Title: Molecular genetics and mid-career economic mobility

Abstract: Reductions in the cost of genetic sequencing have enabled the construction of large datasets including both genetic and phenotypic data. Based on these datasets, polygenic scores (PGSs) summarizing an individual's genetic propensity for educational attainment have been constructed. It is by now well established that this PGS predicts wages, income, and occupational prestige and occupational mobility across generations. It is unknown whether a PGS for educational attainment can predict upward income and occupational mobility even within the peak earning years of an individual. Using data from the Wisconsin Longitudinal Study (WLS), I show that: (i) a PGS for educational attainment predicts wage, income and occupational prestige mobility between 1974 (when respondents were about 36 years of age) and 1992 (when respondents were about 53 years of age), conditional on 1974 values of these variables and a range of covariates; (ii) the effect is not mediated by parental socioeconomic status, is driven primarily by respondents with only a high school education, and is replicated in a within sibling-pair design; (iii) conditional on 1974 outcomes, higher PGS individuals surveyed in 1975 aspired to higher incomes and more prestigious jobs 10 years hence, an effect driven primarily by respondents with more than a high school education; (iv) throughout their employment history, high PGS individuals were more likely to undertake on the job training, and more likely to change job duties during tenure with an employer; and (v) though no more likely to change employers or industries during their careers, high PGS individuals were more likely in 1974 to be working in industries which would experience high wage growth in subsequent decades. These results contribute to our understanding of longitudinal inequality and shed light on the sources of heterogeneity in responses to economic shocks and policy.