Domestic Implications of Debt Relief

Chapter One

Introduction

In recent years, the external debt situation for a number of low-income countries has become extremely difficult, prompting the IMF and the World Bank to design a framework in 1996 to provide special assistance to the heavily indebted poor countries (HIPC). There are 41 HIPCs, including 31 HIPC-LeDCs, which meet the three criteria to qualify for the enhanced initiative. These criteria are (i) a country is only eligible for highly concessional assistance (IDA); (ii) it has an IMF poverty reduction and growth facility supported-programme (PRGF) in place; and (iii) it has agreed to a rescheduling of debts on concessional terms with the Paris Club.

According to the UNDP, the most impoverished and vulnerable countries of the world are grouped under the category of ‘least developed countries’ (LeDCs). Most, but not all, LeDCs are heavily indebted. Furthermore, there are some heavily indebted countries that do not belong to the LeDC category. A country is designated as a least developed country if it meets inclusion thresholds on the following three criteria (African Development Bank 2005):

 1 A low income: income to be below a GDP per capita of US$800.

 2 Weak human resources, measured by the Augmented Physical Quality of Life Index, which is based on indicators of life expectancy at birth, per capita calorie intake, combined primary and secondary school enrolment, and adult literacy.

 3 A low level of economic diversification, measured by the Economic Diversification Index, which is based on the share of manufacturing in GDP. The share of the labour force in industry, annual per capita commercial energy consumption and UNCTAD’s merchandise export concentration index.

The classification of HIPCs seems to be based on a rule of thumb rather than on clear-cut quantitative criteria. In 1996, when the category was introduced, the group of HIPCs consisted of 32 severely indebted low-income countries and nine other countries. To be classified as severely indebted in 1996 a country should have had

1 Present value of debt service to GDP to exceed 80 per cent, or

2 Present value of debt to exports to exceed 220 per cent (UNCTAD, 2002);

Two other common denominators of this group are that the countries only borrow on highly concessional terms from the World Bank (from the Bank’s International Development Association, IDA) and they have negotiated, or are prepared to negotiate, a concessional rescheduling with the Paris Club. Since 1996, the original group of 41 HIPCs has gone through some changes. Nigeria was soon ruled ineligible, as it also borrowed from the World Bank’s non-concessional window, the International Bank for Reconstruction and Development (IBRD). In 1999, Malawi was included and in the summer of 2000, Gambia was added as well.

Equatorial Guinea was declassified as an HIPC in early 2000, as, with the onset of oil production, GDP levels rose above those required for IDA-only assistance. Even though LeDCs are, by definition, very poor and many are also heavily indebted, not all LeDCs are classified as heavily indebted poor countries (HIPCs). The World Bank and the IMF currently classify 41 countries as HIPCs.

The group of 49 LeDCs shows considerable overlap with these HIPCs. More specifically, 31 out of 49 LeDCs are HIPC-LeDCs and ten HIPCs are non-LeDC but HIPCs (Whaites, 2005). Eighteen LeDCs are not classified as HIPCs. Clearly, the classifications of HIPCs and LeDCs are rather arbitrary. Not only because several non-HIPC-LeDCs have unsustainable debts because all HIPCs are very poor and underdeveloped, even though they do not meet the criteria to fit in the LeDC category.

One of the main reasons why development efforts have failed in LeDCs is that instead of investing in social and economic development, these countries are forced to use scarce government resources to finance external debt to foreign creditors. For example, in Burkina Faso, where one out of five children dies before the age of five, in 1998 the government spent as much on debt as on health. And in Niger, where 78 per cent of adult males and 93 per cent of adult women are illiterate, debt service amounted to 3.1 per cent of GNP in 1998, while spending on education was only 2.3 per cent of GNP (Whaites, 2004).

Hjertholm (2000) cites considerable empirical evidence suggesting that there is a strong and significant negative relationship between high debt burdens and poor economic performance, such as low growth, investment and human development. There are several channels through which these occur. But of these, two are particularly important: (i) cash flow effects arising from reduced public expenditures, and (ii) disincentive effects associated with a large debt overhang.

Clearly the magnitude of debt service matters for import capacity in such instances. Import compression can occur both at the balance of payments level and at the budgetary level (through the effects of public debt service on the import-content of government expenditures). Reductions in the import capacity of the government, as a result of debt service, can thus reduce government investment activity, whereby the complementarity effects are lost. Hjertholm (2000) presents evidence to suggest that such cash flow effects have indeed been at work in 23 indebted low-income sub-Sahran African countries.

In addition to cash flow effects are disincentive effects associated with debt overhang. Since governments know that the beneficial effects of economic reforms will largely go to pay external debt, they have little incentive to pursue such reform. Affected by such disincentives are tax reforms as well as investment that cannot be buoyant in a climate of macroeconomic stability.

A public debt overhang can affect macroeconomic stability by increasing the fiscal deficit, depreciating the exchange rate too fast, monetisation of deficit and reliance on seignorage and recourse to exceptional financing such as payments arrears and debt rescheduling, which tends to exacerbate uncertainty about the future debt servicing profile of the public sector. Hence tackling the debt crisis is extremely important for these countries. In Table 1 we give some indication of the burden of the external debt on these countries.

Table 1 clearly indicates the weight of the burden of debt on the poorest countries. In 1998, the total debt stock of LeDCs amounted to US$154 billion. For every single LeDC, the debt stock shows a steady and significant increase since 1980. The majority of the total debt stock is owed by the 31 HIPC-LeDCs with their share being 85 per cent or US$129 billion of the total LeDC debt (World Bank, 1998a). Some LeDCs, such as Mozambique, Angola and the Solomon Islands, owe a significant amount of debt to commercial banks and other private creditors.

But for the majority of LeDCs, the main part of the total debt stock consists of official debt, i.e. bilateral debt owed to governments and multilateral debt owed to multilateral financial institutions. In 1998, 40 per cent of the LeDC debt stock was multilateral debt and 38 per cent bilateral debt. The remaining 22 per cent included private debt and related interest arrears as well as short-term official debt, plus interest arrears.

Debt stock indicators have also worsened for most HIPC-LeDCs since 1980. As a result of rising debt levels, the present values of debt-exports ratio and of the debt-GNP ratio have increased since 1980 for most HIPC-LeDCs, even though exports and GNP show some modest growth. However, for a number of HIPC-LeDCs, debt stock indicators improved from 1980-98. Compared to the performance of other HIPC-LeDCs, these countries’ GNP and export income have improved significantly between 1980 and 1998. In other words, the improved debt indicators are to be attributed to improved economic performance rather than lower absolute levels of debt.

