State‐Level Social and Economic Policies and Their Association With Perinatal and Infant Outcomes

Abstract

Policy Points.

• State‐level social and economic policies that expand tax credits, increase paid parental leave, raise the minimum wage, and increase tobacco taxes have been demonstrated to reduce adverse perinatal and infant health outcomes.

• These findings can help prioritize evidence‐based legislated policies to improve perinatal and infant outcomes in the United States.

Context

Rates of preterm birth and infant mortality are alarmingly high in the United States. Legislated efforts may directly or indirectly reduce adverse perinatal and infant outcomes through the enactment of certain economic and social policies.

Methods

We conducted a narrative review to summarize the associations between perinatal and infant outcomes and four state‐level US policies. We then used a latent profile analysis to create a social and economic policy profile for each state based on the observed policy indicators.

Findings

Of 27 articles identified, nine focused on tax credits, eight on paid parental leave, four on minimum wages, and six on tobacco taxes. In all but three studies, these policies were associated with improved perinatal or infant outcomes. Thirty‐three states had tax credit laws, most commonly the earned income tax credit (n = 28, 56%). Eighteen states had parental leave laws. Two states had minimum wage laws lower than the federal minimum; 14 were equal to the federal minimum; 29 were above the federal minimum; and 5 did not have a state law. The average state tobacco tax was $1.76 (standard deviation = $1.08). The latent profile analysis revealed three policy profiles, with the most expansive policies in Western and Northeastern US states, and the least expansive policies in the US South.

Conclusions

State‐level social and economic policies have the potential to reduce adverse perinatal and infant health outcomes in the United States. Those states with the least expansive policies should therefore consider enacting these evidence‐based policies, as they have shown a demonstratable benefit in other states.

Infant mortality has long been considered a marker for the overall health ¹ of a population. Despite a consistent annual decrease in infant mortality rates, the United States continues to rank high compared with other peer countries. ² While the reason behind this inconsistency is not well known, many hypothesize that it may be attributed to the country's socioeconomic and racial inequalities. ³ , ⁴ , ⁵ , ⁶ A comparison of infant mortality rates in the United States, Austria, and Finland revealed that individuals with a high socioeconomic status had comparable rates of infant mortality across the three countries, but that individuals with a lower socioeconomic status had the highest rates of infant mortality in the United States, and the differences in rates between socioeconomic groups in the United States varied significantly more than those in the two other countries. ⁷ Additionally, this study found that while income is a strong predictor of infant mortality in the United States, geography explained even more of the variation in rates observed between the three countries, suggesting that other, unmeasured, factors could be contributing to this observation. In addition to socioeconomic inequalities, race has consistently been found to predict birth outcomes in the United States, and more recent studies have begun to assess the impact of structural racism on perinatal and infant health. Not surprisingly, greater structural racism in education, work, and homeownership has been found to be associated with increased infant mortality among Black babies. ⁵ , ⁶ A discussion of infant mortality is not complete without mentioning low birthweight and sudden infant death syndrome (SIDS), which were among the top five leading proximal causes of infant death in the United States in 2018, mediated by preterm birth. ⁸ Thus, reducing the incidence of preterm births and low birthweight babies should remain a priority in the effort to reduce infant mortality rates.

Legislated efforts to directly or indirectly reduce rates of preterm birth and infant mortality in the United States by improving the health and quality of life of the mother/parents during either pregnancy or postpartum may include the enactment of certain economic and social policies. Four notable social and economic policies are (1) tax credits like the Earned Income Tax Credit (EITC), which can provide financial relief for working parents; (2) paid parental leave policies, which can increase the amount of time parents are able to spend with their newborns at home while still earning an income; (3) a higher minimum wage, which can reduce poverty in pregnant women and working parents; and (4) a higher tobacco tax, which can reduce maternal smoking rates.

What the literature is lacking is a comprehensive review of these policies focusing on the United States and, more specifically, how states are adopting them and to what extent. As such, we conducted a narrative review of the literature on these four policy changes and whether they reduced preterm births and infant mortality. Approaches for improving birth outcomes include “downstream” behavioral and medical interventions that occur upon engagement with the healthcare system, as well as “upstream” interventions that occur through social and economic policies. ¹ Downstream interventions typically operate on the individual level, and may consist of nutrition programs, breastfeeding support, management of maternal anemia, or the prevention and management of postpartum hemorrhage. ⁹

While insurance‐related policies, including Medicaid, are similarly important to maternal and child health, such policies also exert influence at the point of engagement with the health care system, and therefore a discussion of these policies is beyond the scope of this article, which focuses on upstream social and economic policies. As a secondary aim, we sought to identify a measurement of state profiles based on their adoption of these four policies. Our findings from this review can guide the prioritization of policies that can improve perinatal and infant outcomes and identify any remaining knowledge gaps.

