The Intergenerational Transmission of Welfare Participation: Facts and Possible Causes

Abstract

This article explores two methodological issues in measuring intergenerational correlations in welfare participation. First, a method is proposed that controls for differences in eligibility as well as participation. Second, the use of event history analysis allows all available information on mothers' and daughters' welfare histories to be used. The National Longitudinal Survey of Youth is used to measure the intergenerational correlation among whites, blacks, and Hispanics. Three broad conclusions emerge. First, parental participation in Aid to Families with Dependent Children (AFDC) is correlated with daughters' AFDC participation for whites and Hispanics—daughters living in households that received assistance have higher probabilities of having a child and receiving assistance in the year of first birth than similar daughters whose parents didn't participate (or participated for shorter periods). Second, parents' participation does not seem to be capturing solely the effects of low income which leads to a correlation in mothers' and daughters' eligibility. Finally, the loss of income if the parent does not participate raises the probability that the daughter will receive assistance. The effect of this income loss offsets nearly half of the participation effect for whites.