-lem owner; consultant; and recorder, spendingabout 30 minutes in each.The consultant was asked to use a check-

list to take the problem owner through a set of

2 See Checkland (1989).3 Beer (1981).4 For example a programme on community health ser-

vices at the University of Leeds; an MBA course atKeele University for health managers from develop-ing countries; and a group of public sector managersattending a residential week for MBA students atSheffield Business School.

eight questions based on a simplified versionof the four modes of strategic choice. Mean-while the recorder took notes towards build-ing a ‘problem profile’ to be presented later tomembers of the other groups. The sequenceof questions was as follows:

• Your selected decision problem?• Who’s involved?• Any related decision problems?• Options for your selected decision problem?A, B (or C?)

• Advantages to A versus B?• Uncertainties over preference?• Explorations to reduce uncertainty?• Your suggested progress strategy?

Readers of earlier chapters will recognisethis sequence as based on a stripped downlinear version of the four-mode cycle of thestrategic choice approach, as presented inFigure 8, simplified by the requirement tofocus on just a single decision area.

THE MUTUAL CONSULTING DAYEVALUATED

We had deliberately chosen the mutual con-sulting exercise as an early ice breaking activityfor several important reasons:

• It requires little complicated backgroundtheory;

• It engages all participants and acknowledgesthe contribution of each (democratisation ofthe expert);

• It supports different personalities, allowingexpression without humiliating the less con-fident participants;

• It facilitates the speedy identification ofmajor concerns immediately apparent to par-ticipants;

• It draws on the business experience ofco-participants to cross-examine problem-atic situations.

This exercise was judged a great success.It served not only to identify business prob-lems but also to act as a social catalyst enablingparticipants to quickly learn something of the

328


experiences and skills of other members ofthe network. The process itself created anaudible buzz as participants set about theirdiscussions on their favourite topics. The facil-itator’s role was in this instance mostly oneof co-ordination and time-keeping, calling thechanges within groups every 30 minutes.One of the difficulties of discussing live busi-

ness problems and choices is of course thatof trust. How do you discuss the issues ofgreatest interest to you when potential com-petitors may be in the room? How muchshould such a person learn about your per-sonality or response to strategic decisions?Even more basically, how do you avoid embar-rassment in expressing your views amongststrangers? These issues hadmanifested them-selves at the launch event of the project prior tothe mutual consulting session, with very non-controversial business issues being offered fordiscussion. However, after the overnight stayin a local hotel the atmosphere softened andone of trust and friendship started to emerge.This had a powerful effect on the subsequentengagement with strategic choice and its soft-ware, STRAD.

THE PRESENTATION ANDDEMONSTRATION DAY

A week later, all the participants met againto present wall posters of the various prob-lem profiles developed the previous week, ina ‘gallery’ spread over two walls of the meet-ing room. They were then asked to use smallcoloured dots to place votes against those pro-files, other than their own, that theyweremostinterested in seeing explored further. By a nar-rowmargin, the most votes went to the profileoffered by a sole trader who managed a his-toric working windmill in a small rural town.This business mills floor in the traditional man-ner using awind-poweredmill. Not surprisinglymuch of its business revenue is not from thesale of wholesale flour but from tourism, tea-rooms and speciality products.The manager was invited to come to the

front of the room to work with John Friend,

as organiser of the two-day strategic choicemodule, acting as facilitator. Another parti-cipant volunteered to record the progressmade towards building a commitment pack-age using the STRAD software on a laptopcomputer. The information already displayedin the format of the problem profile was usedat a starting point; but it was now possibleto work on a focus of a few linked decisionareas rather than the single focal decision arearequired by the problem profiling exercise.The picture that gradually developed on the

screen of the laptop was projected on a wallfor all to see. The reliance on a volunteer ratherthan an expert to operate the keyboard andmouse may have slowed progress somewhat(in that each step in the input of new inform-ation had to be spelt out verbally by the leadfacilitator – ‘now click on X, then type in Y’).However it also made the input process moretransparent, as experienced individuals mightwell be tempted to move through the programtoo rapidly for those not accustomed to its lay-out and procedures.Yet, over the course of little more than

an hour, significant progress was made inthe shaping, designing and comparing modes,leading into some discussion of the manage-ment of uncertainty and the creation of a pro-gress package.

COMPUTER-AIDED FACILITATIONREVIEWED

Used in this way, the Strategic ChoiceApproach was well served by the STRADsoftware. Instead of a workshop where allparticipants were already immersed in theproblem under consideration, here was asituation where one key decision-maker wassurrounded by potential co-facilitators andexpert advisers – i.e. the other participants – inaddition to the lead facilitator.This set-up enabled:

• Both broad-brush information and finer detailto be recorded;

329


• The sole trader as decision-maker to controlthe development of the SCA model, jointlywith the consultant;

• The experience and advice of others to beinput at key stages as the decisions werebeing explored.

An interesting feature of the process wasthat it enabled the technical detail to becaptured through the software whilst themore subtle social aspects were managedthrough the social interactions of the group.That is, it was not only the model of thedecision problem itself that was valuable butthe increased understanding gained by the par-ticipants through their interaction with eachother as the picture gradually built up.

FOLLOW UP THROUGHCONSULTING VISITS

It was valuable that the design of the pro-ject enabled the training phase to be followed

by a consulting phase in which the trainingproviders visited the participants at their ownplaces of work. John Friend agreed to workwith four of the participants – visiting the work-ing windmill twice, and also a partnership mak-ing traditional cheeses, a café owner and themanager of a low-price city food shop. Twomembers of the project steering group, bothwith extensive knowledge of the food sec-tor, acted as co-consultants on some of thesevisits.Whilst the visits had the disadvantage of los-

ing the input from all the other participants thathad been possible in the classroom context,it did mean that issues could be explored inmore depth. Now it was possible to overcomeboth the previous time limitations and also thenatural reservations that most people have ofdiscussing theirmost troubling issues in public.Of course, we still do not know what was notbeing shared with us; but we were now dis-cussing investment decisions and even stra-tegic questions of reorganisation and closure.


SCA showed once more that it can be a flexible approach, both in the nature of the decisionsexplored and in the delivery formats and environments used. Its ability to translate quickly fromeveryday experience was appreciated by the business people participating. While SCA had pre-viously been more widely used in public sector and voluntary fields of decision-making thanin the world of business, here was a demonstration that it could also be effective in support-ing small businesses where the challenges of development are primarily to do with managingexternal complexities, rather than with overcoming barriers arising from internal differentiationof roles. Clearly, some difficulties arose from the existence of competitive relations among thosepresent, but these did not jeopardise the success of the exercise. Members of the project steer-ing group also saw a wider potential for use of SCA and STRAD not so much by managers ofsmall businesses themselves as by their business advisers, as a supplement to tools based onfinancial modelling and other more formal quantitative methods. The second author, from hisearlier experience in teaching strategic choice and related methods, saw the mutual consultingworkshop as a valuable addition to the SCA toolbox, especially in community development andother relatively open decision settings (Finlayson, 2004).

330

13.9 The Plutonium Predicament

Managing conflict through strategic action planning

By Richard Harris

This paper is about adapting an SCA tool, the ‘Progress Package’, to a complex, conflictualand long deadlocked discussion. The process, which became known as ‘strategic actionplanning’ (SAP), combines scenario working and strategic planning in addressing a situationof high uncertainty and complexity. The subject was Plutonium (Pu), an important nuclearmaterial, and, more specifically, what to do with the large civil stockpile in the UK that hasbeen built up since the nuclear industry began in the 1940s. This was one of the challengesfor the British Nuclear Fuels Ltd (BNFL) ‘National Stakeholder Dialogue’, a groundbreakingdialogue process independently convened by The Environment Council – an UK charitywhich aims to enable collaboration and sustainable environmental decision-making.

Richard Harris is an independent facilitator, mediator and process consultant working mainlyin environmental conflict management and strategy development. Richard had a first careeras a contractor, consultant and local government officer in forestry and arboriculture, workingin both Australia and the UK. In 1990 a second career developed with a wider environmentalrole in local government management. This quickly led to a fascination in the processesby which people can affect the decisions which impact upon their environment, and thegrowing field of public and stakeholder engagement. In 1998 Richard established his ownpractice. He took over the facilitation of the BNFL National Stakeholder Dialogue fromAllen Hickling in 2001. (See www.the-environment-council.org.uk for all the reports of thedialogue.)

THE CHALLENGE

Plutonium is a thorny subject, not leastbecause of its historic military uses; con-sequent concerns about proliferation; and per-ceptions of risk and hazard in nuclear powergeneration. It is also difficult because of anextensive history of poor relations between thenuclear industry and its critics. This has led tocommunication, when there is any, being char-acterised by confrontation.

The project involves up to 120 stake-holder organisations including the industry,departments of central and local government,regulators, trade unions and campaigning (dis-armament, environment, transport, etc.) andcommunity groups. The dialogue has beenoperating since 1998 and seeks to addressthe long-standing conflicts over nuclear issueswhilst informing future policy. Discussionsare primarily undertaken in facilitated workinggroups (12–20 representative stakeholders)

331


and have, not surprisingly, tackled some verychallenging subjects.

CONFLICT AND UNCERTAINTY

As with most areas of conflict, the positionsadopted by the stakeholders in this case arealmost entirely caused by their reaching dif-ferent sets of assumptions in the face of thesame, or similar, uncertainties. These assump-tions are typically informed by personal valuesand beliefs (which tend to be less negotiable),and also by things such as experience and edu-cation (which offer more potential flexibility).On a simplistic level these differences resultin positional, ‘black vs white’ arguments aboutuncertainties such as: whether one optionis safer than another, or whether the con-sequences of an action are likely to be X or Y.When plutonium management is discussed,

it is common for two broad views to emerge.The anti-nuclear view is that it is a danger-ous waste and should be immobilised1 and dis-posed of without delay. The pro-nuclear lobbyconsiders it a valuable resource which shouldbe reused as a component of fuel for nuclearreactors, thereby realising the energy value forpower generation.So, a key part of the work of the Plutonium

Working Group (PuWG) had to include acknow-ledging and developing a shared under-standing of the uncertainties relating to Pumanagement and perhaps some agreementsabout how these might be reduced or other-wise managed.

LEARNING AND NEGOTIATION

A significant challenge for the facilitation teamlay in developing a process by which a mixedstakeholder group could learn more about therange of strategic options (including their ownpreferences) and their inherent uncertainties,without resorting to positional, well-rehearsedarguments.

1 Immobilisation involves ‘locking up’ the Pu in an inertmaterial such as glass (vitrification).

In this it may help to understand that a use-ful thing to do in deadlocked situations, wherethe parties seem to have run out of flexibilityor room for creativity in their discussions, is tomove from a ‘negotiating approach’, to a ‘learn-ing approach’. This is done in the hope that, byenabling the parties to learn more about eachothers concerns, needs and assumptions, newinsights and perspectives on the problem canemerge. The approach we developed becameknown as SAP.

STRATEGIC ACTION PLANNING(SAP)

The SCA ‘Progress Package’ provides an excel-lent framework for strategic planning in theface of uncertainty. So we adapted it for scen-ario working – an approach which becameknown as SAP.2 This would involve the PuWGgroup in agreeing scenarios which reflectedthe range of ‘possible futures’ and then under-taking high level strategic action planning foreach, before analysing results and develop-ing recommendations to BNFL. The PuWGworked through four key stages,3 as follows.

Strategic option selection (Scenarios)

Four scenarios were chosen:

• Two scenarios concerning Pu immobilisation(one in existing plant, one requiring a newimmobilisation plant to be built); and

• Two scenarios looking at Pu going into newfuel (one for existing nuclear power stations,one for a new fleet of stations).

Together these provided adequate breadthacross the rangeof ‘possible futures’ and itwas

2 Using the term ‘strategic action planning’ (SAP), todescribe the process, clearly appealed to most people.Had we, perhaps more accurately, called it ‘strategicoptions exploration and learning process’ few wouldhave supported the approach. As it was, the team hadalready put much effort into ‘selling’ ‘The Managementof Uncertainty’ with remarkably little success.

3 The nature of theworkwas such that it was not quite solinear as this list may suggest; in fact it involved muchcyclic working within and between stages.

332


SCENARIO: USING Pu IN NEW FUEL RODS Crucial assumption:

Adequate, safe and secure Pu storage will be available over the life time of the fuel manufacture

programme

ISSUEThe heading

under which

uncertainties

arise

DECISION The current

decision

EXPLORATIONSRequired to reduce the

uncertainties

DEFERREDDECISIONS

Decisions which can be

taken later – usually

when the uncertainties

have been reduced

CONTINGENCYPLANS

In the event that the

assumption turned out to

be poor, this activity will be

carried out

NOW LATER

PuSTORAGE

(see

uncertainties

above)

Continue to

use existing

stores

pending fuel

manufacture

By end 2004, assess

whether existing storage is

safe, safeguarded and

structurally sound against

risks, now and throughout

the envisaged programme

By March 2005, decide

how, when and which

actions to undertake to

ensure that safety,

security and safeguards

standards are and will

continue to be met

In the event that facilities

are not adequate in some

way – upgrade existing

stores or build new

stores

Figure 13.3 An Example Showing How an Issue is Managed in an SAP Table.

This is one significant issue within the two scenarios looking at fuel manufacture (i.e. using Pu as a significant

component of new fuel rods for nuclear power stations). The time it would take to process all the stored Pu

(20 years plus) creates uncertainties relating to the nature and availability of storage facilities, such as: Is current

Pu storage capacity adequate? Will it be available over the programme lifetime and will it be able to cope with

demand peaks and troughs? Will it continue to be safe and secure over the programme lifetime?

NB This is a simplified extract and some minor changes to the original content have been made to assist the non-

technical reader. Each SAP table typically consisted of 6–10 such issue ‘rows’ most of which were inter-related. For

example, the above contingency plan (building new stores) connects with uncertainties under the issue heading

‘Regulation’ because of secondary uncertainties such as planning requirements and other regulations.

felt that, although there were other possiblescenarios, this range should provide adequatelearning which could be applied to any otherscenario or variant. For each scenario a set ofinherent issues and uncertaintieswere agreed.

Exploring strategic options

The whole group (usually 16–20 in attendance)worked through each scenario, completing aSAP table (see example below) covering allthe issues and uncertainties (Figure 13.3).In essence, this involved charting a strategicroute, with a focus on short term, robustactions, but mindful of longer-term strategicconsequences. It was essential that the groupkept together since this was a learning focusedprocess and any sub-group working bought

the risk of the learning being confined to afew.4 In order to avoid reversion to positions,the group members adopted an attitude that‘this scenario has been chosen for develop-ment – the SAP table is being devised tofind out what is entailed, not to question itsvalidity (that comes later)’. This drew on thedeveloping trust between group members andprovision of a safe working environment (espe-cially through confidentiality and independentfacilitation). Without either it would not haveworked.Once a SAP table had been completed for a

scenario (typically this took two or three cycles

4 See Brendan Hickling’s contribution concerning‘differential learning’ in Section 13.13 of this chapter.

333


through the process), all the significant dateswere identified and mapped out on a timeline;this gave a graphic view of the strategic optionbeing described. This often revealed criticalpoints in time for example, where a strategicpath was weak or vulnerable to early assump-tions and decisions turning out to be poor(cf. the contribution from Richard Ormerod inSection 13.4 of this chapter).

Analysis of results

Further analysis involved looking at the resultsfrom all the SAPs and timelines together tolook for linkages, convergence and divergence.A number of new and significant issues wererevealed at this stage, such as: (i) commonality(e.g. where no matter which option was pur-sued certain areas of research were needed)and (ii) conflict between strategic options(e.g. when adoption of one sub-option mighteffectively foreclose contingency plans inanother).

Crystallising and preparing to sharethe learning

Effectively, the process now reverted to anegotiating approach, where the commonlearning was applied as the group set aboutthe task of ‘crystallising’ what they had learntinto recommendations and conclusions. Thisresulted in a ‘headline’ recommendation thata mixed strategy should be developed in theshort term to ensure that contingency planscould be well-developed and significant uncer-tainties could be reduced, through an extens-ive research programme.

What stood out was the degree to whichgroup members were able to agree. The com-mon ground was extensive (but not completeof course) and exceeded the expectations ofmany within and outside the group. Clearly,one reason for this was that the SAP work hadgiven each member of the group the oppor-tunity to test the strategic options – both theirown preferences as well as options that theywould not typically support – in an open, fairand consistent way.

BACK TO THE ‘REAL WORLD’

The work was very well received, almostwithout exception. Many who had not beendirectly involved in the Working Group repor-ted that they found the content very transpar-ent and easy to understand, andwelcomed theclarity about assumptions.Commonly people expressed surprise at the

degree of consensus that the Working Grouphad developed (only the members will everreally know how hard won that consensuswas). One senior civil servant described theirreport as the ‘best analysis of Pu managementoptions that he had seen, and the fact that itwas developed by such as wide selection ofinterests only served to make it more power-ful’. The report has been widely distributed inthe nuclear industry and government depart-ments. Only time will show its true impact onfuture policy – but early signs suggest that thekey messages are being taken seriously.


The use of SAP enabled a mixed group of stakeholders to work effectively together on a verydifficult issue about which there was a long history of deep-seated conflict. The work wasquite time-consuming and at times laborious. However, the results show this to have been aworthwhile investment and commensurate with the complexity and the history of the case. Theapproach, as briefly described, shows a way forward for both sharing strategic dilemmas andmanaging strategic decision-making with stakeholders, especially in contentious policy areas.

334


A number of specific learning points concerning the process are worth highlighting:

• A shift from negotiating toward developing mutual learning can overcome deadlock;• Uncertainty is a significant source of conflict;• The language of SCA needs to be adapted for different audiences;• A ‘safe’ working environment and independent facilitation are essential;• Mutual trust, developed with the aid of confidentiality ground rules, is a key to progress;• Timelines to map strategic options can aid comparison between scenarios.

335

13.10 Dealing with the Dumps

Using decision areas and levels in developing national policyfor used tyres

By Rob Angell

This contribution is focused on the use of Analysis of Interconnected Decision Areas(AIDA) in a project designed to inform UK government policy and general practice byindustry and local government over clearing historic used tyre dumps. The work, involvingabout 150 stakeholders, took place within a stakeholder dialogue process convened by TheEnvironment Council – an UK charity which aims to enable collaboration and sustainableenvironmental decision-making. Some of the lessons I will draw out of this centre on theuse of AIDA and the levels of choice in larger groups; and on getting the concepts acrossto stakeholders with differing experiences.

Rob Angell has been the lead facilitator and process consultant for a range of clients involvingstakeholders in formulating environmental decisions or resolving conflicts. Examples includepolicy and strategy development for national government and the private sector on wastemanagement; the decommissioning of nuclear power stations; and flood alleviation schemesfor the Environment Agency and local government. He has designed and run European widee-consultations for WWF, and run training courses for the public, private and communitysectors on participation, facilitation and sustainable development.

THE PROJECT

The project started in early 2001 when theEnvironment Agency for England and Walesearmarked tyres as a priority waste stream.The driver was an impending ban on wholetyres going to landfill sites from mid-2003, andshredded tyres from mid-2006. They wantednot only to improve disposal practice andavoid tyre mountains appearing around thecountry but to increase environmental aware-ness within the tyre industry. What made thispiece of work different was the desire to act-ively involve stakeholders in deciding how thisshould be done.

The key stakeholders were the UK govern-ment (Department of Trade & Industry); thelarge tyre manufacturers (such as Michelin);tyre reprocessors and recyclers; the Britishcement industry (as users of waste tyres asa substitute fuel); English Heritage (as keep-ers/guardians of land and historic sites); Localgovernment; and the Environment Agency (themain environmental regulator).

