While many marry for love, marriage is also a legal and economic institution. Public insurance programs for individuals decrease the relative insurance value of marriage. We explore how this tradeoff affects the marriage market in the context of the Affordable Care Act’s Medicaid expansion in the United States. We show both theoretically and empirically that Medicaid expansion does decrease the marriage rate, even for individuals with high educational attainment, and it also reduces the divorce rate of new marriages, consistent with an increase in match quality. Counter to intuition, there is little effect on divorce rates for already-married couples. Our findings illustrate that even when public insurance reduces the monetary benefits of marriage, the effects on marriage overall may be positive.
Revisiting the breadwinner norm: The effect of the potential relative wage on married women’s labor supply. // [MANUSCRIPT]
In light of recent research suggestive of a male breadwinner norm, I examine the relationship between female breadwinning and women’s labor force participation for opposite-sex married couples in the U.S. I show that using measures based on observed income likely yields bias due to non-classical measurement error and omitted variables. Estimating the relationship with a measure for potential relative earnings unaffected by this bias gives a precise null relationship. These findings suggest aversion to female breadwinning does not impact decisions within marriage, but the results do not rule out possible effects earlier at the margin of marital sorting.
Subsidizing contraception: Effects on take-up of highly effective contraceptives and unintended births. // [Work-in-Progress]
Since the 1990s, Medicaid programs including family planning waiver programs and the recent ACA Medicaid expansion have expanded coverage of prescription contraceptives to millions of women. These expansions may reduce unintended pregnancies if women take up more effective contraceptives. Using data from the State Drug Utilization Program between 1992 to 2019 and several difference-in-difference estimators, I show that these coverage expansions at best increased take-up of highly effective contraceptives by about 40 per 1,000 women of childbearing age per year that would have been eligible for the expansions, although the true effect is likely smaller. Using both U.S. Vital Statistics data and the American Community Survey for births, I find no evidence that either type of expansion led to a significant decrease in fertility, including among subgroups that typically experience higher rates of unintended births. Among the eligible population, confidence intervals suggest the programs could have caused a decrease of up to about 5%, or roughly 4 births per 1000. These results suggest that further expansion of subsidies in states with more limited coverage are unlikely to have substantial impacts on unintended births.