Genetic basis of STEM occupational choice and regional economic performance: A UK Biobank Genome-wide Association Study

doi:10.21203/rs.3.rs-2040131/v1

Science, technology, engineering, and mathematics (STEM) professionals are regarded as the highly skilled labor force that fosters economic productivity, enterprise innovation, and international competitiveness of a country. We conducted a genome-wide association study on the occupational choice of STEM jobs based on a sample of 178,976 participants from the UK Biobank database. We identified two genetic loci significantly associated with participants’ STEM job choices: rs10048736 on chromosome 2 and rs12903858 on chromosome 15. The SNP heritability of STEM occupations was estimated to be 4.2%. We also found phenotypic and genetic evidence of assortative mating in STEM occupations. At the local authority level, we found that the average polygenic score of STEM is significantly and robustly associated with several metrics of regional economic performance. The current study expands our knowledge of the genetic basis of occupational choice and potential regional disparities in socioeconomic developments.

STEM

occupational choice

genome-wide association study

polygenic score

assortative mating

comparative economic development

STEM (Science, technology, engineering, and mathematics) jobs involve the use of technical knowledge and skills to solve daily and societal problems (Boon, 2019). It is widely agreed that STEM professionals make up the crucial part of the workforce that sustains a country’s economic growth, enterprise innovation, and global competitiveness (Deming and Noray, 2018). However, in the global labor market, several countries, such as the United States and the United Kingdom, are increasingly facing a shortage of STEM-related professionals, especially in the government sector and private industry (Xue and Larson, 2015; Radu, 2018). These trends have drawn substantial public attention regarding how to reform and improve STEM education as a mean to promote STEM career choices. A growing body of work has examined a wide range of individual characteristics (e.g., personality traits, mathematical ability, motivation, childhood exposure, and ability beliefs) and environmental factors (e.g., parental occupation, peer effects, and gender stereotypes) that may influence individuals’ choice of STEM jobs (Bell et al., 2019; Van Tuijl et al., 2016; Master et al., 2021). However, only little is known about whether one’s tendency to become a STEM professional has a deeper root—genetic predisposition for occupational choice.

On the other hand, several recent studies have documented the heritability of non-STEM occupations. For example, using data on twins, Nicolaou and Shane (2010) found that the choice to become a teacher, manager, salesperson, or self-employed was highly heritable. Roeling et al. (2017) estimated that the heritability of working in a creative profession was approximately 0.70. More recently, Song et al. (2022) identified nine genetic loci significantly associated with leadership positions based on a large-scale genome-wide association study. Yet to the best of our knowledge, no specific genetic variants have been empirically linked to occupational choice in STEM fields thus far.

The principal objective of this study is twofold. First, we aim to advance our knowledge of the genetic predisposition engaging in occupational choice at the individual level, by conducting a genome-wide association study of STEM occupations using the large-scale UK Biobank (UKB) database. We adopted the general definition used in labor economics studies and measured the STEM phenotype based on the UK Standard Occupational Classification (SOC) 2000 system. We also calculated the heritability of STEM occupations and tested for assortative mating based on the main GWAS results. Second, we investigated the genetic associations between STEM occupations and a number of regional economic performance metrics. This allows us to explore the potential genetic basis of regional socioeconomic disparities discussed in the literature on comparative economic development (Ashraf and Galor, 2018).

Measures of STEM occupations in UKB

Our analysis draws on a unique dataset from the UK Biobank. UK Biobank is a population-based sample that comprises over 500 thousand UK residents recruited since 2006. During a verbal interview conducted by trained staff, participants were asked about their job status and job title categorized by the Standard Occupational Classification 2000 system. We used these questions to construct a phenotypic measure of STEM occupations. In our analysis, all UK Biobank participants with occupations within the sub-major category of “Science and Technology Professionals” were classified as having a STEM job, and we coded these participants as STEM professionals (1) and others as non-STEM professionals (0). This occupation sub-category contains a broad range of science, engineering, information, and communication technology professionals. A detailed list of occupations included in the case-cohort is documented in Supplementary Table 1. Participants with missing occupational data were excluded from the present study (35.3%).

