1 IntroductionAround the world, new spaces and opportunities areemerging for citizen engagement in policy processes,from local to global levels. Policy instruments, legalframeworks and support programmes for promotingthem abound. Yet, despite the widespread rhetoricalacceptance, it is also becoming clear that simplycreating new institutional arrangements will notnecessarily result in greater inclusion or pro-poorpolicy change. Rather, much depends on the natureof the power relations which surround and imbuethese new, potentially more democratic, spaces.

Critical questions are to be asked. Does this newterrain represent a real shift in power? Does it reallyopen up spaces where participation and citizen voicecan have an influence? Will increased engagementwithin them risk simply re-legitimating the statusquo, or will it contribute to transforming patterns ofexclusion and social injustice and to challengingpower relationships? In a world where the local andthe global are so interrelated, where patterns ofgovernance and decision making are changing soquickly, how can those seeking pro-poor changedecide where best to put their efforts and whatstrategies do they use?

Whether concerned with participation and inclusion,realising rights or changing policies, more and moredevelopment actors seeking change are alsobecoming aware of the need to engage with andunderstand this phenomenon called power. Yetsimultaneously, the nature and expressions of powerare also rapidly changing. The very spread andadoption by powerful actors of the language anddiscourse of participation and inclusion confusesboundaries of who has authority and who does not,who should be on the ‘inside’ and who is on the‘outside’ of decision-making and policymakingarenas. Changing governance arrangements, whichcall for ‘co-governance’ and ‘participatory

governance’ challenge our traditional categories ofthe rulers and the ruled, the policymakers and thepublic. The use of terms such as ‘partnership’ and‘shared ownership’ by large, powerful actors like theWorld Bank and the International Monetary Fund(IMF) invite engagement on a ‘level playing field’ butobscures inequalities of resources and power. Theadoption by multinational corporate actors ofnotions of ‘corporate citizenship’, blurs traditional ‘us’and ‘them’ distinctions between economic powerholders and those who might negatively be affectedby their corporate practices. And in the midst of allof this changing language and discourse, rapidprocesses of globalisation challenge ideas of‘community’ and the ‘nation-state’, reconfiguring thespatial dynamics of power, and changing theassumptions about the entry points for citizenaction.

All of these changes point to the need for activists,researchers, policymakers and donors who areconcerned about development and change to turnour attention to how to analyse and understand thechanging configurations of power. If we want tochange power relationships, e.g. to make them moreinclusive, just or pro-poor, we must understand moreabout where and how to engage. This article sharesone approach to power analysis; an approach whichhas come to be known as the ‘power cube’ andprovides some reflections and examples of how thisapproach has been applied in differing contexts.

2 Reflecting on power analysisThough everyone possesses and is affected by power,the meanings of power – and how to understand it– are diverse and often contentious (as the articles inthis IDS Bulletin illustrate). Some see power as heldby actors, some of whom are powerful while othersare relatively powerless. Others see it as morepervasive, embodied in a web of relationships anddiscourses which affect everyone, but which no

Finding the Spaces for Change:A Power Analysis

John Gaventa*

IDS Bulletin Volume 37 Number 6 November 2006 © Institute of Development Studies

23

single actor holds. Some see power as a ‘zero-sum’concept – to gain power for one set of actors meansthat others must give up some power. Since rarelydo the powerful give up their power easily, this ofteninvolves conflict and ‘power struggles’. Others seepower as more fluid and accumulative. Power is nota finite resource; it can be used, shared or created byactors and their networks in many multiple ways.Some see power as a ‘negative’ trait – to holdpower is to exercise control over others. Others seepower to be about capacity and agency to bewielded for positive action.

Power is often used with other descriptive words.Power ‘over’ refers to the ability of the powerful toaffect the actions and thought of the powerless. Thepower ‘to’ is important for the capacity to act; toexercise agency and to realise the potential of rights,citizenship or voice. Power ‘within’ often refers togaining the sense of self-identity, confidence andawareness that is a precondition for action. Power‘with’ refers to the synergy which can emerge throughpartnerships and collaboration with others, or throughprocesses of collective action and alliance building.1

My own view of power was shaped by my ownhistory of engaging with power relations in aparticular context. As a young graduate in politicalscience, I began working with grassroots citizens in aremote mining valley of one of the poorest parts ofthe USA in their efforts to claim political, economicand social rights vis-à-vis government and a London-based corporate mine owner. The conventional viewsof democracy and power in the USA which I hadlearned in my studies failed to explain the reality Iencountered. Though violations of democratic rights,enormous inequalities in wealth and appallingenvironmental living conditions were to be foundeverywhere, there was little visible conflict or actionfor change.

