The causal effects of education on age at marriage and marital fertility

The negative association of education and fertility, over time and between countries, is a central stylized fact of social science. Yet we have scant evidence on whether this is, or is not, causal. Using the universe of vital registration index data from England, 1912 to 2007, I first show that it is possible, using unique names, to construct a demographically and socioeconomically representative sample of 1.5 million women. Historical record linkage of women is typically not attempted but is possible here because of the unique characteristics of English civil registration. I then exploit the natural experiment of sharp discontinuities in who was affected by compulsory schooling law changes in 1947 and in 1972, which exogenously and effectively raised the minimum school leaving age. A Regression Discontinuity design, executed on the individual data, identifies the causal effect of education on age at marriage and fertility. Education may have raised age at marriage in 1972. However one extra year of education at 15 or 16 has a zero causal effect on marital fertility.