THE PROBLEM

Next then, let me describe the problem. Forsome 30 years there have been large dumps –euphemistically called stockpiles – of old usedtyres throughout the UK. A study, carried out

336


as part of this work, listed the locations of13 stockpiles with 0.5–8 million tyres in each.These stockpiles have come about in two

ways. One is the unscrupulous tyre dealer whopretends to set up a tyre recycling businessbut just accumulates the tyres and then ‘doesa runner’. The second is the time that the reg-ulatory system takes in finding out and actingwhen a registered tyre dealer is acting illeg-ally; and then in pursuing matters through thecourts. The end result in both situations is atyre stockpile.The dialogue process dates back to July

2001, when I ran the first workshop for abroad range of stakeholders, who agreed thatthey should ‘come up with a recommendedapproach to clearing the UK’s tyre stockpiles’.By mid-2002 they had decided that this workshould be done by a working group of about15 people (on behalf of the 150 stakeholders).Their task became that of developing a gen-eric strategy to guide those who would haveto carry out a site clearance; those who wouldhave to co-ordinate a site clearance; and thosewho were going to provide funding.

THE APPROACH

The working group started by believing that‘there really wasn’t anything to it – just securethe funding and clearing the stockpiles wouldbe sorted’.So, why use AIDA? Well, my rationale was

based on this: the stakeholders had a goodidea of the underlying issues which have resul-ted in these stockpiles being untouched for solong, and they seemed eager to delve into each‘problem area’. They were not though consid-ering the interconnectedness of their ‘problemareas’; and it was this that I thought was cre-ating the barrier to clearing the stockpiles. Thefirst thing I did was ask them why they weresitting in the room if it was all so easy, and whythe stockpiles were still there after 30 years.After kicking around many ‘issues’, as theysaw it, they started to understand there wasindeed some complexity to the problem.

It was at this stage that I began to talkto them about thinking of their key issuesas areas where decisions would have to bemade by anyone wanting to tackle a tyre stock-pile. They liked this concept because they sawthat if they could work through these decisionareas they would be going a long way to hav-ing their strategy. After much ‘kicking around’we developed an initial set of decision areaheadings:

• Site after-use• Location• Recovery technology• Partners• Leadership• Funding.

Getting to a set of options for each decisionarea proved a lot more difficult for the group toget to grips with. So, at this stage I introducedthem to the concept of levels of choice (cf. Fig-ures 54 and 96). I did this as I sensed that oneof the reasons they were finding it difficult todefine options was that they were talking at anumber of different levels.It turned out to be a really effective move

around the cycle of learning as I could virtually‘see the lights turn on’ for a number of the par-ticipants. Not only did they really seem to getthe concept of levels, but I also felt a changein the group’s attitude towards me. They nowseemed to trust me as a process person andgave me more ‘space’ to take them throughthe AIDA framework.

CYCLIC WORKING

My next move was to help them imagine theyhad resolved the funding issue and had avail-able as much money as they needed. Whatwould they do to clear the stockpile? I foundthe discussion within the group fascinatingbecause for some, practicality and immedi-acy were paramount while for others envir-onmental criteria were more important. Therewas a further question on funding that thegroup had to answer: whether solely to satisfy

337


a funder’s criterion (such as delivering ‘x’ num-ber of full time jobs for ‘y’ period of time), orwhether to combine this with their own criteria(such as environmental).Given that this process was being driven by

the UK Environment Agency, the group didaccept that environmental legislation was atleast the minimum they had to build into theirdecisions.What thismeant in practicewas thatan option of ‘engineered burial’ of a stockpilewas out of the question (essentially, coveringthe tyres as they were with inert material andan upper layer of topsoil). Having taken financeas an unlinked decision area I nowworkedwiththe group on exploring the linkages betweentheir other decision areas. The result was areduction in the number of decision areas thatthey felt they needed to focus on, because itwas these that they realised had the stronglinks between them:

• Leadership• Funding• Method of clearance• Subsequent use of the land.

We spent time working through optionsfor each decision area and working on theuncertainties inherent in these decision areas.I would have liked to have taken the groupthrough the cycle to get to an option tree. How-ever, they were not prepared or interested indoing this as they felt they had got the learningthey needed by this stage.Interestingly, I felt compelled to do an option

tree for myself, outside of the group. I reallywanted to see if it would bring clarity on whatthe group should have recommended be doneand what they should have recommended benot done. I believe it did.Nevertheless, the working group were able

to draft their recommended strategy for clear-ing the historic tyre stockpiles in autumn2003. The paper summarising their work wasapproved by the main stakeholder group and isavailable on The Environment Council’s web-site (www.the-environment-council.org.uk).

TWO TYPES OF PARTICIPANT

During the eight months I felt the group unrav-elled a bit, principally into two segments. Onegot absorbed in the approach to tackling theissue and became enthusiastic about it, whilethe other seemed to get almost bored and a bitdisconnected. I had to work hard to keep themtogether.AIDA is an intense technique that requires

close and collaborative thinking. It magnifiedthe gulf between those I call strategic thinkers,and those I call tactical thinkers. The strategists‘got it’, the tacticians did not. One notableextreme was the consultants who had beencommissioned to carry out the backgrounddata research. They simply could not get thereason for approaching the problem in this wayat all. They thought that their stockpile surveyhad come up with the answers and could notunderstand why the stakeholders did not justsee this.This was highlighted in the extreme when

the working group was writing its report. In anearly draft, in the section on decision areas,what theymeant andwhy theywere importantto the recommendations,what the group even-tually came upwithwas left as a ‘one liner’ – ‘tobe filled in later’. During one discussion I canrecall a participant questioning if they neededto expand on the one line. Thankfully, therewas more than one other who replied, before Icould, that in fact this was a central part of thegroup’s rationale for what they were saying. Itwas one of these stakeholders who went onto write this section.

FOR THE FUTURE

Flowing from this is one lesson that washammered home: that this working group wastoo big to work through AIDA. It is probablyimpossible to find a group of 15 people whoall happen to be strategic thinkers but, in asmaller group (probably of up to seven), therewould have beenmore time,whichwould haveenabled the strategists and tacticians to learnfrom each other.

338


I felt that explaining the concepts in AIDAwas a real struggle until I opened a window(the concept of levels of choice) that allowedthem to see a different way of thinking andultimately reporting on their work.Although you can’t push people to do some-

thing they don’t want to, one thing I could havedone was to work through a reduced numberof decision areas, to get through (at least as

a first pass) this phase quicker. It might haveallowed me to get the group to invest sometime in an option tree for these elements. Hav-ing seen its benefits they might then havehad the motivation to cycle back to the otherdecision areas and on through to a more com-plete option tree. Indeed, you could argue thatthe cycle of learning may not be completewithout an option tree.


From a process practitioner’s perspective, I found the work completely absorbing because I hadto concentrate on the AIDA process as well as helping the group focus on the outcome theyhad been asked to deliver. However, there were two particularly satisfying parts of this work:

• the participants took on board the concepts of AIDA, notably that interconnectedness betweendecision areas is important and something you have to work with (and I know that at least oneof them has gone on to use them in their work);

• the larger group of stakeholders genuinely appreciated the Working Group’s report and therobustness of their work.

339

13.11 Cross-organisational Learning

Sharing insights from managing major construction projects

By Mike Cushman and Alberto Franco

This contribution reports the main lessons drawn from the experience of developing andapplying a strategic choice-based methodology for promoting inter-organisational learningin the construction industry. While the different elements of the resulting methodologyreflect those of the traditional SCA, the focus on project review and learning meant thatmajor changes were needed to the original SCA so that participants could engage inreflection and learning rather than in current decision-making.

Mike Cushman is a research fellow at the London School of Economics and Political Science(LSE). Before joining LSE, he worked in community education in inner London for many years,eventually becoming Head of the Lambeth Adult Education Service. He has used problemstructuring methods in both his academic research and in his consultancy work. Recent workwith strategic choice includes the development of a project review process (with AlbertoFranco) and a major review of the provision of Children’s Health Services (with JonathanRosenhead). He has also worked with John Friend on the development of STRAD.

Alberto Franco is a lecturer in operational research and systems at the University of Warwick.Before moving to the UK from Peru, he worked as a soil mechanics consultant and then as alecturer in mathematics and operations management. Much of his recent research has focusedon exploring how problem structuring methods can assist the development of different typesof inter-organisational collaboration. He has held academic posts at the London School ofEconomics and Political Science, the University of Strathclyde and Kingston University, andconducted research for a number of organisations including Whitbread, Taylor Woodrow,Thames Water and Bombardier.

BACKGROUND

During the mid-1990s, the UK constructionindustry initiated a move from single-tenderedcontracts, where the future behaviour ofother parties is seen as a marginal consid-eration, to partnering arrangements, whereactions have to be weighed against their effecton future interactions. Partnering was seenas playing a key role in the generation of

feedback learning processes, which in turnhad been identified as a critical missing pro-cess in conventional construction arrange-ments (Bennett and Jayes, 1998). Oneway of generating learning in projects isthrough post-mortems, and indeed construc-tion firms often undertake in-house projectreviews. On the other hand, however, thereis little tradition of exchanging perceptionswith other firms – a lack that has impeded

340


Figure 13.4 The Cross-Organisational Learning Approach (Franco, Cushman and Rosenhead, 2004).

learning throughout the industry (Barlow andJashapara, 1998).It was against this background that the

authors participated in an action research pro-gramme, known as the B-Hive project,1 andaimed at increasing the value of constructionprojects. Although both authors were famil-iar with the Strategic Choice Approach beforethe project, they had relatively little experienceregarding its application in multi-organisationalsettings. Yet the partnership environment inwhich the research was developing seemedat the time an appropriate and ‘safe’ settingfor the application of the SCA. That is, therewas a setting of multiple stakeholders within

1 Engineering and Physical Sciences Research Counciland the Department of the Environment Transport andthe Regions funded project under the IMI Link/IDACprogramme, project no: IMI/c/02/013.

a context of broad agreement which needsto be made operational. Towards the end ofthe project, strategic choice had become partof a wider framework for cross-organisationallearning. The framework, labelled the cross-organisational learning approach or COLA (seeFigure 13.4) is aimed at eliciting, reflecting anddistributing formal and tacit knowledge withinand across partner firms.

THE COLA APPROACH

A key aspect of COLA is a project reviewworkshop which draws significantly on theStrategic Choice Approach (Franco, Cushmanand Rosenhead, 2004). One of the main con-straints encountered during the research wasthe lack of time available to conduct one-offproject reviews. These had to be carried out

341


in less than one working day and thus onemajor innovation and input to the workshopwas information collected through a question-naire. These allowed the facilitators to preparea list of candidate discussion areas to save timein the workshop. The questionnaire asks par-ticipants to rank various aspects of the pro-ject and the performance of other firms. Theserankings are supplemented by free text com-ments which usually proved more revealingthan the rankings. The questionnaires typicallytook about 1 hour to complete.A COLA workshop iterates through four

stages: focus, options, plans and commitment,which are derived from strategic choice’sshaping, designing, comparing and choos-ing modes. COLA may make use of someof the distinctive tools of strategic choice,for example decision graphs, or comparat-ive advantage charts. However it is in theemphases of the stages and in the flow ofthe process between them, rather than inthe technology employed within the stages,that the strategic choice influence is mostapparent.During the focus stage, participants identify

key opportunities for improvement. However,and this is another departure from traditionalstrategic choice, they discuss first the pro-ject’s victories and successes. The discussionis informed by the results of the pre-workshopquestionnaire, and the output of this stage is afocus consisting of a set of urgent, importantand interconnected opportunities for improve-ment that is small enough to be manageableduring the workshop.In the options stage, participants are helped

to generate options for improvement withinthe focus chosen in the previous stage.A consideration of the implications of thedistinction between single- and double-looplearning (Argyris, 1999) led to another signi-ficant change in this stage of the process forlater workshops. Options can appear to beself-evidently beneficial, butmore deep-seatedproblems may exist which prevent apparentlyobvious innovations either being implemen-ted or, if implemented, achieving the inten-

ded improvement. A discussion of possibleblockages to action was introduced into thisstage, in which discussion was focused ondevelopment of initiatives to remove theseblockages.A further constraint at this stage was

included in later workshops; actions for debatewere limited to those that could preferablybe initiated by workshop members or, at min-imum, be initiated by their line managersor others that individual workshop memberscould directly influence. This prevented aspir-ations masquerading as actions.The plans and commitment stages closely

reflect strategic choice’s comparing andchoosing modes. The former involves parti-cipants in identifying the value criteria neededfor the comparison of options for improvementand in evaluating the options against these cri-teria – though in the process they commonlyalso uncover uncertainties which stand in theway of identifying a straightforward preferredsolution. The latter enables the group to makeprogress towards agreement in some areasand set up explorations and/or consultations inothers (see Agreed Actions and Explorations inFigure 13.4).

LESSONS DRAWN

During the period of the research WhitbreadHotel Company were in the early stages ofa partnering arrangement with their serviceproviders for a series of refurbishments oftheir hotels to meet the standards of theirrecently acquired Marriott franchise and theconstruction of new Marriott hotels. Three ofthe projects completed during the researchwere the subject of a post-completion reviewusing the COLA workshop. The main lessonsdrawn from these workshops and others heldwith Thames Water and Taylor Woodrow arebriefly described below.The use of a pre-workshop questionnaire

allowed the intended time saving and com-pression of the workshop. However, it intro-duced the risk of a facilitator-imposed agenda

342


and thus it was necessary to explicitly nego-tiate the draft agenda and candidate decisionareas with the workshop participants. Mem-bers may raise issues in a questionnaire thatthey would not raise in the workshop and viceversa. Thus the use of the questionnaire, sup-plemented with additional issues identified atthe workshop, allows themaximum number ofpotential issues to be included in the debate.This pre-description of decision areas is also

possible in a conventional SCA workshop,where it is dealing with issues that havealready been discussed in a number of foraand the facilitator has access to the recordsof the previous activities. This allows a fastentry into activities which leads participants toexperience progress and build confidence inthe benefits of the approach (Cushman andRosenhead, 2004).The early consideration of victories and suc-

cesses helped to building confidence and avoida blame environment and a retreat to defens-ive routines. This is particularly relevant to aproject review environment where there is atradition of adversarial relations and lack oftrust, but is of wider application. An SCA work-shop will often be held when other approacheshave failed and the members bring a his-tory of failure and frustration to the work-shop. Therefore, in order for the workshop tomake progress, it is helpful for the facilitatorto draw early attention to group successes.While involving representatives of competinginterests in the workshop, in the case of COLAdifferent members of a supply chain, maylead to difficulty, it also inhibits loading blameonto others as criticisms must be raised dir-ectly rather than easily passed onto an absentstakeholder.A focus of concern on blockages to action is

an innovation of wider application. In any stra-tegic choice exercise participants may suggestoptions that may be desired but are not easilyattained. A review of possible blockages maycause decision areas to be redrawn to focus

attention on necessary actions to achieve adesired state, or options to be discarded asdesirable but not attainable in the currentexercise. Discussion of blockages may alsodisclose uncertainties of related areas or uncer-tainties about values.Similarly the constraint of focusing on

actions achievable by the workshop parti-cipants, or at least achievable by people orbodies who may be influenced directly bymembers of the study group means that thatitems appearing in the progress package willlead to action rather than failure.There is one final point of novelty in the

application reported here. It can be said thatsome success has been achieved in apply-ing a strategic choice-based methodology in aprocess-detached mode of operation (Friend,1999). This means the use of strategic choiceto develop first an idealised state of affairsfor the project partners (e.g. a ‘zero defects’project) before moving towards the develop-ment of alternative options for action. Strategicchoice provided a useful framework for reflec-tion and learning as well as action. Strategicchoice can thus form the basis of a methodo-logy to support reflective learning, and the fourmodes of SCA facilitate learning-based work-shops as well as workshops designed to planfuture action.

POSTSCRIPT

Two years after B-Hive, COLA is still usedwithin Whitbread Hotel Company as part oftheir project review procedureswithin the part-nership. The COLA processes and tools arenow part of the process manual which everyproject manager should follow, and workshopfacilitation is usually undertaken by Whitbread.Whitbread have extended the use of COLAfrom the Marriott Hotel projects where itwas piloted in B-Hive to the much larger pro-gramme of Travel Inn Reservations.

343



In conclusion, three main lessons regarding this experience can be summarised as follows:

• Collecting information before a workshop (when possible) leads to a more exhaustive work-shop agenda, which can save time and, subject to ratification and amendment by theworkshopparticipants, increase participants’ ownership of agenda. This also allows the early identifica-tion of potential conflict for which appropriate ‘confidence-building’ tasks need to be designedand included early in the workshop.

• The explicit elicitation of blockages to action, and identification of activities to overcome them,increases the feasibility of implementation as well as enables the uncovering of uncertainties.

• The use of strategic choice in process-detached (rather than process-engaged) mode is pos-sible and useful.

REFERENCES

ARGYRIS C (1999) On Organizational Learning (Second Edition) Oxford: Blackwell BusinessBARLOW J and JASHAPARA A (1998) Organizational Learning and Inter-Firm ‘Partnering’ in the UK Construction

Industry, The Learning Organization, 5 (2), pp. 86–98BENNETT J and JAYES S (1998) The Seven Pillars of Partnering: A Guide to Second Generation Partnering. London:

Thomas Telford PartneringCUSHMANM and ROSENHEAD J (2004) Planning in the Face of Politics: Reshaping Children’s Health Services in Inner

London in Handbook of or/Ms Applications in Health Care (F. Sainfort, M. Brandeau and W. Pierskalla eds) Kluwer,Boston, MA, USA

FRANCO LA, CUSHMAN M and ROSENHEAD J (2004) Project Review and Learning in the Construction Industry:Embedding a Problem Structuring Method within a Partnership Context, European Journal of Operational Research,152 (3), pp. 586–601

FRIEND J (1999) Process Engagement as a Characteristic of Problem Structuring Methods. Paper presented to theTriennial Conference of the International Federation of Operational Research Societies. Beijing, 16–20 August 1999

344

13.12 Capacity Building in Venezuela

Using strategic choice methods with policy-makers and students

By Elisenda Vila and Ana María Benaiges

In Venezuela, the Strategic Choice Approach has become increasingly influential in publicplanning and policy-making since it was introduced in the early 1990s by staff of theInstitute of Urbanism in the Central University of Venezuela. It gained added impetus in1999 when the University’s Professor of Planning was appointed as Minister of Planning andDevelopment in the national government. The authors have been involved in applications inthe fields of public services management, restructuring of the sugar industry, post-disasterreconstruction and regional economic development. After reviewing the wider background,they review their experience in an application to the management of solid waste on theextensive campus of the Central University. This project influenced management decisionswhile providing significant learning for the students involved in the work.

Elisenda Vila is a policy adviser to the Venezuelan Ministry of Planning and Development.Through the Venezuelan Institute of Planning IVEPLAN,1 she runs courses in strategic choicefor staff of government institutions including the national oil corporation. In 2002, undercontract to IVEPLAN, she translated the second edition of Planning under Pressure intoSpanish. After her first degree in architecture and a Master’s degree and Doctorate in UrbanPlanning, she became full Professor at the Institute of Urbanism at the Central University ofVenezuela. Here she continues to conduct Master’s courses in town planning, policy-makingand local action.

Ana María Benaiges is a consultant in planning processes and public services managementand a professor at the Institute of Urbanism in the Central University of Venezuela. Shegraduated in computer science then gained a Master’s degree in Urban TransportationPlanning. With Victor Poleo, she first applied strategic choice methods to the management ofwaste in an industrial city in the south of Venezuela. She has since used them in a project ondevelopment of the sugar industry for the Ministry of Commerce, and in a review of wastemanagement for the capital city of Caracas.

INTRODUCING NEW APPROACHESTO PLANNING

During the eighties, an informal group of plan-ners in Venezuela, who had become criticalof traditional thinking about comprehensive

planning, started experimenting with altern-atives based on action research principles,including the Strategic Choice Approach. VictorPoleo and members of this group began to

1 http://www.iveplan.gov.ve.

