Economic Mobility: The Impact of Individual, Parent and Spatial Factors Using National Survey and Administrative Data

University Of Maryland, College Park, College Park MD

Investigators

Abstract

Economic and social mobility across generations has become an important policy issue world-wide. However, researchers have not effectively identified the causes of this lack of mobility, partly because of lack of appropriate data. This project addresses the causes of lack of economic and social mobility by examining the relationship between childhood circumstances and later adult earnings, education, child-bearing, and incarceration. The relationship between parents' earnings and children's later adult earnings is very relevant to concerns about lack of mobility for American children. Using information for millions of children, this research measures economic mobility in a detailed manner. A key contribution of this research is that it combines numerous sources of information from the Census Bureau and other government agencies to identify demographic characteristics of children such as race, neighborhood, family structure, and level of parents' education. The researchers also measure time spent in subsidized housing and its effect on future earnings. Using this information, the researchers can compare mobility levels across childhood circumstances. In addition to employment and earnings, the researchers also look at the effects of these childhood circumstances on schooling, geographical mobility (to different parts of the country), when (and if) they marry, and at what age they have children in their adult lives. The combined information that is developed, including the data, can be used by other researchers for analysis of policies based on evidence. The results of this research will strengthen the U.S. economy by providing policy guidance on improving intergenerational mobility. This project supports the PIs ongoing development and analysis of data infrastructure that merges national survey and administrative data - - especially LEHD data with Census micro data - - that enables them to study the impact of childhood socioeconomic circumstances on adult outcomes such as educational attainment, labor market performance, fertility, and incarceration. The merged data set provides detailed (in terms of both scale and longitudinal scope) information on employment and earnings, demographics and household characteristics, residential location and neighborhood, and program participation, for millions of individuals across the United States. The PIs will the use these data to explore and make causal inferences of the effects of childhood circumstances on intergenerational economic mobility (IEM), including using data on siblings to estimate of the exposure to subsidized housing when young. For the broader population, the PIs will explore how individual, household, and spatial factors interact with parental characteristics to determine the observed distribution of IEM, adult incarceration, educational attainment, and fertility. The data allow the PIs to explore the mechanisms at work using a variety of quasi-experimental approaches employed in the recent literature to identify causal associations. In another approach to identification, the PIs examine the effects of "forced-moves" stemming from public housing demolitions under the Hope IV program---providing another quasi-experimental variation relating housing for children to adult outcomes. Because selected demolitions were in different neighborhoods and metro areas, the PIs can explore the spatial variation in the long-term effect of public housing demolitions on the youth. This treatment effect of heterogeneity, not previously exploited, will help us to better understand how childhood circumstances interact with residential experiences to affect adult well-being. The method used to combine these data sets will provide a guide for other researchers who may need to combine Census and other data sets to create more comprehensive data sets.