Genotyping and imputation

Genotypes of the UK Biobank data set were analyzed with the Affymetrix (Santa Clara, CA, USA) UK Biobank Axiom Array and the UK BiLEVE Axiom Array. Further information about quality control procedures, imputation, and principal components analysis can be found in Bycroft et al. (2018). Participants were removed from the present study based on the following characteristics: non-British ancestry, high missingness, relatedness, quality control failure, and gender mismatch.

Statistical analyses

We performed GWAS using the machine-learning-based method REGENIE (Mbatchou et al., 2021) of the STEM trait across 9,273 cases and 169,703 controls of European descent from the UK Biobank data. REGENIE is an efficient two-step whole-genome regression method for genetic association tests that accounts for both sample relatedness and population structure. Firth logistic regression test was used for the binary trait of STEM jobs. The REGENIE method involves two steps. In step 1, we included directly genotyped variants (a total of 588,699) with a minor allele frequency (MAF) greater than 1%, missingness lower than 10%, and Hardy-Weinberg equilibrium test p-value greater than 1×10^-15. In step 2, we performed the single-variant association testing using the imputed dataset (a total of 7,677,418 autosomal variants) with covariates of age, sex, and the first ten ancestral principal components. Independent significant loci were identified as those with p < 5×10^-8, r² < 0.1, and distance > 250 kb. We then used LDSC (Bulik-Sullivan et al., 2015) to examine genetic correlations across the STEM career choice and a range of behavior and physiological traits. Finally, we conducted a phenome-wide association study (PheWAS) to identify phenotypes that lead SNPs were associated with from previous GWAS work using the GWAS Atlas (Watanabe et al., 2019).

GWAS for STEM occupations

As reported in Table 1, the data involve 178,876 participants of European ancestry that passed quality control from the UK Biobank database, among whom 9,273 (5.2%) had a STEM career. In the pooled sample (N = 178,876), 47.8% were males, 34.9% held a college or university degree, and the average age was 54.3 (s.d. = 7.6). The share of male participants among STEM professionals (N = 9,273) was 83.5%, which is much higher than that among non-STEM workers (45.9%), suggesting a notable gender gap in the professional STEM field in the UK labor market.

Table 1

Summary statistics of the analytical sample in UK Biobank

Variable	(1) Pooled		(2) STEM participants		(3) Non-STEM participants
Variable	Mean	Std. Dev.	Mean	Std. Dev.	Mean	Std. Dev.
Age	54.3	7.6	53.4	7.9	54.4	7.6
Male	47.8%	-	83.5%	-	45.9%	-
College/university degree	34.9%	-	56.9%	-	33.8%	-
STEM occupation	5.2%	-	100.0%	-	0.0%	-
Height (centimeters)	170.8	9.2	175.4	8.1	170.5	9.2
Weight (kilograms)	78.6	15.9	82.9	14.7	78.4	16.0
BMI	27.3	4.7	27.0	4.2	27.3	4.7
Household size	2.6	1.3	2.6	1.3	2.6	1.3
Household income (pounds)	65593.9	46458.2	78087.8	44435.4	64915.4	46469.5

N	178,976		9,273		169,703

We identified two genome-wide significant (p < 5e-8) single-nucleotide polymorphisms for the binary trait of STEM (Figure 1 and Table 2): rs10048736 on chromosome 2 and rs12903858 on chromosome 15. Using LocusZoom (Pruim et al., 2010), SNPs within 500 kb of the lead SNPs were plotted in Figure 1. The genomic inflation factors (λ_cc) was estimated to be 1.07, suggesting no substantial impact of population structure or unmodeled relatedness. The NHGRI-EBI GWAS Catalog showed links between lead SNPs and genes with neuroticism, educational attainment, body mass index, sleep-related phenotypes, corpus callosum mid-posterior volume, and Parkinson’s disease progression.

