Blinding with Science or Encouraging Debate?

19 September 2002 | Report

Alex Wilks and Fabien Lefrançois

Copyright © 2002 by the Bretton Woods Project.

Published by the Bretton Woods Project and World Vision International.

Conditions of use:

The text of this publication may be freely used for brief quotations provided the source is credited. Please send a copy of the publication to the Bretton Woods Project.

Non-profit organisations may reproduce the report for educational purposes without prior authorisation. To get free hard copies contact the Bretton Woods Project info@brettonwoodsproject.

Also available in PDF format

Spanish version (Word)

French version (Word)

Comments welcome. Send to Fabien Lefrançois



Funding for some of the research, design and printing costs was kindly provided by World Vision.

Initial research for this briefing was conducted by Angela Wood. Useful research assistance was also provided by Demelza North. The ‘Governance’ section, pp. 27-28, draws heavily on the work of Elizabeth Drake, Rasha El-Habashy, Ioanna Kotsioni, Ambreen Malik, Vivek Misra, and Ying Xu.

Helpful comments on a draft of this briefing were received from Nancy Alexander, Charlotte Carlsson, Kelly Currah, Lollo Darin-Ericson, Caroline Harper, Barbara Hendrie, Tim Kessler, Paul Ladd, Max Lawson, Elena McCollim, Simon McGrath, Jennie Richmond, Alan Whaites and two World Bank reviewers who wish to remain anonymous.

Responsibility for the final text rests, of course, solely with the authors.

Further comments are welcome. We may post selected comments (with permission) on the Bretton Woods Project’s website. Send to blinding@brettonwoodsproject.org.

About the Authors

Alex Wilks is Coordinator of the Bretton Woods Project and Fabien Lefrançois is its Policy Analyst. They have both written extensively on the World Bank and IMF, and have been actively engaged in analysing and campaigning on the Bank’s and IMF‘s activities for many years

List of acronyms

CAS = Country Assistance Strategy
CSO = Civil Society Organisation
DEEP = Distributive Effects of Economic Policy
DFID = Department for International Development
DTIS = Diagnostic Trade Integration Study
ESW = Economic and Sector Work
HIPC = Heavily Indebted Poor Country
IDA = International Development Association
IF = Integrated Framework
IGR = Institutional and Governance Review
IMF = International Monetary Fund
ITC = International Trade Center
ISODEC = Integrated Social Development Centre
LDC = Least Developed Country
NGO = Non-governmental Organisation
PEAP = Poverty Eradication Action Plan
PRGF = Poverty Reduction and Growth Facility
PRS = Poverty Reduction Strategy
PRSC = Poverty Reduction Support Credit
PRSP = Poverty Reduction Strategy Paper
PSIA = Poverty and Social Impact Analysis
RMSM-X = Revised Minimum Standard Model – Extended
SAPRI = Structural Adjustment Participatory Review Initiative
UNCTAD = United Nations Conference on Trade and Development
UNDP = United Nations Development Programme
WB = World Bank
WTO = World Trade Organisation

Executive Summary

“Far from abandoning aid conditionality, international financial institutions are collaborating to retool the aid regime under the rubric of ‘ownership’ and aid effectiveness. Aid has become increasingly technocratic, with an overwhelming reliance on donor systems of aid management and accountability, implemented by a host of consultants and advisors. The World Bank reports that some 100,000 foreign experts are currently employed in Africa, tending to displace local experts and weaken capacity.” (The Reality of Aid, 2002, p. 8)

“Increasingly, national NGOs and CSOs are expressing concern that the new agenda around PRSPs has failed to deliver both a different way of doing business and a real dialogue around a broader choice of policy options. It is a matter of concern if NGOs and CSOs become disenchanted and disengage at an early stage in the formulation of a PRSP.” (UK Department for International Development, 2001, p. 3)

Despite some pronouncements from Washington and elsewhere, debates about privatisation, liberalisation and other core policy issues are alive and well. Discussions of trade policies at Doha, sustainable development in Bali and water ownership in Ghana demonstrate this clearly. Yet the World Bank and its sister agency, the International Monetary Fund (IMF), often give the impression that there is consensus on the development agenda, and that only details remain to be worked out. The Bank is pursuing this logic through studies in its client countries on a wide range of policy issues. These studies may be more influential than many outsiders realise and partly explain why the Poverty Reduction Strategy (PRS) process has not ushered in real debates about macroeconomic policies.