345


introduce courses in these alternative planningmethods, first at the Naval Staff College thenat the Central University of Venezuela, and totry applying them in practice.Then, in 1999, one of the leading advocates

of this new thinking, Professor Jorge Giordaniof the Centre for Development Research at theCentral University, was appointed by PresidentChavez as Minister of Planning and Develop-ment in his new national government. Laterthat year, he invited Jonathan Rosenhead andJohn Friend to visit Caracas to advise on waysof developing new planning capacities withinhis Ministry.Shortly after this invitation, in December

1999, exceptional rains fell on the Avila moun-tain range separating Caracas from the coast.These rains triggered catastrophic mudslidesthat engulfed several coastal communities,with heavy loss of life and severe destructionof buildings, infrastructure and livelihoods.In March and April 2000, representatives

of many agencies involved in the responseto this disaster were invited to participatein a series of strategic choice workshopsaddressed towards issues of local and regionalreconstruction. John Friend and JonathanRosenhead acted as facilitators, with ourselvesand other associates acting as co-facilitatorsand translators. Photographs fromone of theseworkshops appear in Figure 82.By the time of a further visit by Jonathan

Rosenhead and John Friend in 2001, atten-tion had turned to the challenge of developinga legal and policy framework for invest-ment in special economic development zones(ZEDES). These were designed to regenerateemployment in rural areas that had becomedepopulated by a drift of people to the cities insearch of jobs.

A PROJECT ON CAMPUS WASTEMANAGEMENT

Meanwhile, on the extensive campus of theCentral University of Venezuela, both of ushave been involved in the design and man-agement of a participatory project to improve

the management of solid waste and thusreduce environmental damage. The site cov-ers 115 ha, within which is contained a rangeof urban activities equivalent to those of a cityof almost 100 000 inhabitants.The decision-makers in this case included

a range of actors both within the Univer-sity and in other public and private agencies.Also involved in the project were classes ofMaster’s students in urban planning. Not onlydid they provide a valuable resource for inform-ation collection and analysis; at the same timethey were able to build an understanding ofthe complexity of the decisions to be madeand of the relationships among the variousstakeholders, thus developing their capacitiesas decision-makers and policy advisers of thefuture.The principal agency in the management and

supervision of the solid waste services is theuniversity’s División de Servicios Generales(DSG), which is responsible for collecting allthe non-hazardous waste generated on cam-pus every day through 11 collecting centres,strategically located within the campus. Up tonow, the DSG has had no control over the wayin which users discard waste at these collect-ing centres. Inadequacies in the handling ofwaste could be attributed to lack of containercapacity, difficulties in lifting and of course lazi-ness. The disposable bags provided in hall-ways, on pedestrian routes and on sidewalksare exposed to dogs that rip up the bags andspread the contents. Also the university pop-ulation tends to discard papers, plastic tum-blers and food containers outside the garbagebins, which are sited in open spaces andfrequently full. Demolitionmaterials from refur-bished buildings are one of the biggest prob-lems because this waste is not taken away byFOSPUCA, the municipal collecting company.For several years, sporadic actions had beentaken to deal with the lack of proper and per-manent supervision over all the types of wastegenerated, but with very little success. Then,at the beginning of 1998, a working team wascreated to improve the university’s solid wastemanagement. This brought together the DSG,

346


the university’s Institute of Urbanism and itsDepartment of Sanitary Engineering.Themaster’s students carried out interviews

with the various stakeholders to understandtheir different points of view. The philosophy ofSCA was introduced from the outset, so as toapproach the problem not in holistic terms butin a selective yet co-ordinated and interactiveway.Our concern was to understand the pro-

cesses and to identify the possible actionsopen to the principal stakeholders. Amongthese were identified the FOSPUCA super-visors, the managers and employees of coffeeshops and restaurants, themaintenance chiefsof faculties, the supervisors of green areas, andthe student ecological groups.

IDENTIFYING KEY DECISIONS

We experienced some initial difficulties withthe students in shaping the dimensions andperceptions of the problem in terms of decisionareas. So we tried classifying them by the cat-egories of technical, managerial, political, fin-ancial and behavioural. After discussing thelinks among our initial decision areas, and theirdegrees of importance, urgency and control,we went on to discuss the choice of a problemfocus.We agreed on two main foci: the one tech-

nical/operational, and the other concernedwiththe behaviours of generators of waste. Wedecided to study the latter in a separate way,due to its importanceandour lackof knowledgein that topic area. The initial focus included thefollowing five decision areas:

1. COLLECTION? – how to collect solid wastefrom the collecting centres? OPTIONS: asnow; rent dumper trucks; hire small enter-prises.

2. TRANSFER? – how to deal with waste notpicked up by FOSCUPA? OPTIONS: organ-ise according to type; exclude demolitionwaste.

3. TRANSPORT?– how to pick up wastefrom producers? OPTIONS: organise waste

transport; DGS collect from each producer;relocate collection centres.

4. CONSTRWASTE? – how tomanage demoli-tion waste? OPTIONS: DSG in existing role;redesign of contracts.

5. BEHAVIOUR? –what to do to change users’behaviour? OPTIONS: UCV campaign in thewhole campus; permanent supervision byDSG.

BUILDING A FRAMEWORK OFCOMPARISON

A set of four comparison areas was agreed asa basis for evaluating the decision schemes.These were:

1. COSTS – associated with each scheme.2. CONTROL – level of control required to

guarantee the permanence of the actions.3. COMPLEXITY – could less complex actions

be selected to achieve similar expected res-ults?

4. IMAGE – impact on the overall image of theuniversity campus.

In addition to these five decision areas andfour comparison areas, three further decisionareas were introduced later. These concernedthe organisation of DSG, the number of col-lection centres and the composting of greenfoods. Eight important uncertainty areas wereidentified, concerned with such matters asdanger; legality; payment for service; and thevalue of a crusher. A first progress packagewas agreed with the project steering group,including immediate actions in some decisionareas. For example:

• Restructure! Reorganise and restructurethe cleaning section of DSG. The number ofworkers was increased from 12 to 24 and asupervisor hired for the collection centres.

• Compost! Composting of green waste.• Supervise! Supervision of the collectioncentres and redesign of the collectingsystem.

347


• Contracts! Remodel building contracts toinclude demolition waste removal. Explor-atory options were agreed to reduce thelevel of some of the uncertainties that wereseen as critical to future decisions. Theseincluded:

• TypeUE – Investigate howdangerouswasteis handled now; legality of splitting respons-ibility from municipal collecting service.

• Type UR – How to enforce the remodelledcontracts; how to get owners of commercialbusinesses to pay for waste collection.

• Type UV – Would the authorities agree tocharge for collecting frombusinesses;wouldthey agree to pay for a waste crusher.

RESULTS OBTAINED

When these changes were introduced, healthand safety and operating conditions improvedat the collection centres through the closersupervision from DSG. The accumulation of

waste at the centres was cleared through theseparation of garbage from cafeterias and fac-ulties, and the relocation of waste containers.Also, the loading time of the collecting

trucks was considerably reduced, demonstrat-ing that, with daily collection and supervision,the number of containers was enough to meetdemand. An information system was intro-duced to monitor waste management insidethe campus. Skills were developed withinDSG, DIS and IU that enabled them not onlyto improve daily functioning, but also to com-mit themselves to continuing research to offernew solutions. Foundations were establishedto enable the cleaning section of DSG to takeon other areas of service identified in the actionplan. The project not only contributed to theimproved management of waste on the cam-pus site; it also enhanced the environmentaland aesthetic rehabilitation of an importantnational institution, with the capacity to enrichthe lives and capacities of future generationsof Venezuelan people.

CONCLUSIONS

The example of this project shows how a relatively low-cost application of strategic choice meth-ods can contribute not only to the solution of practical problems of local management, but alsoto the building of shared knowledge of both problems and methods amongmanagers, universitystudents and their tutors. The commitments to action that resulted were easy to perform andnot very expensive, encouraging continuing collaboration among people with different roles andskills.In Venezuela a wider impetus is now being achieved in introducing flexible methods for struc-

turing complex problems into strategic decision processes, through policy workshops and train-ing programmes involving experienced central policy-makers and managers of institutions, aswell as through engaging students in real-time projects at the start of their professional careers.It is hoped that this experience will offer encouragement to innovators in other developingcountries who are motivated to construct platforms for the further transformation of traditionalplanning cultures, at all levels of social organisation from that of local community developmentto that of national policy.

348

13.13 Differential Learning

Managing different rates of progress between participants inextensive projects

By Brendan Hickling

This contribution describes a phenomenon we experience on a daily basis in many differentways. The author has named it differential learning, and it occurs where the understandingof an individual or sub-group accelerates away from that of other participants, creating alearning gap. In working with groups in many different contexts he has learnt that it is veryimportant for anyone using the Strategic Choice Approach to be aware of these learninggaps and to be prepared to take steps to mitigate them if serious misunderstandings areto be avoided. Throughout he draws on experience from the BNFL National StakeholderDialogue, which is a good case in point because it is an extensive dialogue in which thework is progressed by working groups carrying out in-depth analysis before reporting backtheir findings to the main project group.

Brendan Hickling is an independent facilitator, mediator and process consultant working inall areas of development in several European countries including the UK. His first exposureto SCA was working with his father (Allen Hickling, co-author of this book) on participativeprojects and training. More recently he has applied aspects of SCA in helping small tomedium-sized organisations cope with growth.

LEARNING AND OWNERSHIP

Our sense of identity as individuals owesa lot to the learning we have accumulatedthroughout our lives and continue to accumu-late with each passing day. As learning is oftenexperienced with associated feelings, such asachievement, loss, relief, frustration, joy andsadness (to name but a few), we are ableto see ourselves in what we have learnt andto develop ownership over it. Shared learn-ing gives people a common identity which canbind them together and give them collectiveownership over an idea or belief. Little or noshared learning can leave people with the feel-ing that they do not understand each other or

that they have little in common. In the worstcase scenarios of strongly opposing learning,such as when two scientific studies appear togive opposite results, people can be set againsteach other.Planning is a learning experience which can

be shared and therefore owned by as manypeople who take part in the exercise. The Stra-tegic Choice Approach is at its heart a learningprocess (cf. Chapter 4). The principles of cycliclearning and working selectively (cf. pp. 21–22)combined with the extensive tool kit provided,enable individuals and groups to make con-siderable leaps forward in learning in a rel-atively short time. However there can be adanger in this: in the frequent situation where

349


an individual or sub-group is working on behalfof a larger group or constituency, the sub-group’s accelerated learning and ownership ofthe issues often leaves the others behind. Thedanger is that, when the sub-group has fin-ished its work, it will not be understood, andfurther, as they may have developed strongownership over their ideas, they are very likelyto become defensive and closed against anyform of criticism.A startling example of this happened early

on in the BNFL National Stakeholder Dialogue(autumn 1999) when a working group (rep-resenting the full spectrum of constituenciesin the process from environmental groups toindustry) had to report on its progress to themain group after six months working on wastemanagement options. They had made goodprogress and the group had started to developmutual trust and understanding among them-selves, but on the down-side the learninggap between them and the main group hadbecome significant, even although the group’swork was still at an early stage.A progress report was circulated before the

main group meeting and the working groupwas given about an hour to report on pro-gress in the meeting itself. As it turned out thissimply was not enough to bridge the learninggap and members of the main group, not reallyunderstanding what had been done, started tocriticise parts of the report. Some membersof the working group took this personally andstrongly worded exchanges ensued.What makes this a particularly interesting

example is that the most polarised discussionswere between participants who would nor-mally be considered to be on the ‘same side’ aseach other; the differential learning created aperceived shift in loyalties.With somedifficultythe facilitators managed to help the stakehold-ers develop agreement for the working groupto continue with its work but it was a closerun thing and sixmonths of intensiveworkwasnearly lost. This experience made it clear that astrategy for managing the differential learningwas needed if working groups were going to

be an effective tool in this project dealing withvery complex and contentious issues.

MANAGING DIFFERENTIALLEARNING AND SHARINGOWNERSHIP

In considering how to manage differentiallearning, I will focus on the interfaces betweenthe different groups which can exist in anextended project (cf. Figure 84 OrganisationalResponsibilities in Strategic Choice). Althoughdifferential learning does exist in shorter one-off projects, in these circumstances it can usu-ally be managed using workshop facilitationtools which are not the subject of this note.Exactly how one manages the differential

learning in any given project cannot be pre-scribed.However the over-arching principle isthat time must be invested in communicationin order to harmonise learning and broaden theownership base. It can be a considerable chal-lenge for a facilitator to persuade a sub-group,which is making good progress, that it has tostop working on the problem in order to allowsufficient time to consider how they are goingto communicate that progress to the others.I shall now review some of the avenues

of communication which can be employed tokeep differential learning to a minimum. It isnot an either/or list; they should be taken as ele-ments of amulti-level communication strategy.

Newsletters (information giving)

Newsletters are used the world over to keeppeople informed. Their strength is that they canbe used to reach large numbers of people whomay be geographically dispersed. Their verygreat weakness however is that the commu-nication is only one-way. There is no way ofknowing what the readers think or even if theyhave read the newsletter at all. Additionally,as newsletters are by nature very public, theyare often inappropriate in contentious issueswhere it can be very hard to get agreement onwhat should go in them. In the BNFL NationalStakeholder Dialogue it was more than 2 yearsbefore a newsletter became feasible.

350


Newsletters are at their most effectivewhenthey are short (easily read) and frequent(providing constant small reminders of whatis going on). They can be made slightly moreinteractive by inviting feedback or even includ-ing questionnaires, but there are pitfalls, andthe weaknesses mentioned above still apply.

Networking (information giving andgathering, some consultation)

Networking is about taking advantage of thecontacts a working group naturally has. Evena small working group has a huge resourceof potential feedback, which can be accessedby its members talking to their colleaguesand constituents about what they are doingand asking them to do the same with peoplethey meet. Networking is an interactive andresponsive way of disseminating informationand getting feedback; and it has the addedadvantage of keeping members of the group intouch with the people and organisations theyare suppose to represent. The weakness ofnetworking is that the group’s network mightnot include some key figures and groups; if thisis the case then you may need to consider tar-geted networking.

Targeted networking (consultation)

It is very probable that key opinion leaders andother people or groups of influence are goingto be outside of your natural networks and pos-sibly outside the constituencies engaged in theproject. Targeted networking is about identi-fying these people and seeking detailed feed-back from them. As with networking this is notabout bringing more people into a highly inter-active working group; that would defeat theobject of having a small group to do the intenseanalytical work. This is about presenting workin progress and the emerging conclusions, at atime when the overall direction and emphasiscan still be influenced.This kind of communication is the richest and

can give the best indication as to whether on-going work will be understood and accepted.The BNFL National Stakeholder Dialogue used

this approach to good effect when key opinion-formers were not directly participating in theprocess.

Presentation of ideas

When ideas and processes are complex, onemust consider carefully how they are presen-ted. There are many ways of presentinginformation but experience has shown that thevery graphic ways of generating and choosingoptions in the SCA toolkit also work well whenit comes to presenting them. The progresspackage has proved very effective at commu-nicating very complex scenarios and optionsin the BNFL National Stakeholder Dialogue,especially in the transparent management ofuncertainty and exposure of assumptions. (SeeRichard Harris’s contribution for an example,albeit very cut down, of a scenario beingpresented through the commitment packageframework; and also Rob Angell’s contributionfor an example, from another project, in whichthe structure of a simple AIDA analysis wasused as the basis for presentation.)

Induction meetings

In extensive projectswhere participation variesover time, considerable disruption is causedby people joining and rejoining the ongoingprocess afresh. In the BNFL National Stake-holder Dialogue the frustration this caused forthe establishedmembership became a seriousproblem. In the end a ground rule was agreedwhich stated that newcomerswere not permit-ted to join the dialogue unless they underwentan induction process involving both the facilit-ation team and participants from all of the keystakeholder groups.

A COMMUNICATION STRATEGY

I have given here some examples of the aven-ues of communication that can be used andI expect there are many more. However incases where the work is particularly sensitivea communication strategy is needed to work

351


at multiple levels of participation at the sametime. For example: passive information giv-ing, such as newsletters, in parallel with con-sultative feedback loops to constituencies andopinion leaders, can complement the highly

interactive work of a working group. What iscertain is that to do nothing to address thepotential problems can result in good workbeing lost simply because it is not under-stood properly.


Whenever using the Strategic Choice Approach, but particularly in an extended project whereintensive analytical work is carried out by individuals or small groups in isolation from theirconstituencies, the following is generally true:

Small group learning without a communication strategy leads to differential learning

Small group learning with a communication strategy leads to shared learning

The importance of this lies in the different effects of the two:

• The former aggravates misunderstandings between participants and one can expect anongoing process characterised by disharmony and slow progress;

• The latter provides participants with a sense of ownership of the work which is the essentialbasis for a broad-based commitment to an agreed way forward.

352

13.14 Building Commitment in a Rural Community

Use of commitment packages in empowerment throughinformation technology

By Jackie Phahlamohlaka

In community development, it is no simple matter to facilitate a planning process in sucha way that the facilitators and the community see it as a learning opportunity. This isespecially so if the process includes introducing information and communication techno-logies; for their adoption and use in such contexts is not yet well-understood. We foundthat the concept of the commitment package enhances this process of ‘learning together’between facilitators and community members. It fits well within an action research frame-work, enabling both the researcher (facilitator) and the participants (a community) to learntogether as they engage in uplifting the lives of the community. The experience reportedhere involves learning how to improve the lives of a rural community through a local areanetwork (LAN) facility and access to the Internet.

Jackie Phahlamohlaka completed an MSc in Computational and Applied Mathematics atDalhousie University, Canada in 1991 and a PhD in Informatics from the University ofPretoria in 2003. Before joining the Department of Informatics at the University of Pretoriain 1996, where he is currently a senior lecturer, he served in the strategic planning andstatistical services unit in the office of the Premier, Mpumalanga Province. He is a foundermember of the Siyabuswa Educational Improvement and Development Trust (SEIDET) andchair of its executive committee. Within the SEIDET context, an opportunity arose in 1998to apply strategic choice methods alongside John Friend to facilitate a planning workshop asa contribution to the Trust’s own evolving planning process.

INTRODUCTION

The aim here is to share with the reader twokinds of situation in which we found helpfuleither the Strategic Choice Approach, or con-cepts borrowed from its repertoire. The firstoccasion is an actual application of SCA in facil-itating planning for a rural community educa-tion project in South Africa, while the secondis an ongoing process of enquiry and furtherdevelopment of the same project, whose

design is informed by and accompanied thisplanning process.In the language of this book, planning is

viewed as a dynamic process of strategicchoice, under conditions of uncertainty, res-ulting in multiple decision outputs. The coreconcepts of progress package, action scheme,exploratory option and uncertainty area helpin organising the decision outputs achievedthrough the choosing mode. The last two coreconcepts constitute what we regard as an

353


ongoing process of enquiry. Our experiencein using SCA to facilitate a one-day planningworkshop for a rural community educationproject in South Africa partly informed us in thedesign of an ICT action research study in whichwe have been actively engaged since theyear 2001.The community education project referred

to here is the Siyabuswa Educational Improve-ment and Development Trust (SEIDET).SEIDET is a community initiated and com-munity based educational project that startedin 1991/92, based on the rural town of Siy-abuswa in Mpumalanga Province, SouthAfrica.It started as a supplementary tuition project

for high school learners in selected learningareas such as English, Science, Commerceand Mathematics. Over the years, the pro-ject has grown from strength to strength.Presently the SEIDET Community Educa-tion Centre has two satellites hosted byother institutions at KwaMhlanga (TechnikonPretoria, KwaMhlanga Campus) and Vaalbank(Hlalakahle High School). The main centre inSiyabuswa is a multi-purpose facility providinga variety of educational services and develop-mental programmes to the local community inassociation with several other institutions.

AN INITIAL STRATEGY WORKSHOP

On the basis of an annual cycle, SEIDEThas developed a process in which the vari-ous committees and divisions produce theirown plans, and then submit these to theExecutive Committee for consolidation into asingle organisational plan. Through the writ-ings of Rosenhead (1989) and others, the Exec-utive Committee had become aware of thenew generation of participatory problem struc-turing methods developed by managementscientists.The opportunity to build on this awareness

arose when the author was invited by JohnFriend to visit his University campus in the cityof Lincoln, England, in August 1997. A yearlater, in 1998, John Friend paid an exploratory

visit to Zambia and South Africa on behalf of hisCentre at Lincoln. SEIDET took the opportunityto invite him, on the final day of his visit, towork alongside the author in facilitating a plan-ning workshop as a contribution to the Trust’sown evolving planning process.A combination of SCA and the nominal group

technique was used. Figure 13.5 summariseswhat the workshop managed to achieve. Thepositive after effects of the workshop includesan ongoing ICT action research study (Depart-ment of Informatics, University of Pretoria). Inthe next section, I shall focus on the last men-tioned after effect, showing how the conceptof the commitment package, from the stra-tegic choice repertoire, helped us in designingan ICT action research study at SEIDET.