In spite of this, however, these countries’ debts still remain far above the conservative World Bank and IMF sustainability threshold of a 150 per cent present value of debt-export ratio, as is the case for all the HIPC-LeDCs. Also, for 23 out of 28 HIPC-LeDCs for which data are available, the present value of debt-GNP ratio is above 80 per cent, the threshold used by the World Bank in the past to classify a country as ‘severely indebted’.

The group of non-HIPC-LeDCs shows major differences in the level as well as the evolvement of debt indicators. Five out of 14 non-HIPC-LeDCs for which 1998 data are available, have a debt level that exceeds the 150 per cent present value of debt-export threshold.

In one more country, Samoa, the present value of debt-GNP exceeds 80 per cent. Furthermore, for seven out of the 14 non-HIPC-LeDCs for which data are available, debt indicators are worse in 1998 than in the 1980 and 1990 levels, even though GNP and exports have increased compared to 1980. The debt indicators have improved for only two countries, Cambodia and former HIPC Equatorial Guinea. However, Cambodia’s present value of debt-export ratio is still far above the 150 per cent threshold (Devarajan & Go, 2001).

In addition to these debt indicators, the amount of arrears is indicative of the severity of the debt burden. A large number of LeDCs have huge arrears—in particular Congo, Liberia, Somalia and Sudan. In fact nine LeDCs have arrears of over 20 per cent of their total debt stock, and in the case of a further six LeDCs, these arrears amount to more than 15 per cent of the debt stock. There are three non-HIPC-LeDCs for which arrears are extremely high: Cambodia (43 per cent), Comoros Islands (22 per cent) and Equatorial Guinea (45 per cent). The amount of arrears clearly demonstrates that many LeDCs are simply incapable of paying their debt service obligations. For at least 18 LeDCs, including four non-HIPC-LeDCs, annual debt repayments decreased in 1998 compared to 1990, even though the debt stock increased.

The repayments made by LeDCs during 1980-98 are given in Table 1. In 1998, these countries returned a total of US$4,740 million to foreign creditors. HIPC-LeDCs account for 80 per cent of this total. The main receivers of LDC debt repayments are multilateral creditors. It should be pointed out, however, that in addition to the 31 HIPC-LeDCs, many LDCs also have unsustainable debts, even with the rather arbitrary and narrowly defined definition of the World Bank and IMF. Six non-HIPC-LeDCs have debt levels that exceed the WB and IMF debt sustainability thresholds: Bangladesh, Cambodia, Comoros Islands, Haiti, Nepal and Samoa. Furthermore, Equatorial Guinea’s amount of arrears equally suggests that this country is unable to carry its debt burden. Tackling the debt crisis is essential for these countries. If creditors fail to significantly reduce the debt level of LeDCs, social and economic development and poverty reduction remain a mirage.

A question that needs to be raised is: how sustainable are debt levels of HIPC countries given the high levels of poverty that they have? Debt sustainability has been assessed only from the perspective of repayment capacity and the implications of such repayment for the domestic economy have rarely been considered. Some computations suggest that once adjustment is made for their high levels of poverty, most, if not all, HIPC countries should qualify for debt write-off.

The World Bank and the International Monetary Fund (IMF), known as the Bretton Woods Institutions, launched a joint Initiative at the end of 1999 setting the fight against poverty at the heart of development policies. Under this Initiative, low-income countries wishing to apply for financial aid from either of the organisations, or for debt relief under the HIPC (Heavily Indebted Poor Countries) Initiative, are required to draw up poverty reduction programmes known as Poverty Reduction Strategy Papers (PRSP).

Chapter Two

Role of World Bank and IMF

Since then, the Bretton Woods Institutions (BWIs) have mobilised considerable human and financial resources to implement the Initiative and to ensure its success. They began by reorganising and renaming their aid programmes so as to explicitly include the fight against poverty in their key goals. The World Bank’s financial support for low-income countries, handled by the International Development Association (IDA), now comes within the framework of Poverty Reduction Strategy Credits (PRSC) and the IMF’s within the Poverty Reduction and Growth Facility (PRGF).

The BWIs also gave much thought to defining precise guidelines to help low-income countries prepare the PRSPs. The World Bank’s World Development Report 2000/2001 (World Bank 2000a), entitled Attacking Poverty, was duly followed in 2001 by a reference document, or sourcebook, designed as a practical guide for the approximately sixty countries concerned (World Bank 2001a). Finally, unparalleled efforts were made to consult the developing countries’ governments and civil societies, with the organisation of a wide series of international, national and regional seminars.

The increase in importance of poverty-related issues

Today’s acute awareness of the problem of poverty in developing countries is the result of a long process that began at the end of the 1980s. The United Nations, particularly the UNICEF (United Nations International Children’s Emergency Fund) and the UNDP (United Nations Development Programme), were the precursors in this field, whereas the Bretton Woods Institutions took more time to include the issue in their policies. In 1987, the UNICEF published a report that was to have a tremendous impact. Entitled Adjustment with a Human Face, it alerted people of the disastrous effects of structural adjustment policies, and put forward various possible solutions (Cornia et al. 1987).

Throughout the 1990s, the United Nations organised several international conferences that helped create a new awareness of poverty-related issues. One of the most important ones was doubtless the World Summit for Human Development, held in Copenhagen (Denmark) in 1995. The Final Declaration and Action Programme ratified at this summit made poverty reduction a priority of development.

The United Nations General Assembly followed this by declaring 1996 ‘International Year for the Eradication of Poverty’ and the decade 1997-2006 as the ‘First United Nations Decade for the Eradication of Poverty’. Also in 1996, the OECD’s Development Assistance Committee (DAC) decided to focus its attention on poverty and drafted the International Development Goals, which were enlarged and renamed as the Millennium Development Goals (MDGs) by the United Nations in 2000. The first of the MDGs aims to reduce the proportion of the world’s population living in extreme poverty by half between 1990 and 2015.

The World Bank also accompanied this gradual move towards recentring development policies on the question of poverty. First indirectly, by introducing the Social Dimensions of Adjustment Programme, in partnership with the UNDP and the African Development Bank, designed to lessen the short-term negative effects of reforms on vulnerable populations. Then more directly, by devoting the World Development Report 1990 to the subject of poverty. By publishing this report (followed by a second one on the same theme ten years later), the Bank stressed how much importance it attached to the fight against poverty.

It then provided itself with the analytical means of carrying out its mission, by undertaking a series of studies based on data obtained from surveys on this theme, whilst also systematically preparing documents to help define the nature and determining factors of poverty in each country, i.e. the poverty profile, poverty assessment, etc. In fact, only the IMF stayed on the fringe of this movement (World Bank 1995).

