Genetic advantage and equality of opportunity in education: Two definitions and an empirical illustration (read here)
Co-authors: Single authored paper
Status: Under review
Abstract: Recent advances in the literature of social-science genetics have found credible genetic based measures that predict educational attainment - the educational attainment Polygenic Scores (PGS’s). And while the Equality of Opportunity (EOp) literature has long acknowledged the existence of ‘talents’, ‘innate ability’ or ‘genetic ability’, existing frameworks are not well equipped to incorporate a direct measure of genetic advantage. In addition, there exists prevalent disagreement within both kinds of literature on whether genetic components should be perceived as a fair or unfair source of advantage. This paper proposes an EOp framework that explicitly includes genetic advantage and that encapsulates two EOp definitions; one where genetic advantage is counted as fair inequality and one where genetic advantage is counted as unfair inequality. The PGS is used as a measure of genetic advantage while correcting for genetic nurture and accounting for its correlation with other circumstances. An illustration using the US Health and Retirement Study finds that the share of inequality of opportunity for years of education is 26% under the view that genetic components are unfair sources of advantage and 21% otherwise. A cohort analysis finds decreasing returns to genetic advantage for high school completion and increasing returns for college completion and overall years of education. The results suggest an increasing role of genetic advantage in attaining higher educational levels, with genetic advantage explaining 3.3% of the variation of college completion for older cohorts (1920-1929) and 6.8% for younger cohorts (1950-1959). These results show that different classifications of genetic advantage within the EOp framework lead to different estimates of the level and trend of EOp. They also call for a more transparent discussion of the true meaning of merit in education.
Gene-Environment Interplay in the Social Sciences (read here)
Co-authors: Pietro Biroli, Titus Galama, Stephanie von Hinke, Hans van Kippersluis, Niels A. Rietveld, Kevin Thom
Status: Accepted in the Oxford Review of Economics and Finance
Abstract: Nature (one’s genes) and nurture (one’s environment) jointly contribute to the formation and evolution of health and human capital over the life cycle. This complex interplay between genes and environment can be estimated and quantified using genetic information readily available in a growing number of social science data sets. To help the novice reader interested in understanding individual decision making, public policy, and inequality using genetic data, we introduce essential genetic terminology, review the literature in economics and social-science genetics—with a focus on the interplay between genes and environment—and discuss policy implications and future prospects of the use of genetic data in the social sciences and economics.
Intergenerational transmission of education: estimating genetic components
Co-authors: Hans van Kippersluis
Status: Draft stage
Abstract: How much can genetic transmission explain the intergenerational transmission of education? This question has been a topic of debate for a long time among economists. In the early 2000s economists used twin and adoption studies to conclude that genetic transmission explains half of the father-child association and the entire mother-child association. Fast forward 20 years and the direct measurement of one’s genome is now possible. This paper exploits recently released genetic data to quantify how much of the parent-child association is explained by genetic components. We use the Avon Longitudinal Study of Parents and Children (ALSPAC) data set and estimate the association between parental years of education and their offspring grades of Key Stage 4 national exams. The results suggest a much more modest effect of genetic transmission of about 30% and no meaningful difference between father-child and mother-child education associations. Additionally, we find that accounting for genetic components increases the explained variance of test scores by at least four percentage points. This result suggests that countries might fair worse in intergenerational mobility than initially estimated without accounting for genetic factors.
The interplay between maternal smoking and genes in offspring birth weight (read here)
Co-authors: Hans van Kippersluis, Niels Rietveld
Status: Accepted in the Journal of Human Resources
Abstract: It is well-established that both the child’s genetic endowments as well as maternal smoking during pregnancy impact offspring birth weight. In this paper we move beyond the nature versus nurture debate by investigating the interaction between genetic endowments and this critical prenatal environmental exposure – maternal smoking – in determining birth weight. We draw on longitudinal data from the Avon Longitudinal Study of Parents and Children (ALSPAC) study and replicate our results using the UK Biobank. Genetic endowments of the children are proxied with a polygenic score which is constructed based on the results of a recent genome-wide association of birth weight. We instrument the maternal decision to smoke during pregnancy with a genetic variant (rs1051730) located in the nicotine receptor gene CHRNA3. This genetic variant is associated with the number of cigarettes consumed daily, and we present evidence that this is plausibly the only channel through which the maternal genetic variant affects the child’s birth weight. Additionally, we deal with the under-reporting of maternal smoking by using measures of cotinine, a biomarker of nicotine, collected from the mothers during their pregnancy. We confirm earlier findings that genetic endowments as well as maternal smoking during pregnancy significantly affects the child’s birth weight. However, we do not find evidence of meaningful interactions between genetic endowments and an adverse fetal environment, suggesting that one’s genetic predisposition cannot cushion the damaging effects of maternal smoking.
Rank concordance of polygenic indices: implications for personalized intervention and gene-environment interplay (read here)
Co-authors: Stephanie von Hinke, Hans van Kippersluis, Fleur Meddens, Niels Rietveld
Status: Revise & Re-submit on Nature Human Behavior
Abstract: Polygenic indices (PGIs) are increasingly used to identify individuals at high risk of developing diseases and disorders and are advocated as a screening tool for personalised intervention in medicine and education. The performance of PGIs is typically assessed in terms of the amount of phenotypic variance they explain in independent prediction samples. However, the correct ranking of individuals in the PGI distribution is a more important performance metric when identifying individuals at high genetic risk. We empirically assess the rank concordance between PGIs that are created with different construction methods and discovery samples, focusing on cardiovascular disease (CVD) and educational attainment (EA). We find that the rank correlations between the constructed PGIs vary strongly (Spearman correlations between 0.17 and 0.94 for CVD, and between 0.40 and 0.85 for EA), indicating highly unstable rankings across different PGIs for the same trait. Simulations show that measurement error in PGIs is responsible for a substantial part of PGI rank discordance. Potential consequences for personalised medicine in CVD and research on gene-environment (G×E) interplay are illustrated using data from the UK Biobank.
Using Obviously-Related Instrumental Variables to Increase the Predictive Power of Polygenic Scores (read here)
Co-authors: Hans van Kippersluis, Stephanie von Hinke, Pietro Biroli, Titus J. Galama, S.Fleur W. Meddens, Dilnoza Muslimova, Niels Rietveld
Status: Working paper
Abstract: The conventional way of boosting the predictive power of polygenic scores is to increase the GWAS sample size by meta-analyzing GWAS results of multiple cohorts. In this paper, we challenge this convention. Through simulations, we show that Instrumental Variable (IV) regression using two polygenic scores constructed from independent GWAS summary statistics outperforms the typical Ordinary Least Squares (OLS) model employing a single meta-analysis based polygenic score in terms of bias, root mean squared error, and statistical power. We verify the empirical validity of the simulations by predicting educational attainment (EA) and height in a sample of siblings from the UK Biobank. We show that IV regression between-families approaches the SNP-based heritability, and improves the predictive power of polygenic scores by 12% (height) to 22% (EA). Furthermore, IV regression within-families provides the tightest lower bound for the direct genetic effect, increasing the lower bound for EA from 2.0% to 3.4%, and for height from 28.9% to 37.7%.