regualtion initiative working paper series number 41...

1

REGUALTION INITIATIVE WORKING PAPER SERIES

NUMBER 41

ELECTRICITY AND TELECOMMUNICATIONS REAGULATION IN SMALL AND DEVELOPING COUNTRIES

JON STERN

LONON BUSINESS SCHOOL

2000

2

ELECTRICITY AND TELECOMMUNICATIONS REGULATION IN SMALL AND DEVELOPING COUNTRIES1

1. THE NEW UTILITY MODEL AND ITS IMPLICATIONS FOR REGULATION

Over the last 10-15 years, a new standard model has taken hold for the economic framework for the operation of utilities. The new model has as its core that utility services will be:

• provided by a set of commercialised companies;

• monopoly (eg network) elements are separated from potentially competitive elements;

• competition is actively introduced into the potentially competitive elements; and

• private capital is introduced where possible and appropriate, particularly into the competitive elements, typically with privatisation of some or all of the existing assets.

Over the last 10-15 years, this new model has largely replaced the traditional model of utility services being supplied by a state-owned vertically and horizontally integrated monopoly, supervised by the national government and typically operating in a non-commercial or semi-commercialised way. This change has been induced across developed and now across developing countries for the following two main reasons:

(i) a major reduction in the ability of national governments to finance utility investment from tax revenues2; and

(ii) a much greater emphasis on the need to improve efficiency and reduce the costs of infrastructure services (connection as well as service) coupled with (at least in developed countries) the achievement of very widespread if not universal service3.

To support the newly commercialised and privatised utilities, there have been major developments in the theory and practice of regulation. New utility regulatory institutions

1 This paper has benefited from considerable assistance in the provision and processing of data on the size and composition of regulatory agencies and on educational statistics. Particular thanks go to Martin Siner of the LBS Regulation Initiative. I am also grateful for helpful comments from David Newbery and other members of the DAE Cambridge Regulatory Seminar.

2 See Stern (1997). The first major UK privatisation of BT (British Telecom) in 1984 was largely driven by the desire of the government that BT’s major investment programme be financed from private rather than public funds.

3 Universal service was also largely achieved for electricity with widespread service in natural gas and/or district heat in the Communist countries of Central and Eastern Europe and much of the former Soviet Union. But, it was achieved and maintained only by very large-scale subsidy arrangements – mainly implicit subsidies from the state and cross-subsidies. See Stern and Davis (1998).

3

have been discussed and introduced in many countries. In addition, it has been realised that the introduction of competition (and, in particular, the development of competition over networks) has demonstrated the need for new and more complex forms of regulation to support the utility reforms.4

As yet, there seems no sign that the new utility model will lead to the withering away of regulation. Rather, infrastructure regulation has become a major growth area in developed and developing countries alike.

This raises the problem for small and low income countries of whether and how they can establish an effective regulatory capability – the problem of the supply of regulatory services given a large and growing demand for these specialised services. The purpose of this paper is, firstly, to try to establish whether or not there is a problem in the supply of resources for utility regulation in small and low income countries; secondly, to try to estimate its potential severity and, finally, to consider some potential options for how it might be tackled.

The plan of the paper is as follows. In Section 2, we outline the new utility model and its development in the telecommunications and electricity industries. In Section 3, we discuss regulatory and related requirements necessary to sustain the new utility model. In Section 4, we survey the size and high-level staffing share in (a) telecommunications and (b) electricity/energy regulatory agencies from around the world together with some illustrative statistics on the number of regulatory staff employed in regulated companies. In Section 5, we discuss the potential for finding the necessary regulatory staff using statistics stocks and flows of people with post-schooling education. In Section 6, we discuss ways of economising on the need for regulatory staff within the standard national independent regulatory agency model, including some possible alternatives to the standard national regulatory agency model. Section 7 provides a brief summary and some concluding recommendations.

The purpose of this paper is to identify the key issues better so as to be able to identify and assess possible solutions. The range of problems raised is surprisingly wide-ranging, encompassing questions of regulatory (and political) legitimacy and autonomy, the potential role of markets and contracts and other major governance issues.

At stake is the concern that unless this problem is seriously addressed and effectively resolved, the commercialised provision of infrastructure services involving private investment will simply not be sustainable in many small and/or low income countries – at least beyond the richest and most densely populated areas, as occurred in nineteenth century Britain. Ruling out the possibility of providing commercialised and privatised infrastructure services could have potentially serious adverse consequences on the economic welfare of their citizens and the competitiveness of their industries. In such cases, there may simply be no viable alternative to the traditional state-owned (or state underwritten)

4 See Newbery (1999) for a recent major survey.

4

monopoly utility model, which is the fallback model if private provision is unsustainable because of regulatory risk5.

5 See Levy and Spiller (1994) for the classic exposition of this argument. See also Cave and Stern (1998).

5

2. THE COMMERCIALISED MODEL OF UTILITY SERVICE PROVISION

2.1. Telecommunications

The new commercialised model for utility services outlined in Section 1 above has taken firmest hold in telecommunications. In developed countries, not only is there extensive competition between (and private ownership of) multiple mobile service operators and in the provision of value added network services, but competition has now spread considerably to fixed line voice and data services. Currently, only the local loop remains as a potential monopoly service and even this is rapidly becoming unbundled6 and subject to competition eg from cable and wireless based providers. As yet, though, there is still a recognised need for continued economic regulation on some issues. Examples include:

• the balance between requiring new investors to build their own network facilities relative to just allowing them to access existing facilities; and

• on universal service obligations (USOs)..

In general, the problems over access rights, access and interconnection fees and administration of USOs means that there remains a considerable need for the difficult task of regulating competition over networks. This typically involves some combination of specialist regulation and oversight from competition agencies - a pattern seen in the US and the UK, across the EU and more widely. Monitoring competition over networks and regulating network investment plans, access rights and pricing, along with USOs, has become the quintessential regulatory agenda for telecom services. However, it has also increasingly become the dominating set of issues for electricity and increasingly for natural gas, railways, etc as the new utility service model takes hold in those industries.

This new telecom model is spreading very rapidly to developing countries (a) as national telecom companies face competition from mobile and satellite services; and (b) as the cushion of very high profits from protection on overseas calls disappears. The section of the 1998 WTO agreement on telecoms plus subsequent action (eg by the US FCC) has clearly spurred a wide range of middle and low income countries to take urgent action to reform their telecommunications sector7.

A standard strategy (eg Bahamas, Slovakia, etc) is to bring-in strategic foreign partners to upgrade their local telecom incumbent in the temporary period while the national monopoly still remains. The strategic foreign partner often has a sizeable equity stake. In addition,

6 In October 2000, local loop unbundling wa made mandatory by mid-2001 in the countries of the European Union.

7 In this context, it is worth noting just how important it was to the WTO entry agreement for China with the US and the EU to obtain agreement on the conditions and timetable for the liberalisation of the Chinese telecom sector and access to the sector by foreign investors.

6

whether or not it initially has a majority equity stake, it often has managerial control rights plus options for increasing its equity stake. This allows the incumbent time to improve its efficiency and strengthen its position relative to strong external competitors before liberalisation is introduced.

In consequence, the new model for telecom service provision has spread to a wide range of smaller, developing countries. From Latin America, through the Caribbean, across Central and Eastern Europe and in much of Asia, the news is full of the liberalisation of telecom industries and the (usually partial) privatisation of national telecom companies – and in all of these countries choosing this route there is the challenge of:

(i) how to encourage and maintain competition;

(ii) how to provide the economic regulatory arrangements that will provide an acceptable balance between consumers and investors; and

(iii) how to ensure open entry and proper competition between the national incumbent telecom company (which, in many cases, still has a minority if not a majority national government equity stake) and the new operators.

2.2. Electricity

Although telecom services are undoubtedly the pathfinder, the new commercialised utility services model, it has made increasing strides into electricity and, to a lesser extent into natural gas, railways, water and sewage, roads, airports, etc.

Electricity is particularly interesting since the technical requirements of the industry imply that it is totally uneconomic (at least for the foreseeable future) that any electricity area (typically a nation state) has other than one and only one transmission and distribution network with a single centralised dispatch system8. Conversely, technical and other developments over the last two decades (particularly the development of low capital cost combined cycle gas powered generators) have led to a dramatic reduction in the economies of scale in generation and supply (sales) of electricity9.

The result of this is that developed countries have increasingly unbundled their electricity industries, separating generation from transmission, or at least separating generation from dispatch via the introduction of an Independent System Operator (ISO). In addition, competition is being increasingly developed in generation and also in wholesale sales (supply) of electricity and sometimes in retail sales, particularly to large industrial

8 The spread of small-scale distributed generation may, in time, greatly reduce the need for extensive transmission systems with complex dispatch systems, markets and pools but not for the foreseeable future.

9 See Hunt and Shuttleworth (1996). See Posner (1993) for an illuminating discussion of the role of CCGTs on the economics of generation.

7

consumers. Privately owned companies usually dominate generation in such systems and are increasingly important in supply and distribution, albeit not as often in transmission.

Probably the most frequently cited example of such changes is that of England and Wales. But, variants of this basic model have been developed in Norway, New Zealand and are spreading across the US (viz California and New England) and are developing across the EU, fostered by the development of markets in the light of the 1996 EU Electricity Directive.

More importantly, such models have been implemented successfully in some middle income countries and are now developing more widely. The classic examples are in Latin America eg Chile (starting in 1978-79) and Argentina (staring in 1992). More recently, some low income countries have increasingly introduced electricity industry structures with some competition in generation and often with some degree of competition in supply either to distribution companies and/or to large industrial consumers. This trend has spread with similar reforms being undertaken or planned not just through Latin America (eg Bolivia, Ecuador) but also in Asia (eg Thailand, China) and now starting in Africa (eg Senegal, Uganda and Nigeria).

In some cases, (particularly Central and Eastern Europe, China and Africa) transmission has been separated from generation but supply competition is restricted to monopsony purchase by a single buyer who on-sells via a bulk supply tariff to distribution companies. This model has many advantages as a transitional model in countries (a) where distribution companies are not commercialised and/or financially weak10; and (b) where some consumers (typically households and/or small farmers) pay prices for electricity that are substantially below operating costs11. However, the model also has some important disadvantages, not least regarding: (a) the payments risk imposed on the transmission company and (b) the stringent regulatory requirements necessary for efficient operation and investment.12

The spread of electricity restructuring means that it is now increasingly hard to find vertically integrated electricity supply industries, let alone countries where the industry is privatised as a monopoly. Even some very small countries (eg Estonia and Latvia, neither of which has more than 3 million people) are now privatising their electricity industries with some degree of unbundling.

