This project will construct a contextual database to be merged with Wave III of the National Longitudinal Study of Adolescent Health (Add Health). Working in close collaboration with the Add Health project, it replicates selected contextual variables from waves I and II most likely to be of use to researchers of neighborhood and social policy effects on health and well being during the transition to adulthood. Contexts assessed are: 1) Neighborhoods - Census tract data related to concentrated poverty, racial/ethnic heterogeneity, household types, residential stability, etc. 2) Counties - Census and other data sources related to labor market and marriage market opportunities, religious adherents, and vital statistics; and, 3) States - a variety of data sources on state variation in social policies, such as TANF, Medicaid, EITC, divorce, and child support. Longitudinal contextual data will make possible a wide range of new life course analyses with Add Health, including: 1) Duration and timing of neighborhood experiences; 2) Movement of individuals through neighborhoods of varying characteristics; 3) Effects of changes in neighborhoods themselves; 4) Hierarchical panel and instrumental variable models to minimize selection effect biases; and, 5) More accurate measurements and lagged specifications, through interpolation and extrapolation of contextual characteristics for varying years.
