Chapter 3
The changing role of official finance

Back to Contents 

This chapter covers trends in net official finance and aid allocation, efforts to help heavily indebted poor countries achieve sustainable levels of external debt, and the increasing volume and scope of guarantee activities.

Shifting focus of aid

The events of the past two decades have greatly changed the political context of official finance to developing countries. The collapse of centrally planned economies in Eastern Europe and the poor performance of heavily distorted economies in Africa, Latin America, and the Middle East have underlined the importance of providing appropriate incentives for private sector growth. And the need to contain government spending in industrial countries, the end of the cold war, and the evidence of mismanagement of aid flows to some countries have reduced public support for official development assistance. This section shows that:

• The decline in net concessional flows to developing countries in real and nominal terms during the 1990s (including 1997) has been accompanied by efforts to improve the effectiveness of aid by allocating a greater share to countries with better policies. Continued improvements in aid effectiveness will be critical to ensure benefits for developing economies and will provide the most convincing argument for maintaining or increasing concessional assistance in the face of further planned cutbacks by major donors.

Net official finance in 1996–97

Net official development finance, consisting of loans and grants from government agencies and multilateral institutions, has declined in real (and even nominal) terms since the early 1990s. This decline reflects constraints on the availability of concessional resources, the decision by many governments to reduce direct lending as private flows to developing countries have increased, and a shift by some developing countries away from official sources to international capital markets. Net flows also have been reduced by a sharp increase in amortization payments by middle-income countries as the debt deferred by Paris Club creditors during the 1980s has come due.

The decline in net nonconcessional flows was interrupted in 1997 by the international rescue package for Thailand. The package, which included $9 billion of nonconcessional finance, boosted net nonconcessional flows to an estimated $6.9 billion for the year, compared with –$5.5 billion in 1996 (table 3.1). Excluding the Thai package and the early repayment by Mexico of the remaining amounts due under its 1995 rescue package, official nonconcessional flows are estimated at $5.5 billion in 1997, compared with an average of $11.4 billion a year in 1990–92. Net nonconcessional lending by multilateral institutions (excluding the IMF) continued its upward trend over the past few years, a result of strong increases in net flows from the Inter-American Development Bank and the World Bank.

Back to top
Back to Contents

Table 3.1 Net official flows of development finance, 1990–97
(billions of U.S. dollars)

Note: Official concessional finance comprises inflows of official development assistance (excluding technical cooperation grants) and official aid to Eastern Europe and the former Soviet Union. Memo items are not included in preceding aggregates.
a. Preliminary.
Source: World Bank Debtor Reporting System and staff estimates.

Net concessional assistance to developing countries continued its downward trend in 1997. The availability of concessional resources has been constrained by the pressing need for fiscal consolidation in most industrial countries, the declining strategic and military importance of development aid since the end of the cold war, and weak public support for aid in some major donor countries, due in part to skepticism about its effectiveness. Net inflows of official concessional finance, consisting of official development assistance (ODA) and other official aid to low- and middle-income countries, fell by an estimated $2.8 billion in 1997 (following a $5.3 billion decrease in 1996). Dollar appreciation of roughly 10 percent in 1997 explains almost half this decline.1

The decline in development assistance since the 1980s has not been accompanied by any decline in the need for aid. The number of people in low-income countries increased from 2.4 billion in 1980 to 3.2 billion in 1995 (World Bank 1997d). One study showed that the number of people living on less than $1 a day in developing countries rose from 1.2 billion in 1987 to 1.3 billion in 1993 (World Bank 1996). And while life expectancy has climbed in developing countries (from 56 years to 63 for males and from 59 years to 66 for females between 1980 and 1995), access to education and health services has worsened. The number of people 15 and older in developing countries who are illiterate rose from an estimated 848 million in 1980 to 872 million in 1995, and the number of people per hospital bed rose from 921 in 1980 to 950 in 1993 (World Bank 1997d).

Back to top
Back to Contents

The detailed information now available for 1996 confirms the decline in ODA flows. Net ODA flows (including technical cooperation grants) from the OECD’s Development Assistance Committee (DAC) countries fell to 0.25 percent of their GNP in 1996 (figure 3.1), the lowest level recorded since the United Nations adopted a target of 0.7 percent of GNP in 1970. This decline in donors’ ODA effort marks a continuation of the steep fall since the mid-1980s, when total net ODA from DAC members was 0.35 percent of aggregate GNP (figure 3.2).