At the end of the last decade, the Bretton Woods Institutions had three main reasons for enhancing their positions in this field by launching new poverty reduction strategies that put this theme at the very centre of their actions, particularly in the case of the World Bank. The first is related to an increase in poverty in many parts of the world. However uncertain the figures may be, there can be no question that poverty has risen in the past few years, particularly if the figures for China are excluded, as the poor population seems to have been significantly reduced there since economic expansion started in the 1980s.

The increase in poverty has been particularly significant in sub-Saharan Africa and transition countries in Europe and Central Asia. According to the World Bank’s statistics, nearly half the world population lives with less than two dollars per day and a fifth with less than one dollar per day, the amount generally considered as the threshold of absolute poverty.

Perhaps the most worrying phenomenon in this respect is that certain countries appear to have fallen into a ‘poverty trap’, illustrated by the fact that the number of least-developed countries (LDCs) has doubled in the last thirty years, rising from 25 in 1971 when this category of country was first created, to 49 in 2001. Given the demographic trends in these countries, their population will be tripled by 2050 according to United Nations forecasts, rising from 660 million to 1.8 billion, representing nearly 20 per cent of the world population (compared with 11 per cent at present). In this context, urgent action must be taken to prevent entire populations from literally dying of hunger.

The second reason that forced the BWIs to change their policies is related to the failure of structural adjustment policies, and the questioning of the ‘Washington Consensus’ on which they were based, with the three-pronged solution of macroeconomic stabilisation, external and internal liberalisation. The above figures suffice to prove the failure of structural adjustment. After twenty years of recommending these policies, the BWIs cannot produce a single case where they have been successful. Even the ‘Asian miracle’, for many years given as an example to the other developing countries, has been questioned since the 1997 crisis.

The same applies to the programmes applied with the BWIs’ support in other large emerging economies (Argentina, Brazil, Mexico, Turkey, etc.) and transition economies (Russia), which have also experienced serious crises that are still far from being resolved. As for the ‘front-runners’ in Africa, successively praised for their success stories, none has managed to live up to expectations in the long term (it is doubtless rather cruel to point out that the Democratic Republic of Congo, formerly Zaïre, was once ranked in this category). By and large, the ‘Washington Consensus’ failed in most countries, as even the World Bank now admits. It is interesting to hear what Joseph Stiglitz, former chief economist of this institution and Nobel prizewinner for Economics in 2001, has to say on this subject:

The IMF is supposed to guarantee international financial stability. As for the WTO, it is designed to facilitate international trade. Unfortunately, the way these two institutions try to carry out their mandate has probably contributed to increasing poverty… The cocktail obtained from the combination of liberalisation policies and restrictive economic policies imposed by the IMF has had the most dramatic effects for developing countries. (Stiglitz, 2002)

His criticism may spare the World Bank, but our analysis shows that there is no reason why it should be given preferential treatment, at least not until recently. Apart from calling past strategies into question, the failure of structural adjustment policies has led to the multilateral debt crisis, which has an adverse impact both on developing countries and on the BWIs. The crisis is the direct result of twenty years of debt rescheduling without economic growth. The international financial institutions, particularly the World Bank, are on the front line in a situation in which poor countries are heavily indebted and the multilateral element has gradually become preponderant. Faced with the states’ insolvency, it is the principle of the intangibility of the BWIs’ debts that was at stake.

The third factor involved in their change of attitude is the crisis in the legitimacy of the Bretton Woods Institutions. To a great extent, this stems from the other two factors: given the increase in world poverty and the overall failure of their policies, it is quite natural that more and more people should criticise them and call for changes to their policies (some even suggesting to close them down).

These changes were especially a way of responding to aid fatigue in developed countries, where public opinion was beginning to wonder whether it was really worth while continuing to spend so much on development aid. The theme of poverty was used to combat their disenchantment, the perspective of humanitarian solidarity helping to rehabilitate development aid (Severino 2001; Tarp, 2000).

They were also a means of answering criticisms from civil society and the various protest movements. It is important to remember that the decision to launch the PRSPs took place against a background of increasingly strong protests against international organisations, which forced the G7 countries to launch the enhanced HIPC Initiative at the Cologne Summit in 1999, culminating at the Seattle Conference in November 1999 and the G7 Summit in Genoa in June 2001 (Cogneau, Maurin & Pasquier, 2001).

Challenges and future policies

As we have already seen, we can only make a partial assessment of the new poverty reduction initiatives. At this stage the only elements that can be assessed are the processes used to draw up the strategies and their content. The history of development has shown that this is probably not the most important part of the issue. For example, a similar exercise carried out in the 1960s or 1970s would have analysed the options adopted in the long-term plans drawn up in most developing countries.

A large number of economists did just that at the time, but none foresaw their real results. We have to admit that it is unlikely that we are much more clear-sighted today. In retrospect, although the failure of these plans was partly due to mistaken assessments of their future impact, it was at least as much due to the countries’, and particularly their public authorities’, failure to implement their principles. Hence, in the coming years, the challenges will concern both the validity of the policies and the ability to implement them, while monitoring and evaluation tools will be required to both assess the situation and evaluate the results of the policies implemented. On all three levels, participation is supposed to play a key role. It is also the main unknown factor.

On the content of the policies, the fact that the majority of PRSPs simply renewed the previous policies shows that there is strong resistance to any questioning of them, however necessary this may be. Of course, the adoption of new poverty reduction strategies implicitly recognises the failure of the previous policies, even if BWIs have pleaded their own case many times by continuing to put most of the blame for the failure of structural adjustment down to the developing countries’ failure to implement them rather than to their content (World Bank 1994, 2002a).

Nonetheless, as many analysts have suggested, there is reluctant recognition of the need for an innovative approach to the content of the policies in areas such as the need for better balance between the state and markets; taking into account the specific national and local contexts; the relativisation of the benefits of liberalisation, etc. If countries are to do more than simply making a few marginal changes to the content of policies, it is vital that they try to identify viable, coherent alternatives to the previous policies. If this challenge is taken up seriously, the participatory process should be of great help population’s real needs and gaining a better understanding of the positions adopted by each of the parties.

As for the implementation of the strategies, the principle of participation from different members of society opens up new prospects that will have an impact on the way national affairs are led. By favouring respect for the right to information and expression, participation fulfils one objective in that it deals with one of the key factors of poverty, namely exclusion and marginalisation. But the potential impact of this precept goes way beyond this aspect. Participation will only take on its full meaning if it really helps solve the problem of the lack of democracy in poor countries.

It should give extra capabilities and power to intermediate bodies (the media, trade unions, associations, etc.) in drawing up, monitoring, controlling, assessing and redirecting the policies. Information is of course of utmost importance in this respect, and its formative nature must be underlined. It makes public choices explicit and increases transparency in the management of state affairs, whilst offering the different players in society the possibility of exerting pressure, or even taking sanctions in the case of failure. In short, making the state accountable for its actions before its citizens is at stake.

