Home
 COUNTRY BY  COUNTRY
  THE BIG ISSUES
 PROGRESS AND  REGRESSIONS
 DEVELOPMENT  INDICATORS
   | ESPAÑOL | Commitments | Annual Report | News | About  | Site Map Feedback  
 The Big Issues: Reports by commitment

2004
Country Ownership Undone: to Set Policy Directions of Borrowing Governments, the World Bank Uses “One Size Fits All” Scorecard – the “Country Policy and Institutional Assessment” (CPIA)

edited report

Nancy Alexander
Citizens’ Network on Essential Services

Table of Contents

I.                    Introduction to the CPIA

A.      What is the CPIA?

B.      How does the CPIA affect lending levels?

C.      What does the CPIA rate?

D.      Is there a plan to disclose CPIA ratings?

E.      What is the structure of the CPIA rating system?

F.      How is the CPIA used to allocate funds to low-income countries?

G.   What are some of the controversies around the Bank’s role in rating country performance?

H.   What happens when a government flunks the CPIA?

I.    Does the CPIA actually signal success?

J.    How does the CPIA determine operational policies of the World Bank?

K.       How can the CPIA Undercut Sovereignty?

L.        Conclusion

II.         Country Policy and Institutional Assessment (CPIA) Ratings for 2003

A.      Eastern Europe/Central Asia

B.      Latin America and the Caribbean

C.      Africa

D.      South Asia/East Asia/the Pacific

E.      Middle East/North Africa

III. Rating Criteria

A.      Economic Management

B.      Structural Policies

C.      Policies of Social Inclusion and Equity

D.      Public Sector Management and Institutions

Box: Candidates for the U.S. MCA Get Broad Range of Scores from the World Bank

  Figure: IDA Country Performance Rating

I.                    Introduction to the CPIA

 

A. What is the CPIA?

Every year, the World Bank rates the quality of client countries’ policies and institutions with an instrument called the “Country Policy and Institutional Assessment” (CPIA). The Bank carries out this rating for about 100 countries. CPIA ratings are secret for about 20 countries; they are not even disclosed to other multilateral and bilateral institutions. For about 80 other countries, the Bank discloses selected (and partially aggregated) CPIA ratings.

The World Bank uses individual country ratings as diagnostic tools to help allocate its (loan and grant) resources among borrowers[1] and determine the policy thrusts of new operations.

 

B.      How does the CPIA affect lending levels?

 

The World Bank rates each government’s performance by the extent of its compliance with the World Bank's definition of "good policies." In 2002, the Bank allocated about three times more money to governments that were "A students” than to "F students.”

 

C.      What does the CPIA rate?

 

According to the Bank, the purpose of the CPIA is to measure a country’s policy and institutional development framework for poverty reduction, sustainable growth and effective use of development assistance.[2] Another view is that the CPIA rates the extent to which a government has: a) adopted neoliberal economic policies (i.e., liberalization and privatization in the context of strict budget discipline) and b) developed the effectiveness and stability characterized by mature political, social, and economic institutions.

 

The overarching institutional strategy of the World Bank Group is its Private Sector Development (PSD) Strategy.[3] The PSD Strategy expands the privatization “frontier” into basic services -- health care, education, water and energy – especially in poor countries. Pursuant to this Strategy, CPIA-derived policy prescriptions are focused on promoting the private sector. (The IMF, World Bank, and WTO define the “private sector” as including both for-profit firms and not-for-profit agencies or NGOs.)

 

D. Is there a plan to disclose CPIA ratings?

 

Not yet. In October 2003, The Bank’s management proposed a general CPIA disclosure plan to its Board of Executive Directors that involved a gradual phasing-in period. In an initial stage, management proposed that the CPIA be disclosed to multilateral and bilateral development partners. At the same time, management proposed that the Bank should only publicly disclose the CPIA after borrowing governments are thoroughly educated about the scoring system. During this stage, Bank staff would discuss current CPIA ratings with governments. Bank staff may discuss a government’s CPIA ratings with governments that are performing well, but only rarely will such discussions take place with governments that are performing poorly.

