Social Experiments in HousingAbstractIn the past quarter century, HUD has funded eight evaluations of social programs in the housing field using random assignment of households or individuals to treatment and control groups. The existence and results of these rigorous studies often are not well known. The object of this paper is to describe the completed demonstrations, to outline the scope and objectives of those still under way, to discuss unfinished business— important issues that experiments might address—and to summarize the power and limits of the experimental method.The paper summarizes completed demonstrations along the following dimensions: timeframe, treatments tested, outcomes of interest, sample size, target population, number of treatment groups, number and location of sites, major findings, design issues, replicability, generalizability, information sources, and whether there is public access to the data. Reports on the ongoing experiments do not discuss findings, design issues, replicability, or generalizability. Both the power and the limits of the experimental methodology in its application to housing issues are explored. Experimental methods allow identification of the response of individuals to program changes, holding markets constant. They do not, however, capture all the impacts on individual behavior when the program changes significantly affect the operations of the markets themselves. Social Experiments in Housing (*.pdf)
|