Methods

Studies were eligible for inclusion in our review if they investigated the association between at least one of the selected social and economic policy measures and at least one of the selected perinatal or infant outcomes. The four policy measures we chose were (1) tax credits, including the EITC, child care, and child and dependent care; (2) paid parental leave; (3) minimum wages; and (4) tobacco or cigarette taxes. The outcomes of interest were pregnancy loss (including miscarriage and stillbirth), infant mortality (including SIDS), and (low) birthweight. We searched the PubMed and EconLit databases for articles in English based on these criteria without date restriction. We excluded studies that did not measure any of the four policy exposures, that did not measure any of the perinatal or infant outcomes, or that did not include US‐based data. Online Appendix Table 1 shows the details of our search criteria. We conducted this narrative review following the guidelines developed by the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA), ¹⁰ the Scale for the Assessment of Narrative Review Articles (SANRA), ¹¹ and the Academy of Nutrition and Dietetics. ¹²

We then conducted an exploratory latent profile analysis of the four aforementioned policy areas for all 50 states. This latent profile analysis allowed us to create an unobserved categorical variable based on the observed policy indicators, which can be interpreted as the social and economic policy profile of each state. The analysis contained four observed state‐level measures of the policy areas. We obtained tobacco tax data from the Campaign for Tobacco‐Free Kids for 2020, ¹³ and 2020 minimum wage data from the US Department of Labor. ¹⁴ We categorized paid parental leave type as none, legislation proposed, legislation enacted but narrow in scope (e.g., specific to one employer type, such as government), or legislation enacted and broad in scope based on a 2020 report from the National Conference of State Legislators. ¹⁵ We calculated tax credits as the cumulative sum of three dichotomous indicators—any earned income tax credit (yes/no), child tax credit (yes/no), child and dependent care credit (yes/no)—with a greater sum indicating more tax credits available. Our tax credit data came from the nonpartisan Tax Credits for Workers and Their Families initiative for 2020. ¹⁶

To visualize the policy profile obtained from the latent profile analysis, we created a map of the United States with each state shaded according to its latent profile. We obtained a state cartographic boundaries shapefile from the US Census Bureau, which we used for plotting. ¹⁷ As an exploratory aim, we created box plots to visualize the association between the latent policy profiles with infant mortality rate (IMR; mortality in children < 1 year of age per 1,000 live births) and fetal mortality rate (FMR; mortality in utero ≥ 20 weeks of gestation per 1,000 live births [pregnancy denominator was not available]). State‐level IMR for 2018 and FMR for 2014 to 2018 came from CDC WONDER (https://wonder.cdc.gov). All analyses were conducted in R version 3.6.3 (R Foundation for Statistical Computing, Vienna, Austria). The analytic code can be downloaded from https://doi.org/10.5281/zenodo.5013540, and Online Appendix Table 2 contains the data.

Results

Our initial search identified 305 publications (263 from PubMed and 42 from EconLit). After removing the duplicates, we screened 301 articles based on title and abstract. From these, we excluded 240 articles (108 based on birth outcome, 17 based on policy exposure, and 115 based on non‐US data), leaving 61 full‐text articles to be assessed for eligibility. We then removed an additional 34 articles (6 based on outcome, 3 based on policy exposure, 4 based on non‐US data, and 21 based on lack of research). Figure 1 shows the PRISMA flow diagram). We therefore had 27 articles for the final review, which are summarized in Table 1. Of these, nine articles focused on tax credits, eight on paid parental leave, four on minimum wages, and six on tobacco or cigarette taxes. The majority of the studies used a quasi‐experimental design (67%), followed by an ecological design (19%).

Figure 1.

PRISMA Flow Diagram for the Search and Selection of Articles [Colour figure can be viewed at wileyonlinelibrary.com]

Table 1.

Summary of Articles Included in the Review, Grouped by Policy Area: Tax Credits, Paid Parental Leave, Minimum Wages, and Cigarette Taxes