In 1999 the World Bank and IMF introduced the PRS process, which is supposed to improve donor coordination and ensure that governments and civil society groups take the lead in defining policies that the Bank and Fund should support. But many commentators have complained that macroeconomic policy choices have not been adequately debated and that few countries have deviated from standard choices (World Vision, 2002; DFID, 2002). This could be for a number of reasons. The main one often put forward is that the World Bank and IMF hold the purse strings, so they can dictate what a cash-strapped government signs up to in exchange for loans. But the Bank responds that it has moved away from a one-size-fits-all approach to policymaking, partly through decentralising some of its operations and opening up to consultation exercises. Thus understanding why the Poverty Reduction Strategy Paper (PRSP) exercises are proving so dissatisfying for some NGOs and others requires a more detailed examination of the mechanisms by which the Bank continues to promote policy choices.

The Bank most of the time no longer has to rely on its financial clout alone, as it is winning arguments upstream. Through its global and national-level studies, and its extensive network of official, journalist and academic contacts, the Bank has a strong influence on policy debates even where it is not lending. But the Bank is clearly in the most influential position where it can combine its ‘knowledge’ and lending functions, imposing conditions to support its advice.

A recent study on the Uganda PRSP process found that the multi-stakeholder discussions led to very little substantive policy change in key areas. The conditionality in the Bank’s new loans to Uganda did not match the conclusions of the PRSP discussions but appeared to come from an inside track of analysis and discussion (Nyamugasira & Rowden, 2002).

This briefing argues that in many cases like this, the World Bank’s ‘Economic and Sector Work’ – the studies it carries out or commissions in its client countries – may be the source of official views which dominate the policy process at the expense of civil society inputs. The Bank’s studies range from overviews of public expenditure or poverty statistics, to detailed analyses of particular sectors or institutions.

The briefing aims to inform interested parties – particularly NGOs in the South – about the World Bank’s in-country analysis; to examine some concerns about its scope and content; and to discuss some possible strategic approaches. It sets out the areas in which the Bank is carrying out assessments, and ways that NGOs and independent researchers could do more to shape or contest these in areas they find important. It follows other work by the Bretton Woods Project in raising concerns about the World Bank’s increasing “knowledge” work, viewed by many as limited in scope and crowding out others (Bretton Woods Project, 2001; Hildyard, 1998), and about the PRS process (Bretton Woods Project, 1999 & 2000).

Branislav Gosovic of the South Center recently commented: “…global intellectual hegemony should be of special concern to developing countries. Their intellectual dependency means that they tend to rely wholly on a handful of sources in the North for data, analysis, explanation, policy and prescriptions, including in relation to their own national development.” (Gosovic, 2001; p. 135). The Bank’s central role in defining and promoting development orthodoxy is well-documented (George & Sabelli, 1994; Wade, 1996 & 2001). The Bank is very widely respected despite its institutional biases, which frequently lead it to make serious errors in prediction. Partial recognition of errors in predicting debt burdens and commodity prices, for example, has meant that the debt relief obtained under the Highly Indebted Poor Country scheme has proved far from adequate – see table 1 (Jubilee Research, 2002).

The launch of the Structural Adjustment Participatory Review Initiative, and the contributions to the recent World Bank/IMF review of the PRSP process, show that the current approaches for policy analysis and decision-making are not working for all stakeholders. The Bank and Fund recognise some problems with their approaches to PRSPs, but their responses to the PRSP review seem fairly complacent. The IMF‘s role in conducting studies and setting conditions is extremely important alongside that of the World Bank, however this briefing focuses mainly on the Bank.

Many officials and NGOs now pin hopes on the introduction of Poverty and Social Impact Analysis (PSIA). The Bank defines this as “analysis of the distributional impact of policy reform on the well-being or welfare of different stakeholder groups, with particular focus on the poor and vulnerable” (World Bank, 2002a). This is supposed to encourage the use of more diverse analytical approaches as well as greater transparency and accountability in the process of making policy recommendations. The contours of this PSIA are ill-defined. But the introduction of any new exercise or approach at the World Bank needs to be set in the context of previous disappointments. The 1990s saw the launch of a whole series of new Washington initiatives and accompanying acronyms. Many, such as the Participation Flagships and the Country Assistance Strategies and the Environment initiative, had little lasting impact.

Ironically, the Bank is increasing the number of assessments of institutions in developing countries while it suffers from major institutional weaknesses itself. These are not just the perennial ones of political capture by the richest countries and resistance to calls for full accountability and transparency. The reforms initiated by President Wolfensohn in the last seven years have improved some matters, but left others untouched and caused some new problems. The Bank is still more focused on making new loans than on the impacts on poverty, and its internal market management system has caused many tensions.

The Bank constantly talks about capacity-building and listening, but seems reluctant to cede control of policy formulation processes or to recognise contributions or perspectives that diverge markedly from core Washington thinking. The production of the ‘Poverty’ World Development Report, the refusal to engage with the country studies produced by the Structural Adjustment Participatory Review Initiative, and the introduction of its new Private Sector Development Strategy testify to this. Indeed, the Bank often appears to imply that there is a vacuum out there that it needs to fill: that few others are doing serious policy analysis (World Bank, 2001h).

