Advertisement

Racial and Ethnic Abortion Disparities Following Georgia's 22-Week Gestational Age Limit

  • Elizabeth A. Mosley
    Correspondence
    Correspondence to: Elizabeth A. Mosley, PhD, MPH, Georgia State University School of Public Health 140 Decatur St. SE Atlanta, GA 30303. Phone: 470-329-8153; fax: (404) 413-1489.
    Affiliations
    Georgia State University, School of Public Health, Mark Chaffin Center for Healthy Development, Atlanta, Georgia

    Emory University, Rollins School of Public Health, Atlanta, Georgia

    Center for Reproductive Health Research in the Southeast (RISE), Atlanta, Georgia
    Search for articles by this author
  • Sara K. Redd
    Affiliations
    Emory University, Rollins School of Public Health, Atlanta, Georgia

    Center for Reproductive Health Research in the Southeast (RISE), Atlanta, Georgia
    Search for articles by this author
  • Sophie A. Hartwig
    Affiliations
    Emory University, Rollins School of Public Health, Atlanta, Georgia

    Center for Reproductive Health Research in the Southeast (RISE), Atlanta, Georgia
    Search for articles by this author
  • Subasri Narasimhan
    Affiliations
    Emory University, Rollins School of Public Health, Atlanta, Georgia

    Center for Reproductive Health Research in the Southeast (RISE), Atlanta, Georgia
    Search for articles by this author
  • Emily Lemon
    Affiliations
    Emory University, Rollins School of Public Health, Atlanta, Georgia

    Center for Reproductive Health Research in the Southeast (RISE), Atlanta, Georgia
    Search for articles by this author
  • Author Footnotes
    1 Present address: Planned Parenthood of the Great Northwest and Hawaiian Islands, 2001 E Madison St, Seattle, WA, 98122.
    Erin Berry
    Footnotes
    1 Present address: Planned Parenthood of the Great Northwest and Hawaiian Islands, 2001 E Madison St, Seattle, WA, 98122.
    Affiliations
    Center for Reproductive Health Research in the Southeast (RISE), Atlanta, Georgia

    Emory University, School of Medicine, Department of Gynecology and Obstetrics, Atlanta, Georgia
    Search for articles by this author
  • Author Footnotes
    2 Present address: PSI, 1120 19th Street, NW, Washington, DC, 20036.
    Eva Lathrop
    Footnotes
    2 Present address: PSI, 1120 19th Street, NW, Washington, DC, 20036.
    Affiliations
    Emory University, Rollins School of Public Health, Atlanta, Georgia

    Center for Reproductive Health Research in the Southeast (RISE), Atlanta, Georgia

    Emory University, School of Medicine, Department of Gynecology and Obstetrics, Atlanta, Georgia
    Search for articles by this author
  • Author Footnotes
    3 Present address: Population Council Center, 1230 York Ave, New York, NY 10065.
    Lisa Haddad
    Footnotes
    3 Present address: Population Council Center, 1230 York Ave, New York, NY 10065.
    Affiliations
    Center for Reproductive Health Research in the Southeast (RISE), Atlanta, Georgia

    Emory University, School of Medicine, Department of Gynecology and Obstetrics, Atlanta, Georgia
    Search for articles by this author
  • Roger Rochat
    Affiliations
    Emory University, Rollins School of Public Health, Atlanta, Georgia

    Center for Reproductive Health Research in the Southeast (RISE), Atlanta, Georgia

    Emory University, School of Medicine, Department of Gynecology and Obstetrics, Atlanta, Georgia
    Search for articles by this author
  • Carrie Cwiak
    Affiliations
    Emory University, Rollins School of Public Health, Atlanta, Georgia

    Center for Reproductive Health Research in the Southeast (RISE), Atlanta, Georgia

    Emory University, School of Medicine, Department of Gynecology and Obstetrics, Atlanta, Georgia
    Search for articles by this author
  • Author Footnotes
    4 Present address: Columbia University Mailman School of Public Health, 722 W 168th St, New York, NY 10032.
    Kelli Stidham Hall
    Footnotes
    4 Present address: Columbia University Mailman School of Public Health, 722 W 168th St, New York, NY 10032.
    Affiliations
    Emory University, Rollins School of Public Health, Atlanta, Georgia

    Center for Reproductive Health Research in the Southeast (RISE), Atlanta, Georgia
    Search for articles by this author
  • Author Footnotes
    1 Present address: Planned Parenthood of the Great Northwest and Hawaiian Islands, 2001 E Madison St, Seattle, WA, 98122.
    2 Present address: PSI, 1120 19th Street, NW, Washington, DC, 20036.
    3 Present address: Population Council Center, 1230 York Ave, New York, NY 10065.
    4 Present address: Columbia University Mailman School of Public Health, 722 W 168th St, New York, NY 10032.
Open AccessPublished:October 25, 2021DOI:https://doi.org/10.1016/j.whi.2021.09.005

      Abstract

      Introduction

      Georgia's 2012 House Bill 954 (HB954) prohibiting abortions after 22 weeks from last menstrual period (LMP) has been associated with a significant decrease in abortions after 22 weeks. However, the policy's effects by race or ethnicity remain unexplored. We investigated whether changes in abortion numbers and ratios (per 1,000 live births) in Georgia after HB954 varied by race or ethnicity.

      Methods

      Using Georgia Department of Public Health induced terminations of pregnancy data from 2007 to 2017, we examined changes in number of abortions and abortion ratios (per 1,000 live births) by race and ethnicity following HB954 implementation.

      Results

      After full implementation of HB954 in 2015, the number of abortions and abortion ratios at or after 22 weeks (from last menstrual period) decreased among White (bNumber = –261.83, p < .001; bRatio = –3.31, p < .001), Black (bNumber = –416.17, p < .001; bRatio = –8.84, p < .001), non-Hispanic (bNumber = –667.00, p = .001; bRatio = –5.82, p < .001), and Hispanic (bNumber = –56.25, p = .002; bRatio = –2.44, p = .002) people. However, the ratio of abortions before 22 weeks increased for Black people (bLessThan22Weeks = 44.06, p = .028) and remained stable for White (bLessThan22Weeks = –6.78, p = .433), Hispanic (bLessThan22Weeks = 21.27, p = .212), and non-Hispanic people (bLessThan22Weeks = 26.93, p = .172).

