Abstract

An increasingly hostile policy climate has reshaped abortion access in the United States. Recent literature has studied the effects of restrictive abortion policies on reproductive health outcomes. This study is the first to investigate the association between state-level abortion policy hostility and the pregnancy intentions of women with a pregnancy resulting in live birth. Data are from the Pregnancy Risk Assessment Monitoring System survey, merged with a state-level legislative database from 2012–2018 and other state-level controls. Cross-sectional results reveal that a one-unit increase in abortion policy hostility is associated with a relative risk (odds) of having a live birth resulting from an unintended versus intended pregnancy that is 1.02 times as high (RRR = 1.02, 95% confidence interval = 1.01, 1.03). This result corresponds to a 13% increase in the predicted probability of having a live birth resulting from an unintended pregnancy between a zero-hostility and a maximum-hostility state. Models stratified by demographic and socioeconomic characteristics reveal that the association between abortion policy hostility and live birth resulting from an unintended pregnancy is particularly robust among women in younger, less educated, Medicaid, uninsured, and rural populations.

Introduction

Unintended pregnancy and childbearing have long been controversial public health topics in the United States. The national rate of unintended pregnancy declined from 1980 to 2000, rose slightly in the early 2000s, and then decreased again after 2008. The U.S. rates of unintended pregnancy and abortion are at present comparable to those of European countries (Bearak et al. 2022). Still, more than one third of births among U.S. women aged 15–49 surveyed between 2015 and 2017 were either mistimed or unwanted (National Center for Health Statistics 2019). Unintended pregnancy and childbearing can be costly—for both individuals and society—in terms of potential losses to human capital accumulation and labor market participation (Trussell et al. 2013). Recent evidence further reveals that unintended childbearing as a result of abortion denial has direct negative economic and financial consequences for women in terms of significant, sustained financial distress (Miller et al. 2023). Additionally, unintended pregnancy and childbearing affect public health outcomes, as women having unplanned births are less likely to receive early prenatal care and to breastfeed, and their children are more likely to be low birth weight (Kost and Lindberg 2015).

The battle over the legality of abortion in the United States is increasingly a concern from a reproductive justice standpoint, defined by SisterSong (a national multi-ethnic reproductive health collective) as “the human right to maintain personal bodily autonomy, have children, not have children, and parent the children we have in safe and sustainable communities” (Price 2020; Smith 2005). Examination of the “costs” and “choices” regarding unintended pregnancy and childbearing utilizing a reproductive justice framework is important, given that the choice paradigm in the mainstream reproductive rights movement has historically excluded the experiences and needs of marginalized populations (Littlejohn 2021; Price 2020; Smith 2005). Yet these are the very groups already at higher risk of experiencing challenges to attaining reproductive autonomy because of structural and institutional factors; non-Hispanic Black women, Hispanic women, younger women, lower income women, rural women, cohabiting women, and women with lower educational attainment all face an elevated risk of unintended pregnancy (Finer and Zolna 2016; Musick 2002; Musick et al. 2009; Sutton et al. 2019). These disparities also manifest in births. The unplanned birth rate is approximately seven times higher for low-income women than for high-income women (Finer and Zolna 2016), and unintended childbearing is more likely among women without a four-year degree than among their college-educated counterparts (Finer and Zolna 2016; Guzzo and Hayford 2020). In addition, non-Hispanic Black and Hispanic women are more likely to experience unintended pregnancy and births than are non-Hispanic White and Asian women (Finer and Zolna 2016; Guzzo and Hayford 2020; Sutton et al. 2019). Furthermore, among rural and urban Black adolescents experiencing their first live birth between 2002 and 2017, 60% and 51%, respectively, reported that their pregnancy was unintended (Sutton et al. 2019).

States are passing increasingly more restrictive reproductive rights policies, and the passage of these laws has escalated at an unprecedented rate since 2020 (Nash 2021). Further state-level restrictions have been passed since the reversal of Roe v. Wade in Dobbs v. Jackson in the summer of 2022 (Jimenez 2022; Sharfstein 2023). A rich body of empirical research shows that restrictive state-level abortion policy is generally associated with negative outcomes concerning access to reproductive and maternal health care; however, no research to date has examined the effect of restrictive abortion policy and the hostile climate it creates in a state on whether women who carried their pregnancies to term and gave birth are more or less likely to report their pregnancies were unintended. Several different types of abortion restrictions may contribute to an increased prevalence of unintended pregnancies that are carried to term and result in live birth, because such policies can hinder women's ability to exercise reproductive autonomy (Foster 2020; Gerdts et al. 2022; Upadhyay et al. 2021, 2013; White et al. 2019). Restrictive abortion policies—and the hostile climate toward reproductive decision-making and autonomy they create in a state—may be related to the myriad of reasons that lead a woman whose pregnancy was unintended to give birth.

This study uses individual-level survey data of women1 who had a recent live birth and a U.S. state-level legislative database of abortion restrictions from 2012–2018 to empirically examine the association between state-level restrictive abortion policies and the subsequent hostile environment toward abortion they create in a state (henceforth referred to as “abortion policy hostility”) and women's pregnancy intention—focusing on the likelihood that a live birth results from an unintended versus intended pregnancy. The main predictor of interest is an index of abortion policy hostility based on a state's passage of 10 categories of abortion restrictions in a given year. To control for potential unobservable state-level characteristics that may be correlated with both passage of abortion restrictions and live births resulting from unintended pregnancies, we include state-level data on religiosity and political conservativeness. Finally, we estimate our empirical model stratified by sociodemographic subgroups to apply a reproductive justice lens to existing reproductive health policy. Certain populations are at greater risk of both unintended pregnancy and childbearing and may be more vulnerable to abortion policy hostility. Results are interpreted while considering the growing heterogeneity across states in access to reproductive health care and the potential impact this may have on groups who are already the most vulnerable to adverse reproductive health outcomes.

Reproductive Health Care Policy: Challenges and Restrictions

Reproductive health care policy in the United States has increasingly focused on abortion in recent decades. In 2021, states passed 108 abortion restrictions, making it the worst year for abortion rights since 1973, when Roe v. Wade was decided (Nash 2021). Then, on June 24th, 2022, in Dobbs v. Jackson, the U.S. Supreme Court overturned Roe, ruling that there is no longer a constitutional right to abortion in the United States (Dobbs v. Jackson Women's Health Organization 2022).