There was something about power which had lednot only to defeat where voices had been raised, butalso, somehow, over time, the voices had beensilenced altogether.2 Much of my work then shiftedto how citizens recovered a sense of their capacity toact, and how they mobilised to get their issues heardand responded to in the public agenda. For almost20 years, from the mid-1970s to the mid-1990s,while also teaching and researching at the Universityof Tennessee, I was practically engaged with a non-governmental organisation (NGO) working for

grassroots empowerment, the Highlander Centerbased in southern USA. Much of our approachinvolved finding ways to strengthen the capacity ofordinary citizens and to analyse and challenge theinequalities of power which affected their lives.

After joining the Institute for Development Studies(IDS) in the mid-1990s, I continued to work onprocesses of citizen participation and engagement inother parts of the world. In the internationaldevelopment field, I discovered a host of approachesfor participation in research and learning, advocacyand community mobilisation, poverty assessmentsand policy processes, local governance anddecentralisation, and rights-based and citizenship-building approaches. At the same time, with theirincreasing acceptance in mainstream developmentdiscourse, many of these approaches riskedbecoming techniques which did not pay sufficientattention to the power relations within andsurrounding their use. Increasingly, the work of theParticipation Group at IDS and many of ourassociates began to look for approaches which putan understanding of power back in the centre of ourunderstanding of the concepts and practices ofparticipation.

My own work focused mainly on the intersection ofpower with processes of citizen engagement ingovernance at the local, national and global levels.Work with Anne Marie Goetz asked questions aboutthe most important spaces in which citizens couldeffectively engage, and how to move citizen voicefrom access, to presence, to influence (Goetz andGaventa 2001). Work with other colleaguesexamined how citizens participated in policy spacessurrounding poverty reduction, and concluded with acall for moving from ‘from policy to power’ (Brock etal. 2004). Through the Development ResearchCentre on Citizenship, Participation andAccountability, I worked and learned with a researchteam, led by Andrea Cornwall and Vera Coehlo,which was examining the spaces and dynamics ofcitizen participation (Cornwall and Coehlo 2004;2006). Some work, through LogoLink,3 focused oncitizen participation at the local level. Other workfocused on global citizen action (Edwards andGaventa 2001). In all of these areas, the issues ofpower and its links with processes of citizenengagement, participation and deepening forms ofdemocracy were always lurking somewhere close tothe surface.

24 Gaventa Finding the Spaces for Change: A Power Analysis

Increasingly, we began to search for approacheswhich could make the implicit power perspectivemore explicit, and which would help to examine theinterrelationships of the forms of power which wewere encountering in different political spaces andsettings. Building on my previous work based on the‘three dimensions’ of power developed by StevenLukes (Lukes 1974; Gaventa 1980), I began to arguethat Luke’s three forms of power must also beunderstood in relation to how spaces forengagement are created, and the levels of power(from local to global), in which they occur.Understanding each of these – the spaces, levels andforms of power – as themselves separate butinterrelated dimensions, each of which had at leastthree components within them, these dimensionscould be visually linked together into a ‘power cube’(Figure 1). By using this framework, I argued, wecould begin to assess the possibilities oftransformative action in various political spaces.Moreover, the approach could be a tool forreflection by activists and practitioners to map thetypes of power which we sought to challenge, andto look at the strategies for doing so.

While some thought the ‘cube’ image (Figure 1) riskedbeing a bit too static in its portrayal of power, formany practitioners, the approach seemed to havesome resonance. We have used it with donor agenciesas a tool for reflecting on the strategies they use

within developing countries, and to encourage self-reflection on the power which they as donor agenciesexercise (Development Research Centre 2003). I haveshared it in a workshop on political capacity buildingwith NGOs in Indonesia, especially to analyse andreflect on the ways in which they move from workingfor strengthening local participation, to engaging atthe more national level. With my colleagues at JustAssociates – who themselves have long experimentedwith popular education approaches to power analysis– the approach was also used at an internationalworkshop with popular educators, campaigners anddevelopment staff from trade unions and internationalNGOs to discuss how to build links between localknowledge and mobilisation and broader internationaladvocacy work, in order to challenge global economicpower (Just Associates 2006).4 Most extensively, theframework contributed to an evaluation on ‘AssessingCivil Society Participation as supported In-Country byCordaid, Hivos, Novib and Plan Netherlands1999–2004’, which included applications of theapproach in Colombia, Guinea, Guatemala, Ugandaand Sri Lanka (Guijt 2005; Gaventa 2005).

In this article, I elaborate further on the differentsides, or dimensions of the cube, and then examinetheir interrelationships. I conclude by sharing furtherexamples of how this approach has been applied fora critical understanding of power.