 

The opinions of Board members about the proposed CPIA disclosure policy are “all over the map,” including those who want:

  • rapid disclosure;

  • delayed disclosure;

  • no disclosure out of concern that disclosure would jeopardize the ability of some countries to attract foreign direct investment and other financial flows;

  • a more objective and robust CPIA rating methodology prior to the disclosure of ratings. External reviewers of the rating methodology will report to the Board in the next few months.

  • greater country “voice.” Some of the Board members that are concerned that developing countries have no ownership of the CPIA process feel that the Bank’s management should not just teach governments about the CPIA system. They feel that other donors and creditors as well as governments, themselves, should participate in the rating process as “partners” with the Bank.



  • Many Bank borrowers are adamantly opposed to release of the CPIA, such as the Governors of the World Bank from Africa. The twenty higher-income countries with secret CPIA ratings, such as Brazil, will probably maintain that secrecy. However, low-income countries may have little choice about whether their scores will be released or not.[4] This exacerbates the double standard between less and more prosperous countries and violates the provision of the Articles of Agreement calling for equal treatment of borrowing countries.

     

    E. What is the structure of the CPIA rating system?

     

    The CPIA rates countries primarily on the basis of current performance in relation to twenty criteria that are grouped into four categories.

     

    • Economic management, including management of inflation and current account; fiscal policy; management of external debt; and management and sustainability of the development program.
    • Structural policies, including trade policy and foreign exchange regime; financial stability and depth; banking sector efficiency and resource mobilization; competitive environment for the private sector; factor and product markets; and policies and institutions for environmental sustainability.
    • Policies for social inclusion, including gender equity and equality of economic opportunity, equity of public resource use, building human resources, safety nets; and poverty monitoring and analysis.
    • Public sector management and institutions, including property rights and rule-based governance; quality of budgetary and financial management; efficiency of revenue mobilization; efficiency of public expenditures; and transparency, accountability, and corruption in the public sector.

     

    F. How is the CPIA used to allocate funds to low-income countries?

     

    The World Bank’s staff uses a formula to divide up the funds available for low-income countries that includes “need” (income per capita) and “performance.” The CPIA is an important input in calculating a government’s performance rating. The criteria are weighted as follows:

     

    1) The CPIA (comprised of the four categories listed in part “E”) accounts for 80% of a government’s overall rating;

    2) The Bank also rates each government’s performance on the portfolio of outstanding loans. This rating counts for 20% of a government’s overall rating. It measures how well a government manages its loan resources, including how it achieves timely disbursement through efficient procurement practices.

    3)       Finally, the level of grants and loans to which a borrowing government has access will then increase or decrease as a result of the Bank’s application of a “governance factor” to its overall score. [5] The governance factor is the average of seven governance-related criteria that are already in the CPIA and the portfolio performance ratings. Hence, governance criteria are effectively double counted in the final rating. In recent years, the application of a governance factor has reduced the allocation to about twelve countries by as much as 50%.[6]

    Figure: IDA Country Performance Rating

     Source: “Allocating IDA Funds Based on Performance.” International Development Association, March 2003.

     

    G. What are some of the controversies around the Bank’s role in rating country performance?

     

