World Bank and IMF as they pertain to developing-world aid

What is the World Bank and what is the World Bank Group?

The World Bank is a 185-member cooperative (its members are "shareholders," so I assume it is some sort of international corporation) with two arms:

The World Bank Group consists of the IBRD and IDA plus three more institutions that are "closely associated with the World Bank" (see bottom of this page for that quote).

How are the different World Bank Group members funded?

Only IDA appears to have substantial funding in the form of grants (philanthropic purpose). It raises money through periodic replenishments: rich member countries donate to it. Top funders are the US, Japan, Canada, and several European countries (UK being the top funder overall, the only one that gives more than the US). IDA also gets some funding from IBRD and from repayments on its own loans.

The other four:

  • IBRD "raises most of its funds on the world markets."
  • The bottom of this page implies that IFC is primarily self-funding. Its 181 member countries provide its "authorized share capital" of $2.4 billion.
  • MIGA's website (from my scan) says nothing about how it's financed; its site is devoted to promoting/explaining its services, primarily insurance. I didn't see an annual report or anything along those lines.
  • ICSID's "Secretariat's administrative costs are financed out of the World Bank's budget; the costs of ICSID proceedings are borne by the disputing parties." (From this page)

What is the IMF and how is it similar to / different from the World Bank?

IMF overview

The International Monetary Fund (IMF) is a 185-country cooperative whose "main goal is to ensure the stability of the international monetary and financial system." Its activities are listed here; in addition to providing research and technical assistance, it is a "lender of last resort," and also "encourag[es] countries to adopt sound economic policies." It isn't particularly easy to find exactly what they mean by "sound economic policies," but the 2007 decision on bilateral surveillance seems to emphasize discouragement of capital controls and currency manipulation. There is also a focus on corruption-free and transparent government similar to the World Bank's.

The IMF is funded by subscriptions from its member countries, who are also its potential borrowers; more on membership here.

The IMF's publication index is difficult to filter and appears to focus on macroeconomic rather than development issues.

IMF vs. World Bank

Article on IMF vs. World Bank. In a nutshell, the World Bank is focused on helping less wealthy nations by providing services, loans and grants that they could not get otherwise. The IMF is less about assistance/redistribution and more about stability - avoiding global shocks and acting as a lender of last resort when they do occur.

IMF's role in aid to developing countries

The IMF has a concessional lending program - Poverty Reduction and Growth Facility (PRGF) - and how this differs from the lending arm of IDA is not entirely clear to me. The basic goals and approaches seem to overlap, though:

PRGF-supported programs are designed to cover only areas within the primary responsibility of the IMF, unless a particular measure is judged to have a direct, critical macroeconomic impact. Areas typically covered by the IMF include advising on prudent macroeconomic and financial policies and related structural reforms such as exchange rate and tax policy, fiscal management, budget execution, fiscal transparency, and tax and customs administration.

When appropriate, the IMF draws on World Bank expertise in designing PRGF-supported programs, and the staffs of the Fund and Bank cooperate closely on conditionality. The Bank staff takes the lead in advising the authorities in the design of poverty reduction strategies in areas such as poverty assessments, monitoring, structural and sectoral issues, social issues, and costing priority poverty-reducing spending.

"PRGF-supported programs are framed around comprehensive, country-owned Poverty Reduction Strategy Papers (PRSPs)." The IMF hosts the complete set of PRSPs, which give detailed information about the state of (and plans for fighting) poverty in each nation.

The rest of this page focuses on IDA.

IDA is of direct interest to us because it funds developing-world projects without expecting to get its money back (and is financed by donors). This is not true of the other institutions (detailed above).

In some cases IDA & IBRD data is combined (such as for total expenses and "% successful" measures). IDA is broken out when possible.

What is IDA's general philosophy of aid? What is it trying to accomplish and how?

As detailed below (see "details of the IRAI index"), IDA passes specific judgments and has specific opinions on how governments should set policy and run, and uses its loans and grants as incentives to get them to adjust. The allocation process between countries doesn't seem to consider the content of projects - only the track record of success of projects and governments' compliance with IDA's vision.

The allocation process within a country is based on Poverty Reduction Strategy Papers, authored by governments.

What I've read elsewhere in several places (most notably in this paper by Bill Easterly) implies that IDA used to be more "heavy-handed" in terms of imposing strict conditions on governments; failure over time has led to a more flexible, bottom-up approach.

How does IDA select projects and countries?