Table 2

Summary of top loci for STEM trait identified from GWAS

SNP	CHR	BP	A1	A2	EAF	Beta	SE	P	Nearest Gene(s)	Function
rs10048736	2	144239303	A	G	0.36	0.086	0.014	1.46E-09	*ARHGAP15*	Intron Variant
rs12691680	2	144204338	T	C	0.36	0.086	0.014	1.94E-09	ARHGAP15	Intron Variant
rs12903858	15	45972166	C	T	0.43	-0.086	0.014	2.26E-09	*SQOR/SQRDL*	Intron Variant
rs770075436	2	144228576	T	TAAAG	0.35	0.085	0.014	2.45E-09	ARHGAP15	Intron Variant
rs13411140	2	144215811	T	C	0.36	0.085	0.014	2.74E-09	ARHGAP15	Intron Variant
rs28684621	2	144231584	G	T	0.36	0.084	0.014	3.77E-09	ARHGAP15	Intron Variant
rs56081031	2	144197423	C	A	0.27	-0.092	0.016	4.56E-09	ARHGAP15	Intron Variant
rs28380327	2	144232491	T	A	0.36	0.084	0.014	4.59E-09	ARHGAP15	Intron Variant
rs35789697	2	144248718	A	G	0.35	0.084	0.014	4.89E-09	ARHGAP15	Intron Variant
rs2381455	2	144182917	G	T	0.27	-0.089	0.016	1.73E-08	ARHGAP15	Intron Variant
rs6705184	2	144265362	G	C	0.5	0.079	0.014	1.96E-08	ARHGAP15	Intron Variant
rs4662334	2	144241887	G	A	0.36	0.08	0.014	2.20E-08	ARHGAP15	Intron Variant
rs12617059	2	144238363	T	G	0.27	-0.087	0.016	2.91E-08	ARHGAP15	Intron Variant
rs12999615	2	144149614	A	T	0.27	-0.087	0.016	3.08E-08	ARHGAP15	Intron Variant
rs4402695	2	144259542	T	C	0.27	-0.087	0.016	3.94E-08	ARHGAP15	Intron Variant
rs57014442	2	144220812	AAT	A	0.27	-0.086	0.016	4.04E-08	ARHGAP15	Intron Variant
rs35564832	2	144237261	G	A	0.27	-0.086	0.016	4.14E-08	ARHGAP15	Intron Variant
rs13020925	2	144263416	C	T	0.27	-0.087	0.016	4.52E-08	ARHGAP15	Intron Variant

Heritability and genetic correlations of STEM occupations

For the trait of STEM occupation, the LDSC SNP-based heritability (h²) was estimated to be 4.2% (95% CI, 2.8% to 5.6%). The SNP heritability of STEM occupation is smaller than physical traits but generally in accordance with some of the behavioral traits, such as risk tolerance (h² = 4.6%; Linnér et al., 2019) and leadership position (h² ranged from 3% to 8%; Song et al., 2022)

Next, we examined genetic correlations between STEM occupations with 17 potentially linked personal traits using GWAS summary statistics from previous studies, including educational attainment, intelligence, personality, risk preference, height, brain volume, income, and sleep duration. As shown in Figure 2, we found significant positive genetic correlations of STEM occupation with educational attainment (r_g=0.68, 95% CI, 0.55 to 0.82), cognitive ability (r_g=0.62, 95% CI, 0.48 to 0.75), intelligence (r_g=0.60, 95% CI, 0.44-0.75), household income (r_g=0.45, 95% CI, 0.32 to 0.59), noncognitive ability (r_g=0.42, 95% CI, 0.30 to 0.53), and sleep duration (r_g=0.12, 95% CI, 0.00 to 0.23); and significant negative genetic correlations with ever smoked regularly (r_g=-0.41, 95% CI, -0.53 to -0.29), insomnia (r_g=-0.29, 95% CI, -0.43 to -0.15), morningness (r_g=-0.24, 95% CI, -0.35 to -0.12), risk-taking (r_g=-0.22, 95% CI, -0.34 to -0.10), drinks per week (r_g=-0.18, 95% CI, -0.28 to -0.07), neuroticism (r_g=-0.17, 95% CI, -0.28 to -0.07), and number of children (r_g=-0.16, 95% CI, -0.31 to -0.00).

The explanatory power of the STEM polygenic score

To examine the explanatory power of the polygenic score (PGS) of STEM, we constructed the PGS based on the GWAS results (p-value threshold of 0.05; standardized between 0 and 1) and obtained parameters from multiple regression analyses of various socioeconomic outcomes, including STEM jobs, educational attainment (measured by whether a participant holds a college/university degree), and household income (Table 3). All regressions were adjusted for age, age-squared, sex (except for rows b and c), educational attainment (except for column 2), and the first ten genetic principal components (PCs) of each participant.

