THE WORLD BANK IS MOST EFFECTIVE IN COUNTRIES WITH NEWLY ELECTED AND DEMOCRATIC GOVERNMENTS

What determines the success or failure of World Bank reform programmes in the developing world? According to new research by David Dollar and Jakob Svensson published in the latest issue of the Economic Journal, programmes of so-called structural adjustment are far more likely to be successful with relatively new and democratically elected governments. But they are far less likely to be successful in ethnically fragmented societies. What is more, factors that are under the World Bank's control - such as the size of loans or the number of conditions attached to them - have absolutely no relationship with their success or failure.

The researchers note that in the 1980s, foreign aid shifted from financing investments such as roads and dams to promoting policy reform. This change reflected a growing awareness that developing countries were held back more by their own poor policies than by a lack of finance for investment. To promote policy reform, foreign aid came with many strings: detailed 'conditionality' on monetary, fiscal, trade and other policies. Some individual structural adjustment loans of the International Monetary Fund or the World Bank had more than 100 specific conditions.

How effective has detailed conditionality been in promoting real policy change? Dollar and Svensson examine 220 reform programmes supported by the World Bank to understand why they succeed or fail. Of their sample, about one third of the programmes failed to meet their objectives in terms of actual policy change. The researchers find that a few political-economy variables - factors outside of donors' control - can successfully predict whether the reforms will actually take place:

The success of reform is more likely with relatively new governments and democratically elected ones.
Reform is less likely in ethnically fragmented societies, a finding that is consistent with other research, which has found that it is difficult to put sound policies into place in societies that are polarised along ethnic or class lines.
These political economy factors - all of which are known at the time the adjustment loans are approved - correctly predict whether reforms will actually be implemented in 75% of the cases. But factors under the World Bank's control (e.g. the size of the loan, the number of conditions attached to the loan, etc.) have no relationship with success or failure.

The researchers cite the example of Zambia in the 1980s, which entered into four structural adjustment loans with the World Bank and received $212 million to support its policy reform. Evaluation after the fact found that most of the policy measures were not implemented and hence the economic results were poor. The Dollar-Svensson analysis suggests that this result was largely predictable: Zambia at the time had an authoritarian government that had been in power for several decades and as such was not a likely candidate to make major policy reforms.

More generally, the results of this analysis suggest that a key issue for development agencies is to select promising candidates for support. When a poor selection is made, devoting more administrative resources or increasing the number of conditions will not increase the likelihood of successful reform.

Dollar and Svensson conclude, 'If the World Bank and other donor agencies would like to improve their success rate with adjustment programmes, then they must become more selective and do a better job of understanding what are promising environments for reform and what are not. Such a shift would lead to fewer adjustment loans unless there is a significant change in the number of promising reformers.'

Note for Editors: 'What Explains the Success or Failure of Structural Adjustment Programmes?' by David Dollar and Jakob Svensson is published in the October 2000 issue of the Economic Journal. The authors are at the World Bank.

For Further Information: contact David Dollar on 001-202-473-7458 (email: ddollar@worldbank.org); RES Media Consultant Romesh Vaitilingam on 0117-983-9770 or 07768-661095 (email: romesh@compuserve.com); or RES Media Assistant Niall Flynn on 020-7878-2919 (email: nflynn@cepr.org).