Aid and Poverty Reduction
by Oliver Morrissey
Aid effectiveness has attracted considerable attention in the economic development literature since the late 1990s, both in terms of research publications and policy debates. Something of a consensus is emerging that aid does have a positive impact on growth, although debate remains as to whether or not good policy is necessary to ensure aid effectiveness. Recently, research has begun to examine the impact of aid on poverty reduction. This reflects the increasing emphasis being placed on poverty reduction in policy debates, and demands that the objective of reducing poverty requires an increase in aid to poor countries (e.g. the UN Millennium Project and the report of the Commission for Africa). In this context, what do we know about the effects of aid on reducing poverty?
Many will be familiar with arguments along the lines that growth reduces poverty, so if aid is effective in increasing economic growth it will contribute to reducing poverty. Advocates of this argument tend to suggest targeting aid to countries with large numbers of poor people that also display an ability to use aid to increase growth (typically, good policy is accepted as evidence of this ability). This is a reasonable argument (if growth benefits the poor and aid increases growth, then aid benefits the poor), but is not the whole story. On the one hand, growth is not necessarily pro-poor (even if sustained growth is important for sustained poverty reduction). On the other hand, growth is not the only way that aid can benefit the poor; in stagnant economies aid may be able to benefit the poor, and even in a growing economy aid may ensure greater benefits go to the poor.
It is important to allow for the various ways in which aid can impact on the poor. Furthermore, it is important to take a broad concept of poverty. Improvements in aggregate welfare (better health and education for example) may benefit the lives of the poor just as much as reductions in income poverty. Aid that generates income-earning opportunities or that provides social services, such as donor-funded projects in health or sanitation, can increase welfare directly and may be targeted on the poor (even if targeting is imperfect). Aid that contributes to economic growth should lead to long-run increases in aggregate welfare. Furthermore, much aid is directed through government spending, and aid can increase welfare by increasing expenditures towards those social services that contribute to welfare. Box 1 summarizes the main channels.
The research findings discussed here allow for these various channels, especially through the level and allocation of public spending. Although some studies do use data on poverty, comparative cross- country data on poverty over time is extremely scarce, and such data as exist are based on income measures of poverty (which do not capture all dimensions of poverty and are not fully comparable across countries). Researchers avail of the fact that there is a strong correlation between levels of poverty and levels of aggregate human welfare across countries, specifically as measured by the human development index (HDI) and the infant mortality rate (IMR). If there is evidence that aid is associated with higher welfare (higher HDI or lower IMR), then it is likely that aid benefits the poor (at least by improving access to public services).
Evidence on Aid and Welfare
In an effort to try and capture some of the channels through which aid affects welfare, one can include government spending in addition to aid as an explanatory variable. One has to be careful in doing this to avoid double-counting, as much of the aid directly finances government spending (and is therefore 'present in both variables). The results presented here control for this (in effect, by stripping out the aid component of government spend- ing). This still leaves the question of how to measure spending to capture only those components most likely to benefit the poor. The literature suggests that social sector expenditures - health, education, and sanitation - are the most likely to be pro-poor. Social spending not only increases human welfare but tends to do so in a manner that is pro-poor. Thus, one can define 'pro-poor public expenditure (PPE) as the sum of spending on health, education, and water and sanitation (as a share of GDP).
Tables 1 and 2 present results for HDI and IMR respectively. Three explanatory variables are reported. Initial GDP controls for the tendency of richer countries to have higher levels of welfare. The PPE variable captures the efficiency of public spending in increasing welfare. If the coefficient is positive and significant, this implies that public spending (and aid that finances such spending) contributes to increasing welfare. The coefficient on the Aid variable captures the effects of aid, either directly through aid projects or indirectly through growth (as it is lagged).
There is robust evidence that aid improves welfare indicators, HDI and IMR, and this effect is predominantly through direct impacts (aid provides incomes or social services) or growth. This beneficial effect of aid is present for low-income and middle-income countries. We only provide illustra- tive results here, but this is a common finding in a number of recent studies. However, only for middle-income countries is PPE associated with increased welfare, and here only for the HDI measure. This suggests an issue of concern, as government spending does not appear to be very efficient. This may be in part because the levels of PPE spending are quite low, especially in the poorest countries.
Aid effectiveness in benefiting the poor could be enhanced if the efficacy of public spending is increased. Attempts to increase the targeting of expenditure in areas that are more likely to benefit the poor could yield a high pay-off. The use of aid to guide or influence the allocation of government spending offers a way to increase the leverage of aid on poverty alleviation (the final channel in Box 1). Increasingly, aid is being used in the way we consider, to support public spending under the HIPC initiative for example.
Debt Relief, Aid, and the Poor
In most respects debt relief is equivalent to aid and can be treated accordingly. Under HIPC, countries are required to establish a good record of implementing economic and social policy reform and prepare a Poverty Reduction Strategy Paper (PRSP) indicating how they will tackle poverty reduction. The funds made available by debt relief would be then channelled into poverty- reduction, typically through a Poverty Action Fund (PAF) that identifies pro-poor expenditures (i.e. the PRSP has a similar emphasis to PPE above).
To qualify for debt relief countries must demonstrate their ability for sound economic management through satisfactory implementation of policy reforms over three years under IMF and World Bank programmes. The inherent defect with this approach is that the resources to fund pro-poor expendi- tures are not released fully until the end of the process. The essential pro-poor policies in PRSPs can be considered under two headings - those relating to the provision of and access to public services (like PPE) and those relating to the rural sector; increased public spending on the provision of social services is a central element of PRSPs. The role of debt relief itself is to provide increased govern- ment resources to finance these pro-poor policies.
It is generally easier to identify and implement pro-poor expenditures than it is to implement an economic reform programme that includes pro-poor policies. If the primary objective is poverty reduction, therefore, the prior policy is pro-poor expenditures. Pro-poor policies, however desirable, are of secondary priority. The current approach to HIPC conditionality reverses these priorities. This is not necessary, as eligibility for the release of resources (aid and debt relief) could be based on pro-poor expenditure criteria. Reversing these implicit priorities could enhance the provision and effectiveness of debt relief.
Aid can and does contribute to poverty reduction, by contributing to growth, by providing direct benefits to the poor, and by supporting and financing increased social sector spending. It is evident that more can be done, especially in improving the effectiveness of public spending in delivering improvements in welfare, especially of the poor. Aid and donors will continue to play an important role. It is also evident that greater coherence is needed - debt relief should be aligned with aid in tackling poverty, and this means that the emphasis should be on pro-poor spending and providing services. This can be achieved before or in conjunction with macroeconomic policy reform, it does not need to wait until after.
Oliver Morrissey is Professor in Development Economics and the Director of the Centre for Research in Economic Development and International Trade (CREDIT), School of Economics, University of Nottingham.