Author, Year	Study Design and Analysis	Population	Exposure	Outcome	Results
Tax Credits
Bruckner et al. 2013 18	Quasi‐experimental (1); Autoregressive integrated Moving Average (AMIRA) models	Low‐income non‐Hispanic Black women who gave birth in California between 1989 and 1997	Receipt of the EITC in the 2nd or 3rd trimester	Very low birthweight	Odds of very low birthweight increased above expected values two months immediately following tax disbursement.
Hamad & Rehkopf 2015 66	Quasi‐experimental (1); Multivariable linear regressions and Instrumental Variables (IV) analysis	Women surveyed in the 1979 National Longitudinal Survey of Youth that gave birth to children between 1986 and 2000	Size of EITC benefit	Birthweight	Greater EITC payment sizes were associated with increased likelihood of going to term, increased birthweight, and increased likelihood of breastfeeding.
Hill & Gurley‐Calvez 201967	Quasi‐experimental (1); difference‐in‐differences and triple differences	Births in Maryland between 1995 and 2004	Montgomery county EITC	Birthweight, probability of low birthweight	Local EITC policies reduced probability of low birthweight. Effects of the local credit appear to be similar to those generated by federal and state EITC credits.
Hoynes et al. 201521	Quasi‐experimental (1); Difference‐in‐differences analysis	Single mothers aged 18 and older with US singleton births between 1983 and 1999	EITC expansion (pre and post)	Low birthweight	EITC expansion reduces the incidence of low birthweight and increases mean birthweight. For single, low‐education mothers, policy‐induced treatment on treated increase of $1,000 in after‐tax income was associated with a 1.6% to 2.9% reduction in low birthweight rate. Impact was greater for births to Black mothers.
Komro et al. 201919	Quasi‐experimental (1); fixed‐effect panel data models	US singleton births to mothers 18 and older between 1994 and 2013	Presence and generosity of EITC payments for each of 50 states + DC	Birthweight	In states with EITC, birthweight gains were higher for black mothers (16.12g to 37.16g / 0.5% to 1.2% increase) than for White mothers (9.38g 28.40g / 0.3% to 0.9% increase), but not significantly different except among states with high EITC with no refund. There was a larger beneficial effect of EITC among Black mothers compared to White mothers for the probability of low birthweight and gestation weeks. Increases in birthweights for Hispanic mothers in states with EITC range from 1,126g (0.3% to 1.1%) increases in birthweight
Markowitz et al. 201722	Quasi‐experimental (1); multistate, multiyear difference‐in‐differences design	US births between 1995 and 2013	Generosity of state EITCs	Birthweight, probability of low birthweight, gestation weeks	Average birthweights increased by a range of 9g to 27g (0.3% to 0.8%). Larger effects occurred at the lowest birthweights and among states with the most generous EITC. Prevalence of low birthweight births were reduced as EITC generosity increased. Largest increases in birthweight and reductions in low birthweight births were in states with refundable EITCs.
Strully et al. 201020	Quasi‐experimental (1); difference‐in‐differences modeling strategy with state and year fixed effects	Between 1980 and 2002, US singleton births to single mothers with high school degree or less.	Presence of EITC (in state of birth at time of birth), AFCD/TANF	Birthweight	State EITCs increase birthweights by an average 16g. Living in a state with AFDC/TANF benefits in the top quartile increases birthweights by approximately 8g compared to the bottom quartile. Relative to EITC, structure of AFDC/TANF is more complex and variable across time and states.
Wagenaar et al. 201923	Quasi‐experimental (1); interrupted time‐series; Autoregressive integrated Moving average (AMIRA) models	All singleton births from January 1990 through December 2015 in DC and all comparison states	Changes in EITC	Prevalence of low birthweight, mean birthweight, mean gestation weeks	The rate of low birthweight births declined as magnitude of EITC credit increased: a reduction of 1.9 low birthweight births per 100 live births for 10% EITC credit, and a reduction of 4.7 low birthweight births per 100 live births for 40% EITC credit. Mean birthweight increased as magnitude of credit increased: an estimated 48g improvement with implementation of 10% credit, and an estimated 104g improvement with the 40% credit. Mean gestation weeks increased an estimated 0.12 after 10% credit was implemented, and 0.43 after 40% credit was implemented.
Wicks‐Lim & Arno 201724	Quasi‐experimental (1); ordinary least‐squares regression model, generalized linear model, difference‐in‐differences empirical strategy	Births in New York City between 1997 and 2010	EITC benefit changes	Proportion of low birthweight births	Increases in EITC benefits reduce low birthweight rates. A 10 percentage‐point increase in state and local EITC rates resulted in a 0.2 percentage‐point reduction in the low birthweight rate among low‐income neighborhoods.
Paid Parental Leave
Guendelman et al. 200928	Cohort; Cox proportional hazards, linear regression, logistic regression models	Women living in three Southern California counties that delivered live births between July 2002 and December 2003 and were eligible for the state's antenatal leave benefits	Antenatal leave	Gestational age	Women who took antenatal leave did not have longer gestations compared with non‐leave‐takers (but the study was underpowered). Among women whose efforts exceed their occupational rewards, gestation was markedly prolonged.
