| Zusammenfassung: | Through its in-depth study of a randomized evaluation of two programs providing additional support to job seekers, this article examines the objectivity of this method of evaluation and its effects on public policy management. I study the different phases of the trial (from its design to the results it produced and how these were used) and analyze them in light of the institutional logics at play when the programs were implemented. Given that these results are not a measurement but a product of conventions and given that, as they reacted to the evaluation, the actors involved transformed the object being observed, the objectivity of randomization is in fact shown to be relative. Analyzing how the results were used partially, in both senses of the term, offers evidence of the ways in which quantitative data are used to legitimate public policy.
|
|---|