Women entering the workforce en masse is one of the most important developments in the U.S. labor market during the past century. While female labor force participation increased following World War II, the gender wage gap held steady at 40 percent. The 1980s, however, marked a time of steady gains for women: the gender wage gap closed by roughly 10 percentage points. Then in the 1990s and beyond, gender wage convergence again stalled. The ratio of women’s to men’s earnings increased by a mere 2 percentage points in the 20 years between 1990 and 2010. While the reasons for gender wage convergence during the 1980s—including declining unionization, a reduction in gender discrimination, and reduced gender gaps in education, labor market experience, and occupational sorting—are well-understood, the pattern of stagnant wage gains for women in the 1990s is less understood. Our work shows that the introduction of state and federal family-leave policies can explain why gender wage convergence in the United States stalled.

In 1993, President Bill Clinton signed the Family and Medical Leave Act (FMLA), which guarantees 12 weeks of unpaid, job-protected leave to qualified workers for covered family or medical circumstances. Despite the fact that the FMLA does not require firms to pay workers during their leave, it may impose costs on employers whose workers take leave. For 58 percent of employees on FMLA leave, workloads are shifted to another employee, while 6 percent of employees on FMLA leave are replaced by temporary workers. Although the FMLA is a gender-neutral policy, women are more likely to file an FMLA claim; moreover, for those taking FMLA leave, the duration of leave spells are on average 14 business days longer for women than for men. Given the cost of employee leave-taking to firms and the differential use of the FMLA by women, it is plausible that the introduction of family-leave policies could have had differential impacts on wages by gender.

Research demonstrating the unintended effects of family-leave policies and the fact that gender wage convergence stagnated around the same time as the passage of the FMLA makes it conceivable that the FMLA could explain gender wage stagnation during the 1990s. Pinpointing the introduction of family-leave policies as the cause of gender wage stagnation is challenging, however, because there are contemporaneous changes in other federal policies that could also affect the gender wage gap. The earned-income tax credit, for example, was reformed in 1993, the same year the FMLA passed. Welfare reform at the federal level, which created the Temporary Assistance for Needy Families program, passed in 1996, three years after the FMLA passed. To overcome these challenges, we exploited the fact that during the years 1972–1992, 12 states and the District of Columbia enacted antecedents to the FMLA that offered unpaid maternity leave.

We studied the subset of states (including Washington, DC) that passed family-leave policies prior to the FMLA. We leveraged state variation in the timing of family-leave policies that occurred prior to the FMLA. Using wage and demographic data, we show that in the decade before a state family-leave policy passed, the gender wage gap between white women and white men closed at a rate of 0.7 percentage points per year. In the decade after a state family-leave policy passed, the rate of gender wage convergence fell to 0.17 percentage points per year.

Next, we expanded our approach to incorporate all 50 states and Washington, DC, by analyzing the effect of state-specific family-leave policies and the FMLA in the 38 states without a previous family-leave policy. We accounted for state-specific waivers to the welfare system and the federal welfare reform in 1997. When we analyzed all 50 states, we discovered similar estimates to those obtained using only the state-specific policies passed prior to the FMLA. In the decade before a family-leave policy passed, the gender wage gap between white women and white men closed at a rate of 0.7 percentage points per year. In the decade after a family-leave policy passed, the rate of gender wage convergence fell to 0.03 percentage points per year. We found that the introduction of family-leave policies can explain 94 percent of the reduction in the rate of gender wage convergence that is unaccounted for after accounting for changes in observable characteristics of workers.

To discover why family-leave policies affect genders differently, we used data from the Department of Labor for the years 1995, 2000, 2012, and 2018 to identify whether white women or white men take leave more frequently and which gender for longer. We found that white women are more likely to take leave than white men. When we investigated the length of leave, we found that white women take leave spells for the birth or adoption of a child that are four times longer on average than those taken by white men, despite taking leaves of similar lengths to white men for non-child-related reasons. The data suggest that gender differences in leave-taking after childbirth and adoption explain why family-leave policies reduce the rate of gender wage convergence.

Having shown that family-leave policies caused the gender wage gap stagnation, we examined the effects of family-leave policies on annual earnings. We evaluated the differences between actual earnings for workers who work 40 hours per week for 50 weeks and predicted earnings without family-leave policies. The year after a family-leave policy passed, white women’s earnings increased by $488, and white men’s earnings increased by $303. However, in the decade after family-leave policies passed, white women’s actual earnings were not different from what their earnings would have been without a leave policy. On the other hand, the earnings for white men increased consistently. One decade after the policy, we found that the earnings of white men increased by $2,917 and that the earnings of white women were unaffected. Further, we calculated that in the absence of family-leave policies, gender wage parity for white women would have occurred in 2017.

NOTE
This research brief is based on Peter Q. Blair and Benjamin Posmanick, “Why Did Gender Wage Convergence in the United States Stall?,” National Bureau of Economic Research Working Paper no. 30821, January 2023.