Heymann et al. 201125	Ecological; multivariate ordinary least squares regression models	Maternity leave legislation in 178 countries and neonatal/infant mortality statistics in 2007/2008	Leave duration and wage replacement rate	Neonatal mortality rate, infant mortality rate	National guarantees of lengthier paid maternal leave were associated with significantly lower neonatal mortality. Ten additional FTE weeks of maternal leave were associated with 10% lower neonatal mortality rate. National guarantees of lengthier paid maternal leave were associated with significantly lower infant mortality. Ten additional FTE weeks of maternal leave were associated with 10% lower infant mortality rate.
Jou et al. 201829	Cross‐sectional; logistic and multinomial logit regression models	Women aged 19 to 45 who gave birth to singleton infants in US hospitals from July 2011 to June 2012	Use and duration of paid maternity leave	Infant health status, rehospitalization	Women who took partially or fully paid leave experienced nearly 50% reduction in odds of having had their infants hospitalized (AOR 0.53; 95% CI 0.3 to 1.0). Women who took more than 12 weeks of leave saw nearly 75% decrease in the odds of having had their infant rehospitalized (AOR 0.26; 95% CI 0.1 to 0.9).
Khan 202026	Ecological; 2‐way fixed effects model	Country‐level data from 35 OECD countries on paid maternity/paternity leave and neonatal/infant mortality rates from 1990 to 2016	Paid maternity or paternity leave	neonatal mortality, infant mortality	After two years of the adoption of paid maternity leave, there was a statistically significant decrease in neonatal mortality and infant mortality. When accounting for country‐specific time trends, paid maternity leave was associated with 5.2% decrease in neonatal mortality rate and 2.4% decrease in infant mortality rate after two years of adoption.
Montoya‐Williams et al. 202030	Quasi‐experimental (1); difference‐in‐differences analysis	Live births in California, Missouri, and Pennsylvania from 1999 to 2009	2004 Paid Family Leave program in California	Postneonatal mortality	Postneonatal mortality was significantly decreased in California after 2004 (aOR 0.88, 95% CI 0.80to 0.97). This translated to a reduction of 1.94 deaths per 10 000 live births. When models were clustered by state, the authors found lower odds of preterm birth (aOR 0.98, 95% CI 0.97 to 0.98) and increased odds of low birthweight (aOR 1.05 95% CI 1.03 to 1.06) in California after 2004.
Pihl & Basso 201831	Quasi‐experimental (1); difference‐in‐differences analysis	California children under 1 year of age hospitalized between 2000 and 2007; hospitalizations in Arizona, Washington, and New York between 2005 and 2009	Paid family leave (California)	Infant hospitalizations	After CA made 6 weeks of paid leave available, hospital admissions for infants declined by 3% to 6%. Upper respiratory and gastrointestinal admissions declined by 25% to 33%, and 9% to 15%.
Stearns 201532	Quasi‐experimental (1); difference‐in‐differences analysis	Births in California, Hawaii, New Jersey, New York, and Rhode Island from 1972 to 1985	Access to Temporary Disability Insurance (TDI)	Low birthweight births	In four of the five states, implementation of paid maternity leave through TDI generally reduces the share of low birthweight births. In California, the policy reduces low birthweight births by 3.2% from the sample mean. In New Jersey, the policy reduces low birthweight births by 4% from the sample mean. Overall, TDI‐paid maternity leave reduces low birthweight births by 3.2% from sample mean. Paid leave reduces the proportion of early term births by 6.6%. Access to short‐term paid maternity leave had substantial impacts on children of unmarried and Black mothers.
Tanaka 200527	Quasi‐experimental (1); Ordinary least squares regression	Maternity and parental leave policies and infant/child mortality rates in 16 European countries, United States, and Japan	Paid maternity / parental leave	Infant mortality, postneonatal mortality, child mortality, birthweight	A 10‐week extension in paid leave is predicted to decrease the infant mortality rates, postneonatal mortality rates, and child mortality rates by 2.59%, 4.06%, and 3.02%, respectively. Paid leaves significantly decrease infant mortality, while other leaves (unpaid) have no significant effect. Paid leaves significantly decrease low birthweight.
Minimum Wages
Andrea et al. 202034	Quasi‐experimental (1); unconditional linear quantile regression and difference‐in‐differences analysis	Women aged 20 to 45 years of age residing in United States that gave birth to singleton infants between 2004 and 2016	Subminimum wage changes	Infant birthweight for gestational age z‐scores (BWz)	More generous wage policy was associated with increased birthweight (increase of 0.024) for the smallest infants, and decreased birthweight (decrease of 0.041) for the largest infants. The largest estimated potential benefit of subminimum wage increases was in the tails of the BWz distribution.
Komro et al. 201637	Quasi‐experimental (1); difference‐in‐differences analysis controlled for year‐fixed effects and state‐fixedeffects	Births in United States between 1980 and 2011	State‐level minimum wage changes	Low birthweight and postneonatal mortality	Across all models, a dollar increase in the minimum wage above the federal level is associated with a 1% to 2% decrease in low birthweight births and a 4% decrease in postneonatal mortality.
