| This paper contributes to the growing literature analyzing the non-market benefits of education from improved longevity along four veins of research. First, it refines previous estimates of the marginal effect of schooling on mortality using state compulsory schooling laws as instruments and data from the 1983 U.S. National Longitudinal Mortality Study. Second, it translates the estimated marginal effects into life expectancy improvements, monetized survivorship gains and longevity-adjusted income returns by means of a lifetime expected utility maximization model. The distribution of these benefits over the life cycle assigns greater weight to longevity improvements at younger ages since they carry over to future periods and are not yet heavily discounted. As survivorship gains from schooling decline with age in percentage terms, approximately half of monetized longevity gains accrue between the ages of 45 and 64. While whites collect at least a quarter of these benefits after age 65, blacks reap no less than three quarters before retirement. On the whole, an additional year of schooling improves life expectancy at age 25 between 0.15 and 0.32 years and, under a 3% discount rate, raises the magnitude of income returns between 5% and 20%. With the exception of white males, the resulting longevity rates of return decrease with schooling relative to income returns. Third, longevity gains to schooling spill over within the household to children and spouses. Longevity gains accruing to children are modest given their low mortality risk. On the other hand, the longevity benefits that husbands receive from their wives' education are substantially greater than the gains wives receive from their husband's education, even after netting out the benefits from positive assortative mating. Fourth, data from the Integrated Pubic Use Microdata Series reveals substantial longevity gains from schooling through nearly all the second half of the 20th century. Educational disparities in longevity decline among whites from 1940 to 1970, and rise thereafter. Likewise, longevity returns for blacks are sizeable in 1940 but decrease until 1980. Overall, results are sensibly consistent and robust across datasets notwithstanding the different mortality and income measures employed. |