IDS Bulletin Volume 37 Number 6 November 2006 25

Figure 1 The ‘power cube': the levels, spaces and forms of power

FORMS

LEVELS

SPACES

Invisible

VisibleHiddenClosed

InvitedClaimed/

Created

Global

National

Local

3 Understanding the spaces, places and forms ofpowerAs mentioned above, the power cube is aframework for analysing the spaces, places and formsof power and their interrelationship. Though visuallypresented as a cube, it is important to think abouteach side of the cube as a dimension or set ofrelationships, not as a fixed or static set ofcategories. Like a Rubik’s cube,5 the blocks within thecube can be rotated – any of the blocks or sides maybe used as the first point of analysis, but eachdimension is linked to the other. In this presentation,we begin with the dimension of spaces, then moveto the levels of power, then conclude by returning tohow this relates to the three forms of power earlierdeveloped by Lukes.

3.1 The spaces for participation6

The notion of ‘space’ is widely used across theliteratures on power, policy, democracy and citizenaction. Some writers refer to ‘political spaces’ asthose institutional channels, political discourses andsocial and political practices through which the poorand those organisations working with them canpursue poverty reduction (Webster and Engberg-Petersen 2002). Other work focuses on ‘policyspaces’ to examine the moments and opportunitieswhere citizens and policymakers come together, aswell as ‘actual observable opportunities, behaviours,actions and interactions … sometimes signifyingtransformative potential’ (McGee 2004: 16). Otherwork examines ‘democratic spaces’ in which citizenscan engage to claim citizenship and affectgovernance processes (Cornwall and Coehlo 2006).In this article, which takes citizen action andparticipation as its starting point, ‘spaces’ are seen asopportunities, moments and channels where citizenscan act to potentially affect policies, discourses,decisions and relationships that affect their lives andinterests.

As Andrea Cornwall’s work reminds us, these spacesfor participation are not neutral, but are themselvesshaped by power relations, which both surround andenter them (Cornwall 2002). Among others, shedraws upon French social theorists (Lefebvre, Foucault,Bourdieu) for whom the concept of power and theconcept of space are deeply linked. Quoting Lefebvre:‘Space is a social product … it is not simply “there”, aneutral container waiting to be filled, but is a dynamic,humanly constructed means of control, and hence ofdomination, of power’ (Lefebvre 1991: 24).

Inherent also in the idea of spaces and places is alsothe imagery of ‘boundary’. Power relations help toshape the boundaries of participatory spaces, what ispossible within them, and who may enter, with whichidentities, discourses and interests. Using the idea ofboundary from Foucault and others, Haywardsuggests that we might understand power ‘as thenetwork of social boundaries that delimit fields ofpossible action’. Freedom, on the other hand, ‘is thecapacity to participate effectively in shaping the sociallimits that define what is possible’ (Hayward 1998: 2).In this sense, participation as freedom is not only theright to participate effectively in a given space, butthe right to define and to shape that space.

So one dynamic we must explore in examining thespaces for participation is to ask how they werecreated, and with whose interests and what terms ofengagement. While there is much debate on theappropriate terminology for these spaces, our workseems to suggest a continuum of spaces, whichinclude:7

Closed spaces. Though we want to focus on spacesand places as they open up possibilities forparticipation, we must realise that still manydecision-making spaces are closed. That is,decisions are made by a set of actors behindclosed doors, without any pretence of broadeningthe boundaries for inclusion. Within the state,another way of conceiving these spaces is as‘provided’ spaces in the sense that elites (be theybureaucrats, experts or elected representatives)make decisions and provide services to ‘thepeople’, without the need for broaderconsultation or involvement. Many civil societyefforts focus on opening up such spaces throughgreater public involvement, transparency oraccountability.

Invited spaces. As efforts are made to widenparticipation, to move from closed spaces to more‘open’ ones, new spaces are created which may bereferred to as ‘invited’ spaces, i.e. ‘those into whichpeople (as users, citizens or beneficiaries) areinvited to participate by various kinds ofauthorities, be they government, supranationalagencies or non-governmental organisations’(Cornwall 2002). Invited spaces may be regularised,that is they are institutionalised ongoing, or moretransient, through one-off forms of consultation.Increasingly with the rise of approaches to

26 Gaventa Finding the Spaces for Change: A Power Analysis

participatory governance, these spaces are seen atevery level, from local government, to nationalpolicy and even in global policy forums.

Claimed/created spaces. Finally, there are the spaceswhich are claimed by less powerful actors from oragainst the power holders, or created moreautonomously by them. Cornwall refers to thesespaces as ‘organic’ spaces which emerge ‘out ofsets of common concerns or identifications’ and‘may come into being as a result of popularmobilisation, such as around identity or issue-based concerns, or may consist of spaces in whichlike-minded people join together in commonpursuits’ (Cornwall 2002). Other work talks ofthese spaces as ‘third spaces’ where social actorsreject hegemonic space and create spaces forthemselves (Soja 1996). These spaces range fromones created by social movements and communityassociations, to those simply involving naturalplaces where people gather to debate, discuss andresist, outside of the institutionalised policy arenas.