Rosenquist et al. 201936	Ecological; multilevel logistic modeling	Infants born in United States (all 50 states + DC) in 2010	State‐level minimum wage in 2010 and difference in minimum wage from 1980 to 2010	Infant mortality and neonatal mortality	No overall significant association found between minimum wage/change in minimum wage and infant or neonatal mortality. Among infants with non‐Hispanic Black mothers, those who lived in states with high minimum wage were less likely to experience infant mortality compared to those in states with low minimum wage (OR = 0.8, 95%CI = 0.68 to 0.94). Infants with non‐Hispanic Black mothers who lived in states where the minimum wage changed the most were less likely to experience mortality in the first year of life (OR = 0.89, 95%CI = 0.82 to 0.96).
Wehby et al. 201635	Quasi‐experimental (1); difference‐in‐differences analysis	Low‐educated women who gave birth in United States between 1989 and 2012	Minimum wage	Birthweight, gestational age, fetal growth	A $1 increase in minimum wage was associated with a 4.04g (0.1%) increase in birthweight, a 1% decrease in low birthweight, and a 0.075g (0.1%) increase in fetal growth.
Tobacco Taxes
Hawkins et al. 201439	Quasi‐experimental (1); Differences‐in‐differences analysis	Women aged 18 to 50 residing in 28 US states and Washington DC that gave birth to singleton infants between 2000 and 2010	Cigarette tax per pack by state	Birthweight, low birthweight, preterm delivery, small for gestational age (SGA), and large for gestational age (LGA)	Every $1 increase in cigarette taxes was associated with increases in birthweight by 5.4g and 4.0g among White and Black mothers with 0 to 11 years of education, as well as by 1.9g to 0.3g for Black mothers with 12 to 16+ years of education. Among White and Black mothers with a low level of education, every $1 increase in cigarette taxes was associated with reduced numbers of low birthweight infants by 0.08 and 0.12 percentage points, preterm infants by 0.07 and 0.08 percentage points, and SGA infants by 0.31 and 0.30 percentage points. Every $1 increase in cigarette taxes was associated with increased numbers of LGA infants by 0.18 and 0.10 percentage points for both White and Black mothers with a low level of education.
King et al. 201538	Ecological; Poisson fixed‐effects model	Sudden infant death syndrome cases in 23 countries between 1990 and 2009	Cigarette prices per pack (tax included)	Sudden infant death syndrome (SIDS) cases	A $1 increase in the average cigarette price is associated with a statistically significant reduction of 12.5 to 12.8 SIDS deaths.
Levy et al. 201640	Policy Simulation Model; Discrete Markov model	Simulation based on US Census population data from 1965 to 2012	Cigarette tax increase	Low Birthweight (LBW), Preterm Births (PTBs), Sudden Infant Death Syndrome (SIDS), and ectopic pregnancy	By increasing the tax by $1 between 2015 and 2065 (5% change/decline in all outcomes), the prevalence of smoking should decrease by 5.7% and projected number of outcomes should drop by 41,100 for LBW cases, 64,900 for PTBs, 900 for SIDS deaths, 7,200 for placenta previa cases, 10,900 for placental abruption cases, and 20,400 for ectopic pregnancies. By levying $2 more tax, the number of each MCH outcome should be reduced by 10%, about twice as high as a $1 tax increase.
Lien & Evans 200541	Quasi‐experimental (1); differences‐in‐differences analysis	Women who conceived before/after tax hikes in Arizona, Illinois, Massachusetts, and Michigan between 1991 and 1994	Cigarette tax increase	Birthweight, low birthweight births	All states had a statistically significant increase in birthweights and a decrease in the low birthweight rate for conceptions after the tax hike. The largest improvement was in Michigan where the full tax resulted in average birthweights increasing by 11g and the low birthweight rate decreasing by 0.3 percentage points. In Arizona, average birthweight increased by 8g and low birthweight rate decreased by less than 3/10%.
Markowitz 200842	Ecological; Fixed effects Poisson	Sudden infant death syndrome cases in United States between 1973 and 2003	Cigarette prices, cigarette tax	Sudden infant death syndrome (SIDS) cases	Across all models, the results show that higher cigarette prices are associated with fewer SIDS deaths. A $1 increase in cigarette price is associated with average reduction of 7.6 to 8.3 SIDS deaths; in elasticity terms, a 10% increase in the real price of cigarettes reduces SIDS deaths by 6.9% to 7.6 %. Every 1% increase in real cigarette tax reduces SIDS deaths by 0.16% to 0.18%; for price, an average tax increase of 10% (2 cents) would reduce national death rate in 2003 from 0.53to 0.52 per 1,000 births. Total bans in smoking in restaurants are associated with 4 fewer SIDS deaths on average.
Patrick et al. 201643	Time‐series model; Time‐series multivariable regression models	Infant mortality rates by US state between 1999 and 2010	Mean cigarette tax per pack, mean cigarette price per pack	Infant mortality rate	Every $1 increase in cigarette tax per pack was associated with a change in infant mortality rate of −0.19 (95% confidence interval [CI] −0.33 to −0.05) in the population overall. African American infants: −0.46 (95% CI −0.90 to −0.01), White infants: −0.21 (95% CI −0.33 to −0.08). With every $1 increase in cigarette price per pack, −0.19 (95% CI to −0.30 to −0.09) change in infant mortality rates overall, −0.35 (95% CI −0.68 to −0.01) for African American infants, and −0.20 (95% CI −0.29 to −0.11) for White infants.