Back to top
Back to Contents

DAC members provided $55 billion in official development assistance to developing countries and multilateral development banks in 1996, down from $59 billion in 1995, reflecting a decline of 4 percent in real terms.2 Driving this decline was a $5 billion fall in aid from Japan, the largest donor. The sharp decline in Japanese ODA in dollar terms stemmed from a 14 percent depreciation of the yen against the dollar and a 25 percent drop in ODA in yen terms, as difficult economic conditions and pressures for fiscal consolidation constrained Japan’s 1996 ODA budget to 1 trillion yen, down from 1.4 trillion yen in 1995. Also contributing to the decline in Japanese ODA were a delay in payments to the International Development Association (IDA) and other multilateral institutions planned for 1996 due to protracted multilateral negotiations and a rise in repayments of bilateral concessional loans to Japan. Austria, Canada, France, and Portugal also registered significant falls in ODA. But 9 of the 21 DAC member countries increased aid as a share of GNP. ODA flows from the United States rose by 21 percent in real terms in 1996, and those from Italy by 34 percent. In both cases the flows included funds that had been approved but not spent in 1995.

The ranking of countries by their ratio of net ODA to GNP has changed little over the past 10 years, however. Norway, the Netherlands, Denmark, and Sweden continue to have the highest ratios. Among the major donors, Denmark has increased its ODA ratio (from 0.9 percent in 1986 to more than 1 percent in 1996), France has maintained a constant ratio, and the others have seen declines in their ODA ratios of between 15 and 50 percent.

Prospects for ODA allocations are dismal. Preliminary information for 1997 shows that net ODA flows could have declined to as low as 0.21 percent of DAC countries’ GNP.3 In Europe continued efforts to meet the Maastricht target for fiscal deficits (3 percent of GDP) will restrict the scope for real increases in ODA. In Japan, the largest donor by volume, pressure to reduce the fiscal deficit led to a 10 percent cut in the ODA budget for fiscal 1998, and further cuts are expected through 2000. In the United States, which has halved the ratio of net ODA to GNP since the mid-1980s, opinion polls show continued public concern over the funds devoted to foreign aid, in part because of exaggerated perceptions of aid’s share in the budget (currently about 1 percent).4

Many donors are channeling an increasing share of their ODA flows through nongovernmental organizations (NGOs). While in 1980 official sources accounted for less than 10 percent of NGO budgets, by the 1990s their share had risen to 35 percent. But the private resources that NGOs have channeled to developing countries have stagnated at around $6 billion since the mid-1990s and are estimated to have declined to $5.7 billion in 1997.

Aid effectiveness

There is increasing evidence that development assistance is most effective when it supports sound economic policies. Experience over the past 20 years has shown that development projects have little chance of success in an environment of high inflation and severely distorted prices produced by inadequate policies.

In a recent study of World Bank–financed projects, the Bank’s Operations Evaluation Department confirmed that the macroeconomic environment plays a critical part in determining the success of borrowers’ project portfolios (Piciotto 1997; World Bank 1997a). Another study by World Bank researchers provides strong empirical support for the view that sound economic policies are vital to ensure effective use of aid (Burnside and Dollar 1997). It found that large aid flows tend to be associated with faster growth in countries with a sound policy environment—defined as an open trade regime, fiscal discipline, and the absence of high inflation. And it found that aid made little or no contribution to growth in countries with poor policies. In a sample of 41 low-income developing countries in 1970–93, countries with good policies that received large amounts of aid achieved GDP per capita growth rates of 3.5 percent a year, while countries with good policies and small amounts of aid had per capita growth rates of 2 percent. Countries with poor policies achieved virtually no per capita growth, regardless of whether they received large or small amounts of aid.

Back to top
Back to Contents

Other studies have tended to confirm the importance of sound economic policies for aid effectiveness (although their results are sometimes ambiguous). Both Killick (1991) and Kreuger, Michalopoulos, and Ruttan (1989) argued that the effectiveness of economic assistance is strongly affected by the economic policy environment in recipient countries. Boone (1996) found that aid most directly assists the poor (a different focus from Burnside and Dollar, who were more concerned with the impact of aid on growth) in countries that have liberal political regimes and democracies.5

Has aid generally gone to countries with good policies? The evidence is mixed. The literature finds that bilateral aid tends to be allocated more on the basis of strategic considerations than does multilateral assistance. Burnisde and Dollar found that in 1970–93 countries with poor policies received as much bilateral aid relative to income as countries with good policies, reflecting the influence that strategic and political factors have had on bilateral aid allocations. But they found that the allocation of multilateral aid—including assistance from the World Bank’s IDA—favored countries with good policies. Killick (1991), McKinlay and Little (1978), Boone (1996), Kreuger, Michalopoulos, and Ruttan (1989), and Maizels and Nissanke (1984) have also cited the importance of strategic considerations in bilateral aid allocation.