THE LAN AND INTERNET USEPROJECT

This project is using an action researchapproach to investigate the community devel-opment opportunities brought about by theestablishment of a local area network (LAN)and access to the Internet at the communityeducation centre. The purpose of the researchis to make a contribution to the further devel-opment of the SEIDET project by assisting theComputer Committee of SEIDET to identifythe opportunities and possible challenges thatmay arise from the expanded ICT facilities.In a community-interactive way, the researchwill then assist the SEIDET community inanswering sets of questions while providingthe researchers with new perspectives oncommunity development and training.In the short term, the purpose is to empower

the participants and tutors through skillsand knowledge acquisition and to assist instructuring the integration of the LAN andInternet technologies with the activities ofSEIDET in an efficient way (Phahlamohlaka andLotriet, 2002).In the long term, the purpose is to assist the

participants in managing and sustaining LANand Internet-related activities within the con-text of a rural community education centre.

354


What we did

What we did

What we could have

done with more time

What we could have

done with more time

What we did

SHAPING

DESIGNING COMPARING

CHOOSING

What we could have

done with more time

What we could have

done with more time

What we did

Some 20 decision areas were listed by

brainstorming with the executive

committee, while the three tutor groups

worked silently, adapting their views on

issues to those circulated by others

within each group. The results of both

exercises were merged on flipcharts.

Some links were entered on a graph,

then a focus of two key decision areas

was selected.

We made some progress in listing

uncertainty areas before attempting to

build a progress package. Now a

participant came forward as facilitator,

suggesting that the proposed options for

the 1999 strategy could instead all be

seen as things could be done in parallel.

We discussed the main options to be

compared within each of the two

selected decision areas. This led us to

think further about what the decision

areas themselves meant, and to

introduce some changes

We listed six main criteria relating to

the objectives of the Trust, as presented

in the last annual report. We started

drawing up a balance sheet of

alternative strategies in the light of

these criteria, but we made little

progress.

We could have spent

more time discussing the

options in other decision

areas, and exploring

how far they were

compatible.

We could have engaged the

group in further debate of

impacts in relation to

different criteria, perhaps

using stickers to register

personal judgements

We could have introduced

more of the decision areas

we had listed into the

package, and also reviewed

more uncertainty areas

We could have explored

further the relative

urgency and importance

of the decision areas

before choosing a focus

for the day

Figure 13.5 Overview of Progress in the SEIDET Workshop (Source: Phahlamohlaka and Friend, 2004).

The study seeks to respond to the followingset of primary research questions:

• From the perspective of the computer com-mittee (management), what opportunities dothese added ICT facilities present?

• What does the computer committee see asnew challenges in managing the laboratorybrought about by the LAN and access to theInternet?

• What could these developmentsmean to thelocal community and to the functioning ofSEIDET as a whole?

ACTION RESEARCH WORKSHOPS

In seeking responses to the above ques-tions, eight action research workshops have to

date been conducted (March 2001–November2003). This research – which is continuing –has been designed in accordance with actionresearch principles. It follows a cyclic pro-cess of interaction, reflection, action and inter-pretations between the researchers and theparticipants. The participants are seven com-puter committee members of SEIDET. Theresearchers are two faculty members of theDepartment of Informatics.After each session, the participants take the

full set of issues raised, the research questionsand their new technical skill to the rest of theSEIDET tutors. In the process, the tutors alsorespond to the issues and questions raised pre-viously while generating new ones.After a session with the tutors, the parti-

cipants bring back a new set of responses,

355


issues and questions for negotiation and pos-sible action by the researchers and the processis repeated. The sets of agreed upon actionplans are called ‘commitment packages’, using

the SCA term. A commitment package is typ-ically recorded in a tabular format, alongsidethe corresponding questions and actions of theprevious workshop/session.

CONCLUSIONS

The commitment packages produced answers while at the same time generating more ques-tions. The following aspects of the research were experienced as enriching by the researchers:

• The engagement skills on ICT-related issues that were demonstrated by the participants andthe questions that were thrown back to the researchers. These enabled the researchers toreflect on the project.

• The negotiation processes that were entered into each time in order to determine the nextset of research questions.

• The team spirit and the level of trust that exist and are demonstrated from both sides.

Valuable lessons were also learned in the process. It became evident to us that while thecomputer committee has become more aware of the potential that they have to change theircommunity, this increased awareness is also accompanied by an awareness of increasedresponsibility and the critical requirements necessary for further community developmentthrough ICT-enabled intervention.

REFERENCES

PHAHLAMOHLAKA LJ and LOTRIET HH (2002) An investigation into community development imperatives broughtabout by the recent establishment of a local area network and access to the Internet at a rural South African communityeducation centre. Proceedings of the seventh International Working Conference of the International Federation ofInformation Processing (IFIP) WG 9.4. Bangalore, India, 29–31 May 2002

PHAHLAMOHLAKA LJ and FRIEND J (2004) Community Planning for Rural Education in South Africa. European Journalof Operational Research, 152, 684–95. Also published online 2003 and is available from Science Direct

356

13.15 Commitment is the Key

Building inter-agency agreement over the future of an historicestate

By Leny Bregman1

When she was appointed Project Leader for the development of the Haarzuilens estate(Figure 13.6), having the Castle (in Dutch Kasteel) De Haar at its centre, the author wasfaced with many disparate views about its appropriate future use. Plans had already beendeveloped, but no agreement could be reached. She chose to seek consensus over a visionfor the estate as it might be 15 years hence, in a two-day multi-agency workshop using theinteractive style of the Strategic Choice Approach.

After finishing her study in forestry (1982) Leny Bregman worked as a manager for theAmsterdam Woodland Park Forest. The Park is popular: each year 4.5 million visits aremade by 800,000 people of Amsterdam and its outskirts. For the first 7 years she managedthe woodland and the visitor’s centre. Later on her task was to formulate policies for thePark. She also was responsible for external communication. In 1993 she started working forNatuurmonumenten, the largest non-government organisation for nature conservation in TheNetherlands. Her first job was recreation policy. Since 2001 she has been the project leaderfor the restoration of the Haarzuilens estate.

THE PROJECT

‘To each user, be they child or adult, the estatehas become their place of manymemories – oftheir first love; a day of roaming about, a chil-drens’ picnic party; or a romantic tour in a horsedrawn carriage. They have become aware ofthe rich history of the region by experiencingthe many different aspects of the estate; ofnature which has settled there; and of thetransformation of the mainly agricultural land-scape into an idyllic, green oasis in an other-wise urban area.’(This is the opening statement of the 2015

vision for the historic Haarzuilens estate, by

1 translated from Dutch to English by Martha Vahl,Lincoln

Bregman andNatuurmonumenten, 10October2001.)In 2000 Natuurmonumenten acquired the

Haarzuilens Estate and also became a 50 percentpartner in theFoundation thatmanagesthecastle, gardens and grounds (Stichting KasteelDe Haar). The estate comprises 372 ha ofagricultural land with 9 farms and 70 build-ings in the village of Haarzuilens. Baron VanZuylen Nijenvelt de Haar previously owned theestate and the castle privately. For a numberof years the area had been under some pres-sure due to a newhousing development (called‘Leidsche Rijn’) with planning permission for80 000 new houses to the North West ofUtrecht. To provide a recreational zone forthe new residents, the central government

357


Figure 13.6 Part of the Haarzuilens Estate with the

Kasteel de Haar at its Centre, Utrecht, The Netherlands.

had developed a structural development plan(Raamplan) that partly covered the grounds ofthe Haarzuilens estate. As the Baron and theProvince of Utrecht (equivalent to a CountyCouncil in the UK) could not agree over the saleof grounds of the estate, this plan failed to beimplemented.Although Natuurmonumenten was critical of

the plan’s spatial design, it did agree with theprinciples on which its was based. Usually itacquires land fundedbycentral and local author-ities.However, not all the groundswereeligiblefor such funding. This meant that Natuurmonu-menten had to invest a considerable amountof its own capital, even though it was helpedby a large donation from ‘De Postcodeloterij’(which earns its money through a lottery bene-fiting charities). The acquisition was justifiedby the ‘nature-near-the-town policy’, whichNatuurmonumenten had adopted in 1999. Dueto unfamiliarity with this policy, and the largesum involved, the acquisition initially could notcount on much enthusiasm inside the organ-isation of Natuurmonumenten.

BEFORE THE WORKSHOP

As the senior officer accountable for the policydocument in which the new policy for recre-ation and for nature-near-the-town had been

laid down, it was obvious that I would experi-ence the acquisition of the Haarzuilens estateas a great idea. This went deeper, however,the moment I heard about the acquisition isengraved in my memory as a ‘magic moment’.During this ‘magic moment’ all the possibilit-ies that the implementation of the new policywould entail leapt to my mind. I saw imagesof people enjoying a natural landscape. Thiswas a very intense moment. To quote OliverSacks: ‘Inspiration emerges after intense fas-cination and concentration, and after a periodof distancing oneself. Then suddenly there isthe moment of enlightenment, and it becomesclear how things are connected’. (In Psycholo-gie Magazine, March 2002.)This was what happened to me as well. The

only drawback was that I found myself in asituation where very few in the organisationshared my inspiration. On the contrary, peoplewere unsure about the new policy and werecriticising the acquisition of the Haarzuilensestate.In the capacity of project leader I took the

initiative to organise an SCA workshop. Theaim was to build our team, but also to alignthe views of those directly involved, bothfrom Natuurmonumenten and the Foundation(Stichting Kasteel De Haar).While preparing the workshop the Board

of Natuurmonumenten introduced additionalaims. These included developing a creative vis-ion, co-ordinating the responsibilities of theparties involved and the development of afirst action plan. We decided to have a two-day workshop, which would include working insub-groups, but also a half-day fact-finding and‘acclimatisation’ cycle tour across the estate.

A TWO-DAY MULTI-AGENCYWORKSHOP

The workshop started by identifying and nam-ing all the issues. What topics were import-ant? Who were the stakeholders? Next theparticipants started their cycle tour across theestate. A list of questions was used to explorethe surroundings and start the development

358


of a vision. The direct experience of the land-scape as well as the opportunity to exchangeexperiences helped to achieve an open andconstructive atmosphere for discussion.The workshop participants proved able – in

a surprisingly short time – to develop four cre-ative visions on the estate, guided by fourthemes. The visions were:

1. Back to the Middle Ages;2. Country seat of the twenty-first century;3. The Utrecht Forest; and4. Fairytale-Land De Haar.

The following day we were able to get afurther handle on the project by mapping theissues already identified as decision areas.In this way it proved possible to increasinglyreplace most peoples’ initial scepticism bytrust – in a process of specifying and dis-cussing uncertainties such as: ‘Do we haveenough money?’; ‘Do we have sufficient vis-ion?’; and ‘Do we have sufficient experiencewith large projects?’ Elaborating these uncer-tainties in terms of an SCA progress package:decisions to be made, explorations, deferreddecisions, contingency plan (cf. Figure 26),helped to come to terms with the project’scomplex topic. We concluded that we weredealing with not just one but many interrelatedprojects, all part of a long-term process. Therewere issues that required immediate action,but also issues that would not be due for sometime.Making this discovery together proved animportant breaking point. The insight that wewere dealing with many interrelated projectswith different time scales helped us realisethat we wanted to do too much at the sametime. We had to reduce our ambition anddivide the work into manageable sub-parts.

AFTER THE WORKSHOP

After the workshop, the learning, which wasgained through the creation of the four cre-ative visions, was elaborated to constitute atarget vision. This powerful, shared new vis-ion enabled us to inspire and convince other

parties. It helpedNatuurmonumenten tomobil-ise a strong lobby so the county authority (theProvince of Utrecht) decided to allow for thedevelopment of an alternative to the existingstructural development plan.Every phase in a planning process requires

internal commitment, but also external inter-action with a variety of stakeholders. Thisinteraction has taken, and still takes, muchtime and energy. In 2002 a first draft designfor the alternative structural development planwas produced. It came into being in 2003,in collaboration with the Utrecht Municipal-ity (the City of Utrecht) and the Departmentof Rural Areas (a regional organisation forplan implementation), under the responsibilityof the Land Development Committee (Land-inrichtingscommissie). The county authority(Gedeputeerde Staten of Utrecht) is scheduledto discuss the design before the summer of2004. If it is approved, the implementation isto start in 2005.

EVALUATION OF ‘WORKINGTOGETHER’

‘Natuurmonumenten verymuchwants toworkwith others to realise this big project. In the firstplace this concerns the parties involved in theimplementation of the structural developmentplan Utrecht-West, but also the creative inputfrom farmers, citizens, other folk and entre-preneurs. Naturally, designers and landscapearchitects are also involved. The varying inter-play among Van Zuylen the owner, Cuypersthe architect and Copijn the landscape archi-tect at the end of the nineteenth century formsthe basis and source of inspiration for today’swork.’ (This is the concluding statement ofthe vision of Natuurmonumenten for the Haar-zuilens estate by H. Bregman, 10th October2001.)Thedevelopmentof theHaarzuilensestateas

a ‘nature-near-the-town’ facility is a long-termprocess (expected to end in 2013), in whichmany interrelated projects need to be distin-guished and developed in a complex environ-ment of stakeholders from the public aswell as

359


the private sector. Since this process will takemany years, it will be important periodically toevaluate and adjust the coordination betweenthe projects, while keeping in mind one’s owngoals. Communication, including lobbying withthe various stakeholders, has proved to be animportant key towards success and is likely toremain so in the near and distant future.

The current draft design reflects the socialand historical aspects of the estate that Natuur-monumenten wishes to retain much betterthan the original plan. This is in addition to thekind of development of the natural environ-ment it aims to achieve, despite the intens-ive level of recreation that will develop in thearea.


From this experience I have learned two important lessons, which I would like to pass on toothers who might be faced with a similar situation.

• The first is that you do not have to be vastly experienced in SCA to be able to use it. Guidancefrom experts may indeed be very helpful, maybe even necessary from time to time, but itsbasis in common sense is its real strength. Combined with the idea that there is no one rightway of doing it – there are many good ways – much can be achieved relatively easily.

• The second is that even when SCA plays only a relatively small part in a project, it can be veryhelpful. In this case the Strategic Choice Approach laid the ground for the team-building thatallowed for the development of a powerful target vision. With this vision a strong lobby couldbe launched to improve on the original structural development plan.

360

14 The developmental challenge

In this short final chapter, all we wish to do is to draw together some threads from thelessons offered by our contributors to Chapter 13, and from other lessons that the two ofus have drawn from our own recent experiences, so as to offer some pointers to furthershared progress in the decades ahead. This aim is much the same as that of the closingchapter of our second edition, also called The Developmental Challenge; and indeed alsothe final Horizons chapter of our first edition. Yet, in this new twenty-first century edition,a combination of new challenges and new learning points makes it important for us torevisit our earlier speculations about pathways for sustainable future progress.

MAPPING THE CHALLENGESAHEAD

Our final figure – Figure 103 – is a slightlymodified version of one that appeared in theclosing chapter of our second edition. In it weoffer a structured checklist – or what somewould term a conceptual map – in which vari-ous shared agendas for future developmentare shown loosely clustered within eight over-lapping circles, representing the eight broadspheres of research, education, methodo-

logy, facilitation, application, communication,technology and sponsorship.The value of any such picture is merely to

focus discussion among people who approachthese challenges of development from differ-ent directions.Wehave highlightedwith bolderoutlines the cluster of four spheres that wehave positioned in the centre of our map –those of application, facilitation, methodologyand communication – because they have longbeen central to our own primary concerns withaction research. Yet several of the contribut-ors to Chapter 13 write from deeper experi-ence in the four important spheres shown inmore peripheral positions in Figure 103, con-cerned with technology, sponsorship, educa-tion and research of a more critical academickind. Here the accent is more on investment

in longer-term development of foundationsthan on impact on pressing decisions. How,we must ask, can the fruits of such invest-ments underpin and complement the learningof those primarily engaged in the managementof action research projects?The thirty-five more specific agendas of

development that we show at various posi-tions within the map are again intended to pro-voke discussion rather than to assert any firmbeliefs about their relative significance or posi-tions on themap.What matters is not so muchthe validity of any such initial picture, as thekind of debate that it can provoke about pri-orities, linkages, relationships and omissions,as a step towards shared learning by peoplewith different backgrounds of experience andbelief.One brief yet significant opportunity to use

the picture presented in Figure 103 as astarting point for this kind of debate arosetowards the close of an international work-ing conference held in 1997 at the LincolnUniversity campus. Here, John Friend hadinvited twenty consultants, researchers andprogramme managers from twelve differentcountries in Europe, Asia, Africa and theAmericas to gather for a week, along withstaff and postgraduate students of the LincolnSchool of Management, to explore ways of

361

FIGURE

103Mapping the Developmental Challenge: A Semi-Structured Checklist

formation of

new schools

research education

methodology

facilitation

communication

technology sponsorship

doctoral

research

crystallization of

appropriate theory

critical review of

available methods role in public

policy courses

innovative learning

strategies

appeal to

students

inter-generational

succession

role in development

planning courses

role in business

school courses

role in management

science courses

advancement in

methods

synergy between

methods

designs for action

research projects

training in

facilitation

access to range

of case material

consultancy

repertoires

innovation in

group pocesses r

rich accounts

of applications

evaluating quality

of process

cross-cultural

project work

exchanges at

working conferences

receptiveness of

potential clients

evaluating success in

addressing issues

exploiting the

internet

advances in information

technology

champions in

policy roles

translations of

books and software

international sponsorship

for development work

accessible decision

support software

national and regional

support systems

changing policy

agendas

experiments in

distributed meetings

public awareness

of potential

insights into

decision-making

application

The Developmental Challenge

working together in extending the use of thestrategic choice approach and related parti-cipatory methods in the important cause ofbuilding appropriate planning capacities in thedeveloping world.During a closing session, an enlargement

of the map reproduced in Figure 103 wasprojected on a wall to guide discussion andagreement on priorities for further collabora-tion. The most immediate priorities that werethen agreed included steps to support the sixparticipants from five African countries in co-ordinating the contributions they might maketo that continent’s development, supplemen-ted by steps to publish readable accounts ofsuccessful applications in developing coun-tries.1 As a longer-term ambition, a sustainedprogramme of action-oriented research wasalso proposed over a 5- to 10-year period, to linkuniversities and planning institutions acrossthe world in building a broader momentumof practice-based development of theory andmethods on a global scale.After the participants dispersed, it was not

too surprising to those present – most of themwell versed in the challenges of planning underpressure – that the initiatives resulting fromthe conference turned out to be influenced asmuch by unexpected events – sometimes pos-itively, sometimes negatively – as by the morespecific action plan that was agreed amongthose present. Some promising sponsorshipopeningswere pursued, withmodest success;closer working links were forged among thehome institutions of several of the participants;strategic choice workshops and demonstra-tions were arranged in particular developingcountries as opportunities arose; and furtherdevelopments in theory and methods werepresented in papers presented at academicconferences or published in journals (see, forexample, Finlayson, 2000; Friend, 2001; Phah-lamohlaka and Friend, 2004). It is a reflec-tion of the broader products of the conferencethat several of the contributions to Chapter 13

1 It is planned to include some of them in the companionwebsite to this book.

reflect the links formed through the Lincolnmeeting.

