Newly Released World Bank Dataset a Step Forward for Aid Transparency

The World Bank’s Independent Evaluation Group dataset containing evaluations of almost 10,000 World Bank projects since the early 1960s.

November 2, 2011
Ben Buch, Doug Nicholson

The World Bank’s Independent Evaluation Group dataset containing evaluations of almost 10,000 World Bank projects since the early 1960s. This dataset presents a unique opportunity for researchers to examine the determinants of successful project implementation. The dataset includes a measure of the extent to which a project's major objectives were achieved. This indicator is measured on a six point scale, ranging from highly unsatisfactory to highly satisfactory. The dataset also includes measures of the extent to which the World Bank ensured quality at entry of the operation; the extent to which the World Bank supported effective implementation through appropriate supervision; the extent to which the borrower (including the government and implementing agency or agencies) ensured quality of preparation and implementation and complied with covenants and agreements; and the quality of monitoring and evaluation design, implementation, and utilization.

As a first step, we generated some basic descriptive statistics to gain a better understanding of overall trends in the data. We first converted all six point measures to binary variables: we re-categorized all projects as either “successful” (projects rated moderately satisfactory, satisfactory, or highly satisfactory) or “unsuccessful” (projects rated moderately unsatisfactory, unsatisfactory, or highly unsatisfactory).

After converting the data, we broke down project success by region. The regional groupings fall into three tiers.

Europe, East Asia, and Central Asia all have around 80% project success rates. Meanwhile, Latin America and the Caribbean, the Middle East and North Africa, and South Asia all have close to 75% success rates. Africa lags behind, with project success rates just above 60%.Future research might examine the source of this disparity between Africa and other regions. It would also be interesting to explore the linkages between country performance, World Bank monitoring and evaluation, and project success.

After analyzing the data by region, we also disaggregated project outcomes by the decade in which the project was initiated.

The decade with the highest percentage of satisfactory project outcomes was the 1960s, although we note that the sample size for this decade is much smaller. The success rate drops to around 75% in the 1970s and then again to about 67% in the 1980s. It recovers to around 75% in the 1990s, and increases again to near 80% in the 2000s. More sophisticated econometric techniques could shed some light on the causal mechanisms behind these shifts and whether they correspond to changes in World Bank lending practices. In particular, the sudden drop in World Bank lending effectiveness in the 1980s merits further examination.

In the coming weeks, we’ll be releasing some more posts with additional analysis of these data. The World Bank's IEG should be commended for placing these data in the public domain. We are not aware of any other donor that has published such comprehensive project implementation and evaluation data.

This post was contributed by AidData research assistants Ben Buch (William and Mary '12) and Doug Nicholson (William and Mary '12).