In the pooled UKB cohort (Table 3, row a), a 0.1 increase in the STEM polygenic score was associated with a 7.6% increase in the probability of being a STEM professional, a 4.6% increase in the probability of earning a college/university degree, and 634-pound increase in annual household income. Moreover, for the phenotype of STEM jobs, the STEM polygenic score accounted for 3.5% of the variance. This is on top of sex and educational attainment, which accounted for 0.1% and 1.4% of the variance of STEM jobs, respectively. Interestingly, the estimated effects of PGS on the propensities to have a STEM job and a college/university degree were found to be more pronounced in males than in females, but the opposite is found for household income (Table 3, rows b and c).

We also estimated the effects of STEM PGS in five independent UK Biobank cohorts with ethnic backgrounds different from the original discovery dataset (i.e., White British), including Irish (N = 13,108), Indian (N = 5,835), Caribbean (N = 4,420), African (N = 3,308), and Chinese (N = 1,538). As shown in Table 3, rows d to h, the STEM polygenic score is significantly associated with the phenotype of STEM jobs in all five independent UK Biobank subsamples (p < 0.0001). For educational attainment, the PGS of STEM was found to be a significant predictor in Irish (p < 0.0001), Indian (p <0.0001), and African (p = 0.009) subsamples. For household income, the STEM PGS was found to be a significant predictor only in the Caribbean (p = 0.011) and African (p = 0.038) subsamples.

Table 3

Parameter estimates of STEM polygenic score in explaining socioeconomic attainment

	(1) STEM jobs			(2) Educational attainment			(3) Income
UKB Cohort	beta	p	N	beta	p	N	beta	p	N
(a) Pooled	0.764	<0.001	315,400	0.455	<0.001	487,202	6340.0	<0.001	415,766
(b) Males	1.169	<0.001	150,713	0.532	<0.001	223,043	5395.1	<0.001	198,227
(c) Females	0.318	<0.001	164,687	0.379	<0.001	264,159	6744.2	<0.001	217,539
(d) Irish	0.667	<0.001	8,617	0.395	<0.001	12,707	5826.5	0.246	11,133
(e) Indian	0.990	<0.001	4,092	0.459	<0.001	5,660	13444.8	0.150	4,253
(f) Caribbean	0.522	<0.001	3,217	0.185	0.081	4,295	21886.0	0.011	3,305
(g) African	1.399	<0.001	2,212	0.417	0.009	3,202	24189.6	0.038	2,445
(h) Chinese	1.629	<0.001	1,112	0.218	0.264	1,502	615.1	0.977	1,201

Notes: All regressions were adjusted by age, age-squared, sex (except for rows b and c), and the first ten genetic principal components of each participant. Regressions of STEM jobs and income were also adjusted by whether holding a college/university degree.

Assortative mating and intergenerational occupational transmission of STEM occupations

To test for assortative mating of STEM jobs at both genetic and phenotypic levels, we identified 39,985 couples as pairs of unrelated (i.e., genetic relationship < 0.05) opposite-sex individuals matched on several household variables as described previously (Yengo et al., 2018; Cheesman et al., 2020). We found significant correlations between couples for STEM occupations both genetically (ρ = 4.3%, p < 0.0001) and phenotypically (ρ = 11.0%, p < 0.0001), lending support to the existence of assortative mating and economic homogamy from the perspective of occupational choices (Yengo et al., 2018; Gonalons-Pons et al., 2017).

We also identified 3,708 parent-offspring pairs based on the kinship coefficients described in Cheeseman et al. (2020) to examine the intergenerational transmission of STEM occupations. Surprisingly, although we found a significant genetic correlation of STEM jobs between parents and offspring (ρ = 52.9%, p < 0.0001), the phenotypic correlation of actual STEM occupational choices is insignificant (ρ = 4.5%, p =0.21), implying a potential higher intergenerational occupational mobility and more employment opportunities for offspring of STEM parents (Modalsli, 2017).