The World Bank should be far more open about how it is commissioning research and the methods being used. But beyond this, more must be done to break the Bank’s near-monopoly on development analysis by diversifying the commissioners and producers of research. Even if the Bank were to make a major shift towards more open and heterodox research, there would still be many who would not see its conclusions as legitimate and even-handed. Ministers from indebted countries and prominent academics have recently voiced concerns about the conflicts of interest underlying the Bank’s role as analyst and lender (HIPC Ministerial Declaration, 2002, p. 4; Wade, 2002).

Official agencies should be more sensitive to charges that they are crowding out analysis by independent organisations and networks. NGOs considering whether and how to engage in Poverty Reduction Strategy processes should find out more about the World Bank’s analytical work in their country or sector. Then they can consider how to influence the ways in which this research is designed and conducted, or how to use or challenge its findings in their advocacy. The PRS process is often said to be about putting the borrower country governments in “the driver’s seat” of a metaphorical car. But even if governments are allowed to get hold of the steering wheel, it is vital to see who is in charge of producing the maps and deciding on navigation.

This briefing is structured in sections around the following sets of questions:

  • Section 2: What is the World Bank’s Economic and Sector Work? What issues does it cover, how is it carried out and what problems have been identified with it?
  • Section 3: What are the key biases exhibited by World Bank research and analysis, both in general and in particular policy areas (investment climate, trade, governance, poverty measurement and macro-economic modelling)?
  • Section 4: What can be done by official agencies and civil society groups to change World Bank approaches and to support the strengthening of independent ones?

Table 1: 2001 export growth much lower than World Bank projections

Projected growth for 2001 (%)
Revised figure based on actual growth in 2001 (%)
Average for all 24 HIPC Countries

World Bank assessments: What issues, whose voices?

“The World Bank has enormous influence over the shape and pace of Indonesia’s policies and reform in its own right, but also through its production of the economic analysis that serves as the information base on which other creditors and donors rely to make decisions.” (INFID, 2002, p. 1)

“If the PRS process were a government-led process, why would the Bank and Fund send numerous missions to the country to develop the PRS? Why would the Bank develop a 1,000-page Sourcebook to tell developing country groups how to create a PRS?” (Abugre, 2000)

“In attributing ignorance to others, development’s higher-level protagonists ensure that they themselves remain ignorant of others’ knowledge. This reinforces a further kind of ignorance: that of the local background conditions for their own knowledge.” (Lohmann, 1998, p. 2)

The importance of World Bank country studies

In theory, the World Bank is now prepared to let borrowing country governments set their own policies. In practice, it goes to great lengths to influence their choice. Its role, in combination with the IMF‘s, in setting conditions in return for its loans is well-known. The World Bank is also widely recognised as perhaps the most prominent global development think-tank, issuing studies and data on many topics. Less clear to many is how the “lending Bank” and the “knowledge Bank” fit together. This section examines the Bank’s approach to analysis, which forms the basis of its policy conditions.

In every country where it lends, the Bank conducts or commissions a range of studies, covering issues from public spending to tariff reduction, from primary education to natural resource protection. The studies make recommendations that are often taken very seriously by borrower government officials negotiating funding with the Bank and with other aid agencies. Yet these studies are produced without the involvement of many people in governments and civil society groups who are now supposed to be key stakeholders in the Poverty Reduction Strategy process.

The Bank has long conducted analysis in its client countries. The analysis acts as a building block for World Bank Country Assistance Strategies (CASs), which are documents produced every three years setting out the Bank’s programme for each country and policy changes the country must make in exchange for Bank support. The Bank’s so-called “Economic and Sector Work” (ESW) analysis examines economic policy, institutions and structures, and specifies reforms. A briefing from Bank Information Center uses a medical analogy to describe the CAS and the ESW: the latter is the diagnosis that results in the Bank’s prescription for the country, outlined in the CAS. If the cures put forward by the World Bank are limited, it may be because they are based on a narrow range of tests (Bank Information Center, 1999).

Paradoxically, since the introduction of the PRSP process – which is supposed to devolve policymaking to developing countries – the Bank is increasing its output of such studies (IBRD/IDA, 2001, page iv) . And the Bank’s assessments are growing not just in number but in importance. The Bank’s country studies feed into its Country Policy and Institutional Assessment, an exercise that rates countries and determines how much money they will receive. The rating is based on 20 criteria, covering economic management, structural policies, policies for social inclusion, public sector management and institutions. Under this system the Bank rates the extent to which governments comply with the creditor community’s definition of “good policies”, and then allocates more funds to governments that score well. In 2001 the Bank allocated almost five times more money to governments that achieved its “A” rating compared to those rated “F” (Globalization Challenge Initiative, 2002). Some other donors also follow the Bank’s rating system when deciding how to allocate their aid spending.