These are not the only possible spaces – the criticalkinds of spaces for engagement will vary acrosscontext and historical setting. In various applicationsand uses, many other relevant terminologies havebeen added to this continuum, such as ‘conquered’,‘instigated’, or ‘initiated’ spaces. In work with civilsociety actors in Colombia, Pearce and Vela (2005)identified a continuum of spaces which included:

formal by invitation (participation is officiallyoffered in some way)formal by right (participation is mandated orlegislated)created by non-state institutions (e.g. by church,parties, donors)created by civil society organisations (CSOs) (e.g.by NGOs or grassroots organisations)collective transitory action (such as protests orland occupations).

Whatever the terminology, critical though it is whocreates the space – those who create it are morelikely to have power within it, and those who havepower in one, may not have so much in another.

We must also remember that these spaces exist indynamic relationship to one another, and areconstantly opening and closing through struggles forlegitimacy and resistance, co-optation and

transformation. Closed spaces may seek to restorelegitimacy by creating invited spaces; similarly, invitedspaces may be created from the other direction, asmore autonomous people’s movements attempt touse their own fora for engagement with the state.Similarly, power gained in one space, through newskills, capacity and experiences, can be used to enterand affect other spaces. From this perspective, thetransformative potential of spaces for participatorygovernance must always be assessed in relationshipto the other spaces which surround them. Creationof new institutional designs of participatorygovernance, in the absence of other participatoryspaces which serve to provide and sustaincountervailing power, might simply be captured bythe already empowered elite.

The interrelationships of the spaces also createchallenges for civil society strategies of engagement.To challenge ‘closed’ spaces, civil societyorganisations may serve the role of advocates,arguing for greater transparency, more democraticstructures, or greater forms of public accountability.As new ‘invited’ spaces emerge, civil societyorganisations may need other strategies of how tonegotiate and collaborate ‘at the table’, which mayrequire shifting from more confrontational advocacymethods. At the same time, research shows that‘invited spaces’ must be held open by ongoingdemands of social movements, and that moreautonomous spaces of participation are importantfor new demands to develop and to grow. Spanningthese spaces – each of which involves different skills,strategies and resources – is a challenge. In reality,civil society organisations must have the ‘stayingpower’ (Pearce and Vela) to move in and out ofthem over time, or the capacity to build effectivehorizontal alliances that link strategies across thevarious spaces for change.

3.2 Places and levels for participationThe concern with how and by whom the spaces forparticipation are shaped intersects as well withdebates on the places, or levels where critical social,political and economic power resides. While somework on power (especially that on gender andpower) starts with an analysis of power in moreprivate or ‘intimate’ spaces, much of the work onpublic spaces for participation involves the contestbetween local, national and global arenas aslocations of power.8 There are some that argue thatparticipatory practice must begin locally, as it is in the

IDS Bulletin Volume 37 Number 6 November 2006 27

arenas of everyday life in which people are able toresist power and to construct their own voice. Thereare others who argue that power is shifting to moreglobalised actors, and struggles for participation mustengage at that level. In between, there are debateson the role of the nation state, and how it mediatespower; on how the possibilities of local spaces oftendepend on the extent to which power is legitimatednationally, but shared with the locality. A great dealof work in the area of decentralisation, for instance,discusses the dynamics of power between thelocality and the nation state, while other literatureargues for the importance of community- orneighbourhood-based associations as key locationsfor building power ‘from below’.

However, a growing body of literature warns us ofthe dangers of focusing only on the ‘local’, or the‘national’ in a globalising world. Globalisation, it isargued, is shifting traditional understandings of wherepower resides and how it is exercised, transformingtraditional assumptions of how and where citizensmobilise to hold states and non-state actors toaccount (Tarrow 2005; Batliwala and Brown 2006).Concerns with global governance are producing newextra-national fora in which citizens might be seekingto engage, such as NEPAD (The New Partnership forAfrica’s Development) or the African Union.Moreover, rather than being separate spheres, thelocal, national and global are increasingly interrelated.Local forms and manifestations of power areconstantly being shaped in relationship to globalactors and forces, and in turn, local action affects andshapes global power. Local actors may use globalforums as arenas for action (e.g. Narmada Dam;Chiapas) just as effectively – or more effectively –than they can appeal to institutions of localgovernance. Conversely, expressions of global civilsociety or citizenship may simply be vacuous withoutmeaningful links to local actors and local knowledge.