    • Most development practitioners are critical of a “one-size-fits-all” set of characteristics that define “good” policies and “good” institutions. For instance, there is little agreement about what constitutes “good” trade policy. Even where there is agreement on general policy principles, there are still disagreements, even among neoliberal economists, about the pace, sequence and implementation of these policies, as well as their impacts, such as short-term distributional effects.
    • The CPIA exacerbates inequities between richer and poorer countries by imposing policies on countries with less power and resources, while wealthier and more powerful countries do as they please. For example, the IMF and World Bank are requiring governments to adopt some rules of trade agreements that haven’t even been negotiated![7] More prosperous countries would never stand for this.
    • In the context of its practice of directing policy decisions, the Bank’s evaluation of governance -- the government’s accountability to its citizens – has disturbing political implications. When governance scores constrain or determine fundamental decisions about resource allocation or the roles of the public and private sectors, they may violate the World Bank’s Articles of Agreement, which prohibits interference in a country’s domestic affairs.[8] Moreover, the Bank has little capacity to evaluate certain policies, e.g., gender policies or the status of institutional development. Historically, the Bank has an abysmal record of promoting institutional development.
    • Donors and creditors are competing over which one has the best rating system. Some claim that the bilateral U.S. aid program, the Millennium Challenge Account (MCA) has a superior rating system. It is quite plausible that, as the Bank’s largest shareholder, the U.S. could use its MCA rating methodology to drive the evolution of the CPIA rating methodology.

     

    H. What happens when a government flunks the CPIA?

     

    Countries that receive CPIA scores of “D” or “F” are usually also designated as “Low-Income Countries Under Stress” (LICUS). Donors and creditors assume responsibility for many functions of the so-called “failed” governments of LICUS countries. In 2002, there were 30 LICUS countries and, for each of these countries, donors and creditors established Independent Service Authorities (ISAs). An ISA, which is largely independent of the government it claims to serve, generally contracts traditional government services to private providers that are primarily accountable to the donor and creditor community.[9]



     

    Box
    A Study in Contrasts:

    Candidates suggested for the U.S. MCA in 2002
    Get Broad Range of Scores from the World Bank

    When applying the preliminary performance criteria of the Millennium Challenge Account, the 11 countries in the left column are “winners.” The World Bank gives lower ratings to 8 of these countries. World Bank overall and category scores are presented in the columns to the right of MCA Winners.



     

    WORLD BANK RATINGS

     

    Possible
    US MCA
    "Winners"[10]

    Overall
    Rating

    1
    Economic
    Management

    2
    Structural
    Policies

    3
    Social
    Inclusion

    4
    Public
    Sector

    5
    Portfolio
    Performance

    Albania

    B

    A

    A

    A

    C

    C

    Bangladesh

    B

    B

    C

    A

    C

    D

    Bolivia

    B

    B

    B

    C

    C

    C

    Gambia

    D

    C

    C

    C

    F

    F

    Georgia

    D

    C

    D

    B

    F

    D

    Honduras

    A

    B

    A

    A

    A

    B

    Malawi

    C

    D

    C

    B

    C

    F

    Mongolia

    C

    D

    C

    B

    B

    C

    Nepal

    B

    A

    C

    D

    B

    C

    Senegal

    A

    B

    A

    B

    B

    F

    Sri Lanka

    A

    A

    A

    A

    A

    D

     

    I. Does the CPIA actually signal success?

     

    Prominent World Bank economists, such as David Dollar, use CPIA statistics, among other things, to prove that governments with “good policies” (as defined by the CPIA) prosper and make good use of foreign aid and credit. This important claim is the basis for the “selectivity” policies of donors and creditors, through which donors and creditors increasingly allocate lending toward governments that have adopted “good policies.” For instance, World Bank lending concentrates on three of India’s 24 states that are committed to implementing private sector development.

     

    One team of independent economists has been able to question the World Bank’s claim that there is a correlation between “good policies” and economic growth in developing countries. (See box: “Comparing Policies and Outcomes”) William Easterly’s team of economists was able to challenge the World Bank because it had privileged access to the mostly secret CPIA data base. (Until recently, Easterly was a senior economist of the World Bank.) When using an expanded CPIA data set, Easterly et al. reported that they “no longer find that aid promotes growth in good policy environments.” [11]




     

    Comparing Policies and Outcomes

    Yale professor Lawrence King analyzed policies of 12 countries in Eastern Europe using an index that measures the intensity with which each country has embraced neoliberal policies. He found an inverse relationship between the intensity with which these policies were embraced and economic performance. In other words, when a country embraces neoliberalism, its economic performance often declines.