Some countries continue to combine generation and transmission, sometimes with distribution (eg France, Malaysia), sometimes without (eg Italy and Germany). However, even France, the bastion of the state-owned public service monopoly utility is now

10 China, Indonesia and some other Asian economies are in this category. 11 India, many of the countries of Central and Eastern Europe, Russia and the CIS and some African countries are in

this category. 12 The regulatory difficulties with a single buyer model can, in some circumstances, become acute. They seem to be

most serious for state-owned single-buyer companies operating in countries with seriously imperfect markets and governance and/or widescale corruption. Many CIS countries would fall into this category as would eg Indonesia and some African countries.

8

establishing a separately managed ISO as its chosen way of meeting the requirements of the 1996 EU Electricity Directive and has rejected the pure single buyer approach offered in the Directive.13

In developing countries, the Asian IPP model was a major competitor to the unbundled model outlined above. This model covers the situation where the incumbent power company, which owned and managed transmission and dispatch, purchases power on contracts from a small number of independent power producers (IPPs). But, the IPPs compete with the generating plant owned by the incumbent power company, which comprises the bulk of the generating capacity available in the country. This was the pattern in Indonesia, Thailand, Philippines, Malaysia, Pakistan etc.

One point about the Asian IPP model is that it was intended to, and apparently succeeded in economising on regulation. The economic regulation of prices was supposedly handled in the ($-denominated) power purchase agreements (PPAs) between the IPP and the incumbent. In practice, this was wishful thinking. The attempt at regulation by contract produced a crisis of regime when the contracts became unsustainable following the onset of the Asian financial crisis in 1998. The problems arose for two main reasons. Firstly, there was no explicit regulation of anything other than generation prices; and, secondly, there were no procedures in place for handling major shocks. A properly designed and managed regulatory system with an independent regulatory agency would have provided for both of these problems.

The Asian IPP model therefore could not cope with the pressures arising from the local exchange rate devaluations induced by the 1998 Asian financial crisis. Several countries in the region (notably Philippines, Thailand and Indonesia) are now pursuing power sector reforms which will unbundle the generation owned by the incumbent and privatise it. Not co-incidentally, they are also developing new and independent regulatory agencies to support (and push forward) the restructuring.

2.2.1. Implications

The key point from the discussion above is that the new utility model has now sufficiently spread to electricity (a) for large numbers of rich countries and (b) to increasing numbers of developing countries. Hence across the world there is a growing need to establish regulatory arrangements that can manage and regulate the restructured industries and, increasingly, manage and regulate competition across networks.

13 In October 2000, a senior EdF spokesman spoke publicly to the effect that EdF would welcome full liberalisation of supply in France. See FT Power in Europe.

9

For electricity, the technical issues are particularly difficult since:

• the physical flows of electricity must be in balance at every instant so that supply must immediately respond fully to unforeseeable and unforeseen changes in demand;

• there are no currently economic possibilities of competing networks (as in telecoms or, to a lesser extent in natural gas);

• there is no possibility of significant storage (unlike natural gas).

Hence, the regulatory requirements to provide economic signals to ensure continued and efficient network operation and pricing as well as continuous effective operation are very demanding. The economic regulation of decentralised and usually wholly or partially privatised electricity industries appears to have increased rather than reduced both the numbers of people involved in regulation and certainly the complexity and skill levels of the staff involved in regulation. In addition, as we shall demonstrate, this seems to apply at least as much to the regulatory requirements in the regulated companies as to staff in the national regulator.

The difficulties of recruiting and retaining staff with the skills and ability to carry out such regulation are by no means trivial even in the UK. In low income and/or small countries, as we will demonstrate in later sections, the difficulties of obtaining sufficient regulatory staff can be sufficiently acute to threaten the possibility of achieving sustainable unbundling and privatisation of electricity and other utilities.

10

3. REGULATORY AND RELATED INSTITUTIONAL AND STAFFING REQUIREMENTS

In telecoms, the notion that, in some countries, specialist regulation could be replaced by standard competition policy in the reasonably near future is relatively mainstream, but by no means universal. There is, though, much disagreement even among the proponents about the timescale involved let alone the definition of “reasonably near”.

For electricity, the notion that the need for specialist regulation can be dispensed with is not unknown), but is highly unusual14. The mainstream view (which I share) is that specialist regulation will continue to be required for the foreseeable future. Indeed, the EU is having debates about how best to co-ordinate across individual member states’ regulatory systems and the possible need for an EU-level regulator. The latter is unlikely but not impossible to envisage.

The success of the Argentinian and Chilean electricity restructurings is due, in no small part, to the performance of their independent regulators. Similarly, the Regulatory Commission in the Indian state of Orissa has been crucial to the relative success of their ESI reform and privatisation of distribution. These regulatory systems may have their imperfections, but they have played a key role in the initial implementation and the development of the reforms15.

Effective regulation in electricity requires not just substantial numbers of staff, it requires substantial numbers of staff with particular and scare specialist skills eg economists, lawyers, accountants, financial analysts as well as engineers. These services are also needed in the regulated companies. It remains a very open question as to whether poorer and particularly small countries are able to find the necessary numbers of people with these scarce skills available to staff regulatory institutions and to run the new companies and to provide for a policy capacity in the relevant Ministries.

Note that the problem different from finding a Regulator or 3-5 Regulatory Commissioners. The latter have to have relevant experience and weight but do not need to be up-to-date regulatory experts or industry experts. The staff in the regulatory office or commission do need to have such expertise. Further, it is the regulatory staff who provide the critical institutional continuity that provides for the development and responsiveness of the regulatory system.

For developing countries and the countries of Central and Eastern Europe and the CIS, the issue of the supply of sufficient regulatory experts can be critical as to whether regulation

14 Professor Stephen Littlechild, the first UK Electricity regulator is associated with this view. See Littlechild (1983). He has recently re-expressed this view in public, on-the-record discussions.

15 For an appraisal of the Argentine and Chilean electricity reforms, see Spiller and Martorell in Gilbert and Kahn (eds) 1996.

11

can, in practice, effectively be separated from policy and company management. If it cannot be so separated, the question then arises as to the sustainability of unbundled, commercialised and privatised provision of telecom, electricity and other infrastructure services.

The key point to note is that, whatever its other weaknesses, the traditional state-owned monopoly utility minimised the need for specialist regulatory expertise. This was a great advantage for many developing and Transition economies where the supply of these skills is very limited.

It cannot be emphasised strongly enough that the problem of ensuring an adequate supply of regulatory staff is, in practice, critical for establishing effective utility reform in developing and Transitional economies. In my experience it is also a difficult and very much underrated problem. Satisfying the demand for regulation (and regulatory skills) is a non-trivial problem in developed economies like the UK. In developing and Transition economies, there is a massive potential excess demand for these skills and it is very unclear whether and how the supply can be increased sufficiently.

3.1. Micro-Economic Institutional Requirements for a 21st Century Nation State

In terms of micro-economic affairs best–practice modern nation state is advised to have at least the following institutions. Each of these needs to have (a) an adequate legal backing in statute law; (b) an effective head group leadership; and (c) adequate staffing resources in terms of staff numbers and skills.

• Ministries with effective policy-making, monitoring and enforcement resources, operating within a clear budgetary framework (usually supplied by a strong Ministry of Finance or equivalent);

• An effective competition agency;

• Effective regulatory bodies for banks, financial markets etc, for utilities and for other key services;

• An effective (and honest) system of law-courts, particularly for commercial law enforcement, including contract enforcement, bankruptcy provisions, etc.

Besides these agencies, an effective modern nation state needs to be able to provide adequate resources for representation at a sizeable (and increasing range) of international agencies. Developing countries have been playing an increasing role in organisations such as the WTO and the ITU (International Telecommunications Union) as well as in regional groupings such as SADC (the Southern African Development Community), ASEAN and APEC. To provide adequate representation and negotiation also requires the commitment of significant numbers of skilled people. To find and deploy sufficient numbers of such people can be a real problem as shown by the recent moves in the WTO and by some developed

12

countries to fund support for developing countries in their use of WTO committees and dispute panels.

But, deploying highly trained people on these activities means that they are unavailable to run companies or to work in education, health and other services. More directly, the people needed to run competition or regulatory agencies or to represent a country at the WTO need scarce high-level commercial skills. For poor countries (and for many Transition) economies these are the scarcest skills, much scarcer than engineers, teachers or even (in many cases) of doctors.

3.1.1. An Example: Botswana and the Botswana Telecommunications Authority

Botswana provides an interesting example of the problems that can arise. Botswana is not a particularly poor country (on an exchange rate basis, it has a per capita level of GDP of $3,000 per head – much above the average level for Africa). But, it has a population of under 2 million people of whom under 1% have any post-secondary education. In 1997, the annual supply of new graduates from within Botswana was under 2,000 and the total number of new graduates law, commercial studies and the social sciences was 864. Of course, some Botswanan people obtain higher education in South Africa, the US, etc, but not all of them return to Botswana and some Botswanan-educated graduates will emigrate.16

Botswana does not have a separate competition agency. Since 1997, it has had a regulatory agency for telecoms, including the radio spectrum – the BTA (Botswana Telecommunications Authority). The BTA is a decision-making regulatory agency and not just an advisory regulator. Hence, in 1998, it carried out all the main regulatory functions for telecommunications and radio with a single person Agency head (an Executive Chairman) plus 22 staff of whom under half were regulatory professionals. In 1999, the BTA was given the responsibility for the regulation of broadcasting (infrastructure, licensing/franchising and content) but little increase in its staffing levels was expected. (The BTA had already been heavily involved in appraising the bids for Botswana’s first cable television franchise competition.)

It was anticipated that, in due course, the BTA would become an “umbrella” regulator covering electricity, water and possibly other services. However, even if its remit were extended in this way, its chances of doubling or tripling its regulatory staff base are inevitably slim given the number of people available with the necessary skills.