Associations between average STEM polygenic score and regional economic performance

Finally, we investigated whether the regional average STEM polygenic score is associated with the economic performance of local administrative authorities in the UK. Figure 3 shows the geographic distributions of the average STEM polygenic score at the local authority level based on both current home address (a) and birthplace (b) provided in the UK Biobank. We tested the association between the average STEM polygenic score (based on home address) and four publicly available indicators of regional economic performance, including gross domestic product (GDP) from 1998 to 2019, value-added tax (VAT) from 1998 to 2019, counts of business by industry groups in 2016, and employments of business by 17 industry groups in 2016 of 380 local administrative units in UK Biobank dataset. All regional data were collected from the UK’s Office for National Statistics (ONS) website.

As shown in Figure 4, controlling for regional population and aggregate regional indicators (i.e., East Midlands, East of England, London, North East, North West, Scotland, South East, South West, Wales, West Midlands, and Yorkshire and The Humber), the STEM PGS has a statistically significant association with VAT (panel b; 1998-2009, and 2016-2019), but not with GDP (panel a). In terms of local business counts, a 0.1 increase in the average STEM polygenic score is significantly associated with 942 more business counts in professional, scientific & technical, 250 more in information & communication, 245 more in construction, 215 more in arts, entertainment, recreation and other services, 96 more in production, 44 more business counts in education, 35 more in motor trades, and 16 more in public administration & defence. Similarly, for local employments of business, a 0.1 increase in the average STEM polygenic score is significantly associated with 3,415 more employment in professional, scientific & technical, 1,176 more in arts, entertainment, recreation and other services, 1,079 more in agriculture, forestry & fishing, 868 more in education, 783 more in construction, 256 more in motor trades, and 113 more in public administration & defence.

Regional business counts and employment are crucial indicators of the local market and business performance, and the value-added tax has been linked to economic efficiency (Adhikari, 2020). Thus, as discussed by Abdellaoui et al. (2019), the results presented here imply that some of the regional economic outcomes are directionally linked to STEM-associated alleles detected by GWAS conducted in the current study. But such links must be interpreted with caution as they may not necessarily be causal. Other factors such as labor migration and business reallocation might drive these links.

Against the background of an increasing shortage of STEM-related professionals in several countries such as the United States and the United Kingdom, the role of STEM education and jobs in revitalizing economic growth and the labor market in these countries has been subject to heated debate (Xue and Larson, 2015; Radu, 2018). The present study helps to inform this debate by empirically linking individuals’ STEM-related job choices with their genetic markups and estimating the associations between STEM polygenic scores and the indicators of individual socioeconomic attainment and regional economic prosperity. Probing into these previously unexplored topics based on a large-scale dataset provided in the UK Biobank, our GWAS also contributes to the fields of education, labor, and regional economics.

Our study first identified two independent genetic loci associated with participants’ choice of STEM jobs (the most significant SNP in each of the two loci being rs10048736 on chromosome 2 and rs12903858 on chromosome 15). Further estimations revealed strong positive associations between individuals’ STEM PGS and their socioeconomic attainment (measured by college attendance and household income). We also demonstrated statistically significant associations between the average STEM PGS and several indicators of regional economic performance (including local VAT, business counts, and business employment). Pieced together, these findings suggest that driven by STEM-related genetic markups, individuals choosing a STEM-related career may perform better in the labor market (through completing more years of STEM education), which may in turn improve local economic performance.

While most of our findings point in the same direction, important heterogeneity remains. First, there is a significant gender gap in the STEM PGS-socioeconomic attainment association. More specifically, the estimated associations between the STEM PGS and the probabilities of having a STEM job and a college/university degree are much larger in males than in females. This difference might reflect the influence of environmental factors (such as gender stereotyping) that suppress the effect of STEM PGS among females. Interestingly, the STEM PGS-household income association is more pronounced in females, which might reflect factors in the marriage market that we were unable to account for in our analysis.