      Conclusion

      The full implementation of HB954 had differential effects by race/ethnicity and gestational age. Although abortion at 22 weeks or more decreased for all groups, abortion at less than 22 weeks increased among Black people. Additional research should elucidate the possible causes, consequences, and reactions to differential effects of abortion restrictions by race and ethnicity.
      Access to safe and legal abortion services is affected by restrictive state and federal policies (
      • Jones R.K.
      • Jerman J.
      Abortion incidence and service availability in the United States, 2014.
      ;
      • Upadhyay U.D.
      • Weitz T.A.
      • Jones R.K.
      • Barar R.E.
      • Foster D.G.
      Denial of abortion because of provider gestational age limits in the United States.
      ), which can carry disproportionate and inequitable consequences for communities of color (
      • Boonstra H.D.
      Abortion in the lives of women struggling financially: Why insurance coverage matters.
      ;
      • Coles M.S.
      • Makino K.K.
      • Stanwood N.L.
      • Dozier A.
      • Klein J.D.
      How are restrictive abortion statutes associated with unintended teen birth?.
      ;
      • Cook P.J.
      • Parnell A.M.
      • Moore M.J.
      • Pagnini D.
      The effects of short-term variation in abortion funding on pregnancy outcomes [Working Paper].
      ;
      • Upadhyay U.D.
      • Johns N.E.
      • Cartwright A.F.
      • Franklin T.E.
      Sociodemographic characteristics of women able to obtain medication abortion before and after Ohio’s law requiring use of the Food and Drug Administration protocol.
      ). To date, a growing body of evidence has demonstrated associations between the rapid rise in abortion restrictions across the United States—especially in the South (
      • Nash E.
      State abortion policy landscape: From hostile to supportive.
      )—and fewer abortion providers (
      • Jones R.K.
      • Jerman J.
      Abortion incidence and service availability in the United States, 2014.
      ;
      • Jones R.K.
      • Witwer E.
      • Jerman J.
      Abortion incidence and service availability in the United States, 2017.
      ), longer distances to abortion clinics (
      • Gerdts C.
      • Fuentes L.
      • Grossman D.
      • White K.
      • Keefe-Oates B.
      • Baum S.E.
      • Potter J.E.
      Impact of clinic closures on women obtaining abortion services after implementation of a restrictive law in Texas.
      ), and increased demand for and use of self-managed abortion techniques (
      • Aiken A.R.A.
      • Broussard K.
      • Johnson D.M.
      • Padron E.
      Motivations and Experiences of People Seeking Medication Abortion Online in the United States..
      ,
      • Aiken A.R.A.
      • Starling J.E.
      • van der Wal A.
      • van der Vliet S.
      • Broussard K.
      • Johnson D.M.
      • Scott J.G.
      Demand for self-managed medication abortion through an online telemedicine service in the United States.
      ;
      • Grossman D.
      • White K.
      • Fuentes L.
      • Hopkins K.
      • Stevenson A.
      • Yeatman S.
      • Potter J.E.
      Knowledge, opinion, and experience related to abortion self-induction in Texas. Texas Policy Evaluation Project.
      ). Several studies have highlighted that women of color seem disproportionately impacted by abortion restrictions (
      • Coles M.S.
      • Makino K.K.
      • Stanwood N.L.
      • Dozier A.
      • Klein J.D.
      How are restrictive abortion statutes associated with unintended teen birth?.
      ;
      • Cook P.J.
      • Parnell A.M.
      • Moore M.J.
      • Pagnini D.
      The effects of short-term variation in abortion funding on pregnancy outcomes [Working Paper].
      ;
      • Upadhyay U.D.
      • Johns N.E.
      • Cartwright A.F.
      • Franklin T.E.
      Sociodemographic characteristics of women able to obtain medication abortion before and after Ohio’s law requiring use of the Food and Drug Administration protocol.
      ) owing to multiple, intersecting layers of inequity (
      • Crenshaw K.
      Demarginalizing the Intersection of Race and Sex: A Black feminist critique of antidiscrimination doctrine, feminist theory and antiracist politics.
      ) that put women of color at increased risk of 1) needing an abortion and 2) disproportionately experiencing adverse effects of restrictions on how and when they can access abortion care. Yet more evidence is needed to ascertain how abortion policies might affect racial/ethnic disparities in access to and utilization of abortion services, particularly in the South.
      Previous research suggests that Black and Latinx people
      In this manuscript, we use the term “Latinx” when possible to describe the ethnicity of individuals from Latin America, and “people” when describing abortion patients of all genders. “Latinx” is a term that encompasses Latino, Latina, and Hispanic, but is more inclusive of gender diversity. We use the terms “Hispanic” and “women” when citing studies or data that specifically used that language.
      1In this manuscript, we use the term “Latinx” when possible to describe the ethnicity of individuals from Latin America, and “people” when describing abortion patients of all genders. “Latinx” is a term that encompasses Latino, Latina, and Hispanic, but is more inclusive of gender diversity. We use the terms “Hispanic” and “women” when citing studies or data that specifically used that language.
      in the United States are more likely to live in states where abortion is restricted and inaccessible (
      Guttmacher Institute
      State bans on abortion throughout pregnancy. State laws and policies.
      ;
      • Jones R.K.
      • Jerman J.
      Abortion incidence and service availability in the United States, 2014.
      ;
      • Nash E.
      State abortion policy landscape: From hostile to supportive.
      ), experience significant and unique barriers to abortion access (
      • Mosley E.A.
      • Ayala S.
      • Jah Z.
      • Hailstorks T.
      • Scales M.
      • Gutierrez M.
      • Hall K.S.
      ‘All the things that make life harder’: Intersectionality and medication abortion in Georgia.
      ), and have abortions at later gestational ages (
      • Jones R.K.
      • Finer L.B.
      Who has second-trimester abortions in the United States?.
      ) compared with White people. In 2020, 43% of women aged 13–44 years lived in states considered hostile to abortion compared with 7% in 2000 (
      • Nash E.
      State abortion policy landscape: From hostile to supportive.
      ), and three-quarters of states in the South are considered hostile or very hostile to abortion (
      • Nash E.
      State abortion policy landscape: From hostile to supportive.
      ). Black and Hispanic people are more likely to live in states described as hostile or very hostile to abortion (
      • Frey W.H.
      Six maps that reveal America’s expanding diversity.
      ). Further, historical and ongoing reproductive injustices create a legacy of racialized barriers to abortion care. Historically, Black and Latinx women have been targeted and sterilized by unsafe and/or coercive family planning, and, because structural racism makes them more likely to live in poverty, Black and Latinx women seeking abortion services are more likely than White women to do so out of economic necessity (
      • Roberts D.
      Killing the Black body: Race, reproduction, and the meaning of liberty.
      ;
      • Schoen J.
      Choice & coercion: Birth control, sterilization, and abortion in public health and welfare.
      ;
      • Stern A.M.
      Sterilized in the name of public health: Race, immigration, and reproductive control in modern California.
      ). Qualitative research in Georgia has shown how Black and Latinx people face intersectional barriers to abortion care, including past experiences of disrespectful care leading to mistrust, lack of appropriate interpretation services, and risk of deportation, in addition to poverty-related factors such as inability to afford abortion services, lack of transportation, and lack of insurance (
      • Mosley E.A.
      • Ayala S.
      • Jah Z.
      • Hailstorks T.
      • Scales M.
      • Gutierrez M.
      • Hall K.S.
      ‘All the things that make life harder’: Intersectionality and medication abortion in Georgia.
      ). Given this burden, it is plausible that policies limiting access to abortion by gestational age—a common abortion restriction—carry disproportionate consequences for Black and Latinx people.
      The state of Georgia presents a unique case study to better understand racial/ethnic differences in effects of abortion policies, particularly gestational age limits. Historically, abortion access has been greater in Georgia than in surrounding states (
      • Roberts S.
      • Gould H.
      • Upadhyay U.D.
      Implications of Georgia’s 20-week abortion ban.
      ). Research indicates that a substantial proportion of individuals seeking abortion services in Georgia travel to Georgia from out of state (
      • Hall K.S.
      • Redd S.
      • Narasimhan S.
      • Mosley E.A.
      • Hartwig S.A.
      • Lemon E.
      • Cwiak C.
      Abortion trends from 2007 to 2017 following enactment of Georgia’s 22-week gestational age limit.
      ;
      • Roberts S.
      • Gould H.
      • Upadhyay U.D.
      Implications of Georgia’s 20-week abortion ban.
      ), suggesting that restrictive abortion policies enacted in Georgia have implications for access to care expanding to residents of other states. Indeed, both Hall et al. and Roberts et al. note that Georgia's recently implemented 22-week gestational age limit policy (at or after 22 weeks from last menstrual period [LMP]) decreased access to abortion care for out-of-state residents and they emphasize the detrimental implications of this policy for abortion care in the Southeast (
      • Hall K.S.
      • Redd S.
      • Narasimhan S.
      • Mosley E.A.
      • Hartwig S.A.
      • Lemon E.
      • Cwiak C.
      Abortion trends from 2007 to 2017 following enactment of Georgia’s 22-week gestational age limit.
      ;
      • Roberts S.
      • Gould H.
      • Upadhyay U.D.
      Implications of Georgia’s 20-week abortion ban.
      ). Additionally, Georgia's population is racially and ethnically diverse (
      U.S. Census Bureau
      United States Census: Explore data.
      ). In 2010, 32% of the state's 9.7 million residents were Black and 9% were Hispanic/Latino (
      U.S. Census Bureau
      United States Census: Explore data.
      ). The state also has widespread racial/ethnic disparities in unintended pregnancy (
      Healthy Mothers, Healthy Babies Coalition of Georgia
      State of the state report. State of the state report.
      ), and maternal (
      Centers for Disease Control and Prevention
      Pregnancy Mortality Surveillance System. Reproductive Health.
      ) and infant mortality (
      Georgia Department of Public Health
      Reducing infant mortality Georgia 2013 annual report.
      ). In 2012, the Georgia Assembly passed House Bill 954 (HB954) prohibiting abortion at 20 weeks after fertilization (equivalent to 22 weeks since LMP). The bill included exceptions to save the pregnant person's life or to avoid their permanent physical impairment (

      HB954. House Bill 954; General Assembly 2011-2012 Ed 2012.

      ). HB954 was signed into law in April 2012, but quickly enjoined following a legal challenge by the American Civil Liberties Union (Figure 1). The law was then partially implemented in January 2013, allowing abortions up to 24 weeks' gestation from LMP, and it was fully implemented in October 2015, allowing abortions up to 22 weeks' gestation from LMP.
      Figure thumbnail gr1
      Figure 1Timeline of Georgia’s 22-Week Gestational Age Limit on Abortion (HB954).
      Two previous studies examined the implications of Georgia's 22-week gestational age limit for abortion access in the state (
      • Hall K.S.
      • Redd S.
      • Narasimhan S.
      • Mosley E.A.
      • Hartwig S.A.
      • Lemon E.
      • Cwiak C.
      Abortion trends from 2007 to 2017 following enactment of Georgia’s 22-week gestational age limit.
      ;
      • Roberts S.
      • Gould H.
      • Upadhyay U.D.
      Implications of Georgia’s 20-week abortion ban.
      ). Roberts et al. analyzed data from 2012 to 2013 on abortions after 20 weeks' gestation to estimate the potential implications of the new policy. They found that 55% of patients were Black, 7% were Hispanic, and 55% had a high school education or less (
      • Roberts S.
      • Gould H.
      • Upadhyay U.D.
      Implications of Georgia’s 20-week abortion ban.
      ), suggesting that Black people and those of a lower socioeconomic status would be affected disproportionately by HB954, because social determinants of health—including poorer access to clinics and needing more time to gather funds for abortion procedures—push their abortion care later into pregnancy. More recently,
      • Hall K.S.
      • Redd S.
      • Narasimhan S.
      • Mosley E.A.
      • Hartwig S.A.
      • Lemon E.
      • Cwiak C.
      Abortion trends from 2007 to 2017 following enactment of Georgia’s 22-week gestational age limit.
      analyzed Georgia Department of Public Health state-wide abortion data from 2007 to 2017 and found that abortions at more than 21 weeks significantly decreased over time, whereas total abortions and abortions at 21 weeks or less remained stable (
      • Hall K.S.
      • Redd S.
      • Narasimhan S.
      • Mosley E.A.
      • Hartwig S.A.
      • Lemon E.
      • Cwiak C.
      Abortion trends from 2007 to 2017 following enactment of Georgia’s 22-week gestational age limit.
      ). However, neither of these studies investigated differences in the effects of HB954 by race or ethnicity.
      The current study expands on prior studies by measuring racial and ethnic differences in number of abortions and abortion ratios before (2007–2012), during (2013–2015), and after (2016–2017) implementation of Georgia's 22-week gestational age limit. Specifically, we investigated how White, Black, Hispanic, and non-Hispanic abortion numbers and abortion ratios changed during partial and full implementation of Georgia HB954.

      Methods

      Induced Termination of Pregnancy Data

      The Georgia Department of Public Health requires all entities that provide abortions to report each Induced Termination of Pregnancy (ITOP) event. Entities submit data using the Georgia Vital Events Registration System, a vital records software developed by Genesis, Inc. for the DPH State Office of Vital Records (
      Georgia Department of Public Health
      Welcome to the Georgia Vital Events Registration System! Division of Vital Records.
      ). Vital Records collects, maintains, amends, and certifies all vital events that occur in Georgia, including ITOPs (
      Georgia Department of Public Health
      About vital records.
      ). Although state vital records are private and confidential, in aggregate they are publicly available through Online Analytical Statistical Information System with limited query functionality (
      Georgia Department of Public Health
      Induced terminations of pregnancy web query. Online analytical statistical information system.
      ), and available in greater detail when requested through the Public Health Information Portal (
      Georgia Department of Public Health
      PHIP data request.
      ), the method we chose for this study. The institutional review board at GDPH granted exempt status for the ITOP analyses, and the institutional review board at Emory University approved all study procedures including those described in this article.
      We analyzed annual, cross-sectional abortion data from 2007 through 2017 from GDPH's state-mandated ITOP reporting in the Public Health Information Portal (
      Georgia Department of Public Health
      Induced terminations of pregnancy web query. Online analytical statistical information system.
      ). Data were stratified by gestational age, race, and ethnicity. All gestational ages are in LMP dating; gestational age categories included all gestational ages total, less than 13 weeks 0 days, 13w0d to 19w6d, 20w0d to 21w6d, and 22 or more weeks; and a composite category of less than 22 weeks (consisting of all abortions up to 21w6d).

