New Jersey enacted an expansion of its paid family leave program in July 2020, increasing the maximum duration of leave from 6 to 12 weeks and the wage replacement rate from 66% to 85% of weekly earnings. We examined changes in maternal mental health among new mothers in New Jersey before and after the policy took effect.
We used data from New Jersey birth certificates, which include a measure of depression and anxiety based on the Edinburgh Postnatal Depression Scale (EPDS), administered after delivery and prior to hospital discharge. The EPDS ranges from 0 to 30, with higher scores indicating greater risk of depression. A key strength of this approach is the use of population-level data capturing maternal mental health across the state. Our sample included women aged 18–49 with no prior live births; we excluded births resulting in in-hospital death or with severe congenital anomalies.
We used an interrupted time series approach to compare women who were employed in New Jersey (and thus eligible for paid leave, n=101,253) to those employed in Pennsylvania (and thus not eligible, n=4,378). Data were aggregated to the quarter-of-birth level, and we applied Newey–West standard errors with a lag of four to account for autocorrelation.
Figure 1 presents quarterly trends in EPDS scores across three periods: before the initial policy changes (pre-2019), after initial implementation but prior to the expansion in benefit generosity (January 2019–June 2020), and after full implementation (July 2020 onward). Observed mean scores and fitted linear trend lines are shown for each period. Trends in average EPDS scores were relatively flat in both groups prior to 2019. Following initial implementation, EPDS scores increased more sharply among women employed in Pennsylvania compared to those employed in New Jersey, with this divergence persisting into the post-period. The estimated difference between the post-period (after full implementation) and the pre-period is -0.035 (95% CI: -0.058 to -0.012; p-value=0.003), indicating a statistically significant reduction (i.e., improvement) in EPDS scores.
Figure 1. Average EPDS score for NJ births over time, by state of employment.
Trends in mental health by state of employment among NJ births
We next adjusted for a range of potential confounders aggregated at the quarter-year level, including race, ethnicity, maternal age, insurance status, NICU admission, route of delivery, smoking during pregnancy, Apgar score, low birth weight, and preterm birth, and included quarter-of-birth fixed effects. After adjustment, the estimated effect was larger in magnitude (−0.077; 95% CI: −0.148 to −0.005; p=0.036).
Overall, this analysis provides some evidence of improved maternal mental health as a result of implementation of the expansion of paid family leave in NJ. However, further analysis is needed to strengthen causal inference. An additional limitation is that the policy’s implementation closely coincided with the COVID-19 pandemic; longer-term data will be important for assessing whether these patterns persist beyond the pandemic period.