There were many social changes between 1970 and 1980 that could have affected the gains to marriage over the decade. A major change was the national legalization of abortion in 1973. Legal abortions were partially available in some states by 1970. If the partial legalization of abortions in a state reduced the gains to marriage in that state, we would expect to see lower gains to marriage in the early legalizing states relative to later legalizing states in 1970 but not in 1980. Moreover, this difference in difference in the gains to marriage should be concentrated among women of childbearing age. Using marriage rate regressions, Angrist and Evans (1999) showed that the marriage rates of young men and women were lower in early legalizing states relative to later legalizing states in the early 1970s. We show that the stimates of the number of marriages affected are sensitive to whether we use male or female marriage rate regressions. We extend the benchmark model to include whether an individual resided in a state that allowed legal abortions or not as part of the definition of the type of an individual. Methodologically, we extend the standard difference in differences estimator to estimate the effect of a policy change on bivariate distributions. Estimating this extended model, we show that the partial legalization of abortion in some states can explain up to 20 percent of the drop in the gains to marriage among young adults in the 1970s.
That is from "Who Marries Whom and Why," by Eugene Choo and Aloysius Chow, in the January 2006 Journal of Political Economy. Here is an earlier version of the paper.