Overview here; details at this PDF (linked from here). The criteria are:

  • Population
  • Poverty (measured via per capita GNI)
  • Lack of creditworthiness to borrow in conventional markets (used to determine eligibility, not allocation)
  • Country Performance Rating index (ratings by country here). Combines a measure of "portfolio performance" (from the World Bank's Annual Report on Portfolio Performance) and the IRAI, which judges countries on their financial/economic policy and quality of institutions.
  • Exceptions:
    • Countries with partial/potential access to IBRD loans get less.
    • Special post-conflict aid (up to 10 years after conflict)
    • Extra grants can be given to countries "re-engaging with IDA after a prolonged period of inactivity."
    • Special aid for natural disaster aftermath.
    • "selected regional integration projects" (?)
    • Extra allocations can be made to help finance the costs of arrears to IBRD & IDA

First, eligibility is determined by (a) relative poverty in terms of per capita GNI; (b) lack of creditworthiness to borrow in conventional markets. Then, allocations are made based on the Country Performance Rating (CPR) index (country ratings here), which is a combination of "portfolio performance" and the IRAI score. The latter measures "performance in implementing policies that promote economic growth and poverty reduction."


Process description here. Like the IMF, the World Bank relies on Poverty Reduction Strategy Papers designed by governments; strategy is then set by "Bank staff in meetings with government officials, in consultation with country authorities, civil society organizations, development partners and other stakeholders. It takes as a starting point the country's own long-term vision for development and takes into account the Bank Group's comparative advantages in the context of other donor activities."

What are the details of the IRAI index, which judges quality of institutions and policies in a country?

The index is created by a committee; the most detailed explanation I've found is at this PDF. Highlights:

  • Monetary policy: price stability, "external and internal balance," not crowding out private investment
  • Fiscal policy: fiscal balance, provision of public goods, ability to withstand shocks
  • Low debt/debt service burden
  • Trade: low tariffs, low-corruption customs
  • Financial markets: "high ratio of private sector credit to GDP," availability of credit and widespread access to financial services, robustness of banking sector to shocks (hmmm)
  • Regulation of business sector: low barriers to entry (fees for starting enterprise, etc.); low labor market regulation; corporate governance laws
  • Gender equality in terms of access to / use of education; labor force / business participation; equality under the law
  • Progressive taxation; public-sector focus on helping the poor
  • Policies conducive to facilitating widespread access to "(a) health and nutrition services, including population and reproductive health, (b) education, ECD [early child development], training and literacy programs, and (c) prevention and treatment of HIV/AIDS, tuberculosis, and malaria"
  • Social safety net; labor standards; pension/savings systems
  • Pro-environment policies
  • Property rights: clear laws & enforcement, low crime
  • Fiscal budget is reported and adhered to
  • Taxes: consistent and rule-based administration; nondistortive and broad revenue sources
  • Quality of public administration: coordination, organization, merit-based hiring & promotion
  • Structure of public sector: balance of powers, transparency, not dominated by special interests

What is the nature of IDA's funding? (Unrestricted, restricted, grants, loans, etc.)

IDA gives both grants and loans. Its financing also falls into two separate categories in terms of conditionality (outlined here):

  • Investment refers to loans/grants earmarked for specific, pre-determined expenses.
  • Development policy refers to loans/grants given in exchange for policy changes.

I believe that all expenses fall into one of these two categories, i.e., IDA does not fund entities other than governments and does not provide purely unrestricted funding.

The World Bank's most recent Annual Report opens with a summary of expenses implying that "development policy" investments make up about 25% of the total. The IDA has about has had about $11 billion in "commitments" over the last two years, while its grant expenses have grown from $2 billion to $3 billion. It isn't entirely clear to me whether grants are separate from or included in "commitments," but I'm guessing that they're separate since they have a separate line.

How much does the IDA spend each year? How does it break down by country and project type?

Total annual expenses: most recent Annual Report has "commitments" (which I believe are concessional loans - see directly above) going from ~$9 billion to ~$11 billion from 2004-2008. Grants go from ~$1.7 billion to ~$3 billion.


Available on page 56 of the most recent Annual Report, though it breaks down only loans (not sure where we can get similar data for grants).

  • 50% Africa, 25% South Asia, 16% East Asia & Pacific, 9% other.
  • Themes and projects are mostly oriented toward infrastructure and economic environment, rather than direct humanitarian interventions. 8% health, 11% education, 9% "agriculture, fishing and forestry", 9% "water sanitation and flood protection"; the remaining 63% is more infrastructure-oriented.

A more detailed look at the last 10 years is available via the Annual Report of the evaluation arm (IEG), Appendix A.

Regarding history - an article on the IMF vs. World Bank says, "During the first two decades of its existence, two thirds of the assistance provided by the Bank went to electric power and transportation projects. Although these so-called infrastructure projects remain important, the Bank has diversified its activities in recent years as it has gained experience with and acquired new insights into the development process."

What is the track record of the IDA and its different projects?

Major successes/failures

William Easterly's historical overview implies that the World Bank's efforts to change policy and institutions have no real successes to point to, and a lot of failures and changes in direction.

The World Bank is listed as a major player in 3 of Center for Global Development's success stories: treating cataracts in India, controlling tuberculosis in China, controlling onchocerciasis in sub-Saharan Africa.

Project self-evaluation

Each World Bank project is reportedly evaluated by the World Bank's Independent Evaluation Group (IEG). The IEG's Annual Report has summary statistics of % projects with satisfactory outcomes, by project sector and by area, and some very brief descriptions of how it evaluates that don't illuminate much. (See Appendices A and B.)

