Abstract
Evaluations are a potentially important tool for democratic governments: they provide a
basis for accountability and policy learning. To contribute to these key functions, evaluations
must be of sufficient methodological quality. However, this quality is threatened by
both political influences and technical complexities. This article describes and explains the
variance in the quality of ex-post legislative (EPL) evaluations conducted by the European
Commission, which is a frontrunner in this realm. A number of potential political and technical
explanations of evaluation quality are tested with a unique, self-constructed dataset
of 153 EPL evaluations. The results show that the Commission’s EPL evaluations usually
apply a robust methodology, while the clarity of their scope, the accuracy of their data and
the foundations of their conclusions are problematic. The variance in this quality is mainly
explained by the type of evaluator: EPL evaluations conducted by external actors are of
higher quality than evaluations conducted internally by the Commission.
