Strategic (Non)Disclosure: Activation and Avoidance of Social Ties among Women Seeking Abortion

Abstract

The increased politicization of sexual and reproductive health has created barriers to medically necessary care. In absence of formal health care, social ties become critical sources of information and resources, yet the disclosure of stigmatized health needs carries significant risk. How do people navigate the risks and benefits of disclosure when seeking care for stigmatized needs? Drawing on original survey data (N = 153) and in-depth interviews (N = 55) with women who attempted a self-managed abortion, I first describe the distinct roles of weak and strong ties in women’s health-seeking experiences. I then demonstrate how both partial disclosure and nondisclosure are critical tools for obtaining information, resources, and emotional support during periods of health-seeking. Findings advance understanding of disclosure as a continuum that can be strategically wielded by people with stigmatized needs to confront and evade stigma and surveillance from their networks, the state, and the formal health care system.

How do people navigate their social networks when seeking information, resources, and support for stigmatized health needs? The increasing politicization of sexual and reproductive health care, including contraception, abortion, and gender-affirming care, has resulted in laws, policies, discrimination, and misinformation that limit access to medically necessary care, the effects of which are particularly severe for structurally marginalized groups, including minors, immigrants, Black and Hispanic people, and sexual- and gender-minority populations (Abreu et al. 2022; Desai and Samari 2020; Hall, Moreau, and Trussell 2012). In response, alternative pathways to sexual and reproductive care are emerging, including travel across state and national borders and self-managed care, the practice of meeting one’s health care needs outside of formal health care setting (Aiken et al. 2019; Kelly, Biello, and Hughto 2023).

In absence of formal access to care, social networks are critical for health-seeking behaviors and outcomes, yet they represent a double-edged sword (Song et al. 2021). In the case of stigmatized health needs, the potential impact of relationships on health is magnified—social ties may represent the only pathway to health resources and emotional support in a landscape of significant barriers or restrictions. On the other hand, activation of social ties who might deter, coerce, or pass judgement may further constrain health-seeking and result in social, psychological, physical, or legal harm for the person seeking care (Perry and Pescosolido 2015; Song et al. 2021). Despite these divergent outcomes, little is known about people’s awareness of these competing social forces and what strategies they use to navigate their networks in times of health-related crises and uncertainty.

In this study, I examine how women ¹ navigate their social networks—including whom they turn to and whom they avoid—while seeking information and support to end an unwanted pregnancy. Using the case of this legally restricted and stigmatized health behavior, I describe my respondents’ strategic reliance on their social networks for information and support, investigating the characteristics and circumstances of those who confide in weak versus strong social ties during periods of health-seeking. I then demonstrate how partial disclosure and nondisclosure—the intentional withholding of information from specific social ties—is also a critical health-seeking behavior that confers advantages for health and well-being. Together, these findings situate strategic (non)disclosure as an important and undertheorized mode of resistance, autonomy, and protection for marginalized and surveilled populations.

Background

The Activation of Social Ties for Stigmatized Health Behaviors

During periods of health-seeking, stigma operates at both interpersonal and structural levels, shaping access, use, and quality of health care by permeating community-level attitudes and institutional practices (Broussard 2020; Hatzenbuehler 2014; Reid et al. 2014). When facing health-related crises and uncertainty, people with stigmatized health needs often need to strategically activate network ties to obtain information, resources, and support (Chaudoir, Fisher, and Simoni 2011; Cowan 2014; Perry and Pescosolido 2015).

Social networks scholars have long considered the distinctive role of strong versus weak social ties in shaping individual outcomes related to social mobility, health, and well-being (Desmond 2012; Granovetter 1973; Small 2013). Although the measurement of tie strength varies across studies, strong ties are typically characterized as those that are important or special, reflecting a voluntary investment in the relationship and mutual desire for interaction (e.g., family members, intimate partners, friends), whereas distant, temporary, or indirect associations are more often deemed weak social ties (e.g., acquaintances, strangers; Desmond 2012; Granovetter 1973; Small 2013). Researchers examining the role of social networks during periods of health-seeking often emphasize the social and health benefits of tie activation—health discussant partners, particularly strong social ties, offer information and medical resources, reduce stigma, improve health literacy, and promote engagement with the formal health care system (Agner et al. 2023; Perry and Pescosolido 2010, 2015; Raudenbush 2020). By contrast, this research often emphasizes the negative health effects of social isolation or reliance on weak social ties during periods of health crises (Askelson, Campo, and Carter 2011; Perry and Pescosolido 2015; Umberson, Lin, and Cha 2022). For instance, individuals diagnosed with severe mental illness who confide in weak ties are more likely to receive negative responses (Perry 2011) and experience poorer functioning, lower social satisfaction, and lower quality of life later on (Perry and Pescosolido 2015). Researchers also highlight the negative health outcomes associated with social isolation and secret-keeping, including gaps in health knowledge, worse physical and mental health (Askelson et al. 2011; Umberson et al. 2022), and the erosion of access to resources and intimacy within personal relationships (Cowan 2020).

