What are the effects of paid family leave programmes in the United States?

To compensate for the lack of national provision, nine US states plus the District of Columbia have launched their own paid family leave programmes, with four more in the works. Studies highlight the effects of these policies on health, fertility and employment, but there is still much to learn.

The United States is today the only OECD country without a national paid family leave (PFL) programme (OECD, 2024). The benefits provided by such a programme offer workers the opportunity to take paid time off for personal or family circumstances, most commonly around childbirth or to provide care for a loved one (New York State Department of Financial Services, 2022).

The absence of a federal programme means that many Americans continue to miss out on the benefits associated with such policies. US studies detail these to include:

An improvement of 9% in self-reported maternal health (Bullinger, 2019).
A reduction in total infant mortality by 5% (Pihl and Basso, 2019), with a 33% reduction in mortality for upper respiratory diseases.
An increase in the ability of lower-income women to retain employment in the months after childbirth by 5-8% (Byker, 2016).

With a federal PFL programme still subject to congressional debate (Caldwell, 2024), the provision of these benefits instead falls to other bodies. This comes primarily in two forms.

First, a quarter of American workers have access to PFL benefits through their employer (Bartel et al, 2023). Second, and our focus here, are the nine states and the District of Columbia (DC) that have established their own mandatory PFL programmes since 2004. This total will increase to 13 by mid-2026 (abetterbalance, 2024). There are also a further eight states with related voluntary systems (see Figure 1).

Figure 1: Available paid family leave programmes in the United States, 2024

With one exception – New York – states with mandatory programmes use pooled payroll taxes on employees and/or employers to fund their benefits in a social insurance system (Bipartisan Policy Center, 2024).

Those with voluntary programmes, on the other hand, allow private insurance firms to sell PFL benefits as part of health insurance packages. Several states (such as New Hampshire and Vermont) set up and govern this market, while in New York, the purchase of such insurance is mandatory and tightly regulated (Bipartisan Policy Center, 2024).

The benefits offered by these policies are a step up from the unpaid leave available through the federal Family and Medical Leave Act (1993). But this is not to say that they are generous relative to those offered by other countries.

Compared with the OECD average of 51.9 weeks of paid leave that mothers can take around childbirth, state PFL programmes in the United States typically offer only 10-12 weeks of leave (abetterbalance, 2024).

What effects have these programmes had?

Economists have researched the effects of the rollout of state PFL policies. Most studies have looked at California’s programme, which is the oldest and largest. We can group what has been learned into a few key areas:

Health

Improved self-reported maternal health (Bullinger, 2019; Lee et al, 2020; Kim, 2024).
Improved infant health, as reported by parents (Stearns, 2015; Lichtman-Sadot and Bell, 2017; Bullinger 2019).
Increased rates of breastfeeding (Huang and Yang, 2015; Hamad et al, 2019).
Lowered rates of food insecurity (Lenhart, 2021; Kim and Lenhart, 2024).

Employment

Increased job retention for mothers in the short term (Byker, 2016) and longer term (Baum and Ruhm, 2016).
Higher wages and weekly hours worked in the years after childbirth for employed mothers, with both increasing by 10-17% (Rossin-Slater et al, 2013).
No reduction in employers’ satisfaction with their employees’ productivity and morale following the introduction of PFL (Bartel et al, 2023).
Increased leave taking by mothers – by almost five weeks – and corresponding fathers – by two to three days (Baum and Ruhm, 2016).

Reproductive choices

Increased fertility of women, particularly for women in their 30s, Hispanic women (Golightly and Meyerhofer, 2022) and better educated women (Thunell, 2017; Golightly and Meyerhofer, 2022).
A 13% reduction in abortion rates for the 20-39 age group (Allanson et al, 2024).

While the majority of current studies consider the difference that having access to PFL can make to parents’ decisions and outcomes around childbirth, a smaller group of researchers have explored whether a similar impact can be observed when PFL is used to care for a family member (Waldfogel and Liebman, 2019).

One study notably finds that healthy wives are 7% less likely to report leaving employment to care for the household or family if they have access to PFL. This effect is not found among healthy husbands, although these men do reduce their working hours by 3.5 hours per week on average (Coile et al, 2022).

