Advertisement

The Impact of First-Person Abortion Stories on Community-Level Abortion Stigma: A Randomized Trial

Open AccessPublished:June 28, 2022DOI:https://doi.org/10.1016/j.whi.2022.06.006

      Abstract

      Objective

      We aimed to assess the impact of first-person abortion stories on community-level abortion stigma.

      Methods

      Between November 2018 and March 2019, we recruited participants and analyzed data from a nationally representative, probability-based online panel of U.S. adults, randomized to watch three first-person abortion video stories (intervention, n = 460) or three nature videos (control, n = 426). We measured community-level abortion stigma using the Community Abortion Attitudes Scale, Reproductive Experiences and Events Scale, and Community Level Abortion Stigma Scale at baseline, immediately after video exposure, and 3 months later. We dichotomized stigma change scores as decreased stigma compared with no change or increased stigma. Bivariate and logistic regression analysis accounted for complex survey methodology and sample weighting.

      Results

      Sample demographics reflected U.S. Census benchmarks (51% female, 68% White, 47% aged 18–44 years). Most participants (83.1%) completed the 3-month follow-up. Viewing the intervention videos was not associated with decreased stigma measured by Community Abortion Attitudes Scale or Community Level Abortion Stigma Scale immediately (odds ratio [OR], 0.80; 95% confidence interval [CI], 0.59–1.09; OR, 1.28; 95% CI, 0.93–1.75) or at the 3-month follow-up (OR, 0.86; 95% CI, 0.62–1.19; OR, 0.98; 95% CI, 0.70–1.37). Intervention exposure was associated with decreased stigma as measured by Reproductive Experiences and Events Scale immediately (OR, 1.74; 95% CI, 1.23–2.46); however, this association was not observed at the 3-month follow-up (OR, 0.98; 95% CI, 0.70–1.37).

      Conclusions

      Exposure to first-person video stories may not decrease community-level abortion stigma among U.S. adults.
      Although one in four women will have an abortion in their lifetime, abortion remains highly stigmatized (
      • Jones R.K.
      • Kavanaugh M.L.
      Changes in abortion rates between 2000 and 2008 and lifetime incidence of abortion.
      ). Like all forms of stigma, abortion stigma operates on multiple levels, including in individuals, communities, and institutions (
      • Kumar A.
      • Hessini L.
      • Mitchell E.M.H.
      Conceptualising abortion stigma.
      ). Community-level abortion stigma refers to a community's (or population's) negative attitudes, feelings, and anticipated behavior toward people who seek, have had, or provide abortions and their opinions toward policies that govern abortion (
      • Martin L.A.
      • Debbink M.
      • Hassinger J.
      • Youatt E.
      • Eagen-Torkko M.
      • Harris L.H.
      Measuring stigma among abortion providers: Assessing the Abortion Provider Stigma Survey instrument.
      ;
      • McMurtrie S.M.
      • García S.G.
      • Wilson K.S.
      • Diaz-Olavarrieta C.
      • Fawcett G.M.
      Public opinion about abortion-related stigma among Mexican Catholics and implications for unsafe abortion.
      ;
      • Sorhaindo A.M.
      • Karver T.S.
      • Karver J.G.
      • Garcia S.G.
      Constructing a validated scale to measure community-level abortion stigma in Mexico.
      ;
      • Weidner G.
      • Griffitt W.
      Abortion as a stigma: In the eyes of the beholder.
      ). Abortion stigma harms the health of individuals who have or seek abortions and contributes to health inequity and disparity (
      • Major B.
      • Gramzow R.H.
      Abortion as stigma: Cognitive and emotional implications of concealment.
      ;
      • Steinberg J.R.
      • Tschann J.M.
      • Furgerson D.
      • Harper C.C.
      Psychosocial factors and pre-abortion psychological health: The significance of stigma.
      ). Leading health organizations, including the American College of Obstetricians and Gynecologists and the American Public Health Association have highlighted the importance of reducing abortion stigma as part of a public health strategy to increase access to abortion care, improve health outcomes, and secure human rights and gender equity (
      American Public Health Association
      Restricted access to abortion violates human rights, precludes reproductive justice, and demands public health intervention.
      ;
      Committee on Health Care for Underserved Women
      ACOG Committee Opinion No. 613: Increasing access to abortion.