Despite the important insights generated by the literature on the activation of social ties during periods of health-seeking, significantly less empirical attention has been devoted to understanding whom people strategically exclude or avoid as health discussion partners, a behavior that may be beneficial for people with stigmatized needs. As a methodological tool, network analysis is theoretically and structurally motivated to investigate the utility and effects of social tie activation rather than social avoidance or exclusion. As such, previous studies have primarily focused on who hears secrets and stigmatized information, including those assumed to be supportive (Cowan 2014, 2020; Perry and Pescosolido 2015; Raudenbush 2016) and compatriots who, “through the bonds of intimacy or shared transgression,” understand the plight of the stigmatized (Antoniades, Mazza, and Brijnath 2018; Rossier 2007:230). For people with stigmatized needs, however, who often do not have strong ties whom they can trust or depend on for support, nondisclosure and reliance on weak ties may represent the only pathway to avoid stigma and attain a desired health outcome.

In this article, I consider the implications of (non)disclosure for physical, social, and emotional well-being by analyzing the experiences of women seeking abortion. Given the dramatic rise in abortion restrictions across the United States and a pervasive culture of abortion stigma (Norris et al. 2011), the decision of whom to turn to is laden with risk. Over the past 20 years, a significant proportion of criminal investigations of self-managed abortion (SMA) in the United States were brought to the attention of law enforcement by health care providers (39%) and people entrusted with information, including friends, parents, or intimate partners (26%; Huss, Diaz-Tello, and Samari 2023). Strong ties such as friends or family members may be more likely to provide emotional or material support and could potentially shield individuals from outside surveillance or intervention. However, if the threat of shame and stigma within one’s strong personal networks is high, confiding in weak ties may provide a comfortable social distance for obtaining support without damaging close personal relationships. Confiding in weak ties, however, also carries significant risks. Although more “disposable” (that is, a personal relationship need not continue following disclosure; Desmond 2012), weak ties may present greater legal or medical risk for those seeking support outside of the formal health care setting.

Through this case study, I develop and apply the concept of strategic (non)disclosure to describe how individuals engage and avoid their network ties to obtain health-related information and support. Findings advance understanding of disclosure as a continuum that can be strategically wielded by individuals with stigmatized health needs to confront and evade stigma and surveillance from their social networks, the state, and the formal health care system. In the discussion, I consider the implications of these findings for the future of abortion care and other stigmatized health needs.

Data And Methods

Setting and Context

This study was conducted among women living near the United States–Mexico border, a region that has historically offered alternative pathways to abortion (Luker 1984; Murillo 2023). Before abortion was deemed a constitutional right by the U.S. Supreme Court decision Roe v. Wade (1973), thousands of women traveled south of the border in search of care (Luker 1984; Murillo 2023). Individual states continued to pass restrictive abortion laws that disproportionately impacted access for already marginalized groups, including low-income people, women of color, immigrants, and those living in states with hostile abortion policy climates (Desai and Samari 2020; Goyal, Brooks, and Powers 2020; Grossman et al. 2015). Due in part to these restrictions, attempted SMA is higher among these groups (Aiken et al. 2019; Ralph et al. 2020).

Individuals who pursue informal pathways to care are inherently difficult to study given their reduced engagement with the formal health care system. Although state-level abortion restrictions have increased dramatically over the last decade, the year that I concluded data collection, 2021, saw the greatest number of restrictions of any year since Roe v. Wade was decided in 1973 (Nash 2021). Although these restrictions likely increased demand for SMA (Aiken et al. 2020), fears of legal risk may have diminished respondent willingness to participate in research on this topic. To mitigate these challenges, I used a multipronged approach to recruit respondents, drawing from both online and community-based networks. To protect the privacy of all respondents, any personally identifying information was erased. Names presented in this article are pseudonyms, and any personally identifying details have been modified to protect the privacy of respondents and their communities. This study was granted a Certificate of Confidentiality from the National Institutes of Health and was approved by the Institutional Review Board at The University of Texas at Austin.

Before initiating data collection, I traveled to California and Texas to speak with individuals and organizations working for reproductive justice in their communities both on the U.S. and Mexican sides of the border. I spoke with abortion providers and clinic staff members, two student organizations, a feminist group that accompanies individuals seeking abortion in Mexico, and an abortion fund that offers financial support for people seeking clinic-based abortion. With insights gained from these exchanges, I designed an online survey and interview guide to explore the information and social support obtained by women who attempted SMA.

The use of mixed methods was ideal for several reasons. Given the sensitive nature of the topic, the collection of survey data online provided the opportunity for greater anonymity and thus may be more inclusive of individuals who would not be comfortable participating in research in person. Interviews were valuable for exploring strategies of partial disclosure and nondisclosure, information that traditional network approaches are rarely designed or intended to capture. I moved recursively between these two data sources to compare statistical patterns of disclosure with women’s narrative accounts of partial disclosure and nondisclosure.

Recruitment

Survey respondents were recruited online using a targeted social media advertising campaign on Facebook and Instagram from June 2020 to March 2021. Individual users were selected by the campaign to see the recruitment flyer based on their location and demographic characteristics. Digital flyers in English and Spanish were presented to women and girls between the ages of 14 and 49 living within a two-hour drive north of the United States–Mexico border. Women were eligible to participate if they had attempted SMA within the past 10 years. For the purposes of this study, SMA included any attempt to end a pregnancy outside of the formal health care setting, including the use of herbs, abortion medications, supplements, alcohol, or physical methods such as trauma to the abdomen.