The number of abortions obtained in the United States, as well as the abortion rate, peaked in the 1980s and then steadily declined until 2017 (Kortsmit et al. 2020; Nash and Dreweke 2019), though between 2017 and 2020, the annual number of abortions increased 8% (Jones et al. 2022). Higher rates of contraceptive access and use—including increased affordability of contraception because of such policies as the Affordable Care Act—and growing reliance on highly efficacious, long-acting reversible contraceptives such as intrauterine devices and implants are likely driving the general decline in pregnancies, abortions, and births (Nash and Dreweke 2019; Sonfield et al. 2015). Yet, abortion restrictions may disproportionately affect vulnerable women: for example, the proportion of women seeking abortions increased seven percentage points between 2008 and 2014 among those with a family income of less than 100% of the federal poverty line (Jones and Jerman 2017).

A sizable and growing body of research has assessed the effects of restrictive state-level abortion policies on abortion rates, reproductive health outcomes, and maternal mortality (Austin and Harper 2019a; Bossick et al. 2021; Brown et al. 2020; Hawkins et al. 2020). Studies focusing on the effects of targeted regulation on abortion providers (TRAP) laws—a specific category of abortion restrictions that place logistical or structural requirements on abortion-providing facilities (Guttmacher Institute 2020)—found that such laws make providing abortions much more expensive and logistically difficult for clinics and other health facilities. Pursuant to a 2013 TRAP law in Texas, many clinics shut down, resulting in greater travel distances and higher out-of-pocket costs for women seeking abortions (Gerdts et al. 2022). In addition, although there was an overall decrease in total abortions, women were more likely to obtain an abortion in the second trimester—highlighting the law's negative impact on access to care (White et al. 2019). Two other studies assessing the impact of TRAP laws on annual abortion rates (Austin and Harper 2018, 2019a) found mixed and nonsignificant results regarding population-level metrics of abortion rates. Taken together, these recent studies underscore how TRAP laws function differently at the individual level, as shown with microdata, compared with the population level.

Other studies have examined the impact of gestational age limits, or laws that prohibit abortion after a specific point in the pregnancy. The Turnaway Study found that women who were denied wanted abortions experienced increased negative outcomes, such as reduced financial security and increased likelihood of experiencing poverty or intimate partner violence (Foster 2020). Moreover, having a wanted abortion was associated with positive mental health outcomes, including succeeding with aspirational life plans within a year postabortion (Foster 2020), as well as significantly fewer anxiety symptoms and higher self-esteem and life satisfaction in the short run, when compared with women who were denied an abortion—although these differences narrowed over time (Biggs et al. 2017). A six-week gestational age limit policy was also found to be associated with increased demand for self-managed abortion medication from an online provider (Aiken et al. 2022).

Although it is important to examine the effects of a single law, such as gestational age limits or TRAP laws, typically states that impose one of these laws have also enacted other policies that challenge women's reproductive autonomy. Researchers have examined how abortion restrictions collectively contribute to a generally hostile climate in a state and how that environment is associated with reproductive health and access-related outcomes. For example, Brown and colleagues (2020) found that a highly restrictive state legislative climate (based on four categories of abortion restrictions) was associated with a significant 17% decrease in the county-level abortion rate (Brown et al. 2020); however, they did not find an association with distance to an abortion-providing facility, suggesting that it is the restrictive climate itself and not the practical effects of the restrictions that was the barrier to care (Brown et al. 2020). Another study of a sample of women seeking abortion information online found that living in a restricted versus protected abortion access state was associated with 1.7 times greater odds of planning to continue a pregnancy and of continuing to seek an abortion at four weeks' follow-up, although this latter finding was only marginally significant (Upadhyay et al. 2021).

State-level abortion policy that directly reduces access to family planning and reproductive health services is also associated with increased maternal mortality in the United States (Hawkins et al. 2020). Specifically, reducing the proportion of Planned Parenthood clinics by 20% from the state-year mean was associated with an 8% higher maternal mortality rate, and enacting legislation restricting abortion based on gestational age limits was associated with a 38% increase in the maternal mortality rate (Hawkins et al. 2020). Additionally, a meta-analysis found that restricting access to family planning and abortion services and laws that restrict women's reproductive autonomy generally were associated with higher rates of adverse reproductive health outcomes at the state level (Bossick et al. 2021). Furthermore, recent research estimated that a near-total abortion ban would increase pregnancy-related maternal mortality by 21%, with larger increases among racial minority women (Stevenson 2021). This body of work shows that restrictive abortion policy at the state level is generally associated with lower abortion rates and negative health and well-being outcomes for mothers, including mortality. Our study adds to this literature by examining how restrictive abortion policy and thus the hostile climate it creates in a state are associated with live births that result from unintended versus intended pregnancies. We also seek to understand whether hostile abortion policy is associated to a greater extent with live births resulting from unintended versus intended pregnancies among more vulnerable populations.

Methods

Data

Data used in this study came from four sources. Individual-level data were obtained from the Pregnancy Risk Assessment Monitoring System (PRAMS) survey of women who have had a recent live birth; the questionnaire is available in both English and Spanish (Centers for Disease Control and Prevention 2023). The PRAMS survey was created as a joint product between the Centers for Disease Control and Prevention (CDC) and individual state health departments to provide ongoing, state-specific, population-based data about individuals' experiences of pregnancy and infant and postpartum health outcomes during the first few months after the birth. Participating states sample 100–250 women every month who have had a recent live birth. Data collection methodology is standardized across states to allow for multistate analysis, and all states include standardized questions on attitudes and feelings about a woman's most recent pregnancy, preconception care, contraceptive use, prenatal care, risky health behaviors during pregnancy, and infant health care in the months following the birth. The PRAMS data are not publicly accessible and were obtained through an application process to the CDC. The Institutional Review Board (IRB) for Human Participants Research at Cornell University deemed this research exempt from the IRB review process.

Our study relied on two phases of the PRAMS survey: Phase 7 (2012–2015) and Phase 8 (2016–2018). The data were not nationally representative because, in our study period, only 42 states and Puerto Rico both participated in the survey and met the response rate threshold to be released (see online appendix Table A1 for a list of states available from PRAMS during our study period). The total sample size for Phases 7 and 8 was 248,169 respondents. We excluded observations if data were missing for the outcome variable of pregnancy intention (n = 4,742) or had missing demographic data (n = 9,935).2 Observations from Puerto Rico were also dropped as we had no abortion policy hostility data for them (n = 1,553). Our final PRAMS analytic sample contained 231,939 respondents from 41 states. Data were weighted in accordance with PRAMS guidelines to account for survey nonresponse and noncoverage (Shulman et al. 2018).