The BTA operates under a good quality statute. It is an agency that has generated a reputation for acting as an effective agency even though there are inevitably problems,

16 Even allowing for inward migration of graduates, this can cause problems for filling high-level governmental, judicial and regulatory jobs since there is typically a requirement that they be filled by local nationals. Botswana is more relaxed about employing foreign nationals in some jobs than many countries, but even they do not do so at the highest levels.

13

sometimes significant, over its covering its full range of responsibilities under the 1996 Telecommunications Act. The BTA’s staff come from various backgrounds - some are recent graduates; some of the more senior members (not surprisingly) have come from the incumbent telephone company (the Botswana Telecommunications Corporation) or from Ministries. The BTA is working hard to carry out its remit and some of its proposed activities to meet the demand on its services will be discussed later in the paper.

However, the point to be made here is that, already, there has been some impact on Ministry and other capabilities. Thus, the Ministry of Transport and Telecommunications was unable to assign (in late1998) any staff to work on telecoms policy. They relied on the BTA for policy advice and also to provide representation for Botswana at ITU, SADC meetings and for WTO telecom negotiations. These are classic Ministerial functions.

The establishment of an independent regulatory agency is intended to separate policy from regulation. But, in the Botswanan telecoms case, at least initially, it reunited them – but with the new regulatory agency in the leading role. This was not intended and was recognised as unfortunate and something to be remedied. But, given the very limited availability of people with the appropriate skills, it was judged to be unavoidable - at least initially.

This is not an isolated instance. Elsewhere (eg Romanian electricity) it appears that new regulatory agencies appear to take up a policy role because, unlike the relevant Ministry, the new agency is powerful and can attract and retain good staff.

If problems of this kind arise in well-governed and not particularly poor countries like Botswana, what are the prospects for effectively introducing the necessary institutional underpinnings for private sector infrastructure provision eg for all the African and other countries with populations of under 10 million and per capita incomes of under $3,000 (or $1,000) per year? That is the question we address in subsequent sections.

3.2. Discussion and Further Comments

We recognise that shortage of qualified regulatory staff is not the only reason why small and low income countries find it difficult to establish effective regulatory institutions. Other reasons include the following:

1. For very small countries (eg under 3 million people), there is the problem that the political, economic and social elite is typically sufficiently small and regularly interactive that any genuine separation of powers can become virtually impossible. Since regulation (and, in particular, the separation of regulation from policy) depends on the separation of powers, this can make the establishment of separate regulatory agencies a notional rather than a genuine exercise – although this does not appear to be the case in Botswana, which has a population of only around 1 million.

14

2. There is strong evidence that effective regulatory institutions and performance depend critically on effective law courts and judicial independence17. For many countries, the lack of reliable law courts and/or the capacity of the commercial law courts has in the past been or is currently a major impediment.

3. The political economy of the country and its “institutional endowment” must be consistent with the development of commercialised utilities and separate regulatory institutions18. For increasing numbers of countries, this is becoming true, but, in some cases (eg some non or very slowly reforming post-Communist states), it is either not true or any stated commitment is more lip-service than genuine.

4. The country and its utility industries may be sufficiently small or straightforward that regulatory arrangements can be handled by contract (eg by concession contracts) and that competition is not relevant at the country level. This appears to be the case in many of the smaller African countries which have electricity systems with generating capacity of well under 1 GW. But, such countries are developing regional trading arrangements and power pools and the problems are likely to arise at that level.

It is also the case that the need for substantial numbers of regulatory staff can be reduced eg by minimising the scope for regulatory discretion. For various legal and other reasons beyond the need to economise on regulatory staff, minimising regulatory discretion has been a policy pursued in several Latin American countries (eg Chile). This is generally achieved by making the primary law establishing the regulatory process as detailed as possible both on what is covered in it and how it should be interpreted so that neither the regulatory agency nor the courts have any significant interpretative or discretionary responsibilities. However, as shown in the next section, this does not appear to result in small regulatory agencies – some of the largest are in Latin America. (This strategy also creates problems should the structure of the industry change or major issues arise that were unforeseen when the original law was written.)

The use of regulation by contract (eg defined regulatory procedure clauses in PPAs and concession contracts) can also potentially address the problem. The use of explicit and formally written-down regulatory contracts by regulators has also been advocated (eg by Veljanovski). However, both of these come up against the fundamental problem that the major reasons for instituting separate regulatory institutions is because of the inherent difficulties in writing time-consistent, enforceable long-run contracts for 20 plus years ahead that can realistically cover all the necessary contingencies.

For most utility industries, there are clear advantages in relying on defined regulatory processes by which contractual issues can be re-opened at regular intervals or triggered by certain events or by major concerns of the parties rather than on regulatory contracts. This

17 See Bergara, Henisz and Spiller (1998) and the references cited therein. 18 See Stern and Holder (1999). The concept and role of institutional endowments is set out in North (1990).

15

certainly seems to apply to complex industries with long-lived investments such as electricity, gas and telecoms. It may introduce an element of regulatory uncertainty eg by requiring an element of regulatory discretion but this is likely to be rather less than the regulatory uncertainty from incomplete contracts and the issue of whether or not the contract will hold or break in the presence of an unforeseen crisis.19

Regulation by contract may (perhaps) be more feasible for toll roads and similar infrastructure and seems to have some possibilities in telecoms but looks very unpromising for electricity. The force of this issue has been clearly demonstrated by the problems that have arisen over PPAs in Indonesia, Pakistan, Thailand, etc in the face of the 1997 Asian financial crisis20. In any event, there are still substantial legal and other staff needs for the recipient countries in negotiating, writing, monitoring and enforcing the contracts. This is also a frequent but rarely discussed problem in many developing countries.

These and other alternatives make it clear that problems in recruiting and retaining sufficient regulatory staff with the appropriate skills is not by any means the sole reason why small and low income countries have major problems in developing effective regulatory institutions. Nevertheless, the supply of sufficient regulatory staff with the necessary skills remains a major issue worth further investigation. It is a common problem and addressing it satisfactorily is a necessary albeit not a sufficient condition for developing

In the next two sections, we report statistics, firstly, on the size and composition of regulatory agencies in a range of developing countries; and, secondly, on the potential supply of people with the necessary educational qualifications. We will then return to discuss the implications.

19 Consider the analogous discussion of degrees of flexibility in exchange rate regimes and the problem of whether pegged exchange rates can be held in a crisis. Modern international macro-economics tends to argue for either irrevocably fixed exchange rates (EMU, currency boards) or floating rate systems (with varying degrees of flexibility) but not for pegged but adjustable rates. See IMF and other papers following the 1997 Asian Crisis and references therein.

20 See Stern and Holder (1999) for further discussion of this issue.

16

4. TELECOMMUNICATIONS AND ELECTRICITY REGULATORY AGENCIES: THEIR SIZE AND COMPOSITION

In this section, we discuss the size of regulatory agencies for telecoms and for electricity (or energy if combined with natural gas and/or other energy industries. We also pay particular attention, where reported, to the numbers of regulatory professionals employed by the regulatory agency (economists, lawyers, financial analysts, engineers).

The discussion is not intended to be comprehensive of all telecom and electricity regulatory agencies around the world but it should give a good cross-section particularly for developing countries. In particular, we take care to distinguish between decision-making regulatory agencies and advisory regulatory agencies. The former are much more likely to need to be a certain minimum size, because they have specific enforcement powers as well as general legal functions. Advisory regulatory bodies can be any size.

4.1. Telecommunications Regulatory Agencies

Buckle (1999) provides a survey of 13 national telecommunications regulatory agencies. His sample includes 4 from Europe (Estonia, Norway, Sweden and UK), 4 from Asia (Australia, Hong Kong, Malaysia and Singapore), 4 from Central and Latin America (Chile, Colombia, Jamaica and Peru) and 1 from the Middle East (Jordan). The data is mainly taken from the regulatory agencies’ websites. Most of it refers to 1998 although a few of the examples are for 1996.

Of the 13 regulatory agencies, 7 also had responsibility for radio spectrum management and 6 did not. The Norwegian and Swedish regulatory agencies also had responsibility for posts while the Jamaican regulator, the OUR, was an “umbrella” regulator with responsibilities for water and sewage, electricity and bus transport. In 1998, the OUR was primarily a recommending regulator with limited decision-making powers. (It now has more decision making powers, particularly for telecoms). The Swedish telecom regulator was also responsible for supervising radio content (but not television).

The results of the survey are set out in Table 1, taken from Buckle (1999).

On staffing levels, Buckle’s conclusions are as follows for his sample:

• The average number of staff employed by a telecommunications regulator responsible for managing the radio spectrum was 202 employees.

• The average number of staff employed by a telecommunications regulator which is not also responsible for managing the radio spectrum was 107 employees.

17

• Only 2 of the regulators considered in our survey employ fewer than 50 staff, and only 4 employ fewer than 100 staff.

• The experience of Jamaica strongly suggests that adequate decision-making (rather than advisory or reccomendatory) regulation with a staff of only about 25 people is not viable in the long-run.

The detailed results are set out in the table below.

18

Table 4.1: Staff of the National Telecommunications Regulators

Country Authority Number of staff

Spectrum Management?

Competition in sector?

Powers of enforcement?

Existed for >5 years?

Number of Lines

(1997)

Australia Australian Communications Authority (ACA) 450 r 9,350,000 Chile Subsecretaria de Telecomunicaciones (SUBTEL) 130 r 2,600,000 Colombia b Comisión de Regulación de Telecomunicaciones (CRT) 20(+20) a r r 5,334,400 Estonia N/A N/A r r 469,000 Hong Kong Office of the Telecommunications Authority (OFTA) 288 3,646,500 Jamaica Office of Utilities Regulation (OUR) 27 r r r r 353,000 d

Jordan Telecommunications Regulatory Commission (TRC) 65 r r 402,600 Malaysia Jabatan Telekom Malaysia (JTM) 85 c r 4,223,000 Norway Norwegian Post & Telecommunications Authority 170 r 2,325,000 Peru Organismo Supervisor de Inversion Privada en

Telecomunicaciones (OSIPTEL) 125 r r 1,645,900

Singapore Telecommunications Authority of Singapore 158 a r 1,684,900 Sweden National Posts & Telecommunications Agency (PTS) 165 6,010,000 UK Oftel 172 r 31,430,000

Source: Regulators' Websites a These figures are for 1996 (all other figures are for 1998).

b The figure in brackets denotes the number of external consultants used in outsourcing.

c Estimated on basis of 51 professional staff plus support staff. d 1996 data.