      Abortion Ratios

      To account for underlying pregnancy and birth rates of the population, we calculated abortion ratios as the number of induced terminations per 1,000 live births for a given racial or ethnic group, by gestational age from LMP, in a given year or period (
      • Anderson B.
      World population dynamics: An introduction to demography.
      ). Live birth data were obtained from the Georgia Department of Public Health Online Analytical Statistical Information System (
      Georgia Department of Public Health
      Maternal child health: Birth web query. Online analytical statistical information system.
      ).

      Race and Ethnicity

      Racial categories were all racial groups total, White, Black, Other (i.e., Asian, American Indian or Alaska Native, Native Hawaiian or other Pacific Islander, or multiracial), and unknown. Ethnic categories were all ethnic groups total, non-Hispanic, Hispanic (i.e., persons of Mexican, Puerto Rican, Cuban, South or Central American, or other Spanish culture or origin, regardless of race), and other or unknown. We conducted analyses by race and ethnicity separately.
      We analyzed all groups total (including all races and ethnicities), White versus Black, and Hispanic versus non-Hispanic. We included other and unknown racial and ethnic groups in all groups total analyses and excluded them from subgroup analyses. ITOPs abortion numbers for Hispanic people were abnormally high in 2010 and 2011 (more than twice the number of abortions as other years), which called into question the reliability of ethnicity-specific data for that period. For that reason, we excluded 2010 and 2011 data from our primary analyses of Hispanic and non-Hispanic groups, but conducted sensitivity analyses with 2010 and 2011 included (Appendix A).

      Policy Exposure

      Policy exposure was operationalized as a three-category variable: before policy implementation (2007–2012), partial policy implementation (2013–2015), and full policy implementation (2016–2017). This division coincides with the legislative process for HB954, which was enacted in 2012, partially implemented from 2013 to 2015, and fully implemented during 2016 and 2017.

      Changes during and after Policy Implementation

      To examine trends in abortion over time, we first measured changes in the number of abortions and abortion ratios from 2007 to 2017 using linear regression models predicted by year, stratifying by gestational age, race, and ethnicity. To examine how HB954 influenced these trends, we measured changes in number of abortions and abortion ratios using bivariate linear regression models predicted by policy time period: partial policy exposure (2013–2015) and full policy exposure (2016–2017), with pre-policy (2007–2013) as the reference. We then stratified those models by gestational age, race, and ethnicity.

      Results

      ITOPs and Abortion Ratio Trends for 2007 to 2017

      From 2007 to 2017, there were 360,972 abortions in Georgia, with 98% occurring at less than 22 weeks (Table 1). Over time, the total number of abortions for all groups total remained fairly stable, from 33,535 in 2007 to 32,234 in 2017 (b = –298.88, p = .08). They decreased significantly for White people (12,489 to 7,752; b = –322.96, p = .009), and were stable for Black (19,575 to 19,422; b = –76.12, p = .539), Hispanic (2,990 to 2,521; b = –88.11, p = .126), and non-Hispanic groups (26,682 to 29,017; b = –24.04, p = .919). Abortions at 22 weeks or later significantly decreased over time, from 809 to 7 (b = –80.71, p = .001). This finding was consistent across race and ethnicity: White (365–2, b = –31.59; p < .001), Black (404–2; b = –41.55, p = .002), Hispanic (62–0; b = –6.50, b = 0.002), and non-Hispanic (624–6; b = –63.98, p = .007) groups. Notably, abortions among White people decreased significantly for all gestational ages.
      Table 1Number of Abortions in Georgia from 2007 to 2017 with Linear Regression Trends Stratified by Race, Ethnicity, and Gestational Age
      Group20072008200920102011201220132014201520162017TotalbpSE
      All groups all gestations33,53536,09433,42934,85532,93731,63530,86430,19131,10034,09832,23466332–298.88.082152.44
      All groups <13w0d28,72830,15028,21129,49727,89526,84426,56125,78126,76729,99128,48458475135.95.357139.92
      All groups 13w0d–19w6d3,5254,4033,8784,0513,6843,4083,3113,3853,3493,4423,2016643–79.60.01325.72
      All groups 20w0d–21w6d4736455834965104935355424465605421102–2.63.6435.48
      All groups <22w32,72635,19832,67234,04432,08930,74530,40729,70830,56233,99332,22766220–218.17.204159.23
      All groups ≥22w8098967578118488904574835381057112–80.71.00116.34
      White all gestations12,48910,0558,6528,4407,8497,9587,9937,8447,7808,0117,75215763–322.96.00997.68
      White <13w0d10,7268,4957,4097,1706,6976,7896,9446,7936,7767,1046,91314017–245.77.02288.83
      White 13w0d–19w6d1,2181,0938688567577607767717447366951431–41.51.0019.07
      White 20w0d–21w6d180178148154151136145154109144142286–4.08.0161.38
      White <22w12,1249,7668,4258,1807,6057,6857,8657,7187,6297,9847,75015734–291.36.01698.06
      White ≥22w36528922726024427312812615127229–31.59<.0014.28
      Black all gestations19,57520,99619,30320,38417,80317,81818,15717,90318,47920,86419,42240286–76.12.539119.29
      Black <13w0d16,71417,40316,14717,03714,83414,98815,49715,17515,73218,19217,043352354.50.969110.83
      Black 13w0d–19w6d2,1832,7122,4072,6242,2442,0792,0802,1542,1512,2572,0614318–39.11.05918.08
      Black 20w0d–21w6d2743833332772692703093032793493166650.04.9923.74
      Black <22w19,17120,49818,88719,93817,34717,33717,88617,63218,16220,79819,42040218–34.57.788124.80
      Black ≥22w40449841644645648127127131766268–41.55.0029.42
      Hispanic all gestations2,9903,3722,4952,3572,0061,3302,1792,5552,5215076–88.11.12650.81
      Hispanic <13w0d2,5482,8022,0962,0161,7911,1761,9072,3422,3274669–54.85.26445.14
      Hispanic 13w0d–19w6d330430321247182119217185168353–23.22.0045.41
      Hispanic 20w0d–21w6d50623340181620212647–3.54.0101.01
      Hispanic <22w2,9283,2942,4502,3031,9911,3112,1442,5482,5215069–81.61.15050.43
      Hispanic ≥22w62784554151935707–6.50.0021.34
      Non-Hispanic all gestations26,68231,35229,89726,67824,95226,53628,18630,75729,01759774–24.04.919228.15
      Non-Hispanic <13w0d22,91626,19125,30922,65721,58922,64024,20926,94325,5275247067.04.739193.61
      Non-Hispanic 13w0d–19w6d2,7923,8063,3902,8412,5973,0063,0663,1902,9746164–30.17.42735.75
      Non-Hispanic 20w0d–21w6d35056151740539547341752751010373.08.6937.49
      Non-Hispanic <22w26,05830,55829,21625,90324,58126,11927,69230,66029,0115967139.94.868231.18
      Non-Hispanic ≥22w624794681775371417494976103–63.98.00716.91
      Abbreviations: d, days; w, weeks.
      Statistically significant results (p < .05) are bolded and italicized.
      Hispanic and non-Hispanic models omit 2010 and 2011.
      Over the same period, the total abortion ratio (Figure 2) also remained stable, from 222.37 abortions per 1,000 live births in 2007 to 249.57 in 2017 (b = 1.20, p = .32) (Table 2). When stratifying by race and ethnicity, the abortion ratio for White people in Georgia decreased from 139.20 to 104.79 (b = –2.4, p = .008), whereas it remained stable for Black (382.71–426.61; b = 2.83, p = .228), Hispanic (122.17–140.70; b = 1.05, p = .64), and non-Hispanic people (215.36–263.80; b = 1.60, p = .50). Over the entire 2007 to 2017 period, we estimated the abortion ratio for Black people (415.52 abortions/1,000 births) was 3.69 times that of White people (112.74 abortions/1,000 births), whereas the abortion ratio for Hispanic people (121.33 abortions/1,000 births) was 0.48 times that of non-Hispanic people (250.27 abortions/1,000 births).
      Figure thumbnail gr2
      Figure 2Total abortion ratios (abortions per 1,000 live births) from 2007 to 2017 in Georgia stratified by race and ethnicity.
      Table 2Abortion Ratios (Abortions per 1,000 Live Births) in Georgia from 2007 to 2017 with Linear Regression Trends Stratified by Race, Ethnicity, and Gestational Age
      Group20072008200920102011201220132014201520162017TotalbpSE
      All groups all gestations222.37246.44236.53260.76249.07243.14240.17230.86236.80262.41249.57243.191.20.3201.14
      All groups <13w0d190.50205.85199.61220.67210.94206.31206.68197.14203.81230.81220.54208.111.92.0830.98
      All groups 13w0d–19w6d23.3730.0627.4430.3127.8626.1925.7625.8825.5026.4924.7826.70–0.22.3000.20
      All groups 20w0d–21w6d3.144.404.133.713.863.794.164.143.404.314.203.920.04.3740.04
      All groups <22w217.01240.32231.17254.69242.66236.30236.61227.17232.71261.61249.52236.571.74.1621.14
      All groups ≥22w5.366.125.366.076.416.843.563.694.100.810.054.45–0.54.0040.14
      White all gestations139.20124.46109.09113.29106.47108.31110.76105.51103.27108.64104.79112.74–2.40.0080.71
      White <13w0d119.55105.1593.4296.2590.8492.4096.2291.3789.9496.3493.4597.28–1.64.0370.67
      White 13w0d–19w6d13.5813.5310.9411.4910.2710.3410.7510.379.889.989.3997.28–0.36.0010.07
      White 20w0d–21w6d2.012.201.872.072.051.852.012.071.451.951.921.95–0.02.1970.02
      White <22w135.13120.88106.23109.80103.16104.60108.98103.82101.26108.27104.76111.06–2.03.0220.73
      White ≥22w4.073.582.863.493.313.721.771.692.000.370.032.49–0.37<.0010.06
      Black all gestations382.71426.64402.84446.58396.49406.29413.73398.71409.20464.34426.61415.522.83.2282.18
      Black <13w0d326.78353.63336.98373.25330.36341.76353.12337.96348.37404.87374.36352.533.90.0751.94
      Black 13w0d–19w6d42.6855.1150.2357.4949.9847.4147.4047.9747.6350.2345.2749.21–0.33.4410.41
      Black 20w0d–21w6d5.367.786.956.075.996.167.046.756.187.776.946.630.72.3440.07
      Black <22w374.81416.52394.16436.81386.33395.33407.56392.68402.18462.87426.57407.623.65.1382.24
      Black ≥22w7.9010.128.689.7710.1610.976.186.047.021.470.047.15–0.82.0060.23
      Hispanic all gestations122.17131.16101.96136.09118.4377.40122.49142.67140.70121.331.05.642.14
      Hispanic <13w0d104.11108.9985.65116.40105.7468.44107.20130.77129.87105.751.96.3441.94
      Hispanic 13w0d–19w6d13.4816.7313.1214.2610.756.9312.2010.339.3812.24–0.58.0310.21
      Hispanic 20w0d–21w6d2.042.411.352.311.060.931.121.171.451.59–0.10.0690.05
      Hispanic <22w119.63128.13100.12132.97117.5576.30120.52142.27140.70119.581.29.5642.13
      Hispanic ≥22w2.533.031.843.120.891.111.970.390.001.75–0.24.0180.08
      Non-Hispanic all gestations215.36271.47268.80245.00223.70237.04252.08278.73263.80250.271.60.4972.24
      Non-Hispanic <13w0d184.96226.78227.55208.07193.55202.24216.52244.16232.07214.732.16.2841.86
      Non-Hispanic 13w0d–19w6d22.5432.9630.4826.0923.2826.8527.4228.9127.0427.25–0.07.8340.34
      Non-Hispanic 20w0d–21w6d2.824.864.653.723.544.233.734.784.644.090.05.4620.07
      Non-Hispanic <22w210.32264.59262.67237.88220.38233.32247.67277.85263.75246.072.14.3702.23
      Non-Hispanic ≥22w5.046.886.127.123.333.724.420.880.054.20–0.54.0130.16
      Abbreviations: d, days; w, weeks.
      Statistically significant results (p < .05) are bolded and italicized.
      Hispanic and non-Hispanic models omit 2010 and 2011.

