Toan Quoc Nguyen
©2005 International Monetary Fund
[Preface] [Can Debt Relief Boost Growth in Poor Countries?]
Twenty-eight heavily indebted poor countries (HIPCs) were receiving debt relief under the HIPC Initiative by mid-2004, eight years after the Initiative was launched by the IMF and the World Bank and endorsed by governments around the world, and about four years after it was enhanced to provide more substantial and faster debt relief.
The HIPC Initiative, the first coordinated effort by the international financial community to reduce the foreign debt of the world’s poorest countries, was based on the theory that economic growth in these countries was being stifled by heavy debt burdens, making it virtually impossible for them to escape poverty. However, most of the empirical research to date on the effects of debt on growth has lumped together a diverse group of countries, including both emerging market and low-income countries; the literature focusing on the impact of debt on low-income countries (those with 2001 per capita gross national income of less than US$865) is scant.
The paper on which this pamphlet is based, “External Debt, Public Investment, and Growth in Low-Income Countries” (IMF Working Paper No. 03/249, December 2003), addresses this gap in the literature. The paper also appeared as a chapter in a book published by the IMF in 2004, Helping Countries Develop: The Role of Fiscal Policy, edited by Sanjeev Gupta, Benedict Clements, and Gabriela Inchauste. It assesses empirically the effects of external debt on growth in low-income countries and analyzes the channels through which these effects are transmitted, giving special attention to the indirect effects of external debt on growth through its impact on public investment. Readers seeking a more detailed description of our analysis and of the literature on debt and growth are directed to the original working paper, which is available free of charge at www.imf.org/pubs. Brenda Szittya prepared the text for this pamphlet.
The 1996 launch of the Heavily Indebted Poor Countries (HIPC) Initiative by the IMF and the World Bank revived a long-standing debate over the relationship between foreign borrowing and economic growth. The goal of the HIPC Initiative—which provides comprehensive debt relief to poor nations struggling to service heavy foreign debt burdens—is to prevent unsustainable debt burdens from hampering development in the world's poorest nations. Indeed, one of the principal motivations for the HIPC Initiative is concern that a heavy debt burden compromises economic growth.
Analysts have extensively studied the effect of foreign debt on growth, but few of these studies have focused on low-income countries. However, differences between emerging market countries and low-income countries make it likely that foreign debt affects the two groups of countries differently. Unlike emerging market countries, for example, very poor countries have limited access to international capital markets. And their dissimilar economic structures and public sectors may mean that debt affects growth differently in the two groups. Finally, the aid that donors provide to low-income countries could mitigate any negative effects debt service obligations might have on their economic activity. An analysis of the debt/growth relationship in low-income countries could therefore be especially useful in assessing the effectiveness of the HIPC Initiative in enhancing growth.
The theoretical literature on the relationship between external debt and economic growth has focused largely on the harmful effects of a country's "debt overhang"—the accumulation of a stock of debt so large as to threaten the country's ability to repay its past loans, which, in turn, scares off potential lenders and investors. That is, if a country's debt level is expected to exceed the country's repayment ability with some probability in the future, expected debt service is likely to be an increasing function of the country's output level. Thus, some of the returns from investing in the domestic economy are effectively "taxed away" by existing foreign creditors, and investment by domestic and foreign investors is discouraged.
Debt overhang also depresses growth by increasing investors' uncertainty about actions the government might take to meet its onerous debt-servicing obligations. As the stock of public sector debt rises, investors may worry that the government will finance its debt-service obligations through distortionary measures, such as rapidly increasing the money supply (which causes inflation). Amid such uncertainty, wary would-be private investors tend to remain on the sidelines. And even when they do invest, they are more likely to opt for projects with quick returns rather than for projects that enhance growth on a sustainable basis over the long run.
Moreover, debt overhang may also discourage efforts by the government to carry out structural and fiscal reforms that could strengthen the country's economic growth and fiscal position, because a government whose financial position is improving almost inevitably finds itself under increasing pressure to repay foreign creditors. This disincentive to reform would exist in any country with a heavy external debt burden, but it is of special concern in low-income countries, where structural reforms are essential to sustain higher growth.