19

It is at least as important to know the skill levels of the regulatory staff. This information is much harder to obtain, particularly on a standardised basis. Buckle reports the results of some case-studies.

In the UK, Oftel (at end-1997) had 172 employees of whom 129 (75%) were professionals dedicated to regulatory tasks. This is likely to be typical for the European regulators.

In Hong Kong, OFTA (in 1998) had 288 staff, 70% of whom were dedicated to regulatory tasks. In terms of professional skills, OFTA employed 37 professional engineers and 2 economic/accounting professionals.

In Jamaica, the OUR (primarily an advisory regulator in 1998) had 27 staff, of whom 13 (48%) were regulatory professionals, including 2 economists, 2 financial experts and 1 legal counsel.

Developing countries (and transition economies) typically have a larger supply of engineers but very limited supplies of experienced economists, lawyers and accountants. Unfortunately, these are the skills most needed for economic regulation, particularly regulation of privatised companies with competition over networks.

4.2. Electricity Regulatory Agencies

In Table 2 below, we report similar information for electricity regulation. In many cases (and increasingly) regulation of electricity has been or is being combined with the regulation of natural gas. The results have again been compiled from World Bank and regulatory agency websites and refer to the position in the late 1990s. The judgement on the degree of autonomy is taken from the World Bank PSID regulatory database21.

The table is not intended to be comprehensive, particularly for OECD countries but concentrates on developing countries and recently established EU regulatory agencies. Inevitably, the table omits some recent developments, particularly in Africa, which will be discussed below.

The key results are as follows:

In Western Europe, regulatory agencies are typically autonomous, decision-making agencies funded by licence fees or levies.

• In terms of staff numbers, they are well-staffed, 50+ is the norm in Western Europe (and North America);22

21 As found at: www.worldbank.org/psd/ifur 22 Italy with 30 is the lowest in the table. Norway and Sweden (not included in the table) each have 35.

20

• In terms of staff composition, the percentage of professionals is high at 50% or more, particularly lawyers, economists, financial analysts, accountants, etc.

Latin America has several autonomous or semi-autonomous regulatory agencies (Argentina, Brazil, Colombia, Mexico, but Chile has a more Ministerial regulator.

• In terms of staff numbers, the figures are comparable with Western Europe. Brazil with 325 has the highest staffing level of any in the table. The smaller ones are Chile (with 40 staff and covering gas as well as electricity and Colombia with 35). However, the Peruvian electricity tariff regulator apparently has only 10 staff;

• In terms of staff composition, where reported, the percentage of professionals is lower than for Western Europe. In addition, professional level staff tend to be heavily weighted towards engineers rather than the lawyers, economists and accountants found in Western Europe (Note the economist/engineer groups in Argentina, Colombia and Peru which experience suggests are heavily engineer dominated.)

In Asia, the only autonomous regulatory agencies recorded in the table are for Singapore and the Philippines, but and the latter is funded out of general taxation and has a mixed record in terms of its regulatory governance.23

• In terms of staff numbers, staffing levels are generous with Malaysia, Philippines and Singapore all having over 100 staff;

• In terms of staff composition, we do not have access to the relevant statistics but experience of working in the countries suggests that, among the Asian regulators, regulatory professionals are not a large proportion and engineers are the dominant professional group with only very small numbers of lawyers, economists, etc.24

In Africa, the only non-Ministry regulators recorded in the table are in South Africa and Kenya with a semi-Ministerial regulator in Côte d’Ivoire. However, the report of a regional conference held in Gaborone, Botswana in March 2000 recorded that Kenya, Zambia and Malawi also now had functioning autonomous or semi-autonomous electricity regulatory agencies and they were in the process of being developed in Malawi, Mozambique, Namibia, Swaziland, Tanzania and Uganda.25 Other information provides evidence of new electricity regulators at various stages of establishment in Ghana, Senegal and other African countries.

23 Some Indian states (Orissa, now being followed by some others eg Andhra Pradesh and Haryana) are introducing autonomous regulators and Pakistan made abortive moves in that direction in the late 1990s. Thailand and (less strongly) Indonesia are currently (mid-2000) moving in that direction but not Malaysia. See Stern and Holder (1999) for a recent appraisal of Asian utility regulatory trends in these and other developing Asian economies.

24 This is the case for telecoms regulators. 25 Report of Workshop Proceedings to be found at

www.ner.org.za/regional_cooperation/botswana_march_2000.htm

21

• In terms of staff numbers, the new African electricity regulatory agencies are small. According to the table South Africa has only 22 staff and Kenya 27, comparable to the 25 of the Botswana Telecom Authority. The Zambian regulator is apparently smaller still. The number of staff working on regulatory issues in the African Ministry regulators is in some cases under 10, in more than one case covering more than one industry. It is difficult to see how much genuine regulatory oversight they can actually provide

• In terms of staff composition, there is no readily available statistical information on the composition and skill mix of the staff in these agencies. However, evidence from the Botswana Telecom Authority and working experience in Africa suggests that the availability of professional staff, particularly economists, lawyers and accountants is likely to be very limited indeed and this will hamper the new agencies as they begin to build up their regulatory capacity and credibility.

The statistics in table 2 below provide a survey of developing country regulatory experience and capacity. They demonstrate the difficulties that many developing countries have in building up a pool of regulatory professionals, particularly the key professionals needed to support the complex task of regulating competition over monopoly networks. Not surprisingly, the problem seems to become more difficult the poorer and the smaller the country.

Evidence reported in Buckle (op cit) and working experience suggests that it is difficult to provide a fully effective decision-making regulatory capability with fewer than 30-40 staff. There are considerable fixed costs associated with establishing licence (or concession) approval, monitoring and enforcement procedures, particularly where they are subject to enforcement and/or appeal in the law courts. Beyond these fixed costs, regulatory staff requirements seem to depend primarily on the number of companies to be regulated and then on the complexity of the competitive arrangements in the sector.

For the lower income and smaller African countries, the number of companies requiring detailed regulatory oversight (eg transmission, distribution and retail supply) is likely to be low eg 3-6. This is demonstrated by the very small size of the electricity systems reported in the table. Nevertheless, it looks to be a very daunting task to provide effective regulatory oversight with around 20-25 staff of whom fewer than half are likely to be regulatory professionals.

This problem is most apparent in the smaller African countries but also arises in other parts of the world eg in Indo-China, in some of the smaller Latin American countries, in some of the Provinces of China and States of India, in some of the smaller countries of Central and Eastern Europe and the CIS, etc. Personal working experience plus discussions with other practitioners suggests that the problem of the potential supply of regulatory staff can, for a variety of reasons, be a major handicap in utility reform.

22

In subsequent sections, we will consider, firstly, the potential supply of key staff from the educational sector; and secondly, possible strategies for management of the problem given the difficulties of rapid increases in the supply of qualified professionals. In both of these, we will need to remember the potential additional demand for qualified professionals in general (and regulatory and policy staff in particular) from other regulatory and policy needs – as well as from other sectors of the economy, not least the corporate sector.

23

Table 4.2: Staff of National Electricity Regulatory Agencies

Country Authority Type of Agency (1)

Number of Regulatory Staff

(3)

Market Structure?

(1)

Licence/ Concession Administra

tion? (1)

Tariff Regulation?

(1)

Size of System

(GW 1997) (2)

Argentina Ente Nacional Regulador de Electricidad (ENRE)

Semi-Autonomous

(Levy-funded)

141 (of whom 25 lawyers, 42 engineers/ economists)

D D D 20.6

Brazil National Agency of Electrical Energy (ANEEL)

Autonomous (Levy-funded)

325 D D D 60.8

Burkina Faso Ministry of Energy and Mines Ministry (General budget

funded)

10 R D R 0.08

Burundi Ministry of Energy and Mines (includes Water)

Ministry (General budget

funded)

4 R D R 0.04

Chile Comisiόn Nacional de Energia (includes Gas)

Ministry 40 R D 7.4

Congo Ministry of Industrial Development, Energy, Mines, Posts and Telecommunications (includes Water)


funded)

47 D D R 3.2

Eritrea Ministry of Energy, Mining and Water Resources (includes Gas and Water)


funded)

225 R n/a

Colombia Comisión Nacional de Regulación de Energie y Gas (includes Gas)

Semi-Autonomous

35 (of whom 5 lawyers, 15

economists/ engineers)

10.8

Côte d’Ivoire Autorité Nationale de Régulation de Section d’Electricité

Semi-Autonomous

38 Planned P D P 1

24



(3)

Market Structure?

(1)


tion? (1)

Tariff Regulation?

(1)

Size of System

(GW 1997) (2)

Gambia Ministry of Works Communication and Information (includes Gas, Telecommunications, Water)

Ministry 4 0.03

Italy Autorita per l’Energia Ellectrica e il Gas

Autonomous (Levy-funded)

Approx 30 (Max 120 allowed)

A R D 61.4

Kenya Electricity Regulatory Board 27 0.8 Mali Ministry of Mines, Energy and

Water (includes Water) Ministry

(General budget funded)

11 D D A 50.1

Malaysia Department of Electricity and Gas (includes Gas)


funded)

150 R D R 11.8

México Comisión Reguladora de Energia (includes Gas)

Autonomous (General budget

funded)


economists/ accountants, 29

engineers)

D D D 37.6

Mozambique Direccao Nacional de Energia (includes Gas)


funded)

23 R D, R D, R 2.4

Namibia Ministry of Mines and Energy (includes Gas)


funded)

35 D D D n/a

Peru Comisión de Tarifas Eletricas Autonomous (Levy funded)

10 (of whom 1 lawyer, 6

economists/ engineers)

D 4.3

Philippines Energy Regulatory Board (includes Gas)

Autonomous (General budget

200 P D D 8.7

25



(3)

Market Structure?

(1)


tion? (1)

Tariff Regulation?