The majority of the players do not seem to be fully aware of these new prospects, as they lack guidelines and are used to being excluded from decision-making circles. Also, the real influence of the participatory process on economic policy decisions has yet to be defined. No help on this matter can be found either in the World Bank’s sourcebook on PRSPs or in practice in the field (Whaites, 2004), although the BWIs’ and the governments’ natural inclination probably works in favour of making it a purely consultative process. Once the PRSPs have been drawn up by the countries and validated by the BWIs, can we hope to see an attempt to institutionalise these processes of popular expression?

If this intention does exist, the real problems will arise at that stage, because the participatory process approach challenges the way in which the so-called representative institutions usually work in developing countries. If things are to change, they will need to have a great deal of courage, or, more probably, there will need to be an enormous amount of pressure from the public. The players’ ability to define adequate policies depends on whether they have access to information on the problems as they stand and on the real or expected impact of the options which have been implemented or are planned.

Unfortunately, we only have very fragmented information on poor country economies, and lack relevant, reliable data. In this context, it is very difficult, if not impossible, to assess the impact of the policies. It follows that it is now urgent to mobilise all the means required to deepen our knowledge of the situation and the mechanisms in force in poor countries, both in economic and socio-political terms, given that these two factors are very closely related. Defining an effective strategy to fight against poverty particularly requires a careful assessment of the distributive impact of the economic policies that are planned or implemented.

The World Bank’s and IMF’s Poverty and Social Impact Analysis (PSIA) programmes respond to this need. These recent programmes represent an immense project in terms of operational research and applications. They will be faced with three types of difficulties, described hereafter. First of all, the natural order of things would have been to use these techniques from the beginning when drawing up the PRSPs and defining policy priorities. However, the PRSP system was set up in such haste that this was not possible. More fundamentally, there are no instruments available at the present time to enable precise assessment (ex ante and ex post) of the impact of the policies on poverty (Joseph, 2003).

Certain innovative paths of research, such as micro-simulations techniques, represent promising methodological progress in this direction. But two fundamental problems remain unsolved. How can the multidimensional nature of poverty be taken into account when analysing the policies’ distributive effects? How can the impact of alternative measures be measured in terms of poverty (with a given budget, is it better to construct rural tracks or improve the quality of primary schooling)? A programme of this sort must be based on an analysis of what are still widely unknown interactions between the different forms of poverty and the wide range of alternative policies (Whaites, 2004).

Finally, assuming that researchers solve these analytical difficulties and that the data required to implement the new methods is produced, the constraints of local technical and institutional capabilities will need to be lifted, and the means of dialogue between those ‘in the know’ and society will need to be rethought, in order to bring the key principles of ownership and participation, empowerment and accountability into being. Although this very ambitious project is clearly inaccessible in the short term, it is nonetheless vital to make a firm commitment in this direction. Creating and transferring appropriate instruments to be mobilised for this project will be a final challenge, and by no means the easiest.

In the end, the BWIs have opted for a path over which they have no control. In a way, they have opened a ‘Pandora’s box’ by creating expectations that they may be unable to satisfy, running the risk of disappointment and retreat. They may be taken at their word by the southern countries and have to adapt to radical changes that they have not fully anticipated. In any event, new paths have been opened by the PRSP Initiative. It remains to be seen whether the numerous obstacles and constraints that still exist can be overcome, and whether the different players at national and international levels are aware of the real stakes involved and are ready to mobilise the resources needed to make the process a success.

Chapter Three

Diagnosis and Reflections on Poverty Reduction Policies (Africa & Developing Regions)

The launching of the new international poverty reduction strategies is an enormous challenge for the international community. The initiative is based, amongst other things, on two key observations: the scale of poverty in some regions such as sub-Saharan Africa where nearly half of the inhabitants live in extreme poverty; the progression of the affliction, with the deterioration of living conditions in many countries. According to the World Bank, the number of poor people (based on the threshold of two dollars per day) increased throughout the world in the 1990s, except in the Latin American region where it stagnated and in eastern Asia and the Pacific where it fell.

Apart from being able to stir people into action, global statements of this sort are of little interest, particularly as some of the results are marred by uncertainties. Up until now, not enough attention has been paid to the fact that, if appropriate strategies are to be drawn up, they must be based on a precise description of the current state of affairs, drawing lessons from thinking on the nature and evolution of poverty and from past poverty reduction policies. Cases of policies crowned with success are the exception, particularly in the poorest regions. Nonetheless, the decision-makers act as if they had a wide range of effective, appropriate measures at their disposal and that there is no question of their ability to fight against poverty.

We must face the facts: at the present time there are still very serious gaps in our knowledge of the characteristics, evolution and determining factors of poverty, and also of the means of curbing it. Based on specific case analyses, various paths are suggested to enable us to understand the different dimensions of poverty, to improve our knowledge of changes over time, but also to find a way of making policies more effective. The aim is not so much to give a comprehensive overview of the situation of poor people and of ways of improving their conditions, as to cast a new light on a certain number of essential points that are very often neglected.

Poverty and recession in sub-Saharan Africa

Sub-Saharan Africa is the region in which poverty is at once most acute and paradoxically at the same time least understood. This state of affairs can be explained to a large extent by the lack and/or doubtful quality of available data. For the most part, analyses of poverty based on statistical information are merely monographic and only very rarely take account of the temporal dimension (African Development Bank 2005). To go beyond merely anecdotal observations to obtain an overview of poverty in all its complexity it is necessary to produce an analysis of its evolution over time which takes into account the macroeconomic and social context of the country under consideration.

We have therefore concentrated on African capitals in which the incidence of poverty has markedly progressed, and in which, concomitantly, a major restructuration of the socio-economic environment is taking place. On the basis of a certain number of factors derived from statistical data and illustrated by specific examples, we will attempt to draw up an inventory of the situation in the continent’s major cities and evaluate the impact of the recessionary trend of national economies on their populations’ living standards (Whaites, 2005).

Africa deprived of its inheritance

Since the wave of independence of the 1960s, sub-Saharan Africa’s position vis-à-vis the rest of the world has been on a continual downwards curve. Unlike in other developing regions, per capita income has declined over the last 30 years. In constant 1987 dollar terms, it has dropped from $525 in 1970 to $336 in 1997; in other words, an average decrease of 36 per cent.

During the same period, per capita income increased by 88 per cent in South Asia and by 355 per cent in East Asia. Even South America – where the 1980s were branded the ‘Lost Decade’, so severe was the crisis – enjoyed 55 per cent growth over the period (see Table 2). Of 30 African economies for which long-term figures are available, more than half have declined, and some have shrunk by more than 50 per cent (World Bank, 2004).