We used two sources of data to obtain state-level covariate measures. Data on religiosity were obtained from the 2014 Pew Research Center Religious Landscape Study, and data on the proportion of Republicans elected to the U.S. House of Representatives were obtained from the Library of Congress (Library of Congress 2022; Pew Research Center 2014). The state-year–level legislative data on abortion restrictions were obtained from the Guttmacher Institute through a data-sharing agreement that provided access to legislative archives used in the earlier version of a previously published article by Elizabeth Nash, at the time a principal policy associate at the Guttmacher Institute (Nash 2019). This database was compiled using legislative records and is a tabulated list of the number of categories of abortion restrictions every state had in place on January 1st of a given year from 2012 to 2018. The ten categories of abortion restrictions include: (1) trigger laws (laws preemptively passed to immediately ban abortion in the event Roe was overturned); (2) 20-week abortion ban or unconstitutional postviability abortion restriction (gestational age limit); (3) provision of inaccurate or misleading counseling; (4) in-person counseling that requires two separate trips to the facility; (5) mandated ultrasound; (6) abortion coverage restricted in all private health insurance; (7) abortion coverage restricted in Medicaid, except in cases of rape, incest, or life endangerment; (8) restrictions on medication abortion (via either telemedicine or FDA protocol); (9) parental involvement in a minor's abortion (notice, consent, or both); and (10) TRAP laws, specifically clinic regulations and admitting privileges.

Measures

Outcomes

The PRAMS dataset includes a measure of pregnancy intention for a woman's most recent pregnancy that resulted in live birth, with five mutually exclusive categories based on survey items. Respondents are asked, “Thinking back to just before you got pregnant with your new baby, how did you feel about becoming pregnant?” Response options for the years in this study include “I wanted to be pregnant then,” “I wanted to be pregnant sooner,” “I wanted to be pregnant later,” “I didn't want to be pregnant then or at any time in the future,” and “I wasn't sure what I wanted.” Consistent with past research (Kost 2015; Sutton et al. 2019), responses of “I wanted to be pregnant sooner” and “I wanted to be pregnant then” were consolidated into a category of “intended pregnancy.” Responses of “I wanted to be pregnant later” and “I didn't want to be pregnant then or at any time in the future” were consolidated into a category of “unintended pregnancy.” The answer choice of “I wasn't sure what I wanted” is indicative of pregnancy ambivalence or uncertainty and is distinctly separate from either intendedness or unintendedness (Maddow-Zimet and Kost 2020), and thus was considered as a separate category. We recognize the limitations of the language of pregnancy “intentions” (Kost and Zolna 2019; Potter et al. 2019), and we interpret our results more broadly in the framework of reproductive justice by paying close attention to the association between reported intentions and which groups are most likely to experience births.

Individual-Level Controls

Individual-level demographic controls from the PRAMS data include maternal age, education, race and ethnicity, household income, marital status, insurance status at the time of conception, urbanicity, and birth order. Age was coded as a categorical variable split into levels of age 19 or younger, 20–24, 25–29, 30–34, and 35 or older. Educational attainment was categorized by years of schooling as less than a high school diploma (0–11 years), high school diploma (12 years), some college (13–15 years), and college degree or more (≥16 years). Race and ethnicity were coded as mutually exclusive categories of Hispanic, non-Hispanic White, non-Hispanic Black, and non-Hispanic other; this consolidation of racial categories follows precedent from Mark and Cowan (2022).

Across survey waves, household income was siloed into disparate ranges that could not be adequately consolidated. Instead, the median value for each range was taken, and from those values, income categories were created from the distribution of the median values that had been standardized to 2018 dollars using the Bureau of Labor Statistics Consumer Price Index Research Series (U.S. Bureau of Labor Statistics 2023; U.S. Census Bureau 2022). Six household income categories were created from the 10%, 25%, 50%, 75%, and 90% values of the distribution, which corresponded to the values of household income of $8,000 or less, $8,001–18,000, $18,001–35,630, $35,631–79,000, $79,001–85,296, and $85,297 or more, again standardized to 2018 dollars. For the 18,588 respondents in the analytic sample with missing data on household income (8.18%), an income category of “did not report income” was created.

Health insurance status at the time of pregnancy was included as an additional control variable, as previous research has shown an association between being uninsured and experiencing an unintended pregnancy (Geiger et al. 2021; Kost et al. 2012). The health insurance variable was based on insurance status in the month before pregnancy and includes five categories: private insurance from employer, private insurance paid for by someone else, Medicaid, other state-specific options (vary by state-specific questionnaire), or no insurance coverage. Of the respondents in the analytic sample, 52,362 (23.5%) were missing data on health insurance status. An additional category of “did not report health insurance status” was created. Urbanicity was a dichotomous variable of either urban or rural. Birth order was a categorical variable with values of first birth order and second or later birth order, and as was done for household income and health insurance status, an additional category was created for those who had a missing response for the birth order question in the PRAMS survey.

State-Level Controls

Three state-level control variables were included in the analysis. Two variables catalog average state-level attendance at religious services and were generated from the 2014 Pew Research Center Religious Landscape Study (Pew Research Center 2014). The first variable was the proportion of individuals in the state that self-report that they attend religious services weekly, and the second was the proportion of individuals that self-report that they attend religious services less than weekly but more often than seldom or never (i.e., once or twice a month or at least a few times a year). The third state-level control variable was the proportion of U.S. House of Representatives seats in that state-year held by Republican Party members, generated from data in the Library of Congress (Library of Congress 2022). Controlling for state-level religiosity and party affiliation of representatives serves as a proxy for state-level cultural norms and political climate that likely influence how individuals schematize their stance on abortion.

Policy Exposure Measure

The state-level abortion policy hostility index exposure measure was the primary independent variable of interest in the estimation model and was based on the legislative data received from the Guttmacher Institute. The measure is a continuous scale (0–10) of state-level abortion policy hostility, based on how many abortion restrictions a state had in place each year. The measure was compiled by the Guttmacher Institute using the 10 classifications of abortion restrictions noted earlier. The average number of state-level restrictions during the 2012–2018 study period is four (see column 4 of online appendix Table A1), with numbers ranging from zero in such states as Connecticut and New Jersey to 10 in Louisiana and Mississippi.

Analytic Approach

Our empirical model was based on the conceptualized individual- and state-level relationships with pregnancy intention among women with a pregnancy resulting in a live birth, as shown in online appendix Figure A1. The empirical analyses were conducted using multinomial logistic regression analysis. The model is of the following general form:
Yist=β0+β1×legislationst+β2×Ii+β3×Sst+β4×Tt+εist,
(1)

where Yist reflects the individual multinomial outcome of pregnancy intention for individual i experiencing a live birth living in state s in year t; legislationst reflects the continuous measure of the state-level abortion policy hostility index; Ii reflects the vector of individual-level demographic and socioeconomic status (SES) control measures; Sst reflects the state-level time-constant and time-varying control measures; and Tt reflects the year fixed effects. Lastly, all models included a random error term, εist.

