Experimenting with Welfare Reform: Policy Evaluation Research Methods and the Politics of Social Policy

Abstract This article examines the history of randomized experiments in social policy evaluation research, focusing on an early wave of influential “social experiments” pertaining to the welfare reform debates of the 1980s. Existing literature does not account for the sudden rise of this methodology in policy contexts that had previously been considered too complex for conducting randomized experiments. I argue that the early proliferation of social policy experiments involved the enterprising efforts of a new policy research contractor that adapted this methodology in response to changing market demands and a new political context. Using archived research reports, memoirs of organizational leaders, and a survey of methodological literature, I document a subtle shift in the scientific norms and practices of evaluation researchers to illustrate how a politically unaligned body of expertise filled a niche in service of conservative welfare reform. This historical account offers a unique case study of the relationship between expertise and political change, and it locates the rise of randomized trials within broader critiques of economistic approaches to social policy research that ascended after the 1960s.