Exploring Attitudes About the Legality of Self-Managed Abortion in the US: Results from a Nationally Representative Survey

Twenty-four states have at least one law in place that could be used to prosecute people for self-managed abortion (SMA), or the termination of a pregnancy outside of the formal healthcare system. We investigated factors associated with public attitudes about SMA legality and legal access to abortion more generally. In August 2017, we surveyed a nationally representative sample of English- and Spanish-speaking women ages 18–49 years in the United States (US) using Ipsos Public Affairs’ KnowledgePanel. Unadjusted and adjusted multinomial logistic regression estimates identify characteristics associated with believing that SMA should not be against the law, compared to should be against the law, with weighting to account for sampling into the panel. Overall, 76% (95% CI: 74.3%-77.1%) and 59% (95% CI: 57.3%-60.4%) of participants (n = 7,022, completion rate 50%) reported that abortion and SMA, respectively, should not be against the law; 1% and 19% were unsure. Among those living in a state with at least one law that could be used to prosecute an individual for SMA, the majority (55%, 95% CI: 52.7%-57.9%) believed SMA should not be against the law. Factors associated with believing SMA should not be against the law, compared to should be against the law, included prior abortion experience and higher levels of education and income. Most reproductive age women in the US believe that SMA should not be criminalized. There is more uncertainty about SMA legality than about the legality of abortion more generally. US laws that criminalize SMA are not supported by the majority of the people living in their jurisdictions.