Tax Credits

Nine articles assessed the association between the EITC and one or more of the outcomes. Classification of the EITC varied, with some studies comparing the EITC's presence or absence in the state or county of birth, and others comparing the size or extent of payments. Some studies differentiated between refundable and nonrefundable tax credits (refundable tax credits allow the eligible individual to receive a refund if the amount of the credit is larger than the tax owed). Although the outcomes were based on birthweight, the studies had various classifications, ranging from difference (in grams), percent change, or probability of a low or a very low birthweight birth. Some studies also included weeks of gestation as an outcome.

All but one study found a significant inverse relationship between the EITC and low birthweight (< 2500 grams), ¹⁸ the exception being a study that examined very low birthweight (< 1500 grams) as the outcome, with no statistical relationship observed. The presence of the EITC in the state of birth was associated with an increase in mean birthweight and a lower incidence of low birth weight births compared with states without the EITC. ¹⁹ , ²⁰ Those studies that measured the differences found a 9g to 38g increase in birthweight with the presence of the EITC. Komro and colleagues found a larger beneficial effect on birthweight gains in babies born to Black mothers compared with babies born to White mothers, but an insignificant difference for babies born to Hispanic mothers. ¹⁹

In the rest of the studies, an increase in the amount of tax credit was associated with an increase in birthweight or a decrease in the incidence of low birthweight births. ²¹ , ²² , ²³ , ²⁴ Wagenaar and colleagues found a reduction of 1.9 low birthweight births per 100 live births when the EITC credit was 10%, and a reduction of 4.7 low birthweight births per 100 live births at 40%. ²³ Markowitz and colleagues also found that the greatest increases in birthweights were in those states with the most expansive EITCs and in those with refundable EITCs. ²² One study also measured birthweight increases in states with Aid to Families with Dependent Children (AFDC) or Temporary Assistance for Needy Families (TANF) benefits, and revealed an average increase of 8 grams in birthweight for those states with top‐tier benefits compared with those with lower‐tier benefits. ²⁰

Paid Parental Leave

Eight articles assessed the association between paid parental leave and one or more of the perinatal or infant outcomes. Three of the studies used country‐level data, comparing several countries’ parental leave policies and infant or neonatal mortality rates. ²⁵ , ²⁶ , ²⁷ The remainder of the studies focused on state‐level policies in the United States. ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² As of September 2020, nine US jurisdictions (California, Connecticut, Massachusetts, New Jersey, New York, Oregon, Rhode Island, and Washington, plus the District of Columbia) that had paid family and medical leave laws. Ten additional states (Arkansas, Delaware, Idaho, Indiana, Kansas, Missouri, North Carolina, New Mexico, Tennessee, and Virginia) had more restrictive policies that covered some or all state employees. ¹⁵ The majority (60%) of the US‐based studies used California's Paid Family Leave program as a primary exposure, either by comparing outcomes in California before and after the program was implemented or by comparing outcomes in California with those in states without a paid leave policy. ²⁸ , ³⁰ , ³¹ Even though it was not specifically developed with new parents in mind, one study used access to Temporary Disability Insurance (TDI) as a proxy for paid parental leave. ³² Outcomes for these studies varied, with the majority focused on neonatal / postneonatal and infant mortality rates, followed by birthweight / low birthweight births, infant hospitalizations, and gestational age.

All but one study found a significant inverse relationship between paid parental leave and adverse perinatal and infant outcomes. Guendelman and colleagues found that women who took antenatal leave did not have longer gestations than women who did not take leave. ²⁸ This association was significantly different only when the authors restricted their analysis to women whose efforts at work exceeded their occupational rewards. They did mention, however, that their study was underpowered, which could explain the absence of statistical significance.

International studies found that the implementation or extension of a national paid parental leave policy was associated with a decrease in infant mortality rates by a range of 2% to 10%. For example, when Heymann and colleagues looked at maternity leave legislation and neonatal/infant mortality rates in 178 countries, they found that an additional ten full‐time equivalent weeks of maternal leave was associated with 10% lower neonatal and infant mortality rates. ²⁵ Tanaka and colleagues also predicted a decrease in infant and postneonatal mortality rates following a 10‐week extension in paid leave, as well as a decrease in low birthweight births. Their study did not, however, find a significant effect of unpaid leave on infant mortality rates. ²⁷

For the US‐based quasi‐experimental studies, paid parental leave was associated with a reduction in postneonatal deaths, lower odds of preterm births, and fewer infant hospital admissions. This was true for both studies that used access to a paid family leave program (i.e., living in California) as the outcome, ³⁰ , ³¹ as well as the study that used a personal paid leave experience (i.e., survey response) as the outcome. ²⁹ Access to TDI was associated with an average 3.2% reduction in low birthweight births, with significant effects observed in four of the five states providing TDI. ³²

Minimum Wage

Four articles assessed the association between minimum wage and one or more of the outcomes. Three of the studies focused on state‐level minimum wage changes, and one study on state‐level subminimum wage changes for tipped workers. According to the US Department of Labor, employers are required to pay their tipped workers only $2.13 an hour if their combined wage and tips equal the federal minimum wage. ³³ Outcomes varied across the studies but primarily included birthweight and neonatal/postneonatal and infant mortality. One study calculated infant birthweights for gestational age z‐scores, ³⁴ and another study included gestational age and fetal growth in their models. ³⁵

