A quarterly magazine of the IMF

June 2005 , Volume 42, Number 2

How donor support for primary education can be made more effective

Providing primary school education for all the world´s children will be a protracted and costly undertaking. The good news is that major donors have significantly increased their official development assistance (ODA) for education in general and for primary education in particular since 1990. Donor commitments for all education purposes reached $6.7 billion in 2003, more than double the amount in real terms committed just six years earlier. Commitments for primary education have been rising even faster, quadrupling to $1.9 billion between 1990 and 2003. Aid for primary education has risen much faster than total ODA, with its share in commitments rising from just 0.4 percent in 1990 to 1.9 percent in 2003 (see Charts 1 and 2).

The emphasis that individual donors place on primary education varies widely (see table). Five donors allocated more than 4 percent of their ODA commitments to primary education in 2003, while four others provided less than 1 percent. These variations partly reflect different levels of prominence donors place on education, and partly reflect some donor specialization in different activities in an effort to avoid duplication.

Besides ODA, private foundations, nongovernmental organizations (NGOs), faith-based organizations, and other charities make important contributions to education in many countries. For example, some of the most effective primary schools in Africa and Asia, particularly in rural areas, are run by religious organizations. The Organization for Economic Cooperation and Development (OECD) reports $10.2 billion in private aid donations in 2003 for all purposes, but, by all accounts, this figure is too low, probably by a factor of two or three. There are no valid estimates for the share of these funds devoted to primary education, but it is plausible that $1 billion or more is provided annually for primary education from private sources, in addition to the $1.9 billion provided by official sources in 2003.

Nevertheless, while the amount of assistance committed to primary education has been rising, most analysts believe the current level remains too low, particularly to achieve the second Millennium Development Goal of universal primary education by 2015. According to the Task Force on Education and Gender Equality of the Millennium Project, reaching this goal would require an additional $7 billion to $17 billion per year (estimates vary), with perhaps half coming from domestic resources and the rest from foreign sources. Some analysts, such as Michael Clemens at the Center for Global Development (CGD), argue that even with large amounts of funding universal primary education is unlikely to be achieved in all countries, although substantial progress toward the goal is certainly possible.

While there are many competing demands for donor and public sector resources, the development community has long stressed the importance of both expanding primary education to previously excluded groups and improving the quality of teaching, given the expected high economic and social returns. Of course, money alone will not achieve the goals, but it can help to relieve constraints on training teachers, retaining them with adequate compensation, building schools, supplying those schools with books and other materials, and reducing or eliminating school fees.

Much depends on the actions of developing countries themselves, but donors can create efficiencies as well. To make aid smarter, donors can learn some lessons from successes in other fields, such as health, where donors have experimented with results-based incentives. This article examines how effective aid is in general and highlights potential lessons that can be applied to ODA for primary education from the substantial institutional innovations in aid delivery of the past decade. It concludes by pointing to directions donors might take to increase the impact of ODA for primary education, based on our increased understanding of the conditions conducive to aid effectiveness and of how to improve learning outcomes.

How effective is aid?

The effectiveness of foreign aid in achieving development objectives, especially in supporting economic growth, has been a subject of controversy for many decades. There are three main viewpoints that have support in the research literature:

• Aid has no effect on growth, and may even have a negative effect, because it is wasted on bad projects, engenders corruption, or undermines private investment.
  
  
• Aid has a conditional impact on growth, working only (or at least better) in countries with good policies and institutions. This has become the most influential viewpoint, although some of the results have come under question.
  
  
• Aid on average has a positive impact on growth with diminishing returns—meaning that aid has not worked everywhere, and not necessarily only in countries with good policies and institutions, but overall it has had a positive impact on growth.

In a recent CGD study, Michael Clemens, Steven Radelet, and Rikhil Bhavnani took a new approach. They disaggregated ODA and focused on the types of aid that are actually aimed at affecting growth directly and relatively quickly—aid for infrastructure, agriculture, and industry, as well as budget support—as distinct from aid for humanitarian relief, political development, or health and education (the latter of which might affect growth, but less directly and over a longer time frame). They found a strong causal relationship between this subset of aid and economic growth (with diminishing returns), a conclusion that withstood a wide variety of statistical tests. The relationship was stronger in countries with better institutions, but did not depend on strong institutions. These results suggest that aid has been more effective in supporting growth than many analysts had previously thought.