Introduction
Rapid changes in the United States (US) abortion landscapeincluding tightening of facility-based abortion access and increasing availability of medication abortion pills outside of the formal healthcare system-have sparked a newfound interest in what is referred to as self-managed abortion (SMA) or self-induced abortion. In recent years, discussions about SMA have extended beyond clinical and research communities and into the mainstream media and policy arenas (Aiken et al., 2018;Astor, 2019). Renewed attention to the topic raises many questions about the prevalence of SMA (Ralph et al., 2020), common methods used for attempting SMA, safety and effectiveness of SMA, and whether attempting to terminate a pregnancy on one's own should be against the law.
In this study, we define SMA as any attempt to end a pregnancy on one's own without the supervision of a clinician. This is in contrast to obtaining an abortion at a clinic or facility and in contrast to inducing an abortion at home by taking mifepristone and misoprostol prescribed by a registered provider. SMA may include self-sourcing mifepristone and misoprostol or misoprostol alone (online or in person), ingesting herbs or plants (such as parsley) or vitamins/supplements (such as vitamin C), taking over-the-counter pain relievers in high doses, using drugs and alcohol, or attempting physical harm, among other methods (Aiken et al., 2018;Moseson et al., 2020;Grossman et al., 2010).
Since the Supreme Court ruled in Roe v. Wade (1973), states have passed over 1,100 abortion restrictions, with more than one-third (34%) of these laws occurring in the last decade, specifically between 2011 and 2017 (Nash et al., 2018). As of 2019, 24 states have at least one law in place that could be used to prosecute people who attempt SMA or help others to do so (If/When/How, 2019). These include explicit bans on SMA (in six states), laws criminalizing harm to fetuses (in 10 states), and criminal abortion laws (in 14 states). Since 2000, at least 21 people have been arrested for SMA, and criminal investigations have occurred in 20 states (If/When/ How, 2019). The criminalization of SMA is likely to have disproportionate impacts on people of color and those living on low incomes, who face both larger barriers to abortion access within the formal healthcare setting and the strictest policing in the US (Flavin, 2008;Jones & Jerman, 2017;Robinson, 2017). Given the responsibility of elected officials to respond to the interests of their constituents, public opinion plays an important role in shaping public policy, including policies related to abortion and SMA (Norrander & Wilcox, 1999). In 2020, one-quarter of voters said they would vote only for a candidate who shares their views on abortion (Brenan, 2020). It is therefore important to understand and document what the public thinks about the legality of abortion and SMA, and whether current laws concur or conflict with public opinion.
Little is understood about what the public thinks about the legality of SMA. One representative survey conducted in Texas found that 26% of women said SMA should not be against the law, 14% believed SMA should be against the law, and the remainder fell somewhere in between (Grossman et al., 2015). To our knowledge, no previous studies have investigated predictors of SMA attitudes, particularly at a national level. There is, however, a wealth of data on predictors of abortion attitudes more generally (Adamczyk et al., 2020). National polling data suggest that 49% of US residents support access to abortion for any reason (Smith et al., 2018) and 61% believe abortion should be legal in all or most cases (Pew Research Center, 2019). The General Social Survey, a biennial, nationally representative survey conducted by NORC at the University of Chicago and designed to monitor changes in social characteristics and attitudes in the US, finds little difference between men and women in their attitudes toward abortion, with men being slightly more supportive of legalization (Smith & Son, 2013). Support for legal abortion is associated with higher levels of education and better overall health status. Those over age 65 are the least supportive, and those middle aged (35-49) are the most supportive, followed by those under 35 (Smith & Son, 2013). Studies also show that higher income and non-religious individuals are more supportive of legal access to abortion compared to their counterparts (Hoffmann & Johnson, 2005;Semega et al., 2016). African Americans were less likely than whites to support legal abortion in the two decades after Roe v Wade (1973); however, this gap later narrowed and was shown to be primarily attributable to differences in religiosity and education levels (Jelen & Wilcox, 2005). By 2017, African Americans were more likely than whites (62% versus 58%) to support legal access to abortion in all or most cases (Pew Research Center, 2019). Hispanics are less likely than non-Hispanics to support legal access to abortion (Smith et al., 2018), though support varies by place of birth with US-born Hispanics more likely to be supportive of legal access to abortion than foreign-born Hispanics (Pew Research Center, 2014). Women's own reproductive experiences have also been shown to influence their views on abortion; for example, The Turnaway Study showed that women denied an abortion were significantly less likely to support the legal right to abortion after abortion-seeking than women who had received an abortion (Woodruff et al., 2018).
Prior research also demonstrates that attitudes toward abortion are multi-faceted; people may think differently about the morality and legality of abortion, and this distinction is not always clear-cut (Henshaw & Martire, 1982;Woodruff et al., 2018). Consistently, evidence shows that people shift their position on whether an abortion should be permitted based on various circumstances, including clinical and social factors (Hans & Kimberly, 2014;Smith & Son, 2013). It is also possible that the factors that predict abortion attitudes in general may not necessarily be the same as those that predict attitudes about SMA. For example, one's attitudes about SMA may be more likely to be influenced by perspectives about privacy, lack of trust in clinicians, demedicalization of health care services, and prior experiences seeking and obtaining reproductive health services within and outside of the formal healthcare setting-all factors that have been previously documented as reasons individuals attempt SMA (Aiken et al., 2019(Aiken et al., , 2018Ralph et al., 2020). Perspectives about these factors may further vary across sociodemographic and geographic characteristics.
In this exploratory study, we aimed to (1) provide a national estimate of the proportion of reproductive age women living in the US who think SMA should not be against the law (compared to should be against the law), (2) identify factors associated with attitudes about the legality of SMA, and (3) assess how these factors compare to those associated with attitudes about the legality of abortion more generally. In the present study, we focus on attitudes about legality specifically, rather than attitudes about morality, because legality is more specific and relatively easier to measure, and because we recognize that public opinion can inform policymaking (Norrander & Wilcox, 1999). Better understanding the public's attitudes about the legality of SMA could help to inform new and existing legislation related to SMA, particularly in a context where people are facing increasing barriers to facility-based abortion and legal consequences for attempting or helping others to attempt SMA.