Of course, not all foreign borrowing dampens investment and growth. At low levels of debt, additional foreign borrowing could stimulate growth, to the extent that the additional capital financed by this new borrowing enhances the country's productive capacity. Higher output, in turn, would make it easier for a country to service its debt. As debt and the capital stock increase, however, the marginal productivity of investment falls. Some analysts have argued that only above a certain threshold will additional foreign loans have a negative impact on growth, owing to the debt-overhang considerations explained above. That is, up to a certain threshold, increased borrowing makes repayment of debt more likely. But, beyond that threshold, further increases in foreign debt reduce the prospects of creditors being repaid. As a country's access to loans drops, its ability to accumulate capital suffers, and growth may slow. In short, the negative effects of debt overhang are likely to take effect only after a certain threshold level has been reached.
The empirical literature has found mixed support for the debt-overhang hypothesis. Most models of the determinants of growth have presumed that the stock of debt affects growth both directly (by reducing a government's incentives to undertake structural reforms) and indirectly (by dampening investment). But relatively few studies have assessed the direct effects of the debt stock on investment in low-income countries econometrically. A 2001 review of studies on the debt-overhang hypothesis by Geske Dijkstra and Niels Hermes found the empirical evidence on this issue to be inconclusive ("The Uncertainty of Debt Service Payments and Economic Growth of Highly Indebted Poor Countries: Is There a Case for Debt Relief?" (unpublished; Helsinki: United Nations University)). And few studies have been able to determine how large the stock of external debt has to be, relative to GDP, for the debt overhang to have an effect.
A 2002 study of 93 developing countries between 1969 and 1998, "External Debt and Growth," by Catherine Pattillo, Helene Poirson, and Luca Ricci (IMF Working Paper No. 02/69), found strong support for the debt-overhang hypothesis, however. The authors found that external debt began to have a negative impact on growth when its net present value exceeded 160–170 percent of exports and 35–40 percent of GDP. Their simulations suggest that doubling the average stock of external debt in these countries would slow down annual per capita growth by between a half and a full percentage point.
In a follow-up study in 2004, "What Are the Channels Through Which External Debt Affects Growth?" (IMF WP/04/15), Pattillo and her coauthors applied a growth-accounting framework to a group of 61 developing countries over 1969–98 and found that doubling their average external debt level reduces growth of both per capita physical capital and total factor productivity by almost 1 percentage point. In other words, large debt stocks negatively affect growth by dampening both physical capital accumulation and total factor productivity growth.
In theory, the service of external public debt (the payment of interest and repayment of principal)—to be distinguished from the stock of external debt—may also affect growth by discouraging private investment or altering the composition of public spending. Higher external interest payments can increase a country's budget deficit, thereby reducing public savings. This, in turn, may either drive up interest rates or crowd out the credit available for private investment, depressing economic growth. Larger debt-service payments can also inhibit growth by squeezing the public resources available for investment in infrastructure and human capital. Indeed, such nongovernmental organizations as Oxfam International see high external debt service as a key obstacle to meeting basic human needs in developing countries. But relatively few empirical studies have tested these hypotheses by assessing the effects of debt service payments on private investment or on the composition of public spending, and the available empirical evidence is mixed.
Using data for 1970–99 for 55 low-income countries (see box on page 11) classified as eligible for the IMF's Poverty Reduction and Growth Facility, which provides concessional loans at low interest rates (0.5 percent a year), we estimated equations to identify the key determinants of the growth of real per capita income (GDP).
To account for the role of debt, we augmented the standard growth model with four widely used debt variables—the face value and the net present value of the stock of external public debt, each as a share of GDP and as a share of exports of goods and services.
In principle, the net present value of debt reflects the degree of concessionality of loans—long repayment periods and below-market interest rates—and is therefore a more accurate measure of the expected burden of future debt-service payments than the face value of debt.
In the model, the chief determinants of growth of real per capita income (GDP) are
Estimated results provide some support for the debt-overhang hypothesis and suggest a threshold of about 30–37 percent of GDP. Beyond that threshold, higher external debt is associated with lower growth rates for per capita income, independent of any effect debt may have on gross domestic investment.