Other studies have focused on the relationship between aid and other development outcomes, particularly health. For example, the CGD study "Millions Saved" by the What Works Working Group and funded by the Bill & Melinda Gates Foundation, documents 17 major health success stories—large-scale health interventions that improved health outcomes over a sustained period—almost all of which received substantial ODA. Little such systematic research on aid for education exists. However, there are certainly many successful aid-financed education initiatives—such as the girls´ education initiative in Balochistan, Pakistan; system-wide education reforms in Ethiopia in the 1990s; Uganda´s expansion of primary education; and Indonesia´s construction of over 61,000 primary schools in the mid—1970s, which MIT economist Esther Duflo showed led to significant increases in education attainment and increased wages for school graduates.

Shifting approaches

Along with these new perspectives on aid, donor practices have shifted markedly since the early 1990s as a result of the end of the Cold War and the resolution of the widespread macroeconomic crises that led to the ascendancy of IMF and World Bank stabilization and structural adjustment programs. There is a new emphasis on providing aid more flexibly to well-performing countries, and a stronger focus on achieving measurable results to show the effectiveness of aid-supported programs. Two broad funding patterns have emerged.

First, some donors have moved to providing broad-based program, budget, or sector support, usually for a select group of countries considered to have shown a stronger commitment to sound development policies. Several European donors and multilateral development banks now offer financing for education or health through Sector-Wide Approaches (SWAps), which provide pooled funding to support sector-specific strategies. Similarly, the World Bank has introduced its Poverty Reduction Support Credits to provide support for economywide policy and institutional reform and poverty reduction programs.

These "horizontal" approaches provide greater flexibility for recipients to allocate funding to a range of activities, allow funding for recurrent costs, and focus on building institutions and broader systems. Although these instruments are appropriate in some countries—providing flexibility and discretion makes sense in countries with a strong commitment to good development policies—they will be less so in countries with widespread corruption and poor or destructive policies. Thus, donors have made this kind of financing available in only a relatively small number of countries. The new U.S. Millennium Challenge Account, which will provide support for economic growth initiatives, shares some of the characteristics of these approaches, particularly in its strong country selectivity and focus on local leadership in determining funding allocations. However, it is too early to tell how it will operate, whether it will be used for budget support in some cases, and whether it will provide significant finance to primary education.

Second, several new initiatives provide funding for very specific purposes. The most prominent of these are for health, especially the Global Fund to Fight AIDS, Tuberculosis, and Malaria (the Global Fund) and the Global Alliance for Vaccines and Immunizations (GAVI). In education, a Fast Track Initiative (FTI) has been established to facilitate attainment of education for all. The FTI includes a Catalytic Fund (see "Designing a Global Compact on Education" on page 38 of this issue) to help ensure resource availability to countries neglected by major donors, although it is far less structured, with more general goals than these other funds.

The Global Fund is a Geneva-based foundation set up in early 2002 to stimulate a substantial increase in financing for programs to fight the three diseases. It has grown quickly: as of early 2005, it had approved funding for 294 programs in 129 countries—a larger profile than almost any other donor—with signed two-year grants agreements totalling $2.3 billion and disbursements of over $1 billion. The fund´s structure differs from traditional aid institutions in several ways. It has a very small bureaucracy with no in-country staff, acts only as a financing instrument (it does not implement programs or provide technical assistance), and has a strong focus on results. It relies on recipient-driven strategies with an open and participatory approach and a strong role for civil society groups in both program design and implementation.

GAVI (along with its funding arm, the Vaccine Fund) was founded to boost childhood immunization coverage, which began to fall in the 1990s after making important gains in the 1980s. GAVI is a public-private partnership: its largest backer is the Bill & Melinda Gates Foundation, augmented by donations from several international agencies. Partners include the World Health Organization (WHO), UNICEF, the World Bank, bilateral donors, technical agencies (such as the U.S. Centers for Disease Control and Prevention), and the private sector, including both multinational and developing country vaccine suppliers. It coordinates action and policies in global immunization—for example, by working to align the advice given by WHO and UNICEF to ministries of health about priorities for their immunization programs. It also advances important new funding: in some countries GAVI´s efforts have led to a doubling or tripling of the amount spent on immunization. GAVI´s resources initially support increased coverage in traditional core vaccines and capacity strengthening; over time it will increasingly support the introduction of new vaccines.

A special feature of GAVI is its "reward for performance," through which it provides incentives for expanded coverage in two ways. First, GAVI ceases funding if countries do not achieve specified immunization coverage targets. While many donors make this pledge, GAVI is an exception in that it carries it out. In late 2003, it terminated grants to 10 countries that had failed to achieve specified goals. It has also increased funding for 23 countries that achieved their immunization coverage goals. Second, in some programs it directly rewards performance by simply paying $20 per additional fully immunized child, approximately equivalent to the cost of fully immunizing a child under typical developing country conditions. Half of the money is paid in advance, with the remainder paid upon proof of immunization. There are no strings attached to the use of the funds, other than showing proof of having achieved specified immunization targets.

