Family leave policies are widely championed as a means to close the gender gap in wage earnings (Council of Economic Advisers 2016). California?s first-in-the-nation paid family leave statute has been in place for over a decade, but data limitations have prevented strong conclusions about the policy?s effects on the careers of mothers and fathers. This project will address this question using novel, large-scale administrative IRS-tax data to understand how California?s paid leave policy affected mothers? and fathers? long-term labor-force attachment, job and employer transitions, and wage growth.
Intellectual Merit
The proposed project will use a differences-in-differences research design to provide the richest and most dynamic intention-to-treat analysis of U.S. paid leave on careers to date, exploring both the costs and the benefits of California?s paid leave policy for millions of families. This project will contribute to the scientific and policy literature by using large-scale administrative IRS-tax data with samples at least three orders of magnitude larger than in other labor-force surveys (e.g., Current Population Surveys (CPS) and American Community Surveys (ACS)). These large sample sizes allow us to refine many statistical zeros in the literature to compute much more precise estimates of the effects of these policies on labor-force outcomes. They also allow more detailed distributional analyses, including heterogeneity analyses for women and men by baseline family income (including poverty status), marital status, and age of mother at first birth.
A second contribution is that the IRS-tax data provide a 10-year, individual panel of outcomes, allowing a ten year perspective on the outcomes of affected individuals. Because other labor-market surveys are repeated cross-sections (CPS, ACS), short panels (SIPP), or have very small sample sizes (PSID, NLSY), the IRS-tax data are novel in being able to describe the career dynamics of mothers and fathers gaining access to paid through California?s policy change. For instance, we can describe how short-term losses in wages may be offset by longer-term gains over ten years following childbirth.
A third contribution is that we will provide new evidence regarding how paid leave policies affect family leave-taking?including husband?s labor supply, compensation, and long-term childcare specialization decisions. Modelling family decisions is possible because tax data provide information on husband?s outcomes before he marries and becomes a father as well as for the 10 years after he first becomes a father, including periods where he may no longer be married to his wife and before they married. This data feature allows us for the first time to quantify the effects of paid leave policies on long-term marital stability and long-term within-household specialization in work and home production.
Broader Impacts
The proposed research will have broader impacts on public policy, education of economics students and research by other social scientists. First, the U.S. lags behind all other advanced countries in providing basic work-place accommodations for parenthood and paid leave (Council of Economic Advisors 2014). This project will provide estimates of the long-term impact of the first state-level paid family leave statute?as well as heterogeneity by subgroups?that could be used to inform cost-benefit analyses of proposed policies. We will disseminate our results in a peer-reviewed paper targeted to social scientists, which we will present at professional conferences and circulate through working paper series. Second this project will have two educational components. Undergraduate students in economics, one at the University of Michigan and one at Middlebury College, will contribute to the project by conducting complementary data analyses of the American Community Surveys (ACS) and Survey of Income and Program Participation (SIPP) as part of thesis projects. In addition, this
