Adverse childhood experiences (ACEs) are potentially traumatic childhood events associated with significant negative health outcomes throughout the lifespan (Anda et al., 2006; Felitti et al., 1998; National Center for Injury Prevention and Control, 2019, 2020). ACEs include abuse and neglect, and household challenges. Household challenge ACEs focusing on household challenges include living with someone who has substance use or mental health issues, witnessing domestic violence, and experiencing parental/caregiver¹ separation or incarceration. Children who experience household challenges or other ACEs are at greater risk for emotional, cognitive, and behavioral problems. They are more likely to have poorer educational outcomes, psychiatric diagnoses, and criminal involvement (Haj-Yahia et al., 2019; Mueller & Tronick, 2019; Peleg-Oren & Teichman, 2006; Rasic et al., 2014; Wakefield & Wildeman, 2013; Weaver & Schofield, 2015; Yamamoto & Keogh, 2018). The prevention of ACEs is a public health priority with potentially far-reaching implications for children’s short- and long-term health outcomes. Policies that support reductions in the occurrence of any household challenge ACEs may play an important role in improving public health.

Paid Family Leave (PFL) is a state-level policy intended to aid health and economic outcomes in families with children (State of California, n.d).; State of Washington, n.d.). Presently, 20 states have enacted and implemented PFL and an additional four states (Minnesota, Maine, Delaware, and Maryland) will implement PFL by 2026 (Bipartisan Policy Center, n.d.). Existing PFL policies allow employees to take 4 to 12 weeks of partially paid leave to address personal health problems, care for an ailing family member, or, more commonly, care for a newborn or newly adopted child. In 2025, Rhode Island’s Temporary Caregiver Insurance provides up to six weeks of partial wage replacement (i.e., 60% of an eligible employee’s wage up to a maximum of $1070 per week) and job protection during this time off, while New York State Paid Family Leave offers up to 12 weeks of job-projected time off at 67% of an eligible employee’s average weekly wage up to a maximum of $1,177.32 per week.² Each state’s implementation of PFL varies and may change from year to year.

The ability to take leave following childbirth is particularly important. It allows new parents time to recover from childbirth, address health concerns for themselves and their children, and have time to bond with their children (Bullinger, 2019; D’Inverno et al., 2018). In turn, this increases protective factors such as positive parenting practices and reduces risk factors of violence such as poor mental and behavioral health and negative parenting practices. Additional pathways through which PFL may prevent or reduce intimate partner violence, teen dating violence, and child maltreatment include maintaining household income thereby reducing financial stress which could prompt violence, as well as increased egalitarian parenting and domestic practices that support a mother’s return to work and her career and income advancement (D’Inverno et al., 2018). This ultimately promotes equality, which has been associated with reductions in intimate partner violence and sexual violence (Kearns et al., 2020; Willie & Kershaw, 2019). Additionally, these positive outcomes may strengthen the relationship between parents, reducing the incidence of caregiver separation. By preventing a child from becoming a victim of violence or witnessing violence, PFL helps to break the intergenerational transmission of violence and to reduce the number of ACEs a child experiences.

Studies have found links between PFL and a range of improved caregiver and child health outcomes. Findings suggest PFL is associated with improved overall child health(Bullinger, 2019). This includes decreases in low birthweight and preterm birth rates, particularly among African American and unmarried mothers (Chen, 2021; Ruhm, 2000; Stearns, 2015; Tanaka, 2005). Furthermore, the introduction of PFL was associated with decreases in infant mortality and hospitalizations (Klevens et al., 2016; Montoya-Williams et al., 2020; Pihl & Basso, 2019) and increases in on-time infant vaccinations (Choudhury & Polachek, 2021), as well as time mothers spent breastfeeding (Baker & Milligan, 2010; Hamad et al., 2019; Huang & Yang, 2015; Pac et al., 2019) and engaging in childcare activities (Trajkovski, 2019). Since birth and neonatal outcomes are strongly correlated with longer-term child well-being (Ip et al., 2007; Moster et al., 2008; Oreopoulos et al., 2008; Stein et al., 2006), policies such as PFL may have long-lasting benefits for children including educational attainment levels, income, and mortality and medical disability rates (Litchman-Sadot & Bell, 2017).

PFL is also associated with benefits for caregivers, including reduced alcohol use in fathers, and decreased psychological distress, particularly for single and younger mothers (Doran et al., 2020; Irish et al., 2021; B. C. Lee et al., 2020). These early findings on caregiver well-being are important, given the evidence that high caregiver stress can predict hostile, less responsive parenting, and poorer child well-being (Lee et al., 2018; Rominov et al., 2016; Ward & Lee, 2020). In addition, caregiver well-being has been linked to children’s experiences of ACEs and maltreatment (O’Donnell et al., 2015; Reupert & Maybery, 2016). Therefore, policies that support caregivers may contribute to preventing a wide range of negative health outcomes associated with early childhood adversity.