All but one study found an overall significant association between minimum wage change and the outcomes. Rosenquist and colleagues found that an increase in minimum wage was associated with a significant reduction in infant mortality, but only among infants with non‐Hispanic Black mothers. ³⁶ The other studies we reviewed revealed significant overall improvements with minimum wage increases. For example, Komro and colleagues and Wehby and colleagues showed that on average, every $1 increase in the minimum wage led to reductions of approximately 1% to 2% in low birthweight births and 4% in postneonatal mortality, and increases of 0.1% in birthweight and fetal growth. ³⁵ , ³⁷ In addition, Andrea and colleagues found that a more expansive wage policy was associated with a higher birthweight for the smallest infants and a lower birthweight for the largest infants. ³⁴

Tobacco Tax

Six articles assessed the association between tobacco taxes and one or more of the outcomes. One study assessed the association between outcomes and cigarette prices on a global scale, ³⁸ and the other five looked at differences by state within the United States. ³⁹ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³ All the studies looked at changes or differences in the cigarette taxes themselves, and some studies also included the overall price of cigarettes (tax included) in their models. ³⁸ , ⁴² , ⁴³ Outcomes varied across the studies, with the majority focused on birthweight / low birthweight births. A few studies looked at infant mortality rates, and three included the more specific outcome of SIDS.

All the studies found an overall statistically significant association between the tobacco tax and the outcomes. The studies focusing on birthweight found statistically significant increases in birthweight as cigarette taxes increased and statistically significant decreases in the rates of low birthweight births. For example, Lien and colleagues looked at average birthweights before and after cigarette tax increases in four US states and found the largest improvement in Michigan, where the tax increase resulted in the average birthweights rising by 11 grams and the low birthweight rate falling by 0.3 percentage points. ⁴¹ When Hawkins and colleagues looked at 10 years of tobacco tax changes in all 28 US states and Washington, DC, they found the most significant improvements in babies born to mothers with low education compared with mothers with high education. For example, they found that for every $1 increase in cigarette tax, birthweight increased by 4 to 5.4 grams for White and Black mothers with 0 to 11 years of education, and only 0.05 to 1.5 grams for White and Black mothers with ≥ 13 years of education. ³⁹ Infant mortality (including SIDS) was found to decrease as the tobacco tax increased. When studying at infant mortality rates by race/ethnicity, Patrick and colleagues found a change in the infant mortality rate of ‐0.19 in the population overall, ‐0.46 in infants born to Black mothers, and ‐0.21 in infants born to White mothers. ⁴³ The studies focusing on SIDS deaths observed a reduction of between 7 and 13 deaths with every $1 increase in the price of cigarettes. ³⁸ , ⁴²

State Social and Economic Policy Profiles

The distribution of the four observed policy indicators is summarized in Supplemental Table 2. The states had the following tax credit laws: earned income tax credit (n = 28, 56%), child tax credit (n = 6, 12%), and child and dependent care (n = 22, 44%). Seventeen states did not have any tax credit laws, and only four states had all three tax credit laws. Thirty‐two states did not have any parental leave laws; nine had a law that was narrow in scope; and nine had a law that was broad in scope. Five states did not have a state minimum wage law. For the others and compared with the federal minimum wage of $7.25, two were lower, 14 were equal, and 29 were higher (for states with minimum wage laws lower than the federal minimum, the federal minimum applies). Of the states with a higher federal minimum wage, the average state's minimum wage was $10.49 (standard deviation = $1.43) per hour. The average state cigarette tax was $1.76 (standard deviation = $1.08).

Based on these variables, the latent profile analysis identified three policy profiles as the optimal solution (entropy = 0.97). Policy profile 1 generally had the lowest cigarette tax, lacked a state minimum wage, and tended to have the least amount of paid parental leave and fewer tax credit laws. Policy profile 3 generally had the highest cigarette tax, the highest minimum wage, the greatest amount of paid parental leave, and more tax credit laws. Policy profile 2 was an average of profiles 1 and 3. Figure 2 depicts the distribution of these policy profiles by state. Policy profile 1 states were located in the southern United States, and policy profile 3 states tended to be on the West Coast or in the Northeast, with the exception of Colorado. As shown in Supplemental Figure 1, states in policy profile 1 tended to have a higher average IMR and FMR, and those states in policy profile 3 tended to have a lower average IMR and FMR, although we caution against overly interpreting these associations, as they are not adjusted for potential confounding.

Figure 2.

Map of the United States Depicting Three Social and Economic State Policy Profiles Based on Four Indicators: Tax Credits, Paid Parental Leave, Minimum Wage, and Tobacco Tax

Policy profile 1 generally had the lowest tobacco tax, lacked a state minimum wage, and tended to have the least amount of paid parental leave and tax credit laws. Policy profile 3 generally had the highest tobacco tax, the highest minimum wage, and the greatest amount of paid parental leave and tax credit laws. Policy profile 2 was an average of profiles 1 and 3.