Including state-level covariates in the model allows for estimates that account for potentially confounding time-constant and time-varying state-level characteristics, to help parse out the direct association between abortion policy hostility and pregnancy intention of women with a pregnancy resulting in live birth. We used controls for state-level covariates of religiosity and political leaning rather than including state-level fixed effects in our models because of the limited variation in state abortion policy in our sample over time (see online appendix Table A1, column 3). Year-level fixed effects were included to control for systematic variation in observed time units (2012–2018). The model was also estimated for samples stratified by race and ethnicity, age, educational attainment, insurance status, and urbanicity. We report the relative risk ratios (RRR)—the exponentiated multinomial logit coefficients—which indicate the risk of being in the group of women who have a live birth resulting from an unintended pregnancy, compared with the reference group of women whose live birth resulted from an intended pregnancy. We report the corresponding 95% confidence intervals (CI) for the RRRs. We also computed postestimation predicted probabilities of having a live birth resulting from an unintended pregnancy at each level of the state-level abortion policy hostility index. For our stratified analysis, nonlinear joint Wald tests were run on the main estimates of interest (association of abortion policy hostility with pregnancy intention) for each group within a given stratum to test for significant differences within strata. All analyses were conducted in Stata/SE version 17.0 (StataCorp 2021).

Results

Descriptive Statistics

Table 1 presents the descriptive statistics for the analytic sample. The majority (57.8%) of respondents reported having a live birth resulting from an intended pregnancy; yet more than a quarter (28.0%) of respondents experienced a live birth resulting from an unintended pregnancy, and a smaller minority (14.2%) reported ambivalent intention about their pregnancy. During the study period, the proportion of live births from intended pregnancies remained relatively stable, while the proportion from unintended pregnancies declined slightly and the proportion from ambivalent pregnancies increased slightly. These trends are in line with recent literature reporting declines in unintended pregnancy nationally over the last 15 years (Finer and Zolna 2016; Kost 2015; Kost et al. 2023; Tapales and Finer 2015).

The percentage of live births resulting from unintended pregnancies was not evenly distributed across the country (Figure 1). States with the highest levels of unintended birth were concentrated primarily in the South and Midwest regions, while the lowest levels were among states in the Northeast and West.

In terms of demographic and socioeconomic characteristics, the modal respondent in the sample was a 25- to 29-year-old married, non-Hispanic White, college-educated woman, living in an urban area, with a household income of $35,631–79,000 (in 2018 dollars), who was receiving health insurance from her job. Notwithstanding the rise in childbearing at older ages, more than three quarters of respondents in the sample were concentrated in the 20–34 age range, and most (62.0%) were married. Three fifths (60.8%) of the weighted analytic sample was non-Hispanic White, followed by almost equal proportions of Hispanic (15.8%) and non-Hispanic Black (13.8%) respondents. More than one third (35.6%) of women in the sample had a college degree, while another 27.6% had completed some college; women with only a high school diploma accounted for one quarter (24.0%) of respondents. Most respondents in our sample—85.1%—lived in urban areas.

Respondents relied on a range of health insurance options. Less than a third (31.6%) reported having health insurance from their job, although the next largest share (24.9%) did not disclose their health insurance status. Nearly a quarter of respondents—23.1%—were on Medicaid, while a sizable minority (14.5%) reported having no health insurance in the month before pregnancy. For the vast majority (92.6%) of respondents, this was their first live birth. In terms of the state-level controls, the respondents in our sample lived in states where, on average, 35% of the population attended religious services weekly and 33% attended less than weekly but more often than seldom or never. At the state-year level, respondents lived in states where, on average, about half (53.5%) of the elected U.S. Representatives were from the Republican Party.

Regression Results: Main Model

Table 2 reports the main result from the empirical multinomial logistic regression analysis. A one-unit increase in the state-level abortion policy hostility index was associated with a 1.02 times greater relative risk (odds) of having a live birth resulting from an unintended pregnancy relative to an intended pregnancy (RRR = 1.02; 95% CI = 1.01, 1.03). As state-level abortion policy hostility increases, the risk of a woman having a live birth resulting from an unintended pregnancy is predicted to increase. To help interpret the results, Figure 2 shows the predicted probabilities of having a live birth resulting from an unintended pregnancy at each level of the state-level abortion policy hostility index. In a state where no hostile legislation toward abortion was in place, the probability of having a live birth resulting from an unintended pregnancy was 26.7%; in a state where the abortion policy hostility index was at the maximum (10), the probability of having a live birth resulting from an unintended pregnancy was 30.1%. Thus, there was a 13% higher predicted probability of having a live birth resulting from an unintended pregnancy between a maximum-hostility state and a nonhostile state.

Table 2 also shows the full output for the multinomial logistic regression model for the demographic and SES covariates and state-level controls. Compared with women aged 25–29, younger women—especially those 19 or younger—were almost three times as likely to have a live birth resulting from an unintended pregnancy relative to an intended pregnancy (RRR = 2.94; CI = 2.71, 3.18), while older women (aged ≥35) were less likely (RRR = 0.85; CI = 0.80, 0.89), all else equal. Relative to non-Hispanic White women, women in all other racial and ethnic groups except Hispanics were also more likely to have a live birth resulting from an unintended versus intended pregnancy, ceteris paribus. These findings are generally consistent with past research (Finer and Zolna 2016; Guzzo and Hayford 2020; Sutton et al. 2019).

In terms of educational attainment, women at the tails of the spectrum (less than a high school diploma and college degree or more) were less likely than women with a high school diploma to have a live birth resulting from an unintended pregnancy relative to an intended pregnancy, ceteris paribus. Women with some college education were more likely than those with a high school diploma to have a live birth resulting from an unintended pregnancy relative to an intended one, all else equal. Additionally, the observed income effect was quite robust: relative to middle-income women and all else being equal, low-income women were more likely and high-income women were less likely to have a live birth resulting from an unintended pregnancy relative to an intended pregnancy. Our results thus provide additional evidence that less advantaged women are more likely to carry an unintended pregnancy to term than are their higher income counterparts, who more frequently report that their pregnancies (and subsequent births) were intended.