      Changes during and after Policy Implementation

      Compared with the pre-implementation period, the total number of abortions temporarily decreased during partial implementation between 2013 and 2015 (b = –3029.17, p = .012), but returned to baseline levels during full implementation of the policy in 2016 and 2017 (b = –581.50, p = .607) (Table 3). When stratified by race and ethnicity, the total number of abortions for each group did not significantly differ between baseline and full implementation (bWhite = –1359.00, p = .271; bBlack = 829.83, p = .385; bHispanic = –265.50, p = .492; bnon-Hispanic = 1234.75, p = .497). When analyzing abortion numbers at different gestational ages, we found significant decreases in abortions at 22 weeks or later, for all groups combined (b = –779.17, p < .001) as well as for White (b = –261.83, p < .001), Black (b = –416.17, p < .001), Hispanic (b = –56.25, p = .002), and non-Hispanic (b = –667.00, p < .001) groups specifically.
      Table 3Linear Regression Models of the Number of Abortions by Policy Exposure (Before-Reference, Partial Implementation, and Full Implementation) to Georgia's 22-Week Gestational Age Limit, Stratified by Race, Ethnicity, and Gestational Age
      GroupPartial Policy

      2013–2015
      Full Policy

      2016–2017
      bpSEbpSE
      All groups (n = 11, F = 5.26, p = .035)–3029.17.012941.44–581.50.6071087.08
      All groups <13w0d (n = 11, F = 5.94, p = .026)–2184.50.018733.97683.33.443847.51
      All groups 13w0d–19w6d (n = 11, F = 3.69, p = .073)–476.50.053209.69–503.33.071242.13
      All groups 20w0d–21w6d (n = 11, F = 0.81 p = .479)–25.67.55741.8617.67.72448.33
      All groups <22w (n = 11, F = 4.70, p = .045)–2686.67.021933.61197.67.8591078.04
      All groups ≥22w (n = 11, F = 170.56 p < .001)–342.50.00137.48–779.17<.00143.28
      White (n = 11, F = 1.28, p = .329)–1368.17.206995.04–1359.00.2711148.98
      White <13w0d (n = 11, F = 0.89, p = .448)–1043.33.260859.83–872.50.405992.85
      White 13w0d–19w6d (n = 11, F = 2.06, p = .190)–161.67.165105.76–209.83.124122.12
      White 20w0d–21w6d (n = 11, F = 1.57, p = .266)–21.83.12912.92–14.83.34914.91
      White <22w (n = 11, F = 1.00, p = .408)–1226.83.242970.58–1097.17.3561120.73
      White ≥22w (n = 11, F = 36.99, p < .001)–141.33.00127.93–261.83<.00132.35
      Black (n = 11, F = 2.03, p = .194)–1133.50.185781.41829.83.385902.30
      Black <13w0d (n = 11, F = 3.41, p = .085)–719.17.294640.571430.33.089739.66
      Black 13w0d–19w6d (n = 11, F = 1.77, p = .231)–246.50.130145.98–215.83.236168.56
      Black 20w0d–21w6d (n = 11, F = 0.59, p = .575)–4.00.88827.4831.50.35031.74
      Black <22w (n = 11, F = 2.45, p = .148)–969.67.247776.261246.00.202896.34
      Black ≥22w (n = 11, F = 106.53, p < .001)–163.83<.00125.06–416.17<.00128.94
      Hispanic (n = 9, F = 4.64, p = .061)–965.17.024320.36–265.50.492363.26
      Hispanic <13w0d (n = 9, F = 4.41, p = .066)–740.83.032265.93–31.00.921301.54
      Hispanic 13w0d–19w6d (n = 9, F = 7.45, p = .023)–159.33.01446.28–155.50.02552.48
      Hispanic 20w0d–21w6d (n = 9, F = 9.32, p = .014)–28.25.0076.95–22.75.0287.88
      Hispanic <22w (n = 9, F = 4.55, p = .063)–928.42.025313.49–209.25.578355.47
      Hispanic ≥22w (n = 9, F = 17.44, p = .003)–36.75.0079.02–56.25.00210.23
      Non-Hispanic (n = 9, F = 1.87, p = .233)–2094.00.2141507.071234.75.4971708.86
      Non-Hispanic <13w0d (n = 9, F = 3.08, p = .120)–1455.58.2541153.441966.75.1831307.87
      Non-Hispanic 13w0d–19w6d (n = 9, F = 0.57, p = .169)–317.58.312287.82–125.25.714326.36
      Non-Hispanic 20w0d–21w6d (n = 9, F = 0.92, p = .447)–29.92.61155.7360.25.37763.19
      Non-Hispanic <22w (n = 9, F = 2.24 p = .188)–1803.08.2671474.811901.75.2991672.28
      Non-Hispanic ≥22w (n = 9, F = 58.15 p < .001)–291.17.00254.98–667.00<.00162.34
      Abbreviations: d, days; w, weeks.
      Statistically significant results are bolded and italicized (p < .05);
      Hispanic and non-Hispanic models omit 2010 and 2011.
      Overall, the total abortion ratio did not change during partial (b = –7.11, p = .385) or full (b = 12.94, p = .186) policy implementation, but we noted an increasing abortion ratio before 13w0d (b = 20.02, p = .024) and a decreasing abortion ratio at 22 weeks or later (b = –5.60, p < .001) during full policy implementation (Table 4, and Figure 3, Figure 4). For each group, the abortion ratio at 22 weeks or later decreased during both partial and full policy implementation. Notably, the decrease in abortion ratios at 22 weeks or later was significantly greater for Black people (b = –8.84, 95% confidence interval [95% CI], –10.70 to –6.98) than White people (b = –3.31, 95% CI, –3.95 to –2.67) and marginally greater for non-Hispanic (b = –5.66, 95% CI, –7.17 to –4.60) compared with Hispanic people (b = –3.03, 95% CI, –4.87 to –1.19). Moreover, during full policy implementation, Black people saw an increase in the abortion ratio at less than 22 weeks (b = 44.06, p = .028) that was not observed in the White (b = –6.78, p = .433), Hispanic (b = 21.27, p = .212), or non-Hispanic groups (b = 26.93, p = .172).
      Table 4Linear Regression Models of Abortion Ratios (Abortions per 1,000 Live Births) by Policy Exposure (Before-Reference, Partial Implementation, and Full Implementation) to Georgia's 22-Week Gestational Age Limit Stratified by Race, Ethnicity, and Gestational Age
      GroupPartial Policy