DRAWING LESSONS TOGETHER:APPLICATION AND FACILITATION

Because so many of the contributors toChapter 13 report lessons from their ownexperiences in facilitating workshops, in verydiverse fields of application, it seems relevantto begin drawing together threads by focus-ing on the central horizontal axis of Figure 103,where the two spheres of facilitation andapplication come together.One striking learning point from several

of the applications described in Chapter 13is that, through various extensions in facilit-ation and process management methods ofthe kind described in Chapter 11, the Stra-tegic Choice Approach has shown it can beadapted to support high-level policy develop-

ment projects that bring together stakehold-ers with widely divergent accountabilities inaddressing high-level issues where their posi-tions have long been opposed. This is despitethe way in which SCA challenges conventionalplanning norms, by advocating a selectiverather than a comprehensive view of problemscope. Yet recent successes in adapting theStrategic Choice Approach to broader policyagendas by Allen Hickling and his associates,through programmes successively sponsoredby the Netherlands government, by interna-tional agencies and by the UK EnvironmentCouncil, have been demonstrating practicalways in which the approach can be adapted tosupport collaborative work on a broader policycanvas, demonstrating practical ways in whichthe acknowledged limitations of conventionaltop-down planning can be overcome.Meanwhile, other contributors to Chapter 13

describe shorter engagements that have beennegotiated in a more opportunistic way aroundmore specific local problems. There will con-tinue to be wide scope for more modestaction research engagements of this nature insuch fields as participatory community devel-opment, small enterprise management and

363


partnership for local action. For here, as muchas in broader policy arenas, those in facilit-ation roles can expect to face challenges ofbuilding shared commitment among repres-entatives of stakeholder groups with oftendeeply conflicting accountabilities. Together,the contributions to Chapter 13 and the illus-trations from practice at the end of Chapters 5,6, 7, 8 and 12 give some flavour of theway in which the variety of applications hascontinued to increase. We continue to beaware too of applications by other people inother parts of the world. A few of thesehave been written about in published journ-als, sometimes with quite specialised reader-ships. In the list of references will be foundEnglish language accounts of applications inIndonesia (van Steenbergen, 1990), in Belize(White, 1994), in Thailand (Kammeier, 1998)and in the favelas of Rio de Janeiro in Brazil(Bredariol and Magrini, 2003). As further newapplications are published, it will be pos-sible to keep the list up to date throughthe Planning Under Pressure companionwebsite.

DRAWING LESSONS TOGETHER:METHODOLOGY ANDCOMMUNICATION

Turning to the vertical axis of the centralcluster of spheres in Figure 103, concernedwith methodology and communication, thereare other significant learning points to be dis-cussed. In terms of methodology, one signi-ficant learning point concerns the importanceof learning to reach out towards understand-ing of, and accommodation with, the methodsto support decision-making that are already inuse by other participants in a planning process.These methods may range from the relativelystructured and formal approaches espoused byprofessionals in many corporate organisationsand business consultancies, to the more parti-cipatory and informal methods that are increas-ingly advocated in relatively open fields suchas that of local community development.

This is where the challenges of reachingout – of outreach – across cultural, linguisticand other divides can become central tothe overall communication challenge. It hasbecomeevident that not only do those involvedin high-level policy development work needtools to enable them to reach out towardsgreater understanding of the complexitiesand uncertainties of decision-making faced bythose addressing specific problems at a morelocal level; those local decision-makers them-selves also need tools to help them in reachingout towards understanding of the complexit-ies – and the uncertainties – that arise in theoften intricate and obscure politics of policydevelopment at the more generic level (Friend,2001).

FOUNDATIONS FOR FURTHERPROGRESS: SPONSORSHIPAND TECHNOLOGY

Although we believe that the action researchfocus reflected in the cluster of four spheresin the centre of Figure 103 should continue tobe viewed as a leading driver of future pro-gress, we also recognise that developments inthe two supporting spheres of sponsorship andtechnology shown at the base of Figure 103can both be viewed as important foundationsfor sustaining long-term progress.Investment in sustained sponsorship of pro-

jects and programmes to support innovationin decision-making will continue to be import-ant if the lessons from future action researchengagements are to be fully absorbed andwidely shared. It is only realistic to assumethat this kind of sponsorship will only benegotiable if future innovators can exercisepersuasiveness in attracting support fromchampions in policy roles, and flexibility inaddressing the imperatives of changing policyagendas.In the story so far of the Strategic Choice

Approach, periods of sustained sponsorshipover several years have provided foundationsfor much of the continuing innovation repor-ted both in Chapter 13 and in earlier chapters

364


of this edition. Among the most significantsources of investment have been charitablefoundations and government research councilsin the earlier years; then departments of theBritish and Netherlands governments, then dir-ectorates of international agencies such as theEuropean Commission and the Organisationfor Economic Co-operation and Development.More recently an important role has beenplayed by the UK Environment Council, whichis an independent agency drawing wide spon-sorship from varied governmental, commercialand voluntary sources to pioneer mediation,training and related services through which tobring together potentially conflicting interests,whether to agree national policy guidelines orto negotiate accommodations on a more localscale.2

In reviewing the advancement of Techno-

logy to support future progress, the most strik-ing change to be acknowledged is the way inwhich the spectacular recent advances in elec-tronic communication technologies have ledto a much higher intensity of global commu-nications through the internet, through e-mailand through mobile communication networks.The rapid diffusion of access to these tech-nologies – from backroom specialists to famil-ies and communities, with young people in thevanguard – could scarcely have been imaginedeven a generation ago.One consequence of the ever-widening

spread of access to communication andinformation technologies is that computer sup-port to strategic choice has so far foundits most promising applications in relativelylocal applications, rather than in more extens-ive policy projects. Here is a rich field forexperiment in the development of what wasdescribed in Chapter 4 as ‘open technology’ forstrategic choice, recognising that there haverecently been several other significant innova-tions in open technology to support dialogueamong decision-makers.3

2 www.the-environment-council.org.uk.3 For example, see www.dialoguebydesign.net.

QUESTIONS FOR DEEPERRESEARCH: CONVERGENCE ORDIVERGENCE IN APPROACHES?

Moving to the sphere of research, which weplaced in the top-left corner of Figure 103,deeper questions arise about sustaining sci-entific progress through time. As the approachto which we have given the name strategic

choice becomes subjected to a variety of inter-pretations by others, we can ask whether weshould expect it to dissolve into some evolvingmainstream of knowledge about ways of sup-porting difficult decisions. Or is there some dis-tinctive core of ideas – or indeed of emergenttheory – that might be recognised and debatedmorewidely, as a contribution both to decision-making practice and to the growth of humanknowledge in a broader sense?Some pointers towards answers to these

questions about distinctiveness can be foundby referring back to Chapter 1, where thefoundations of the Strategic Choice Approachwere discussed. The approach was presentedas challenging the four familiar managementand planning norms of linearity, objectivity, cer-tainty and comprehensiveness. Instead, a casewas made for learning to work with four altern-ative precepts: those of cyclicity, subjectivity,uncertainty and selectivity.The development of useful theories in man-

agement and planning entails striving to drawout common features and differences amongspecific approaches and methodologies at amore generic level, so that they can be sub-jected to intensive critical debate. In the caseof the Strategic Choice Approach, we recog-nise significant contrasts with some of theother approaches that have emerged in par-allel with ours to support difficult decisions –among them those approaches that reflect themore comprehensive perspective that was toemerge from the work of systems scientists inthe second half of the twentieth century (Floodand Jackson, 1991).With its less synoptic emphasis on the mak-

ing of decisions through time in the faceof multiple sources of uncertainty, we can

365


place the Strategic Choice Approach as closerto the recognised world of decision sci-

ence or decision theory than to that of sys-tems science. For the origins of SCA lie inopportunities to observe closely, and struggleto understand more clearly, the challengesfaced by people attempting to address toughdecisions under uncertainty, while workingtogether under insistent yet continually shift-ing action pressures. Then the more fullydeveloped repertoire of SCA, as discussedin Chapters 5 to 12, has taken shape throughwide experiences in active facilitation roles,alongside a growing range of associates andpartners in several countries.

THE DEVELOPMENT OF DECISIONS:AN EMERGENT DOMAINOF THEORY?

As a means of drawing out differencesbetween the approach thatwe have developedand others based on more synoptic perspect-ives, the phrase developmental decision mak-

ing was suggested by Friend (1995). Hepresented the possibility that an awarenessof this as an important field of human activ-ity might lead to the recognition of an emer-gent domain of developmental decision theory

or developmental decision science: a domainwithin which others in addition to ourselvesmight be seen as having played a pioneeringrole.These phrases were introduced in a specu-

lative vein in the final chapter of our secondedition. Since then there has been somedebate of alternative phrases that might beless open to misinterpretation – for example atheory of decision development or of decisiondynamics. The important thing is that any suchlabel should convey the richness of the con-tinually shifting relationships among decisionproblems, and among decision-makers, thatmarks out the intended field of study, with itsimplications for planning practice.Certainly, this kind of language seems

to us more apt than any alternative lan-guage emphasising systemic relationships in

addressing the social and political aswell as themore technical dynamics of the generic field ofhuman activity with which we have been con-cerned over the last 40 years. We can certainlycontend that it was the observable practices

of developmental decision-making – or ofdecision development – that have remainedour focus ever since our action research began.We can also argue that the methods and pro-cess guidelines that we have introduced sincethen could be described as forming a prac-tical approach to, or methodology for, devel-opmental decision support. So could we claima place among the pioneers of an emergentbody of distinctive theory in this hitherto little-explored field?The purpose of introducing language of this

kind is not to add to the range of alterna-tive labels that have already been adopted byothers in their interpretations of the particularset of methods that we have been describ-ing as the Strategic Choice Approach. Rather,it is to facilitate constructive debate amongproponents, students and users of all the dif-

ferent approaches that have been starting tocolonise parts of the same emergent field –varying, for example, in their emphasis on thepolitics of inter-agency negotiation or conflictmanagement. How far, it can be asked, mightany of these offer perspectives or insights thatcan add to what others have done, therebyenriching the wider domain? How far toocan they explore in more depth the funda-mental differences among different bodies oftheory?4

We have found that the case for a develop-mental perspective is especially valuable whenworking in inter-organisational and other relat-ively open arenas of decision-making. For in

4 For example, it can be hypothesised that one criticaldifference between the decision orientation and thesystems orientation lies in the different concepts ofexistence of problems that is highlighted in Spanish andsome other languages by the use of two different verbs‘to be’. The Spanish verb ser applies to the essence ofthe subject, whereas the verb estar applies to a moretransient state of being in some particular situation, andis therefore more consistent with a decision orientationthan a systems orientation.

366


such settings it becomes more realistic totreat objectives, purposes and values not somuch as foundations to be explored at theoutset of a process, but rather as matters tobe addressed as and when they emerge, andas relevant to the building of common groundand the sustaining of progress towards agreeddecisions. These precepts remain central toour practice and our philosophy. Yet, in theearly twenty-first century, we continue to findourselves engaging with people whose man-agement responsibilities or personal beliefsdraw them towards more tightly structuredprecepts of management and planning, withroots in the more conventional norms of linear-ity, objectivity, certainty and comprehensive-ness – and with little or no recognition of thecase for learning toworkwithmore challengingyet rewarding alternatives.

CONNECTING LOCAL ANDBROADER PERSPECTIVES ONSUSTAINABLE DEVELOPMENT

One policy field of rising global significancewhere there is a growing awareness of thecase for a developmental approach to decision-making in that of building local capacities forsustainable development in more impover-ished countries. Another closely related policyfield is that of achieving sustainable urban andregional regeneration in older industrial eco-nomies. Both fields present similar challengesof learning to link central policy perspectiveswith the perspectives of disadvantaged localcommunities, through developing capacitiesfor outreach in both directions. The challengeof helping central policy-makers to learn howto reach out towards clearer appreciation oflocal concerns is becoming widely recognised(Chambers, 1983; 1997). Less widely recog-nised, though just as significant, is the recip-rocal challenge of developing capacities ofoutreach from local decision-makers towardscentral policy sources; so that they can learnto appreciate the uncertainties and the limitsof influence, as well as the tangible resources,

which all people with generic policy responsib-ilities must contend (Friend, 2001).

EMPOWERING FUTUREDECISION-MAKERS

Turning lastly to the sphere of education, whichwe have positioned in the upper-right cornerof Figure 103, issues of inter-generational suc-cession in sustaining the pace of developmenthave inevitably become of growing signific-ance to both of us, as early pioneers of theStrategic Choice Approach, as we have stead-ily moved towards retirement age.One source of potential encouragement lies

in the dramatic growth among young peoplestill at school in access to new information andcommunication technologies (ICT), often leav-ing those who have grown upwith earlier tech-nologies well behind. Many of these youngpeople demonstrate an impressive capacity toput their technological capacities to imaginat-ive use in their leisure pursuits. Yet could thesecapacities also be harnessed in the causeof facilitating difficult collective decisions, ifeducation and career development channelscould be adapted accordingly?What strategiesmight be devised to steer young peoplewho combine motivations to influence sig-nificant decisions with impressive ICT skillstowards career paths involving expertise in thefacilitation of developmental decision-making,whether close to their own local communitiesor in wider policy arenas? What is the poten-tial to link such people together in imaginativeforms of mutual support networks, reachingacross levels of choice and across cultures?The concept of empowerment of people

who start from a position of little power toinfluence the developmental decisions thatimpinge on their future lives has become apotent driver of policy change in many partsof the world. Yet one lesson from our ownrecent action research experiences, as well asfrom the experiences of several of our contrib-utors, is that it is not only members of disad-vantaged local communities that tend to feela lack empowerment in decision-making; so

367


too do many people in central policy roles whostruggle to find effective means of exercisinggreater leverage over intricate local decisions.This explains the concern of many policy-makers and policy advisers with the designof increasingly sophisticated procedures offormal performance measurement and policyevaluation; yet it is apparent that they faceserious conceptual and practical difficulties inshaping such procedures in order to engagewith complex issues of a more subtle develop-mental nature.It will not be enough to pursue new ways

forward either solely through high-level policydebates, or solely through debates within localdecision networks on issues of more specificlocal concern. The facilitators of the future willhave to be sufficient in number, and widelyenough diffused through civil society, to beable to develop and sustain rich connectionsboth between andwithin levels, and across cul-tural divides, if the base of shared understand-ing is to continue to expand.The interlocking challenges that we have

attempted to draw together in Figure 103can only realistically be addressed throughthe influence and inspiration of people inmany kinds of role – managers, planners,teachers, researchers, students, consultants,political leaders, community entrepreneurs.

Such a diffusion will be important if successivegenerations of decision-makers and facilitat-ors are to be helped to make imaginativeuse, selectively and adaptively, of innovativeapproaches to interactive planning and learningof the kind that we have presented on thesepages.If the dialogue that we have opened up by

inviting a number of other people to contributeto this third edition of Planning Under Pressure

can be sustained into the future through otherchannels – not least, through the companionwebsite that our publishers have offered – thenthe dynamics of action research and actionlearning has the potential to gathermomentumin ways that match both the resources forglobal communication now open to us, and thescale of the global issues calling for collectiveaction.For the various pointers to more appro-

priate practice, theory and learning that wehave brought together here are built on manypeoples’ experiences in planning under pres-sure in many settings and at many levels. Theworking lives of all these people are finite; yetthere can be no respite in the challenges facingthem and their successors who will come toshare responsibility for the sustainability of lifeon our troubled planet as this new centuryunfolds.

368

Access to further information

MAKING USE OF OURCOMPANION WEBSITE

The second edition of Planning under Pressure

concludedwith a short section entitledAGuide

to Further Reading, followed by another evenshorter section entitled Points of Contact. Inev-itably, the information in both these sectionsshowed a tendency to become out of datewiththe passage of time.However, In this new twenty-first century

edition, we now have a means to surmountthis problem, through the resource of a com-panion website that has been afforded us byour publisher Elsevier. This means that, when-ever you call up the address http://books.

elsevier.com/companions/0750663731, youwill be able to find references to further read-ing and points of contact which we intend tokeep continually up to date.One of our contributors to Chapter 13,

Brendan Hickling, has agreed to act as our linkwith the Elsevier website and as coordinator ofwhatever new contributions to it may come infrom time to time.Should readers wish to offer any learning

points from their own experience that theythink might be useful to other readers, theyare invited to contact Brendan by e-mail at theaddress given on the website. Because wecannot predict how many readers will respondin this way, we cannot guarantee that whatyou say will find a place on the website. How-ever, if you are motivated to write briefly aboutyour experiences and what you have learntfrom them, or to send in questions addressedto other readers, please get in touch withBrendan by e-mail. You may, of course, make

reference to otherwebsiteswhere further sup-porting information will be found.On the main menu of the website will be

found the following options:

• About the book – basic information on con-tents and translations.

• Action learning forum – including latestnews on applications of Strategic ChoiceApproach and discussion of learning points.

• Books – updated guide to further reading,and additions to the bibliography.

• Contacts – latest contact details and newsabout the two authors, the 21 invited contrib-utors and related institutions - including linksto other websites.

• Learning resources – this section includesanswers and comments relating to the exer-cises printed at the end of Chapters 2 and 3.It also includes guidance on ways of organ-ising short courses for practitioners, andinteractive learning programmes for post-graduate and undergraduate students. It isplanned too to include references to othersites from which handouts can be orderedon realistic small group exercises from differ-ent fields of planning and management thathave been designed by the authors and theirassociates for courses that they have run invarious countries.

• Software – latest information on the availab-ility of supporting software.

FURTHER READING: GENERAL

All references to published books and papersthat were cited in the lists of references inearlier editions have been retained in the bibli-ography that follows this section. Most of the

369


publications that are not already cited in earlierchapters will be cited in this guide.Starting with publications in the English lan-

guage, earlier accounts of the Strategic ChoiceApproach will be found in Friend and Jessop(1969, second edition 1977); in Friend, Powerand Yewlett (1974); in Hickling (1974) and inRosenhead andMingers (eds) (2001) – inwhichStrategic Choice Approach is presented in twochapters by ourselves together with chapterson other interactive problem-structuring meth-ods. Accounts of relevant early work by col-leagues in the Tavistock Institute will be foundin Higgin and Jessop (1965), Harary, Jessop,Luckman and Stringer (1965); Crichton (ed.)(1966); Luckman (1967) Stringer (1967) andMorgan (1971).1

Later overviewsof thework of IORwithin theTavistock Institutewill be found inFriend,Norrisand Stringer (1988); Friend (1997) and Friend,Bryant, Cunningham and Luckman (1998).Any future additions to this list will be posted

on the companion website.

FURTHER READING: APPLICATIONS

References to early applications of the Stra-tegic Choice Approach will be found in Friend,Wedgwood-Oppenheim et al. (1970); Bunker(1974); Bather, Williams and Sutton (1976);Hickling (1978); Decker et al. (1978); Dekkerand Mastop (1979) and van der Graf (1985).Other published applications which we wereable to include as references in the second edi-tion included Hickling (1989); van Steenbergen(1990);Ling(1990);Moullin (1991);Thunhurstetal. (1992); KhakeeandStromberg (1993); Swan-son (1994) Holt (1994); White (1994); Friend(1994) and Ormerod (1995, 1996a, 1996b).Not all these sources are easily accessed.

Two important additions that we can make to

1 A full set of reports and other documents resultingfrom the work of IOR and its successor unit theCentre for Organisational andOperational Research hasbeen lodged in the Modern Records Centre at theUniversity of Warwick, along with other OR recordsco-ordinated by the Archives Committee of the Opera-tional Research Society.

the list as this edition goes to press – both pub-lished in Elsevier journals – include Kammeier(1998) and Bredariol and Magrini (2003), refer-ring to recent applications in Thailand and Brazilrespectively. Other published applications arealready referenced here in contributions toChapter 13: Ormerod (1998, 1999); Horlick-Jones, Rosenhead et al. (2001); Kain (2003)and Phahlamohlaka and Friend (2004).It is planned to extend the list of references

to applications further after this edition goes topress, by short direct accounts of applicationsin the Action Learning Forum section of thewebsite and by additions to the bibliography.