(1)

Size of System

(GW 1997) (2)

(includes Gas) (General budget funded)

Sao Tome and Principe

Ministry of Social Equipment and Environment (includes Telecommunications, Aviation, Ports, Roads and Water)


funded)

6 0.01

Singapore Public Utilities Board (PUB) Autonomous (Licence funded)

101 D D D 5.6

South Africa National Electricity Regulator Semi-Autonomous (Levy funded)

22 P A D 35.2

Spain Comisión del Sistema Eléctrico Nacional

Autonomous (Levy funded)


economists, 28 engineers

P D P 41.7

UK Offer (now Ofgem, including gas) Autonomous (Licence funded)

252 (of whom 106 lawyers, economists,

accountants, and other senior

professionals)

P D D 70.5

Sources: (1) http://wbln0018.worldbank.org/psd/ifur.nsf/, (2) www.eia.doe.gov/emeu/international/electric.html#Capacity, (3) Various websites and author knowledge.

Definitions: D Decision-making A Advisory R Recommendatory P Proposes

26

4.3. Implications

One response to the arguments and data on regulatory capabilities set out above is to say:

“So what?

All regulators, including those in the richest countries, always complain about inadequate staff resources, the difficulties of recruiting and retaining specialist staff and the implications for their work. All economic regulation is imperfect and that is why economists always advocate the use of markets in preference to regulation wherever possible.

No-one with any sense would expect economic regulation in developing countries to be anything other than very limited in scope and skills. But, in most cases, even with the gross imperfections, it must be better than the mess created by state-owned utilities which have awful service records, tend to be grossly inefficient and which, in most cases, are in a dreadful financial state.

Further, even when some competitive elements are unbundled and open to private investment (eg mobile telecoms, electricity generation), it may be possible to provide the regulation by appropriate contracts (eg simple PPAs in electricity), rather than require an explicit regulator.”

This is clearly one view that can be taken and, in essence, it makes the case that the perfect must not be the enemy of the good. However, it is not the end of the story for the following reasons:

(i) It assumes that unbundled, privatised utilities are always (or virtually always) the superior solution. For the small-scale electricity industries of many of the African countries in Table 3.2, that is not obviously correct. For instance, the size of the market may well be too small to support competition in electricity or even in telecoms (let alone water, railways, etc).

(ii) For many small, low income countries, public ownership of some sort (including municipal, not necessarily classic state owned enterprises) may well be more suitable, particularly for ownership of major networks. This can usually be combined with franchised (private sector) management to increase commercialisation.

(iii) A sufficiently good regulatory system seems to be essential to support private investment in most utility industries – particularly for slow-growing economies with poor macro-economic reputations and/or low credit worthiness26. It may not be

26 See Levy and Spller (1994) and Stern (1994).

27

essential for roads or mobile telecoms, but regulation by contract for electricity generation is never likely to be very satisfactory, particularly beyond the simplestand smallest African systems.

What, of course, is difficult to judge is how good is sufficiently good. In other words, what is the minimum standard of regulatory coverage and competence necessary to support private investment and avoid the need for the fall-back vertical monopoly, public ownership option.

These are not easy questions to answer empirically. It is quite possible that many small and low income countries will be unable to provide a sufficiently strong regulatory capability to support other than the most limited unbundling, competition and use of private investment.

We will return to these questions at the end of the paper.

28

5. POTENTIAL STOCKS AND FLOWS OF QUALIFIED REGULATORY STAFF: EDUCATIONAL INFRASTRUCTURE

Chile has an electricity and regulatory agency with 40 staff and a telecoms regulator with 130 staff. Among developing countries, it was a pioneer in commercialising and unbundling its utilities – and in developing explicit (if less than fully independent) regulation. In 1998, Chile had a level of per capita income of $5,00027. In addition, latest available education statistics show that, in 1992, 12% of Chile’s population had some post-secondary education and that its annual number of new graduates in 1997, was 42 thousand. Within the 1997 graduate flow, it had 10 thousand graduates in commercial and business subjects and a further 4 thousand in social science subjects (including economics).28

Contrast this with Kenya and Sri Lanka as representative examples of low income, moderate sized African and Asian countries. Kenya had a population of 29 million and a per capita income of $350. In terms of educational qualifications, 1.3% of its population had any post-secondary education. It produces around 10,000 graduates per year of whom 750 were classified as in commerce and business subjects and 650 in social science subjects. Sri Lanka had a population of 19 million and a per capita income of $810. In terms of educational qualifications, 1.1% of its population had any post-secondary education. It produces around 8 thousand graduates per year of whom 1,500 were classified as in commerce and business subjects and a further 200 in social sciences.

These figures are typical. In Sub-Saharan Africa, Zimbabwe had the highest proportion of the population with post-secondary education of 4.9%29 and only Nigeria and South Africa produced more than 5,000 business and commerce and social science graduates per year. In Asia, for countries with per capita income levels of $5,000 or less, only Malaysia (7%), Philippines (22%) and Thailand (5%) had 5% or more of its population with any post-secondary education and only China, India, Indonesia and Pakistan (all with populations of over 100 million) plus the Philippines and Thailand produced more than 5 thousand new graduates per year in business/commerce and social science subjects.

It is very striking how different this is from Latin America. Brazil is among the lowest with 6% of its population having some post-secondary education, Mexico has 9.2%, Colombia 10.3%, Argentina 12%, Ecuador 13% and Peru 20%. Brazil, however, produces over 60

27 Exchange rate basis. World Bank Atlas method. 28 See Annex tables. All educational data cited in this paper from Unesco and UN sources with ratios calculated by

Martin Siner. The latest data for educational stocks is, from estimates made within the last 5 years but much of the data on graduate flows, is unfortunately, from the 1980s or early 1990s. The authors are grateful to the World Bank for providing access to these data.

29 South Africa may well be higher but we have not been able to find a comparable figure.

29

thousand business/commerce and social science graduates per year plus 27 thousand new law graduates per year.

On the basis of the numbers quoted above, it is very noteworthy how the countries that have been in the forefront of developing effective and autonomous regulatory institutions have had 5% or more of their population with some post-secondary education and/or a large annual output of new graduates, particularly in business/commercial and social science graduates. This is true particularly in Latin America but also in Asia.

The implications of this are serious for Africa and the smaller Asian developing countries (also for the smaller and/or low income Provinces/States in China and India).

Taking the World Bank classification of income groups, of 31 low income countries (per capita income under $760 per year), only 3 had more than more than 5% of their population with any post secondary education. Of the 31 lower middle income countries (per capita incomes of $761-$3,030), 18 had more than 5% of their population with this level of education. In terms of graduate flows, only 16 out of 47 low income countries produced more than 10 thousand graduates per year.

For regulatory staffing and competing activities, it is most useful to consider the flow of graduates in law, business/commerce and social science graduates. Of the 32 low income countries for whom the relevant data are available, only 7 produced more than 3 thousand of these graduates per year and only 2 produced more than 10 thousand per year. Of the 13 low income countries with populations of 10 million or under, only 2 produced more than 3 thousand law, business/commerce and social science graduates. Even for the lower middle income countries, of the 16 with populations of 10 million or under, only 7 produced more than 3 thousand law, business/commerce and social science graduates

Taking Chile as a benchmark, we computed the numbers of countries with more than 10% of its population having some post-secondary education and producing at least 10 thousand graduates per year. Only 2 out of 22 low income countries met this benchmark. For lower middle income countries, 9 out of 25 countries met the benchmark but only 3 out of the 13 lower middle income countries with populations of 10 million or less. In this context, it is worth recognising that Argentina, Brazil, Chile and Mexico (but not Colombia) are all well into the World Bank’s upper-middle income group, unlike almost all of Sub-Saharan Africa and much of Asia.

It is against this background that the capacity for establishing credible regulatory institutions must be made. For the countries of sub-Saharan Africa, the low income and small countries of Asia and other areas, the implication is that they may well find it very hard to establish and maintain credible infrastructure regulatory agencies - even making the greatest possible use of multi-sector ‘umbrella’ regulators.

30

It is crucial to remember the demand for these skills from commercial and financial companies as well as Government Ministries and international agencies, the law courts, competition authorities, financial sector and other regulators, etc. The reality is that, in reality, the number and qualification level of the necessary staff to run effective infrastructure regulatory agencies may not be available for many years in most low income developing countries, particularly those with populations of under 10 million people.

In the next section, we will consider what can be done to alleviate the problem.

31

6. OPTIONS FOR ALLEVIATING THE PROBLEM OF INSUFFICIENT REGULATORY STAFF IN SMALL AND DEVELOPING COUNTRIES

In Section 3.2, we discussed some general strategies on how small and developing countries could economise on the need for regulatory staff including writing regulatory laws to minimise interpretation and discretion and regulation by contract. However, it was also pointed out (a) that these strategies had significant regulatory costs and (b) that minimising regulatory discretion did not, apparently reduce the observed size of regulatory offices eg to under 30-40 staff.

In the light of the data reported in the last two sections, we are still left with the problem of how many developing countries are supposed to provide the skilled staffing resources for one (or several) utility regulators as well as financial regulators, competition agency staff, legal system staff as well as qualified company senior managers, high-level civil servants, etc. In this section, we will briefly review some alternatives that have been suggested as potential options for tackling the problem.

6.1. Use of Consultants

One strategy that has been proposed is that small regulators contract out much of their regulatory work eg to consultants.

UK and many other established regulators use consultants. However, they usually do so for one-off pieces of work, for instance, specific studies that are required in the course of a regulatory review. As such, the consultants are used as a complement to existing staff resources.

To use consultants rather than regular staff for on-going day-to-day work is very different, particularly if there are issues of legal enforcement processes. It is also the case that far more resources are needed than generally realised to set up consultancy projects, choose the consultants, work with and monitor their performance, digest the results and implement them (or not). The amount of regulatory office staff time saved by employing external consultants is much less of the consultancy time-input than might be imagined.

In addition, the use of consultants as a substitute for existing staff means that the agency is much less likely to gain the benefits of building up institutional memory.

6.2. Use of Multi-National Regional Regulatory Agency

Another proposal is to create teams of a utility regulator covering several countries, say electricity regulators in West Africa or the Mekong River area.

32

It is clear how this could operate in terms of regulatory commissioners. It is much less clear how this would work (or economise on the need for) regulatory staff who need the detailed expert knowledge and understanding in the circumstances and data for the various countries in the region. It is possible to imagine teams of regulatory staff from the countries attached to and working out of a regional institution or regional development bank (the ADB or the African Development Bank). However, it does require considerable co-operation and trust between the countries and governments concerned. It is doubtful whether these are sufficiently present in many of the areas where it might be a useful possibility.