      2013–2015
      Full Policy

      2016–2017
      bpSEbpSE
      All groups (n = 11, F = 2.02, p = .195)–7.11.3857.7412.94.1868.94
      All groups <13w0d (n = 11, F = 4.80, p = .43)–3.10.6326.2320.02.0247.19
      All groups 13w0d–19w6d (n = 11, F = 1.08, p = .385)–1.82.2521.48–1.90.2961.70
      All groups 20w0d–21w6d (n = 11, F = 0.81 p = .479)0.06.8280.290.42.2420.33
      All groups <22w (n = 11, F = 3.14, p = .099)–4.86.5377.5318.54.0668.70
      All groups ≥22w (n = 11, F = 90.24 p < .001)–2.24<.0010.37–5.60<.0010.42
      White (n = 11, F = 1.34, p = .314)–10.29.1957.28–10.09.2648.40
      White <13w0d (n = 11, F = 0.69, p = .528)–7.09.2926.29–4.71.5357.26
      White 13w0d–19w6d (n = 11, F = 2.56, p = .138)–1.36.1540.86–2.00.0791.00
      White 20w0d–21w6d (n = 11, F = 0.67, p = .538)–0.16.2820.14–0.07.6780.16
      White <22w (n = 11, F = 0.85, p = .461)–8.61.2617.12–6.78.4338.23
      White ≥22w (n = 11, F = 77.61, p < .001)–1.68<.0010.24–3.31<.0010.28
      Black (n = 11, F = 2.52, p = .141)–3.04.84114.6635.22.07116.92
      Black <13w0d (n = 11, F = 6.41, p = .022)2.69.81911.3845.82.00813.14
      Black 13w0d–19w6d (n = 11, F = 0.55, p = .597)–2.82.3903.10–2.73.4673.58
      Black 20w0d–21w6d (n = 11, F = 1.30, p = .326)0.27.6170.520.97.1460.60
      Black <22w (n = 11, F = 3.87, p = .067)0.15.99214.3144.06.02816.52
      Black ≥22w (n = 11, F = 61.23, p < .001)–3.19.0020.70–8.84<.0010.81
      Hispanic (n = 9, F = 2.39, p = .173)–16.73.26713.5918.84.27015.52
      Hispanic <13w0d (n = 9, F = 3.34, p = .106)–9.99.43712.0026.53.09913.60
      Hispanic 13w0d–19w6d (n = 9, F = 5.77, p = .040)–4.44.0251.50–4.54.3701.70
      Hispanic 20w0d–21w6d (n = 9, F = 7.27, p = .025)–0.99.0100.27–0.72.0570.31
      Hispanic <22w (n = 9, F = 2.62, p = .152)–15.42.29413.4221.27.21215.22
      Hispanic ≥22w (n = 9, F = 14.28, p = .005)–1.31.0190.41–2.44.0020.47
      Non-Hispanic (n = 9, F = 1.60, p = .278)–12.55.45617.8721.10.28217.87
      Non-Hispanic <13w0d (n = 9, F = 2.84, p = .136)–7.74.55212.2726.28.10813.91
      Non-Hispanic 13w0d–19w6d (n = 9, F = 0.36, p = .709)–2.16.4572.72–0.40.9903.08
      Non-Hispanic 20w0d–21w6d (n = 9, F = 1.02, p = .415)–0.18.7440.530.69.2910.60
      Non-Hispanic <22w (n = 9, F = 2.10 p = .203)–10.01.53515.3326.93.17217.38
      Non-Hispanic ≥22w (n = 9, F = 38.36 p < .001)–2.46.0060.59–5.82<.0010.67
      Abbreviations: d, days; w, weeks.
      Statistically significant results are bolded and italicized (p < .05);
      Hispanic and non-Hispanic models omit 2010 and 2011.
      Figure thumbnail gr3
      Figure 3Abortion ratios (abortions per 1,000 live births) at less than 22 weeks of gestation (from last menstrual period) from 2007 to 2017 in Georgia, stratified by race and ethnicity.
      Figure thumbnail gr4
      Figure 4Abortion ratios (abortions per 1,000 live births) at 22 weeks of gestation or more (from last menstrual periods) from 2007 to 2017 (excluding 2010 and 2011) in Georgia, stratified by race and ethnicity.

      Discussion

      Differential Policy Effects on Abortion at 22 Weeks or Later by Race and Ethnicity

      Our results suggest that HB954 carried differential effects on abortion numbers and ratios depending on race and ethnicity. First, we observed significant decreases in abortion at 22 weeks or later for all groups—with steeper decreases for Black people compared with White people and for non-Hispanic people compared with Hispanic people. This is due, at least in part, to higher baseline abortion ratios for Black compared with White people (7.09 abortions at or after 22 weeks/1,000 live births for Black people vs. 4.07 for White in 2007) and non-Hispanic compared with Hispanic people (5.04 abortions at or after 22 weeks/1,000 live births for non-Hispanic people vs. 2.53 for Hispanic in 2007). By 2017, all abortion ratios at 22 weeks or later had decreased to nearly zero. These findings build on previous research (
      • Hall K.S.
      • Redd S.
      • Narasimhan S.
      • Mosley E.A.
      • Hartwig S.A.
      • Lemon E.
      • Cwiak C.
      Abortion trends from 2007 to 2017 following enactment of Georgia’s 22-week gestational age limit.
      ) that showed overall number of abortions at or after 22 weeks decreased significantly after both partial and full implementation of HB954. Our results also support the hypothesis of
      • Roberts S.
      • Gould H.
      • Upadhyay U.D.
      Implications of Georgia’s 20-week abortion ban.
      , who anticipated that HB954 would disproportionately affect Black people because they made up the majority of people receiving abortions at or after 20 weeks in Georgia. Notably, it is challenging to interpret our results on ethnic differences in abortion, because the non-Hispanic group contains all races (e.g., White and Black).
      The constructs of “race” and “ethnicity” are demographic variables collected by the ITOP system as reported by clinic staff. They therefore might serve as markers for discrimination and inequities inherent in our society and health care system. Owing to racism and ethnocentrism, Black and Latinx individuals are more likely to be unemployed (
      Department of Labor Women’s Bureau
      Unemployment rates.
      ), live in households headed by a single woman (
      U.S. Census Bureau
      Historical living arrangements of children.
      ), or experience poverty-related barriers to accessing abortion services (
      • Blount L.G.
      Poor women suffer most from restrictive abortion policies. Reproductive Health Reality Check.
      ;
      • Mosley E.A.
      • Ayala S.
      • Jah Z.
      • Hailstorks T.
      • Scales M.
      • Gutierrez M.
      • Hall K.S.
      ‘All the things that make life harder’: Intersectionality and medication abortion in Georgia.
      ;
      • Semega J.L.
      • Fontenot K.R.
      • Kollar M.A.
      Income and poverty in the United States: 2016. Current Population Reports.
      ;
      • Upadhyay U.D.
      • Johns N.E.
      • Cartwright A.F.
      • Franklin T.E.
      Sociodemographic characteristics of women able to obtain medication abortion before and after Ohio’s law requiring use of the Food and Drug Administration protocol.
      ). The Turnaway Study, a longitudinal prospective study, assessed 956 women across 21 states from 30 abortion facilities that are located more than 150 miles from another facility that could provide abortions at a later gestational age limit. Women were recruited if they presented for abortion during the first trimester, presented up to 2 weeks before the gestational age limit (based on limits of the clinician, facility, or legal restrictions), or up to 3 weeks after the gestational age limit and were denied an abortion (
      • Upadhyay U.D.
      • Weitz T.A.
      • Jones R.K.
      • Barar R.E.
      • Foster D.G.
      Denial of abortion because of provider gestational age limits in the United States.
      ). Although the researchers did not note significant racial/ethnic differences, they found that people were more likely to seek an abortion at or after 20 weeks if they were younger, unemployed, single mothers, or having trouble accessing the clinic owing to financial, transportation, or other barriers (
      • Foster D.G.
      • Kimport K.
      Who seeks abortions at or after 20 weeks?.
      ). Furthermore, other studies have found that people of color are more likely to access abortion care in the second trimester (
      • Jones R.K.
      • Finer L.B.
      Who has second-trimester abortions in the United States?.
      ). Therefore, it is likely that gestational age limits inherently carry greater consequences for people of color.
      Our finding that HB954 had stronger effects on abortion at or after 22 weeks among Black people echoes previous research that shows Black individuals can be affected disproportionately by abortion restrictions. Because Black individuals are more likely to rely on Medicaid for health insurance (
      • Boonstra H.D.
      Abortion in the lives of women struggling financially: Why insurance coverage matters.
      ), they are more affected by the Hyde Amendment, which restricts federal funding for abortion. In North Carolina, researchers found that the absence of Medicaid funding for abortion was associated with 11% fewer abortions for Black individuals compared with 1% fewer abortions for White individuals (
      • Cook P.J.
      • Parnell A.M.
      • Moore M.J.
      • Pagnini D.
      The effects of short-term variation in abortion funding on pregnancy outcomes [Working Paper].
      ). A national study by
      • Coles M.S.
      • Makino K.K.
      • Stanwood N.L.
      • Dozier A.
      • Klein J.D.
      How are restrictive abortion statutes associated with unintended teen birth?.
      similarly found that Black adolescents are four times more likely to have an unintended birth when living in states without Medicaid funding for abortion. More recently,
      • Upadhyay U.D.
      • Johns N.E.
      • Cartwright A.F.
      • Franklin T.E.
      Sociodemographic characteristics of women able to obtain medication abortion before and after Ohio’s law requiring use of the Food and Drug Administration protocol.
      evaluated an Ohio law that required patients to make four visits to receive a medication abortion; the authors found that Black individuals were less likely to access medication abortion under the law, representing 24% of medication abortion patients before the law, but only 16% after the law.