Qualitative evidence of the effect of state PFL programmes is sparse, although one study does interview mothers in PFL states about their decisions around returning to work. Of the low-income mothers in this study, 71% were able to return to employment. These mothers stated that both the financial support and the weeks of leave afforded by access to PFL aided this return and the wider transition into parenthood (Winston et al, 2019).

Research conducted in the United States on PFL complements existing evidence from other industrialised nations, reinforcing its legitimacy. For example, this can be seen with maternal health, as the US evidence on improvements in this area (discussed above) aligns with similar effects previously seen in Denmark (Beuchart et al, 2016) and Norway (Bütikofer et al, 2021).

The policy setting of the United States does lend itself particularly well to economic research on this issue, as individual states are able to create unique legislation that is distinct from that of other states and the federal government. This allows for some novel studies to be conducted, such as one demonstrating that New York’s PFL programme reduced abortion rates in the state, particularly for younger women (Allanson et al, 2024).

In sum, state PFL programmes can make a powerful contribution to the wellbeing of women, children and families. Existing evidence demonstrates benefits to key areas including health and employment across many PFL-providing states, with no loss to employers (Bartel et al, 2023).

Is this evidence credible?

The evidence provided by the studies surveyed above is highly credible, generated using contemporary methods of causal inference. This field of study uses robust statistical and economic techniques to separate the true causal effects of policies from other factors and background noise. Studies of the effects of paid family leave that employ these methods are often published in peer-reviewed economic journals.

Many studies use some configuration of the ‘difference-in-differences’ method, wherein the change in an outcome over time is compared between a state that provides PFL and an otherwise closely comparable state(s) that does not. This produces robust estimates of the effect of having access to such benefits.

Are there factors limiting further research?

A major bottleneck in the study of state PFL programmes, and indeed all economic research, is the availability of applicable data. The robust methods employed by many of the studies referenced here often require data on an outcome of interest for each and every state for a period of many years. This can be a tall order when attempting to explore effects on particular aspects of individual behaviour.

Economists also often demand additional information on ‘control’ variables that serve to parse out true effects from the influence of other background factors, intensifying data constraints. For example, information on the ethnic, income and poverty profiles of states is often required to rule out whether it is the unique characteristics of a state that explain observed effects, rather than the policy itself.

For some of the more exotic techniques employed in this area, such as the ‘synthetic control’ (Golightly and Meyerhofer, 2022) or ‘synthetic difference-in-difference’ (Allanson et al, 2024) methods, such data may also be required on a consistent basis.

These methods compare the state of interest against a ‘model’ version of itself, constructed by creating a weighted average of the untreated control states that is as closely comparable to the treated state as possible. This process requires annual data to perform, which renders impossible study of factors with spotty state reporting. This problem was particularly prevalent around the time of the Covid-19 pandemic.

As a result, there is significant room for further learning in areas where consistent state reporting is lacking. This is most relevant in terms of surveys, which economists often employ as sources of data (Kim and Lenhart, 2024).

Take our current understanding of the effect of state PFL programmes on reproductive decisions as an example. While economists have already explored the effect on fertility outcomes, the study of fertility intentions is hamstrung by a lack of state-by-state survey data on this topic. Access to such data would help to explain the determinants of changes to fertility outcomes already detailed (Thunell, 2017; Golightly and Meyerhofer, 2022). Specifically, it would help to untangle how and why access to PFL has influenced women’s reproductive choices.

A distinct but related issue concerns the ability of economists to replicate results across multiple PFL-providing states. While a study may demonstrate an effect following the launch of one specific PFL programme, it can remain unclear whether we should expect a similar effect associated with each of the other respective policies (abetterbalance, 2024).

Again, data limitations constrain economists’ ability to abate this concern. To take abortion as an example, not all states consistently report the abortion counts for different age groups, including PFL-providing California. This prevents researchers from confirming if the decline in abortions following the launch of New York’s PFL programme followed a potential decline in the golden state several years prior (Allanson et al, 2024).

Due to the recency of many of these programmes, it may simply be a matter of time before economists have the data to determine if effects observed previously with older initiatives are seen again with those more recent.

In the meantime, researchers will continue to learn all they can about these important policies as they continue to rollout. With a federal programme still a distant dream – the house bipartisan working group has established an action framework but there is no funding plan at present (Caldwell, 2024; Wang, 2024) – learning all we can about the set of state programmes will continue to be essential.