Table 2. Comparative performances of sub-Saharan Africa and other developing regions

Finding convincing long-term explanations for this phenomenon is not an easy task. At a time when the rest of the world economy is undergoing globalisation, Africa is becoming sidelined. Most African countries still rely on a small number of primary commodities, and the fact that their economies are still characterised by limited diversification renders them extremely vulnerable (Berthélémy and Söderling 2004). Moreover, the unbridled growth of the population – in spite of the beginnings of a demographic transition – and the rapid degradation of the environment have effectively mortgaged the future. On the political and social front, the picture is almost as grim.

Among 50 countries, there are no fewer than 14 armed conflicts, with their attendant loss of life, long-term injuries and refugees. After the enthusiasm engendered by the impressive wave of democratisation which began in the late 1980s, confidence is on the wane. Progress made has yet to be consolidated and the risk of destabilisation in various regions cannot be discounted. At the same time, the extent of corruption (with the dubious award for world’s most corrupt country going to Cameroon in 1999, and Nigeria in 2000) highlights a generalised deficit of good governance and the failure of the state (Transparency International 2004).

While undeniable progress has been made in the fields of health and, particularly, education, Africa is still lagging behind. Around 250 million Africans do not have access to drinking water, and 200 million are deprived of basic healthcare. Once again, Africa is the only region in which nutrition has not improved – every year, nearly 2 million children die before reaching their first birthday. The terrifying effects of the Aids pandemic will surely be felt in the years to come, but its impact is still little understood. At the present time, Africa accounts for 70 per cent of the world’s cases (Demery, 1999).

Average life expectancy, which is presently just over 50 years, has already declined in southern Africa (the worst hit area), and some forecasts predict the appalling prospect of a drop in the figure of 20 years, a brutal collapse which would completely wipe out all the progress made since the 1950s. Evidently, these adverse economic conditions have had serious consequences in terms of poverty in Africa. Allowing for the scarcity of sources, it can be stated that, in 1998, nearly 50 per cent of Africans lived beneath the extreme poverty line, which corresponds to $1 per capita per day in purchasing power parity.

It is estimated that, even if growth rates in the 1990s were relatively favourable compared to the preceding decade, the number of people in these circumstances increased from 240 million to 300 million in ten years (World Bank 2001e; World Bank 2002a). Consequently, nearly one in four of the world’s poor is African, as compared to only 18 per cent in 1987 (Atkinson & Stiglitz, 1989). This is the result of a combination of poor economic performance and particularly high levels of inequality. Indeed, South America is the only region in which inequality is more pronounced.

The urban dynamics: cities suffer most

In this general context of recession, the very rapid development of cities in sub-Saharan Africa is one of the most important developments of the last 40 years. Urban growth was particularly marked between 1960 and 1975 (rising at an annual rate of 6 or 7 per cent). Since then, the rate has slowed considerably to between 4 and 5 per cent (Joseph, 2003). More than the size of the urban population, which can still be considered moderate at the present time (at around 30 per cent in sub-Saharan Africa, as against 75 per cent in South America and 46 per cent in world terms), it is the speed of urban growth which gives cause for concern since it has not been accompanied by an equivalent economic dynamism (Booth & Lucas, 2002).

To take two precise examples: in West Africa, the number of people living in urban areas increased from 12 to 78 million between 1960 and 1990, with cities absorbing nearly two-thirds of total demographic growth (Snerch 1994). Today, the level of urbanisation is over 40 per cent, whereas in 1960 it was only 13 per cent. This substantial increase in the urban population is just as apparent in relatively ‘rich’ countries such as Nigeria and the Côte d’Ivoire – in which (in the same period) the urban population rose from 15 to 50 per cent – as it is in ‘poor’ countries such as Mauritania (9 to 42 per cent) and Chad (6 to 24 per cent) (Booth, 2002).

The phenomenon is all the more worrying since sub-Saharan Africa is a unique case in world terms in that the region’s rapid urbanisation has not been accompanied by economic growth (Hicks 1998). Between 1975 and 1998, the urban population increased at an annual rate of 5.2 per cent, while per capita GDP decreased almost continuously on average by 0.4 per cent per annum (see Figure 1).

While it is true that there was a reversal in the trend in 1995, this recent dynamic did little to compensate for the decline in per capita income, and seems, in any case, to have been ephemeral. Improvements in macroeconomic management and structural reforms have encouraged recovery, but the upturn depended at least as much on exceptionally high world commodities prices between 1995 and 1997 (World Bank 1998a).

Figure 1 Evolution of the rate of urbanisation and per capita GDP.

Source: World Development Indicators, 2000, authors’ calculations.

Non-agricultural GDP can be used to provide an overall idea of the development of economic activity in the urban context. The picture becomes yet bleaker for the cities of sub-Saharan Africa, in that the nonagricultural GDP divided by the urban population – which gives an approximation of per capita urban GDP – has decreased even more dramatically (World Bank 2000a). Even though the latter was still 2.5 times higher than total per capita GDP for the region as a whole as late as 1998, it fell by 43 per cent between 1975 and 1998 (in other words, an average annual decrease of 2.4 per cent) (World Bank 1998a). The absence of economic growth in the urban context is thus wholly apparent.

In view of this recessionary trend, the future of cities in sub-Saharan Africa gives real cause for concern. The urban population is set to more than double by 2020. By that date, over 60 per cent of the region’s population will live in cities (Devarajan, Dollar & Holmgren, 2001). Taking into account the low living standards of city-dwellers and the limited development of urban infrastructures (see Table 3), the magnitude of the challenge to be met if cities are not to collapse into extreme poverty is clear for all to see. The preceding analyses lead to an unambiguous conclusion: over the long term, Africa’s poor economic performances have led to a fall in living standards and a growth in poverty of which city-dwellers have been the main victims.

However, the reliability of the macroeconomic data used to measure the evolution of the real well-being of the population is open to question. First, due to the ongoing deterioration of national statistical offices, official figures are to be viewed with caution. The international sources for Africa on which analyses are based are unreliable, incomplete, and often contradictory (World Bank 2000d).

Second, living standards and, to an even greater degree, poverty, are difficult phenomena to quantify. For example, an indicator such as per capita GDP gives no more than a very vague approximation of the purchasing power of individual households (Ghura, Leite, & Tsangarides, 2002). Faced with such limitations there is a great temptation to call into question the catastrophic vision evoked by official figures. The decline in living standards could merely be a product of the lamentable quality of national accounts.