As in the example of the spaces of participation, thisvertical dimension of the places of participationshould also be seen as a flexible, adaptablecontinuum, not as a fixed set of categories. As in thetypes of spaces, the relevance and importance oflevels and places for engagement varies according tothe purpose of differing civil society organisationsand interventions, the openings that are beingcreated in any given context, etc. In work inColombia, for instance, civil society organisationsidentified eight different levels of civil society

engagement in the public sphere, each of which hasits own types of spaces, including the international,national, departmental, regional/provincial, municipal,communal and neighbourhood levels (Pearce andVela 2005). Many of these are shaped by the relevantlegal frameworks of governmental administration,and may differ across rural and urban communities,yet increasingly, extra-local arenas seem to grow ascentres of power and decision making.

For civil society, the changing local, national andregional levels of power pose challenges for whereand how to engage. Some focus at the global level,waging campaigns to open the closed spaces ofgroups like the World Trade Organization (WTO).Other focus more on challenging economic powerlocally. Yet, the interrelationships of these levels ofpower with one another suggest that the challengefor action is not only how to build participatoryaction at differing levels, but how to promote thedemocratic and accountable vertical links across actorsat each level. As Pieterse (1997) puts it, ‘this involves adouble movement, from local reform upward andfrom global reform downward – each level ofgovernance, from the local to the global, plays acontributing part’ (quoted in Mohan and Stokke2000: 263). At the same time, a growing concernamong civil society organisations has to do with thelack of such vertical links between those organisationsdoing advocacy at an international level, often led orsupported by international NGOs, with thoseworking to build social movements or alternativestrategies for change at the more local levels(Batliwala 2002).

This subject was pursued further in a 2005 workshopat IDS on ‘Citizen Action, Knowledge and GlobalEconomic Power’, where the power cube approachwas to used to reflect and analyse the different kindsof civil society initiatives going on at different levels.In doing so, the disconnections between those whoare speaking at the global level and those who areexperiencing problems of poverty or economicinjustice at the local level were highlighted. Theworkshop then focused on ways of overcoming thedisconnections.

3.3 The forms and visibility of power across spacesand placesAs we examine the relationships of place and spacevis-à-vis participation, we must also examine thedynamics of power that shape the inclusiveness of

28 Gaventa Finding the Spaces for Change: A Power Analysis

participation within each. Here, much of theliterature of power is concerned with the degree towhich conflict over key issues and the voices of keyactors are visible in given spaces and places. In earlierwork, building on work by Lukes (Lukes 1974;Gaventa 1980) I explored the differences between:

more pluralist approaches to power, in whichcontests over interests are assumed to be visiblein public spaces, which in turn are presumed to berelatively opena second form of power, in which the entry ofcertain interests and actors into public spaces isprivileged over others through a prevailing‘mobilisation of bias’ or rules of the game; anda third form of power, in which conflict is moreinvisible, through internalisation of powerlessness,or through dominating ideologies, values andforms of behaviour.

In more recent work which in turn builds upon thisapproach, VeneKlasen and Miller argue more simply

for distinguishing between the visible, hidden andinvisible (or internalised) forms of power (see Box 1).

The importance of this for how we analyse thedynamics of participation in differing spaces andplaces is relatively obvious. Historically, many pluraliststudies of power have mainly examined power in itsvisible manifestations. One looked at whoparticipated, who benefited and who lost in order tosee who had power. But as we have seen, power inrelationship to place and space also works to putboundaries on participation, and to exclude certainactors or views from entering the arenas forparticipation in the first place. Or power, in its moreinsidious forms, may be internalised in terms of one’svalues, self-esteem and identities, such that voices invisible places are but echoes of what the powerholders who shaped those places want to hear. Suchpower analysis points again to the importance ofestablishing the preconditions of participation inorder for new institutional spaces to lead to changein the status quo. Without prior awareness building

IDS Bulletin Volume 37 Number 6 November 2006 29

Box 1 Forms of power

Visible power: observable decision makingThis level includes the visible and definable aspects of political power – the formal rules, structures,authorities, institutions and procedures of decision making … Strategies that target this level areusually trying to change the ‘who, how and what’ of policymaking so that the policy process is moredemocratic and accountable, and serves the needs and rights of people and the survival of the planet.

Hidden power: setting the political agendaCertain powerful people and institutions maintain their influence by controlling who gets to thedecision-making table and what gets on the agenda. These dynamics operate on many levels toexclude and devalue the concerns and representation of other less powerful groups … Empoweringadvocacy strategies that focus on strengthening organisations and movements of the poor can buildthe collective power of numbers and new leadership to influence the way the political agenda isshaped and increase the visibility and legitimacy of their issues, voice and demands.

