Why education expanded but opportunity remained unequal

By Dr. Priyom Bose, Ph.D.Reviewed by Lauren HardakerJan 30 2026

As education expanded and gender gaps closed, new genetic evidence shows that family advantage continues to determine who gets ahead in school.

Study: Genetic associations with education have increased and are patterned by socioeconomic context: Evidence from 3 studies born 1946–1970. Image credit: Iryna Inshyna/Shutterstock.com

A recent study in Proceedings of the National Academy of Sciences investigated whether predictors of educational attainment (EA) have changed across time.

Limitations in strategies to study factors influencing educational attainment

An individual’s level of education is strongly associated with health and wealth outcomes. Multiple studies have shown that EA is influenced by both personal characteristics and other factors, such as DNA, gender, and social background. Assessing how these predictors change over time could help measure patterns related to equality of opportunity. While gender gaps in education have narrowed or reversed in many countries, socioeconomic gaps remain substantial.

Traditional survey-based studies of educational change suffer from measurement error, reverse causation, and generational variability. In contrast, genetically informed studies using polygenic indexes (PGIs) offer multiple advantages. For instance, genetic variation evolves slowly over generations, so analyzing cross-cohort changes in genetic predictors may reflect societal rather than biological shifts. DNA’s constancy throughout life rules out reverse causation, and genetic data’s precision reduces measurement error.

In theory, a larger genetic role and declining environmental barriers could suggest increased equality of opportunity. A stronger association between PGI and education among women born in later cohorts was consistent with reduced gender barriers in the 20th century. However, evidence on socioeconomic background is mixed, with genetics playing a bigger role in education in countries where people can move up the social ladder.

Gene-by-environment interactions may produce different returns to genetic predisposition across social classes. Some studies find stronger genetic effects among advantaged groups, others among disadvantaged groups, and some find no pattern. The differential findings could be attributed to the use of diverse data sources with varying representativeness.

Study design and data sources

The current study focuses on how PGI-education associations have changed over time and vary by gender and socioeconomic background, using three UK birth cohorts: the 1946 National Survey of Health and Development, the 1958 National Child Development Study, and the 1970 British Cohort Study.

Each study cohort was analyzed separately and combined to estimate cohort-specific and stable patterns, using multiple imputation and inverse probability weighting to account for missing data and selection bias. This study improves cross-cohort comparability by harmonizing PGIs using statistical methods to address missing data and selection bias. Both education and cognition PGIs were examined to distinguish education-specific from cognition-related genetic influences over time.

This study assessed the effects of sex, parental occupational social class at birth, and parental education as effect modifiers. Full genotyping details were available on the Center of Longitudinal Studies (CLS) Genomics Data GitHub and Resource Profile. Blood samples were collected with informed consent at ages 53 (1946c), 44 (1958c), and 46 (1970c). PGIs were generated for education and cognition.

Social advantage augments genetic effects on education

Analyzing the three cohorts, researchers observed that the average years of education rose from 16.51 in 1946c to 17.18 in 1958c and 17.90 in 1970c, driven by increasing university degree attainment, 8.85 %, 20.91 %, and 31.32 %, respectively. Parental education and occupational class also improved across cohorts, reflecting broader societal changes.

Genetic associations with education strengthened over time, with education PGI associations increasing from 0.44 years in 1946c to 0.67 in 1970c, explaining 3.5 % to 5.1 % of the variance. Cognition PGI associations remained stable at 0.23,0.27 years, explaining less than 1.5 % of variance.

EA PGI associations increased from 0.44 years in 1946c to 0.67 years in 1970c, with significant cohort differences. Cognition PGI associations remained stable across cohorts at 0.23,0.27 years with overlapping confidence intervals and no cohort interactions.

Genetic predisposition explained 3.5 % to 5.1 % of educational differences across cohorts, while socioeconomic background explained 10,16 %, which is approximately two to three times more. Cognition PGI remained low, ranging from 0.7 % to 1.4 %.

Over the years, gender gaps have narrowed and reversed. Women shifted from having half a year less education than men in 1946 to slightly more by 1970, driven by sharper increases in university degrees for women, from 6 % to 34 %, versus 11 % to 29 % for men.

Children from advantaged families had higher genetic scores for both education and cognition. Social class gaps remained large across all cohorts. Those from the highest social class completed 3.5,4 more years of education than those from the lowest class, while those with degree-educated parents completed about 3 more years than those whose parents had no qualifications.

Social background strongly shaped how much genetics mattered. People with the highest genetic scores but basic-educated parents achieved the same education level as those with the lowest genetic scores but degree-educated parents. In other words, having advantaged parents compensated for low genetic predisposition, while disadvantaged backgrounds often prevented high genetic predisposition from translating into educational success. This pattern was weaker for cognitive ability.

Conclusion

Genetic associations with education strengthened across British cohorts born between 1946 and 1970, yet social background remained around twice as predictive. Crucially, genetic predisposition translated into educational gains primarily for those from advantaged backgrounds. Despite educational expansion and narrowing gender gaps, social barriers continue to limit equality of opportunity, a pattern the findings are consistent with rather than directly testing, highlighting the need for policies addressing structural inequalities.

Download your PDF copy now!