Studies have also provided some support for the view that multilateral aid has been less responsive to strategic considerations. Maizels and Nissanke (1984) found that the allocation of aid by multilateral donors in the late 1980s (though not in the late 1960s) was based on countries’ needs. Trumbull and Wall (1994) found that the needs of recipients played a significant part in the allocation of ODA by multilaterals, although Frey and Schneider (1986) concluded that World Bank lending was best explained by a combination of economic and political indicators rather than by such considerations as need or development potential.

But these studies mostly explain the allocation of aid before the 1990s. The increasing recognition of the importance of the policy environment for aid effectiveness has recently led donors to focus more attention on efforts to increase aid to good performers. One such effort, the Special Program of Assistance for Africa, helps coordinate aid efforts to ensure that African countries that are effectively implementing economic reform programs receive the quick-disbursing assistance they need. The program identifies funding gaps in performing countries and provides timely information on the status of reform programs so that donors can adjust the allocation of assistance accordingly. The program reflects donor countries’ commitment to improving the effectiveness of aid through better coordination of aid delivery at the national and local levels.

Donor countries also are working to coordinate the goals of aid. The DAC has set out a strategic agenda of goals to be achieved in developing countries by 2015 with the support of official assistance. The agenda includes reducing by half the proportion of people living in extreme poverty, ensuring universal primary education in all countries, eliminating gender disparity in primary and secondary education, reducing by two-thirds the mortality rates for infants and children under age five, reducing by three-fourths maternal mortality, providing access to reproductive health services for all individuals of appropriate age, and reversing the current trend in the loss of environmental resources.

Has the increased attention to aid effectiveness and policy performance led to any improvement in donors’ allocation of aid to good performers in recent years? There is some evidence that it has. Between 1990 and 1995 the share of net ODA going to the best-performing countries—the first and second quintiles in a ranking of policy performance by the World Bank—rose from 38 percent to 45 percent (table 3.2). Most of the change stemmed from a decline in the share of the fourth quintile.

Table 3.2 Recipients’ share of official development assistance by policy performance, 1990 and 1995
(percent)

Note: The categorization by performance is based on World Bank ratings that evaluate the effectiveness of economic policies in supporting growth and reducing poverty. The ratings for each year are used, so the composition of each group differs between 1990 and 1995. a. Excluding China, India, and Pakistan, which also borrow from IBRD and other nonconcessional sources of official development finance.
Source: OECD data and World Bank staff estimates.

Back to top
Back to Contents

Donors might use other allocation criteria to enhance the development impact of aid that could lead to less observable improvement in aid allocation by policy performance. For example, donors might allocate more aid to countries heavily dependent on concessional flows than to those with access to nonconcessional flows, or more aid to lower-income than to middle-income countries. But there does not appear to have been any improvement in aid allocation by these criteria during the 1990s. The share of net ODA going to countries with access to private flows (countries with bond ratings from international credit rating agencies) was 33 percent in 1996, essentially the same as it was in the mid-1980s. And the share of net ODA going to low-income countries has declined in the 1990s, from 50 percent in 1991 to 46 percent in 1995. This decline reflects a fall in the share of bilateral ODA to low-income countries (from 45 percent in 1991 to 28 percent in 1996), due in part to the fact that in many large recipients civil war or economic or political conditions have precluded substantial official assistance. The share of multilateral ODA to low-income countries has remained unchanged over the decade, but it has shifted sharply toward humanitarian support (which averaged close to 20 percent in 1994–96).

Another way to determine whether concerns over income levels or capital market access affect the data on aid allocation by policy performance is to look at the allocation among low-income countries (see table 3.2).6 The best-performing 40 percent of these countries received 53 percent of ODA in 1995, up from 36 percent in 1990. But this increase in ODA to the top two quintiles came at the expense of the third quintile; the share allocated to the worst performers changed little. In short, aid allocation did improve among low-income countries, but it is unclear whether it improved more than among all developing countries. And there is inevitably a problem of “adverse selection” in the allocation of some forms of aid. For example, the significant share of aid provided primarily for emergency assistance and humanitarian reasons will go disproportionately to countries where political or economic conditions may not be conducive to good policies.