Invisible power: shaping meaning and what is acceptableProbably the most insidious of the three dimensions of power, invisible power shapes thepsychological and ideological boundaries of participation. Significant problems and issues are not onlykept from the decision-making table, but also from the minds and consciousness of the differentplayers involved, even those directly affected by the problem. By influencing how individuals thinkabout their place in the world, this level of power shapes people’s beliefs, sense of self andacceptance of the status quo – even their own superiority or inferiority. Processes of socialisation,culture and ideology perpetuate exclusion and inequality by defining what is normal, acceptable andsafe. Change strategies in this area target social and political culture as well as individual consciousnessto transform the way people perceive themselves and those around them, and how they envisagefuture possibilities and alternatives.

Adapted by Just Associates from VeneKlasen and Miller (2002).

so that citizens possess a sense of their own right toclaim rights or express voice, and without strongcapacities for exercising countervailing power againstthe ‘rules of the game’ that favour entrenchedinterests, new mechanisms for participation may becaptured by prevailing interests.

As in the other dimensions of the power cube, themultiple forms of power also pose challenges for civilsociety actors trying to change power relations. Somegroups may focus on advocacy approaches, challengingthe visible forms of power in visible arenas throughpublic debate, informed research and working toinfluence public representatives. Others may focus onmobilising and collective action strategies, which workto challenge barriers which prevent certain actors andforms of knowledge from entering public arenas inthe first place. Yet, others may focus more onchanging the invisible, internalised forms of power,through awareness and consciousness-buildingcampaigns. While often these are different strategiesinvolving different organisations and interventions tochange power, in fact strategies are also needed whichlink across them. For instance, a policy victory in thevisible arena of power may be important, but may notbe sustained, if those outside the arena are not awarethat it has occurred and how it relates to theirinterests, or are not mobilised to make sure that otherhidden forms of power do not preclude itsimplementation.

4 The interrelationship of the spaces, places andforms of powerAs argued above, the dynamics of power dependvery much on the type of space in which it is found,the level at which it operates and the form it takes.Moreover, as has been suggested, along eachdimension, any sustained and effective changestrategy must concern itself with how to build andsustain effective change across the full continuum.Transformative, fundamental change happens, Isuggest, in those rare moments when socialmovements or social actors are able to workeffectively across each of the dimensionssimultaneously, i.e. when they are able to link thedemands for opening previously closed spaces withpeople’s action in their own spaces; to span acrosslocal and global action, and to challenge visible,hidden and invisible power simultaneously. Just likethe Rubik’s cube, successful change is about gettingeach of the pieces on each dimension of the cube toalign with each other, simultaneously.

‘Alignment’ of strategies for change is a hugechallenge, both across each of the dimensions of thecube, but also made more difficult by theirinteraction. For instance, along the spaces dimension,while many groups seeking action work either onopening closed spaces through demanding moretransparency or supporting internal reform, or onbuilding social movements and mobilisation inclaimed space, much research suggests that it iseffective when horizontal alliances are built acrossthese spaces that real change occurs. Similarly,advocacy and change strategies must often buildvertical alliances across local, national and globallevels to make sure that changes are meaningful ateach level. And, those seeking not only to influencepolicies in the public arena, but also to changepower relations more fundamentally, mustsimultaneously think about winning the issue,mobilising to broaden the political space, andbuilding awareness of those who are excluded.Rather than any single strategy, an ensemble ofstrategies, which work together and not againsteach other, are required to fully challenge these setsof power relationships.

Moreover, while it is difficult enough for thoseseeking change to work across the range of anysingle dimension of the cube, in fact the dimensionsare simultaneously interacting to affect the other.Strategies for alignment along one axis maycontribute to misalignment on another. Those whostudy the Rubik’s cube argue that there are literallybillions of different positions9 that the blocks of thecube may have, illustrating the complexity andpermutations in which power can take across space,place and form in any given context. The local,national and global agenda affects the opening andclosure of invited spaces; the visibility of power isshaped by who creates the space; in turn priorparticipatory experiences which have helped toovercome forms of invisible and hidden power, maystrengthen the possibilities for success of newinstitutional designs for participation.

For any given issue or action, there is no singlestrategy or entry point. Much depends on navigatingthe intersection of the relationships, which in turncan either contribute to new misalignments anddistortions of power, or simultaneously creates newboundaries of possibility for strategic action. Forinstance, linking local–national–global campaigns toopen up previously closed spaces may be important,

30 Gaventa Finding the Spaces for Change: A Power Analysis

IDS Bulletin Volume 37 Number 6 November 2006 31

but in so doing, they may re-enforce forms of hiddenand invisible power, if they simultaneously excludecertain potential actors or forms of knowledge. Onthe other hand, the opening of previously closedlocal spaces can contribute to new mobilisations andconscientisation, which may have the potential toopen other spaces more widely, and to createmomentum for change at national or global levels.Like a Rubik’s cube, sometimes the dimensions mayappear chaotic, random and confused; at other timesthey may appear as if alignments will be possible.