Secondly, there are also differences in the STEM PGS-socioeconomic attainment association across ethnic subgroups. Take the Chinese subgroup, for example. While the STEM PGS-STEM job linkage among Chinese is highest among all five ethnic subgroups examined, the STEM PGS-college degree and STEM PGS-income linkages (—the latter is statistically significant) among Chinese are nearly the lowest among the five. These ethnic disparities might be due to cultural factors. For example, while the genetic makeup induces Chinese individuals to pursue STEM education, the possible lack of entrepreneurship in the Chinese culture may dampen the effect of STEM PGS on their income. A potentially segmented labor market in the UK could also be a factor. These ethnic disparities suggest that while promoting STEM education may be able to revitalize economic progress and promote individual welfare in general, policies tailored to different subgroups’ cultural and ethnic backgrounds may be needed to fully realize the potential of STEM education and STEM jobs.

Before closing, a note on the limitations of this study is in order. First, despite the wealth of information provided in the UK Biobank database, our study still lacks sufficient controls for environmental factors. Thus, the associations between STEM PGS and economic indicators we found might not necessarily reflect causation; environmental factors might moderate this association.

Second, data limitations may have also prevented us from depicting some of the associations of interest. For example, the UK Biobank contains data only on household income rather than wage or labor income at the individual level. As such, household composition and factors from the marriage market may affect the estimated effect of STEM PGS on individuals’ labor productivity.

Finally, as with other GWAS, a fundamental methodological challenge is that population stratification can occur even in homogeneous populations (Uffelmann et al., 2021). In that case, the associations we found might have been driven by demographic changes rather than genetic factors.

Despite these limitations, we believe that our analysis has provided novel and valuable information to advance knowledge about genetic predisposition engaging in STEM job choices. We also hope our study can attract more GWAS or related studies on the nexus of genetic disposition, educational attainment, and labor market performance.

Funding: This research was supported financially by the National Natural Science Foundation of China (Nos. 72103187, 71973136, and 71973134), the Academy of Global Food Economics and Policy (AGFEP), the Beijing Food Safety Policy and Strategy (FSP) Research Base, and the 2115 Talent Development Program at China

Agricultural University.

Conflicts of interest: The authors declare no conflict of interest.

Ethics approval: The UK Biobank received ethical approval from the North West Multi-Centre Research Ethics Committee (16/NW/0274). Participants provided written informed consent before enrolment in the study, which was conducted in accord with the principles of the Declaration of Helsinki.

Consent to participate: Not applicable.

Consent for publication: Not applicable.

Availability of data and material: The data that support the findings of this study are available from UK Biobank but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the corresponding authors upon reasonable request and with permission from UK Biobank.

Code availability: Codes are available from the corresponding authors upon reasonable request.

Authors’ contributions: All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Chen Zhu. The first draft of the manuscript was written by Chen Zhu and Qihui Chen, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Acknowledgements: This research study was conducted using the UK Biobank database. The authors would like to thank the participants of the UK Biobank study.