More research is needed to establish the role of PFL in improving caregiver outcomes and preventing ACEs in early childhood. In addition to being proximal to the time in which caregivers would utilize PFL, early childhood is also an important developmental period in which adverse, traumatic experiences can alter brain development and subsequent long-term health (Merrick et al., 2019; Teicher et al., 2016). This underscores the critical need to prevent early childhood ACEs. The bulk of existing evidence for PFL and health effects used data from California, which was the first state to enact the policy (Huang & Yang, 2015; Pihl & Basso, 2019; Stein et al., 2006; Trajkovski, 2019). Studies that investigate the impacts of PFL in other states can strengthen the evidence for the policy’s impact on health and improve the generalizability of inferences. This study adds to the existing literature by examining the effects of PFL on household challenge ACEs in New York and Rhode Island, two states that have implemented PFL. We employed causal inference methods to estimate PFL effectiveness in reducing household challenge ACEs for families with children. It is proposed that paid family leave can decrease family and caregiver stress and increase economic stability, ultimately decreasing risk for household challenge ACEs that focus on household challenges.

Table 1 provides a descriptive summary of measures used in the analysis, including outcome summaries from a pre- and a post-intervention year for the two intervention states and 46 control states and Washington, DC (N_{control states} = 47). The average proportions of children who experienced household challenge ACEs in intervention states were similar in magnitude to the average proportions of children experiencing household challenge ACEs in the control states before PFL was enacted (2011). About 12% of children experienced at least one household challenge ACE in both intervention and control states in 2011. In 2019 post-PFL, 5.5% of children experienced at least one household challenge ACE in the intervention states compared to 9.1% of children in the control states.

Figure 1 shows the effects of PFL on Any Household challenge ACE outcome in the BART adjusted model. The Y axis represents the difference between the observed outcome in NY and RI and the expected outcome in those states in absence of the intervention (based on the control states and the statistical model). The X axis is the time in years relative to the start of the intervention (with year 0 as the intervention start year). Note that the difference between the observed outcome and the outcome expected in absence of the intervention should be zero before the intervention starts and represents the estimated effect of the intervention afterwards. The figure suggests adequate model fit (i.e., near zero average total treatment effects (ATT) estimates that are not statistically significant in the six years’ pre-intervention period) and negatively trending ATT estimates post-intervention. Other outcomes and base model estimations followed similar trends in model fit and direction of the treatment effect (data not shown). One notable exception was the BART estimation of mental health, where the model fit was compromised by the significant ATT estimates in the pre-intervention period (data not shown).

Table 2 provides the ATT of PFL on household challenge ACEs across intervention states estimated using replication from BART and TWFE methods. Base models do not include state-level covariates. Results showed several significant effects. PFL implementation in intervention states, relative to controls, was associated with significant reductions in the proportion of young children experiencing any household challenge ACE (adjusted ATT = −0.04, 95%CI = −0.06; −0.01), caregiver incarceration (adjusted ATT = −0.02, 95%CI = −0.04, −0.01), and caregiver separation (adjusted ATT = −0.04, 95%CI = −0.06, −0.02). PFL impact on caregiver substance use was only statistically significant in the adjusted model (adjusted ATT = −0.01, 95%CI = −0.03; −0.00³). Caregiver mental health outcome results were suppressed as unreliable due to poor model fit.

Base TWFE models showed a significant negative effect of PFL on household caregiver substance use (ATT = −0.01, 95%CI = −0.02; −0.003), domestic violence (ATT = −0.01, 95%CI = −0.01; −0.003), caregiver separation (ATT = −0.03, 95%CI = −0.05; −0.02), and caregiver incarceration (ATT = −0.02, 95%CI = −0.02; −0.02). However, after the inclusion of state-level covariates, only household caregiver separation (adjusted ATT = −0.04, 95%CI = −0.06; −0.01) and caregiver incarceration (adjusted ATT = −0.02, 95%CI = −0.03; −0.02) ACEs estimates remained statistically significant.

As a robustness check, we conducted placebo analyses for adjusted models where we identified significant effects of PFL in which the two intervention states were replaced by two randomly selected control states. Placebo analyses showed no significant effects of PFL on household challenge ACEs (see Table 3), further validating the significant findings above.