      ).
      Contact theory, which proposes that connecting stigmatized people with nonstigmatized people can break down stereotypes and other prejudices (
      • Pettigrew T.F.
      • Tropp L.R.
      A meta-analytic test of intergroup contact theory.
      ,
      • Pettigrew T.F.
      • Tropp L.R.
      Does intergroup contact reduce prejudice? Recent meta-analytic findings.
      ), has been used previously to decrease stigma toward several marginalized populations, such as persons with HIV, individuals with psychiatric illness, or members of the LGBTQI community (
      • Corrigan P.W.
      • Penn D.L.
      Lessons from social psychology on discrediting psychiatric stigma.
      ;
      • Herek G.M.
      • Capitanio J.P.
      “Some of my best friends” intergroup contact, concealable stigma, and heterosexuals’ attitudes toward gay men and lesbians.
      ). Deep canvassing—whereby activists engage with voters in two-way conversations to persuade opinion change—has demonstrated potential to increase support for several progressive social issues, but its demonstrated impact on the topic of abortion is mixed (
      • Kalla J.L.
      • Broockman D.E.
      The minimal persuasive effects of campaign contact in general elections: Evidence from 49 field experiments.
      ;
      • Kalla J.L.
      • Levine A.S.
      • Broockman D.
      Personalizing moral reframing in interpersonal conversation: A field experiment.
      ). Previous research has shown that contact theory may be applicable to abortion stigma in more intimate settings. A study by
      • Cockrill K.
      • Biggs A.
      Can stories reduce abortion stigma? Findings from a longitudinal cohort study.
      illustrated how both exposure to first-person essays about abortion in a book club setting and subsequent disclosure of an abortion from a group member resulted in more positive feelings toward people who have abortions and abortion providers. Knowing the abortion story of a partner or close friend also increases the likelihood of supporting abortion rights (
      • Cowan S.K.
      Secrets and misperceptions: The creation of self-fulfilling illusions.
      ;
      • Jones R.P.
      • Cox D.
      • Laser R.
      Committed to availability, conflicted about morality: What the millennial generation tells us about the future of the abortion debate and the culture wars.
      ).
      Data on the impact of abortion stories shared in less intimate settings are sparser. Nevertheless, in recent years, several reproductive health and rights organizations have promoted storytelling as a way to change people's opinions about abortion (
      • Cruz C.
      How women sharing their abortion stories are shattering stigma around the procedure.
      ). Answering these calls, individuals are sharing their stories in public forums—for instance, on social media platforms using hash tags such as #shoutyourabortion, #1in3, #youknowme—as well as in the halls of congress to demonstrate support for an increasingly threatened constitutional right (
      • Kissling E.
      From a whisper to a shout: How social media activism is challenging abortion stigma.
      ;
      • Stolberg S.G.
      A congresswoman's story: Raped at 17, “I chose to have an abortion.
      ). Media outlets increasingly promote the impact of storytelling on stigma reduction as promising, but there is a paucity of empirical data on the potential benefits and harms of publicly sharing abortion stories on community-level stigma (
      • Kissling E.
      From a whisper to a shout: How social media activism is challenging abortion stigma.
      ;
      • Stolberg S.G.
      A congresswoman's story: Raped at 17, “I chose to have an abortion.
      ).
      We sought to test whether viewing first-person abortion video stories (shared by people unfamiliar to participants) decreased community-level abortion stigma among the general public. We hypothesized that compared to watching a neutral, time-controlled video, watching a video of people describing the circumstances, considerations, and decision-making that led them to have an abortion would decrease community-level abortion stigma scores. We aimed to assess the impact of first-person abortion stories on community-level abortion stigma scores among a representative sample of adults in the United States. Secondary aims included assessing their impact on other aspects of stigma, including the degree of warmth toward common abortion scenarios and expectations of secrecy and privacy surrounding people's abortion experiences.