In fact, those who study the Rubik’s cube also tell usthat any cube can be re-aligned in a maximum of 29moves, taken in the right sequence. While notwanting to reduce social change to a formulaicsolution, this does suggest that those seeking tochallenge power in all of its spaces, levels and formsneed to search not for one solution, but to buildmultiple, linked strategies and in different sequences,depending on the starting point in any given context.The challenge is to understand what these strategiesmight be, and how they can be linked to realign allof the dimensions of power. That is whentransformative change might really occur.

The power cube approach does not do this for us. Itis not a checklist, for example the idea is not tocheck off each box, as the significance, dynamics andinterrelationships of the dimensions are constantlychanging and vary enormously from one context toanother. Nor is it prescriptive, valorising local,claimed spaces over the distant global, closed spaces,for it suggests that each has forms of visible, hiddenand invisible power within them. Rather, the powercube is meant to be more of an analytical device,which can be used – along with other approaches –to reflect on and analyse how strategies for changein turn change power relations.

In numerous settings, we have used the approach toencourage such reflections with development actorson their strategies for change, and indeed their ownpower and position within them. In work in Nigeria,the use of an early formation of the approach helpedus to see how struggles to open up the nationalPoverty Reduction Strategy Paper (PRSP) process,which itself came through global interventions,

contributed to challenging traditionally closed andtop-down approaches to poverty policy. At the sametime, by remaining disconnected from many localactors and from many existing movementschallenging poverty in their own spaces, the donoragenda for more participation and inclusion in thenational policy process had the inadvertent affect ofcreating a greater sense of exclusion for some otherlocal and social movement actors (Brock et al. 2004).

In using the power cube with international NGOs,reflections using the power cube approach helped tosee the need and possibilities of greaterinterconnections between global campaigning, oftendone by professional advocates working in a rapidlychanging policy environment, and local developmentwork, where strategies were more likely toencourage people to mobilise and speak forthemselves, and to challenge invisible andinternalised forms of power.

In an evaluation of the civil society participation workof Dutch international non-governmentalorganisations (INGOs), the approach was used withlocal civil society groups in Sri Lanka, Uganda andColombia to encourage local groups to reflect onthe kinds of spaces in which they were participating,the strategies they used in each and how theyinteracted (Dutch CSO). And in work with donors,the approach has been used to reflect on what kindsof change donors sought to support, and thenplacing themselves in the power cube, whatstrategies they actually engaged with, and how indoing so, they themselves became part of the powerequation (Guijt 2005).

In conclusion, even if the power cube approach doesnot tell us how to align our efforts, if it can be usedby actors seeking to change the world to reflect onwhere and how they do so, and how they workacross boundaries with others who are also workingfor change, then perhaps the alignment of effortsfor transforming power will become more possible.In this sense, reflections on power, and reflections bychange agents on how their work affects powerrelationships in all of its dimensions, is perhaps thefirst step in making more visible, power’s mosthidden and invisible forms.

32

Notes* A similar version of this article has been prepared

for the Dutch CFA evaluation, ‘Assessing CivilSociety Participation’, coordinated by Irene Guijt(2005) of Learning by Design, and supported byCordaid, Hivos, Novib and Plan Netherlands andthe Power, Participation and Change Programmeof the Participation Group at the Institute ofDevelopment Studies. My thanks to the manycolleagues from the ‘Civil Society Participation’evaluation, the Participation Group, JustAssociates, and others from whom I have learnedin using and discussing the ‘power cube’approach.

1 For further development of these debates, seeVeneKlasen and Miller (2002) and Kabeer (1994).

2 For the account of this work, see Gaventa (1980).3 See the LogoLink website for work on citizen

participation and local governance,www.ids.ac.uk/logolink/index.htm (accessed7 August 2006).

4 Workshop on ‘Citizen Action, Knowledge andGlobal Economic Power: Reflecting on CurrentPractices and Challenges Ahead, organised by JustAssociates, Action Aid and the IDS ParticipationGroup, 1–3 August 2005 (Just Associates 2006).

5 Rubik’s cube refers to a mechanical cube inventedin 1974 by the Hungarian sculptor and professorof architecture Ernö Rubik, and marketed widelyas a puzzle. The sides of the cube can rotate,though the whole remains intact.

6 The following sections draw heavily on earlierpapers on the power cube, as cited in VeneKlasenand Miller (2002) and Kabeer (1994).

7 These ideas have developed from Cornwall(2002); Brock et al. (2001) and Brock et al. (2004).

8 In this article, the ‘power cube’ focuses primarilyon power in the ‘public sphere’, while recognisingthat this approach is incomplete.

