Why organizations (do not) evaluate: a search for necessary and sufficient conditions

The wide acceptance of evaluation in this evidence-based society might hide significant variation in the extent of evaluation activeness between public sector organizations. In explaining these differences, evidence is only fragmentally available. Admittedly, multiple explanatory factors can be identified in the evaluation community, mainly in the evaluation capacity building literature. Yet, common to the practical character of the field, insights are mainly of anecdotic nature and have seldom been systematically tested. Thus far, the only certainty is that ‘contingency’ matters. The inherently contingent nature of evaluation practices may not discourage us, however, from collecting more systematic insight in explaining differences in the extent of evaluation activeness. It is not clear, indeed, to which degree the contingency reigns. The question is whether more parsimonious patterns can nonetheless be discerned, when attacking the complexity. The present paper takes up this challenge. Via a systematic comparison of 27 public sector organizations of the Flemish administration (Belgium) through the application of several configurational comparative techniques (MSDO/MDSO & csQCA), the analysis identifies a range of necessary and sufficient (combinations of) conditions for the (non)conduct of evaluations.