Several clinics and community groups also assisted with recruitment: Two abortion clinics in Texas and California posted recruitment flyers in their waiting rooms. Clinic patients accounted for about 6% of respondents. Several organizations located near the border shared information about the study on social media. Ultimately, the largest share of respondents (78%) reported learning about the study from an announcement or advertisement on social media. I also invited survey respondents to refer other people they knew who had attempted SMA. Respondents received a $5 gift card for completing a survey and an additional $5 gift card if they referred somebody to the study who also completed the survey. After data collection began, it became apparent that people with a greater number of confidants required more time to complete the survey, so compensation for the survey was increased to a $10 gift card. Respondent-driven sampling resulted in an additional 19 respondents (12% of the sample). To minimize the potential for survey bots or multiple user submissions, the survey link was only shared with individuals who contacted the project directly via email. They were then provided with a unique four-digit code to enter upon beginning the survey.

Survey respondents were invited to participate in a follow-up interview about their experience either via Zoom or over the phone. After completing 30 interviews, I selectively reached out only to respondents who contributed greater diversity to the sample. For example, I purposefully recruited respondents who were foreign-born, and I invited both people who completed SMA using abortion medications or other methods. I also sought people who reported no confidants and those who reported a high number of confidants (four or more). In total, I conducted 55 in-depth interviews (36% of the survey sample). Most chose to participate by phone (n = 51) rather than Zoom. I conducted interviews in English or Spanish, and they lasted between 40 minutes and 1 hour and 50 minutes. Interview respondents received a $30 gift card for their time.

Network Measures

To gain insight into whom respondents turned to when seeking abortion, the survey included a name generator to construct egocentric discussion networks (see Appendix A in the online version of the article). Respondents were first asked “Did you talk to anybody about your plans to end your pregnancy?” and then “How many people?” Only two respondents (<3%) reported speaking with more than three people. All respondents then received the following instructions: “We would like to know more about who you turned to for information and support when seeking abortion. Please list these people in the spaces below.” The number of spaces provided in the survey corresponded to the number of people to whom they reported speaking (hereafter referred to as “confidants”), with a maximum of four entry spaces. This name generator gave respondents the option of listing their confidants using only initials or pseudonyms to protect their identity. Respondents then reported the characteristics of each confidant, including their gender, race and ethnicity, country of birth and residence, and whether the person had themselves previously sought and/or attempted SMA. Respondents were also asked whether they conceived the pregnancy with this person, the quality and type of relationship they had (e.g., parent, partner, friend), and the kinds of support they received from them. To assess the strength of social ties, the survey included a validated scale known as the Unidimensional Relationship Closeness Scale (URCS; Dibble, Levine, and Park 2012). Respondents provided a number from 1 (strongly disagree) to 7 (strongly agree) for 11 statements measuring relationship closeness (see Appendix A in the online version of the article).

The interview guide was designed to explore women’s social context broadly and then more narrowly explore the sources of information and support that they obtained when seeking abortion. To explore partial and nondisclosure during abortion-seeking, I asked respondents what information they shared with people from whom they sought help and whether there was anybody in their life whom they avoided or kept their abortion secret.

Sample and Analytic Approach

Table 1 presents the sample characteristics of 153 survey respondents and 55 women who participated in a follow-up interview. Summary statistics broadly reflect the demographic composition and socioeconomic disadvantage often present in border communities. Most respondents were Hispanic, and almost a quarter were foreign-born or only spoke Spanish at home. The distribution of characteristics was similar across both the survey respondents and interview participants, with a few notable exceptions. On average, interview respondents were older and more likely to be foreign-born, to have completed college, and to be employed.

Table 1.

Characteristics of Women Living near the United States–Mexico Border Who Attempted a Self-Managed Abortion within the Past 10 Years.

	Survey Respondents		Interview Participants
	N = 153	(%)	N = 55	(%)
Demographic characteristics
Age at time of survey (years)
14–17	13	(8)	1	(2)
18–24	50	(33)	19	(35)
25–29	34	(22)	12	(22)
30–34	35	(23)	12	(22)
35+	21	(14)	11	(20)
Country of birth
United States	117	(77)	39	(71)
Mexico	32	(21)	14	(25)
Other	3	(2)	2	(4)
Place of residence a
Arizona	35	(23)	13	(24)
California	40	(26)	19	(35)
New Mexico	9	(6)	0	(0)
Texas	59	(39)	23	(42)
Mexico	12	(7)	0	(0)
Race-ethnicity
Hispanic	104	(68)	34	(62)
White	28	(18)	11	(20)
Multiracial and other	21	(14)	10	(18)
Language spoken at home
English only	79	(52)	28	(51)
Spanish only	36	(24)	11	(20)
English and Spanish	38	(25)	16	(29)
Highest education attained
High school or less	62	(41)	21	(38)
Some college	58	(38)	18	(33)
Completed college	32	(21)	16	(29)
Employment
None	73	(48)	23	(42)
Part-time	44	(29)	19	(35)
Full-time	35	(23)	13	(24)
Abortion experience
Gestational age at time of SMA attempt
10 weeks or less	98	(64)	36	(65)
11–16 weeks	31	(20)	13	(24)
Over 16 weeks	2	(1)	0	(0)
Unsure	21	(14)	6	(11)
When attempted SMA
Past year	42	(28)	14	(25)
Past 5 years (>1 year ago)	73	(48)	31	(56)
Past 10 years (>5 years ago)	38	(25)	10	(18)
Methods of SMA attempted a
Abortion pills	79	(52)	35	(64)
Herbal remedies	76	(50)	27	(49)
Vitamin C	47	(31)	19	(35)
Alcohol	44	(29)	18	(33)
Physical exercise	46	(30)	20	(37)
Physical injury	28	(18)	9	(16)

Note: One survey respondent (0.6%) had missing data for country of birth, highest education attained, employment, has children, gestational age, and had a previous abortion. Percentages may exceed 100 due to rounding. SMA = self-managed abortion.