Additionally, relative to their married counterparts, unmarried women were almost two and a half times as likely to have a live birth resulting from an unintended pregnancy relative to an intended pregnancy, ceteris paribus (RRR = 2.42; CI = 2.33, 2.52). Urbanicity was not significantly associated with the likelihood of a live birth resulting from an unintended versus intended pregnancy (RRR = 0.98; CI = 0.94, 1.03), all else equal. Compared with women who had health insurance from their jobs in the month prior to their pregnancy, women with Medicaid coverage, other health insurance, or no health insurance were more likely to have a live birth resulting from an unintended pregnancy relative to an intended pregnancy. Finally, higher parity women (those having their second or higher order birth) were less likely than women having their first birth to have a live birth from an unintended pregnancy relative to an intended pregnancy.

We also find evidence that state characteristics—particularly residing in states with higher levels of religious service attendance and greater shares of Republican U.S. House Representatives—were associated with a greater likelihood of having a live birth resulting from an unintended versus intended pregnancy. While living in a state with high or moderate average weekly religious service attendance elevated the odds of having a live birth from an unintended (versus intended) pregnancy compared with states where religious service attendance was rare, this association was significant only for states where religious service attendance was high, all else equal (for high, RRR = 1.70; CI = 1.19, 2.42; for moderate, RRR = 1.40; CI = 0.88, 2.21). We also find that women in states with a greater proportion of Republican House representation had greater odds (RRR = 1.10; CI = 1.01, 1.20) of experiencing a live birth resulting from an unintended versus an intended pregnancy.

Overall, the observed patterns in sociodemographic associations are almost completely in accordance with prior research. Our results show that disadvantaged women—those who are younger, racial minorities, less educated, low-income, uninsured, or on Medicaid—were more likely to experience unintended pregnancy and childbearing. However, this model does not support an urban–rural difference or find significant differences between Hispanic and non-Hispanic White women (Finer and Zolna 2016; Guzzo and Hayford 2020; Kost et al. 2012; National Center for Health Statistics 2019; Sutton et al. 2019).

Stratified Models

We next stratified our models by demographic and SES characteristics, testing whether there are significant differences in the strength of the association between abortion policy hostility and unintended pregnancy outcomes for different groups within a sociodemographic stratum. Table 3 provides support for a moderation effect by age, educational attainment (proxy for SES), insurance status, and urbanicity, although not for race and ethnicity. These results underscore that not only are some sociodemographic groups more at risk for unintended childbearing as observed in the full model, but their risk is also differentially higher when faced with higher abortion policy hostility.

As shown in panel a of Figure 3, there was a prominent moderating effect by age between state-level abortion policy hostility and having a live birth resulting from an unintended relative to intended pregnancy, which diminished in magnitude with increasing age. Among women aged 19 or younger, a one-unit increase in state-level abortion policy hostility was associated with a 1.04 times greater relative risk of having a live birth resulting from an unintended pregnancy. This means that in a nonhostile state, young women's probability of having a live birth resulting from an unintended pregnancy was 55.1%, and in a maximum-hostility state, it was 60.0%.

Panels b and c of Figure 3 display the predicted probabilities of live birth resulting from unintended pregnancy at every level of the abortion policy hostility index by educational level and insurance status, respectively. Educational attainment—a proxy for SES—had a robust moderating effect. Women with lower levels of educational attainment—a high school diploma or less—had more robust associations between state-level abortion policy hostility and live birth resulting from unintended pregnancy than did women with some college or a college degree. In terms of insurance status, a one-unit increase in abortion policy hostility was most significantly associated with increases in live births resulting from unintended versus intended pregnancies for women on Medicaid and women without health insurance.

While there was not an observed urban–rural difference in the main model, a one-unit increase in abortion policy hostility was associated with significantly higher relative risk of having a live birth resulting from an unintended versus an intended pregnancy for rural women (RRR = 1.03; CI = 1.01, 1.05) compared with urban women (RRR = 1.02; CI = 1.01, 1.03; χ2 = 5.19; p = .02). Indeed, for rural women, as shown in panel d of Figure 3, there was a 17% increase in the probability of having a live birth resulting from an unintended pregnancy in a nonhostile state versus a highly hostile state (28.2% vs. 33.0%). Nonlinear joint Wald tests revealed that race and ethnicity had no significant moderating effect on the association between abortion policy hostility and live births resulting from unintended relative to intended pregnancies (χ2 = 3.35; p = .19). Figure A1 of the online appendix displays the predicted probabilities of the stratified models for racial and ethnic groups. Overall, the results from the stratified models highlight the importance of better understanding the very real heightened risk of unintended pregnancy and childbearing that particular marginalized groups may face regarding state-level abortion policy hostility.

Discussion

To our knowledge, this study is the first to investigate how state-level policy hostility toward abortion is associated with women's live births resulting from unintended versus intended pregnancies. Past research has examined the associations and causal effects of state-level abortion policy hostility on such reproductive health outcomes as abortion rates, maternal mortality, and distance to an abortion-providing facility (Austin and Harper 2019b; Bossick et al. 2021; Brown et al. 2020; Hawkins et al. 2020; Nash and Dreweke 2019). The literature (Aiken et al. 2022; Foster 2020; Gerdts et al. 2022; Haas-Wilson 1996; Upadhyay et al. 2013; White et al. 2019) suggests that state-level abortion policy hostility may have a robust relationship with live births resulting from unintended pregnancy, given documented relationships between hostile abortion policy and decreased access to and use of formal abortion services and increased requests for telemedicine self-managed abortion medication. Moreover, certain abortion restrictions, such as gestational age limits and restricted Medicaid funding, directly limit abortion access and reproductive autonomy and, thus, may in turn lead women to carry pregnancies to term that they otherwise would have aborted. Drawing from this research, we hypothesized that the state-level abortion policy hostility climate—constructed from the passage of restrictive abortion laws—would be positively associated with a higher likelihood of women's live births resulting from unintended pregnancy relative to intended pregnancy and that this relationship would be moderated (higher) for more disadvantaged women.

Our findings confirm our hypotheses. We find that a one-unit increase in state-level abortion policy hostility—that is, the passage of an abortion restriction in one additional category of the index—was associated with a 1.02 times as high relative risk of having a live birth resulting from an unintended versus intended pregnancy, after controlling for individual- and state-level covariates and year fixed effects. We show that this translates to a 13% higher predicted probability of a woman having a live birth resulting from an unintended pregnancy from living in a maximum-hostility state versus a zero-hostility state (30.1% vs. 26.7%), revealing that the state abortion policy climate where one lives is meaningfully associated with unintended pregnancy outcomes. Our findings are consistent with the results of Brown et al. (2020), who found that a highly restrictive state legislative climate was associated with a lower county-level abortion rate. Our findings—based on cross-sectional data stacked over time—reinforce the current scientific evidence that restrictive state-level abortion policy may limit women's access to reproductive health care, reproductive autonomy, and pregnancy options, and as a result, put women's well-being and health at risk (Brown et al. 2020; Foster 2020; Gerdts et al. 2022; Hawkins et al. 2020; Stevenson 2021; Upadhyay et al. 2021; White et al. 2019).

