Working Paper 56 - A Statistical Analysis of Determinants of Project Success: Examples from the African Development Bank

This paper analyzes some of the critical factors that determine project success inthe African Development Bank on the basis of data available on 146 projectscompleted by 1995. The analysis relies heavily on these observed data and doesnot take into account issues of project quality and implementation which haveassumed greater importance since 1995. The paper looks at the relationship betweenthe economic rates of return at appraisal and at completion for some 56 projectswhere such data were available. First, simple ordinary least squares models areused to explain some of the divergences in the economic rates of return at appraisaland completion stages. This is done by regressing economic rates of return atcompletion on the rates determined at appraisal. The model is then extended toinclude sector dummies and time and cost overruns. An alternative specificationuses simple probit modeling techniques to assess the effect of a number of variableson the probability of project success. The variables included in the analysis areGDP, population size, project size (proxied by cost), time and cost overruns andsector dummies.The findings of the study suggest that the rates of return at appraisal are at bestweak indicators of project performance. There are invariably huge disparities betweenrates of return at appraisal and those at completion, suggesting that if not qualifiedthe former might not be good policy guides. Further, the results suggest that a goodpolicy environment, as shown by rates of economic growth, inflation and thecountry’s level of development, is as important for project success as are the projectspecific characteristics, such as size of project or its sectoral description.