How Costly is Welfare Stigma? Separating Psychological Costs from Time Costs

This paper empirically decomposes the costs associated with participation in welfare by developing a structural model of labor supply and multiple program participation to better understand the barriers of participation in social insurance programs. The well-documented fact that many individuals who are eligible for welfare choose not to participate implies that there is a sizable cost associated with participation, but little is known about the composition of this cost. Prior estimates have not differentiated psychological costs of participation, or “stigma,” from the time and effort required to become eligible and maintain eligibility (time costs). However, the relative size of these two costs has implications for welfare reform. Applying our model to data from the Survey of Income and Program Participation, 1996 (SIPP96), we find that psychological costs are substantially larger than the time costs associated with participation and we find that psychological costs increase with educational attainment and are lower for women of a minority race. To determine whether the psychological cost of participation borne at the individual-level reduces social welfare by a commensurate amount, this paper considers a model in which psychological costs may discourage the use of government assistance as a substitute for work in the context of asymmetric information about an individual’s ability. We estimate a negative correlation between the preference for leisure and psychological costs, which suggests that the incidence of psychological costs is inconsistent with these costs acting as an effective screening mechanism. Carlson School of Management, 321 19th Avenue South, Minneapolis, MN 55455. [email protected] Department of Economics, 403 West State Street, West Lafayette, IN 47907. [email protected] Acknowledgments: We would like to thank Gopi Goda, Giacomo De Giorgi, Tom MaCurdy, Anita Alves Pena, John Pencavel, Luigi Pistaferri, and Frank Wolak for their help on this project as well as participants in the Applied Economics seminar at the University of Minnesota and seminar participants at Purdue University. In addition, we would like to thank Tae-Youn Park for research assistance.