The Bank argues that it must carry out certain assessments to ensure that its funds are properly used and will be repaid. These include reviews of public expenditure, financial accountability and procurement (IBRD/IDA, 2001, Annex 3, p. 29). But these “fiduciary assessments” are complemented by an ever-widening range of Bank studies about particular issue areas. These have a confusing set of names and overlapping remits (see Box 1, below) and clearly take the Bank into areas where other organisations, such as UN specialised agencies and national research institutes, are seen to have more expertise and legitimacy. It is important to question whether the Bank should be doing studies in these areas, and, if so, who it should work with to set questions, gather data and finalise the policy conclusions. Does Bank research question the fundamental appropriateness of policies, or does it seek to “fine-tune” a pre-determined set of policies? Is the analysis of cross-cutting issues such as gender and the environment best served by stand-alone studies, or by integration into other analysis?

Box 1: Assessment overload

The World Bank conducts two main categories of analytical report in its client countries: core reports which are national in scope and carried out for all countries, and sector- or issue-specific reports which are done only in selected countries. The Bank also produces some regional reports and less formal policy notes, often resulting from workshops and conferences.

The main types of World Bank in-country analytical reports are:

Core reports

Poverty assessments – Aim to provide in-depth analysis of poverty issues, and evaluate the effects of economic and social policies on the poor.

Country economic memoranda or Development policy reviews – Provide an overall assessment of a country’s economic and sectoral policies and development path. Development policy reviews are gradually replacing Country economic memoranda.

Public expenditure reviews – Analyse the equity and efficiency of public expenditure and assess the effectiveness of public expenditure management processes in achieving fiscal discipline and enabling cost-effective public service provision.

Country financial accountability assessments – Diagnose a country’s private and public financial management systems, assess the strengths and weaknesses of public sector accountability arrangements and identify risks any weaknesses pose to the use of Bank funds.

Country procurement assessment reviews – Analyse a country’s public sector purchasing procedures, and establish the need for an action plan to improve a country’s system for procuring goods, works and services.

Sector or issue reports

  • Institutional and governance review
  • Social protection, health, and education sector review
  • Rural development assessment
  • Social analysis
  • Country gender assessment
  • Country environmental analysis
  • Financial sector assessment
  • Investment climate assessment
  • Diagnostic trade integration study
  • Country infrastructure framework report
  • Corporate governance assessment
  • Energy-environment review
  • Financial stability assessment
  • Economy-wide assessment
  • Review of spending priorities
  • Ownership assessment

Sources: World Bank, 2001i, and www-wds.worldbank.org

Opening debate or pushing conclusions?

Despite the fact that their views may be eclipsed by the Bank’s studies, few civil society organisations seem to know much about them, still less to be involved in them in a substantive way. A number of groups have been involved in or are aware of some of the Bank’s Poverty Assessments, many of which have deliberately aimed to be participatory. But a recent World Bank evaluation recognised that “there has been less participation in Public Expenditure Reviews and other aspects of economic and sector work”. (World Bank, 2001a, p. 62) Other official Bank reviews also recognise that their analytical work has not changed sufficiently in line with the principles of ownership and participation, which are supposed to underlie the Poverty Reduction Strategy (PRS) approach (World Bank, 2001a, p. 62 & World Bank, 2000b).

The Bank has often been criticised for its dogmatic approach to policymaking. Many parts of the Bank still show little interest in listening to other perspectives. The resignation of Ravi Kanbur as lead author of the World Development Report, the early departures of Joe Stiglitz as Chief Economist and William Easterly as a senior economist illustrate the difficulties facing alternative perspectives. Bank reports frequently make reference to the “right policies” for countries, or a “good policy environment” as if these can be conclusively assessed and asserted (World Bank, 1998b). The Bank very often appears to be pushing policy ‘solutions’ rather than contributing to a debate and listening to diverse evidence and arguments. Indeed the Bank is often reluctant to listen to opinions even from some of its own researchers. David Ellerman, formerly senior adviser to the Bank’s Chief Economist, wrote recently in the Bank’s Staff Association Newsletter of the danger that in such an atmosphere “Experimentation, debate, and the exercise of critical reason are curtailed to stay within the safe boundaries of Official Wisdom” (Ellerman, 2001, p. 3).