^aRespondents could select multiple responses.

I began my analysis by examining the demographic characteristics, socioeconomic status, and reproductive health experiences of my sample, presenting averages separately for survey respondents and interview participants. I then analyzed experiences of disclosure using two separate approaches. First, I considered the experience of disclosure from the perspective of respondents, examining the number and characteristics of people whom they turned to for information and support (Table 2). Second, using data produced by the name generator, I created a confidant-level data set to examine the relationship characteristics and support provided by each confidant. Accounting for the clustered nature of the data (each respondent could name zero to four confidants), I calculated the average relationship closeness score (URCS) according to relationship type. Mean values of relationship closeness with 95% confidence intervals in Figure 1 reveal a division at the midpoint value of 3.5. For all quantitative analyses, I categorized confidants with a mean URCS above this value as “strong ties” and those with URCS below this value as “weak ties.” Figure 1 presents the average relationship closeness reported by respondents according to relationship type. As might be expected, confidants who were family members, friends, and partners tended to have values above 3.5, whereas colleagues, acquaintances, and other relationships tended to fall below this threshold and thus were more likely to be categorized as a weak tie. Chi-square tests were used to compare the type of support offered by weak versus strong ties.

Table 2.

Social Networks and Experiences of (Non)Disclosure among Women Who Attempted a Self-Managed Abortion (N = 153).

	Mean	SD	Range
Talked to somebody about their plans to end their pregnancy	.45		0, 1
Number of confidants	.84	1.28	0–10 a
Among those who disclosed their plans (n = 69):
Turned to a strong tie (URCS > 3.5)	.74		0, 1
Turned to a weak tie (URCS < 3.5)	.41		0, 1
Among those who did not disclose to anyone (n = 84 a ):
Would you have preferred to speak with somebody about your plans to end your pregnancy?
Yes	.51		0, 1
No	.49		0, 1
If yes → why were you not able to speak to anybody?
Did not know if anyone would have been supportive	.76		0, 1
Did not want anyone to know about abortion	.67		0, 1
Was worried about getting in trouble with the law	.62		0, 1
Other	.05		0, 1

Note: URCS = Unidimensional Relationship Closeness Scale.

^aResponses were truncated at 10 or more confidants.

^bOne respondent who did not turn to anyone did not provide their preference for disclosure.

Figure 1.

Average Relationship Closeness According to Relationship Type (Estimates with 95% Confidence Intervals).

Note: Average relationship closeness score (URCS) adjusted for the clustered structure of the data (66 respondents named 119 confidants). Relationship closeness was calculated with the URCS, a series of 11 questions about the quality of the respondent’s relationship to their confidant (see Appendix A in the online version of the article). Confidants with URCS values > 3.5 were designated as strong ties, and those with URCS values < 3.5 were designated as weak ties. The red line at the midpoint value of 3.5 demonstrates how various relationship types were distributed across these categories. URCS = Unidimensional Relationship Closeness Scale.

Regression analysis

Multiple logistic regression was used to investigate the characteristics and circumstances associated with respondents’ (non)disclosure while seeking abortion, including whether they confided in a weak or strong tie (Table 3). Independent variables included the respondent’s age at time of survey, country of birth (coded as 1 if born in Mexico), and whether they currently resided in a place with a hostile abortion policy climate as defined by the 2018 Guttmacher Classification (Nash et al. 2018). The four U.S. states that share a border with Mexico reflected opposite ends of this spectrum, with California and New Mexico categorized as “supportive” and Texas and Arizona as “extremely hostile.” For this analysis, women living in Mexico (n = 10) were coded as living in a hostile policy environment because abortion was not legally available in any northern Mexican state during the study period. To evaluate socioeconomic status, I included educational attainment (reference = high school or less) and employment status (1 = unemployed) in the model. I accounted for circumstances related to the respondent’s experience seeking abortion with three variables: recency of the respondent’s attempted abortion (coded as 1 if attempted within the past year, 0 otherwise), gestational age when SMA was attempted (second trimester or later = 1), and whether the respondent attempted SMA using abortion pills, the most common method used by women in the sample. All statistical analyses were conducted using Stata 16.

Table 3.

Factors Associated with Nondisclosure, Disclosure to a Strong Tie, and Disclosure to a Weak Tie while Seeking Information and Support for Abortion. ^a

	Nondisclosure			Disclosure
				Strong Tie			Weak Tie
Characteristics	OR	Significance	95% CI	OR	Significance	95% CI	OR	Significance	95% CI
Age (at survey)	1.07	*	1.00–1.14	.93	*	.87–1.00	.93	+	.85–1.00
Born in Mexico	1.31		.54–3.20	.64		.24–1.70	.78		.26–2.41
Hostile policy climate	.60		.28–1.26	1.17		.54–2.56	3.08	*	1.01–9.40
Educational attainment
High school or less (reference)	—		—	—		—	—		—
Some college	.57		.25–1.28	1.73		.72–4.17	1.20		.39–3.65
Graduated college	.24	*	.09–.66	4.67	*	1.64–13.30	4.17	*	1.23–14.19
Unemployed	.81		.40–1.64	.94		.45–1.97	2.02		.78–5.25
Attempted SMA within past year	1.29		.54–3.07	.86		.34–2.15	.33	+	.09–1.18
Second trimester or later	.88		.42–1.83	1.12		.52–2.45	2.31	+	.89–5.99
Attempted SMA with abortion pills	.33	*	.16–.70	2.47	*	1.15–5.32	3.00	*	1.14–7.92
Constant	.87		.15–5.02	1.21		.19–7.67	.19		.02–2.23
N	151			148			148
Model χ2	18.69*			15.70+			17.47*
Pseudo R2	.09			.08			.12

Note: OR = odds ratio; CI = confidence interval; SMA = self-managed abortion.