The results from the sociodemographic covariates in the full model and the test of moderation of these covariates in the stratified models are also largely in accordance with our hypotheses. Furthermore, these results build on and provide nuance for past descriptive research (Finer and Zolna 2016; Guzzo and Hayford 2020; Sutton et al. 2019), which found that disadvantaged women face higher rates of unintended pregnancy, as well as unplanned births. However, while White–Black disparities in reproductive health outcomes are well established (Kusunoki et al. 2016; Musick 2002; Sutton et al. 2019), and we find that non-Hispanic Black women were significantly more likely than non-Hispanic White women to have a live birth resulting from an unintended pregnancy, the results of our stratified models show that race and ethnicity alone did not significantly moderate the association between state-level policy hostility toward abortion and live birth resulting from unintended pregnancy. Our findings suggest that while non-Hispanic Black women are more vulnerable to the experience of unintended childbearing, they are not more vulnerable directly because of a heightened association with state-level abortion policy hostility.

Our findings most prominently point to an age effect and SES effect, as proxied by educational attainment and insurance status. That is, higher state-level policy hostility toward abortion was most strongly associated with live birth resulting from unintended pregnancy for younger women, women with lower levels of educational attainment, and women who had health insurance coverage prior to their pregnancy through Medicaid or who lacked insurance. In fact, the association between state-level abortion policy hostility and experiencing a live birth resulting from an unintended pregnancy among women aged 19 or younger was almost three times that of the average woman in the sample. We find that it is generally women in physically or financially resource-constrained groups for whom state-level abortion policy hostility correlates most strongly with live birth resulting from unintended pregnancy. These results are generally consistent with past research that found age, education, insurance status, and rural residence to be both strong correlates of unplanned childbearing and factors that make women more vulnerable to abortion restrictions (Guzzo and Hayford 2020; Haas-Wilson 1996; Musick 2002; Musick et al. 2009; Sutton et al. 2019; Upadhyay et al. 2021).

Many of the abortion restrictions that compose our hostility index are specifically focused on making it more practically or logistically difficult to obtain an abortion. That is especially the case for women with few resources to travel far or pay high out-of-pocket costs for the procedure. Such restrictions include the requirement of in-person counseling (requiring two separate trips to the facility), parental consent or notification laws, restricting abortion coverage in insurance, medication abortion restrictions, and TRAP laws. This might explain why we find significant moderation effects for sociodemographic variables that are more closely tied to practical or tangible resource availability. These variables include being young, having low educational attainment, being on Medicaid, being uninsured, and living in a rural area. Overcoming the large logistical and practical barriers created by the abortion restrictions that contribute to a hostile policy climate may be greater for younger, low-SES, and rural women, who are more physically resource-constrained. The pattern of these moderation effects may also emerge because there is a very direct link between the abortion restrictions examined in this study and the sociodemographic indicators of practical disadvantage—such as parental consent laws and age, or Medicaid abortion restrictions and insurance status—compared with other overlapping spheres of oppression experienced by racial minority women.

Limitations

One prominent limitation of our study is that we could only observe outcomes for women who carried their pregnancies to term. Future research should seek to replicate this study design, or implement one like it, with data from pregnant people regardless of their pregnancy outcome (e.g., Sutton et al 2019), as our study suggests the need to further explore how state-level policy hostility toward abortion is associated with all pregnancy outcomes. Our findings are also likely conservative estimates because of data availability and release restrictions from the PRAMS data. As shown in Figure 2 and online appendix Table A1, we do not have data from nine states in the sample. The five states that are excluded are those with high levels of policy hostility toward abortion or states that experienced increases in hostility during the study period; thus, their exclusion may have limited our ability to identify the full impact of abortion restrictions on unintended pregnancy outcomes. We also relied on survey data for our individual-level data, and our main outcome of interest—pregnancy intention—is asked not at the time of conception or pregnancy, but after a live birth; thus, our survey data are susceptible to hindsight and recall bias. These data are also self-reported, so may be subject to measurement and reporting error as survey participants may be hesitant to report negative feelings they experienced about the pregnancy. Finally, we had access only to the aggregated policy data that show the total number of categories of restrictions a state had in place each year, leaving us unable to decompose the abortion policy hostility index by type of policy.

Conclusion

The landscape of abortion policy and access in the United States is rapidly changing, and heterogeneity in state-level abortion policy has only increased given the Dobbs ruling (Jimenez 2022; Sharfstein 2023). Decomposing the index of abortion policy hostility and testing specific parameters for their relation to pregnancy intentions—as well as the extent to which they differ by sociodemographic subgroups—may provide additional insight into the policies driving the association between state-level hostility and live births resulting from unintended pregnancy. Such an approach could help clarify which policies may more robustly impact women who are already at a disproportionately high risk for unintended pregnancy and maternal mortality. Indeed, this future work would build on our important finding that not only are some vulnerable groups more at risk for having a live birth resulting from an unintended pregnancy, but their risk is also differentially higher when faced with a more hostile state-level abortion policy climate. Overall, this research reinforces the finding that the cumulative effect of a hostile policy climate in and of itself is robustly associated with increases in both negative and medically dangerous reproductive health outcomes, including—as this study has found—increases in live births resulting from unintended pregnancies.

Acknowledgments

The authors thank the PRAMS Working Group and the CDC for access to the PRAMS dataset. We thank Ms. Elizabeth Nash and the Guttmacher Institute for access to their legislative database. We also thank Dr. Laura Lindberg for reviewing a draft of this paper, as well as the three anonymous reviewers for their feedback. A Summer Undergraduate Research Grant from the Cornell University College of Human Ecology supported Julia Eddelbuettel in this work during the summer of 2021.

Notes

1

Throughout this article we use the term “women” to be consistent with both the cited literature and our data source (the Pregnancy Risk Assessment Monitoring System survey). We recognize that not all pregnant people or people capable of becoming pregnant identify as female, and that a limitation of this article is that we cannot capture the experiences of all pregnant people, including transgender or gender nonbinary individuals.

2

Missing data resulted in Vermont being dropped from the analytic sample.