We question whether the World Bank should undertake all this analysis. The studies it continues to conduct should be reoriented to help different stakeholders to engage more meaningfully in debates about policy options. Analysis commissioned by the Bank should be defined and produced in an open manner and, ideally, be carried out by national researchers and civil society groups. The studies should aim to review evidence and set out issues and options for debate, not resolve debates and make firm proposals. The IMF and the Bank have agreed in principle that they should provide menus of potential reforms from which the country can select. It remains to be seen whether the Bank will work with others to offer a diverse and tailored range of dishes for countries to select from, or whether the menu will continue to be formulated very narrowly according to the tastes of Bank staff.

Poverty and Social Impact Analysis: what’s new?

The Bank, under pressure from NGOs and some governments, has announced new approaches, in particular Poverty and Social Impact Analysis (PSIA). In the UK, Oxfam and the UK‘s Department for International Development are among those who have recommended that the Bank make available further analysis of the likely poverty impact of proposed policies to help foster an informed national debate about the macro programme, and make explicit and accessible the economic logic underpinning policy proposals (Oxfam International, 2001; DFID, 2001, p. 3). The World Bank and IMF have made commitments to implement some further poverty and social analysis and in April 2002 the Bank produced a draft User’s Guide to Poverty and Social Impact Analysis (World Bank, 2002a). This is a lengthy document describing different methodologies for evaluating individual reforms as well as economy-wide changes. The techniques described range from rapid qualitative analysis to complex data-intensive spreadsheets and models. Box 2 below briefly outlines some of these tools.

But it is still unclear what PSIA is and how it fits with existing work by the Bank, IMF and others. Is PSIA intended to outline likely impacts of reforms before they are agreed and implemented, or to help evaluate the effects of previous reforms? Is the User’s Guide to PSIA just a shopping list of methodologies, or will it lead to Bank staff really changing the type and nature of the studies they conduct or commission? Will it persuade them to listen more to analysis produced by different organisations? Will PSIA‘s impact be limited to changing the timing or sequencing of reforms, or will it encourage a first principles examination of what are the appropriate policies?

The Bank has stated that “over time, it is expected that social and environmental analysis would increasingly be integrated into a systematic review undertaken as part of the upstream preparatory work of the PRSP” (World Bank, 2001b, p. 11). The IMF has pledged that “the distributional impacts of major macroeconomic or structural reforms should be considered and reported on in PRGF [Poverty Reduction and Growth Facility] documents together with any countervailing measures to offset the impact of these reforms on the poor” (IMF 2002, p. 39). But it is worrying that in the same document, the IMF goes on to claim that more than half of its programmes in low-income countries already refer to some sort of poverty and social impact analysis and two-thirds include measures to offset the impacts of reforms on poorer people.

In the context of the PRS process, assessing the anticipated benefits and costs of a menu of policy options for consideration by citizens, the government and others could enable greater engagement of civil society in decision-making processes, leading to better policy design and ownership. Open deliberation with multiple stakeholders now appears to be welcomed in principle by everyone from the World Bank President to academics and NGO staff. It can also help ensure that Bank and Fund staff have to argue in detail what they believe will be the results of the reforms they are suggesting, not just present aggregate data and economic hypotheses borrowed from textbooks or other countries. Coupled with an effective process to monitor implementation, up-front impact assessments could help hold creditors and donors to account and enable the discussion of changes to programmes when significant negative social outcomes occur.

Who will conduct the analysis?

It is unclear whether some existing Bank Economic and Sector studies will be renamed Poverty and Social Impact Analysis or whether there will be a new approach. If it is the former, it will clearly still be the World Bank or the IMF doing the analysis. Rosalinda Quintanilla of the World Bank says, “We can’t do good ESW without locals – I mean government and technical people from the country. Generally, it’s not true that the World Bank knows better than nationals, local people know what’s best for them. But the World Bank knows a great deal about the international experience which gives it a unique comparative advantage in the analysis it undertakes in collaboration with counterparts. It is the joint effort, the leveraging of local and international knowledge, that brings the highest quality and best analysis for consideration of the policy makers.” (Quintanilla, 2001). However, Jim Stephens, who works on ESW in the World Bank’s Operational Policy and Country Services central team, clarified: “In most cases ESW represents just the Bank’s view. It has corporate sign-off and is intended to influence the policies of client countries or other donors working in the client country. In specific cases the contents of ESW are discussed with the government and made publicly available” (Stephens, 2002).