One area where progress has been made is in donor efforts to increase the concessionality of aid and to minimize the use of ODA loans (as opposed to grants), particularly for countries with an unsustainable debt burden. Several donors now provide all their ODA in the form of grants, most provide all their ODA to heavily indebted poor countries as grants, and many have canceled the ODA obligations of these countries. As a result several donors (Australia, Norway, and the United Kingdom) no longer have any outstanding ODA claims.

Yet the terms on ODA from some donors are hardening, owing to an increase in the share of loans. Despite the fall in total net ODA, new commitments of concessional loans from bilateral donors rose by 15 percent between 1995 and 1996. Because of the lag in actual disbursements, however, this trend will not start to show up in net ODA flows until 1997. The tendency to mix ODA and commercial loans in a single package also appears to be increasing, reversing the trend of the early 1990s. Fully a third of ODA lending committed in 1996 had associated commercial funds, underlining the continued importance of tied aid and the influence of commercial interests on aid flows.

Progress under the HIPC Debt Initiative

The Debt Initiative for Heavily Indebted Poor Countries (HIPCs) was launched in September 1996 out of a recognition that despite considerable efforts by countries and creditors, many of the poorest developing economies still faced unsustainable debt burdens. This section explains that:

The HIPC Debt Initiative has played an important part in the evolution of donor policies and in the increased effectiveness of aid by making the achievement of debt sustainability its explicit goal and by involving all creditors in debt relief efforts. In its first year seven countries that had established the required track record of good economic performance were considered for additional debt relief under the initiative, and agreements were reached to reduce the debt of four of these countries by $1.2 billion in present value terms. The HIPC Debt Initiative is geared to encouraging improved economic and social policies, facilitating the provision of interim finance to strong performers, contributing to a more productive relationship between creditors and debtors, and, ultimately, enabling countries to exit from repeated debt rescheduling exercises.

The HIPC Debt Initiative was formed against a background of weak economic performance and high debt levels in a large number of poor countries. Economic reform programs undertaken by many of these countries have improved economic performance. And debt relief efforts—through such mechanisms as the Paris Club, the IDA Debt Reduction Facility, and the World Bank’s Fifth Dimension Program—have provided substantial benefits. But for many countries the debt relief available through these mechanisms could not ensure that debt sustainability could be achieved within a reasonable period even with good policy performance.

Back to top
Back to Contents

Economic performance and debt of the HIPCs

Forty-one countries are classified as heavily indebted poor countries.7/ Mostly in Africa, these countries account for 12 percent of the debt of developing countries, but less than 5 percent of their exports and only 3 percent of their GNP (table 3.3). Though economic and policy performance has varied widely among the HIPCs, average GDP growth for the group was only 2.2 percent between 1985 and 1990. The GDP growth rate fell to 1 percent during 1990–95, well below the population growth rate of 2.7 percent. The volume of exports from HIPCs rose by 2.7 percent a year in 1985–90, but the increase slowed to 2.2 percent in 1990–95 despite more vigorous reform efforts (table 3.4). The HIPCs’ poor average growth can be attributed in large part to civil disturbances, weak governance, unstable macroeconomic policies, and severely distorted prices in many of the countries and to declining terms of trade during the late 1980s in some. Large debt burdens also have played a part in depressing investment and growth.8/

Table 3.3 Heavily indebted poor countries relative to all developing countries, 1996
(percent)

a. Data are for 1995.
Source: World Bank data.

Table 3.4 Economic performance of the heavily indebted poor countries, 1980–95
(average annual percentage change or percent)

a. Net transfers from official sources.
Source: World Bank data.

The median ratio of the present value of debt to exports among HIPCs is 340 percent, and some have ratios of more than 1,000 percent (the average ratio of debt to exports for other developing countries is 130 percent). Debt burdens of this size raise concern that debt may not be sustainable, meaning that the debtor is unlikely to be able to service the debt and, ultimately, to repay it without resorting to further rescheduling or the accumulation of arrears (box 3.1).

Continue with The changing role of official finance

Back to top
Back to Contents