References

Aiken, A. R. A., Starling, J. E., Scott, J. G., & Gomperts, R. (
2022
).
Association of Texas Senate Bill 8 with requests for self-managed medication abortion
.
JAMA Network Open
,
5
,
e221122
. https://doi.org/10.1001/jamanetworkopen.2022.1122
Austin, N., & Harper, S. (
2018
).
Assessing the impact of TRAP laws on abortion and women's health in the USA: A systematic review
.
BMJ Sexual & Reproductive Health
,
44
,
128
134
.
Austin, N., & Harper, S. (
2019a
).
Constructing a longitudinal database of targeted regulation of abortion providers laws
.
Health Services Research
,
54
,
1084
1089
.
Austin, N., & Harper, S. (
2019b
).
Quantifying the impact of targeted regulation of abortion provider laws on U.S. abortion rates: A multi-state assessment
.
Contraception
,
100
,
374
379
.
Bearak, J. M., Popinchalk, A., Beavin, C., Ganatra, B., Moller, A.-B., Tunçalp, Ö., & Alkema, L. (
2022
).
Country-specific estimates of unintended pregnancy and abortion incidence: A global comparative analysis of levels in 2015–2019
.
BMJ Global Health
,
7
,
e007151
. https://doi.org/10.1136/bmjgh-2021-007151
Biggs, M. A., Upadhyay, U. D., McCulloch, C. E., & Foster, D. G. (
2017
).
Women's mental health and well-being 5 years after receiving or being denied an abortion: A prospective, longitudinal cohort study
.
JAMA Psychiatry
,
74
,
169
178
.
Bossick, A. S., Brown, J., Hanna, A., Parrish, C., Williams, E. C., & Katon, J. G. (
2021
).
Impact of state-level reproductive health legislation on access to and use of reproductive health services and reproductive health outcomes: A systematic scoping review in the Affordable Care Act era
.
Women's Health Issues
,
31
,
114
121
.
Brown, B. P., Hebert, L. E., Gilliam, M., & Kaestner, R. (
2020
).
Association of highly restrictive state abortion policies with abortion rates, 2000–2014
.
JAMA Network Open
,
3
,
e2024610
. https://doi.org/10.1001/jamanetworkopen.2020.24610
Centers for Disease Control and Prevention
. (
2023
,
March
28
).
PRAMS
. Retrieved from https://www.cdc.gov/prams/index.htm
Dobbs v. Jackson Women's Health Organization, 19 U.S. 1392
(
2022
). Retrieved from https://www.supremecourt.gov/opinions/21pdf/19-1392_6j37.pdf
Finer, L. B., & Zolna, M. R. (
2016
).
Declines in unintended pregnancy in the United States, 2008–2011
.
New England Journal of Medicine
,
374
,
843
852
.
Foster, D. G. (
2020
).
The Turnaway Study: Ten years, a thousand women, and the consequences of having—or being denied—an abortion
.
New York, NY
:
Scribner
.
Geiger, C. K., Sommers, B. D., Hawkins, S. S., & Cohen, J. L. (
2021
).
Medicaid expansions, preconception insurance, and unintended pregnancy among new parents
.
Health Services Research
,
56
,
691
701
.
Gerdts, C., Fuentes, L., Grossman, D., White, K., Keefe-Oates, B., Baum, S. E., . . . Potter, J. E. (
2022
).
Impact of clinic closures on women obtaining abortion services after implementation of a restrictive law in Texas
.
American Journal of Public Health
,
112
,
1297
1304
.
Guttmacher Institute
. (
2020
,
January
22
).
Targeted regulation of abortion providers (TRAP) laws
. Retrieved from https://www.guttmacher.org/node/28379/printable/print
Guzzo, K. B., & Hayford, S. R. (
2020
).
Pathways to parenthood in social and family contexts: Decade in review, 2020
.
Journal of Marriage and Family
,
82
,
117
144
.
Haas-Wilson, D. (
1996
).
The impact of state abortion restrictions on minors' demand for abortions
.
Journal of Human Resources
,
31
,
140
158
.
Hawkins, S. S., Ghiani, M., Harper, S., Baum, C. F., & Kaufman, J. S. (
2020
).
Impact of state-level changes on maternal mortality: A population-based, quasi-experimental study
.
American Journal of Preventive Medicine
,
58
,
165
174
.
Jimenez, L. (
2022
,
August
24
).
60 days after Dobbs: State legal developments on abortion
.
Brennan Center for Justice
. Retrieved from https://www.brennancenter.org/our-work/research-reports/60-days-after-dobbs-state-legal-developments-abortion
Jones, R. K., & Jerman, J. (
2017
).
Population group abortion rates and lifetime incidence of abortion: United States, 2008–2014
.
American Journal of Public Health
,
107
,
1904
1909
.
Jones, R. K., Philbin, J., Kirstein, M., Nash, E., & Lufkin, K. (
2022
,
June
15
).
Long-term decline in U.S. abortions reverses, showing rising need for abortion as Supreme Court is poised to overturn Roe v. Wade
.
Guttmacher Institute
. Retrieved from https://www.guttmacher.org/article/2022/06/long-term-decline-us-abortions-reverses-showing-rising-need-abortion-supreme-court
Kortsmit, K., Jatlaoui, T. C., Mandel, M. G., Reeves, J. A., Oduyebo, T., Petersen, E., & Whiteman, M. K. (
2020
).
Abortion surveillance—United States, 2018
(MMWR Surveillance Summaries, Vol.
69
No.
SS-7
). https://doi.org/10.15585/mmwr.ss6907a1
Kost, K. (
2015
).
Unintended pregnancy rates at the state level: Estimates for 2010 and trends since 2002
(
Guttmacher Institute report
). Retrieved from https://www.guttmacher.org/report/unintended-pregnancy-rates-state-level-estimates-2010-and-trends-2002
Kost, K., Finer, L. B., & Singh, S. (
2012
).
Variation in state unintended pregnancy rates in the United States
.
Perspectives on Sexual and Reproductive Health
,
44
,
57
64
.
Kost, K., & Lindberg, L. (
2015
).
Pregnancy intentions, maternal behaviors, and infant health: Investigating relationships with new measures and propensity score analysis
.
Demography
,
52
,
83
111
.
Kost, K., & Zolna, M. (
2019
).
Challenging unintended pregnancy as an indicator of reproductive autonomy: A response
.
Contraception
,
100
,
5
9
.
Kost, K., Zolna, M., & Murro, R. (
2023
).
Pregnancies in the United States by desire for pregnancy: Estimates for 2009, 2011, 2013, and 2015
.
Demography
,
60
,
837
863
. https://doi.org/10.1215/00703370-10690005
Kusunoki, Y., Barber, J. S., Ela, E. J., & Bucek, A. (
2016
).