      Methods

      We conducted a randomized trial from November 30, 2018 through March 22, 2019, using the Ipsos KnowledgePanel, a nationally representative, probability-based online panel of U.S. adults (
      GfK Group
      Knowledge panel design summary.
      ). We included English-speaking male and female U.S. citizens, aged 18 years or older. To minimize subject awareness bias, the invitation did not mention the study's focus on abortion stigma, but rather that the study would assess opinions “about issues facing the country today … some of [which] are personal and sensitive.” After successfully passing an audiovisual test to ensure the videos could be viewed and heard, respondents completed a baseline pretest survey and were then randomized by a random number generator function to either intervention or control group. Yale University's Human Investigation Committee approved the study before recruitment. Respondents were offered Ipsos KnowledgePanel incentive points that could be redeemed for cash or other rewards.
      We chose the Community Abortion Attitudes Scale (CAAS)—a scale originally validated with a sample of more than 1,500 women seeking abortion in six states across the United States—to measure our primary outcome of community-level abortion stigma (
      • Cutler A.S.
      • Lundsberg L.S.
      • White M.A.
      • Stanwood N.L.
      • Gariepy A.M.
      Characterizing community-level abortion stigma in the United States.
      ). Our secondary outcomes included stigma scores as measured by the abortion subscale of the Reproductive Experiences and Events Scale (REES), which measures degrees of warmth toward various scenarios involving women who have abortions (
      • Cockrill K.
      • Biggs A.
      Can stories reduce abortion stigma? Findings from a longitudinal cohort study.
      ), and the silence subscale of the Community Level Abortion Stigma Scale (CLASS), which assesses expectations of silence and secrecy surrounding abortion experiences (
      • Shellenberg K.M.
      • Hessini L.
      • Levandowski B.A.
      Developing a scale to measure stigmatizing attitudes and beliefs about women who have abortions: Results from Ghana and Zambia.
      ). We measured all three stigma scores at three time points: at enrollment (baseline) just before video exposure, immediately after video exposure, and 3 months later (Figure 1). We chose these three scales to capture multiple dimensions of community-level abortion stigma: judgment (CAAS), context (REES), and secrecy/silence (CLASS) (
      • Cutler A.S.
      • Lundsberg L.S.
      • White M.A.
      • Stanwood N.L.
      • Gariepy A.M.
      Characterizing community-level abortion stigma in the United States.
      ).
      Figure thumbnail gr1
      Figure 1CONSORT diagram. CASS, Community Abortion Attitudes Scale; CLASS, Community Level Abortion Stigma Scale; REES, Reproductive Experiences and Events Scale.
      Ipsos fielded our study in two waves: at baseline/enrollment (November 30, 2018, through December 7, 2018) and 3 months later (March 8, 2019, through March 22, 2019). Ipsos provided data on respondents' age, gender, race/ethnicity, education, employment, relationship status, metropolitan status of residence, and geographic region. In the baseline wave, respondents in both groups were asked to complete the same questionnaire, comprising all three stigma scales (Supplemental Materials). Respondents randomized to the intervention then watched three first-person abortion video stories participants assigned to the control group watched three narrated nature videos. Respondents in both groups then immediately answered a questionnaire comprising the same three scales, as well as additional questions about the demographic characteristics of religious affiliation, political leanings, political party affiliation, voting practices, social media use, reproductive history (abortion, children, parenting experience, desire for children, miscarriage, infertility), proximity to someone who had an abortion, and sexual history (history of heterosexual intercourse, sexual assault, and intimate partner violence). Of the intervention group only, we also asked a series of questions (“process measures”—previously administered to five pilot participants to ensure understandability) that aimed to test whether participants thought our intervention videos in fact captured the qualities we intended them to capture (Supplemental Materials). In the 3-month follow-up wave, respondents were given the same questionnaire, comprising the three stigma scales as well as series of questions (“interim measures”) to explore potential exposure to any confounding events in the three months since baseline/enrollment (Supplemental Materials).
      To help us select three first-person abortion video stories to use as our intervention, we first convened a racially and ethnically diverse advisory committee comprising nine individuals with either professional or personal expertise in abortion stigma and storytelling. Among the committee members were four abortion storytellers who were compensated for their participation in the committee. All other committee members agreed to participate in kind as part of their professional work at their respective nonprofit organizations. The study intervention videos were chosen in a systematic manner with priority given to stories that centered the most common reasons for seeking an abortion, and storytellers who reflected the demographics of people who most commonly seek abortions and who highlighted how their intersecting identities were central to their experience. Because the aim of this study was not to design an intervention, but rather to test a phenomenon already occurring, we did not choose stories based on how well we thought they would sway opinion. We did, however, choose stories that advisory board members scored as authentic, trustworthy, and easy to follow and that held the potential to make participants think about abortion in a new way.
      Given that public support for abortion in the context of rape, incest, threats to maternal life, and fetal anomalies is already relatively high, we intentionally excluded stories about those topics. The final three videos selected for the trial featured a story by a Latina woman in the military whose previous efforts to obtain contraception had been dismissed by her superiors, a story by a Black woman who described the circumstances surrounding two different abortions and the ways they enabled her to have her desired family, and a story by a self-identified biracial woman who described the logistical barriers she faced when trying to obtain an abortion in Texas. The videos were edited to be 2 minutes each; care was taken to edit them in a way that did not substantially alter the original story. For the control group, we selected neutral, narrated, time-controlled videos sourced from the National Park Service to be used as the video exposure. The control videos were edited to be 2 minutes long each to match the length of the intervention videos.
      Because we had no published benchmark with which to power our study, we used the original CAAS study data (mean CAAS score, 1.114 ± 0.824; range, 0–4) to calculate our sample size (
      • Cockrill K.
      • Nack A.
      I’m not that type of person”: Managing the stigma of having an abortion.
      ). We aimed to allocate 394 per group to allow 80% power to detect a standard effect size of at least 0.2 (translating to a clinical change of 0.16 on the 5-point scale) and an alpha error of 0.05.
      We performed an exploratory factor or inter-reliability analysis on all measures to confirm their reliability in our sample. Detailed descriptions of survey sampling, participant recruitment, and post hoc weighting have been published previously (
      • Cutler A.S.
      • Lundsberg L.S.
      • White M.A.
      • Stanwood N.L.
      • Gariepy A.M.
      Characterizing community-level abortion stigma in the United States.
      ). To measure the effect of our intervention, we created a dichotomous variable of the change in stigma scores for CAAS, REES, and CLASS. Measures of change were calculated for the time periods of baseline (before the intervention) to immediately after the intervention and from baseline to 3-month follow-up. Dichotomized measure of scale score change was defined as decreased stigma or neutral/increased stigma. We examined decreased stigma as the outcome of interest. We used a bivariate analysis to examine the distribution of participant characteristics and intervention status using a χ2 test of association. We performed logistic regression modeling to evaluate the association between the intervention and decreased stigma for each dichotomized measure of scale change. Odds ratio (OR) and 95% confidence interval (CI) estimates were generated for unadjusted logistic models; additional adjustment was performed for baseline stigma scale scores. As sensitivity analyses, we ran linear regression models with continuous dependent variables; we modeled both baseline-immediate follow-up and baseline–3-month follow-up CAAS scores in which we defined our dependent variable as the continuous difference in those measures, and modeled all postintervention scores adjusted for baseline scores.
      Analyses were performed using SAS 9.4 (Statistical Analysis Software, Cary, NC) and applying a survey methods approach to account for specific sampling weight and nonresponse. All results are presented using weighted data.