^aRelationship closeness was calculated with the Unidimensional Relationship Closeness Scale (URCS), a series of 11 questions about the quality of the respondent’s relationship to their confidant (see Appendix A in the online version of the article). Confidants who received an average score of relationship closeness <3.5 were designated as weak ties, and those with URCS averages >3.5 were designated as strong ties. Two respondents with missing data for covariates are excluded from analysis. Three respondents indicated that they disclosed to another person and thus are included in the first model, but they did not provide URCS values and are excluded from the second and third models. SMA = self-managed abortion.

⁺p < .10. *p < .05 (two-tailed tests).

In-depth interviews

Interviews were coded and analyzed using a flexible approach (Deterding and Waters 2018). I transcribed four interviews (two in English and two in Spanish) to develop the coding guide, and all remaining interviews were transcribed by a professional transcription service. I first assigned codes for participant attributes and then indexed the transcripts using broad codes reflecting the concepts that motivated the research and questions asked in the interview. I wrote analytic memos at two specific time points: immediately following each interview and after assigning index codes to each transcript. Using these memos, I then developed more fine-grained codes related to people’s experiences seeking support within their social networks. These analytic codes were applied to subsections of the interview where respondents discussed their decision to engage their networks when searching for abortion care. I coded and analyzed all interviews in their original language using Atlas.ti 9 and translated all Spanish quotations presented in this article.

Results

Disclosure Experiences

Survey respondents engaged in limited disclosure when seeking abortion; less than half told another person about their plans to end their pregnancy (45%), and only 9% told more than two people (not shown). Table 2 separately describes the experiences of respondents who disclosed and did not disclose their abortion plans. Among those who confided in another person, most turned to a strong tie (74%), yet a substantial minority turned to a weak tie (41%), confirming that ties of varying social proximities are utilized in the context of abortion-seeking.

Among those who did not disclose their abortion, just over half said they would have preferred to speak with somebody, but they either did not perceive their networks to be supportive or feared potential repercussions of disclosure. Two respondents provided open-ended responses in the survey that revealed the gravity of their dilemma, writing, “I didn’t want to risk losing my living space at my parents’ house. I needed to keep it a secret,” while another revealed that her abusive partner prohibited her from talking to other people: “[He] held me against my will, and he cut off all communication with others.”

The role of weak and strong ties

Although respondents turned to both weak and strong ties when seeking information and support for abortion, it is unclear whether the type and quality of support received differed according to tie strength. To compare the support obtained across confidants, Figure 2 displays the type of support provided for strong versus weak ties, revealing several patterns.

Figure 2.

Type of Support Provided by Strong versus Weak Ties (N = 119).

Note: Confidants may have provided more than one type of support. Confidants are designated as weak or strong ties using the Unidimensional Relationship Closeness Scale (URCS; range = 1–7). Strong ties received an average URCS > 3.5, and weak ties had an average URCS < 3.5. SMA = self-managed abortion.

Results of chi-square test comparing support provided by strong versus weak ties: +p < .10. ***p < .001.

First, a significantly larger percentage of strong ties provided emotional support, accompaniment, and money for respondents compared to weak ties. These patterns support previous research that suggests strong social ties, such as friends and family, play an important role in health discussion networks (Perry and Pescosolido 2010; Wellman 2014). Ximena’s experience, shared during an interview, illustrates the distinct roles of strong and weak ties among women seeking abortion (Figure 3). Ximena was born in the United States but spent most of her early life in Mexico and knew that abortion pills were available in pharmacies south of the border. Like many other women I interviewed, Ximena spoke with only a few people about her abortion, including two close friends and an acquaintance:

One of my friends was very, very sad because he was very much in love with me. So, he told me that he completely supported me, that he would give me money, whatever I needed. . . . My girlfriend was of the position “Whatever you want, I support you.” The other person who I talked to—in reality, I only talked to him because he could get me the pills—because he had gotten the pills for other people. That’s why I told him. [Translated]

Figure 3.

Illustrative Sociogram of Weak and Strong Ties Utilized by a Woman (Ximena) Who Sought and Completed a Self-Managed Abortion.

Second, although strong ties are often framed as being primary sources of affective care and support, most weak ties also provided emotional support (56%). Like Ximena, respondents initially approached weak ties for instrumental reasons, yet many nevertheless received emotional support during these exchanges. For example, Astrid went to Mexico to buy abortion pills and spoke about the personalized care she received:

I didn’t have nobody to help me out or anything, so I just crossed over [the border], and it was in the pharmacy . . . I gave them the paper of the medication that I had written down and the tea. He said he had to go a couple of blocks down because his other location had it. He just came, he gave me a medication and told me [sic] if I knew how to drink it . . . he never interrogated me, never questioned me. I actually felt like he was there to help me out . . . without no ifs or buts. He was very helpful in providing all that information.