Ministers in the HIPC Finance Ministers Network (which brings together ministers from 33 Heavily Indebted Poor Countries) recently urged the World Bank and IMF to “dramatically accelerate PSIA in HIPCs, [since] analysis of the links between macroeconomic and structural policies and poverty reduction remains among the weakest areas of most PRSPs”. They were clear, however, that they did not mean that the World Bank and IMF should do more, or better, studies. They argued that “it is essential to equip countries with the tools to conduct their own PSIAs rather than depending on outside assistance. These tools should have input from the Bretton Woods Institutions and donors, but be administered and disseminated by independent capacity-building sources, to avoid conflict of interest for partners in the negotiation process of PRGF and PRSC [Poverty Reduction Support Credit] frameworks” (HIPC Ministerial Declaration, 2002, p. 4).

The heads of the IMF and World Bank replied saying that they understood “the need for a broader and deeper discussion with all stakeholders of macroeconomic frameworks and policies, including on current policy choices and trade-offs. A concerted international effort will be required to assist the countries in undertaking more systematic PSIAs of major policies. The Bank and the Fund in cooperation with other partners are committed to help provide the necessary technical and financial support” (Köhler & Wolfensohn, 2002, p. 2). This raises important questions of who will fund, commission and carry out this work, but implies that the Bank and Fund have accepted in principle the need for it to be done independently, not by themselves.

Some have argued that although it is important for the World Bank to stop doing all this analysis by itself and allow governments to take the lead, there is some concern that the Bank and Fund might try to distance themselves from independently-produced research findings (Oxfam International et al., 2001, p. 4). Thus it is proposed that impact analysis be carried out by teams involving the Bank and Fund, the government, and civil society representatives (WWF, 2000). If PSIA is to be a tool to enhance the accountability of the World Bank and IMF, this may be appropriate, but if the intention is to enhance in-country capacity and responsibility for analysing options, this may be less successful.

What will be assessed using which methods?

The Poverty and Social Impact Analyses currently being piloted (separately) by the Department for International Development and the World Bank are focusing on specific policy actions that are expected to have particularly significant social impacts. These include: tax increases, subsidy reforms, exchange rate changes, civil service downsizing, energy price reforms and the size of the fiscal deficit. The Bank has stated that it proposes to introduce analyses of country policies and institutions and their capacity to mitigate adverse effects of reforms. This would apply to “all key sectors, including agriculture, education, energy, forestry, health, mining, social protection, transport and water” (World Bank, 2002e, p. 10).

The PSIA User’s Guide contains such a wide range of possible methods that many Bank staff will be able to say that they are already complying with its intentions. A former World Bank staff member has described the Users’ Guide as “a huge and rambling guide containing a hodge-podge of qualitative and quantitative methodologies” (World Bank Interviewee A, 2002). And Bank management is very reluctant to force its staff to use the PSIA User’s Guide or indeed to introduce any new mandatory procedures. Thus it is likely that there will be only a slow uptake of new approaches by Bank staff, who face many competing incentives and initiatives (Bosshard, 2002). A former Bank consultant concludes that: “there is nothing about PSIA which forces analysts to question policy from the outset. PSIA does not change institutional biases towards orthodoxy within the Bank itself”.

It is still far from certain that the Bank’s current PSIA initiative will mark a break from narrow, extractive and technical approaches to understanding poverty and planning policy responses. However, the comment by the heads of the World Bank and IMF in reply to HIPC Finance Ministers is encouraging. They stated that the recent reviews of the PRSP/PRGF “underscore the need for a broader and deeper discussion with all stakeholders of macroeconomic frameworks and policies, including on alternative policy choices and trade-offs within the overall poverty reduction strategies. The Bank and Fund, in cooperation with other partners are committed to help provide the necessary technical and financial support to low-income countries in undertaking PSIA” (Köhler & Wolfensohn, 2002, p. 2). But the detail of what Bank and Fund staff are bringing forward does not meet these objectives; still less do they achieve what Andrea Cornwall, a prominent analyst of participatory approaches, recommends: “a more deliberative process: one that engages policy actors in critical reflection on pervasive policy discourses and the accepted wisdom of prevailing policy narratives, rather than simply in finding out about poor peoples’ perceived needs” (Cornwall, 2001, p. 64).

Box 2: PSIA methodologies

The World Bank defines Poverty and Social Impact Analysis (PSIA) as “…analysis of the distributional impact of policy reforms on the well-being or welfare of different stakeholder groups, with particular focus on the poor and vulnerable”. There are seven key aspects: what particular reform is being analysed; which welfare indicator is being assessed; which social group is being analysed; what are the impacts on employment and wages, prices, market access, assets, and transfers and taxes; how do institutions affect these outcomes; when do the impacts materialise; what are the risks of an unexpected outcome.