Despite the lack of clarity about what exactly constitutes a "satisfactory outcome," this data is still interesting.

  • Overall "% projects with satisfactory outcomes" (henceforth "success rate") has gone from between 70-75% (1992) to around 85% (current) (pg 78).
  • Pg 82: success rate has been highest (>90% 2003-2007; 80% 1998-2002) in East Asia and the Pacific; closer to 80% in other regions; and lowest (~72% 2003-2007; 60% 1998-2002) in Africa. Latin America/Caribbean is the only region where performance has been worse recently (80% 2003-2007; ~85% 1998-2002).

    You can see the evaluations for specific projects via this page - click a topic and select a project. Example report on an AIDS program in Ghana. It's harsh:

    While the prevalence of HIV in Ghana fell, trends in HIV prevalence are not a valid measure for changes in the number of new infections (incidence). A decline in HIV prevalence could conceivably be the result of high AIDS mortality or unsuccessful treatment efforts, while an increase could signal more widespread and successful treatment. On the other hand, positive trends in the knowledge, attitudes and behaviors among the general population and among high-risk groups are consistent with a reduction in HIV incidence. Unfortunately these trends are not available because M&E was not undertaken as planned. While some baselines were established for some groups, they were not updated at the end of the project. Field visits suggest an increase in the knowledge of modes of HIV transmission and prevention methods among the general population and among youth and this is likely attributable in part to GARFUND support. However, there was no evidence of change in the levels of fear, stigma and discrimination associated with HIV/AIDS. Some RACs and DACs did report an increase in condom availability. Indeed, annual condom sales increased from 24 to 30 million, between 2002 and 2005, a result largely attributable to social marketing efforts financed by USAID. GARFUND investment in condom distribution was very modest and the bulk of prevention efforts were focused more on information, education, and communication (IEC) than on behavior change.

    What other relevant information/research is available from IDA?

    Research publications

    Complete publication database: they tried hard to make this usable but I find it unusable. There are too many publications and while you can search by topic, you can't distinguish broad reviews from highly specific papers, so there's no way to find papers of interest except by going through mountains of papers. If we know exactly what subtopic we're interested in, this could be useful in some cases. For example, I took a look at nutrition and found 3 things that look worth reading:

    Issue briefs: these are short. I'd like to read about 10 of them.

    World Development Reports: each year is a big report on a specific topic. Most are on highly macro/policy topics that don't look directly relevant for our purposes. The 2008 report on agriculture could give an important perspective on agriculture.

    Policy Research Reports: seem intended to be accessible/broad. There seem to be 16 total. Most pertain to macro/policy; some that are of interest have already been found/referred to/noted elsewhere. We may want to look at the Assessing Aid - What Works, What Doesn't, and Why paper though I can't see a way to download it.

    Global Monitoring Reports focus on progress toward Millennium Development Goals.

    Development Impact Evaluation Initiative (DIME): mentioned in Making Aid Work as the more results-oriented part of the World Bank. I haven't looked at this yet.

    Topic pages: these are deep and would take forever to really go through. Potential areas of interest include:

    Research center has more huge lists of resources, including lists of books authored and edited by World Bank researchers.

    General data on country populations

    The World Bank hosts the Living Standards Measurement Study, which collects a variety of data on living standards in different parts of the world.

    IDA has indicators by country and region, available in a variety of formats (I like the table). Most are in GapMinder; here are the ones that aren't.

    • Poverty headcount ratio at $1.25 a day (PPP) (% of population)
    • Cost of business start-up procedures (% of GNI per capita)
    • Public financial management (number of HIPC benchmarks met)
    • Time required to start a business (days)
    • Access to an all-season road (% of rural population)
    • Fixed line and mobile phone subscribers (per 100 people)
    • Household electrification rate (% of households)
    • Agriculture, value added (constant 2000 US$ millions)

    There is a seemingly much larger set of indicators available through online databases. Subscription required. Lists of potentially relevant indicators here (World Development Indicators) and here (XLS - African Development Indicators).

    • A lot of it is probably available elsewhere: through Gapminder, OECD-DAC (free data on aid flows) or standard financial/balance-of-payments sources.
    • Includes data on food production / arable land.
    • Includes a lot of stuff on "Governance & partnership" and "Doing Business" that seems related but not identical to IRAI ratings. Control of Corruption, Government Effectiveness, etc. (not clear how they quantify these but it looks like some sort of survey.)
    • Lots of infrastructure indicators: KM of rail lines, passenger cars per 1000 people, etc.
    • Human development: data on how many people have/use bednets, are treated with antimalarials, give birth with skilled attendants.
    • Income inequality: income share held by each quintile (and top/bottom 10%)
    • Lots of labor force indicators: employment & child employment in agriculture, employment by education level & gender

    African Development Indicators costs $100/yr for an individual. World costs $200.