      Shifts in the Timing of Abortion after HB954 by Race and Ethnicity

      After full implementation of HB954, the less than 13w0d abortion ratio for Black people increased significantly, whereas those for other groups were stable or marginally declining. The increase for Black people is likely driving the increase we observed for abortion before 13w0d among all groups combined and among non-Hispanic people in our ethnic sensitivity analyses. This finding suggests that even with the new restrictive abortion policy, Black people exhibit continued resilience in overcoming barriers to obtain the health care they need. In contrast, the overall White abortion ratio was stable, and we actually saw marginal (but not statistically significant) decreases in the White abortion ratio at 13w0d to 19w6d (b = –2.00, p = .08). In other words, abortions among White people may be decreasing over time, but this is not likely because of Georgia's restrictions on abortion at 22 weeks or more.
      This finding points toward national trends, where overall abortion rates are decreasing (
      • Blount L.G.
      Poor women suffer most from restrictive abortion policies. Reproductive Health Reality Check.
      ;
      • Boonstra H.D.
      Abortion in the lives of women struggling financially: Why insurance coverage matters.
      ;
      Guttmacher Institute
      State bans on abortion throughout pregnancy. State laws and policies.
      ;
      • Jones R.K.
      • Jerman J.
      Abortion incidence and service availability in the United States, 2014.
      ). From 1995 to 2014, the U.S. abortion ratio decreased from 25.9 to 18.8 abortions per 1,000 live births in response to a variety of factors, including improved access to all methods of contraception, better education and use of contraceptives, lower rates of unintended pregnancy, changes in sexual behavior, and possibly greater use of self-managed abortion outside of the traditional health system (
      • Foster D.G.
      Dramatic decreases in US abortion rates: Public health achievement or failure?.
      ;
      Guttmacher Institute
      State bans on abortion throughout pregnancy. State laws and policies.
      ;
      • Jones R.K.
      • Jerman J.
      Abortion incidence and service availability in the United States, 2014.
      ;
      • Ralph L.
      • Foster D.G.
      • Raifman S.
      • Biggs M.A.
      • Samari G.
      • Upadhyay U.
      • Grossman D.
      Prevalence of self-managed abortion among women of reproductive age in the United States.
      ).
      Other studies might explain why earlier gestation White abortion ratios are stable (if not declining) as Black ratios increase. Researchers in North Carolina similarly found that the state's total Black abortion ratio increased slightly from 1980 to 2004 after the implementation of Medicaid funding restrictions for abortion, whereas the White abortion ratio decreased significantly (
      • Alvey J.
      • Bryant A.G.
      • Curtis S.
      • Speizer I.S.
      • Morgan S.P.
      • Tippett R.
      • Perreira K.
      Trends in abortion incidence and availability in North Carolina, 1980–2013.
      ). They concluded that “access to contraception may be driving the reduction in abortion rates in some demographic groups, but not others” (
      • Alvey J.
      • Bryant A.G.
      • Curtis S.
      • Speizer I.S.
      • Morgan S.P.
      • Tippett R.
      • Perreira K.
      Trends in abortion incidence and availability in North Carolina, 1980–2013.
      p. 721). For example, Black women have decreased access to contraceptive services, higher levels of unintended pregnancy, and higher poverty burdens that may result in terminating pregnancies they cannot afford to keep (
      • Davis A.
      Racism, birth control, and reproductive rights.
      ;
      • Dehlendorf C.
      • Harris L.H.
      • Weitz T.A.
      Disparities in abortion rates: A public health approach.
      ;
      • Finer L.B.
      • Zolna M.R.
      Declines in unintended pregnancy in the United States, 2008–2011.
      ;
      • Masinter L.M.
      • Feinglass J.
      • Simon M.A.
      Pregnancy intention and use of contraception among Hispanic women in the United States: Data from the National Survey of Family Growth, 2006–2010.
      ;
      • Ross L.
      African-American women and abortion.
      ,
      • Ross L.
      Understanding reproductive justice.
      ). Reproductive justice leaders have cited the relatively higher abortion rates and ratios among Black women as a reproductive injustice (
      • Luna Z.
      • Luker K.
      Reproductive justice.
      ;
      • Ross L.
      Understanding reproductive justice.
      ;
      • Ross L.
      • Solinger R.
      ;
      • Silliman J.
      • Fried M.G.
      • Ross L.
      • Gutiérrez E.R.
      Undivided rights: Women of color organize for reproductive justice.
      ). They expand on the framework of reproductive rights, which emphasizes access to contraception and abortion, to also include the human rights to have children if desired, and to raise those children in safe and healthy environments. Although access to abortion and contraception in Georgia are fundamental reproductive and human rights, so is access to social and economic conditions that enable parenting with dignity regardless of race or ethnicity.
      Those who face policy restrictions and poverty-related barriers to abortion care sometimes attempt to self-manage their abortions (
      • Aiken A.R.A.
      • Broussard K.
      • Johnson D.M.
      • Padron E.
      Motivations and Experiences of People Seeking Medication Abortion Online in the United States..
      ,
      • Aiken A.R.A.
      • Starling J.E.
      • van der Wal A.
      • van der Vliet S.
      • Broussard K.
      • Johnson D.M.
      • Scott J.G.
      Demand for self-managed medication abortion through an online telemedicine service in the United States.
      ;
      • Gerdts C.
      • Fuentes L.
      • Grossman D.
      • White K.
      • Keefe-Oates B.
      • Baum S.E.
      • Potter J.E.
      Impact of clinic closures on women obtaining abortion services after implementation of a restrictive law in Texas.
      ;
      • Grossman D.
      • White K.
      • Fuentes L.
      • Hopkins K.
      • Stevenson A.
      • Yeatman S.
      • Potter J.E.
      Knowledge, opinion, and experience related to abortion self-induction in Texas. Texas Policy Evaluation Project.
      ;
      • Moseson H.
      • Herold S.
      • Filippa S.
      • Barr-Walker J.
      • Baum S.E.
      • Gerdts C.
      Self-managed abortion: A systematic scoping review.
      ). An estimated 7% of U.S. women have attempted self-managed abortion in their lifetime (
      • Ralph L.
      • Foster D.G.
      • Raifman S.
      • Biggs M.A.
      • Samari G.
      • Upadhyay U.
      • Grossman D.
      Prevalence of self-managed abortion among women of reproductive age in the United States.
      ), with Black and Hispanic women being three times more likely to self-terminate than White women (
      • Grossman D.
      • Holt K.
      • Peña M.
      • Lara D.
      • Veatch M.
      • Córdova D.
      • Blanchard K.
      Self-induction of abortion among women in the United States.
      ,
      • Grossman D.
      • Ralph L.
      • Raifman S.
      • Upadhyay U.D.
      • Gerdts C.
      • Biggs M.A.
      • Foster D.G.
      Lifetime prevalence of self-induced abortion among a nationally representative sample of U.S. women.
      ). Qualitative research with U.S. women who have self-managed their abortions emphasizes how hostile policy environments have created logistical and financial barriers to care that push women toward methods that can be less safe—particularly lower income women/people and women of color, who had fewer resources to overcome those barriers (
      • Aiken A.R.A.
      • Broussard K.
      • Johnson D.M.
      • Padron E.
      Motivations and Experiences of People Seeking Medication Abortion Online in the United States..
      ). Most commonly, people self-manage their abortion using misoprostol (
      • Ralph L.
      • Foster D.G.
      • Raifman S.
      • Biggs M.A.
      • Samari G.
      • Upadhyay U.
      • Grossman D.
      Prevalence of self-managed abortion among women of reproductive age in the United States.
      ), one of two medicines used for medication abortion (
      Guttmacher Institute
      Medication Abortion.
      ). It causes uterine contractions, bleeding, and expulsion of the pregnancy (
      U.S. Food and Drug Administration
      Misoprostol (marketed as Cytotec) Information.
      ). International studies suggest misoprostol is 96% effective up to 9 weeks' gestation or 71% after 12 weeks’ gestation and is safe (
      • Foster D.G.
      Dramatic decreases in US abortion rates: Public health achievement or failure?.
      ;
      • Moseson H.
      • Herold S.
      • Filippa S.
      • Barr-Walker J.
      • Baum S.E.
      • Gerdts C.
      Self-managed abortion: A systematic scoping review.
      ). Self-managed abortion with misoprostol can be made safer and more effective through online resources and support hotlines that provide information and/or medications, although some people still experience stress, find it harder to seek advice from a clinician, and might turn to other less safe termination options (
      • Aiken A.R.A.
      • Broussard K.
      • Johnson D.M.
      • Padron E.
      Motivations and Experiences of People Seeking Medication Abortion Online in the United States..
      ,
      • Gerdts C.
      • Hudaya I.
      Quality of Care in a Safe-Abortion Hotline in Indonesia: Beyond Harm Reduction..
      ). Some forms of self-managed abortion—including ingesting poisons, inserting objects into the uterus, or causing trauma to the abdomen—can be dangerous and increase risk of maternal morbidity and mortality (
      • Moseson H.
      • Herold S.
      • Filippa S.
      • Barr-Walker J.
      • Baum S.E.
      • Gerdts C.
      Self-managed abortion: A systematic scoping review.
      ).
      Abortion restrictions like HB954 can also lead to unintended births (
      • Upadhyay U.D.
      • Weitz T.A.
      • Jones R.K.
      • Barar R.E.
      • Foster D.G.
      Denial of abortion because of provider gestational age limits in the United States.
      ), potentially increasing risk of poor maternal and child health outcomes like delayed prenatal care, substance use, and child maltreatment (
      • Coles M.S.
      • Makino K.K.
      • Stanwood N.L.
      • Dozier A.
      • Klein J.D.
      How are restrictive abortion statutes associated with unintended teen birth?.
      ;
      • Mosher W.D.
      • Jones J.
      • Abma J.C.
      Intended and unintended births in the United States: 1982–2010. National health statistics reports, 55.
      ). The Turnaway Study demonstrated that abortion-seeking patients “turned away” owing to gestational age limits were more likely to be living in poverty 7 years later (
      • Foster D.G.
      • Biggs M.A.
      • Ralph L.
      • Gerdts C.
      • Roberts S.
      • Glymour M.M.
      Socioeconomic outcomes of women who receive and women who are denied wanted abortions in the United States.
      ), to have negative mental health outcomes (
      • Biggs M.A.
      • Upadhyay U.D.
      • McCulloch C.E.
      • Foster D.G.
      Women’s mental health and well-being 5 years after receiving or being denied an abortion: A prospective, longitudinal cohort study.
      ), and to remain in violent relationships (
      • Roberts S.C.
      • Biggs M.A.
      • Chibber K.S.
      • Gould H.
      • Rocca C.H.
      • Foster D.G.
      Risk of violence from the man involved in the pregnancy after receiving or being denied an abortion.
      ) compared with counterparts who received an abortion. Our results suggest that, because gestational age limits seem to have stronger effects on abortion at 22 weeks or later among Black people, HB954 has the potential to exacerbate existing racial disparities in poverty (
      • Boonstra H.D.
      Abortion in the lives of women struggling financially: Why insurance coverage matters.
      ), mental health (
      • Berger M.
      • Sarnyai Z.
      ‘More than skin deep’: Stress neurobiology and mental health consequences of racial discrimination.
      ), and pregnancy-associated intimate partner violence including homicide (
      • Kivisto A.J.
      • Mills S.
      • Elwood L.S.
      Racial disparities in pregnancy-associated intimate Partner Homicide.
      ;
      • West C.M.
      Black women and intimate partner violence: New directions for research.
      ). However, the increase we observed in the Black abortion ratio before 13w0d suggests Black people might be successfully mitigating these potential consequences by accessing abortion at earlier gestational ages.

      Implications for Practice and/or Policy

      Abortion restrictions like HB954 decrease access to a safe and needed health service without addressing the underlying social conditions that increase the need for abortion. A public health approach to abortion would focus on equitable access to safe abortion care, high-quality contraceptive services, and anti-poverty measures that improve material and social support resources in preparing for and raising children if desired (
      • Bongaarts J.
      • Westoff C.
      The potential role of contraception in reducing abortion..
      ;
      • Davis A.
      Racism, birth control, and reproductive rights.
      ;
      • Dehlendorf C.
      • Harris L.H.
      • Weitz T.A.
      Disparities in abortion rates: A public health approach.
      ;
      • Ross L.
      African-American women and abortion.
      ).

