The international community has recently embarked on an ambitious quest to trigger a substantial acceleration of economic growth and development, in order to try crushing poverty, especially in lower-income countries. Ambitions are centered on an international consensus on a concrete, albeit broad set of development goals, the Millennium Development Goals (MDGs). For low income countries, documents such as the Poverty Reduction Strategy Papers (PRSPs) should translate these broad targets into country-owned development strategies; from the donor side, it has provoked a substantial redrawing of the overall aid architecture, at such a scale that some authors refer to it as a true paradigm shift (Renard, 2006). Apart from the scaling up of aid, improving the effectiveness of aid is a crucial element of concern in this new paradigm: results could be improved, not only by scaling up efforts, but also by rechanneling scarce aid resources towards where they would be most productive in generating growth and reducing poverty. Moreover, it may be the case that aid effectiveness is not independent of the choice of aid modalties or instruments, leading to the importance of a disaggregated analysis of aid flows.
Traditionally, economists have tried to answer the aid effectiveness question using standard cross sectional macro-economic growth models, through the inclusion of aid flows as an explanatory variable. Perhaps not surprisingly, the results were inconclusive, reflecting the heterogeneous way in which aid is both given and used by different governments. More recently, especially since the World Bank Policy Research Report on Assessing Aid (1998), a new consensus, albeit with continuing dissent voices, seems to emerge: If aid is given to countries characterized by good governance, there is a positive impact on economic growth (Van de Walle and Johnston (1996), Burnside and Dollar (2000)). This seems to confirm that, as in Franco-Rodriguez et al. (1998) “the core deficiency of this “aid-growth” literature is that it fails to recognize explicitly that aid is given primarily to the government, and that hence any impact of aid on the economy will depend on government behaviour, in particular how fiscal decisions on taxation and expenditure are affected by aid revenues” (p.1242). Again, the issue is raised that different aid instruments and modalities might provoke very different fiscal reactions of the recipient public sector (Mavrotas, 2002, 2005).
During the last decade, debt relief has slowly made its way as an important element of the international ‘Financing for Development’ agenda, through initiatives such as the Heavily Indebted Poor Countries (HIPC) Initiative, and more recently, at last year’s G8 summit in Gleneagles, where the G8 proposed to go beyond HIPC in granting additional multilateral debt relief to a specific set of countries, which is now known as the Multilateral Debt Reduction Initiative (MDRI). It is important to note that these debt relief initiatives nicely fit into the new aid architecture and the aid effectiveness debate: since debt relief to low-income countries is almost exclusively for debt owed to official creditors, these creditors are the same as those that provide (traditional forms of) aid to those countries. As such, debt relief is just one of the instruments of donor intervention (Berlage et al., 2003). So the crucial question is to what extent, and under which circumstances, debt relief is a more promising instrument than the more ‘traditional’ modes of aid delivery (project aid, program aid, either in the form of concessional loans or grants, technical assistance, etc.). Again here, differences in modalities may lead to differences in effectiveness, and differences in fiscal response behaviour of the recipient public sector. The idea that debt relief might be a more optimal form of aid delivery is in fact implicit in MDRI as this additional debt relief is very similar to budget support-type of aid, and will be granted to the beneficiary countries, i.e.post-HIPCs (only), partly as some substitute for it (Cassimon & Renard, 2006).
As noted by Chauvin and Kraay (2005), the arguments in favour of debt relief have ranged from the moral to the mundane. To concentrate on the latter, economic arguments are at least threefold. First of all, there is the direct effect that debt relief will free up public resources that would otherwise have been earmarked for debt service, i.e. the “fiscal space” (Heller, 2005) argument of debt relief. But even is this is not the case, and debt relief is merely “virtual”, i.e. an ‘accounting clean up’ of historic (and future) arrears accumulation, debt relief might still be an optimal reaction of donors for at least two good reasons. One is the so-called debt overhang argument, affecting both public and private sector behaviour. For the private sector, an excessive debt stock might deter investment for fear of future taxation; for the public sector, Krugman (1988) shows that high debt service obligations reduce the incentive of debtors to engage in policy reforms that raise revenues available for debt service, since part of the additional revenues accrue to the creditor. As a result, on top of the actual debt relief savings, debt relief, by removing debt overhang, may help in relieving some key bottlenecks for growth acceleration in low-income countries. Finally, to the extent that donors may succeed in receiving truly “additional” development budget resources, by using the aid instrument of debt relief, debt relief may transform itself into the optimal instrument of donor policy to the extent that the intervention benefits the recipient country. Again, to correctly measure the impact of debt relief, one should make a detailed analysis of the fiscal response behaviour of a recipient government on donor interventions of the debt relief type. Whatever the theoretical arguments, the actual impact of debt relief essentially remains an empirical question.
Disaggregated analysis of the fiscal response of aid has so far tended to disregard the potential effect of debt relief as a separate aid category, which is rather surprising, as debt relief is becoming a substantial part of aid efforts and because of the reasons explained in the previous paragraph. This paper aims at contributing to the existing aid effectiveness literature by looking at the fiscal effects of debt relief, relative to other types of interventions, for a panel of 28 Heavily Indebted Poor Countries (HIPCs), that have reached at least decision point in the HIPC, and are currently receiving (at least interim) debt relief within the HIPC Initiative. We focus on debt relief granted through the HIPC Initiative (only). In doing so, the paper uses recent insights and methodologies from different parts of the aid effectiveness literature in general, and the recent fiscal response literature in particular.
This paper is organized as follows. The next section provides a brief review of the recent literature related to the aid effectiveness, debt relief as well as fiscal response literature. Section three describes our empirical strategy for the estimating the panel vector autoregressive model and deals with data issues. The fourth section presents the estimation results and the last section concludes, while also discussing some preliminary policy consequences.
CONTENTS
Abstract
1. Introduction
2. Aid effectiveness, debt relief and their fiscal responses: a review of recent literature and practices
- 2.1. Aid effectiveness and fiscal response
2.2. Fiscal response effects of debt relief
2.3. A particular application to HIPC debt relief
3. Model Estimation, empirical strategy and data issues
- 3.1. A panel data VAR for fiscal response modelling
3.2. Data issues
4. A discussion of empirical results
5. Conclusions and policy implications
References