      Results

      We fielded the first wave of our study in November 2018. A random sample of 1,800 individuals meeting eligibility criteria (ages ≥18 years, English-speaking, U.S. citizenship) was drawn from Ipsos's KnowledgePanel. Reminders to nonresponders were sent on day 4 and again on day 6 after the initial email invitation. Of the 1800 invited, 984 (excluding breakoffs) opened the study link and answered a screening question to assess technical ability to participate in the study (54.6% response rate). Of the 984, 886 (90% completion rate) completed the baseline assessment, which included a preintervention questionnaire, exposure to intervention or control, and an immediate postintervention questionnaire. Sixteen participants were excluded from these analyses owing to missing immediate postintervention CAAS data. Our final baseline analytic sample comprised 870 participants (Figure 1). In March 2019, we fielded the second wave of our study, a 3-month follow-up questionnaire sent to 875 of the original sample (excluding 11 individuals who had left the KnowledgePanel in the interim 3 months). Email reminders to nonresponders were sent on days 3, 6, and 9 of the field period. Of the 875 contacted, 736 (excluding breakoffs) responded (84.1% response rate) and 736 completed the 3-month follow-up survey (100% completion rate). One participant was excluded from analysis owing to missing CAAS data, yielding a final analytic sample of 735 (Figure 1). A sensitivity analysis comparing the weighted and unweighted data revealed no difference in the directionality or magnitude of the findings. All results presented use weighted data.
      Sample demographics reflected U.S. Census benchmarks (
      U.S. Census Bureau
      2018 American Community Survey 1-year estimates.
      ). The average participant age was 47.3 years and 51.0% were women, 69% identified as White, 12% as Black non-Hispanic, and 13% as Hispanic. Approximately one-third identified as Democrats (33%). Most identified as Protestant, Evangelical, or Catholic (59%); were currently registered to vote (87%); and reported proximity to someone who had an abortion (61% “know someone” and 71% “am or have been close to someone”). With the exception of gender, experimental groups were balanced across all demographic variables (Table 1).
      Table 1Sociodemographic, Political, and Pregnancy-Related Characteristics of Analytic Sample by Intervention and Control Status
      CharacteristicTotal (%)Intervention (%)Control (%)p Value
      (n = 870)(n = 455)(n = 415)
      Gender.01
       Male48.844.553.6
       Female51.257.046.4
      Gender identityNAC
       Cis99.298.899.7
       Trans0.50.70.3
       Other0.30.60.0
      Age (years), median (95% CI)47.3 (44.7–49.9)46.8 (43.5–50.3)47.9 (44.2–51.6).69
      Age (years).89
       18–2921.422.620.2
       30–4425.124.925.4
       45–5925.525.026.1
       ≥6028.027.628.4
      Race/ethnicity.73
       White, non-Hispanic68.770.167.1
       Black, non-Hispanic11.610.413.0
       Hispanic12.512.412.5
       Other, non-Hispanic5.96.75.6
       ≥2 races, non-Hispanic1.31.01.7
      MSA.94
       Non-MSA14.714.814.6
       MSA85.385.285.4
      Region.05
       Northeast18.017.718.3
       Midwest21.825.717.5
       South37.034.839.4
       West23.221.924.7
      Marital status.12
       Married or living with partner66.763.570.2
       Widowed or divorced or separated13.515.611.3
       Never married19.820.918.5
      Education.11
       Less than high school8.57.010.2
       High school28.026.929.2
       Some college28.832.325.0
       Bachelor's degree or higher34.733.835.7
      Employment> 0.99
       Working (paid employee or self-employed)63.163.063.2
       Not working (temporarily laid off, looking for work, disabled, other)18.518.518.5
       Retired18.418.518.3
      Head of household.93
       No23.023.122.8
       Yes77.076.977.2
      Household Income (US$).64
       <40,00024.723.426.3
       40,000–74,99923.123.522.7
       75,000–124,99925.727.323.9
       ≥125,00026.425.827.2
      Religion.84
       Catholic22.522.622.3
       Evangelical or Protestant36.436.436.5
       Other15.214.016.4
       No Religion25.926.825.0
      Religiosity - attend services….50
       Once or more a week28.526.830.4
       Less than weekly/more than yearly23.524.921.9
       Once a year or less15.614.716.5
       Never32.433.531.2
      Political identity.24
       Strong Republican or Republican24.425.723.0
       Independent/leans/other/no preference43.140.346.2
       Strong Democrat or Democrat32.534.030.7
      Political ideology.80
       Liberal29.930.928.8
       Moderate35.435.335.5
       Conservative34.733.935.6
      Voting practice
       Currently registered to vote86.587.385.5.51
       Voted in 201676.077.174.7.51
      Clinton47.249.644.4.40
      Trump38.936.641.4
      Other/didn't vote14.013.814.2
      Abortion
       I have had an abortion.7.96.79.1.22
       I know someone who has had an abortion.61.261.061.3.94
       I am or have been close to someone who has had an abortion.70.570.470.7.94
      Parenting (mutually exclusive categories)
       I have biological children.64.064.763.2.69
       I have nonbiological children.14.413.315.6.35
       I have other parenting experience.18.414.023.0.07
       I have ever wanted children.61.260.362.1.78
       I might someday want children.22.324.419.9.62
      Pregnancy
       Self or partner currently pregnant1.52.20.8.13
       Self or partner with history of pregnancy61.361.561.0.90
       Self or partner with history of infertility15.316.114.5.54
       Self or partner with history of miscarriage24.524.924.0.78
      Sexual activity
       I have had heterosexual intercourse.78.077.478.5.73
       Personal history of experiencing sexual misconduct.22.322.921.5.66
       Personal history of intimate partner violence.14.014.713.2.57
      Social media use (ever)
       Facebook76.676.776.4.94
       Instagram31.330.832.0.72
       Twitter23.223.223.2.99
      Abbreviations: MSA, Metropolitan Status Area; NAC, not able to calculate.