This issue is related to the question of perceived regulatory legitimacy. This is probably more of a problem if the multi-national regulatory agency is thought of in terms of a body of regulatory commissioners, but staffing is also relevant. Debates within the European Union have clearly revealed the sensitivities from national regulatory decisions being taken by (or being significantly constrained by) EU institutions. In recent years, the pressures have been for more subsidiarity to national governments within EU minimum requirement frameworks and away from the notion of EU-wide regulatory agencies. The former is the path adopted in both the recent EU Electricity and Gas Directives. More widely, much of the backlash against the WTO can readily be seen as a lack of willingness to grant the WTO regulatory legitimacy to override national decisions.

In this context, a decision by (say) a Cambodian-led regulatory commission on cost grounds to increase electricity prices to Vietnamese households by twice the current rate of inflation is likely to be of very doubtful regulatory legitimacy to the Vietnamese government and people. Indeed, it may well be a necessary condition that such decisions would be made by a single regulator or a regulatory commission head from the country under review.

6.3. Multi-National Regulatory Collaboration

A less ambitious variant on the multi-national regional regulatory agency is for the development of shared information and possible pooling of resources between regulators in neighbouring countries.

This option is one that has been adopted in the European Union where regulators (and regulatory agencies) meet to exchange views and information and, where appropriate to help develop common approaches. In developing countries, something similar appears to be developing in Southern Africa. The March 2000 regional conference of Southern African electricity regulators referenced in Section 3.2 seems to be an example of this. Similar co-operation is apparently proceeding between Southern African telecoms regulators (including and heavily involving the Botswana Telecommunications Authority). This pattern may well be present among other groups of developing countries eg in the Caribbean and/or Central or Southern America.

33

This option looks potentially very attractive. It does not create problems of regulatory legitimacy and it enhances rather than diminishes the development of institutional memory. It can also develop flexibly as needed in terms of both depth and breadth of coverage. For instance, one can imagine a regional grouping where (say) one agency built-up a high level of expertise in transmission access and pricing and provided this to the other regulators while a second country agency built up expertise in (say) regulatory issues affecting rural electrification.

Over time, such arrangements could, if required become the basis of establishing more formal multi-national regulatory institutions between countries. However, the main benefits could well come from the informal co-operation. Further, it is neither necessary nor obviously desirable (not least for the reasons discussed above) that such co-operation should result in single multi-country regulators. In the US, the individual states maintain their state-level Public Utility Commissions as well as having the FERC and the FCC at Federal level.

A major benefit is that the degree of regulatory collaboration between national regulators can be primarily market-driven. Clearly the net benefits of regional regulatory collaboration increase as inter-country trade levels and market integration increases. In addition, the potential for providing and developing regulatory underpinnings may well be very advantageous for such market expansion and hence for the sustainability of a greater degree of competition than can be supported in small electricity or telecom markets.

The only obvious disadvantage is that progress will be limited to the speed at which the slowest wish to proceed. But, while we are discussing sovereign countries, this may actually be as much of an advantage as a disadvantage. It also mirrors the governance and enhancement of integration by consensus that exists in other multi-national organisations, including the European Union and the WTO.

6.4. Discussion and Further Work

This section has discussed some potential practical possibilities for alleviating the staffing difficulties of regulatory agencies in small and developing countries besides the generic options discussed in Section 3. It would be surprising if there were not a number of other options that had been tried out or discussed.

There is also the option of recognising that regulatory capabilities will inevitably remain very limited in these countries and just “making do” eg concentrating on the most important issues and providing minimal (or no) coverage of less than crucial issues. This is the implication of the “So what?” discussion in Section 4.3 above. In practice, this may well be the approach realistically taken by many developing country regulators and, in many cases, it may be sufficient to support considerable private investment. This Chilean electricity system, where the regulatory system is not independent and has well-recognised major imperfections is a case in point where there are major and recognised deficiencies in the

34

regulatory system for electricity, but it has been “good enough” to maintain privatised industries and to support considerable amounts of new private investment.

It would be very interesting to see the results of a more thorough-going and systematic study on the use of these (and all other relevant options) for utility regulation in developing countries. Any such study would probably need to take the form of an interview survey of regulators and others in a sample of developing countries, possibly an in-depth interview study. This would, among other things, provide a data-base for countries in addressing these issues which, this paper has argued must be resolved if utility reform is to continue and develop in middle and lower income countries.

To test the “So what?” argument and help establish the necessary minimum standard of regulation to support competition and private investment, any survey of regulatory institutions should be supplemented by a survey of potential investing counties to establish their attitudes on ensuring a sufficiently safe economic environment for major investments in new capacity and/or existing companies. This would best be done in the context of specific developing countries and their (usually recent or embryonic) regulatory institutions.

35

7. SUMMARY AND CONCLUSIONS

This paper has argued that the new commercialised, unbundled utility model, based on private investment and competition where appropriate and feasible has been, and is continuing, to spread through the telecommunications and electricity industries of developing countries, including many lower income and small developing countries. However, the sustainability and success of this model depends considerably on the establishment of effective and autonomous regulatory institutions.

These institutions and their performance were crucial in the pioneering Latin American electricity reform movements. The presence (or, in many cases, the absence) of effective autonomous regulatory agencies has also been important in the Asian utility industries eg their relative ability to respond effectively to the Asian financial crisis.

This creates the problem of how to ensure an adequate supply of regulatory staff. The success of regulatory regimes depends to a considerable extent on creating autonomous institutions that are sustainable and which possess and develop an effective regulatory memory. Other factors are also important, but for small and low income countries (eg in Sub-Saharan Africa and in Asia) this looks to be – and is in practice – a very serious problem.

Data assembled and reported in Section 4 of this paper demonstrate the numbers of staff working in regulatory agencies and, where possible, the numbers of senior professionals. For many African countries (unlike Argentina, Brazil, Chile and Mexico) the numbers are regulatory staff are very small indeed.

Section 5 provides data on income levels and educational resources. The results of this analysis clearly demonstrate the relatively strong educational infrastructure of the Latin American utility reform pioneers on the basis of which the new regulatory bodies could recruit sufficient staff – and could recruit a reasonable number of senior regulatory professionals – without imposing serious costs on the rest of the economy. The position is very different in the low income countries of Africa and Asia, where the stocks and flows of people with the necessary skills is much lower.

The paper has also explored the degree to which this problem can be alleviated. It seems very doubtful whether the problem can be avoided by the use of regulation by contract. Multi-sector regulators within a country help and are useful to economise on staff whose skills can be used in more than one regulated industry but they still require quite a lot of staff to be effective.

Regional, multi-national regulatory bodies were explored but, as shown by EU and other experience, the problem of regulatory legitimacy with multi-national regulators creates serious difficulties even for countries with considerable levels of integration and political trust. Contracting out day-to-day regulatory work also provides limited assistance.

36

The most promising development identified in this paper is the idea of informal exchanges of information and pooling of resources between national regulators. This seems to be developing in a very interesting and potentially productive way in Southern Africa both for telecommunications and electricity. It is also the route being followed in the European Union. A particular advantage is that informal pooling can be market-driven, both responding to and helping encourage trade, integration of markets and networks and increasing the scope for competition.

However, these conclusions are only provisional since they are based on a limited sample of data. Further, their importance is not obvious and, even if correct, the implications are by no means obvious.

It would be very interesting to see the results of a more thorough-going study on the use of the options identified (and all other relevant) options for utility regulation in developing countries. Any such study would probably need to take the form of an interview survey of regulators and others in a sample of developing countries, possibly an in-depth interview study. This would, among other things, provide a data-base for countries in addressing these issues which, this paper has argued must be resolved if utility reform is to continue and develop in middle and lower income countries.

But, to test the importance of regulatory capability for reform and the introduction of competition and private investment, any survey of developing country regulatory institutions should be supplemented by a survey of potential investing counties to establish their attitudes on ensuring a sufficiently safe economic environment for major investments in new capacity and/or existing companies. This would best be done in the context of specific developing countries and their (usually recent or embryonic) regulatory institutions. That would help establish the essential minimum regulatory capability necessary to support competition and private investment in the infrastructure industries of such economies.

JON STERN London Business School and NERA

November 2000

37

References

Bergara, M. E., Henisz, W.J., Spiller, P.T., 1998. Political institutions and electricity utility investment: a cross-nation analysis. Californian Management Review 40 (2), 18-35.

Buckle, A. 1999. International experience of telecommunications regulation. NERA, London.

Cave, M., Stern J., 1998, Regulatory institutions and regulatory policy for economies in transition. In: Regulating Utilities: Understanding the Issues. Institute of Economic Affairs/London Business School.

Hunt, S., Shuttleworth G., 1996. Competition and Choice in Electricity. John Wiley & Sons.

Levy, B., Spiller, P.T., 1994. The institutional foundations of regulatory commitment: a comparative analysis of telecommunications regulation. Journal of Law and Economics and Organisation 10 (2), 201-246.

Littlechild, S., 1983. Regulation of British Telecommunications Profitability. HMSO.

Newbery, D.M., 1999. Privatisation, Restructuring and Regulation of Network Utilities. The MIT Press.

North, D.C., 1990. Institutions, Institutional Change and Economic Performance. Cambridge University Press.

Posner, M., 1993. Electricity and the primary fuels: technology, market structure and prices. National Institute Economic Review No. 145, 3/93, 64-86.

Spiller, P.T., Martorell, L.V., 1996. How should it be done? Electricity regulation in Argentina, Brazil, Uruguay and Chile. In International Comparisons of Electricity Regulation. Eds: Gilbert, R.J., Kahn, E.P. Cambridge University Press.

Stern, J., 1994. Economic regulation in central and eastern Europe. Economics of Transition, September, 391-397.

Stern, J. 1997. What makes an independent regulator independent? Business Strategy review, Summer, 67-75.

Stern, J. and Davis, J.R., 1998. Economic Reform of the Electricity Industries of Central and Eastern Europe. Economics of Transition, Volume 6 (2), 427-60.