After selling her the pills, the pharmacist also made himself available by phone after she went home and suggested that she return for follow-up care if needed. Emotional support also sometimes came from health care providers with whom women interacted when seeking follow-up care after their abortion. When Gracia ended up seeking care at a hospital, she said:

I was very scared to go to the doctor because if they knew what I did, they would put me in jail. . . . But it turns out that the doctor who was taking care of me . . . we went to high school together. When he was checking me, he just said, “I’m going to put that you had an ectopic pregnancy.” I don’t know if it was an ectopic pregnancy from the very beginning, and that’s why the pills didn’t work, or if he just did that so he didn’t have to put any other explanation, and then [I would] have to go to jail. . . . He did such a good job on my stitches, on my little scar. You cannot see it. It looks almost like a plastic surgeon did it. It was really pretty. [laughs]. I was very lucky. So, to this day, I don’t know if he did it to protect me.

Although she did not disclose her abortion to her doctor, a weak tie who could have reported her to legal authorities, she nevertheless described the protection and emotional support she felt while receiving care.

The survey data provide some insight into factors that may affect disclosure patterns among those seeking care for a stigmatized need. Table 3 presents the results of multiple logistic regressions that describe the relationship between respondent characteristics and (non)disclosure, with separate analyses showing characteristics associated with those who confided in a strong versus weak social tie. The overall model predicting nondisclosure showed that for every additional year of age, respondents had 7% higher odds of not disclosing their abortion. College graduates had significantly lower odds (OR = .24, p < .05) of engaging in nondisclosure compared to their less educated counterparts, as did those who attempted SMA with abortion pills (OR = .33, p < .05). Factors associated with disclosure to a strong or weak tie were similar, reflecting patterns that contrast those of nondisclosure: Younger respondents, college graduates, and respondents who attempted SMA with abortion pills were more likely to disclose to a strong or weak tie. Several factors were notably different for respondents who confided in weak social ties. Those living in a policy climate that was hostile toward abortion access (Texas, Arizona, or Mexico) were over 3 times as likely to confide in a weak tie compared to those living in more supportive abortion policy climates (New Mexico and California). Respondents further along in their pregnancies (second trimester or later) had over 2 times greater odds of confiding in a weak tie, while those who attempted SMA within the past year had 67% lower odds of confiding in a weak tie. Building on research that highlights the importance of strong social ties, these findings highlight the importance of weak-tie disclosure for people with stigmatized health needs, particularly for those living in hostile policy climates.

Using a traditional approach to egocentric network data collection, the survey yielded an abundance of data about individuals who spoke to another person and the individuals in whom they confided. By prioritizing disclosure, however, this approach yielded little information about the social contexts of people who were isolated or independent when seeking abortion or whom they avoided or excluded from their plans. As such, these data failed to account for the use of nondisclosure as a potential resource and health-seeking strategy. Ximena’s experience is again illustrative; although she confided in two strong ties and a weak tie when seeking abortion, nondisclosure was also an important part of her experience. When asked if she avoided speaking with anyone, she said:

I didn’t tell [the father] anything. I didn’t speak with him at all. I didn’t tell him until—when I had already aborted, I told him what had happened because we had stopped seeing each other completely . . . the father, definitively, I didn’t want to talk to him for anything, and with my family, I never mentioned any aspect . . . to my sisters or my parents. . . . I think they would have felt disappointed.

In the following section, I consider nondisclosure—the absence of information flow and exchange—highlighting women’s motivations for avoidance and secrecy and how nondisclosure was a similarly vital tool for obtaining information and resources.

Nondisclosure as Strategy and Continuum

Although family members are often an important source of support for matters of health, particularly among Hispanic populations (Almeida et al. 2009; Diaz and Niño 2019), the women I spoke with often explicitly avoided their familial ties, citing fears of feeling shame and guilt, destroying relationships, or risking physical violence. Over two-thirds of interview respondents specifically wanted to keep their abortion a secret from their mothers (Table 4). Fathers, intimate partners and spouses, other family members, and siblings were also listed as people with whom 22% to 36% of respondents avoided disclosure. Beyond the family, friends and health care providers were also avoided or excluded.

Table 4.

Strategic Nondisclosure: Individuals with Whom Interview Respondents Avoided Disclosing Their Abortion (N = 55).

Individual	Number of Respondents Who Avoided Disclosure	(%)
Mother	37	(67)
Father	20	(36)
Intimate partner/spouse	20	(36)
Other family members	17	(31)
Health care provider	16	(29)
Sibling	12	(22)
Friend	12	(22)

Note: Interview respondents could name multiple people whom they avoided or excluded.

Participants described an array of motivations for nondisclosure. Similar to the reasons people activate their social ties during times of need, nondisclosure and secrecy were also motivated by the need for instrumental and expressive support. Women kept their abortion a secret for the expressive reasons of (1) managing stigma and (2) accessing emotional support and the instrumental reasons of (3) ensuring physical safety, (4) safeguarding reproductive autonomy, and (5) avoiding medical surveillance.