FURTHER READING: LANGUAGESOTHER THAN ENGLISH

An abridged German translation of Local Gov-ernment and Strategic Choicewas published in1973 (Friend and Jessop, 1973), while the firstedition of Planning under Pressure was trans-lated into Japanese by Hirotaka Koike et al.(Friend and Hickling, 1991). This was followedby a Spanish translation of the second edition,sponsored by IVEPLAN in Venezuela (Friendand Hickling, 2002).In addition, briefer guides to the Strategic

Choice Approach have been published in sev-eral other languages. The earlier guides wereadapted from an English language handbookby Allen Hickling, first published in Britain(Hickling, 1974) and subsequently in exten-ded form in Canada (Hickling, 1976). The lan-guages involved in these handbooks wereDutch (Hickling, Hartman and Meester, 1976);French (Hickling, Wilkin and Debreyne, 1980)and Portuguese (Hickling, 1985).Later guides to the approach have been

published in Swedish (Strömberg, 1986);Bahasa Indonesia: (Ismail, Setiabudi and vanSteenbergen, 1989); and again Dutch (Hicklingand de Jong, 1990). Any future additions to thislist will be posted on the companion website.

WIDER READING

Some of the other publications included in thebibliography serve to set the Strategic Choice

370

Access to Further Information

Approach in thecontext ofwider developmentsin such varied fields as strategic businessmanagement, problem structuring methods,urban and regional planning theory, operationalresearch; project management; and participat-ory community development.These include such books as Bryant (1989);

Eden and Radford (eds)(1990); Ritchie, Taketand Bryant (eds)(1994); Wilcox (1994); Trist,Emery and Murray (eds)(1997); Mingers andGill (eds)(1997); Wyatt (1999); Salet andFaludi (eds)(2000); Rosenhead and Mingers(eds)(2001); Webb (2003) and Midgley andOchoa-Arias (eds)(2004).Between 1977 and 1983, several reports

on advances in applications and ideas relat-ing to inter-organisational planning appeared

in a series of seven issues of the news-letter LINKAGE published through the Tav-istock Institute of Human Relations (Friend,Laffin, Norris and Ogden (eds)(1977–83).LINKAGE was initially published in associ-ation with a research programme on inter-organizational relations supported by a grantfrom the Social Science Research Council(now ESRC).The contributions of the work of the ‘IOR

School’ to the development of planning the-ory are reviewed from both European andNorth American perspectives in a special issueof the journal Planning Theory (Mandelbaum,ed., 2004) (Bryson, Ackerman and Eden, 2004)(Burns, 2004) (Faludi, 2004) (Friend, 2004b)(Needham, 2004).

371

Bibliography

ACKOFF RL (1970) A Concept of Corporate Planning.New York: Wiley

ACKOFF RL (1974) Redesigning the Future. New York:Wiley

ACLAND AF (1990) A Sudden Outbreak of CommonSense: Managing Conflict through Mediation. London:Hutchinson Business Books

ACLAND AF (1995) Resolving Disputes without Going toCourt. London: Business Books

ALEXANDER C, ISHIKAWA S and SILVERSTEIN M (1977)A Pattern Language. Towns – Buildings – Construction;New York: Oxford University Press.

AMES SC (1993) A Guide to Community Visioning: Hands-on Information for Local Communities, Oregon Vis-ion Project, American Planning Association (OregonChapter); Oregon, Portland.

ARGYRIS C (1999) On Organizational Learning. Oxford:Blackwell

ARGYRIS C and SCHÖN DA (1973) Organisational Learn-ing: A Theory of Action Perspective. Menlo Park:Addison-Wesley Publishing

ASHBYWR (1956) An introduction to Cybernetics. London:Chapman & Hall

BALBO M (1975) La Pianificazione come Processo diScelte Strategiche. In: P. Ceccarelli (ed.) Potere e PianiUrbanistica. Milano: Franco Angeli Editore

BARLOW J and JASHAPARA A (1998) OrganizationalLearning and Inter-Firm ‘Partnering’ in the UK Construc-tion Industry, The LearningOrganization 5 (2), pp. 86–98.

BATHER NJ, WILLIAMS CM and SUTTON A (1976) Stra-tegic Choice in Practice: The West Berkshire StructurePlan Experience. Reading, UK: University of Reading,Geographical Paper No. 50

BATTY SE (1977) Game-Theoretic Approaches to UrbanPlanning and Design. In: Environment and Planning B 4,211–39

BEER S (1966) Decision and Control. Chichester: WileyBEER S (1981) Brain of the Firm. Chichester: WileyBENNETT J and JAYES S (1998) The Seven Pillars of

Partnering. A Guide to Second Generation Partnering;London: Thomas Telford Partnering.

BENNETT P (1980) Hypergames: Developing a Model ofConflict. In: Futures 12, 489–507

BENNETT P (1985) On Linking Approaches toDecision Aid-ing: Issues and Prospects. In: Journal of the OperationalResearch Society 36, 659–69

BENNETT P and HUXHAM CS (1982) Hypergames andWhat They Do: A ‘Soft OR’ Approach. In: Journal of theOperational Research Society 33, 41–50

BENNETT P, CROPPER S and HUXHAM C (1989) Mod-elling Interactive Decisions: The Hypergame Focus.In: Rational Analysis for a Problematic World, ed.J.V. Rosenhead. Chichester: Wiley

BION WR (1961) Experiences in Groups. London:Tavistock Publications

BREDARIOL CS andMAGRINI A (2003) Conflicts in Devel-oping Countries: A Case Study from Rio de Janeiro.In: Environmental Impact Assessment Review 23,489–513

BREUREA andHICKLINGA (1990) Copingwith Unconven-tional Projects: A ‘Socio-Technical’ Approach. In: Hand-book of Management Projects, ed. R. Garies. Vienna:Manz

BRYANT JW (1989) Problem Management. Chichester/New York: Wiley

BRYANT JW (1993) OR Enactment: The TheatricalMetaphor as an Analytical Framework. In: Journal of theOperational Research Society 44, 551–61

BRYSON JM, ACKERMAN F and EDEN C (2004) Con-tributions of Planning Under Pressure. In: PlanningTheory 3, 3

BUNKER R (1974) Making Decisions in St Albans. In: BuiltEnvironment 1974, 316–18

BUNKER BB and ALBEN BT (1997) Large Group Interven-tion: Engaging the Whole System for Rapid Change.Jossey-Bass

BURNS T (2004) A Practical Theory of Public Planning: TheTavistock Tradition and John Friend’s Strategic ChoiceApproach. In: Planning Theory 3, 3

CARTWRIGHT TJ (1992) STRAD: A New Role for Com-puters in Planning. In: Computers, Environment andUrban Systems, 16, 77–82

CHAMBERS R (1983) Rural Development: Putting the LastFirst. Harlow: Longman

CHAMBERS R (1997) Whose Reality Counts? Putting theFirst Last. London: Intermediate Technology Press

CHECKLAND PB (1981) Systems Thinking, SystemsPractice. Chichester: Wiley

CHECKLAND PB (1989) Soft Systems Methodology.In: Rational Analysis for a Problematic World, ed.J.V. Rosenhead. Chichester: Wiley

373

Bibliography

CHILD J (1972) Organisational Structure, Environment andPerformance: The Role of Strategic Choice. In: Soci-ology 6, 2–22

CLARK AW (1976) Experimenting with Organisational Life:The Action Research Approach. New York: Plenum

CRICHTON (ed.) (1966) Interdependence and Uncertainty:A Study of the Building Industry. London: TavistockPublications

CUSHMAN M and ROSENHEAD J (2004) ‘Planning in theFace of Politics: Reshaping Children’s Health Servicesin Inner London’ in Handbook of OR/MS Applications inHealth Care (F. Sainfort, M. Brandeau and W. Pierskallaeds) Kluwer, Boston, MA, USA

DALKEY NC (1969) The Delphi Method: An ExperimentalStudy of Group Opinions. Memorandum RM-5888-PR.Santa Monica, Calif: Rand Corporation

DEKKER F and MASTOP P (1979) Strategic Choice: AnApplication in Dutch Planning Practice. In: PlanningOutlook 22, 87–96

DEKKER F et al. (1978) A Multi-Level Application ofStrategic Choice at the Sub-Regional Level. In: TownPlanning Review 49, 149–62

DELBECQ AL, VAN DE VEN AH and GUSTAFSON DH(1975) Group Techniques for Programme Planning.Glenview, Illinois: Scott Foresman and Company

DELLO P (1985) Strategic Choice and Evaluation: SomeMethodological Considerations. In: A. Faludi andVoogd H (eds) Evaluation of Complex Policy Problems.Delft: Delftsche Uitgevers Maatschappi

DOYLE M and STRAUS D (1976) How to Make MeetingsWork: Jove

EDEN C (1989) Using Cognitive Mapping for StrategicOptions Development and Analysis (SODA). In: RationalAnalysis for a Problematic World, ed. I.V. Rosenhead.Chichester: Wiley

EDEN C and ACKERMAN F (1998) Making Strategy: TheJourney of Strategic Management. London: Sage

EDEN C and JONES S (1984) Using Repertory Grids forProblem Construction. In: Journal of the OperationalResearch Society 35, 779–90

EDEN C, JONES S and SIMS D (1983) Messing about inProblems. Oxford: Pergamon

EDEN C, JONES S, SIMS D and SMITHIN T (1981)The Intersubjectivity of Issues and Issues ofIntersubjectivity. In: Journal of Management Studies18, 34–47

EDEN C, WILLIAMS H and SMITHIN T (1986) SyntheticWisdom: The Design of a Mixed Mode Modelling Sys-tem for Organisational Decision Making. In: Journal ofthe Operational Research Society 37, 233–42

EDEN C and RADFORD J (eds) (1990) Tackling StrategicProblems. London: Sage

EMERY FE and TRIST EL (1972) Towards a Social Ecology:Contextual Appreciation of the Future in the Present.London: Plenium

ETZIONI A (1968) The Active Society: A Theory of Societaland Political Processes. New York: The Free Press

FALUDI A (1973) Planning Theory. Oxford: PergamonFALUDI A (1984) The Return of Rationality. In: M.J.

Breheny and A.J. Hooper (eds) Rationality in Planning.London: Pion Press

FALUDI A (1986) Critical Rationalism and PlanningMethodology. London: Pion Press

FALUDI A (1987) A Decision Centred View ofEnvironmental Planning. Oxford: Pergamon

FALUDI A (2000) The Performance of Spatial Planning.In: Planning Practice and Research, Vol. 15, No. 4,pp. 299–318

FALUDI A (2004) The Impact of a Planning Philosophy. In:Planning Theory 3, 3

FALUDI A and MASTOP JM (1982) The IOR School –The Development of a Planning Methodology. In:Environment and Planning B 9, 241–56

FALUDI A and VALK AJ van der (1994) Rule and Order:Dutch Planning Doctrine in the Twentieth Century.Dordrecht: Kluwer Academic Publishers

FERGUSON MJ (1979) A Strategic Choice Approach toRecreation Site Resource acquisition. In: Town PlanningReview 50, 325–45

FINLAYSON DE (2000) Internal Legitimacy, CommunityDevelopment, Action Research and DevelopmentEvaluation. In Proceedings of 44th Annual Meeting ofthe International Society for the Systems Sciences,Toronto.

FINLAYSON DE (2004) Embedding Research andEvaluation in Community Regeneration Projects. In:Proceedings of 48th Annual Conference of theInternational Society for the Systems Sciences.California

FISHER R and BROWN S (1989) Getting Together: Build-ing a Relationship that Gets to Yes. London: BusinessBooks

FISHER R and URY W (1981) Getting to Yes. London:Hutchinson

FLOOD RL and JACKSON MC (1991) Creative ProblemSolving: Total Systems Intervention. Chichester: JohnWiley

FLOYDM (1978) Structure PlanMonitoring: Looking to theFuture. In: Town Planning Review 49, 476–85

FRANCO LA, CUSHMAN M and ROSENHEAD J (2004)Project Review and Learning in the ConstructionIndustry: Embedding a Problem Structuring Methodwithin a Partnership Context. In: European Journal ofOperational Research 152 (3), pp. 586–601

FRIEND JK (1976) Planners, Policies and OrganisationalBoundaries: Some Recent Developments in Britain. In:Policy and Politics 5, 25–46

FRIEND JK (1977) The Dynamics of Policy Change. In:Long Range Planning 10, 40–7

FRIEND JK (1980) Planning in a Multi-Organisational Con-text. In: Town Planning Review 51, 261–9

FRIEND JK (1988) Building on Public Sector Wisdom.In: Management Development in the Public Sector:A European Perspective, R. Kakabadse, P. Brovetto andR. Holzer (eds). Aldershot: Avebury

374

Bibliography

FRIEND JK (1989) Planning as Responsible Scheming.In: Operational Research and the Social Sciences, ed.M.C. Jackson, P. Keys and S.A. Cropper. New York:Plenum

FRIEND JK (1989) The Strategic Choice Approach.In: Rational Analysis for a Problematic World, ed.J.V. Rosenhead. Chichester: Wiley

FRIEND JK (1990) Handling Organizational Complexity inGroup Decision Support. In: Tackling Strategic Prob-lems: The Role of Group Decision Support, C.L. Edenand J. Radford (eds). London: Sage

FRIEND JK (1990) Planning Concepts, Planning Contexts.In: Systems Practice 3, 195–206

FRIEND JK (1992) NewDirections in Software for StrategicChoice. In: European Journal of Operational Research61, 154–64

FRIEND JK (1993) Searching for Appropriate Theory andPractice in Multi-Organisational Fields. In: Journal of theOperational Research Society 44, 585–98

FRIEND JK (1993) The Strategic Choice Approach inEnvironmental Policy Making. In: The EnvironmentalProfessional 15, 164–75

FRIEND JK (1993) Planning in the Presence of Uncertainty:Principles and Practice. In: Journal of InfrastructurePlanning and Management (Japan), 4761 IV-21, 1–9

FRIEND JK (1994) Community Involvement in HealthStrategy in Tower Hamlets. In: Community Works, edsC. Ritchie, A. Taket, J. Bryant. Birmingham: OperationalResearch Society

FRIEND JK (1995) Supporting Developmental DecisionProcesses: The Evolution of an OR Approach. In: Inter-national Transactions in Operational Research 2, 225–32

FRIEND JK (1996) Strategy on the Run: Decision Sup-port for Hard-Pressed Managers. In: IT Support in theProductive Workplace, ed. J. Chapman. Cheltenham:Stanley Thornes

FRIEND JK (1997) Connective Planning: from Practice toTheory and Back. In: The Social Engagement of SocialScience. Volume III: the Socio-Ecological Perspective,eds E.L. Trist, F. Emery and H.Murray. Philadelphia: TheUniversity of Pennsylvania Press

FRIEND JK (1998) Operational Choices and StrategicSpatial Planning. In The Revival of Strategic Spatial Plan-ning, eds W. Salet and A. Faludi. Amsterdam: RoyalNetherlands Academy of Arts & Sciences

FRIEND JK (1999) Process Engagement as a Character-istic of Problem Structuring Methods. Paper presentedto the triennial conference of the International Federa-tion of Operational Research Societies, Beijing, August1999. (Journal of the Operational Research Society 56,forthcoming)

FRIEND JK (2001) Engaging with Transient Complexity inDevelopment Projects. In: Understanding Complexity.Kluwer/Plenum, London for International Society for theSystems Sciences

FRIEND JK (2004a) Perspectives of Engagement inCommunity Operational Research. In: Community

Operational Research: OR and Systems Thinkingfor Community Development, eds G.R. Midgley andA.E. Ochoa-Arias. New York: Kluwer/Plenum

FRIEND JK (2004b) A Future for the Non-Academic Plan-ning Theorist? In: Planning Theory 3, 3

FRIEND JK and HUNTER JMH (1970) Multi-OrganisationalDecision Processes in the Planned Expansion of Towns.In: Environment and Planning 2, 33–52

FRIEND JK and JESSOPWN (1971) Hvad er Planlaegning.In: BYPLAN 23, 134–40

FRIEND JK and JESSOP WN (1973) Entscheidungss-trategie in Stadtplanung und Verwaltung. German trans-lation by Zwirner WGO. Dusseldorf: Bertelsmann

FRIEND JK and JESSOPWN (1977) Local Government andStrategic Choice: An Operational Research Approachto the Processes of Public Planning (Second Edition).Oxford: Pergamon [First Edition (1969) – London:Tavistock Publications]

FRIEND JK and HICKLING A (1991) The Strategic ChoiceApproach. Japanese translation of Planning under Pres-sure by H. Koike et al. Tokyo: Gihodo

FRIEND JK, POWER JM and YEWLETT CJL (1974) Pub-lic Planning: The Inter-Corporate Dimension. London:Tavistock Publications. Reprinted (2001) in the Inter-national Behavioural and Social Science Library,Routledge

FRIEND JK, LAFFIN MJ and NORRIS ME (1981) Compet-ition in Public Policy: The Structure Plan as Arena. In:Public Administration 59, 441–63

FRIEND JK, NORRISME and STRINGER J (1988) The Insti-tute for Operational Research: An Initiative to Extend theScope of OR. In: Journal of the Operational ResearchSociety 39, 705–13

FRIEND JK, BRYANT DT, CUNNINGHAM JB andLUCKMAN J (1998) Negotiated Project Engagements:Learning from Experience. In: Human Relations 51,1509–42

FRIEND JK, WEDGWOOD, OPPENHEIM F et al. (1970)The LOGIMP Experiment: A Collaborative Exercise inthe Application of a New Approach to Local PlanningProblems. London: Centre for Environmental Studies(refer to Tavistock Institute of Human Relations)

GIANDRANDE A and MORTOLA E (1996) L’ipertesto perl-applcazione de Strategic Choice a Venezia. In: La Qual-ita dell’ambiente (ed. Mortola). Milano: Francoangeli

GLASER B and STRAUSS AL (1967) The Discovery ofGrounded Theory. Chicago: Aldine

GTZ (1991) Methods and Instruments for Project Plan-ning and Implementation. Eschborn: Geselischaft frTechnische Zusammenarbeit

GUPTA SK and ROSENHEAD J (1968) Robustness inSequential Investment Decisions. In: ManagementScience 15, 13–18

HALL P (1980) Great Planning Disasters. London:Weidenfeld & Nicholson

HAMMERSLEY M and ATKINSON P (1995) Ethnography:Principles in Practice, 2nd edition. Routledge

375

Bibliography

HARARY F, JESSOP WN, LUCKMAN J and STRINGER S(1965) Analysis of Interconnected Decision Areas: AnAlgorithm for Prohect Development. In: Nature 206, 118

HART DA, HICKLING DA, NORRISME and SKELCHER CK(1980) Regional Planning Methodology. London:Tavistock Institute of Human Relations. Internal Paper2T/436

HAWKINS L and HUDSONM (1986) Effective Negotiation.Melbourne: Information Australia

HICKLING A (1974) Managing Decisions: The StrategicChoice Approach. Rugby: Mantec (refer to TavistockInstitute of Human Relations, London)

HICKLING A (1976) Aids to Strategic Choice (SecondEdition). Vancouver: University of British Colombia,Centre for Continuing Education

HICKLING A (1978) AIDA and the Levels of Choice inStructure Plans. In: Town Planning Review 49, 459–75

HICKLING A (1979) Aids to Strategic Choice Revisited.London: Tavistock Institute of HumanRelations. InternalPaper 2T/226

HICKLING A (1981) Abordagerm da Escolha Estra-tegica. Sao Paulo: Fundacao do DesenvolvimentoAdministrativo

HICKLING A (1982) Beyond a Linear Iterative Process. In:B. Evans, J. Powell andR. Ralbot (eds) ChangingDesign.Chichester: Wiley

HICKLING A (1985) Evaluation is a Five Finger Exer-cise. In: A. Faludi and H. Voogd (eds) Evaluation ofComplex Policy Problems. Delft: Delftsche UitgeversMaatschappij

HICKLINGA (1989) Gamblingwith Frozen Fire? In: RationalAnalysis for a Problematic World, ed. J.V. Rosenhead,Chichester/New York: Wiley, pp. 159–92

HICKLING A (1990) Decision Spaces. In: Tackling StrategicProblems, C. Eden and J. Radford. eds London: Sage

HICKLING A and DE JONG A (1990) Mens en Beleid.Leiden/Antwerpen: Stenfert Kroese Uitgevers

HICKLING A, HARTMAN R and MEESTER JG (1976)Werken met Strategische Keuze. Alpen aan den Rijn:Samson Uitgeverij