38

Stern, J., Holder, S., 1999. Regulatory governance: criteria for assessing the performance of regulatory systems. An application to infrastructure countries in the developing countries of Asia. Utilities Policy 8, 33-50.

39

APPENDIX A. SUMMARY TABLES ON EDUCATION AND INCOME BY COUNTRY

Table A.1: No. of countries (by income and population) with more than 5% and 10% per capita post-secondary education rates. (figures in brackets are no. of observations)

Population 0-2m 2-5m 5-10m 10-20m 20m+ Total Income >5% >10% >5% >10% >5% >10% >5% >10% >5% >10% >5% >10% Low 0 0 (2) 1 1 (6) 1 1 (6) 1 0 (4) 0 0 (13) 3 2 (31) Lower middle 2 1 (8) 3 2 (4) 4 2 (5) 3 2 (5) 6 4 (9) 18 11 (31) Upper middle 3 3 (12) 4 1 (5) 2 1 (2) 1 1 (1) 7 3 (7) 17 9 (27) High 14 10 (15) 6 5 (6) 6 4 (6) 1 0 (1) 4 3 (5) 31 22 (33) Total 19 14 (37) 14 9 (21) 13 8 (19) 6 3 (11) 17 10 (34) 69 44 (122)

Table A.2: No. of countries (by income and population) with more than 5,000 and 10,000 graduates per annum. (figures in brackets are no. of observations)

Population 0-2m 2-5m 5-10m 10-20m 20m+ Total Income >5k >10k >5k >10k >5k >10k >5k >10k >5k >10k >5k >10k Low 0 0 (1) 7 3 (11) 5 4 (12) 4 2 (10) 10 7 (13) 26 16 (47) Lower middle 2 1 (5) 5 4 (8) 5 4 (6) 6 5 (6) 11 11 (11) 29 25 (36) Upper middle 6 5 (11) 4 2 (4) 2 2 (2) 1 1 (1) 8 8 (8) 21 18 (26) High 3 2 (9) 6 6 (7) 7 7 (7) 3 3 (3) 6 6 (6) 25 24 (32) Total 11 8 (30) 22 15 (30) 19 17 (27) 14 11 (20) 35 32 (38) 101 83 (141)

40

Table A.31: No. of countries (by income and population) with more than 3,000 and 10,000 law, business and social science graduates per annum. (figures in brackets are no. of observations)

Population 0-2m 2-5m 5-10m 10-20m 20m+ Total Income >3k >10k >3k >10k >3k >10k >3k >10k >3k >10k >3k >10k Low 0 0 (1) 1 0 (7) 1 0 (5) 1 0 (8) 4 2 (11) 7 2 (32) Lower middle 0 0 (4) 4 0 (8) 3 0 (4) 4 1 (6) 11 9 (11) 22 10 (33) Upper middle 0 0 (5) 2 0 (4) 2 0 (2) 1 1 (1) 6 5 (7) 11 6 (19) High 1 0 (8) 6 1 (7) 7 2 (7) 3 2 (3) 4 4 (4) 21 9 (29) Total 1 0 (18) 13 1 (26) 13 2 (18) 9 4 (18) 25 20 (33) 61 27 (113)

Table A.4: No. of countries (by income and population) with more than 10% per capita post-secondary education rates and 10,000 graduates per annum. (figures in brackets are no. of observations)

Population 0-2m 2-5m 5-10m 10-20m 20m+ Total Income Low 0 (0) 1 (5) 1 (5) 0 (2) 0 (10) 2 (22) Lower middle 0 (4) 1 (4) 2 (5) 2 (4) 4 (8) 9 (25) Upper middle 0 (9) 0 (3) 1 (2) 1 (1) 1 (5) 3 (20) High 1 (7) 5 (6) 4 (6) 0 (1) 3 (5) 13 (25) Total 1 (20) 7 (18) 8 (18) 3 (8) 8 (28) 27 (92)

41

Source for tables A.1-A.4:

Population and Income data - World Bank (2000), World Development Indicators Education data - UNESCO Statistical Yearbook.

Income Categories:

High income - $9,361 or more Upper middle income - £3,031 to $9,360 Lower middle income - $761 to $3,030 Low income - $760 or less

Note:

1. In table 3, countries were included if data was available at least two of the three educational categories (i.e. law, commerce and business, and social science)

42

APPENDIX B. SOURCE TABLES ON EDUCATION AND INCOME BY WORLD REGION AND COUNTRY

Table B.1: Africa

Population GNP per capita Post-secondary Total Graduates Law Commerce Social Total COUNTRY (millions) (Atlas method, $) (%) & Business Sciences Algeria 30 1,550 32,557 2,140 1,967 3,660 7,767 Angola 12 380 383 15 101 116 Benin 6 380 1.3 933 130 95 225 Botswana 2 3,070 0.9 1,971 53 465 306 824 Burkina Faso 11 240 3,613 257 857 1,114 Burundi 7 140 0.6 1,192 104 257 361 Cameroon 14 610 1,804 Cape Verde 0.4 1,200 Central African Republic 3 300 2 650 Chad 7 230 1,184 145 145 Comoros 0.5 370 Congo, Dem. Rep. 48 680 1.2 3,460 Congo, Rep. 3 110 3 1,095 Cote d'Ivorie 14 700 8.7 Dijbouti 0.6 H Egypt 61 1,290 4.6 107,512 11,494 20,284 1,638 33,416 Equatorial Guinea 0.4 1,110 Eritrea 4 200 559 26 155 181 Ethiopia 61 100 7,448 111 1,340 243 1,694 Gabon 1 4,170 730 Gambia 1 340 Ghana 18 390 2,739 40 247 269 556 Guinea 7 530 891 Guinea-Bissau 1 160 0.1 Kenya 29 350 1.3 10,701 135 751 647 1,533

43

Population GNP per capita Post-secondary Total Graduates Law Commerce Social Total COUNTRY (millions) (Atlas method, $) (%) & Business Sciences Lesotho 2 570 1,930 58 632 57 747 Liberia 3 I 483 Madagascar 15 260 6,341 781 1,829 621 3,231 Malawi 11 210 3,815 18 197 157 372 Mali 11 250 1,195 Mauritania 3 410 1.3 733 157 10 196 363 Mauritius 1 3,730 1.9 1,729 22 175 54 251 Mayotte 0.1 G Morocco 28 1,240 24,501 4,172 318 2,743 7,233 Mozambique 17 210 0.1 327 8 23 31 Namibia 2 1,940 4 1,760 168 168 Niger 10 200 436 Nigeria 121 300 31,322 1,999 1,988 4,271 8,258 Rwanda 8 230 724 63 98 73 234 Sao Tome and Principe 0.1 270 0.3 Senegal 9 520 4,917 Seychelles 0.8 6,420 4.6 23 Sierra Leone 5 140 1.5 Somalia 9 I 8,145 South Africa 41 3,310 147,391 4,391 25,206 11,115 40,712 Sudan 28 290 0.8 7,588 875 1,273 821 2,969 Swaziland 1 1,400 3.3 988 74 226 51 351 Tanzania 32 220 2,413 129 967 199 1,295 Togo 4 330 1.3 5,254 931 1,009 1,940 Tunisia 9 2,060 2.8 14,565 1,134 2,869 760 4,763 Uganda 21 310 0.5 9,959 71 2,142 463 2,676 Zambia 10 330 0.4 4,374 Zimbabwe 12 620 4.9 14,087 143 2,324 405 2,872 Chile 15 4,990 12.3 41,548 678 10,314 3,880 14,872 United Kingdom 59 21,410 25 456,180 18,705 88,440 37,731 144,876

44

Table B.2: Asia

Population GNP per capita Post-secondary Total Law Commerce Social Total Country (millions) (Atlas method, $) (%) Graduates & Business Sciences Afghanistan 25 I 3 2,429 157 148 87 392 Bangladesh 126 350 1.3 Bhutan 0.8 470 Cambodia 11 260 1 China 1239 750 2 1,040,135 31,198 199,105 230,303 China, Hong Kong SAR 7 23,660 14.5 23,515 739 5,305 1,616 7,660 Cyprus 0.8 11,920 17 2,482 18 766 51 835 Georgia 5 970 28,431 1,620 1,022 1,877 4,519 India 980 440 2.5 1,213,387 36,021 234,405 270,426 Indonesia 204 640 2.3 200,583 70,617 70,617 Japan 126 32,350 20.7 1,127,500 306,738 306,738 Korea, Dem. People's Rep. of 23 I Korea, Republic of 46 8,600 21.1 380,571 90,842 90,842 Lao People's 5 320 2,817 206 206 Macau 0.5 F 5.9 1,656 40 579 67 686 Malaysia 22 3,670 6.8 20,886 467 579 2,319 3,365 Maldives 0.3 1,130 1.7 Myanmar 44 I 2 35,898 439 216 2,125 2,780 Nepal 23 210 0.6 Pakistan 132 470 2.5 93,570 3,590 6,026 9,616 Philippines 75 1,050 21.9 286,545 1,967 86,585 88,552 Singapore 3 30,170 7.6 20,603 194 4,564 4,758 Sri Lanka 19 810 1.1 7,995 227 1,550 223 2,000 Thailand 61 2,160 5.1 130,223 10,131 62,608 72,739 Turkmenistan 5 I Uzbekistan 24 950 Vietnam 77 350 2.6 Chile 15 4,990 12.3 41,548 678 10,314 3,880 14,872 United Kingdom 59 21,410 25 456,180 18,705 88,440 37,731 144,876