Managing stigma

Women were often acutely aware of the personal and religious beliefs of their loved ones regarding abortion. Mothers and older female family members, such as aunts and grandmothers, were often perceived to be sources of shame and stigma that many women wished to avoid. For example, Lena recalled a previous conversation with her mother about abortion: “[My mother] told me that if one day I decided to do that, I had to forget that I had a family and that I had her, like, ‘If you do that, I’m dead to you.’ So, the fear of that was what was instilled in my mind.” Similarly, Bernadina was just 19 years old and had recently immigrated to the United States to work as a nanny when she discovered she was pregnant. After a failed attempt to end her pregnancy at home, she asked her mother to send her money but decided not to disclose her situation, saying “I didn’t want to . . . it’s something that I’m going to take to the grave. . . . I’m the little girl of the house, the youngest. There would always be a shadow over me for telling them this.” For these women, nondisclosure was a tool to manage stigma and prevent what they perceived to be irreparable damage to the most important relationships in their lives.

Accessing emotional support

Other times, women were strategic about how they disclosed their experience to obtain love, sympathy, and support from their friends and family. After resuming a relationship with an abusive boyfriend, Gracia discovered she was pregnant and attempted SMA with pills she purchased at a pharmacy. She ended up seeking care at the hospital for complications. When I asked her how everyone responded when she was in the hospital, she said:

The thing is, nobody knew it was an abortion. I said that it was my appendix removed. . . . So yes, I felt supported because there were people around taking care of me and showing care. Even the bad guy [her abusive partner] showed up. He was in the honeymoon phase! It was good.

By telling everyone that her appendix had been removed, Gracia was able to take on the role of a normal health care patient receiving standard care for an emergency medical procedure. Free of the stigma associated with abortion, she received the sympathy and support of her friends and partner that she desired.

More often, women chose to disguise their abortion as a miscarriage, particularly when disclosing their experience to other women in their families. Willamina described her stepmom’s response after telling her that she had a miscarriage: “[My stepmom] said that she had a miscarriage in 2000, and she understood how painful and unsettling it could be. [We] bonded over the loss of motherhood, essentially. That was something I couldn’t have done if I told her that I’d done it on purpose.” Taken together, these strategies of partial disclosure and nondisclosure not only offered protection from judgement and interpersonal stigma but also allowed them to access the emotional support they desired.

Ensuring physical safety

Some participants explicitly avoided telling their partners for instrumental reasons. Several were experiencing intimate partner violence and knew that secrecy was critical for their futures and their personal safety. After taking pills that she purchased from a nurse in Mexico, Florencia told her boyfriend that she was having a miscarriage so that he would take her to the hospital. Later, she said:

He never found out. . . . I think that he would have done something to me because he was—he was involved in drug trafficking at the time . . . and I believe that my life would have been in danger if he had found out. . . . He would have beat me, I believe, until I was dead.

Altogether, about a quarter of the women I interviewed (n = 13) were in abusive relationships when they discovered they were pregnant. One woman, Pilar, expressed ambivalence about not being able to confide in her boyfriend of three years:

I wish I could’ve talked to him more about it, but I wasn’t able to because I knew how he would have reacted. . . . He would have been absolutely livid. There’s been times when I fell asleep too early, and my phone would be blown up. He’d knock on my door, and he’d drag me to his house. I knew it would have been way worse than that. There’s been a lot of instances with violence, and if he would have known the whole story, it would’ve been a really big deal.

The threat of physical violence deterred many women from confiding in partners who, due to typical behaviors of abusers, were often their strongest social ties at the time. Nondisclosure to partners ultimately allowed most women to ensure their physical safety, end an unwanted pregnancy, and eventually leave an abusive relationship.

Safeguarding reproductive autonomy

Nondisclosure was also a strategy for some women to safeguard their autonomy as they sought abortion. These women explicitly avoided disclosing to their family members and partners because they feared these individuals would intervene. Minors who attempted SMA, like Myka, were especially concerned that their parents would force them to continue an unwanted pregnancy: “I remember always growing up, [my mom] would be like, ‘If you ever have a child at a young age, you’re going to keep it, and you’re going to give it up for adoption.’” Sole was 20 years old when we spoke and still closely under the scrutiny of her parents. She described them as recent immigrants from Mexico, “Catholic and extremely religious” and very conservative when it came to sex and childbearing. Upon learning that she was sexually active with her boyfriend, they urged her to leave home in California and move in with his family in Arizona to begin planning for marriage. When she found out she was pregnant several months after the move, she said:

It was really hard for us since we couldn’t talk to our parents about it. One, because my parents would not let me do that . . . if I told them [I was planning to have an abortion] . . . I feel like they would go to the extent of coming back for me, coming all the way from California to Arizona, literally, right then and there when they found out.

Avoiding medical surveillance

Nondisclosure was an important strategy for women during their interactions with health care providers. Pharmacists, nurses, and doctors could provide information, medications, and life-saving care in the event of complications, but the fear of coercion or legal retribution left individuals unable to access abortion openly, depriving them of resources necessary to avoid dangerous risks. Noemi, who had lived most of her life in Juárez and was accustomed to crossing the border, described health care in Mexico as “faster, cheaper, and sometimes better.” She knew and trusted her doctors in Mexico for general health needs but feared disclosing her need for abortion. Instead, she went to a pharmacy anonymously, saying “I never gave my information or name, everything was false.” This kind of strategic nondisclosure allowed her to minimize personal risk while gaining access to medications. Another respondent, Astrid, said she preferred SMA because it enabled her to evade the surveillance and administrative recordkeeping of her clinic in Texas:

Once I got the name of the medications and [learned] that we could find these medications in Mexico with no one asking questions or interrogating you—no prescription or having to see a doctor—to me, that was way better. . . . I get to do it in my house without talking too much, with somebody asking me questions. I know what I had to do. I have my medicine ready, my own medicine. [It was] more private. I wasn’t going to leave nothing like a folder that was in my history.