9 More specifically, about 43 quintillion, or43,252,003,274,489,856,000, according toWikipedia!

Gaventa Finding the Spaces for Change: A Power Analysis

ReferencesBatliwala, Srilatha (2002) ‘Grassroots Movements as

Transnational Actors: Implications for Global CivilSociety’, Voluntas: International Journal of Voluntaryand Nonprofit Organizations 13.4: 393–409

Batliwala, Srilatha and Brown, David L. (eds) (2006)Transnational Civil Society: An Introduction,Bloomfield: Kumarian Press

Brock, K., Cornwall, A. and Gaventa, J. (2001) Power,Knowledge and Political Spaces in the Framing ofPoverty Policy, IDS Working Paper 143, Brighton:IDS

Brock, K., McGee, R. and Gaventa, J. (eds) (2004)Unpacking Policy: Actors, Knowledge and Spaces inPoverty Reduction, Kampala: Fountain Press

Cornwall, A. (2002) Making Spaces, Changing Places:Situating Participation in Development, IDS WorkingPaper 170, Brighton: IDS

Cornwall, A. and Coehlo, V. (eds) (2006) Spaces forChange? The Politics of Citizen Participation in NewDemocratic Arenas, London: Zed Books

Cornwall, A. and Coehlo, V. (eds) (2004) ‘NewDemocratic Spaces?’, IDS Bulletin 35.2 (see alsowww.drc-citizenship.org)

Development Research Centre, Rights and PowerWorkshop Report (2003), www2.ids.ac.uk/drccitizen/docs/r&pworkshopreportfinal.pdf(accessed 7 August 2006)

Edwards, M. and Gaventa, J. (2001) Global CitizenAction, Boulder; Colorado: Lynne Reinner Publishers

Gaventa, J. (2005) ‘Reflections on the Uses of the‘Power Cube, Approach for Analyzing the Spaces,Places and Dynamics of Civil Society Participationand Engagement’, prepared for Assessing CivilSociety Participation as Supported In-Country byCordaid, Hivos, Novib and Plan Netherlands1999–2004, The Netherlands: MFP BreedNetwerk

Gaventa, J. (2004) ‘Towards Participatory Governance:Assessing the Transformative Possibilities’, inS. Hickey and G. Mohan (eds), From Tyranny toTransformation, London: Zed Books

Gaventa, J. (1980) Power and Powerlessness:Quiescence and Rebellion in an Appalachian Valley,Oxford: Clarendon Press

Goetz, A.M. and Gaventa, J. (2001) From Consultationto Influence: Bringing Citizen Voice and Client Focusinto Service Delivery, IDS Working Paper 138,Brighton: IDS

Guijt, I. (2005) ‘Synthesis Report of Dutch CFAProgramme Evaluation’, Assessing Civil SocietyParticipation as Supported In-Country by Cordaid,Hivos, Novib and Plan Netherlands 1999–2004, TheNetherlands: MFP Breed Netwerk

Hayward, C.R. (1998) ‘De-Facing Power’, Polity31.1: 1–22

IDS Bulletin Volume 37 Number 6 November 2006 33

Just Associates (2006) Citizen Engagement and GlobalEconomic Power Workshop Report, Washington DC:Just Associates

Kabeer, N. (1994) Reversed Realities: Gender Hierarchiesin Development Thought, London: Verso

Lefebvre, H. (1991) The Production of Space, London:Verso

Lukes, S. (1974) Power: A Radical View, London:Macmillan (reprinted 2004, Basingstoke: PalgraveMacmillan)

McGee, R. (2004) ‘Unpacking Policy: Actors,Knowledge and Spaces’, in K. Brock, R. McGeeand J. Gaventa (eds), Unpacking Policy: Actors,Knowledge and Spaces in Poverty Reduction,Kampala: Fountain Press: 1–26

Mohan, G. and Stokke, K. (2000) ‘ParticipatoryDevelopment and Empowerment: The Dangersof Localism’, Third World Quarterly 21.2: 247–68

Pearce, J. and Vela, G. (2005) ‘Colombia CountryReport for the Dutch CFA Programme

Evaluation’, Assessing Civil Society Participation asSupported In-Country by Cordaid, Hivos, Novib andPlan Netherlands 1999–2004, The Netherlands:MFP Breed Netwerk

Pieterse, J. (1997) ‘Globalisation and Emancipation:From Local Empowerment to Global Reform’,New Political Economy 2.1: 79–92

Soja, E. (1996) Third Space: Journeys to Los Angeles andOther Imagined Places, Cambridge, MA: Blackwell

Tarrow, S. (2005) The New Transnational Activism,Cambridge: Cambridge University Press

VeneKlasen, L. and Miller, V. (2002) A New Weave ofPeople, Power and Politics: the Action Guide forAdvocacy and Citizen Participation, Oklahoma:World Neighbors

Webster, N. and Engberg-Petersen L. (2002) (eds) Inthe Name of the Poor: Contesting Political Space forPoverty Reduction, London: Zed Books