In the study sample, higher genetic scores associated with emotional stability were linked to higher earnings and household wealth, while a score associated with conscientiousness was also linked to greater wealth. A one-standard-deviation increase in the emotional-stability score was associated with €1,420 more in annual earnings, or 1.38 percentile points in the earnings ranking. The same increase was associated with €6,322 more in household net wealth, or 0.72 percentile points. For conscientiousness, the wealth association was €4,344, or 0.53 percentile points.
The study reports associations, not a settled causal explanation. Its family-based design was meant to address confounding, but residual pleiotropy cannot be ruled out, meaning genetic variants may be related to earnings or wealth through pathways other than the personality traits being studied.
A family-based genetic comparison
The researchers combined genetic data from the Lifelines Cohort Study with Statistics Netherlands tax records on individual earnings and household net wealth. The records span 2006 through 2024. The analysis was limited to people of European genetic ancestry aged 30 to 65 and included approximately 39,000 individuals. The main earnings analyses used 574,690 person-year observations from 38,936 individuals, while the wealth analyses used 566,810 observations from 38,934 individuals.
Rather than comparing unrelated people alone, the study used the random transmission of parental genetic copies and conditioned on the parents' corresponding polygenic indexes. That approach isolates differences in offspring scores within the family-based design. When both biological parents had not been genotyped, the researchers used Mendelian imputation to recover parental genotype sums.
Big Five polygenic indexes, or PGIs, are genetic scores for openness, conscientiousness, extraversion, agreeableness and emotional stability. They were built with a PUMAS-ensemble method using genome-wide association study summary statistics that excluded Lifelines participants. The regression included all five scores and their corresponding parental scores, plus gender, birth-year and calendar-year fixed effects. Standard errors were clustered at the individual level, with heteroskedasticity-robust errors for time-invariant outcomes.
What the scores captured
As a validation check, the researchers compared the scores with measured personality. In 37,768 participants, a one-standard-deviation increase in the conscientiousness, extraversion or emotional-stability PGI corresponded to 0.11, 0.09 or 0.14 standard deviations in the matching survey measure. All three associations had P values below 0.001.
For earnings, a one-standard-deviation higher emotional-stability PGI was associated with €1,420 higher annual earnings, equal to 3.3% of the sample's €43,385 mean. In a version that ranked people rather than using euro amounts, the association was 1.38 percentile points, with a standard error of 0.18 and P < 0.001.
For household wealth, the emotional-stability association was €6,322 per one-standard-deviation increase, or 3.2% of the sample's €195,543 mean. The rank estimate was 0.72 percentile points, with a standard error of 0.20 and P < 0.001; the euro estimate had a standard error of €2,219 and P = 0.004.
The conscientiousness score also showed a wealth link: €4,344 per one-standard-deviation increase, with a rank estimate of 0.53 percentile points. The standard errors were €2,190 and 0.20, and the P values were 0.047 and 0.007. Openness, extraversion and agreeableness did not show a consistent pattern across outcomes and model specifications.
The links changed with other pathway controls
Adding a genetic score for educational attainment weakened the emotional-stability earnings association but did not remove it. The euro estimate fell from €1,420 to €957, while the rank estimate fell from 1.38 to 0.98 percentile points. Both adjusted estimates remained positive, with P < 0.001.
The researchers then added genetic controls for educational attainment, cognitive performance and major depressive disorder. The emotional-stability earnings association remained 0.66 percentile points, with a standard error of 0.20 and P < 0.001. The conscientiousness wealth association was 0.44 points, with a standard error of 0.19 and P = 0.025. The emotional-stability wealth estimate was 0.53 points after educational-attainment and cognitive-performance controls, but fell to 0.02 points after the major-depressive-disorder control was added, with a standard error of 0.21 and P = 0.92.
A separate analysis added measured personality traits one at a time. Measured emotional stability reduced the emotional-stability PGI associations with earnings and wealth, while measured conscientiousness reduced the conscientiousness-PGI association with wealth. The authors explicitly describe this as descriptive and non-causal, not a formal mediation estimate.
Why the result remains provisional
Even with the family-based comparison, residual pleiotropy cannot be ruled out. Controls for education, cognitive performance or depression may miss some alternative pathways or remove effects that run through personality. Adjusted estimates may therefore understate total associations.
The sample also limits how far the findings can be carried. It consists of European-ancestry participants in the Netherlands, so the results may not apply to other ancestries, countries or environments. The findings therefore describe associations in this setting, not a guarantee that the same pattern holds elsewhere.
Paper data and sources
Original title: Random Genetic Variation Links Personality to Earnings and Wealth
Authors: Leandro Carvalho, Sjoerd van Alten, Silvia Barcellos et al.
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-10-05
DOI: Not available
Original paper · Full text