Black–White differences in sex and contraceptive use among young women
.
Demography
,
53
,
1399
1428
.
Library of Congress
. (
2022
,
January
2
).
Current members of the U.S. Congress
. Retrieved from https://www.congress.gov
Littlejohn, K. E. (
2021
).
Just get on the pill: The uneven burden of reproductive politics.
Oakland
:
University of California Press
.
Maddow-Zimet, I., & Kost, K. (
2020
).
Effect of changes in response options on reported pregnancy intentions: A natural experiment in the United States
.
Public Health Reports
,
135
,
354
363
.
Mark, N. D. E., & Cowan, S. K. (
2022
).
Do pregnancy intentions matter? A research note revisiting relationships among pregnancy, birth, and maternal outcomes
.
Demography
,
59
,
37
49
. https://doi.org/10.1215/00703370-9710311
Miller, S., Wherry, L. R., & Greene Foster, D. (
2023
).
The economic consequences of being denied an abortion
.
American Economic Journal: Economic Policy
,
15
(
1
),
394
437
.
Musick, K. (
2002
).
Planned and unplanned childbearing among unmarried women
.
Journal of Marriage and Family
,
64
,
914
929
.
Musick, K., England, P., Edgington, S., & Kangas, N. (
2009
).
Education differences in intended and unintended fertility
.
Social Forces
,
88
,
543
572
.
Nash, E. (
2019
,
August
29
).
State abortion policy landscape: From hostile to supportive
.
Guttmacher Institute
. Retrieved from https://www.guttmacher.org/article/2019/08/state-abortion-policy-landscape-hostile-supportive
Nash, E. (
2021
,
October
4
).
For the first time ever, U.S. states enacted more than 100 abortion restrictions in a single year
.
Guttmacher Institute
. Retrieved from https://www.guttmacher.org/article/2021/10/first-time-ever-us-states-enacted-more-100-abortion-restrictions-single-year
Nash, E., & Dreweke, J. (
2019
).
The U.S. abortion rate continues to drop: Once again, state abortion restrictions are not the main driver
.
Guttmacher Policy Review
,
22
,
41
48
.
National Center for Health Statistics
. (
2019
,
November
8
).
Key statistics from the National Survey of Family Growth—I listing
.
Centers for Disease Control and Prevention
. Retrieved from https://www.cdc.gov/nchs/nsfg/key_statistics/i.htm
Pew Research Center
. (
2014
).
Pew Research Center 2014 U.S. Religious Landscape Study
. Retrieved from https://www.pewresearch.org/religion/dataset/pew-research-center-2014-u-s-religious-landscape-study/
Potter, J. E., Stevenson, A. J., Coleman-Minahan, K., Hopkins, K., White, K., Baum, S. E., & Grossman, D. (
2019
).
Challenging unintended pregnancy as an indicator of reproductive autonomy
.
Contraception
,
100
,
1
4
.
Price, K. (
2020
).
What is reproductive justice?: How women of color activists are redefining the pro-choice paradigm
.
Meridians
,
19
(
S1
),
340
362
.
Sharfstein, J. (
2023
,
January
4
).
Abortion access after the Dobbs decision
.
Johns Hopkins Bloomberg School of Public Health
. Retrieved from https://publichealth.jhu.edu/2023/abortion-access-after-dobbs
Shulman, H. B., D'Angelo, D. V., Harrison, L., Smith, R. A., & Warner, L. (
2018
).
The Pregnancy Risk Assessment Monitoring System (PRAMS): Overview of design and methodology
.
American Journal of Public Health
,
108
,
1305
1313
.
Smith, A. (
2005
).
Beyond pro-choice versus pro-life: Women of color and reproductive justice
.
NWSA Journal
,
17
(
1
),
119
140
.
Sonfield, A., Tapales, A., Jones, R. K., & Finer, L. B. (
2015
).
Impact of the federal contraceptive coverage guarantee on out-of-pocket payments for contraceptives: 2014 update
.
Contraception
,
91
,
44
48
.
StataCorp
. (
2021
).
Stata statistical software: Release 17
[Computer software].
College Station, TX
:
StataCorp LLC
.
Stevenson, A. J. (
2021
).
The pregnancy-related mortality impact of a total abortion ban in the United States: A research note on increased deaths due to remaining pregnant
.
Demography
,
58
,
2019
2028
.
Sutton, A., Lichter, D. T., & Sassler, S. (
2019
).
Rural–urban disparities in pregnancy intentions, births, and abortions among U.S. adolescent and young women, 1995–2017
.
American Journal of Public Health
,
109
,
1762
1769
.
Tapales, A., & Finer, L. (
2015
).
Unintended pregnancy and the changing demography of American women, 1987–2008
.
Demographic Research
,
33
,
1257
1270
. https://doi.org/10.4054/DemRes.2015.33.45
Trussell, J., Henry, N., Hassan, F., Prezioso, A., Law, A., & Filonenko, A. (
2013
).
Burden of unintended pregnancy in the United States: Potential savings with increased use of long-acting reversible contraception
.
Contraception
,
87
,
154
161
.
Upadhyay, U. D., McCook, A. A., Bennett, A. H., Cartwright, A. F., & Roberts, S. C. M. (
2021
).
State abortion policies and Medicaid coverage of abortion are associated with pregnancy outcomes among individuals seeking abortion recruited using Google Ads: A national cohort study
.
Social Science & Medicine
,
274
,
113747
. https://doi.org/10.1016/j.socscimed.2021.113747
Upadhyay, U. D., Weitz, T. A., Jones, R. K., Barar, R. E., & Foster, D. G. (
2013
).
Denial of abortion because of provider gestational age limits in the United States
.
American Journal of Public Health
,
104
,
1687
1694
.
U.S. Bureau of Labor Statistics
. (
2023
,
March
20
).
Consumer price index: R-CPI-U-RS homepage
. Retrieved from https://www.bls.gov/cpi/research-series/r-cpi-u-rs-home.htm
U.S. Census Bureau
. (
2022
,
September
12
).
Current versus constant (or real) dollars
. Retrieved from https://www.census.gov/topics/income-poverty/income/guidance/current-vs-constant-dollars.html
White, K., Baum, S. E., Hopkins, K., Potter, J. E., & Grossman, D. (
2019
).
Change in second-trimester abortion after implementation of a restrictive state law
.
Obstetrics & Gynecology
,
133
,
771
779
.
This is an open access article distributed under the terms of a Creative Commons license (CC BY-NC-ND 4.0).

Supplementary data