      Strengths and Limitations

      We analyzed state-level data on the number of abortions as well as abortion ratios, which are more appropriate for studying group differences given underlying differences in population size and fertility. However, the available ITOPs data include people coming to Georgia for abortion, although live birth rates represent only people who gave birth in Georgia. Researchers have previously calculated that 10% of abortions in Georgia are for patients from neighboring states (Alabama, Florida, North Carolina, South Carolina, and Tennessee), and out-of-state patients are more likely to have an abortion after 20 weeks' gestation compared with Georgia residents (
      • Shapiro R.
      • Erhardt-Ohren B.
      • Rochat R.
      Comparison of women from Georgia and contiguous states who obtained abortions in Georgia, 1994–2016.
      ). Moreover, because ITOPs data are only available by year (not by month) and because HB954 was implemented in phases over time, it is challenging to define our policy exposure categories accurately. For example, HB954 was fully implemented in October 2015. We made the decision to include all of 2015 in our “partial exposure” category, but this might not accurately reflect the reality of October to December in 2015. Second, unlike studies that rely on data from certain clinics, our study used abortion data collected from all facilities providing abortions in Georgia. However, we do not have data on people who might have self-managed their abortions. Third, this study focused on racial and ethnic differences in abortion after implementation of the gestational age limit policy and, therefore, crucially expands on previous evaluations of HB954 (
      • Hall K.S.
      • Redd S.
      • Narasimhan S.
      • Mosley E.A.
      • Hartwig S.A.
      • Lemon E.
      • Cwiak C.
      Abortion trends from 2007 to 2017 following enactment of Georgia’s 22-week gestational age limit.
      ;
      • Roberts S.
      • Gould H.
      • Upadhyay U.D.
      Implications of Georgia’s 20-week abortion ban.
      ) and similar policies. Unfortunately, we could not disaggregate by race and ethnicity at the same time using existing datasets, because DPH is careful to only share de-identified and un-identifiable data. Disaggregating by both race and ethnicity simultaneously when looking at abortions after 20 weeks (of which there are relatively very few) decreases the subsamples to less than 5, which is the cut-off for DPH. Similarly, it is generally unknown how racial and ethnic identity is determined on both ITOP and live birth records, which may result in the misrepresentation of birthing people's identities. One previous study in Georgia found that, rather than through self-report, clinic staff assess race and ethnicity based on physical appearances (
      • Chelko O.
      • Rochat R.
      • Herold J.M.
      • Carter J.
      • Lavoie M.
      An evaluation of abortion data and abortion reporting in Georgia.
      ). However, research has shown that external perceptions of race and ethnicity shape the treatment people receive; thus, although the fact that ITOPS data likely rely on clinic staff's assessment of race and ethnicity is a limitation, it can help to capture the racism and ethnocentrism that people perceived as Black and Hispanic experience. Our decision to remove Hispanic and non-Hispanic data from 2010 and 2011 as outliers could reduce our ability to draw conclusions about ethnic differences in the data. Moving forward, researchers could request and DPH could report combined years of ITOPs data (e.g., 2007–2008, 2009–2010, 2011–2012, etc.) to disaggregate by race and ethnicity without going below the five-patient threshold for maintaining confidentiality. Finally, given the start-and-stop implementation of HB954, it is challenging to specifically define the periods of “partial” and “full” implementation. For example, Figure 4 shows significant decreases in abortion at or after 22 weeks LMP between 2012 and 2013 although the law had not been implemented. It is possible that patients perceived the law to be in effect based on media coverage during that time. Qualitative studies are needed to further explore this notion.

      Conclusions

      Although Georgia's 22-week restriction was associated with decreased number of abortions and abortion ratios at 22 weeks or later for all racial and ethnic groups, it was associated with increased abortion ratios before 13w0d for Black people. This suggests that Black people are seeking abortions at earlier gestational ages in response to the law. The decrease in relatively uncommon abortions at 22 weeks or later (<2%), however, still reflects decreased access to a safe and vital reproductive health service for all racial and ethnic groups. We also found notable racial differences between Black and White people. For one, the associated decrease in abortions at 22 weeks or later was steeper for Black people compared with White people, suggesting that the restrictive abortion policy had stronger effects on Black communities. And although the policy was associated with stable and marginally decreasing White abortion ratios at earlier gestational ages, it was associated with increased Black abortion ratios at less than 22 weeks. This finding could reflect ongoing racial reproductive inequities, including Black people's poorer access to contraception and the socioeconomic resources needed for childrearing.

      References

        • Aiken A.R.A.
        • Broussard K.
        • Johnson D.M.
        • Padron E.
        Motivations and Experiences of People Seeking Medication Abortion Online in the United States..
        Perspectives on Sexual and Reproductive Health. 2018; 50: 157-163
        • Aiken A.R.A.
        • Starling J.E.
        • van der Wal A.
        • van der Vliet S.
        • Broussard K.
        • Johnson D.M.
        • Scott J.G.
        Demand for self-managed medication abortion through an online telemedicine service in the United States.
        American Journal of Public Health. 2020; 110: 90-97
        • Alvey J.
        • Bryant A.G.
        • Curtis S.
        • Speizer I.S.
        • Morgan S.P.
        • Tippett R.
        • Perreira K.
        Trends in abortion incidence and availability in North Carolina, 1980–2013.
        Southern Medical Journal. 2017; 110: 714-721
        • Anderson B.
        World population dynamics: An introduction to demography.
        Pearson Higher, Upper Saddle River, NJ2014
        • Berger M.
        • Sarnyai Z.
        ‘More than skin deep’: Stress neurobiology and mental health consequences of racial discrimination.
        Stress. 2015; 18: 1-10
        • Biggs M.A.
        • Upadhyay U.D.
        • McCulloch C.E.
        • Foster D.G.
        Women’s mental health and well-being 5 years after receiving or being denied an abortion: A prospective, longitudinal cohort study.
        JAMA Psychiatry. 2017; 74: 169
        • Blount L.G.
        Poor women suffer most from restrictive abortion policies. Reproductive Health Reality Check.
        (Available:)
        • Bongaarts J.
        • Westoff C.
        The potential role of contraception in reducing abortion..
        Studies in Family Planning. 2000; 31: 193-202
        • Boonstra H.D.
        Abortion in the lives of women struggling financially: Why insurance coverage matters.
        Guttmacher Policy Review. 2016; 19: 7
        • Centers for Disease Control and Prevention
        Pregnancy Mortality Surveillance System. Reproductive Health.
        (Available:)
        www.cdc.gov/
        Date: 2019
        Date accessed: September 1, 2021
        • Chelko O.
        • Rochat R.
        • Herold J.M.
        • Carter J.
        • Lavoie M.
        An evaluation of abortion data and abortion reporting in Georgia.
        Conference Oral Presentation at the Annual Meeting of the American Public Health Association, Atlanta, GA2001
        • Coles M.S.
        • Makino K.K.
        • Stanwood N.L.
        • Dozier A.
        • Klein J.D.
        How are restrictive abortion statutes associated with unintended teen birth?.
        Journal of Adolescent Health. 2010; 47: 160-167
        • Cook P.J.
        • Parnell A.M.
        • Moore M.J.
        • Pagnini D.
        The effects of short-term variation in abortion funding on pregnancy outcomes [Working Paper].
        National Bureau of Economic Research. 1996; 18: 241-257
        • Crenshaw K.
        Demarginalizing the Intersection of Race and Sex: A Black feminist critique of antidiscrimination doctrine, feminist theory and antiracist politics.
        University of Chicago Legal Forum, 1989: 139-169
        • Davis A.
        Racism, birth control, and reproductive rights.
        in: Feminist postcolonial theory: A reader. Routledge, New York, NY2003: 353-367
        • Dehlendorf C.
        • Harris L.H.
        • Weitz T.A.
        Disparities in abortion rates: A public health approach.
        American Journal of Public Health. 2013; 103: 1772-1779
        • Department of Labor Women’s Bureau
        Unemployment rates.
        U.S. Department of Labor, 2020 (Available:)
        • Finer L.B.
        • Zolna M.R.
        Declines in unintended pregnancy in the United States, 2008–2011.
        New England Journal of Medicine. 2016; 374: 843-852
        • Foster D.G.
        Dramatic decreases in US abortion rates: Public health achievement or failure?.
        American Journal of Public Health. 2017; 107: 1860-1862
        • Foster D.G.
        • Biggs M.A.
        • Ralph L.
        • Gerdts C.
        • Roberts S.
        • Glymour M.M.
        Socioeconomic outcomes of women who receive and women who are denied wanted abortions in the United States.
        American Journal of Public Health. 2018; 108: 407-413
        • Foster D.G.
        • Kimport K.
        Who seeks abortions at or after 20 weeks?.
        Perspectives on Sexual and Reproductive Health. 2013; 45: 210-218
        • Frey W.H.
        Six maps that reveal America’s expanding diversity.
        Brookings Institute, 2019 (Available:)
        • Georgia Department of Public Health
        Reducing infant mortality Georgia 2013 annual report.
        (Available:)
        • Georgia Department of Public Health
        Induced terminations of pregnancy web query. Online analytical statistical information system.
        (Available:)
        https://oasis.state.ga.us/oasis/webquery/qryITOP.aspx
        Date: 2020
        Date accessed: October 12, 2021
        • Georgia Department of Public Health
        Maternal child health: Birth web query. Online analytical statistical information system.
        (Available:)
        https://oasis.state.ga.us/oasis/webquery/qryBirth.aspx
        Date: 2020
        Date accessed: September 1, 2021
        • Georgia Department of Public Health
        Welcome to the Georgia Vital Events Registration System! Division of Vital Records.
        (Available:)
        https://gavers.dph.ga.gov
        Date: 2021
        Date accessed: September 1, 2021
        • Georgia Department of Public Health
        About vital records.
        (Available:)
        https://dph.georgia.gov/about-vital-records
        Date: 2021
        Date accessed: October 12, 2021
        • Georgia Department of Public Health
        PHIP data request.
        (Available:)
        https://dph.georgia.gov/phip-data-request
        Date: 2021
        Date accessed: October 12, 2021
        • Gerdts C.
        • Fuentes L.
        • Grossman D.
        • White K.
        • Keefe-Oates B.
        • Baum S.E.
        • Potter J.E.
        Impact of clinic closures on women obtaining abortion services after implementation of a restrictive law in Texas.
        American Journal of Public Health. 2016; 106: 857-864
        • Gerdts C.
        • Hudaya I.
        Quality of Care in a Safe-Abortion Hotline in Indonesia: Beyond Harm Reduction..
        American Journal of Public Health. 2016; 106: 2071-2075
        • Grossman D.
        • Holt K.
        • Peña M.
        • Lara D.
        • Veatch M.
        • Córdova D.
        • Blanchard K.
        Self-induction of abortion among women in the United States.
        Reproductive Health Matters. 2010; 18: 136-146
        • Grossman D.
        • Ralph L.
        • Raifman S.
        • Upadhyay U.D.
        • Gerdts C.
        • Biggs M.A.
        • Foster D.G.
        Lifetime prevalence of self-induced abortion among a nationally representative sample of U.S. women.
        Contraception. 2018; 97: 460
        • Grossman D.
        • White K.
        • Fuentes L.
        • Hopkins K.
        • Stevenson A.
        • Yeatman S.
        • Potter J.E.
        Knowledge, opinion, and experience related to abortion self-induction in Texas. Texas Policy Evaluation Project.
        (Available:)
        • Guttmacher Institute
        State bans on abortion throughout pregnancy. State laws and policies.
        Guttmacher Institute, New York2019 (Available:)
        • Guttmacher Institute
        Medication Abortion.
        Guttmacher Institute. New York, New York2021 (Available:)
        • Hall K.S.
        • Redd S.
        • Narasimhan S.
        • Mosley E.A.
        • Hartwig S.A.
        • Lemon E.
        • Cwiak C.
        Abortion trends from 2007 to 2017 following enactment of Georgia’s 22-week gestational age limit.
        American Journal of Public Health. 2020; 110: 1034-1038
      1. HB954. House Bill 954; General Assembly 2011-2012 Ed 2012.

