Evaluations are spreading over a number of policy areas and are being integrated into administrative routines on many levels in organizations. Yet, a central finding in evaluation research is that the utilization of evaluations is limited. As an alternative to conventional rational and political approaches the paper explores how an institutionalist perspective explains this paradox. This article argues that the institutionalist perspective is useful in showing how evaluation procedures are symbolically appropriate in our age of "reflexive modernization"; explaining a number of empirical findings about the actual (non-)utilization of evaluations; and suggesting an alternative horizon of effects of evaluations.
Evaluators must institutionalize their effortsif they are to produce a steady output --Aron Wildawsky 1 IS]olutions, if they are large enough, turn into future problems.--Aron Wildawsky 2If we roughly think of "evaluation" as any systematic, formal, and retrospective collection of data that concerns the merits of organized activity such as public policy programs, 3 and that is believed to provide feed back to decisionmaking about that activity, then we quite clearly live in an age of evaluations. The a m o u n t of evaluations of public programs increased dramatically during the last thirty years or so (Shadish, Cook, and Leviton, 1991). Evaluations are being integrated into administrative routine on m a n y levels and in different institutions (Hellstern, 1986, p. 305). Evaluation skills are mandatory for y o u n g professionals in the public sector (Vedung, 1991,