      Effect of the Intervention

      Exposure to the intervention was not associated with decreased stigma measured by CAAS (judgement) or CLASS (secrecy/silence), either immediately (OR, 0.80 [95% CI, 0.59–1.09] and OR, 1.28 [95% CI, 0.93–1.75], respectively) or at the 3-month follow-up (OR, 0.86 [95% CI, 0.62–1.19] and OR, 0.98 [95% CI, 0.70–1.37], respectively) compared with controls. Intervention exposure was associated with decreased stigma as measured by REES (context for decision-making) immediately after the intervention (OR, 1.74; 95% CI, 1.23–2.46); however, this association was not observed at the 3-month follow-up (OR, 0.98; 95% CI, 0.70–1.37) (Table 2). None of the sensitivity analyses yielded substantially different results; for example, exposure to the video intervention was not associated with continuous measures of CAAS scores and adjusting for baseline stigma scores also did not change the findings appreciably.
      Table 2Association of Video Intervention and Decreased Stigma as Measured by CAAS, REES, CLASS
      ScalesBaseline to Immediate-Post: Decreased Stigma (vs. Neutral/Increase)Baseline to 3-Month Follow-Up: Decreased Stigma (vs. Neutral/Increase)
      OR (95% CI)NOR (95% CI)N
      CAAS (judgement)0.80 (0.59–1.09)8700.86 (0.62–1.19)735
      REES (context)1.74 (1.23–2.46)5970.90 (0.59–1.38)419
      CLASS (secrecy/silence)1.28 (0.93–1.75)8670.98 (0.70–1.37)725
      Abbreviations: CASS, Community Abortion Attitudes Scale; CI, confidence interval; CLASS, Community Level Abortion Stigma Scale; OR, odds ratio; REES, Reproductive Experiences and Events Scale.
      We performed a stratified analysis by gender because the groups were not balanced by this variable and found no statistical difference in our findings. We did find an interesting signal suggesting that the (nonstatistical) effect of the intervention is stronger for men than for women and works in the opposite direction. In other words, the abortion video stories lead to a smaller decrease in stigma for men across all three measures and, unlike women, men were less likely to have decreased stigma at the 3-month follow-up (Table 3).
      Table 3Association of Video Intervention and Decreased Stigma as Measured by CAAS, REES, CLASS, Stratified by Gender
      ScalesBaseline to Immediate-Post: Decreased Stigma (vs. Neutral/Increase)Baseline to 3-Month Follow-Up: Decreased Stigma (vs. Neutral/Increase)
      OR (95% CI)OR (95% CI)
      Male
       CAAS (judgement)0.74 (0.48–1.14)0.68 (0.43–1.07)
       REES (context)1.62 (1.01–2.58)0.64 (0.37–1.12)
       CLASS (secrecy/silence)1.12 (0.72–1.73)0.83 (0.53–1.32)
      Female
       CAAS (judgement)0.83 (0.54–1.29)1.11 (0.69–1.81)
       REES (context)1.89 (1.13–3.18)1.29 (0.68–2.45)
       CLASS (secrecy/silence)1.45 (0.92–2.28)1.17 (0.71–1.92)
      Abbreviations: CASS, Community Abortion Attitudes Scale; CI, confidence interval; CLASS, Community Level Abortion Stigma Scale; OR, odds ratio; REES, Reproductive Experiences and Events Scale.

      Process Measures

      Participants in the intervention group overwhelmingly agreed or strongly agreed with statements affirming that our selection of stories had conveyed the intended qualities; namely, that the stories expressed common reasons/circumstances for having an abortion (72%), showed more than one element of the storytellers’ identities (76%), included information about how easy or hard it was to obtain an abortion (80%), were easy to follow from beginning to end (88%), and were true (65%) and that the storytellers were the central focus of their stories (73%), seemed authentic (78%), and seemed trustworthy (69%). Notably, only 18% of participants agreed or strongly agreed with the statement that the stories made them “think about abortion in a new way” (Supplemental Materials).

      Interim Measures

      We found no difference between the groups in exposure to potential confounding events (e.g., “Since, 2018; have you or your partner had an abortion?”) between the baseline and immediate follow-up or 3-month follow-up (Supplemental Materials).