HICKLING A, WILKIN L and DEBREYNE F (1980)Technologie de la Decision Complexe: Des Aidespour l’Elaboration des Choix Strategiques. Bruxelles:Universite Libre de Bruxelles, Centre E Bernheim pourI’Etude des Affaires

HIGGIN G and JESSOPWN (1965) Communications in theBuilding Industry: The Report of a Pilot Study. London:Tavistock Publications

HILLM (1968) A Goals–Achievement Matrix for EvaluatingAlternative Plans. In: Journal of the American Instituteof Planners 34, 19–28

HILL M (1985) Decision-Making Contexts and Strategiesfor Evaluation. In: A. Faludi and H. Voogd (eds) Evalu-ation of Complex Policy Problems. Delft: Delftsche Uit-gevers Maatschappij

HOLT J (1994) Disarming Defences. In: OR Insight 7 (4),19–26. Birmingham: Operational Research Society

HORLICK-JONES T, ROSENHEAD J, GEORGIOU I,RAVETZ J and LÖFSTEDT R (2001) Decision Support forOrganisational Risk Management by Problem Structur-ing. Health Risk and Society 3 (2), 141–65

HOWARD N (1971) Paradoxes of Rationality: The Theoryof Metagames. London: MIT Press

HOWARDN (1989) TheManager as Politician and General:the Metagame Approach to Analysing Cooperation andConflict. In: Rational Analysis for a Problematic World,ed. J.V. Rosenhead. Chichester: Wiley

HUXHAMC (ed.) (1996) Creating Collaborative Advantage.London: Sage

ISMAEL M, SETIABUDI T and VAN STEENBERGEN F(1989) Modul Pendekatan Pilihan Strategis (in BahasaIndonesia). Aceh: Bappeda Provinsi Daerah IstimewaAceh

JACKSON MC (1992) Systems Methodology for theManagement Sciences. New York: Plenum

JAGO L and NORRIS M (1981) Training in inter-Organisational Working: An Experimental Course. In:Linkage six. London: Tavistock Institute of Human Rela-tions

JAGO L et al. (1983) Strategies for Working betweenOrganisations: A Training Experience Reviewed. In:Linkage seven. London: Tavistock Institute of HumanRelations

JONES S and EDEN C (1980) OR in the Community. In:Journal of the Operational Research Society 32, 335–45

KAIN JH (2003) Sociotechnical Knowledge. An Operation-alised Approach to Localised Infrastructure Planning andSustainable Urban Development. Chambers Universityof Technology, Göteborg

KAMMEIER HD (1998) A Computer-aided StrategicApproach to Decision-making in Urban Planning: AnExploratory Case Study in Thailand. In: Cities 15 (2)pp. 105–19

KELLY GA (1955) The Psychology of Personal Constructs.New York: Norton

KELLY GA (1972) A Theory of Personality. New York:Norton

KHAKEE A and STROMBERG K (1993) Applying FuturesStudies and the Strategic Choice Approach in UrbanPlanning. In: Journal of the Operational Research Soci-ety 44 (3), 213–24

KOLB DA (1984) Experiential Learning. New Jersey:Prentice-Hall

KRAVATZKY AJ (1995) Hiview and Equity: Decision Ana-lysis for Environmental Management. In: PARKS. TheInternational journal for Protected Area Managers 5 (1),32–45

LEITH M (1997) The CLGI Guide to Large Group Inter-ventions. Amsterdam: The Centre for Large GroupInterventions

LICHFIELD N (1966) Cost-Benefit Analysis in Town Plan-ning. A Case Study: Swanley. In: Urban Studies 3,215–49

376

Bibliography

LICHFIELD N (1985) From Impact Assessment to ImpactEvaluation. In: Evaluation of Complex Policy Problems.Delft: Delftsche Uitgevers Maatschappij

LICHFIELD N, KETTLE P and WHITBREAD M (1975) Eval-uation in the Planning Process. Oxford: Pergamon

LINDBLOM CE (1965) The Intelligence of Democracy:DecisionMaking throughMutual Agreement. NewYork:Free Press, London: Collier-MacMillan

LING M (1990) Making Sense of Planning Evaluations:A Critique of Some Methods. In: Planning and Develop-ment (Hong Kong), 6 (1), 14–20

LUCKMAN J (1967) An Approach to the Management ofDesign. In: Operational Research Quarterly 18, 345–58

MANDELBAUM S (ed.) (2004) A Symposium on the Workof John Friend. In: Planning Theory 3, 3

MASON RD and MITROFF II (1981) Challenging StrategicPlanning Assumptions. New York: Wiley

MIDGLEY GR and OCHOA-ARIAS AE (eds) (2004)Community Operational Research: OR and SystemsThinking for Community Development. New York:Kluwer/Plenum

MILLER EJ and RICE AK (1967) Systems of Organization.London: Tavistock Publications

MINGERS J and GILL A (eds) (1997) Multimethodology.Chichester: Wiley

MOCKLER RJ and DOLOGITE DG (1991) Using ComputerSoftware to Improve Group Decision-Making. In: LongRange Planning 24, 4447

MOORE CM (1987) Group Techniques for Idea Building.London: Sage

MORGAN JR (1971) AIDA – A Technique for the Manage-ment of Design. IOR Monograph No. 2. London: Tavis-tock Institute of Human Relations

MORTOLA E (ed.) (1996) La Qualita dell’ambiente. Milano:Francoangeli

MOULLINM (1991) Getting Planners to takeNotice. In: ORInsight 4 (1), 25–9. Birmingham: Operational ResearchSociety

NEEDHAMB (2004) John Friend: Advising and Theorising.In: Planning Theory 3, 3

NEUMANN JE, HOLTI R and STANDING H (1995) ChangeEverything at Once! Didcot: Management Books 2000

NORRIS ME (1985) Operational Research and the SocialSciences: Review of Local Government and StrategicChoice. In: Journal of the Operational Research Society36, 870–2

NUNAMAKER JF, DENNIS AR, VALACIO JS, VOGEL DRand GEORGE JF (1991) Electronic Meeting Systems toSupport Group Work. Communications of the ACM 34

(7), 40–61OPENSYSTEMSGROUP (1981) SystemsBehaviour (Third

Edition). London: Harper & RowOPENSHAW S and WHITEHEAD P (1986) The Decision

Optimising Technique: History, Development andProgress Towards a Future Machine Based PlanningSystem. In: K.G. Willis (ed.) Contemporary Issues inTown Planning. Aldershot: Gower

ORMEROD R (1983) Corporate Planning and its use ofOR in the NCB: A Personal View. In: Journal of theOperational Research Society 34, 461–7

ORMEROD R (1995) Putting Soft OR Methods toWork: Information Systems Strategy Development atSainsbury’s. In: Journal of the Operational ResearchSociety 46, 277–93

ORMEROD R (1996a) Information Systems StrategyDevelopment at Sainsbury’s Supermarkets using ‘Soft’OR. In: Interfaces 26, 102–30

ORMEROD R (1996b) Putting Soft OR Methods to Work:Information Systems Strategy Development at RichardsBay. In: Journal of the Operational Research Society 47,1083–97

ORMEROD R (1998) Putting Soft OR Methods toWork: Information Systems Strategy Development atPalabora. In: Omega International Journal of Manage-ment Sciences 26, 75–98

ORMEROD R (1999) Putting Soft OR Methods toWork: The Case of the Business Improvement Projectat PowerGen. In: European Journal of OperationalResearch 118, 1–29.

OWEN HH (1992) Open Space Technology – A User’sGuide, Cabin John, MD: Abbot Publishing

PAULEY GS and ORMEROD R (1998) The Evolution ofa Performance Measurement Project at RTZ. In: Inter-faces 28, 94–118

PHAHLAMOHLAKA LJ and FRIEND JK (2004) Com-munity Planning for Rural Education in South Africa. In:European Journal of Operational Research 152, 684–95

PHAHLAMOHLAKA LJ and LOTRIET HH (2002) Aninvestigation into community development imperativesbrought about by the recent establishment of a localarea network and access to the Internet at a rural SouthAfrican community education centre. Proceedings ofthe seventh International Working Conference of theInternational Federation of Information Processing (IFIP)WG 9.4: Bangalore, India, 29–31 May 2002

PHILLIPS LD (1982) Requisite Decision Modelling. In:Journal of the Operational Research Society 33, 303–12

PHILLIPS LD (1990) Decision Analysis for Group DecisionSupport. In: Tackling strategic problems: the Role ofGroup Decision Support, C. Eden and J. Radford (eds).London: Sage

RAIFFA H (1968) Decision Analysis. Reading. Mass:Addison-Wesley

REVANS RW (1983) The ABC of Action Learning. Bromley:Chartwell-Bratt

RICE AK (1965) Learning for Leadership – Inter-Personal and Inter-Group Relations. London: TavistockPublications

RICKARDS T (1974) Problem Solving through CreativeAnalysis. Epping: Gower Press

RITCHIE C, TAKET A and BRYANT J (1994) CommunityWorks: 26 Case Studies of Community OR in Action.Birmingham: Operational Research Society

377

Bibliography

ROBINSON I (ed.) (1972) Decision-Making in Urban Plan-ning: An Introduction to New Methodologies. London:Sage

ROSENHEAD JV (1978) An Education in Robustness.In: Journal of the Operational Research Society 29,105–11

ROSENHEAD JV (1980) Planning under Uncertainty:II A Methodology for Robustness Analysis. In: Journalof the Operational Research Society 31, 33–42

ROSENHEAD JV (ed.) (1989) Rational Analysis for a Prob-lematic World. Chichester/New York: Wiley

ROSENHEAD JV and MINGERS J (ed.) (2001)Rational Analysis for a Problematic World Revisited.Chichester/New York: Wiley

SAATY T (1980) The Analytical Hierarchy Process.New York: McGraw-Hill

SAGAR T (1993) Paradigms for Planning: a Rationality-Based Classification. In: Planning Theory 9, 79–118

SCHARPF FW (1972) Komplexitat als Schranke derPolitischen Planung. Politische Innovation undGesellschafliche Wandel: Politische Vierteljahresschrift

SCHNELLE E (1973) Metaplan: Op Zoek naarDoelstellingen Leerproces an Medewerkers Betrokken.Amersfoort: DHV

SCHON DA (1971) Beyond the Stable State: Public andPrivate Learning in Changing Society. London: TempleSmith

SHAKUN MF (1995) Restructuring a Negotiation withEvolutionary SystemsDesign. In: Negotiation Journal 11(2), 145–50

SÖDERBERG H and KAIN J-H (2002) Integrating Know-ledge or Aggregating Data – Multi-Criteria Approachesfor SustainableWater andWaste Systems. Proceedingsof 3rd International Conference on Decision-Making inUrban and Civil Engineering, London

STRINGER J (1967) Operational Research for Multi-Organisations. In: Operational Research Quarterly 18,105–20

STRÖMBERG K (1986) Planera met Onsakerheter!(in Swedish) Stockholm: Byggforskningsradet

STRÖMBERG K (2001) Facilitating Collaborative DecisionDevelopment in Urban Planning. Scandinavian Journalof Architectural Research 4

SUSSKIND A and CRUIKSHANK J (1987) Breaking theImpasse: Consensus Approaches to Resolving PublicDisputes. New York: Basic Books

SUTTON A and WILKIN L (eds) (1986) Related Socio-Technical Approaches to the Management ofUncertainty. Dordrecht: Martinus Nijhoff

SWANSON J (1994) Sending out the Right Packages. In:The Surveyor (UK), 18 August

TAKET AR and WHITE LA (2000) Partnership and Parti-cipation: Decision-Making in the Multiagency Setting.Chichester: Wiley

THUNHURST C, RITCHIE C, FRIEND JK and BOOKER P(1992) Housing in the Dearne Valley: Doing CommunityOR with the Thurnscoe Tenants Housing Co-operative.Part I – the Involvement of the Community OR Unit. In:Journal of the Operational Research Society 43, 81–94

TRIST EL, HIGGIN GW, MURRAY H and POLLOCK AB(1963) Organizational Choice. London: Tavistock

TRIST EL and MURRAY H (eds) (1990) The SocialEngagement of Social Science. Volume I: TheSocio-Psychological Perspective. London: Free Associ-ation Books

TRIST EL and MURRAY H (eds) (1993) The SocialEngagement of Social Science. Volume II: TheSocio-Technical Perspective. Philadelphia: The Univer-sity of Pennsylvania Press

TRIST EL, EMERY, F and MURRAY H (eds) (1997)The Social Engagement of Social Science. VolumeIII: the Socio-Ecological Perspective. Philadelphia: TheUniversity of Pennsylvania Press

URY W (1991) Getting Past No: Negotiating with DifficultPeople. London: Business Books

VAN AKEN JE (2004) Management Research based on theParadigm of the Design Sciences: the Quest for Field-tested and Grounded Technological Rules. In: Journal ofManagement Studies 41–2, 219–46

VAN DE GRAAF R (1985) Strategic Choice in LPGPolicy. In: A. Faludi and H. Voogd (eds) Evaluation ofComplex Policy Problems. Delft: Delftsche UitgeresMaatschappi

VAN GUNDY AB (1981) Techniques of Structured ProblemSolving. New York: Van Nostrand Reinhold

VAN STEENBERGEN F (1990) The Strategic ChoiceApproach in Regional Development Planning. In: ThirdWorld Planning Review 12 (3), 301–04

VICKERS G (1965) The Art of Judgement. London:Chapman & Hall

VOGEL DR (1993) Electronic Meeting Support. In: TheEnvironmental Professional 15, 198–206

WATSON SR and BUEDE DM (1987) Decision Synthesis.Cambridge: Cambridge University Press

WEBB A (2003) The Project Manager’s Guide to HandlingRisk. Aldershot: Gower

WEDGWOOD-OPPENHEIM F (1972) Planning UnderUncertainty. In: Local Government Studies 2, 53–65

WHITE L (1994) Development Options for a RuralCommunity in Belize – Alternative Development andOperational Research. In: International Transactions inOperational Research 1 (4), 453–62

WILCOX D (1994) The Guide to Effective Participation.Brighton: Partnership Books

WORLD BANK (1996) The World Bank ParticipationSourcebook. Washington DC: The World Bank

WYATT R (1999) Computer-Aided Policymaking: Lessonsfrom Strategic Planning Software. London: E & FN Spon

378

Index

accountability, 7, 17, 64, 81, 88, 151, 179, 225,

239, 263

accountable roles, 225

Ackoff, Russell, 75

action forms, 290, 291

action gallery, 280, 287

action learning, 295–6

action research workshops, 355–6

action scheme, 58, 60, 69, 181, 195

flexibility, 60, 190

robustness, 60

timing decisions, 190–1, 193

actions, as element in commitment package, 43,

63–4, 69, 72

adaptation, 88, 273

customising SCA, 282–3

engaging participants in identifying issues,

273–5

facilitating extensions to common ground, 276,

277–9

framing way forward in face of uncertainty,

279–80

language to context, 283

in practice, 282–93

securing commitment to action proposals, 280

setting structure by differentiating levels of

choice, 275–7

socio-technical design, 280–2

adoption of progress package, in STRAD, 61, 181,

201, 208, 209, 252, 268, 269, 278–80, 295,

305, 312, 314, 320, 324, 331, 332, 343, 347,

353, 355, 359

advantage comparison, 48, 50, 52, 69, 147, 235

agenda building, 91

analysis of interconnected decision areas (AIDA),

23, 33, 40, 43, 109, 119,127, 129, 215, 245,

277, 308, 309, 336–9

Angell, Rob, 336

approach, 69, 71

appropriate technology for strategic choice, 19,

43, 75, 217, 321, 365

areas of concern, 55, 295, 309

arena, political, 7, 179, 223, 226, 241, 243, 258,

259, 261, 263, 264

Argyris, C., 295

Ashby, Ross, 153

assessment:

choice of method within comparison area,

155–7

expressing feelings of uncertainty, 157, 159–61

options within critical critical uncertainty areas,

189–90

A-TOPP (Approach-Technology, Organisation,

Process, Product), 69, 71, 76, 87, 229, 282

balanced evaluation framework, 150, 151–2

Beer, Stafford, 328

Benaiges, Ana Maria, 345

Bennett, Peter, 195, 340

brainstorming, 281

Bregman, Leny, 357

Bryant Jim, 370, 371

Bunker, Raymond, 370

calibration of advantage scale, 161, 162

campus waste management project, 346–7

capacity-building, 345–8

Carousel, 280–1, 287

Carsjens, Gerrit Jan, 315

categories:

of decisions, 77, 93–5, 105, 110

of impact, 151

of uncertainty, 9, 17, 55, 59, 77, 183, 205

Checkland, Peter, 75, 299, 314, 328

choice, see strategic choice

levels of, 27, 93–4

strategic vs simple, 241

choosing mode, 17, 18, 43, 51, 53, 57, 61, 63,

64, 69, 82, 83, 89, 97, 110, 121, 137, 161,

167, 169, 179, 201–203, 213, 217, 226, 237,

342, 353

Clark, A.W., 212

co-ordination, 9, 11, 79, 105, 223, 226, 252,

321, 329

379

Index

co-ordinator role, 91, 225, 327

coalescing decision areas, 129–33

combining options, 167

commitment, 357–60

commitment package, 61, 62, 69, 83, 85, 181,

275, 281

alternative, 197, 199

design of, 197

empowerment through IT, 353–6

recording, 234

common grounding, 276, 277, 289

communication, 364

extensive projects, 261

group, 218–21

strategy, 349–50

communicative learning, 303–305

comparative advantage, 159–60

comparing mode: 15, 16, 109, 117, 167, 201,

209, 245, 246–55, 261, 317, 319–21, 324,

327, 329, 330, 340

comparison area, 44, 45, 69, 147

assessing consequences within, 47, 49

building a framework, 347–8

choice of method of assessment within,

155–7

combining advantage judgements, 161–3

concept, 47–8

developing set of relevant, 148, 149, 151

dynamic, 64

further extensions, 167

judging advantage between alternatives,

49–51

multi-criteria approach, 45

multiple, 166–7


restricting focus among decision schemes, 51–3

uncertainty, 53–7

comparison of alternatives, 152–3, 155

compatibility matrix, 35

compatibility of options, 37

complexity, 312–14

composite options, 122, 123, 128, 129, 137

comprehensiveness, 13, 17, 19, 67, 71, 365, 367

computer methods, 39, 76, 110, 135

confidence, 189

conflict management, 331–5

connectedness of decisions, 3, 29, 89, 101,

115, 339

connections, lateral, 72, 77, 87

consensus-building, 259

consistency:

between decisions and policies, 135, 279

consultancy:

computer-aided, 19, 254

mutual, 213

projects, 213, 325–8

in strategic choice, 72, 113, 215

context of project, 291, 293

adapting language to, 283

appropriate workplace, 268

committed group, 268

experienced project leader, 268

external sponsor, 268

internal sponsor, 268

long enough project, 268

non-hostile culture, 268

significant set of issues, 268

successful, 268–9

contingency planning, 61, 193–5, 197, 200, 239

continuity:

