Unemployment, the Great Recession, Fragile Families and Child Development

DESCRIPTION (provided by applicant): This proposal seeks funds to study the short-term effects of aggregate unemployment-particularly high unemployment resulting from the Great Recession-on family resources and child development. The specific aims of the study are to: (1) Describe the associations between aggregate unemployment and family resources and child well-being, as well as the associations between family resources and child well-being; (2) Describe the heterogeneity of these associations across different subgroups; and (3) Estimate the impact of aggregate unemployment on family resources and child development and the robustness of individual level associations to the inclusion of previously unobserved variables. Our project builds on Elder's (1974) classic study of the Children of the Great Depression and on Elder and Conger's (2000) study of the Iowa farm crisis and tests hypotheses set forth in their 'family stress model' with better data and methods. The Fragile Families Study (FFS) is particularly well-suited for this analysis. The FFS is a nationally representative, longitudinal, birth cohort survey of children born between 1998 and 2000 in large US cities (population of 200,000 or more), with a large oversample of births to unmarried couples-a very disadvantaged population. Three quarters of the parents have no more than a high school education and nearly 70 percent are black or Hispanic. These data are uniquely rich in measures of the family stress model, including four areas of family resources-economic well-being, parental relationships, parental heath, and parenting behavior-and three areas of child well- being-physical health, emotional health, and cognitive development. Children in the FFS experience very different economic environments at the same age because they live in 20 different cities and because data collection began in different cities at different times and continued for up to a year within each city. To obtain an exogenous measure of the economic environment, we append city level unemployment rates to each observation in accord with the interview date and city residence. Though one analysis estimates the full family stress model, most focus on one or two particular components of the model: child health, parenting, parent relationships, parent health, or family economic well-being. Each uses outcomes from the year 9 survey and stacked data from multiple years and controls for individual level differences in parent and child fixed characteristics. We estimate reduced form models of the impact of city level unemployment with city or individual level fixed effects models for each family resource and child health outcome. To test the robustness of associations between family resources and child health, we add co-variates that have not been measured in previous studies, add lagged measures of the child outcome variables, and estimate individual fixed effects models. By estimating both the impact of aggregate unemployment on family resources and child development and the robustness of individual level associations to the inclusion of previously unobserved variables we come closer than previous estimates to causal estimates of the pathways of the family stress model.