      Discussion

      Our study provides evidence that a single exposure to three abortion video stories representing the most common demographics of people who have abortions and the most common reasons why people seek abortions, shared by people unknown to their audience, does not decrease community-level abortion stigma scores in a general population. These findings challenge the narrative that sharing stories about one's abortion experience with the general public—through mainstream media outlets, public service campaigns, or through viral social media posts—may lead to decreased community-level abortion stigma. This finding is important; although people who have had abortions may want to share their stories for any number of personal reasons, we lack evidence to suggest that doing so can change opinions among members of the general public. It is important that advocates of abortion rights know what the impact (or lack thereof) of sharing abortion stories with the public may be—especially in light of well-established risks associated with disclosure of concealed and stigmatized identities (
      • Cockrill K.
      • Nack A.
      I’m not that type of person”: Managing the stigma of having an abortion.
      ;
      • Hardon A.
      • Posel D.
      Secrecy as embodied practice: Beyond the confessional imperative.
      ).
      Our study has several strengths. This randomized trial used a large, nationally representative sample of U.S. adults, powered to detect a small effect in community-level abortion stigma change. The study used validated stigma measurement tools (
      • Cutler A.S.
      • Lundsberg L.S.
      • White M.A.
      • Stanwood N.L.
      • Gariepy A.M.
      Characterizing community-level abortion stigma in the United States.
      ). Our intervention was developed systematically and carefully by a diverse group of people with expertise in abortion stigma. We used process measures to test and confirm that our intervention videos reflected what we intended them to reflect. We collected many demographic and other characteristics of our study participants and additionally explored whether any events that happened in between the two study time points might have confounded our results. Our methodology ensured respondents’ anonymity and invitations to participate did not mention abortion, both of which decrease the likelihood of social desirability bias, confidentiality concerns, and differential participation owing to topic.
      Our sample was limited to English speakers and U.S. citizens, so is not generalizable to the entire U.S. population. Additionally, Ipsos KnowledgePanel participants are selected by random chance and include household members who do not have readily available internet access or cell phones. Despite these efforts and the fact that the panel is maintained to be representative of the U.S. population, it is possible that our study is limited by selection bias and over-recruitment of professional respondents who volunteer to participate in internet panels. Although we lack data on how many invited panelists received or opened the invitation, we had high completion and follow-up rates consistent with those of other KnowledgePanel research studies (
      • Bearak J.M.
      • Jones R.K.
      Did contraceptive use patterns change after the Affordable Care Act? A descriptive analysis.
      ;
      • Grossman D.
      • Grindlay K.
      • Li R.
      • Potter J.E.
      • Trussell J.
      • Blanchard K.
      Interest in over-the-counter access to oral contraceptives among women in the United States.
      ;
      • White K.
      • Potter J.E.
      • Stevenson A.J.
      • Fuentes L.
      • Hopkins K.
      • Grossman D.
      Women’s knowledge of and support for abortion restrictions in Texas: Findings from a statewide representative survey.
      ).
      In considering why we did not see decreased stigma scores after viewing the abortion video stories, it is also possible that the effect of storytelling on the general public is dose dependent, and that the general public needs a minimum amount and frequency of exposure before an effect is seen. Our study is inherently limited by the three specific videos we chose as our intervention. A different selection of videos comprising different storytellers—such as physicians or those whose racial and ethnic identities match those of the viewers’—might have yielded different results (
      • Seewald M.
      • Martin L.
      • Simon A.
      • Whaley J.
      • Javaid N.
      • Harris L.
      O7 The role of complexity and ambiguity in shaping abortion attitudes.
      ). Along similar lines, our study did not examine the impact of sharing these stories with known contacts, such as family members, friends, or social acquaintances. It is worth noting that sharing abortion stories in other settings—for example, with known acquaintances at a book club or through face-to-face deep canvassing—has been shown to warm people's opinions toward abortion (
      • Cockrill K.
      • Biggs A.
      Can stories reduce abortion stigma? Findings from a longitudinal cohort study.
      ;
      • Kalla J.L.
      • Levine A.S.
      • Broockman D.
      Personalizing moral reframing in interpersonal conversation: A field experiment.
      ). Additionally, forthcoming work suggests that sharing abortion stories informed by evidence-based messaging research can also increase support for abortion (
      • Seewald M.
      • Martin L.
      • Simon A.
      • Whaley J.
      • Javaid N.
      • Harris L.
      O7 The role of complexity and ambiguity in shaping abortion attitudes.
      ; Lisa Harris, MD, personal communication, 2021). Although we intentionally chose stories that involved unexceptional reasons for having an abortion, it is possible that stories involving circumstances known to garner higher levels of public support (e.g., rape, fetal anomalies) may have had a greater impact. We also cannot know how the identities of the storytellers, including race/ethnicity, affected our findings and our study was not designed to examine these factors. Given that people of color already face greater prejudice and stigmatization compared with their White counterparts as a result of racism, further understanding the role of racism and implicit bias in community-level abortion stigma is a crucial direction for future research in this field. Understanding the interactions between abortion stigma and racism is especially critical if a goal is to decrease barriers to abortion care for people of color, who have disproportionately high rates of abortion in our country and are thus disproportionately harmed the most by stigma (
      • Jones R.K.
      • Jerman J.
      Population group abortion rates and lifetime incidence of abortion: United States, 2008–2014.
      ;
      • Jones R.K.
      • Kavanaugh M.L.
      Changes in abortion rates between 2000 and 2008 and lifetime incidence of abortion.
      ). Finally, despite efforts to be comprehensive and to ensure that real-world events at the time did not affect results, our survey may not capture the full or current picture of community-level abortion stigma in the United States, especially given the rise of high-profile court cases challenging the right to abortion and the subsequent uprising of activism around abortion stigma reduction.
      Access to abortion care—an exceedingly common life event for individuals of reproductive age—is a fundamental public health right (
      American Public Health Association
      Restricted access to abortion violates human rights, precludes reproductive justice, and demands public health intervention.
      ). Several health organizations, including the American Public Health Association, call for reducing abortion stigma as one means of decreasing barriers to abortion care (
      • Espey E.
      • Dennis A.
      • Landy U.
      The importance of access to comprehensive reproductive health care, including abortion: A statement from women’s health professional organizations.
      ). Evidence-based stigma reduction interventions are crucial for reproductive rights and public health advocates who seek to sway public opinion on abortion-related policy, for abortion care providers who endeavor to decrease abortion stigma toward their patients and themselves, for individuals who seek abortion care, and for their loved ones and friends who help them to navigate that experience. Despite the growing popularity of abortion storytelling, this randomized trial does not offer evidence that sharing first-person abortion stories with the general public decreases community-level abortion stigma.