    Although King uses different (and narrower) indicators than the CPIA, his results are nonetheless instructive.

    Industrialized countries often reject these policies or adopt them (on a selective basis) when they reached a fairly advanced level of development and, even then, they were only selectively adopted.

    Source: Lawrence P. King, “The Emperor Exposed: Neoliberal Theory and De-Modernization in Post communist Society.” Yale University, 2002.

    The World Bank’s own evaluators issue a warning against interpreting any internal Bank research as finding that “good policies” as measured by the CPIA from 1977 to 2000 help explain good economic growth. [12]

     

    J.       How does the CPIA determine operational policies of the World Bank?

     

    While the CPIA measures past performance of borrowing governments, its rating criteria also represent implicit “one-size-fits all” policy prescriptions. The World Bank offers a government a programmatic (adjustment) loan to addresses weaknesses in government performance as assessed by the CPIA.

     

    The World Bank’s Country Assistance Strategy (CAS) for each country identifies the need and rationale for these adjustment loans. The CAS outlines prospective Bank investments over a medium-term (e.g., three-year) time horizon for each borrowing country. [13] Each CAS stipulates which policy actions, or “triggers,” the government must take to gain access to a higher-level of resources or to prevent losing access to resources.

    In some cases, when governments fail to comply with World Bank (or IMF) policy conditions, donors and creditors may terminate assistance to an entire sector or country. They also suspend debt relief when a government (e.g., Nicaragua, Benin, Mali, Chad, and Nigeria) fails to expedite specified market reforms, such as privatization.

    CAS triggers are largely determined by the CPIA. A 2003 Bank paper stated that the main policy prescriptions included in the CAS are: “increasingly focused on aspects of the CPIA that are shown to be weak. The triggers can also include policy targets from PRSP, to the extent that they are expected to strengthen policy and institutional performance.”[14] Hence, the CAS may reflect different priorities than a government espouses.

    K.      How can the CPIA Undercut Sovereignty?

    The IMF and World Bank require that each low-income country prepare a Poverty Reduction Strategy Paper (PRSP) — a three-year “national development strategy” — in order to qualify for external financing and debt relief. In preparing a PRSP, governments often solicit input from a wide variety of domestic constituencies.

    The IMF and World Bank promised that the PRSP would provide a framework to set forth a government’s own priorities to guide operations financed by donors and creditors. However, the sovereignty of governments is compromised because:

     

    • the institutions have broken this promise. In practice, the World Bank’s CPIA can be more influential than the PRSP in shaping key economic policies in borrowing countries. The World Bank will require a government to remedy weaknesses in its CPIA rating in order to qualify for more financing or debt relief. Moreover, the IMF dictates the framework and the resource “envelope” for the priorities identified by a PRSP.
    • donors and creditors (which often adopt CPIA priorities) play a leading role in preparation of PRSPs (more prominent than parliamentarians). The process of formulating PRSPs can displace more “homegrown” policy-making processes.[15]
    • CPIA-related policy prescriptions can override the policy priorities of citizens and elected officials. Indeed, these constituencies usually lack knowledge of their government’s CPIA ratings and the implications of those ratings for public policies.

     

    L. Conclusion

     

    In general, country “ownership” of the development process is a mirage, except where countries are large and can take an independent stand. Such countries, like China, often borrow significant sums from the IFIs and lack crippling debt burdens.

     

    Donors and creditors dominate the policy-making of low-income countries more than ever before. The CPIA represents a policy straightjacket. No matter what a country's own development strategy (or Poverty Reduction Strategy Paper) says, a country is likely to adhere to CPIA-derived policy prescriptions if it expects to retain external support. Governments are in a double bind if citizens and elected officials choose a path other than that specified by CPIA priorities.