45

Table B.3: Central and Eastern Europe and the Former Soviet Union

Population GNP per capita Post-secondary Total Graduates Law Commerce Social Total Country (millions) (Atlas method, $) (%) & Business Sciences Armenia 4 460 9,055 185 820 331 1336 Azerbaijan 8 480 27,249 5 3,135 3140 Belarus 10 2,180 12.5 68,663 10 11,831 11841 Bosnia & Herzegovina 4 H Bulgaria 8 1,220 15 36,463 1,317 6,925 1,501 9743 Croatia 5 4,620 6.4 12,467 641 2,316 550 3507 Czech Republic 10 5,150 8.5 30,640 2,072 7,359 395 9826 Estonia 1 3,360 13.7 6,236 217 1,164 269 1650 Hungary 10 4,510 10.1 32,393 2,607 3,666 1,836 8109 Kazakstan 16 1,340 12.4 106,008 2,684 9,608 978 13270 Kyrgyz Republic 5 380 Latvia 2 2,420 13.4 8,188 241 1,360 287 1888 Lithuania 4 2,540 12.6 17,704 943 2,663 405 4011 Macedonia, FYR 2 1,290 3,292 186 621 807 Moldova 4 380 11.3 11,810 686 3,062 3748 Mongolia 3 380 7,394 190 714 427 1331 Poland 39 3,910 7.9 134,367 3,916 20,256 5,924 30096 Romania 23 1,360 5.6 70,841 7,830 11,926 1,709 21465 Russian Federation 147 2,260 14.1 950,488 15,537 72,651 107,573 195761 Slovak Republic 5 3,700 9.5 12,799 399 1,895 292 2586 Slovenia 2 9,780 10.4 8,557 563 2,793 275 3631 Tajikistan 6 370 11.7 10,198 98 98 Ukraine 50 980 140,218 1,537 12,899 14436 Yugoslavia 11 H 19,050 1,273 321 2,516 4110 Chile 15 4,990 12.3 41,548 678 10,314 3,880 14,872 United Kingdom 59 21,410 25 456,180 18,705 88,440 37,731 144,876

46

Table B.4: Central America

Population GNP per capita Post-secondary Total Graduates Law Commerce Social Total Country (millions) (Atlas method, $) (%) & Business Sciences Antigua and Barbuda 0.1 8,450 Aruba 0.1 F 0.7 Bahamas 0.2 F 13.5 Barbados 0.3 G 3.3 Belize 0.2 2,660 6.6 Bermuda 0.1 F 18.4 BVIslands 13.6 Cayman Islands 0 F 14.1 Costa Rica 4 2,770 9,813 623 3,124 550 4,297 Cuba 11 H 5.9 27,502 246 667 822 1,735 Dominica 0.1 3,150 1.7 197 13 28 41 Dominican Republic 8 1,770 19,276 2,040 4,535 1,233 7,808 El Salvador 6 1,850 6.3 7,897 1,932 1,932 Grenada 0.1 3,250 1.5 Guadeloupe - 5.2 Guatemala 11 1,640 2.2 Haití 8 410 0.7 Honduras 6 740 3.3 1,649 102 329 128 559 Jamaica 3 1,740 2.6 1,575 204 448 652 Martinique - 5.6 Mexico 96 3,840 9.2 238,553 15,758 43,699 10,544 70,001 Montserrat 5.8 Netherland Antilles 0.2 F 8.8 Nicaragua 5 370 6,915 512 608 349 1,469 Panama 3 2,990 13.2 5,744 198 1,347 233 1,778 Puerto Rico 4 G 28.7 St Kitts and Nevis 0 6,190 2.3 183 18 18 St Lucia 0.2 3,660 1.3 313 St Vincent and the Grenadines 113 2,560 1.4 Trinidad and Tobago 1 4,520 3.4 998 78 228 306

47

Population GNP per capita Post-secondary Total Graduates Law Commerce Social Total Country (millions) (Atlas method, $) (%) & Business Sciences Turks and Caicos Islands 7.7 US Virgin Islands 0.1 F 24.4 278 158 8 166 Chile 15 4,990 12.3 41,548 678 10,314 3,880 14,872 United Kingdom 59 21,410 25 456,180 18,705 88,440 37,731 144,876

48

Table B.5: Middle East

Population GNP per capita Post-secondary Total Graduates Law Commerce Social Total Country (millions) (Atlas method, $) (%) & Business Sciences Bahrein 0.6 7,640 10.3 1,382 372 372 Brunei Darussalam 0.3 F 9.4 420 15 39 54 Iran 62 1,650 4.1 83,385 32,244 83,463 4,868 120,575 Iraq 22 H Israel 6 16,180 11.2 15,573 506 4,946 5,452 Jordan 5 1,150 22,395 794 4,109 1,212 6,115 Kuwait 2 F 10.1 3,341 371 570 19 960 Lebanon 4 3,560 9,653 1,281 1,885 1,731 4,897 Libya 5 G 2.7 Oman 2 G 733 21 21 Palestine 4,510 63 595 75 733 Qatar 0.7 F 13.3 1,360 114 114 Saudi Arabia 21 6,910 37,901 282 1,864 853 2,999 Syrian Arab Republic 15 1,020 24,202 1,940 2,609 4,549 Turkey 63 3,160 124,861 2,590 13,573 19,413 35,576 United Arab Emirates 3 17,870 1,638 32 187 219 West Bank & Gaza 3 1,560 Yemen 17 280 9,080 558 1,697 110 2,365 Chile 15 4,990 12.3 41,548 678 10,314 3,880 14,872 United Kingdom 59 21,410 25 456,180 18,705 88,440 37,731 144,876

49

Table B.6: Oceania

Population GNP per capita Post-secondary Total Graduates Law Commerce Social Total Country (millions) (Atlas method, $) (%) &

Business Sciences

American Samoa 0.1 G 22.6 Australia 19 20,640 141,478 4,951 27,301 5,549 37,801 Fiji 0.8 2,210 4.5 French Polynesia 227 F Guam 0.1 F 39.9 Kiribati 0.1 1,170 Marshall Islands 0.1 1,540 Micronesia 0.1 1,800 Micronesia 0.1 1,800 New Caledonia 0.2 F 7.5 New Zealand 4 14,600 39.1 34,218 905 7,243 1,750 9,898 Northern Mariana Islands 0.1 F Pacific Islands 10.9 Palau 0 G Papua New Guinea 5 890 Samoa 1,070 2.7 Solomon Islands 0.4 760 Tonga 0.1 1,750 2.8 Tuvalu 6.9 Vanuata, Rep. Of 0.2 1,260 Chile 15 4,990 12.3 41,548 678 10,314 3,880 14,872 United Kingdom 59 21,410 25 456,180 18,705 88,440 37,731 144,876

50

Table B.7: South America

Population GNP per capita Post-secondary Total Graduates Law Business Social Total Country (millions) (Atlas method, $) (%) & Business Sciences Argentina 36 8,030 12 Bolivia 8 1,010 9.9 4,348 496 882 396 1,774 Brazil 166 4,630 5.5 240,269 26,535 43,209 21,268 91,012 Chile 15 4,990 12.3 41,458 678 10,314 3,880 14,872 Colombia 41 2,470 10.3 92,219 10,568 27,707 38,275 Ecuador 12 1,520 12.7 11,722 737 2,112 1,155 4,004 French Guiana - 6.4 Guyana 0.8 780 1.8 1,554 27 186 90 303 Paraguay 5 1,760 6.6 3,669 432 623 174 1,229 Peru 25 2,440 20.4 27,428 1,955 5,730 3,690 11,375 Suriname 0.4 1,660 Uruguay 3 6,070 10 5,930 650 353 604 1,607 Venezuela 23 3,530 10.1 United Kingdom

59 21,410 25 456,180 18,705 88,440 37,731 144,876

51

Table B.8: Western Europe

Population GNP per capita Post-Secondary Total Graduates Law Commerce Social Total COUNTRY (millions) (Atlas method, $) (%) & Business Sciences Albania 3 810 3,845 299 462 169 930 Andorra 0.1 F Austria 8 26830 6.1 21,357 1700 2413 1437 5550 Belgium 10 25380 59,894 2390 17887 20277 Channel Islands 0.1 F Denmark 5 33040 19.6 30,243 654 6079 1366 8099 Faroe Islands 0 F Finland 5 24280 15.4 28,625 436 1751 1477 3664 France 59 24210 11.4 389,820 Germany 82 26570 336,473 11681 21412 48592 81685 Greece 11 11740 8.7 29,268 1334 3357 223 4914 Holy See 3,047 48 393 441 Iceland 0.3 27830 1,624 Ireland 4 18710 13.1 35,380 859 7760 1268 9887 Isle of Man 0.1 G Italy 58 20090 3.8 175,489 16822 19661 10015 46498 Liechtenstein 0 F Luxembourg 0.4 45100 10.8 Malta 0.4 10100 1,332 178 297 20 495 Monaco 0 F Netherlands 16 24780 80,675 3733 11510 17246 32489 Norway 4 34310 17.9 52,990 1910 7294 8051 17255 Portugal 10 10670 7.7 39,116 2213 7205 3581 12999 Spain 39 14100 8.4 177,737 25732 31224 6913 63869 Sweden 9 25580 21 35,243 965 4549 2172 7686 Switzerland 7 39980 11.5 37,210 1717 1957 1114 4788 Chile 15 4,990 12.3 41,458 678 10,314 3,880 14,872 United Kingdom 59 21410 25 456,180 18705 88440 37731 144876

52

Sources for Tables B.1-B.8

Income and population data - World Bank (2000), World Development Indicators.

Education data - UNESCO Statistical Yearbook.

Population and income data is for 1998.

‘Education’ data is taken from the most recent year available (italics are used to indicate data greater than ten years old).

Data on post-secondary education rates is the % of the population with a post-secondary education whilst the other data are the number of graduates (within the defined groups) per annum.

Data definitions:

Post-secondary education defined by the International Standard Classification of Education (ISCED) and accepted by UNESCO is 'provided at universities, teachers' colleges, higher professional schools, which requires as a minimum condition of admission, the successful completion of education at the second level, or evidence of the attainment of an equivalent level of knowledge.' (UNESCO)

Law - Law, programmes for 'notaires', local magistrates, jurisprudence.

Commerce and Business - Business administration and commercial programmes, accountancy, secretarial programmes, business machine operation and electronic data processing, financial management, public administration, institutional administration.

Social Sciences - Social and behavioural science, economics, demography, political science, sociology, anthropology, psychology, geography, studies of regional cultures.

Total = sum of law, commerce and business and social sciences.

53

Income Codes (where figure unavailable):

F – estimated to be high income ($9,361 or more) G – estimated to be upper middle income (£3,031 to $9,360) H – estimated to be lower middle income ($761 to $3,030) I – estimated to be low income ($760 or less)

Ref: Anne Mace/16/29-Nov-00/F:\JONS\MSOFFICE\WINWORD\LBS\Small Country Reg\SmalCtryPaperFINNov00.doc

regualtion initiative working paper series number 41...

Documents