cyclic, 72, 81, 82, 87, 89, 264, 268

in planning, 241

control, 40, 77, 81, 82, 85, 88, 91, 95, 103, 119,

143, 151, 153, 177, 182, 195, 223, 229, 255,

258, 279, 301, 303, 306, 314, 320, 321, 330,

346, 347

core concepts, 43, 63, 67, 69, 181, 211, 215, 217

core group activity, 230–1

corporate planning, 13, 73, 255

counterpart role, 13, 43, 225

Coventry project, 5, 7

criterion, 45

Cropper, Stephen, 195

cross-organisational learning approach (COLA),

341–2

curriculum planning, 315–16

Cushman, Mike, 340

cyclic continuity, 72, 81, 82, 87, 89, 264, 268

loop/switch, 82, 88

cyclicity vs linearity, 19, 365

de Jong, Arnold, 143, 207, 370

decision analysis, 35, 39

decision area, 24, 25, 80–1, 90

altering sequence in a tree, 133–5

building relevant, 91–3

coalescing, 128, 129, 131, 133

composite options, 123

concept, 25–7

highlighting important, 101–103

identifying sets of options, 121–3

links between, 27–9, 97–9

option bars, 125–7

options within, 33–5

380

Index

recording, 124, 125, 157, 234

reformulating, 95–7, 123

sorting by categories/levels, 93–5

testing range of options, 121

decision delay, 189

decision graph, 29, 30, 67, 97, 99

rearranging, 99–101

recording, 236

superimposing organisational

responsibilities on, 103–105

decision link, 26, 29

decision-makers, 365–6

decision-making, 1, 364–5

choices in timing, 59–60

collective, 3, 4–5

complexity, 312–14

designing/comparing courses of action, 15, 17

difficulty in, 7–9

facilitating public–private, 303

focus on, 72, 73

human settings for, 13–15

identifying key, 347

incremental practice, 5

individual/group, 15

inputs/outputs, 17, 18

interconnected agendas, 11

models of, 17

shape of issues, 5

technical/political, 5

uncertainties in, 9

uncertainty areas in perspective, 182–5

urgency/uncertainty, 5

and use of STRAD software, 255–6

decision option, 31, 33, 34, 67

compatibility of interconnected areas, 33

decision scheme, 37, 38, 67, 69

feasible, 130, 131, 133, 135

decision stream, 136

decision support system, 254, 315, 366

decision-takers, 199–201, 225–6

vs decision makers, 7, 198, 199, 200, 201, 203,

225, 226, 235, 237

decision theory:

developmental, 366

decision tree, 39

decisions:

focus on, 72, 73, 87, 217, 309

immediate, 61, 195, 196, 278

timing of, 59, 60, 190

delay:

consequences of, 193

in decision making, 59, 189, 190

Delphi method, 161, 267

democracy, 306

design skills, 119, 121, 136–9


designing mode, 15, 16, 31, 33, 67, 89, 95, 101,

105, 107, 109, 110, 127, 131, 136, 137, 138,

139, 169, 185, 249, 323

development challenge:

application/facilitation, 361–2

convergence/divergence in approaches, 365–6

decision-making, 366–7

empowering future decision-makers, 367–8

local/broader perspectives on sustainable

development, 367

mapping, 361–3

methodology/communication, 364

pointers, 361

sponsorship/technology, 364–5

developmental decision making, 366, 367

differential learning, 349–52

dilemmas of practice, 5–7

focused/synoptic, 5

reactive/interactive, 5

reducing/accommodating, 5

simplifying/elaborating, 5

dispute resolution, alternative, 258

documentation, 87, 125, 139, 144, 199, 235, 237,

243, 266, 301, 324

doing, as mode, 241, 242

dominated decision schemes, 165, 166, 167

dummy decision area, 126, 129

dynamic comparison, 64

dynamic grouping, 229, 266

Eden, Colin, 85, 371

educational challenge, 361

effectiveness, 88

efficiency, 305–306

elaboration vs simplification, 47, 49, 64, 76, 101,

137, 169, 211

empowerment, 367–8

ends/means analysis, 94, 110, 114, 115, 145, 295

Environment Council, 331, 336, 338, 365

environmental:

policy, 257, 308–11

project, 290, 291

evaluation, 43, 51, 57, 59, 63, 65, 71, 81, 88, 134,

136, 141, 147, 149, 150–3, 161, 167, 169,

179, 189, 211, 216, 217, 220, 229, 238, 239,

243, 267, 269, 281, 301, 304, 305, 312, 313,

314, 322, 359, 368

Evers, Frans, 308

381

Index

expert techniques, 71

exploratory options, 56, 57, 69, 181, 189–90,

195, 235

team, 57, 69, 209, 211, 258, 267, 291, 332, 351

extensive projects:

application of concept, 261–4

collaborative methods, 257

communication, 258

complexity of issues, 257

creating context for success, 268–9

ISCRA model, 263–5

management skills, 266–8

participants, 258

policy trends, 257

political context, 257–9

project leader, 265

reference/representative groupings, 265–6

stakeholder analysis, 265

time-frames, 258

twin-track technical/policy streams, 259–61

facilitator role, 117, 145, 173, 177, 209, 220, 225,

227, 231, 233, 238, 247, 253, 254, 266, 267,

275, 281, 283, 300, 301, 313, 314, 323, 329,

331, 336, 342, 343, 346, 349, 350, 353,

355, 368

Faludi, Andreas, 315, 316, 371

Finlayson, Dennis, 327

flexibility, 152–5, 190

focus for comparison, 53, 156

focus on decisions, 72, 73

forecasting, 7, 57, 77, 81, 323

formal vs informal channels, 7, 79

Franco, Alberto, 340

Friend and Jessop, 5, 157, 370

Friend, Dave, 319

Friend, John, 5, 245

games, theory of, 255

Giangrande, Alessandro, 322

Giordani, Jorge, 346

goals, 7, 59, 151, 275, 309, 325, 360

good practice, 212–13

graph:

decision, 28–31, 33, 35, 37, 40, 41, 67, 89, 95,

97, 99–106, 109, 110, 112, 113, 115, 116,

117, 127, 133, 182, 183, 219, 235, 249,

299, 319, 324, 342

option, 35, 36, 37, 39, 40, 69, 124, 125, 127,

129, 131, 139, 142, 143, 145, 195

uncertainty, 183–7, 190, 204, 205, 219

group communication/interaction, 218–21

group process, 231

grouping and linking, 221, 225, 229

groupings vs groups, 223, 227, 235, 237,

243, 280

guided tour, 287

guidelines, operational, 73, 87, 88, 211

Harris, Richard, 331

Hawkins, Leo, 267

Herron, Rebecca, 327

Hickling, Allen, 115, 143, 145, 173, 207, 241, 243,

277, 295, 306, 308, 309, 317, 331, 349,

363, 370

Hickling, Brendan, 349

hierarchy, 79, 134, 225, 277

Howard, Nigel, 195

Hudson, Michael, 267

Hunter, Hadley, 129, 245

Huxham, Christine, 195

idea presentation, 351

implementation, 142, 143, 226, 237, 252, 257,

265, 271, 278, 280, 282, 289, 291, 293, 301,

306, 315, 344, 358, 359

incompatibility of options, 35, 125, 127,

137, 141

incremental progress, 72, 83

individual working, 71

induction meetings, 351

information:

giving, 350–1

relevant, 77

supporting, 237

information technology (IT), 245, 353–6

Institute for Operational Research, 245

inter-agency agreement, 357–60

inter-agency working, 226

inter-departmental working, 226

interaction:

adapting to pressure/maintaining momentum,

230, 231

continuity and the individual, 239–43

convergence of view, 238

core group task, 230

gather feedback, 229

group, 218–21

jargon, 235

managing, 221–9

recording, 234–5

report progress, 229

visible products, 235

382

Index

interactive, 5, 19, 64, 71, 72, 75, 76, 77, 79, 81,

82, 85, 87, 88, 93, 101, 110, 121, 124, 127,

130, 131, 137, 139, 148, 149, 154, 158, 162,

180, 207, 213, 217, 218, 219, 220, 221, 225

interactive participation, 77

interconnectedness of decision areas, 89, 337, 339

interests, representation of, 179, 223, 265

internet, 353

interpretation, 87

intra-group roles:

co-ordinator, 225

facilitator, 225

leader, 223

recorder, 225

invisible products:

documentation of, 87, 237–8, 266, 267, 268

IOR, see Institute for Operational Research

ISCRA model, 262, 263–4

extended, 264–5

Jackson, Michael, 365

jargon, interpretation of, 87, 235, 316

Jessop, Neil, 5

Jones, S., 85

judgements:

in choice mode, 201–203

in comparing mode, 161–3, 167–9

in designing mode, 136–8

in shaping mode, 107–10

justification of proposals, 85, 137, 235

Kain, Jaan-Henrik, 303

knowledge, sources of, 159, 161

Koike, Hirotaka, 370

Kolb, D.A., 295

Laffin, Martin, 371

language, 283–4

lateral connections, 72, 77

learning, 330, 347–8

by doing, 215, 317

curve, 231, 233

process, 72, 229–33

levels of choice, 93–5, 308–11

exploring consistency in, 135–6

link, see decision link

linking and grouping, 79, 88, 225, 229

local area network (LAN), 353–4

Local Government and Strategic Choice, 157, 370

LOCOMOTIVE project, 304–306

LOGIMP experiment, 212, 238

looping and switching, 169

Luckman, John, 33, 131, 370

management:

challenges to, 17–19

linearity, objectivity, certainty,

comprehensiveness, 17, 19

of organisation, 227–8

metaplan, 281

Microsoft, 247

mistakes, learning from, 215, 253

modes of strategic choice, 21, 83, 247, 328

money, 217

as a resource, 59, 217, 278

monitoring, 88

Moore, C.M., 281

Morgan, James, 370

Mortola, Elena, 322

multi-agency workshop, 357–60

multi-organisational group, 239

multiple comparison areas, 165–7

multiple option bars, 126, 127, 129

Murray, Hugh, 280

mutual consulting workshop, 327–30

mutually exclusive options, 33, 121, 123, 145

national policy, applications to, 203, 206, 213, 297,

336, 348, 365

negotiation, 227, 332

skills, 267

networks, 266, 351

newsletters, 350–1

nominal group technique, 281

Norris, Michael, 69, 370, 371

Notting Hill Carnival, 298–302

objectives, 7, 9, 59, 81, 151, 152, 267, 299, 300,

355, 367

open technology, 71, 213

managing in practice, 217–21

operational guideline, 72–3, 77

operational research, 176, 177, 245, 296, 297,

298, 327, 340, 370

Operational Research Society, 144, 145, 312

operations, field of, 243, 259, 261, 265

option bar, 34, 35, 67, 69, 124, 125, 127

multiple, 126, 127–9

option graph, 36, 37, 38

option tree, 39, 131, 133, 135, 136

options, see decision option

composite, 122, 123

identifying sets of, 121, 123

for mutual gain, 275, 277

383

Index

organisation, 3–5, 69, 211

extensions to, 265–6

management check-list, 228

managing interactive participation, 221–9

responsibilities, 224

organisational choice, 76–81

bilateral linking, 79

levels of response to uncertainty, 78

multi-lateral grouping, 79, 87

structuring of arrangements, 81

orientation, 71–2, 86

interpreting, 72–3

Ormerod, Richard, 312

ownership, 349–50

of ideas, 271

paper, use of, 75, 76, 91, 93, 94, 105, 181,

196, 218–21, 234, 247, 249, 264, 281,

287, 346

participation, 71, 72, 76, 77, 79, 81, 87, 220, 221,

227, 229, 230, 233, 237, 267, 299, 304, 316,

317, 318, 324, 325, 336, 351, 352

partitioning of a decision graph, 133

Pattern Language, 322–6

Phahlamohlaka, Jackie, 353

philosophy of planning, 1, 4, 9, 72

photo-record, 144, 234

picture gallery, 287

planning, 235, 349–50

as craft, 3

as strategic choice, 1, 3

as universal activity, 3

challenges to, 17–19

incremental progress in, 11–13

linearity, objectivity, certainty,

comprehensiveness, 17, 19

new approaches, 345–8

notion of, 4

philosophy of, 1, 4

realities of, 2

responding to difficult, 7–9

plutonium, 331–5

policy choice, 154

policy-making, 308–11

political stream, 259–61

combined with technical stream, 259–61, 263

see also twin-track approach

political–technical U-loop, 261

Power, John, 161, 239, 370

pre-coordination, 226, 280

pressure, working under:

fast and effective, 216

quick and dirty, 216

problem focus, 30, 31, 67

choice of, 31–3

considerations in choice of, 105–107

problem-solving, 72

problems:

categories of, 77

difficulty, 77

process, 69, 72, 85, 211, 229–33

process choice, 295

process management, 225, 258

checklist, 232

choice, 81–3

extensions to skills, 266–8

product, 69, 72, 211

(in)visible, 234–5, 238–9

management checklist, 240

managing incremental progress in practice,

233–9

recording, 234–5

product choice, 83–7

(in)visible products, 85–6

intersubjectivity, 85

justification of action, 85

recording/interpreting, 85, 87

progress:

incremental, 11, 61, 72, 73, 83, 87, 89, 169,

191, 233

progress package, 278, 279, 280, 295, 305

see also commitment package

project group, 226, 265

project leader, 268

project manager, 225

promenade, 287

proposals, 280

ranking, 51, 53, 165, 166–7, 342

rationality, 82, 237, 316, 337, 338

reactive vs interactive working, 5, 81, 221,

264, 266

recommendations, 57, 82, 97, 103, 109, 197, 199,

207, 209, 212, 230, 235, 237, 238, 249, 263,

283, 312, 332, 334, 338

recording:

interpretation, 234–5

loose-leaf, 234, 235

photo-record, 234

reference/representative groupings, 226, 235, 265

reformulating decision areas, 95, 122, 123

relative assessment, 46, 47, 69, 147

repertoire, personal, 215

representative:

roles, 151, 223

384

Index

resources, 189

Revans, R.W., 295

Rickards, T., 281

robustness, 60, 152, 193, 251, 279, 298, 339

roles:

in a working group, 239

responsible, 266

Rosenhead, Jonathan, 298

routine procedures, 72

scanning, as mode, 241, 242

scheduling, 82

scheme

action, 58–61, 69, 190–1, 193, 195

decision, 38, 39–41, 43, 47

Schemers, 4

Schnelle, E., 281

Schön, D.A., 295

selectivity, 87

in planning, 88

sensitivity analysis, 186, 187

shaping mode, 17, 18, 89–107


process judgements in, 107–10

shortlists, see Working shortlist

recording, 234, 236

simplification vs elaboration, 47, 49, 64, 76, 137,

169, 211

Sims, D., 85

skills:

conflict mediation, 267

effective negotiation/participation, 267

extensions to, 266–8

information flows, 267

network management, 267

organisational/group, 266

process design, 266

project evaluation, 267–8

project planning, 268

Smithin, T., 85

socio-political work streams, 259

socio-technical design, 280–2

soft systems methodology/approach, 75, 299, 314

software:

designing electronic support, 319–21

see also STRAD software

South Side Story, 23, 29, 40, 64, 73, 87, 181, 182,

201, 203

specificity vs generality, 73, 94, 149, 152

sponsorship, 226, 268, 301, 361–5

SSM, see soft systems methodology/approach

stakeholder, 223

analysis, 265

grouping, 226–7

workshops, 296–300

steering group, 226

STRAD software, 245–7

advances in technology/extensions in access,

254–5

as aid to progress in small informal meetings,

253–4

as interactive learning tool, 254–5

challenges in design, 320

comparing/choosing, 250, 251–2

computer-aided consulting, 254

development, 320

early work on, 319–20

experiences in using, 252–3

implication of widening technology access,

255–6

keeping pace with change, 320–1

shaping/designing, 249–51

windows of access, 246, 247–9

strategic action planning, 331–5

strategic choice, 1, 119, 179–99, 211

adaptation of, 258

as continuous process, 242

base of experience, 212–13

categories of, 93–5

choosing tools under pressure, 216–17

combining with other methods, 322–6

comparison, 47, 49

customising, 282–3

deferred, 191, 197

dilemmas of practice, 5–7

exploratory/decisive balance, 6, 11

future procedures, 237

helping decision-takers, 199–201

keeping it simple, 243

linear framework, 237

modes of, 21, 83, 247, 328

open technology, 217–20

organisationl responsibilities, 224

organisational contexts, 3–5

orientations/guidelines, 70, 72

planning as, 1, 3, 4, 7, 9, 11, 13, 15, 43, 72, 147

post-rationalisation, 237


process judgements in, 201–203

recording/interpreting, 235–9

responsible, 199

setting levels of, 274

strategic level, 1, 3

technology of, 19, 21

385

Index

strategic choice, (Continued)

toolbox analogy, 213–16

vocabulary, 68

strategic products, 73, 83, 84

Stringer, John, 319, 370

Strömberg, Knut, 303

sub-groups, 227, 229

substance vs process in products, 83, 87, 225,

233, 238, 239

surprise limit method, 159–60

sustainable development, 255, 296, 303, 306,

336, 365, 366, 367

Sutton, Alan, 216, 233, 370

Sutton Principle, 216–17

switching/looping, 82, 88, 229–33, 237

synoptic vs focused treatment of problem scope,

5, 9

systems perspective, 73

targeted networking, 351

Tavistock Institute of Human Relations, 371

team facilitating, 231

technical stream, 259

see also twin-track approach

technical support team, 266

technological choice, 73–5

interacting/facilitating, 88

technology, 69, 71, 211, 364–5

management check-list, 222

managing, 217–21

theory, 4, 5, 77, 88, 157, 162, 315, 317, 328, 362,

363, 365, 366, 368, 371

timing, 187

choice of action, 190–1, 193

extensive projects, 257

toolbox, 214, 259, 273, 311, 315, 320

adaptiveness, 214–15

selectivity, 214, 215

trade fair, 287

training, 212–13

Trist, Eric, 280

twin-track approach, 259–61

application to project planning, 261–4

tyre dumps, 336–9

UE, uncertainties about the working environment,

9, 15, 21, 55, 57, 59, 76, 77, 175, 180–3, 185,

189, 190, 195, 205, 207, 267, 278, 320

U-loop, political-technical, 261, 263

uncertainty, 8, 161, 275, 277

about guiding values, see UV, uncertainties

about guiding values

about related decisions, see UR, uncertainties

about related decisions

about the working environment, see UE,

uncertainties about the working

environment

environment

assessing comparison areas, 49

assessing exploratory options, 189–90

choosing incrementally, 63–4

comparing alternative responses, 188

expression of, 47, 157, 159–60

extending problem focus, 12, 13

graph, 183, 185

levels of response, 78, 79

managing, 9–11, 44–5, 88, 258

preparing for alternative contingencies,

193–5

uncertainty area, 53, 54, 57, 69

intractable, 185

putting into clearer decision perspective,

182–5

recording, 182, 235

relevant, 183

restructuring composite, 185–9

working list of, 180, 181–2

uncertainty graph, 183, 184, 185, 186, 187, 190,

204, 205, 219

UR, uncertainties about related decisions, 9, 11,

13, 15, 21, 55, 57, 59, 76, 77, 79, 136, 175,

180, 181, 182, 183, 185, 189, 190, 195, 205,

207, 278, 320, 348

urgency:

of decision areas, 31, 113, 155, 191, 313, 324,

347, 355

UV, uncertainties about guiding values, 9, 15, 21,

45, 55, 57, 59, 76, 77, 79, 175, 180, 181, 182,

183, 185, 189, 190, 195, 205, 207, 278,

320, 348

Valk, Arnold van der, 315

value:

clarification of, 4

judgement, 50, 65, 137, 161

van Gundy, 281

Vila, Elisenda, 345

visible/invisible products, 233–5, 237–8

visioning, 322–6

walls, paper, pens, 218–21

Wedgwood-Oppenheim, Felix, 370

White, Leroy, 364, 370

Wilcox, David, 265, 371

386

Index

Wilkin, Luc, 370

work bites, 262, 264, 265, 267

working environment, 181

working group:

advance scheduling of meetings, 230

between-session meetings, 230

conflicts of interest, 239

coordination, 226

core membership, 225

decision-making roles, 223

decision-takers, 225–6

inter-departmental/inter-organisational

membership, 238–9

intra-group roles, 223, 225

managing learning process, 229–33

mutual commitment, 239

negotiation, 226

participants, 227

permeable boundaries, 229

regular/ad hoc participants, 223

representative/stakeholder, 226–7

semi-formal, 223

sub-divisions, 227, 228

team facilitating, 231, 233

working group roles:

accountable for decisions, 223, 225–6

management of project, 223

reference, 223

representative, 223

responsible for conduct, 223

stakeholders, 223

working on project, 223

working shortlist, 51, 52, 69, 147

applying constraints in choice of, 163–5

workshop, strategic choice, 110, 112, 113, 115,

117, 145, 173, 174, 175, 177, 203, 204, 205,

208, 209, 247, 249, 251, 253, 258, 266, 267,

273, 293, 296, 298, 299, 301, 309, 320, 323,

324, 325, 327, 329, 330, 337, 341–4, 346,

348, 350, 353, 354, 355–9, 363

Yewlett, Christopher, 161, 239, 370

387

planning under pressure, third edition (urban and regional planning series)

Documents