Sample
Data for this study were collected in August 2017 from female-identified members of the web-based Knowl-edgePanel®, a nationally representative, probabilitybased online household panel administered by GfK (currently known as Ipsos Public Affairs). GfK selects panel members using a probability sample of all noninstitutionalized US addresses from the latest Delivery Sequence File of the US Postal Service (a database with full coverage of all delivery points in the US). The administrators initially invite adults from sampled households to join the panel through a series of mailings, including an invitation letter, a reminder postcard, and a subsequent follow-up letter. GfK regularly invites all panel members to participate in survey activities online on a range of topics and provides them with the technology to do so as needed (Ipsos, 2020).
After pilot-testing our survey with 25 participants to assess functionality and confirm survey length, GfK invited eligible panel members to complete our 53-item survey on "women's experiences and opinions related to reproductive health care." We designed the survey to estimate the prevalence of SMA in the US and added exploratory survey questions about abortion and SMA attitudes adapted from a previous study estimating attitudes of Texas residents (Grossman et al., 2015). Eligibility criteria required participants to be self-identified female, 18 to 49 years old, residing in the US, and have English or Spanish language proficiency. GfK invited eligible participants to participate in the survey over the course of two weeks in August 2017.
The analytic sample included a total of 7,022 participants who completed the online survey, or 50% of the 14,151 women approached in the panel. GfK sent automatic email reminders to non-responders three and eight days following initial contact. Responders completed the survey in 11 minutes on average. Participants were reimbursed for their participation through the panel's points program, where panelists receive cash-equivalent checks ranging from $4 to $6 per month depending on their level of participation in the panel. The University of California, San Francisco, Institutional Review Board approved all study activities.

Measures
The primary outcome of interest was attitudes about the legality of SMA, measured by the following survey question designed for this study: "What is your opinion of women trying to end a pregnancy without medical assistance?" There were six response options to the SMA question (see Table 1).
The following text, developed using cognitive interviews (Moseson et al., 2019), was included earlier in the survey to provide a definition for SMA: Women make different choices about how to end an unwanted pregnancy. Some women may go to a hospital, clinic, or doctor's office to have an abortion. Other women may do something to try to end a pregnancy without medical assistance. For example, they may get information from the internet, a friend, or family member about pills, medicine, or herbs they can take on their own, or they may do something else to try to end the pregnancy.
Our secondary outcome of interest was attitudes about the legality of abortion more generally, for which we asked respondents: "Which of the following statements about the issue of abortion comes closest to your own view?" There were four response options to the abortion question. For analysis, we collapsed responses to both questions (about SMA and abortion respectively) into four categories: should be against the law, should not be against the law, unsure, and other; see Table 1.
For both SMA and abortion, those who initially marked "Other" were given the opportunity to provide open text responses. Based on text responses, these individuals were re-categorized into one of the four main outcome categories (should be against the law, should not be against the law, unsure, or other) for analysis. Four researchers independently reviewed and coded a subset of respondents who initially answered "Other" into one of the four outcome categories above, then met to discuss and refine coding criteria. After finalizing coding criteria, the primary coder (SR) reviewed and coded all "Others" and secondary coders (MAB, LR, KE) each coded one-third of the "Others." If the primary and respective secondary coder agreed, then re-categorization was complete. If they disagreed, then a third coder blindly reviewed and coded the text: if two out of three coders agreed, the re-coding was final, but if all three coders disagreed, then the coders discussed until they came to an agreement. There was 90.4% agreement on the coding of SMA open-text responses, with a kappa of 0.81.
Panel participants provide socio-demographic information annually to GfK. We included the following variables in this analysis, based on previous studies of factors predicting abortion attitudes (Smith et al., 2018;Smith & Son, 2013): age, race/ethnicity/language, education, marital status, religion, religiosity, and income. We used a three-part categorical federal poverty level (FPL) variable based on the US 2016 Census Bureau's thresholds (Semega et al., 2016), household composition, and household income (100% FPL, 100-199% FPL, and ≥ 200% FPL). Although panel members were asked about their place of birth and citizenship status, 6% of participants in our sample had not yet been administered this question and were missing responses. Language is highly related to acculturation; therefore, we decided to assess the language in which participants chose to complete the survey (English or Spanish). We accounted for policy environment by including a binary variable for whether the participant lived in a state with a law that could be used to criminalize SMA (If/When/How, 2019) and for political party affiliation (Republican, undecided/independent/other, or Democrat). We also included lifetime abortion experience, categorized as no experience, any SMA, previous medication abortion (and no SMA), or previous surgical abortion (and no SMA or medication abortion). These categories were chosen based on the assumption that one's own experience with abortion and SMA may influence their perspectives on the legality of abortion and SMA (Woodruff et al., 2018). We hypothesized that those with medication abortion or SMA experience would be more knowledgeable about abortion medications and SMA methods, whereas those with surgical abortion experience only or no abortion experience may assume that all abortion includes surgery and is therefore dangerous. Those who had previous experience with both medication and surgical abortion were categorized in the medication abortion group. Women who reported only using emergency contraception prior to confirming pregnancy as an SMA method were not considered to have SMA experience.