Discussion

The complications of infant mortality and preterm birth observed in the United States are long‐standing, complicated, and multifactorial. Social determinants of health are frequently recognized as being the key driver for high rates of preterm birth and infant mortality in marginalized communities, particularly communities of color. ⁴⁴ , ⁴⁵ , ⁴⁶ More recently, the dialogue has been elevated to include not decades but centuries of historical factors and policies that have left these communities vulnerable and prone to adverse perinatal and infant outcomes. ⁵ , ⁴⁷ , ⁴⁸ These complex but related outcomes cannot be blamed on a single gene, social factor, personal choice, or policy.

Over the past few decades, both health care spending and policymaking have consistently been topics of high priority for Americans and American politicians. ⁴⁹ Aside from the high burden of morbidity and mortality on the American population, preterm and low birthweight births are extremely costly. The Institute of Medicine estimated that in 2005, the yearly cost of premature births in the United States was $26.2 billion. ⁵⁰ In addition, a 2019 study of more than 700,000 infants born in the United States revealed that during the first six months of life, preterm infants incurred medical expenditures of an average of $76,153; low birthweight infants incurred an average of $114,437, and infants born at 24 weeks gestation incurred an average of $603,778. ⁵¹ Public policies aimed at reducing preterm and low birthweight births will not only save lives and improve perinatal and infant outcomes, it will also help relieve the existing financial burden.

This narrative review supports the proposal that tax credits (most notably the EITC), paid parental leave policies, a higher minimum wage, and a higher tobacco tax are associated with improved perinatal and infant outcomes in the United States. We observed that rates of infant mortality and low birthweight births were lower overall in those states that adopted the EITC, those with paid parental leave policies, those with a higher minimum wage, and those with a higher tobacco tax. There are many ways in which these policies are thought to affect perinatal and infant outcomes, for example, by reducing poverty, increasing the time new parents spend with their child, and reducing rates of maternal smoking. While a full investigation into these pathways is beyond the scope of our article, the literature does offer detailed descriptions of these mechanisms. ²⁶ , ²⁹ , ⁵² , ⁵³ , ⁵⁴ , ⁵⁵ , ⁵⁶ , ⁵⁷ , ⁵⁸ , ⁵⁹

A key limitation of our narrative review is that it does not cover all the policies influencing infant health outcomes. The social and economic policies we selected are those most commonly studied in the field of perinatal and infant health. While health care and insurance policies such as Medicaid and the Patient Protection and Affordable Care Act are important to maternal and infant health in the United States, these policies operate at the point of care and thus were outside our review of upstream social and economic policies. As with any research study making associational claims, we should acknowledge the feasibility of changing the exposure of interest in the target population, which would be much more difficult for Medicaid and related health care policies than the four social and economic policies described in our review. Nonetheless, apart from the literature on these associations, ⁶⁰ , ⁶¹ , ⁶² , ⁶³ extrapolating from the limited resources available is beyond the scope of our review.

Most of the studies in our review have quasi‐experimental designs. These methods (specifically difference‐in‐differences and interrupted‐time‐series analyses) are increasingly common in studies with population‐level exposures, such as policy interventions. Effect estimates from observational quasi‐experiments have been shown to come close in causal strength to those obtained through clinical trials, owing to their ability to control for unobserved confounding in effect size estimation. ⁶⁴ Ensuring valid causal inference through quasi‐experimental studies requires a rigorous understanding and examination of certain assumptions, such as parallel trends and the stable unit treatment value assumption. ⁶⁵ Such studies may suffer from residual confounding, possibly inducing the observed associations. The associations we observed were consistent across multiple studies. As with any review of the literature, the results of our review may be skewed away from the null owing to publication or other reporting biases.

The pathway between legislation and perinatal and infant health outcomes is complex. The pathway may have many direct and indirect factors, and so we were unable to identify these factors. A strength of our work is the consideration of multiple policies concurrently as well as the development of state social and economic policy profiles. Our findings may prove useful as an evidence base for agencies and elected officials to prioritize perinatal and infant health policies.

Infant mortality and preterm birth are the beacons that have been blinking for years, signaling the health effects of our collective history. As we have demonstrated, certain state‐level policies such as child or dependent tax credits, paid parental leave policies, a higher minimum wage, and a higher tobacco tax have the potential to reduce these adverse maternal and child health outcomes in our country. As displayed in our policy profiles, there is a legislative disparity between states. Those states with the least expansive policies (such as those in policy profile 1) should consider enacting these evidence‐based policies, as they have shown a demonstrable benefit in other states. Future research should examine additional social, economic, and health‐related policies, as well as how the combination of these legislative efforts has collectively affected maternal and child health outcomes in the United States.

Funding/Support: None.

Conflict of Interest Disclosures: All authors have completed the ICMJE Form for Potential Conflicts of Interest. No conflicts were reported.

Supporting information

Supplemental Table 1. PubMed and EconLit Search Terms

Supplemental Table 2. State‐Level Measures of the Four Policy Areas and Infant and Fetal Mortality

Supplemental Figure 1. State‐level crude associations between three social and economic state policy profiles and infant (panel A) and fetal (panel B) mortality.