      Implications for Policy and/or Practice

      As state houses prepare for the possible fall of Roe v Wade, which currently constitutes the constitutional and federal right to abortion, it is imperative to know whether abortion storytelling is effective at changing the opinions of voters and legislators. Our findings suggest that abortion rights advocates should seek to know more about the relationship between storytelling and stigma before it is promoted as an effective intervention at the community level.

      Acknowledgments

      The authors thank our advisory committee members who offered invaluable expertise from their professional and lived experience, including Kate Cockrill, Renee Bracey Sherman, Julia Reticker-Flynn, Amelia Bonow, Sara Edwards, Holly Bland, Kenya Martin, Stephanie Piñeiro, and Jack Qu'emi Gutiérrez. We are also grateful to The Guardian, Public Square Media, and the National Park Service for use of their videos, and to the people whose stories we used for our intervention.

      References

        • American Public Health Association
        Restricted access to abortion violates human rights, precludes reproductive justice, and demands public health intervention.
        (Available:)
        • Bearak J.M.
        • Jones R.K.
        Did contraceptive use patterns change after the Affordable Care Act? A descriptive analysis.
        Women’s Health Issues. 2017; 27: 316-321
        • Cockrill K.
        • Biggs A.
        Can stories reduce abortion stigma? Findings from a longitudinal cohort study.
        Culture, Health, Sexuality. 2018; 20: 1-16
        • Cockrill K.
        • Nack A.
        I’m not that type of person”: Managing the stigma of having an abortion.
        Deviant Behavior. 2013; 34: 973-990
        • Committee on Health Care for Underserved Women
        ACOG Committee Opinion No. 613: Increasing access to abortion.
        Obstetrics and Gynecology. 2014; 124: 1060-1065
        • Corrigan P.W.
        • Penn D.L.
        Lessons from social psychology on discrediting psychiatric stigma.
        American Psychologist. 1999; 54: 765-776
        • Cowan S.K.
        Secrets and misperceptions: The creation of self-fulfilling illusions.
        Sociological Science. 2014; 1: 466-492
        • Cruz C.
        How women sharing their abortion stories are shattering stigma around the procedure.
        (Available:)
        • Cutler A.S.
        • Lundsberg L.S.
        • White M.A.
        • Stanwood N.L.
        • Gariepy A.M.
        Characterizing community-level abortion stigma in the United States.
        Contraception. 2021; 104: 305-313
        • Espey E.
        • Dennis A.
        • Landy U.
        The importance of access to comprehensive reproductive health care, including abortion: A statement from women’s health professional organizations.
        American Journal of Obstetrics and Gynecology. 2019; 220: 67-70
        • Grossman D.
        • Grindlay K.
        • Li R.
        • Potter J.E.
        • Trussell J.
        • Blanchard K.
        Interest in over-the-counter access to oral contraceptives among women in the United States.
        Contraception. 2013; 88: 544-552
        • GfK Group
        Knowledge panel design summary.
        Author, Palo Alto, CA2013
        • Hardon A.
        • Posel D.
        Secrecy as embodied practice: Beyond the confessional imperative.
        Culture, Health, Sexuality. 2012; 14: S1-S13
        • Herek G.M.
        • Capitanio J.P.
        “Some of my best friends” intergroup contact, concealable stigma, and heterosexuals’ attitudes toward gay men and lesbians.
        Personality Social Psychology Bulletin. 1996; 22: 412-424
        • Jones R.K.
        • Jerman J.
        Population group abortion rates and lifetime incidence of abortion: United States, 2008–2014.
        American Journal of Public Health. 2017; 107: 1904-1909
        • Jones R.K.
        • Kavanaugh M.L.
        Changes in abortion rates between 2000 and 2008 and lifetime incidence of abortion.
        Obstetrics and Gynecology. 2011; 117: 1358-1366
        • Jones R.P.
        • Cox D.
        • Laser R.
        Committed to availability, conflicted about morality: What the millennial generation tells us about the future of the abortion debate and the culture wars.
        Public Religion Research Institute, Washington, DC2011
        • Kalla J.L.
        • Broockman D.E.
        The minimal persuasive effects of campaign contact in general elections: Evidence from 49 field experiments.
        American Political Science Review. 2018; 112: 148-166
        • Kalla J.L.
        • Levine A.S.
        • Broockman D.
        Personalizing moral reframing in interpersonal conversation: A field experiment.