Debt service, in contrast with the stock of debt, has no direct effect on growth, perhaps because its influence is realized through its impact on investment, which is also included as an explanatory variable in the model and is thus held constant. Gross investment has a significant positive impact on growth. Lagged GDP has a statistically significant negative impact. The central government's fiscal balance has a significant positive effect, consistent with recent research that found links between sound fiscal policy and economic growth, while population growth and terms of trade shocks are statistically significant and negative. Openness and secondary school enrollment have no discernible effect. The insignificance of the latter could be due to the modest range of educational attainment levels in our sample of low-income countries. Thus, despite the fact that a relationship between education and growth may exist for developing countries as a group, it was not possible to quantify such a relationship for the low-income countries chosen for this study.
Reestimating the growth equations, but disaggregating gross investment into private and public investment, suggests that it is public investment that affects growth in low-income countries. For each percentage point of GDP increase in public investment, annual per capita growth rises 0.2 percentage point. High levels of public investment that increase budget deficits do not necessarily lead to faster growth, however, because larger budget deficits have a dampening effect on economic activity. In the reestimated equations, changes in the terms of trade, population growth, and openness have no significant effect on growth. As before, debt service has no direct effect. With respect to the stock of debt, the results are once again consistent with the debt-overhang hypothesis—the marginal impact of external debt on growth becomes negative beyond a threshold ratio of debt-to-GDP of about 50 percent of GDP for the face value of debt and 20–25 percent of GDP for the estimated net present value of debt.
These findings have important implications for the impact of HIPC debt relief. The weighted average net present value of external debt to GDP for the 27 HIPCs that had reached the "decision point" as of July 2003—reaching this point means they have been initially approved for partial HIPC debt relief—is projected to decline from 60 percent before debt relief at the decision point to 30 percent by 2005. By that time, most of the HIPCs are expected to have reached their "completion points"—that is, they will have implemented key policy reforms, maintained macroeconomic stability, and implemented a poverty reduction strategy for at least one year, making them eligible for full HIPC debt relief. (Fourteen countries had reached their completion points by the end of July 2004.) Based on our estimates, this debt reduction would, other things being equal, directly add 0.8–1.1 percentage point to these countries' annual per capita GDP growth rates.
These findings imply a more powerful relationship between debt and growth in poor countries than researchers have found in developing countries generally. And the effect of debt on growth is greater when the effects of debt on public investment and the central government's fiscal balance, both of which influence growth, are taken into account.
Analysts have as yet done relatively little research on the determinants of public investment in low-income developing countries. In 2001, Jan-Egbert Sturm at the University of Kostanz in Germany modeled public investment in developing countries using three sets of determinants: (1) structural variables such as urbanization and population growth; (2) economic variables such as real GDP growth, government debt, budget deficits, and foreign aid; and (3) politico-institutional variables to measure, for example, political stability. Sturm found the politico-institutional variables less significant than the structural and economic variables. (We did not include them in our empirical analysis of public investment, not only because institutional variables have not been found to be significant in explaining public investment in developing countries but also, and more important, because of lack of data.) We model public investment as being a function of (1) urbanization, (2) total debt service as a share of GDP, (3) foreign aid, (4) openness, (5) lagged real per capita income, and (6) the same four indicators of the stock of external debt used in the growth model.
The theoretical impact of urbanization on public investment is ambiguous. On the one hand, it could be argued that, as a society becomes urbanized, the provision of services like education and health care shift from the family to the government; thus, one might expect urbanization to increase public investment. On the other hand, most public capital spending concerns physical infrastructure, for which rural areas have a relatively greater need. Thus, urbanization could weaken demand for physical infrastructure while, perhaps, strengthening demand for public consumption spending; as a result, public investment would decrease.
We measured total debt service as a percentage of GDP rather than as a share of exports because this appears, at least intuitively, to be the measure that would most affect government decisions about public investment. However, the relationship between debt service and public investment is not necessarily a linear one. It is plausible that low debt-service payments have no perceptible impact on public investment, but that, as debt service absorbs a growing share of national income, it begins to crowd out public investment. It could be that crowding-out occurs only after debt service exceeds a certain threshold.
Increased foreign aid would be expected to enable governments to spend more on public investment. Openness would be expected to increase public investment because more open economies often compete for foreign direct investment, for example, for financing infrastructure projects. As before, we used the lagged per capita income variable as a proxy for the level of economic development.
Estimates of the public investment equation yield some interesting results. The openness of a country's economy and the receipt of foreign aid seem to boost public investment substantially, while urbanization dampens it considerably. Higher real per capita income also boosts public investment significantly in low-income countries.