Does the health sector experience provide lessons for ODA in education? GAVI´s performance-based budget support may have wider applicability for education initiatives, as recipients could be rewarded for increasing student hours in school, teacher-student ratios, number of classrooms, or other measurable intermediate outcomes.

The Global Fund and GAVI have also helped with donor resource mobilization. The growth in ODA for primary education suggests that resources can be mobilized through traditional institutions, although substantially more funding will be needed, including for the FTI´s Catalytic Fund, to achieve universal primary education. The tight focus of vertical programs facilitates the diffusion of best practice, accountability, performance measurement, and communication to political stakeholders. Such programs help ensure that ODA resources actually expand a critical activity rather than substituting for government resources that may be reallocated to altogether different objectives. Budgetary support or SWAps for primary education are, in essence, already vertical programs that can provide knowledge transfer when traditional donor staff are technically up to date. Hence large-scale new enterprises like GAVI and the Global Fund may be less important for primary education, although their success does suggest the value of ensuring that FTI has strong technical capacity and that the Catalytic Fund has adequate resources.

Making aid more effective

Recent initiatives in aid delivery have helped put the focus on making aid more effective and on improving its quality. To reinforce these trends, we believe donors should focus on the following:

ODA needs to be more long-term, more predictable (so long as specified goals are being accomplished) and more flexibly disbursed. Donors have stated their commitment to move in this direction. This is particularly important in the context of program funding, SWAps, and other cases in which ODA is used to fund recurrent expenditures (such as teachers´ pay). Where donor funds are shorter term or less predictable, it is more risky to fund recurrent costs, and in these cases, ODA should focus on technology transfer and capital investment—for example, by revising and printing textbooks, training teachers, developing systems for testing, monitoring and evaluation, upgrading facilities, and improving financial management and procurement capacity.

Budget support for primary education could be more performance-based. As with immunization, coverage of primary education is measurable (public expenditures, enrollment rates, contact-hours per student per year). If ODA is disbursed against indicators like these, it is possible to provide the flexibility of budget support with the accountability of achieving specific objectives. Dean Jamison has outlined mechanisms to ensure that the level of budget support can remain sensitive to the level of performance—that is, incentives can be maintained as GAVI has done—even if the overall level of budget support declines with income growth. Tore Godal, the founding Executive Secretary of GAVI, has recently pointed to the importance of careful—and independent—monitoring and evaluation to judge performance both for release of funds and for learning about approaches that work.

Special funds are needed to fill gaps. The positive experience with special purpose funds in health points to the importance of ensuring adequate resources for the Catalytic Fund of the Education for All Fast Track Initiative.

ODA should support diffusion of best practices. This is likely to require new investment in the technical competence of the staff in ODA agencies. One key purpose of country-specific ODA in education is to diffuse appropriate technology and best practices to improve learning outcomes (test scores) by pointing both to the economic returns to better quality (See "Why Quality Matters in Education" on page 15 of this issue) and to well-established means for improving it. Central governments—supported by ODA—can facilitate diffusion of best practice to school-level decision makers. The box on the previous page reviews areas where best practices have been identified but insufficiently diffused. The interventions listed overlap but go beyond the suggestions of the Millennium Project´s Task Force on Education and Gender Equality.

Research, development, testing, and evaluation (RDT&E) in education should be ODA-funded. Most elements of RDT&E are international public goods. When they are worth financing, they are best financed for the most part from international resources. ODA-financed institutions in agriculture and in health have allocated substantial resources to basic research and new product development. In contrast, the ODA community in education, with some important exceptions, has placed relatively low priority on generation of new knowledge and products (such as educational technologies). This low priority might suggest an assessment that the returns to investment in education RDT&E are low, but a strong case could be made to the contrary. International comparative testing to measure performance in mathematics and science has become a staple of assessing the relative performance of education systems of high-income countries and, with World Bank support, a few low- and middle-income countries have participated as well (see "Picture This" on page 24 of this issue for results from some countries). The case is strong for allocating an increased fraction of ODA for primary education to RDT&E. Possible priority areas include: generation of internationally comparable data on education resource use (national education accounts), on contact hours of instruction and on test scores in key subjects; development and evaluation of improved textbooks, educational technologies, school health interventions, and mechanisms for teaching in multi-grade and multi-lingual environments; research on determinants of enrollment, attendance, repetition and persistence in schooling; and research on determinants of the distribution of cognitive outcomes in age-cohorts of students.

* * * * *

In sum, there is potential for substantial progress by 2015 toward achieving universal primary education. Challenges will remain in terms of how much education is received (contact-hours of instruction per year), in terms of quality, and in terms of the need to go beyond primary education. Well-designed ODA programs have important roles to play in meeting these challenges.