        Journal of Politics. 2021; 84: 1239-1243
        • Kissling E.
        From a whisper to a shout: How social media activism is challenging abortion stigma.
        Watkins Media Limited, London2018
        • Kumar A.
        • Hessini L.
        • Mitchell E.M.H.
        Conceptualising abortion stigma.
        Culture, Health, Sexuality. 2009; 11: 625-639
        • Major B.
        • Gramzow R.H.
        Abortion as stigma: Cognitive and emotional implications of concealment.
        Journal of Personality and Social Psychology. 1999; 77: 735-745
        • Martin L.A.
        • Debbink M.
        • Hassinger J.
        • Youatt E.
        • Eagen-Torkko M.
        • Harris L.H.
        Measuring stigma among abortion providers: Assessing the Abortion Provider Stigma Survey instrument.
        Women Health. 2014; 54: 641-661
        • McMurtrie S.M.
        • García S.G.
        • Wilson K.S.
        • Diaz-Olavarrieta C.
        • Fawcett G.M.
        Public opinion about abortion-related stigma among Mexican Catholics and implications for unsafe abortion.
        International Journal of Gynaecology and Obstetrics. 2012; 118: S160-S166
        • Pettigrew T.F.
        • Tropp L.R.
        Does intergroup contact reduce prejudice? Recent meta-analytic findings.
        Reducing Prejudice and Discrimination. 2000; : 93-114
        • Pettigrew T.F.
        • Tropp L.R.
        A meta-analytic test of intergroup contact theory.
        Journal of Personality and Social Psychology. 2006; 90: 751-783
        • Seewald M.
        • Martin L.
        • Simon A.
        • Whaley J.
        • Javaid N.
        • Harris L.
        O7 The role of complexity and ambiguity in shaping abortion attitudes.
        Contraception. 2020; 102: 275
        • Shellenberg K.M.
        • Hessini L.
        • Levandowski B.A.
        Developing a scale to measure stigmatizing attitudes and beliefs about women who have abortions: Results from Ghana and Zambia.
        Women Health. 2014; 54: 599-616
        • Sorhaindo A.M.
        • Karver T.S.
        • Karver J.G.
        • Garcia S.G.
        Constructing a validated scale to measure community-level abortion stigma in Mexico.
        Contraception. 2016; 93: 421-431
        • Steinberg J.R.
        • Tschann J.M.
        • Furgerson D.
        • Harper C.C.
        Psychosocial factors and pre-abortion psychological health: The significance of stigma.
        Social Science & Medicine. 2016; 150: 67-75
        • Stolberg S.G.
        A congresswoman's story: Raped at 17, “I chose to have an abortion.
        (Available:)
        www.nytimes.com/2021/09/30/us/politics/cori-bush-abortion.html
        Date: 2021
        Date accessed: December 21, 2021
        • U.S. Census Bureau
        2018 American Community Survey 1-year estimates.
        (Available:)
        • Weidner G.
        • Griffitt W.
        Abortion as a stigma: In the eyes of the beholder.
        Journal of Research in Personality. 1984; 18: 359-371
        • White K.
        • Potter J.E.
        • Stevenson A.J.
        • Fuentes L.
        • Hopkins K.
        • Grossman D.
        Women’s knowledge of and support for abortion restrictions in Texas: Findings from a statewide representative survey.
        Perspectives on Sexual and Reproductive Health. 2016; 48: 189-197

      Biography

      Abigail S. Cutler, MD, MPH, is an Assistant Professor at the University of Wisconsin School of Medicine and Public Health where she practices general obstetrics and gynecology and complex family planning. Her research interests include abortion stigma, reproductive decision-making and expanding health equity through increased access to abortion care.
      Lisbet S. Lundsberg, PhD, MPH, is an Associate Research Scientist and perinatal epidemiologist. Her research interests include patient-centered issues in family planning and evaluation of antenatal risk factors with maternal and neonatal morbidity.
      Nancy L. Stanwood, MD, MPH (she/her), is a Complex Family Planning specialist and Chief Medical Officer of Planned Parenthood of Southern New England with interests in access to abortion care and decreasing health disparities.
      Marney A. White, PhD, MS, is a clinical psychologist and epidemiologist at the Yale School of Public Health. Her clinical and research focus is in eating disorders and women's health.
      Aileen M. Gariepy, MD, MPH, MHS, is the Director of Complex Family Planning, Department of Obstetrics and Gynecology, at Weill Cornell Medicine. Her research focuses on developing and testing innovative, multi-faceted strategies for decreasing unintended pregnancy in marginalized populations.