Analysis
During the process of re-categorizing the "Other" survey responses to the four main outcome categories, coders made note of key themes considered by participants in deciding whether SMA and abortion should or should not be against the law. We organized and summarized coder notes on the themes that emerged in order to provide additional insight into quantitative survey responses. SR reviewed the notes from all coders and summarized them into four main themes, then solicited feedback from the team. A description of the themes and selection of open text responses, organized by theme and identified by age and state of residence of the participant, are included in the results section. We did not use a formal qualitative analysis approach to identify these themes due to the large number of short open-text responses.
For descriptives, we generated estimates of weighted proportions who believed SMA and abortion "should not be against the law" and estimates of weighted proportions who were "unsure." For bivariable and multivariable analyses, we conducted weighted multinomial logistic regression to first determine factors associated with believing SMA and abortion "should not be against the law" and then to assess factors associated with being "unsure." For all analyses, we assigned "should be against the law" as the reference group because of our focus on understanding predictors of believing that SMA and abortion "should not be against the law." When referring to the main outcome throughout this manuscript, we use the phrase "should not be against the law" rather than the more affirmative "should be legal" because this was the language used in the original survey options presented to participants. We also believe that there is a difference between believing that a behavior should not be against the law (absence of criminalization) and believing it should be legal (presence of legal protection). For all analyses, we generated estimated risk ratios and applied sampling weights provided by GfK and based on US Census data, so that results are representative of the US population of noninstitutionalized adult reproductive age women ages 18-49 and account for the probabilistic sampling frame and any differential nonresponse. Results from a previously published descriptive comparison of the demographic profile of the study sample to reproductive aged women in the National Survey of Family Growth 2013-2015 indicates that the present study sample slightly overrepresents married women and underrepresents women living below 100% of the federal poverty level (Ralph et al., 2020).

Results
Sample characteristics are presented in Table 2. Most respondents had no prior personal experience with abortion or SMA (86%, 95% CI: 84.7%-86.8%) and had never given birth (nulliparous) (58%, 95% CI: 56.8%-60.1%). Thirty-six percent (95% CI: 34.4%-37.4%) lived in a state with at least one law that could be used to prosecute an individual for SMA. The majority of the sample believed that SMA (59%, 95% CI: 57.3%-60.4%) and abortion in general (76%, 95% CI: 74.3%-77.1%) should not be against the law (Table 3). Bivariable analyses showing which factors are statistically associated with believing SMA and abortion should not be against the law are presented in Table 4. Multivariable analyses show that prior abortion experience and being 200% or more above the federal poverty level, more educated, older, and less religious were associated with believing that SMA should not be against the law (Table 5). These findings were similar for attitudes about the legality of abortion more generally. However, SMA experience and age were not significantly associated with attitudes about general abortion. Non-Hispanic Black participants were more likely than non-Hispanic white participants to believe that abortion should not be against the law; this was not true for SMA in particular (Table 5).

