Performance Assessment, Vulnerability, Human Capital, and the Allocation of Aid Among Developing Countries

Summary

Developing country performance with respect to economic policies and institutional behavior is a common criterion for the allocation of aid among recipient countries. This paper examines how performance is used, arguing that performance is too narrowly defined. A more appropriate definition is one that controls for the economic vulnerability and human capital of developing countries. Econometric analysis of cross-section and panel data is presented that supports this contention. The paper also contends that performance and exogenous economic shocks are likely to be pro-cyclical. This implies a double punishment when aid is allocated according to performance. Evidence of such punishment is also provided. The paper concludes by arguing that economic vulnerability and human capital variables should augment performance measures in aid allocation decision-making.

Introduction

Performance is an important criterion for the allocation of aid among developing countries. Some donors, such as the World Bank and the African Development Bank, allocate aid among recipient countries according to formal prescriptive models that incorporate performance as a determining variable. The amount of aid prescribed by these models for a given recipient country is an increasing function of its performance. The World Bank International Development Association (IDA) performance-based allocation system is perhaps the best-known model of this type. This system relies heavily on the well-known World Bank Country Institutional and Performance Assessment (CPIA).¹ Other donors allocate aid in a less systematic manner but often provide more aid to given countries than would otherwise be the case on the basis of superior performance. Performance, in this context, is defined in terms of the efficacy of economic and social policies and public sector management and institutions. The efficacy of public sector management is typically based on factors including the perceived quality of budgetary and financial management, revenue mobilization and public administration, and on the transparency and accountability of the public sector.

The rationale for basing aid receipts on performance, as defined, is twofold. The first reflects a widely held view in the donor community, that the incremental impact of aid on recipient country economic growth is an increasing function of the performance of these countries. Ensuring that the allocation of aid among recipient countries is an increasing function of their performance is consistent with maximizing the global economic growth and poverty reduction efficiency of development aid. In short, it is thought to maximize aid effectiveness from these perspectives. This is consistent with the very well-known Burnside and Dollar (2000) and Collier and Dollar, 2001, Collier and Dollar, 2002 studies. The second rationale follows from a strategy of ex-post conditionality. Better economic policies and institutional performance are (rightly) thought to be better for economic growth. Rewarding high-performing countries with more aid promotes further growth in them, and provides positive incentives for other countries to improve their performance.

There is much to be said in support of basing aid allocation on recipient country performance using an indicator such as the CPIA. Above all, it ensures that developmental criteria determine aid allocation rather than political and commercial criteria, and imposes order on what can otherwise be a chaotic decision-making process that results in illogical outcomes. If it is transparent recipient countries can understand the amounts of aid they have been allocated and what they need to do to receive a larger share of aid from the donor in question in future. It is not without weakness, however. Three weaknesses are especially apparent.

First, its assessments of performance are somewhat narrow, ignoring what might be described as initial country conditions. An important such condition is human capital. It is reasonable to expect that countries with low human capital levels are likely to have low performance scores. They may have difficulties formulating economic and social policies and achieving high-quality budgetary and financial outcomes, mobilizing revenues and achieving public transparency. A country with low human capital might achieve relatively low performance assessments, despite the best of intentions and huge efforts. Penalizing such a country with less aid than would otherwise be the case is not consistent with a strategy for providing positive incentive to improve performance.

Second, performance-based allocative approaches are highly reductionist with respect to aid effectiveness criteria. Aid effectiveness will be contingent on a number of factors in addition to performance. The literature on aid and growth has pointed to a number of contingencies, including economic vulnerability. Guillaumont and Chauvet (2001) show that the incremental impact of aid on growth is contingent on recipient country structural economic vulnerability, with the former being an increasing function of the latter.² Allocating more aid to economically vulnerable countries than would otherwise be the case is consistent with maximizing the global growth and, by implication, poverty efficiency of aid. Ignoring this criterion reduces this efficiency. Put differently, not taking into account vulnerability means the effectiveness of aid in promoting global growth, and in turn reducing global poverty, will be lower than would otherwise be the case.³

The third weakness relates to the impact of exogenous economic shocks on performance. There is strong reason to expect that performance will be partly driven by these shocks or, more specifically, will be pro-cyclical with respect to them. A negative exogenous shock can be bad enough in its own right, but will be made all-the-worse if accompanied by lower aid. This is a case, therefore, of double punishment.

This paper empirically assesses these criticisms. While each might make sound intuitive sense a priori, and might be supported by country-specific or anecdotal evidence, they have more veracity if the behavioral relationships they describe can be observed across reasonably large samples of developing countries. This paper provides a multivariate econometric analysis of developing country performance. Performance is measured using the World Bank CPIA and Country Performance Rating measures, which are used to allocate IDA funds among eligible developing countries. The explanatory variables of interest are economic vulnerability, human capital and measures of exogenous economic shocks.

The paper consists of three further sections. Section 2 provides a brief review of the literature on the determinants of performance. Section 3 examines whether performance is determined by vulnerability, human capital and income. It reports the methods used by and results from an econometric analysis of these relationships. Section 4 explores the issue of pro-cyclicality of performance with respect to exogenous shocks, reporting the results of an econometric analysis. Section 5 concludes.