PSIA aims to make more transparent the links between policy and poverty to improve public understanding of the logic behind policy choices. A policy analyst has a choice of a variety of tools when carrying out PSIA. The choice will be dependent on the analyst’s view of the nature of impacts of the reform in question and their data, time and capacity constraints. Examples of proposed methods include:

  • Social Impact Assessment – Assesses how the costs and benefits of reforms are distributed among different stakeholders and over time. Has been used to analyse the privatisation of state-owned enterprises and the reform of basic services.
  • Social Capital Assessment – Measures institutions and networks that determine access to resources of individuals and groups. Surveys are used to identify associations that have facilitated policy reform. Used in Bosnia and Herzegovina in the reform of social welfare systems and public service provision for returning refugees.
  • Demand/supply analysis – Estimates the likely response of consumers and producers to changes in price of goods or services. Has been used when considering increases in electricity prices in Armenia and market liberalisation in Mexico.
  • 1-2-3 PRSP model – A comprehensive model of an economy that can be combined with household data to simulate the impact of policy changes on poverty and welfare. Has been used in the Philippines to simulate the impact of macroeconomic changes on households’ nutritional status.
  • Simulations of Social Indicators and Poverty (SimSIP) – Examines how changes in growth rates will affect poverty and income distribution in particular countries by estimating the likelihood that a particular policy – such as an education sector reform – will achieve its objective.

Source: World Bank (2002a).

Problems in the Bank

Ironically, given that it does a lot of institutional and governance assessments, the Bank as an institution functions very poorly in many respects. The reforms introduced by the current World Bank President, James Wolfensohn, were supposed to improve the Bank’s mix of skills and ensure that the different parts of the Bank could share information better and collaborate more. However there is growing evidence that success has been patchy and that some of the reforms have been counter productive.

Non-economic social scientists in the Bank are less influential than their economist peers. Despite a rise in their numbers they often have to adapt their research methods and jargon to get their points across (Bebbington et al., 2002). The people working on “empowerment” in the Bank are in a separate team from those working on “social development” who in turn are distinct from the key “poverty” analysts.

The Bank has decentralised – moved some 2,000 staff including about half of its Country Directors – to its client countries. Catherine Weaver, a researcher at the Brookings Institution, comments: “the degree to which decentralisation has led to an espoused ‘listening culture’ is undermined by staff acknowledgement that reality has not quite lived up to this rhetoric”. She cites a World Bank internal report, which finds that “while staff are encouraged to listen to their clients in the field, they frequently find resistance in Washington to tailoring Bank approaches to heed what they have heard. And still, to an apparently excessive degree, they find themselves pressing their clients to use Bank guidelines, policies, systems and ways of planning” (World Bank, 2001d, p.14).

Another plank of the Bank’s reforms was the introduction of an “internal market”, which aimed to ensure that the best researchers were selected for each piece of in-country research. Bank Country Directors now have increased power over budgets and can contract in Bank analysts in response to what they think is needed in a country. This has not had the desired effect of creating healthy competition for research assignments. Instead a recent Bank review found that it has “created stress, job insecurity and poor morale as staff competed with each other for work. It also resulted in protectionism with budgetary disincentives to using people from other units” (World Bank, 2001d, p. 42). The system also frustrated the objectives of knowledge-sharing as “budget downsizing and job insecurity create incentives for staff to hold onto knowledge as a form of power” (World Bank, 2001d, Annex 3, p. 11). As a result, the “Knowledge Bank” has remained “unhealthily supply-oriented” and the incentive for staff is still the approval of their superiors in the Bank (ibid, p. 24). Bruce Rich, a long-standing analyst of the World Bank, argues that the Bank has to stop trying to be all things to all people and to decide who are its real clients – governments or poverty-stricken people. This would involve “focusing on quality not quantity in its lending, and rewarding staff first and foremost for ensuring that its policies relating to poverty alleviation, participation and the environment are carried out” (Rich, 2000).

One of the activities of any large organisation, and of units within it, is to maintain and expand its own remit. To do this the Bank has to make loans, its primary function. So the incentive system in the organisation, despite all the rhetoric about participation and about poverty reduction being its overriding objective, is still geared primarily towards large volume lending (World Bank, 2001e, p. vii).

Whose knowledge counts?

A number of commentators argue that the Bank is structurally incapable of understanding and acting on poorer peoples’ concerns. This is not just because it is politically dominated by the USA and other G7 countries (Griffith-Jones, 2002) but for more fundamental reasons. Larry Lohmann presents two. Firstly “development’s need to cross physical and social boundaries inherently complicates its efforts to predict, to manage and to translate power and knowledge from one place to the other”. Development officials have to defend their own and their institutions’ positions by denying “the existence of a reality in principle unmanageable from an office or institution” (Lohmann, 1998, pp. 1-2). Lohmann points out that development agencies’ reaction, when faced with criticisms or concerns about their interventions, will be to commission another study. But inevitably the studies conclude that the same development agency that is implicated in causing the problems also has the expertise and potential to resolve them (ibid., p. 6). In this light the Bank’s proliferation of studies is less than encouraging, reflecting the Bank’s ability to reinvent itself and expand its mandate in the light of external criticism.

