Using Aid To Reduce Poverty In The Developing World

07 Jun 2004

A series of articles in the June 2004 Economic Journal examines how to make the best use of aid in reducing poverty in the developing world:
· Is a good policy environment conducive to aid effectiveness or does that depend on the institutional framework and the geographic characteristics of the country – and in any case, what constitutes good policies?
· Should aid be given conditional on policy reform, thus enhancing its impact, or is this simply ''pie in the sky'' optimism?
· And how might aid be used for other objectives, for example, in reducing conflict?
 

The effectiveness of aid
Until recently, the dominant view among economists was that aid had largely failed to have a beneficial impact on growth and poverty. Research by World Bank economists David Burnside, Paul Collier and David Dollar has led to a major revaluation of this view.

They argue that aid does work, particularly when targeted at low-income countries, but a good, primarily macroeconomic, policy environment is critical for its success. This is something which, to a considerable extent, aid agencies around the world are beginning to take on board and holds the promise of doubling the effectiveness of aid in taking people out of poverty.

The significance of geography
Carl-Johan Dalgaard, Henrik Hansen and Finn Tarp in this series of articles argue that yes aid can work, even in a bad environment, but it is the geographical realities of the country rather than whether it follows good policies which determines its effectiveness. Hence, and this is critical, a reward/penalty-based aid allocation system based on countries'' policy records is both unfair and inefficient.

Making aid conditional on good policies
Paul Mosley, John Hudson and Arjan Verschoor are more concerned with another aspect of the World Bank position, namely the view that giving aid conditional on the promised pursuit of good policies simply does not work. They question this on the grounds that conditionality has not failed so comprehensively as the World Bank claims and that in any case it can be made to work better by the introduction of graduated penalties for reneging on agreed policy reform. If true, conditionality will allow the development of a dialogue with developing countries and hence facilitate a more tailored set of policy reforms than the World Bank''s approach.

They also argue that, with poverty reduction in mind, the net of good policies needs to be cast wider to include ''pro-poor expenditure'' (PPE). This is government expenditure particularly beneficial to the poor, for example, education and spending on agriculture. Aid has a direct impact on PPE, which in turn is a powerful agent for combating poverty.

Paul Collier and David Dollar meet head on the criticism that the World Bank''s definition of good policies is too narrowly drawn and make use of the Bank''s Country Policy and Institutional Assessment index (CPIA). This evaluates a country''s policy and institutional framework in 20 dimensions, including, for example, economic management and policies for social inclusion. It is a significant step forward.

Using aid to reduce conflict
Collier and Dollar look at corruption, the possibility of working around governments rather than through them and the possible use of aid in reducing conflict and then in coping with post-conflict situations. Thus, they estimate that increasing aid by $1 a day per capita for the average country, reduces the risk of conflict by 10% in five years. With policy reforms also included, this figure rises to 30%.

Dalgaard et al counter this by arguing that the CPIA correlates with the proportion of land in the tropics and it is this, rather than good policies, which acts as the catalyst for aid effectiveness. At the heart of their paper is a much
more rigorous theoretical framework for analysing aid''s effectiveness than has traditionally been the case. This emphasises that strong fundamental structural characteristics, such as institutions and climate-related circumstances, may compensate for a bad policy environment.

This then raises the question as to what extent institutions are determined by accidents of history, geographical factors or respond to current realities including efforts to reform or change them. This may not lead to exciting sound bites but is of critical importance in enhancing our understanding of the growth process.

''Aid and Development'' edited by John Hudson is published in the June 2004 issue of the Economic Journal. Hudson is at the University of Bath.

John Hudson

07798-502891 | j.r.hudson@bath.ac.uk