Being "Unsure"
Nineteen percent of the sample was unsure what they thought about the legality of SMA (19%, 95% CI: 17.7%-20.2%) and 1% were unsure what they thought about the legality of abortion more generally (1%, 95% CI: 0.77%-1.5%). In bivariable analyses (Table 4), factors associated with being unsure about the legality of SMA included any prior experience with abortion, older age, non-Hispanic Black compared to non-Hispanic white race/ethnicity, being 200% or more above the federal poverty level, religious affiliation and higher frequency of religious practice, and Republican Party affiliation. There was some overall lack of correspondence in attitudes about the legality of SMA as compared to the legality of abortion. Of those who believed abortion in general should not be against the law, 20% (95% CI: 19%-22%) were unsure about the legality of SMA, 6% believed SMA should be against the law, and 5% said "other." Of those who believed abortion should be against the law, 60% (95% CI: 55.5%-62.4%) also believed that SMA should be against the law, 28% believed SMA should not be against the law, 9% were unsure, and 4% said "other" (Table 3).
Results from multivariable adjusted models (Table 5) highlight some similarities and differences in factors associated with being unsure about the legality of SMA and abortion. Factors associated with being unsure about both the legality of SMA and abortion included past experience with medication or surgical abortion (compared to no abortion experience) and non-Hispanic Black race/ethnicity (compared to non-Hispanic white). Factors associated with being unsure about the legality of SMA-but not about the legality of abortion-included being age 40-49 years (compared to 19-24 years) and affiliated with a non-Christian religion (compared to no religion).

The "Others"
Initially, 561 (8%) and 515 (7.3%) of respondents chose "other" when asked whether SMA and abortion, respectively, should be legally available. Of these, 86% and 89% provided some open text responses, though not all responses were relevant to the question. The team reviewed and recategorized 202 and 407 of these responses for SMA and abortion, respectively, leaving 359 (5.2%) and 108 (1.6%) responses in the "other" categories for analysis. Many of those who submitted text responses discussed their general beliefs about abortion, such as "I am against abortion" (30s, Ohio) or "I fully support a woman's right to choose the best option for them in their circumstances" (40s, Missouri). The re-categorization process highlighted four key themes that arose in participants' open text responses about SMA, including their consideration of morality versus legality, safety concerns, criminalization versus pros-ecution, and who (if anyone) should be held accountable for SMA. Though not prompted by the survey question, some respondents chose to express attitudes related to both morality and legality. In one example, the participant's perspectives on morality and legality were aligned: "I am against abortion. ANY type of abortion. Once the heart starts beating, the baby should be given all rights as a person. ANY attempt should be prosecuted" (20s, Texas). In another case, the respondent's beliefs about the morality and legality of SMA appeared to differ: "I think it is morally wrong to end an unwanted pregnancy. However, I don't think making it a crime is going to stop people from doing it…but I sure wish people would simply decide for themselves not to do it" (30s, Texas). Safety concerns about SMA were mentioned in many open text responses about the legality of SMA. One participant explained, "The only reason to make these practices illegal is to prevent women from doing irreparable harm to themselves" (30s, Massachusetts). Another said, "If women try their own method, they could still remain pregnant and damage themselves and potentially the fetus" (30s, Oregon). Some participants distinguished between legality of SMA and the prosecution of people who attempt SMA. For example, one respondent explained, "I think selfterminating abortions are dangerous and therefore should be illegal. However, 'prosecution' seems extremely harsh" (20s, California). Another said, "While I don't think abortions should be legal, I don't think jail time is a good consequence" (40s, Colorado). Some respondents explained that providers, rather than people seeking abortion, should be legally responsible for SMA, but did not clarify whether they agreed with the implications of this in a scenario where the pregnant person herself is procuring (or "providing") the abortion method, as in the case of SMA. One participant wrote, "Women should not be penalized for getting abortions on their own; practitioners or people who facilitate these abortions should be penalized for conducting these or providing these services without regulations" (20s, New York). Another respondent explained, "Yes, I believe abortion is wrong. Yes, I think it should be against the law. No, I do not think the woman should be prosecuted, unless she has done this multiple times. Prosecute and convict the abortion providers" (40s, Mississippi).

