Together with an ex-student of mine I wrote a short piece about when and how the World Bank evaluates its projects. We use examples from the Western Balkans and find that the decision to evaluate is non random. The selectivity of who gets evaluated has also implications for the overall assessment of the Bank’s and the recipients’ performance. In particular, after controlling for who gets evaluated, we find that the Bank’s and the recipients performance are positively correlated and arguably mutually influence each other. We end with some policy recommendations about how to improve the evaluation process.
You can find our working paper here: http://cps.ceu.edu/sites/default/files/publications/cps-working-paper-world-bank-evaluation-2015.pdf