Table 3. Development indicators in various African cities

However, by pooling over twenty surveys carried out over various periods in five African capitals (Abidjan, Bamako, Dakar, Antananarivo, and Yaoundé), the comparison of aggregated data from national accounts and the results of surveys on household consumption (two independent sources), widely confirm the conclusions described above (Transparency International 2004). On the one hand, in the five countries in question, and independently of their specific characteristics, the two kinds of sources converge.

While they do not present exactly the same nuances, the trends point in the same direction; this is true not only for the long-term recession which lasted until the mid-1990s, but also for the subsequent recovery. On the other hand, it seems that the major cities (in this case, the capitals) have been by far the worst affected by the chronically poor economic conditions. Thus, a converging swathe of data confirms the hypothesis according to which living conditions in Africa, especially in the urban context, have declined dramatically.

One explanation for the less pronounced decline in the living standards of people in the country might be that the rural economy is less dependent on market fluctuations, largely due to the low rate of monetisation of agricultural activities and the very large proportion of food products consumed in situ. But in spite of this, average purchasing power is still systematically higher – and poverty less widespread – in the cities than in the country.

A comparison of monetary poverty, based on data from surveys carried out in seven African countries shows that the incidence of poverty is twice as high in rural areas as in urban ones (Demery 1999), with a fairly high variations in ratio: from 1 to 1.2 in Nigeria in 1996, to 1 to 3 in Burkina Faso (1998), Uganda (1997), and Zimbabwe (1996). Sahn and Stiefel (2000) obtain figures even less favourable to rural areas based on non-income poverty indicators including quality of housing, possession of durable goods, and human capital in thirteen African countries (World Bank 2004).

Recession and poverty: case studies

The overall picture that has just emerged leaves no room for doubt about the catastrophic decline of the situation of households in African cities. However, in order to present this assessment more clearly, and analyse the dynamic, the extent, and the characteristics of poverty, we will now concentrate on the specific cases of a number of capital cities.

A first series of examples shows the marked increase of the incidence of poverty in three cities – Bamako, Abidjan, and Dakar (DIAL 2000). The perspective is then widened to include all the capital cities of the West African Economic and Monetary Union (WAEMU) region. Lastly, the example of Madagascar – in which, according to the criteria employed, poverty appears in various guises and affects distinct groups – sheds light on the complexity and gravity of the phenomenon (World Bank 2004).

Chapter Four

Conclusion and Recommendations

Policy design post evaluation

Designing and evaluating economic policies are two closely related processes. From an optimistic standpoint, the history of development policies can be seen as a learning process, with each new policy drawing lessons from the failures or disadvantages of former policies, in a more or less radical manner. For instance, following the financial crisis that hit the Latin American countries at the beginning of the 1980s, structural adjustment policies were drawn up as a remedy for the supposed shortcomings of the previous policies, based on the substitution of national production for imports (Bratton & van de Walle, 1997).

The need to restore fiscal balance and balance of payments figures not only led to the introduction of corrective macroeconomic stabilisation policies, but also to a concern to find more structural policies designed to reduce state intervention and promote exports. Worries about unequal growth patterns that had already been voiced at the beginning of the 1970s, when growth was high, resurfaced at the end of the 1980s and led to the reorientation of structural adjustment towards poverty reduction.

The Russian and Asian crises also called into question the role of state and non-state institutions in the workings of free market economies. They inspired greater prudence in reform policies, and the idea of more detailed evaluations of the policies was keenly promoted. As a result of this, the ‘new policies’ were based in part on a more or less precise a poste-riori or ex post evaluation of the previous policies. This critical assessment meant that new normative criteria could be put forward (private sector efficiency, poverty reduction, etc.), and new policies designed to improve these criteria could be proposed. This brought up the difficult issue of a priori or ex ante evaluation.

From an epistemological standpoint, the terms ex post and ex ante are not clear-cut or sufficient in themselves, because there is also a border between experimental evaluation and prospective evaluation, or between experiment and simulation. In fact, econometric theory is now relatively good at delimiting the gradation between direct experiment methods by random sample and those that provide a structural representation of the agents’ behaviour and enable prospective simulations.

Strictly speaking, ex post evaluation seeks to check retrospectively whether the (in principle clearly identified) objectives of a policy (already put into practice) have been met, with a positive approach. It comes close to a pharmacological type of question, such as: ‘is the drug effective?’ which suggests an experimental approach. However, experimental or pseudo-experimental evaluations cannot be applied to a great number of policies because it is practically impossible to form a control group (Decaluwé, Patry, Savard & Thorbecke, 1999). This is the case for all nontargeted or non-graduated policies and, even more so, for any policies that, although targeted, may have external or macroeconomic effects on the population as a whole.

Let us take the case of a policy fixing a minimum wage level and address the question of its impact on employment and poverty. This is typically a non-targeted policy, so the evaluation cannot be made by an econometrically controlled comparison of certain agents a priori concerned by the minimum wage and others who are not. From the moment that there is a single minimum wage, it is not possible either to establish a scale between agents to whom a certain level is applied and others to whom another level is applied, which would enable a similar comparison.

We can observe, this time over a period of time, how increases in the minimum wage affected employment in different labour categories, particularly for individuals whose wages were initially close to the minimum wage (then considered as a targeted group). We can also model the numbers of individuals leaving poverty and try to identify the impact of earned income received in t -1 on the state of poverty in t.

In this case, the economic situation applicable to the different periods must be carefully controlled, in order to avoid attributing economic factors that have no relation whatsoever to the level of the minimum wage to the variations in employment and income observed (Whaites, 2005). Furthermore, if the increase in minimum wage has a strong macroeconomic impact, whatever its characteristics, purely microeconomic evaluations can prove to be completely misleading.

Now let us take the case of a public works policy, offering low-paid jobs, for instance at half of the current minimum wage, and again look at the question of its impact on poverty. This time certain agents can be selected (or select themselves) and others not, and it is possible to control differences between the two groups, either a priori by selecting beneficiary agents amongst a population of candidates at random, or a posteriori, by jointly modelling participation in the programme and its impact on the participants’ income.

However, if the programme is sufficiently wide in scale, the choices made by the programme’s beneficiaries and the resulting redistribution of income have consequences for the rest of the labour market and on overall demand for goods and services. The two above examples demonstrate that, in economics, an evaluation process can be far removed from the pharmacological model of applying a treatment to a group of sick people. In many cases, both ex post and ex ante evaluations require representations of the workings of the economy as a whole.

Obviously, another problem with ex ante evaluation is that it involves making prospective rather than retrospective simulations. When a policy results in a significant, long-term change in the structure of a population, as, for instance, in the case of education or health policies (fighting Aids, for example), the prospective dimension is of fundamental importance.