Discussion
Our findings show that most respondents believed that terminating a pregnancy without medical assistance should not be against the law. Laws used to criminalize SMA are therefore misaligned with the views of the majority of selfidentified females ages 18-49 living in the US, those most directly affected by these policies. These laws are also misaligned with The American College of Obstetricians and Gynecologists (ACOG), which opposes efforts to criminalize pregnant people for a variety of behaviors that allegedly could cause harm or risk to the fetus, including self-induced abortion (ACOG, 2015(ACOG, , 2017. Our results correspond to national polling data on support for legal access to abortion and also support findings that abortion attitudes often depend on particular circumstances of a given case, with most people favoring legal access to abortion for some but not all indications (Smith & Son, 2013). Results from this study substantiate evidence indicating that support for legal abortion access is higher among those with higher education, income, and prior abortion experience.

Potential Consequences of Criminalizing SMA
There are many potential unintended and harmful consequences of criminalizing SMA. Laws that criminalize SMA may lead to avoidance of health care-seeking for fear of being reported, particularly among patients undergoing spontaneous miscarriage or other unexpected complications in pregnancy unrelated to SMA but which may result in a similar clinical presentation (Harris & Grossman, 2020). In the US, policies targeting substance use during pregnancy have been shown to be associated with decreased care-seeking and worse rather than improved health outcomes (Roberts & Pies, 2011;Subbaraman et al., 2018). In addition, criminalization of SMA could require medical providers to take on policing roles for which they are underprepared and could further deteriorate expectations of patient-provider confidentiality (ACOG, 2017; If/When/How, 2020). In international settings, the criminalization of abortion has been shown to lead health professionals to breach patient confidentiality, compromise their patients' trust, and offer suboptimal care particularly when treating pregnancy or post-abortion complications (McNaughton et al., 2006;Ramm et al., 2020).

Lack of Correspondence in Attitudes About SMA and Abortion
We found that not all individuals have the same attitudes about the legality of SMA as they do about the legality of abortion more generally. Specifically, fewer respondents supported SMA legality (59%) as compared to abortion legality (76%), some respondents believed that SMA but not abortion should be against the law (6%), and a larger group of respondents believed that abortion but not SMA should be against the law (28%). Being unsure about the legality of SMA was more common than being unsure about abortion more generally (19% compared to 1%). These discrepancies may be due to differences in the wording of the questions about abortion and SMA, and specifically the option to choose "unsure" for the SMA question but not for the abortion question. There was an "other" option for both questions, so it is possible that those who were "unsure" about abortion chose "other" and were then re-categorized as "unsure" in the coding process described in the methods. Concerns that SMA is less safe and/or effective than facility-based abortion services, two   themes which arose in open-text responses from respondents, may also explain the apparent discrepancies in attitudes about abortion and SMA legality. Some of those who believed SMA but not abortion should be against the law may have assumed that SMA is more dangerous than clinic-based abortion. Others who said that abortion but not SMA should be against the law may have assumed that SMA includes the use of all-natural alternatives for controlling one's fertility, such as edible herbs, and therefore concluded that it should not be regulated by the law. Data from the American Values Atlas, a nationally representative household sample of US adults, indicate that beyond demographic factors, perceptions of clinic safety emerged as the single strongest predictor of views on the legality of abortion (Jones & Cox, 2012). Using data collected from questions included in the present study's survey (beyond those used for this analysis), Biggs et al. (2019) found that between 40% and 60% of the sample were concerned about the safety of alternative provision of medication abortion models, including online ordering, over-the-counter access, and obtaining the medication from a clinician in advance of pregnancy. Though the beliefs of respondents who supported SMA but not abortion legalization appear contradictory, this finding corroborates results from another study, which found that abortion patients who attempted SMA before coming to the clinic were more likely to have anti-legal attitudes about facilitybased abortion, compared to those who did not attempt SMA . It is possible that some people may prefer private alternatives for managing their fertility because of, rather than despite, their greater ambivalence or lack of support for legal abortion protections.
Confusion about the difference between SMA and abortion, and a lack of consistent knowledge and understanding about what constitutes SMA and whether it is safe, may also contribute to a high number of "unsure" responses in this study. It is likely that many in our sample had not heard of SMA before completing the survey, and those who had heard of SMA may have been exposed to misinformation (Aiken et al., 2020;Berglas et al., 2017). Though drug and alcohol use and physical harm may be unsafe, more common SMA methods, such as self-sourced highly effective and safe medication abortion pills (mifepristone and misoprostol or misoprostol alone) or natural herbs, are relatively low risk (Grossman et al., 2010;Moseson et al., 2019;Murtagh et al., 2018). It is likely unknown to most people that patients who obtain medication abortions from a provider in clinic often swallow one or both medications at home rather than in front of a clinician (Greene & Drazen, 2016) and have equally high rates of effectiveness (Aiken et al., 2017). This information could alter public perceptions about whether SMA should be against the law; a recent randomized study in Texas found that informational statements about abortion safety reduced support for restrictive abortion laws purported to increase safety (White et al., 2017). Given growing media and policy attention to SMA, there is a need to better inform the public about this topic. Our findings related to high uncertainty about the legality of SMA underscore a potential opportunity to change people's attitudes about SMA through accurate information and education efforts.
Another possible source of uncertainty about the legality of SMA may be a desire for more nuanced discussions-and survey questions-that distinguish between morality and legality and between criminalization and prosecution. Some were uncomfortable with people terminating a pregnancy without medical assistance, either for moral or safety reasons, but did not believe the behavior should be regulated by law. One might expect that an individual who believes SMA should be against the law would also support the prosecution of individuals who attempt SMA; yet open-text responses demonstrated more nuanced views. Some believed that SMA should be against the law but did not support prosecution and imprisonment of people who attempt it. Research shows that abortion attitudes are increasingly complex and context specific (Jozkowski et al., 2018). These nuances are important to identify to better understand attitudes about abortion and SMA in particular.

