Does Children's Education Improve Parental Health and Longevity? Causal Evidence from Great Britain

Abstract

Parents with better-educated children are healthier and live longer, but whether there is a causal effect of children's education on their parents' health and longevity is unclear. First, we demonstrate an association between adults' offspring education and parental mortality in the 1958 British birth cohort study, which remains substantial-about two additional years of life-even when comparing parents with similar socioeconomic status. Second, we use the 1972 educational reform in England and Wales, which increased the minimum school leaving age from 15 to 16 years, to identify the presence of a causal effect of children's education on parental health and longevity using census-linked data from the Office for National Statistics Longitudinal Study. Results reveal that children's education has no causal effects on a wide range of parental mortality and health outcomes. We interpret these findings discussing the role of universal health care and education for socioeconomic inequality in Great Britain.

Keywords: causal inference; education; health; intergenerational relationships; mortality.

Figure 1.

The 1972 Educational Reform Greatly Reduced the Share of Individuals Leaving School at Age 15 for Those Born after September 1st, 1957. Source: Health Surveys for England, 1991–2013 (Mindell et al. 2012), own calculations. Note: Years of full-time education by year of birth.

Figure 2.

Better-Educated Children Have Parents Who Live Longer, Even When Comparing Parents from the Same Social Class and with the Same Educational Attainment. Source: National Child Development Study (Power and Elliott 2005), own calculations. Note: Hazard ratios are based on the models reported in Table A2 in the online version of the article. Spikes denote 95% confidence intervals.

Figure 3.

The Proportion of Deceased Parents Decreased Approximately Linearly over a Period Spanning Two Decades. Source: Office for National Statistics Longitudinal Study (Shelton et al. 2019), own calculation.

Figure 4.

Regression Discontinuity Estimates of Educational Reform Stratified by Parental Sex. Source: Office for National Statistics Longitudinal Study (Shelton et al. 2019), own calculations. Note: Error bars denote 95% confidence interval based on robust standard errors. Controls comprise parental education, number of children, occupational class, and year of birth. Estimates shown in Table A3, available in the online version of the article.