In practice, ex ante evaluation is also required to compare several alternative policies. Whereas ex post evaluation is part of a positive process: ‘such and such a policy more or less achieved its objectives’, ex ante evaluation is closer to a normative approach: ‘this policy is better than that one’. In that case, the greater the political stakes, the more it is vital to ensure an independent evaluation (

Foster et. Al 2004). There are numerous examples where the theoretical structure of the models used postulates more or less explicitly that a recommended policy will have a positive impact, with their empirical application serving merely to quantify its scale. For instance, the structure of many computable general equilibrium (CGE) models built to assess structural adjustment or free trade policies tended to put the emphasis on static gains in efficiency obtained by reducing domestic or external taxes.

However, they minimised the potentially considerable short-term Keynesian impact of a reduction in public spending, the contractionary consequences of a rapid fall in the protection of the economy, or a certain number of dynamic longer-term gains in efficiency depending on the supply of public goods or the protection of local innovations. Conversely, policies in the 1970s were evaluated using fixed-price planning models which ignored the problems arising from a heavily distorted relative prices structure.

The success or failure of policy experiments therefore provides lessons not only for the new policies but also for the corresponding methods of evaluation. The first of these lessons is clearly modesty. The ‘participatory’ approach to drafting these policies results in more freedom of manoeuvre, which should hopefully encourage policies that are better suited to democratic demands, but will also lead to greater heterogeneity. The new methods of negotiating international aid, using medium and long-term schedules of objectives, should also provide enough time for a full evaluation of the results, stage by stage.

Putting the accent on poverty reduction fixes obvious targets for the evaluation methods. They must be capable of providing satisfactory estimates of the policies’ distributive impact, in terms of income distribution but also distribution of all the other primary factors of well-being such as access to public goods, health and education. This is highly ambitious and implies that the heterogeneity of the different populations must be taken into account far more than in the past, both in terms of differences in resources (income, capital, education, health), and in terms of the economic agents’ characteristic behaviour and the environments in which they are immersed.

Certain elements of stabilisation and structural adjustment policies that are still applied must be examined, no longer just in macroeconomic and financial terms or in terms of efficiency, but also in distributive terms. For instance, international institutions now tend to accept that public finance or monetary policies that could possibly lead to a significant increase in poverty are no longer tenable and that, in cases where they are nonetheless considered inevitable, they must be backed up by palliative policies such as safety nets, tariffs or subsidies to protect certain vulnerable groups. Suitable evaluation methods must therefore be capable of simulating the distributive impact of macroeconomic policies at the same time as the impact of targeted palliative policies.

However innovative they may be, labour market policies generally raise the same type of difficulties. Hence, labour-intensive public works programmes or workfare policies must be evaluated taking into account the heterogeneity of the agents’ behaviour regarding labour supply and their job opportunities, but also the potential macroeconomic effects of these interventions.

The same is true in this respect for more traditional policies for farm prices or wage policies. In addition, policies dealing with domestic and external indirect taxes must take into account heterogeneous consumption habits and the existence of an informal sector with little taxation (Ghura, Leite & Tsangarides, 2002). Policies for education, health and, more generally, the provision of public goods (e.g. water supplies) call on slightly different methods. First of all, it is not possible to start from an income and prices generation model as certain fairly specific elements of behaviour must be taken into account.

Conclusion

The domestic implications of debt relief are apparently far reaching. Objections to an IMF-supported suspension of payments are based on the moral hazard problem and on disapproval of interfering with the efficient operation of the market. It has been argued that such a proposal could distort incentives and lead to excessive borrowing. However, it seems unlikely that countries would take excessive risks because of the possibility of a debt moratorium given the extremely painful consequences for a country which experiences a crisis.

Furthermore, as Cogneau and Grimm 2002 point out, a suspension of payments would reduce the moral hazard that encourages lending by financial institutions that expect to be bailed out by an IMF-led rescue. Here, however, the danger would be that IMF-sanctioned moratoria might throw out the baby of capital flows to emerging markets in general with the bath water of more speculative or less sustainable flows (Cogneau et al. 2002).

There are also objections to debt moratoria based on the argument that the cost of capital could rise for all borrowers if they were used too often. Moreover, the issue of contagion implies that the involvement of the private sector in the resolution of the problems of one country could lead to capital outflows from other countries (Gautier 2002). Reportedly, the possibility of an IMF-supported orderly workout in one Asian country was not adopted because of fears that the crisis could spread to other regions. Such fears may always be there and inhibit the use of IMF-sanctioned payment standstills.

Despite the inherent difficulties, however, the international community is agreed that there is a need for new procedures for the resolution of crises in heavily indebted countries. The experience of the 1980s and 1990s has shown that, with the changes to the character of international financial markets, the existing mechanisms for crisis resolution are no longer adequate.

The principal benefit from the establishment of orderly workout procedures would be that priority could be given to dealing with the domestic implications of a crisis, rather than to paying back investors and creditors. This would be of particular value in situations where the basic fundamentals of the country concerned are sound, and the problem is more one of illiquidity than insolvency. In such cases, the weakness is usually a structural problem, such as high debt or a weak banking system, which will take time to be corrected.

Resources

Atkinson, A.B. and Stiglitz, J.E. (1989) Lectures in Public Economics, New York: McGraw-Hill International Editions.

African Development Bank (2005) Achieving the Millennium Development Goals in Africa. Progress, Prospects, and Policy Implications, Global Poverty Report 2002, in collaboration with the World Bank, Abidjan: African Development Bank.

Booth, D. and Lucas, H. (2002) Good Practice in the Development of PRSP Indicators

Booth, D. (eds) (2002) Fighting Poverty in Africa: Are PRSPs Making a Difference? , London: Overseas Development Institute for the Strategic Partnership with Africa.

Bratton, M. and van de Walle, N. (1997) Democratic Experiments in Africa. Regime Transitions in Comparative Perspective, London: Cambridge University Press.

Berthélémy, J.C. and Söderling, L. (2004) Fighting Poverty in Africa: Are PRSPs Making a Difference? , London: Overseas Development Institute for the Strategic Partnership with Africa.

Cornia, G.A., Jolly, R. and Stewart, F. (eds) (1987) Adjustment with a Human Face: Protecting the Vulnerable and Promoting Growth, Paris: Oxford University Press/UNICEF.

Cogneau, D. and Grimm, M. (2002) The Distribution of AIDS over the Population in Africa: Hypothesis Building from Individual Answers to a Demographic and Health Survey with an Application to Côte d’Ivoire, Working Paper DT2002-02, March, Paris: DIAL.

Cogneau, D., Maurin, E. and Pasquier, L. (2001) Social Mobility in Africa, a Comparative Analysis of Five sub-Saharan Societies (mimeo), Presented at the meeting of the RC28 group of ISA in Mannheim, May.