A more favourable interpretation, however, is that not all Bank studies suffer from the same institutional pressures. Some Bank staff have managed to experiment with approaches involving genuine collaboration with outsiders. The Uganda Participatory Poverty Assessment Project, the Vietnam Public Expenditure Review and the Voices of the Poor national consultations stand out as oft-cited examples. The dynamics of these studies, the factors that made them relative successes, and their problems are explored in a new book Knowing Poverty: Critical Reflections on Participatory Research and Policy (Brock & McGee, 2002). Rosemary McGee concludes that one important factor is whether the policy “spaces” are created by “powerful actors from above” or are “autonomously created [where] less powerful actors set agendas” (p. 190).

Towards transparency and accountability?

Many NGOs and bilateral donors have welcomed the Bank’s intention to support governments who need help with social impact analysis. The UK‘s Department for International Development (DFID) has cautioned, however, that “the Bank and Fund have an obligation to manage the PSIA in a way that reflects PRS principles, including the promotion of national ownership and a more inclusive policy process” (DFID, 2001, p. 3).

In line with this, the selection of policies to be assessed should be made by the country, following consultations. In principle the Bank and Fund agree with this, but it is not clear who will be considered “key stakeholders” and whether consultation will occur both at the stage of considering what research to commission and at the stage of agreeing research conclusions and possible policy responses (World Bank & IMF, 2001).

Nor is it clear whether all such studies will be made public. Economic and Sector studies are reviewed by the Bank’s board, the government concerned and sometimes by parliaments. While the Bank in general says it supports publication, it is happy to accept government arguments that particular reports or sections of reports should not be published if they contain particularly “sensitive” information (World Bank, 2002b, p. 7).

Donor government representatives negotiating the financing of the Bank’s IDA lending recently urged the Bank “to look for ways in which more information can be made publicly available, including relevant economic and sector work underpinning the CAS and information on projects and PRSCs [Poverty Reduction Support Credits] under preparation” (IDA, 2001). Some governments have backed a call for aid agencies to publish a clear statement of what research they are commissioning, from whom, for what purpose, on what timetable. DFID and NGOs have urged that analysis of the likely impacts of reforms be made public well in advance of the PRSC being brought to the Bank and Fund boards (Oxfam International et al, 2001, p. 3). Indeed, since one of the intentions of PSIA is to make explicit the logic behind policy reform proposals, it is essential that the rationale for commissioning and conducting impact studies be brought into the open. The topics, the approach, the people, the timetable and the purpose of this publicly-funded research should all be made public.

The World Bank currently categorises its Economic and Sector Work in terms of intended audience and objective. The audience categories are ‘government’, ‘bank’, ‘donor’, and ‘public dissemination’, while the objectives are ‘knowledge generation’, ‘problem-solving’ and ‘public debate’. A surprising number of reports are not intended for public dissemination or to foster public debate. For example, the following 1998 studies on Bolivia, which raise issues of wide public interest, were categorised as non-public: Nutrition; Food Security and Rural Water; Rural Participatory Investments: Impact Assessment; Secondary and Higher Education; and Poverty (World Bank, 1998a, Annex D).

Even when documents are nominally made public, there is often a problem for Southern civil society organisations to obtain or make sense of them. The Internet is not accessible to many, so Southern NGOs may only be able to access documents if they travel to the nearest Bank office. The technical style of Bank documents is also daunting, and currently the documents are written in English – the Bank’s official working language – with occasional translations of executive summaries. This limits the potential for broad national dialogue, as many civil society organisations that have participated in the PRS process have pointed out. A recent Christian Aid briefing complained: “How can we expect participation to work when sometimes even the basic building blocks are not in place? Often it is very simple things, such as the language in which the document is published, or the lack of information about what a PRSP is, that prevent real and effective involvement of local groups” (Christian Aid, 2001, p. 14). Cambodian organisations similarly recorded: “Language is another important issue. [Key documents] have been drafted in English. As a result the plans have been discussed by foreigners while most Cambodians are not able to access them at all” (NGO Forum, 2001). The Bank argues that it has limited resources for translation, but this is a question of priorities: the Bank could, for example, easily save money on some of its self-promotional publications.

The Bank and Fund will have to do more to show that they are really prepared to make transparent the whole process of assessing policy options. Whilst some changes have been introduced in recent years, the “knowledge” work done by these agencies in their client countries still often conveys a traditional “we know best” mentality. And the Bank’s processes mean that much energy is being wasted on seeking internal approval rather than engaging with outsiders.

part II (110 Kb)