The Economics and Econometrics of Gene-Environment Interplay

Abstract

Economists and social scientists have debated the relative importance of nature (one's genes) and nurture (one's environment) for decades, if not centuries. This debate can now be informed by the ready availability of genetic data in a growing number of social science datasets. This paper explores the potential uses of genetic data in economics, with a focus on estimating the interplay between nature (genes) and nurture (environment). We discuss how economists can benefit from incorporating genetic data into their analyses even when they do not have a direct interest in estimating genetic effects. We argue that gene-environment (G × E) studies can be instrumental for (i) testing economic theory, (ii) uncovering economic or behavioral mechanisms, and (iii) analyzing treatment effect heterogeneity, thereby improving the understanding of how (policy) interventions affect population subgroups. We introduce the reader to essential genetic terminology, develop a conceptual economic model to interpret gene-environment interplay, and provide practical guidance to empirical researchers.

Statistical methods for gene–environment interaction analysis

“…Future research should strive to develop innovative methodologies that integrate genetic, environmental, and functional genomic data to elucidate the functional implications of G × E interactions. Fourth, the current gold-standard of PGS × E analysis involves performing regression within families, which effectively eliminates all bias from population stratification, environmental confounding, assortative mating, and other sources ( Biroli et al, 2022 ). It is an open question whether the methods discussed might have biases for the reasons mentioned above when applied to population-level data, and if so, whether they can be used for within-family analyses.…”

“…Another method, often referred as empirical PGS × E analysis, has become increasingly popular in the field of G × E research ( Biroli et al, 2022 ; Domingue et al, 2020 ; Li et al, 2019 ; Schmitz et al, 2022 ). This approach involves a two-step procedure that begins by summarizing the genetic predisposition of each individual into a polygenic score (PGS).…”

Overcoming Attenuation Bias in Regressions using Polygenic Indices: A Comparison of Approaches

“…In sum, within-family analyses are the gold standard to estimate direct genetic effects, free from bias arising from the omission of parental genotype. However, when using a family fixed effects strategy on the basis of a PGI from a conventional GWAS that did not include parental genotype, the direct genetic effect is biased downward as a result of measurement error, genetic nurture effects and social genetic effects (see also 69). Therefore, this approach provides a lower bound estimate on the direct genetic effects.…”

Reimagining Gene-Environment Interaction Analysis for Human Complex Traits

“…Other studies perform stratified GWAS in different environments and then test differential heritability and/or imperfect genetic correlation between the environments 26,27,32 . Further, PGSxE studies have gained popularity in the GxE literature 3,[33][34][35] . It is a two-step approach that first summarizes each individual's genetic predisposition into a PGS, and then tests the interaction between PGS and the environment 1,[36][37][38] .…”