Our study found that PFL implementation in the two intervention states of Rhode Island and New York was associated with the reduction in the proportion of young children who experienced certain household challenge ACEs. Statistically significant reductions in caregiver separation and caregiver incarceration were consistent across BART and TWFE methods, suggesting particularly robust treatment effects for these outcomes.

Our findings indicate that PFL policy implementation prevented on average 4 out of every 100 children from experiencing caregiver separation, and 2 out of 100 children from experiencing caregiver incarceration in the states each year post-intervention. In addition, BART estimations suggested a significant negative effect of PFL on the proportion of young children experiencing at least one household challenge ACE. Each year following PFL policy implementation, approximately 4 out of every 100 children were prevented from experiencing any household challenge ACE in the two states, a reduction of over 40% compared with the average rate that would have been observed had the policy not been implemented (i.e., in control states). These reductions would have large population implications with respect to preventing ACEs in early childhood if PFL was implemented in all states.

These findings carry potential implications for children’s health. Caregiver separation is linked to short- and long-term issues with anxiety, depression, self-esteem, school problems, suicide attempts, alcohol and drug use, and smoking (Auersperg et al., 2019; Cherlin et al., 1998; Størksen et al., 2005). Researchers found that the risk factors of separation, such as intercaregiver conflict and decline in socioeconomic level, drive some of these effects (Nunes-Costa et al., 2009). Similarly, caregiver incarceration is tied to a wide range of negative child health outcomes, including increased risk for other ACEs (Turney, 2018; Wildeman et al., 2018). The U.S. Department of Justice reported 22% of state inmates and 16% of federal inmates have children ages four years and younger (Glaze & Maruschak, 2010). Separated caregivers may face parenting and childrearing challenges if the other caregiver is incarcerated. The negative impact may be especially pronounced in children experiencing these two ACEs. More than 40% of jailed mothers of minor children are single caregivers (Glaze & Maruschak, 2010), leaving the childrearing responsibilities to other family members, adults, or social services. Policies like PFL that support caregivers with young children can be especially relevant for improving child and caregiver health outcomes.

There are important data limitations to consider. NSCH did not include ACEs assessment prior to 2011, hindering our ability to add California and New Jersey (that had implemented PFL in 2004 and 2009, respectively) as intervention states in the analysis. NSCH does not contain data on place or type of employment (i.e., full-time vs. part-time vs. hourly) and whether a caregiver was employed or eligible for PFL. While individual-level PFL eligibility may confound associations, this study’s adjusted models included state-level unemployment rates as covariates to partially account for the effects of employment status. Additionally, the number or percentage of people in each state that are eligible or have used PFL, as well as jobs that offer PFL is unknown. As of 2023, only 27% of private sector workers in the U.S. had access to PFL, and among the lowest wage workers, 95% had no access to PFL (U.S. Department of Labor, 2024). In some states like Rhode Island, federal, state and some municipal employees are not eligible to receive PFL benefits. This lack of knowledge prohibits understanding of the full scope of the issue and generalizations to other states. There are also methodological limitations worth noting. The study design cannot account for omitted variable bias for unmeasured characteristics that vary over time. As the current study utilized data from 2011–2019, it is possible that exposure change may be confounded by time-varying characteristics. Additionally, inferences from TWFE are vulnerable to biases when the number of time periods is limited or when the stringent parallel assumptions are violated.(T. D. Hill et al., 2020a, 2020b; Strumpf et al., 2017) Finally, BART computes inference using posterior credible intervals, as opposed to frequentist confidence intervals used by TWFE methods. (J. Hill et al., 2020a, 2020b) This study employed additional placebo checks to complement and increase the credibility of findings.

As more data on household challenge ACEs from additional intervention states become available, future studies can incorporate richer data to better estimate PFL impacts on ACEs. Additionally, examining variations in PFL implementation by length of time and benefits can contribute to a better understanding of the impacts of these policies on families.

ACEs may have tremendous negative effects on children’s health. Policies such as PFL that support families shortly after childbirth may be particularly relevant in preventing or reducing children’s experiences of household challenges. The results showed that implementation of state-PFL was consistently associated with reductions in two specific household challenge ACEs, caregiver separation and incarceration, as well as experience of any household challenge ACE. Given robust literature linking child health outcomes to household challenges, PFL may be an important policy for promoting caregiver well-being in the months surrounding childbirth, preventing ACEs, and improving children’s health outcomes. Existing research on the effects of PFL largely used data from California (Lee et al., 2018). This study expanded the literature by investigating PFL effects in two more states that implemented PFL. As PFL continues to be enacted and implemented in more states and at the federal-level, researchers can build on existing evidence to estimate PFL effects on young children’s health and well-being.