The Impact of Job Protection on Paid Leave Use and Post-Leave Employment: Evidence from Administrative Data

• In Event: Paid Leave Across the States: Access, Utilization, and Employment Outcomes

Abstract

Thirteen states and the District of Columbia have passed paid family leave programs (Donovan 2023). A growing literature finds that paid leave supports the economic stability and health of workers (Bartel et al. 2023). However, no studies evaluate whether job protections affect the use of paid leave or post-leave employment. Using WA Paid Leave administrative data for 2019-2025 and a regression discontinuity design, we will evaluate the impact of job protections on paid leave use and post-leave employment. We will also assess heterogeneity in the effect of job protection by worker and employment characteristics.

Job protection is a colloquial term for the legal right to reinstatement, a guarantee that an employee can come back to the same or similar job after they take leave. The federal Family and Medical Leave Act (FMLA) provides about half of American workers with 12 weeks of unpaid, job-protected leave (Rossin-Slater and Uniat 2019). State paid leave programs vary in their offering of job protection.

Our identification strategy will leverage a 1,250 work-hour job protection eligibility threshold in WA Paid Leave using sharp regression discontinuity designs (RDD). Until 2026, WA workers with 820 work hours in four of the prior five calendar quarters were eligible for paid leave wage replacement, but workers with at least 1,250 hours for a single employer were eligible for job-protected leave. We will compare worker outcomes close to the 1,250 hour threshold. Outcomes for our analysis will include paid leave take up (binary), leave spell duration (weeks), leave use intensity (avg hours per week), post-leave employment (binary), return to pre-leave employer (binary), post-leave work hours (hours per quarter), and quarterly post-leave earnings.

In a descriptive analysis of job protection in WA Paid Leve, we found that 53% percent of paid leave-eligible Washington workers would be job protected if they took leave. In addition, job protection was associated with a 64% (0.69 percentage point) increase in the likelihood of taking leave. Among leave users, job protection was associated with a 5% (3.3 percentage point) increase in the likelihood of being employed following leave. Our RDD analysis will build on our descriptive results to estimate the causal impact of job protection.

Theoretically, employment protections should reduce the real and perceived risk to employees of taking leave. Reductions in the perceived risk of leave should increase leave use. Reductions in the actual risk of leave should increase workers’ post-leave attachment to employers. However, job protections are not proactively enforced by the Department of Labor (for the FMLA) or state agencies (for paid leave programs), with one study estimating FMLA firm compliance at 54% to 77% (Wing, Armenia, and Gerstel 2014). Research is needed to understand whether existing job protections empower workers to use leave and support post-leave attachment to employers. Our project will be the first to study job protections in the context of paid leave, providing causal evidence related to the design of paid leave programs and employment law in the United States.

Author

• Tom Lindman, Harvard University
  Presenting Author