Decaluwé, B., Patry, A., Savard, L. and Thorbecke, E. (1999) Poverty Analysis within a General Equilibrium Framework, CREFA WP 9909, Québec: Université Laval.

Devarajan, S. and Go, D.S. (2001) A Macroeconomic Framework for Poverty Reduction Strategy Papers (with an application to Zambia) (mimeo), Washington, DC: World Bank.

Devarajan, S., Dollar, D. and Holmgren, T. (2001) Aid and Reform in Africa; Lessons from Ten Case Studies, Washington, DC: World Bank.

Demery, L. (1999) Poverty Dynamics in Africa: An Update, Poverty Reduction and Social Development Unit, Washington, DC: World Bank.

DIAL (2000) The 1-2-3 Survey: A Poverty Monitoring Tool, May, Paris: DIAL.

Foster, M., Fozzard, A., Naschold, F. and Conway, T. (2004) How, When and Why does Poverty get Budget Priority; Poverty Reduction Strategy and Public Expenditure in Five African Countries, Synthesis Paper, Working Paper 168, London: Overseas Development Institute.

Ghura, D., Leite, C.A. and Tsangarides, C. (2002) Is Growth Enough? Macroeconomic Policy and Poverty Reduction , IMF WP 02/118, July, Washington, DC: International Monetary Fund.

Hjertholm, P. (2000) ‘Analytical History of Heavily Indebted Poor Country (HIPC) Debt Sustainability Targets’, University of Copenhagen, Institute of Economics Discussion Paper 00/03.

Hjertholm (1997) ‘Wage Indexation and the Effect of Inflation Uncertainty on Employment: An Empirical Analysis’, American Economic Review 76 (2):235-43.

Joseph, A. (2003) Resolving the Debt Problem: From the HIPC Initiative to Cologne, Technical Document 163, August, Paris: OECD Development Centre.

Stiglitz, J. (1974) ‘Alternative theories of wage determination and unemployment in LDCs: the labor turnover model’, Quarterly Journal of Economics, 88(2): 194-227.

Stiglitz, J. (2002) Globalisation and its Discontents , London: Penguin.

Severino, J.-M. (2001) ‘Les fondements stratégiques de l’aide au développement au XXIe siècle’, Critique Internationale, 10(1): 75-99.

Snerch, S. (1994). Synthesis of the study on long-term prospects in West Africa, Paris: OECD/ADB/CILLS.

Tarp, F. (ed.) (2000) Foreign Aid and Development; Lessons Learnt and Directions for the Future, London and New York: Routledge.

Thin, N., Underwood, M. and Gilling, M. (2001) Sub-Saharan Africa’s Poverty Reduction Strategy Papers from Social Policy and Sustainable Livelihoods Perspectives, A report for the Department for International Development, Oxford: Oxford Policy Management.

Transparency International (2004) Annual Report 2000, Berlin: Transparency International.

UNAIDS (2000) AIDS Epidemic Update , Joint United Nations Programme on HIV/AIDS, Geneva: World Health Organisation, December.

UNCTAD (2002) The Least Developed Countries Report 2002; Escaping the Poverty Trap, New York and Geneva: United Nations.

Vandemoortele, J. (2002) ‘Are the MDGs feasible?’, UNDP, Bureau for Development Policy, New York, Paper presented to the ABCDE-Europe conference organised by the World Bank in Oslo, 24-26 June.

Whaites, A. (2005) PRSP: Good News for the Poor? Social Conditionality, Participation, Poverty Reduction, Monrovia: World Vision International.

Whaites, A. (2004) Masters of their Own Development? PRSPs and the Prospects for the Poor , Monrovia: World Vision International.

World Bank (1980) World Development Report 1980, Washington, DC: World Bank.

World Bank (1981) Accelerated Development in Sub-Saharan Africa: An Agenda for Action, Washington, DC: World Bank.

World Bank (1990) World Development Report 1990: Poverty, Washington, DC: World Bank.

World Bank (1994) Adjustment in Africa: Reforms, Results and the Road Ahead, World Bank Report on Development Policy, Washington, DC: World Bank.

World Bank (1995) A Continent in Transition: Sub-Saharan Africa in the Mid-1990s , Africa Region, Washington, DC: World Bank.

World Bank (1997) World Development Report 1997: The State in a Changing World, Washington, DC: World Bank.

World Bank (1998a) Assessing Aid: What Works, What Doesn’t and Why, World Bank Policy Research Report, New York: Oxford University Press.

World Bank (1998b) Urban in the Regions. Snapshot of Sub-Saharan Africa: Regional Urban Statistics and Projections (mimeo). Available online: http://www.worldbank.org/external/urban/urbanDev (accessed on 27 April 2006).

World Bank (1999) Core Welfare Indicators Questionnaire: Handbook and CD ROM, Washington, DC: World Bank.

World Bank (2000a) World Development Report 2000/2001: Attacking Poverty , New York: Oxford University Press.

World Bank (2000b) Global Economic Prospects and the Developing Countries, Washington, DC: World Bank.

World Bank (2000c) Can Africa Claim the 21st Century ? Washington, DC: World Bank.

World Bank (2000d) The Quality of Growth, New York: Oxford University Press.

World Bank (2001a) Poverty Reduction Strategy Sourcebook: Chapters and Related Materials, April. Available online: http://www.worldbank.org/poverty/strategies/chapters (accessed on 27 April 2006).

World Bank (2001b) Global Economic Prospects and the Developing Countries, Washington, DC: World Bank.

World Bank (2001c) African Forum on Poverty Reduction Strategies, Quotations from Day 1. Available online: http://www.worldbank.org/wbi/attackingpoverty (accessed on 27 April 2006).

World Bank (2001d) World Development Report 2002: Building Institutions for Markets , New York: Oxford University Press.

World Bank (2001e) RMSM-X for PRSPs. A New Module to Measure the Impact of Macroeconomic Policies and Programs on Poverty and Income Distribution. A Research Proposal (mimeo), Washington DC: World Bank.

World Bank (2002a) The Role and Effectiveness of Development Assistance: Lessons from World Bank Experience, Research Paper from the Development Economics Vice-Presidency of the World Bank, Washington, DC: World Bank.

World Bank (2002b) Overview of Poverty Reduction Strategies, Washington, DC: World Bank. Available online: http://www.worldbank.org/poverty/strategies/overview.htm (accessed on 27 April 2006).

World Bank (2002c) World Development Report 2003: Sustainable Development in a Dynamic World: Transforming Institutions, Growth and Quality of Life , New York: Oxford University Press.

World Bank (2002d) World Development Indicators 2002, Washington, DC: World Bank.

World Bank (2004) A User’s Guide to Poverty and Social Impact Analysis (mimeo), Washington DC: World Bank.