Abdellaoui, A., Hugh-Jones, D., Yengo, L., Kemper, K. E., Nivard, M. G., Veul, L., ... & Visscher, P. M. (2019). Genetic correlates of social stratification in Great Britain. Nature Human Behaviour, 3(12), 1332-1342.
Adhikari, B. (2020). Does a Value‐Added Tax Increase Economic Efficiency?. Economic Inquiry, 58(1), 496-517.
Ashraf, Q. H., & Galor, O. (2018). The macrogenoeconomics of comparative development. Journal of Economic Literature, 56(3), 1119-55.
Bell, A., Chetty, R., Jaravel, X., Petkova, N., & Van Reenen, J. (2019). Who becomes an inventor in America? The importance of exposure to innovation. The Quarterly Journal of Economics, 134(2), 647-713.
Boon Ng, S. (2019). Exploring STEM competences for the 21st century. UNESCO International Bureau of Education. https://unesdoc.unesco.org/ark:/48223/pf0000368485.locale=en (Accessed 19 August 2022).
Bulik-Sullivan, B., Finucane, H. K., Anttila, V., Gusev, A., Day, F. R., Loh, P. R., ... & Neale, B. M. (2015). An atlas of genetic correlations across human diseases and traits. Nature Genetics, 47(11), 1236-1241.
Bycroft, C., Freeman, C., Petkova, D., Band, G., Elliott, L. T., Sharp, K., ... & Marchini, J. (2018). The UK Biobank resource with deep phenotyping and genomic data. Nature, 562(7726), 203-209.
Cheesman, R., Coleman, J., Rayner, C., Purves, K. L., Morneau-Vaillancourt, G., Glanville, K., ... & Eley, T. C. (2020). Familial influences on neuroticism and education in the UK Biobank. Behavior genetics, 50(2), 84-93.
Deming, D. J., & Noray, K. L. (2018). STEM careers and the changing skill requirements of work (No. w25065). National Bureau of Economic Research.
Gonalons-Pons, P., & Schwartz, C. R. (2017). Trends in economic homogamy: Changes in assortative mating or the division of labor in marriage?. Demography, 54(3), 985-1005.
Linnér, R., Biroli, P., Kong, E., Meddens, S. F. W., Wedow, R., Fontana, M. A., ... & Beauchamp, J. P. (2019). Genome-wide association analyses of risk tolerance and risky behaviors in over 1 million individuals identify hundreds of loci and shared genetic influences. Nature Genetics, 51(2), 245-257.
Master, A., Meltzoff, A. N., & Cheryan, S. (2021). Gender stereotypes about interests start early and cause gender disparities in computer science and engineering. Proceedings of the National Academy of Sciences, 118(48), e2100030118.
Mbatchou, J., Barnard, L., Backman, J., Marcketta, A., Kosmicki, J. A., Ziyatdinov, A., ... & Marchini, J. (2021). Computationally efficient whole-genome regression for quantitative and binary traits. Nature Genetics, 53(7), 1097-1103.
Modalsli, J. (2017). Intergenerational mobility in Norway, 1865–2011. The Scandinavian Journal of Economics, 119(1), 34-71.
Nicolaou, N., & Shane, S. (2010). Entrepreneurship and occupational choice: Genetic and environmental influences. Journal of Economic Behavior & Organization, 76(1), 3-14.
Pruim, R. J., Welch, R. P., Sanna, S., Teslovich, T. M., Chines, P. S., Gliedt, T. P., ... & Willer, C. J. (2010). LocusZoom: regional visualization of genome-wide association scan results. Bioinformatics, 26(18), 2336-2337.
Radu, S. (2018). STEM Worker Shortage at a Crisis, Survey Shows. US News. https://www.usnews.com/news/best-countries/articles/2018-08-23/americans-think-they-have-a-shortage-of-stem-workers (Accessed 19 August 2022).
Roeling, M. P., Willemsen, G., & Boomsma, D. I. (2017). Heritability of working in a creative profession. Behavior Genetics, 47(3), 298-304.
Song, Z., Li, W. D., Jin, X., Ying, J., Zhang, X., Song, Y., ... & Fan, Q. (2022). Genetics, leadership position, and well-being: An investigation with a large-scale GWAS. Proceedings of the National Academy of Sciences, 119(12), e2114271119.
Uffelmann, E., Huang, Q.Q., Munung, N.S., de Vries, J., Okada, Y., Alicia R. Martin,A.R., Martin, H.C., Lappalainen, Posthuma, T.D. 2021. NATURE REVIEWS 1, Article ID 59.
Van Tuijl, C., & van der Molen, J. H. W. (2016). Study choice and career development in STEM fields: an overview and integration of the research. International Journal of Technology and Design Education, 26(2), 159-183.
Watanabe, K., Stringer, S., Frei, O., Umićević Mirkov, M., de Leeuw, C., Polderman, T. J., ... & Posthuma, D. (2019). A global overview of pleiotropy and genetic architecture in complex traits. Nature Genetics, 51(9), 1339-1348.
Xue, Y., & Larson, R. C. (2015). STEM crisis or STEM surplus? Yes and yes. Monthly Labor Review, 2015.
Yengo, L., Robinson, M. R., Keller, M. C., Kemper, K. E., Yang, Y., Trzaskowski, M., ... & Visscher, P. M. (2018). Imprint of assortative mating on the human genome. Nature Human Behaviour, 2(12), 948-954.

No competing interests reported.

220907SupplementaryTables.docx

Genetic basis of STEM occupational choice and regional economic performance: A UK Biobank Genome-wide Association Study

Status:

Version 1

Abstract

Figures

1. Introduction

2. Methods

3. Results

Conclusion and Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1