The Heterogeneous Effects of Education on Health and Productivity

点击翻译按钮获取中文摘要

PROJECT SUMMARY This application responds to PAR-16-080, an NIH FOA to advance scientific understanding of “the nature of the causal relationship between education and health.” We propose to study how education has shaped the life course of health and economic well-being for the generation of Americans born in the 1920s and 1930s. The premise of our research agenda is that education is determined by three factors: First, it is shaped by parents' decisions and circumstances. Second, local institutions and community characteristics also matter. And third, these contextual factors can interact with family background, strengthening or weakening the intergenerational transmission of health and economic well-being. For example, compulsory schooling laws and child labor laws compel parents to keep their children in school through some minimum age. To the extent the laws are enforced, they can potentially narrow the education gap between children from richer and poorer families, leading to changes in the degree of inequality in health and social outcomes in later life. We propose a research program that studies how parental circumstances and contextual factors jointly affect educational attainment in childhood and traces its effects on health and economic outcomes over the life course. To accomplish this, we need a data set that (1) links individuals across generations; (2) can be merged with local contextual variables; and (3) includes life-course measures of economic and health outcomes. Fortunately, a recent collaboration between the Census Bureau and academic scholars, the Core Longitudinal Infrastructure Project (CLIP), provides the backbone of our data needs. CLIP allows us to match 1940 US Census records on children and youth to near-population data from the Social Security Administration's Master Beneficiary Record File (MBRF) and the Current Population Survey (CPS). To gain health outcomes, we propose to link the National Health Interview Survey (NHIS) to this effort. The 1940 Census contains 42.7 million children aged 0-18, most of whom were living with parents and siblings at the time. The MBRF includes measures of lifetime earnings, disability status, and age at death for 95% or more of people in these birth cohorts, while the CPS has data on schooling and occupational attainment, and the NHIS on health behaviors and outcomes for large subsamples. We will use these data to describe the correlations between family background, education, longevity, and lifetime earnings, and the between-sibling correlations in the latter outcomes. We will document how these correlations differ by gender, race, and ethnicity, and vary across regions of the United States. We will also merge information on schooling laws and local school characteristics, and conduct causal studies of the effect of education on health and lifetime earnings. All files will be made available to other researchers through the Census and NCHS Research Data Centers. These data—which will span one hundred years—will be well-suited for our proposed research on how education shapes lifelong well-being in terms of health and socioeconomic metrics.