Strengths and Limitations
This study is one of the first to explore attitudes about SMA in a large nationally representative sample of reproductive age women. However, this study has some important limitations, including challenges related to the quantitative measurement of attitudes about abortion and SMA and the lack of public knowledge about SMA. Given the exploratory nature of this study, and the fact that the instrument was designed specifically to estimate the prevalence of SMA in the US and not attitudes about SMA, the survey questions used were not validated or formally piloted. The survey also did not include reading or attention checks to screen for comprehension; however, the participants are part of a survey panel, so they are somewhat accustomed to taking surveys. The survey questions were not equipped to tease apart legal versus moral perspectives regarding SMA and abortion, which have been shown to be incongruent for some individuals (Woodruff et al., 2018). Further, we did not clearly ask about participants' attitudes related to legal consequences (i.e., prosecution) for individuals who attempt SMA. These nuances are important to articulate and measure in future surveys about SMA and abortion attitudes, particularly in a time when legal protections for abortion are under threat and pregnant people are turning to alternatives outside of the legal healthcare setting. We acknowledge that language spoken by the participant is an imperfect measure of acculturation and that the decision to combine language and race/ethnicity variables in the model was driven by concerns about multicollinearity. Finally, the sample includes only selfidentified women of reproductive age, which limits our ability to generalize our findings to self-identified men in the US; however, evidence shows little difference in general abortion attitudes by sex (Smith et al., 2018).

Conclusions
Most reproductive age women living in the US believe SMA should not be against the law. Laws that criminalize SMA are not aligned with the opinions of women living in the US more broadly nor with the opinions of women living in the states where these laws are in place. These findings may help to support efforts to decriminalize SMA. Relative uncertainty about the legality of SMA compared to facility-based abortion highlights an opportunity for public outreach and education on SMA in an effort to shift attitudes about policies that criminalize it.
Ultimately, shifting attitudes could lead to even more support for the decriminalization of SMA through policy reform.
Funding This study was funded by The David and Lucile Packard Foundation.

Declarations
Ethical Approval The study was approved by the University of California, San Francisco Human Research Protection Program. The procedures used in this study adhere to the tenets of the Declaration of Helsinki.
Informed Consent Informed consent was obtained from all individual participants included in the study.

Conflict of Interest
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.