     

    As noted, the PRSP is supposed to be the “roadmap” for achievement of the MDGs, which aim to halve the number of people in poverty by the year 2015. However, the influence of the CPIA over the PRSP (among other things[16]) underscores the lack of ownership that governments have over their own development future. Moreover, the achievement of MDGs significantly depends upon whether the neoliberal policy preferences embedded in the CPIA can help overcome poverty and deprivation. There is much more evidence to rebut this causal connection than to support it.[17] Accordingly, it is legitimate to ask who is responsible for achieving the MDGs: governments which, if they wish to maintain external support, are pressured to adopt policies derived from CPIA ratings; or the donors and creditors that drive the development process “from behind”?

     

    The CPIA rating system undercuts democracy in borrowing countries by constraining the policy choices available to citizens and their elected officials. If donors and other multilateral creditors adopt the CPIA rating system, then a policy cartel wielding most aid and credit will have an even more profound consequence for democracy and development.

     

    The CPIA straight jacket is one indicator of the increasingly ideological approach to policy-making. Harvard Professor Dani Rodrik stated that, “The broader the sway of market discipline, the narrower will be the space for democratic governance… International economic rules must incorporate “opt-out” or exit clauses [that] allow democracies to reassert their priorities when these priorities clash with obligations to international economic institutions. These must be viewed not as 'derogations' or violations of the rules, but as a generic part of sustainable international economic arrangements." Occasionally, such exits from obligations are possible for large borrowers from the IMF and World Bank, but the institutions discriminate against low-income countries.

     

    The Rodrik approach is minimalist insofar as it would allow governments to opt out of commitments which they made, freely or under duress, to the IFIs. However, the ideal --so often proclaimed and so little practiced -- would invite governments and their citizens to authorship as well as ownership of their national strategies.

     

    To put these heretical ideas into perspective, one might reasonably ask what kind of CPIA rating industrialized country governments might receive? Developing country governments aren't given the same flexibility that their wealthier counterparts claim for themselves when determining whether or when to liberalize, privatize or exercise greater budgetary discipline. For instance, if the US and EU were subject to CPIA review, their current fiscal policies would result in austerity measures that are politically unimaginable. From blatant protectionism, to market distorting subsidies and ballooning deficits, everyday policies of the governments that control the IFIs reveal a shocking double standard that makes a mockery of national sovereignty for most of the world’s countries.

     

    II. CPIA Ratings for 2003[18]

     

    This section presents CPIA ratings from Borrowing countries in all of the Bank’s geographic divisions. About 20 countries with CPIA ratings do not appear, because of their ratings are secret. The Bank discloses each country’s overall rating, as well as ratings for each of the five broad categories from which the overall rating is derived. The five ratings are those for: economic management, structural policies, policies for social inclusion and equity, public sector reform, and portfolio performance. For each of these ratings, the Bank applies numerical ratings from 1 (low) to 6 (high). The charts in this paper convert these numbers to only five “letter” grades. The reason for presenting the data in this way is that the Bank itself places each country in one of five quintiles based upon the CPIA performance score. While the letter grades are therefore not perfectly precise representations of the numerical scores, they are still highly indicative.

     

    A. Europe/Central Asia

     

     

    IDA

    Country

    Rating

    A
    Gover-nance

    B
    Overall
    CPIA

    Rating

    1
    Economic
    Manage-ment

    2
    Structural
    Policies

    3
    Social
    Inclusion

    4
    Public
    Sector

    5
    Portfolio
    Perfor-mance

    Albania

    C

      C

      C

    C

    B

    C

    D

    B

    Armenia

    A

    B

    A

    A

    A

    B

    B

    A

    Azerbaijan

    B

    B

    B

    A

    C

    D

    C

    A

    Bosnia/

    Herzegovina

    B

    B

    B

    B

    B

    C

    C

    B