        • Healthy Mothers, Healthy Babies Coalition of Georgia
        State of the state report. State of the state report.
        Healthy Mothers, Healthy Babies Coalition of Georgia, Atlanta, Georgia2019 (Available:)
        • Jones R.K.
        • Finer L.B.
        Who has second-trimester abortions in the United States?.
        Contraception. 2012; 85: 544-551
        • Jones R.K.
        • Jerman J.
        Abortion incidence and service availability in the United States, 2014.
        Perspectives on Sexual and Reproductive Health. 2017; 49: 17-27
        • Jones R.K.
        • Witwer E.
        • Jerman J.
        Abortion incidence and service availability in the United States, 2017.
        Guttmacher Institute, New York2019
        • Kivisto A.J.
        • Mills S.
        • Elwood L.S.
        Racial disparities in pregnancy-associated intimate Partner Homicide.
        Journal of Interpersonal Violence. 2021; (886260521990831)
        • Luna Z.
        • Luker K.
        Reproductive justice.
        Annual Review of Law and Social Science. 2013; 9: 327-352
        • Masinter L.M.
        • Feinglass J.
        • Simon M.A.
        Pregnancy intention and use of contraception among Hispanic women in the United States: Data from the National Survey of Family Growth, 2006–2010.
        Journal of Women’s Health. 2013; 22: 862-870
        • Moseson H.
        • Herold S.
        • Filippa S.
        • Barr-Walker J.
        • Baum S.E.
        • Gerdts C.
        Self-managed abortion: A systematic scoping review.
        Best Practice & Research Clinical Obstetrics & Gynaecology. 2020; 63: 87-110
        • Mosher W.D.
        • Jones J.
        • Abma J.C.
        Intended and unintended births in the United States: 1982–2010. National health statistics reports, 55.
        National Center for Health Statistics, Hyattsville, MD2012
        • Mosley E.A.
        • Ayala S.
        • Jah Z.
        • Hailstorks T.
        • Scales M.
        • Gutierrez M.
        • Hall K.S.
        ‘All the things that make life harder’: Intersectionality and medication abortion in Georgia.
        Contraception. 2020; 102: 280
        • Nash E.
        State abortion policy landscape: From hostile to supportive.
        Guttmacher Institute, 2019 (Available:)
        • Ralph L.
        • Foster D.G.
        • Raifman S.
        • Biggs M.A.
        • Samari G.
        • Upadhyay U.
        • Grossman D.
        Prevalence of self-managed abortion among women of reproductive age in the United States.
        JAMA Network Open. 2020; 3: e2029245
        • Roberts D.
        Killing the Black body: Race, reproduction, and the meaning of liberty.
        Random House LLC, New York1997
        • Roberts S.C.
        • Biggs M.A.
        • Chibber K.S.
        • Gould H.
        • Rocca C.H.
        • Foster D.G.
        Risk of violence from the man involved in the pregnancy after receiving or being denied an abortion.
        BMC Medicine. 2014; 12: 1-7
        • Roberts S.
        • Gould H.
        • Upadhyay U.D.
        Implications of Georgia’s 20-week abortion ban.
        American Journal of Public Health. 2015; 105: e77-82
        • Ross L.
        African-American women and abortion.
        in: Abortion wars: A half-century struggle. University of California Press, Berkeley, California1998
        • Ross L.
        Understanding reproductive justice.
        SisterSong, Atlanta, Georgia2006 (Available:)
        • Ross L.
        • Solinger R.
        Reproductive justice: An introduction. Vol. 1. University of California Press:Oakland, California, California2017
        • Schoen J.
        Choice & coercion: Birth control, sterilization, and abortion in public health and welfare.
        University of North Carolina Press Chapel Hill, North Carolina, 2005
        • Semega J.L.
        • Fontenot K.R.
        • Kollar M.A.
        Income and poverty in the United States: 2016. Current Population Reports.
        U.S. Census Bureau, Washington, DC2019 (Available:)
        • Shapiro R.
        • Erhardt-Ohren B.
        • Rochat R.
        Comparison of women from Georgia and contiguous states who obtained abortions in Georgia, 1994–2016.
        Maternal and Child Health Journal. 2020; 24: 299-309
        • Silliman J.
        • Fried M.G.
        • Ross L.
        • Gutiérrez E.R.
        Undivided rights: Women of color organize for reproductive justice.
        2nd ed. Haymaker Books, Chicago2016
        • Stern A.M.
        Sterilized in the name of public health: Race, immigration, and reproductive control in modern California.
        American Journal of Public Health. 2005; 95: 1128-1138
        • Upadhyay U.D.
        • Johns N.E.
        • Cartwright A.F.
        • Franklin T.E.
        Sociodemographic characteristics of women able to obtain medication abortion before and after Ohio’s law requiring use of the Food and Drug Administration protocol.
        Health Equity. 2018; 2: 122-130
        • Upadhyay U.D.
        • Weitz T.A.
        • Jones R.K.
        • Barar R.E.
        • Foster D.G.
        Denial of abortion because of provider gestational age limits in the United States.
        American Journal of Public Health. 2013; 105: 2557-2563
        • U.S. Census Bureau
        Historical living arrangements of children.
        2019 (Census.Gov. 2019. Available:)
        • U.S. Census Bureau
        United States Census: Explore data.
        2020 (Data.Census.Gov. 2020. Available:)
        https://data.census.gov
        Date accessed: October 12, 2021
        • U.S. Food and Drug Administration
        Misoprostol (marketed as Cytotec) Information.
        2015 (Available:)
        • West C.M.
        Black women and intimate partner violence: New directions for research.
        Journal of Interpersonal Violence. 2004; 19: 1487-1493

      Biography

      Elizabeth A. Mosley, PhD, MPH, is a public health demographer and full-spectrum doula. She completed a postdoctoral fellowship in family planning and reproductive health equity at Emory University with RISE, where she was a Co-Investigator on the Gestational Age Policies study.
      Sara K. Redd, PhD, MSPH, is a postdoctoral fellow in the Department of Health Policy and Management and with the RISE Center at the Emory University Rollins School of Public Health, where she is a Co-Investigator on the Gestational Age Policies study.
      Sophie A. Hartwig completed an MPH in Behavioral Sciences and Health Education, with a certificate in socio-contextual determinants of health. She serves as a Director of Research Projects (including the Gestational Age Policies study) and Co-Director of Administration at RISE.
      Subasri Narasimhan, PhD, MPH, is a public health social scientist and Assistant Professor in the Hubert Department of Global Health. She was a co-Investigator on the Gestational Age Policies study during her Postdoctoral Fellowship with the RISE Center.
      Emily Lemon, MPH, is a doctoral student in the Department of Behavioral, Social, and Health Education Sciences at the Emory University Rollins School of Public Health, and worked as a PhD-level research assistant with RISE.
      Erin Berry, MD, MPH, completed a residency in obstetrics/gynecology and fellowship in family planning. She is co-Investigator for the Gestational Age Policies study at RISE.
      Eva Lathrop, MD, MPH, completed a residency in obstetrics/gynecology and fellowship in family planning. She is co-Investigator for the Gestational Age Policies study at RISE.
      Lisa B. Haddad, MD, MS, MPH, completed a residency in obstetrics/gynecology and fellowship in family planning. Her clinical and research focus centers on reproductive health in high-risk populations. She is co-Investigator for the Gestational Age Policies study at RISE.
      Roger Rochat, MD, is a Professor of Global Health. He studied demography at Princeton, was Director of Reproductive Health at the CDC, and founded the Global Elimination of Maternal Mortality from Abortion Program at Emory. He is a Co-Investigator at RISE.
      Carrie Cwiak, MD, MPH, completed a residency in obstetrics/gynecology and fellowship in family planning. She is Division Director for Family Planning and Director of the Family Planning Fellowship Program at Emory, and co-PI for the Gestational Age Policies study at RISE.
      Kelli Stidham Hall, PhD, MS, BS, completed a postdoctoral fellowship in social epidemiology, demography, and reproductive health. She researches multilevel social determinants of reproductive health. She was the RISE Founding Director, and is co-PI for the Gestational Age Policies project.