Minors similarly feared medical surveillance and often avoided the formal health care system for this reason. Yoana, who was 26 when we spoke, described her experience seeking care 10 years prior:

I asked in a clinic what is needed to go and get treated and all that stuff. I didn’t say why or anything. I just wanted to get information, and they told me I had to have an adult with me. . . . That’s the only place I knew of that could probably help me, but they told me that for pregnancies or anything related to that, since I was only 16, I would have to go with an adult.

As an undocumented immigrant, Yoana also faced surveillance from the state and was unable to cross the border to Mexico. Instead, she spent months trying alternative abortion methods at home and ultimately failed to end her pregnancy. Instead, she gave birth at the age of 16.

Like Yoana, the fear of being caught or punished for attempting abortion led others to take dangerous risks. Gracia, the woman who received medical care from a doctor with whom she attended high school, first went to the hospital when she began experiencing pain and heavy bleeding after taking pills. Fearing legal risks of SMA, she sat for hours in the waiting room, pretending that she didn’t know she was pregnant, and eventually had to have emergency surgery. I asked if she ever considered telling her doctor about the pills:

I would’ve not risked it. I keep denying it. I don’t know if that’s good or bad, but I think it’s bad. Because you should be able to tell your doctor what’s really going on with you. . . . But if you’re in shame, you cannot just tell them, and you’re just hoping they figure out what’s happening so they can save you.

Pharmacists, nurses, and doctors from Mexico often played important roles in participants’ ability to obtain medications and treatment, but given the social and legal risks of SMA, women were careful about what they disclosed to health care providers. Although this sometimes increased their health risks, nondisclosure helped them gain access to medical resources and support while shielding them from legal and social risk.

Discussion

In the year following the Dobbs v. Jackson Women’s Health Organization decision by the U.S. Supreme Court in June 2022, 11 states imposed complete bans on abortion, leaving nearly one in three reproductive-age women in the United States without access to abortion care (Center for Reproductive Rights 2022). In absence of formal care, the knowledge, support, and experience of friends, family members, or strangers within a person’s social network may become critical determinants of health access and equity. Yet the decision of whom to turn to for support presents serious risks for individual health and well-being.

Previous research of health-seeking behavior emphasizes the positive impact of disclosure—particularly to strong social ties—on health outcomes. For stigmatized health needs, I argue that people’s social networks play an especially critical role in health-seeking outcomes; while most respondents confided in a strong tie, the activation of weak ties and the avoidance and exclusion of social ties via nondisclosure were also critical health-seeking tools. Several factors were associated with weak-tie, but not strong-tie, disclosure: Respondents who lived in hostile policy climates, those who attempted SMA within the past year, and those seeking abortion at later gestational ages were more likely to confide in a weak tie. These groups may be more desperate to obtain care due to heightened stigma, limited options for care, and time constraints, thereby increasing their need to turn to people beyond their core networks.

Moreover, qualitative findings revealed that social avoidance, enacted through partial disclosure and nondisclosure, is also a critical tool for ensuring access to desired care. By strategically managing their activation of social ties, respondents sought to reduce stigma, access emotional support, ensure physical safety, safeguard reproductive autonomy, and avoid medical surveillance.

This study highlights the role of social networks among women seeking abortion, a common reproductive event subject to legal restrictions and intense forms of interpersonal and structural stigma. As such, their experiences shed light on the potential health-seeking strategies of other populations whose needs are stigmatized, including sexual and gender minorities, undocumented immigrants, survivors of intimate partner violence, people living with HIV, and those suffering from addiction, depression, or mental illness. Building on previous research by medical sociologists and network scholars that emphasize the health benefits of disclosure generally and the support and resources provided by strong social ties specifically, this case study draws attention to the potential harms that may be imposed by social tie activation during periods of health-seeking. People with stigmatized health needs may often choose to avoid disclosure to preserve their physical, social, and emotional well-being. Future network studies should employ name generators to account for nondisclosure, asking questions such as “Is there anyone that you avoided speaking to?” or “Were you worried that anyone in particular would find out?”

People with stigmatized health needs also experience “parallel processes of coercive institutional disclosure,” that is, they are obligated to share extensive personal information in exchange for needed services (Cho 2021; Hughes 2018:114; Page and Polk 2017; Watkins-Hayes 2013). Under such constraints, people may be increasingly motivated to conceal information from health care providers or even avoid the formal health care system altogether. Future research should extend beyond individual relationships to examine institutional engagement and avoidance during periods of health-seeking. By acknowledging both the risks and benefits of disclosure for health and well-being, researchers, policymakers, and health care providers may begin to better understand the ways (non)disclosure is wielded as a health-seeking tool and implement new strategies to support those seeking resources and support for stigmatized health needs.

Author Biography

Kathleen Broussard is an assistant professor in the Department of Sociology at the University of South Carolina. Her research investigates how people make meaning of reproduction and health care and navigate reproductive governance imposed by the state and formal health care system. Her work appears in journals such as Social Science & Medicine, Population Studies, JAMA Network Open, and American Journal of Public Health. Prior to joining the University of South Carolina in 2022, she completed her PhD in sociology at the University of Texas at Austin.