Experimenting With Welfare Reform: The Political Boundaries Of Policy Analysis

Since the emergence of the policy analysis profession in the 1960's, social policy initiatives increasingly have been linked to analytic projects designed to inform policymakers about the potential effects of proposed reforms. In poverty policymaking and other areas, these linkages have taken the form of "controlled analysis" in which experiments and demonstration projects are used to test policy proposals before they are legislated nationally. Although the function of controlled analysis is manifestly nonpolitical, we seek to illuminate its political functions as an institutional channel for poverty politics, albeit a "shadow institution" that operates in the interstices of formal policymaking channels. Reviewing the now-extensive history of controlled analysis in welfare policymaking, we explore how political contests over reform have occurred within its boundaries: how analysis is employed in the politics of agenda-setting, how it articulates and interprets social concerns, and how it is used tactically to divert or incubate policy proposals. We challenge the conventional view of analysis as antidote to politics, contending that it is, in important respects, a creature of politics. In the case of poverty politics, we suggest that controlled analysis narrowed the boundaries of policymaking, reinforcing conservative tendencies that have limited the range of policy possibilities.