The stock of external debt has no significant effect on public investment; public investment seems to be driven more by the gov-ernment's current fiscal position and the availability of resources than by factors that affect fiscal sustainability over the longer term. However, the results support the hypothesis that higher debt service (as opposed to the stock of external debt) crowds out public investment. The relationship is nonlinear, with the crowding-out effect intensifying as the ratio of debt service to GDP rises. On average, for every percentage point of GDP increase in debt service, public investment declines by about 0.2 percentage point of GDP. In some sense, the modest magnitude of this decline is surprising, indicating that large debt burdens have not seriously hampered public investment in low-income countries. More important, it implies that, all things being equal, debt relief by itself cannot be expected to lead to large increases in public investment. In most cases, debt relief leads either to greater public consumption or, if used to reduce government deficits or to lower taxes, to greater private consumption and investment.
If only a small share of debt relief is channeled into public investment, the corresponding impact on growth will also be modest. For example, a reduction in the ratio of debt service to GDP from 8.7 percent (the average in 2000 of the seven most heavily indebted poor countries) to 3.0 percent (roughly the average debt service-to-GDP ratio for all highly indebted poor countries in 2002) would increase public investment by 0.7–0.8 percentage point of GDP and indirectly raise real per capita GDP growth by 0.1–0.2 percentage point annually. Still, this small boost to growth (in absolute terms) is roughly equal to the actual growth in per capita incomes achieved by heavily indebted poor countries during the 1990s. Moreover, if half (instead of a fifth) of this debt service relief were channeled to public investment, annual per capita growth would rise quite significantly (about 0.5 percentage point a year). Under all scenarios, the positive impact of greater public investment on growth will be offset, in part or in full, if financed through larger budget deficits.
Although high levels of debt can depress economic growth in low-income countries, external debt slows growth only after its face value reaches a threshold level estimated to be about 50 percent of GDP (or, in net present value terms, 20–25 percent of GDP). These findings imply that the substantial reduction in external debt projected for the countries participating in the HIPC Initiative would directly add 0.8–1.1 percent to their per capita GDP growth rates. Indeed, the positive effects of debt relief may already be reflected in some of the healthier growth rates achieved by these countries in the past few years relative to their poor performance in the 1990s. (Annual GDP growth averaged 1.2 percent in 2000–02, compared with 0.2 percent during the 1990s.)
External debt also affects growth indirectly through its effect on public investment. Although the stock of public debt does not appear to depress public investment, the cost of servicing the debt does. The relationship is nonlinear, with the crowding-out effect intensifying as the ratio of debt service to GDP rises. On average, every percentage point increase in debt service as a share of GDP reduces public investment by about 0.2 percentage point, implying that reducing debt service by about 6 percentage points of GDP would raise public investment by 0.75–1.0 percentage point of GDP, which, in turn, would result in a modest increase of about 0.2 percentage point in growth. But if a greater share of this debt relief— say, about half—could be channeled into public investment, growth could increase by 0.5 percentage point a year.
While each low-income country participating in the HIPC Initiative determines its use of debt relief in the context of its own poverty reduction strategy, the findings here suggest that one way for country authorities to raise growth and combat poverty would be to allocate a substantial share of debt relief to public investment. As noted earlier, the full benefits of higher public investment will be reaped only if greater public spending on capital outlays is not associated with increasing budget deficits.
These findings have important implications for the design of adjustment programs in countries receiving debt relief. Reducing the stock of debt alone—rather than immediately reducing debt service— is unlikely to induce governments to increase their spending on public investment. And, although cutting debt-service obligations can provide countries with the breathing space they need to increase public investment, debt relief by itself is likely to raise public investment only modestly. In practice, most countries participating in the HIPC Initiative have been raising public investment while receiving financial support for their reform programs from the IMF's Poverty Reduction and Growth Facility (PRGF). (On average, they have targeted an increase of 0.5 percentage point of GDP in public investment expenditure relative to the year before they started their PRGF-supported reform program.)
Given the significance of debt's indirect effect on growth through public investment, it may be useful for researchers to focus on other indirect channels through which debt affects growth. In particular, the finding that stronger central government fiscal balances contribute to growth suggests that the relationship between debt and public sector deficits merits further examination.