Pregnancy and infant health outcomes in the US continue to lag behind other countries with similar economic performance, with rates of infant and neonatal mortality and preterm birth typically 2 to 3 times higher than that of other high-income nations. 1 These adverse outcomes also differentially occur in patients by race, ethnicity, socioeconomic status, or rural vs urban residence. 2,3 As a result, there have been myriad programs and policies enacted on the state or local levels to address, and hopefully improve, the ultimate health of pregnant patients, newborns, and infants. Unfortunately, there is little information to help guide policymakers, advocates, or clinicians on the population health impact of such programs. The article by Chang et al 4 provides some evidence that state-level policies may influence pregnancy and newborn outcomes. This study calculated yearly state and local government expenditures per person with low income in multiple social programs, including state refundable earned income tax credits, cash assistance, childcare assistance, housing and community development, and public health. They found that, for every increase of $1000 per person with low income, overall rates of preterm birth, defined as delivery of a live-born infant at a gestational age of less than 37 weeks, was decreased by 1.4%. This effect was larger in infants of Black mothers and for expenditures in cash assistance, housing, and community development. Such research suggests that public policies on social programs may have short-and long-term impacts on infant health and may impact specific infant populations to a greater degree.However, this work also highlights the real challenges in developing and analyzing the effects of specific policies for pregnant people and their infants. The first issue is how to analyze such policies. Since it is extremely uncommon and often impossible to perform a randomized study of public policies, researchers are left with 2 primary methods to perform such analyses. The first is a difference-in-difference method, where states or regions that implement a policy are compared with states that did not, examining the change in the difference in outcomes between these groups after the policy is implemented. This method controls for both secular trends in outcomes and characteristics of the states or regions that do not change around the time of the policy implementation. Although this is a gold standard method of testing the effect of a policy, such work typically only studies 1 or a few states that implemented a specific policy; requires detailed data at the patient level, which can be challenging to obtain; and must pass multiple assumptions for the change in outcomes to be attributed to the policy change. 5