Sage Journals: Discover world-class research

Abstract

In the United States, sexuality and sexual relationships were foundational concepts for understanding women's prisons that generally receded from carceral research as the nation's incarceration rates climbed. This study reevaluates classic themes of sexuality and women's prison informal organization using social network and health data collected in two Pennsylvania women's prison units (n = 118). Sampled women perceived a high prevalence of prison-based sexual relationships and overwhelmingly endorsed the rights of sexual minorities. Additionally, network analyses found that sexual minority women had similar friendships, social status, and self-reported health as their heterosexual peers across both units, suggesting an absence of sexual stigmatization or group segregation processes. However, older, longer-term incarcerated women were more likely than their peers to hold unfavorable views toward prison sex. This was particularly evident in a “good behavior” unit holding many older, long-term residents where overt sexual behaviors could jeopardize the unit's stability and privileges.

Keywords

Sexuality incarceration networks social integration exponential random graph models health

Introduction

From its outset, carceral research in the United States emphasized sexuality and same-sex relationships as central for understanding the informal social organization of women's prisons (Ford, 1929). Merging the deprivation and importation perspectives of men's prison societies (Irwin and Cressey, 1962; Sykes, 1958), early scholars of women's prisons argued that same-sex relationships provide primary means for women to adapt to the pains of incarceration and help ameliorate the loss of valued pre-prison roles (Giallombardo, 1966; Ward and Kassebaum, 1965). Subsequent penological research continued to document the overrepresentation of same-sex relationships in women's prisons and connected these to prison informal social organization, to include the role of pseudo families and kinship networks for prison adaptation (Giallombardo, 1974; Heffernan, 1972; Propper, 1982). As the nation's incarceration rates spiked in the late twentieth century, American carceral research simultaneously declined (Simon, 2000; Wacquant, 2002) and scholars of women's prisons shifted their focus to the effects of punitiveness on women's prison experiences, with some arguing that the country's harsh and overcrowded prisons eroded peer trust, reduced same-sex relationships, and weakened prison pseudo families (Greer, 2000; Kruttschnitt and Gartner, 2005). Moreover, feminist scholars critiqued earlier emphases on prison homosexual relationships as stereotypical and discounting the complexity and diversity of incarcerated women's sexuality and gender identities (Bosworth, 1999; Carr et al., 2020; Kunzel, 2002; Owen et al., 2017).

More recently, and contemporaneous with the U.S. Government's Prison Rape Elimination Act of 2003, American prison scholars moved beyond same-sex relationships to examine incarcerated women's sexual attitudes (Blackburn et al., 2011; Hensley, 2000; Hensley et al., 2002). Although nascent, this literature consistently documents older, longer-term, and heterosexual women are least likely to support prison-based consensual sexual relationships and sexual behaviors. Moreover, research of contemporary women's prisons suggests that a half-century of hyper-incarceration increased the number of older, longer-term, “prisonized” (Clemmer, 1940), “colonized” (Goffman, 1961), or “adapted” (Kruttschnitt and Gartner, 2005) women viewing prison as home. These women are often critical of younger, shorter-term, women perceived as rambunctious and inattentive to prison informal social order (Owen et al., 2017). Simultaneously, recent surveys find sexual minority incarcerated women are significantly more likely to be incarcerated than their heterosexual peers (Jones, 2021; Meyer et al., 2017). The intersection of an aging prison population and overrepresentation of sexual minority women may alter prison society, yet limited access to state prisons and lack of relational data reduce researchers’ ability to rigorously investigate the phenomena.

In this study, we examine sexual identity, sexual attitudes, self-reported health, and prison informal social organization using unique data collected in a contemporary women's prison. Using data collected in 2018 from over 100 women incarcerated in a Pennsylvania medium/maximum security prison, we first replicate Ward and Kassebaum's (1965) analysis of the perceived prevalence of prison-based sexual behaviors with a more recent sample. We then compare self-reported sexual attitudes across two prison units: one a general population unit composed primarily of younger women serving shorter sentences, and the other a “good behavior” unit largely housing older and longer-term women. Using network data and analyses of prison friendship and status nominations, we investigate the social positions of heterosexual and sexual minority women and connect these to self- and peer-reported sexual attitudes and health. Following recent prison attitudinal research, we expect that older, longer-term residents are more likely to hold negative attitudes toward sexual minority women and same-sex sexual behaviors. If accurate, such patterns may result in health and social inequities among an already marginalized sexual minority population and hold important implications for gender-responsive prison policies.

Literature review

Ward and Kassebaum's (1964, 1965) seminal study of California's Frontera Prison was among the first to document a high prevalence of same-sex sexual relationships among incarcerated women and connect these to the informal social organization in women's prisons. The prison staff and incarcerated women the authors surveyed estimated that over 50% of imprisoned women were involved in prison-based sexual behaviors. Furthermore, and in contrast to men's adherence to an inmate code that enabled a solidary prison society (Sykes, 1958), the authors argued that same-sex relationships served as a primary means for women to adapt to prison's deprivations by relieving the loss of pre-prison familial ties. Accordingly, social organization in women's prisons was presented as “a non-cohesive aggregate of homosexual dyads and friendship cliques…[where] the needs of female prisoners are most often met by another [incarcerated] individual” (Ward and Kassebaum 1965:78). The authors thus placed same-sex relationships and sexual behaviors as compartmentalizing prisoner social structures and destabilizing prisoner cohesion, establishing clear distinctions to contemporaneous research of men's prisons (Irwin and Cressey, 1962; Sykes, 1958).

Simultaneous to Ward and Kassebaum (1964, 1965), Giallombardo (1966) examined the informal social organization within West Virginia's Alderson Federal Reformatory for Women. She also found that same-sex relationships and sexual behaviors were highly prevalent among her incarcerated participants. Of equal importance for the prison's informal social organization, however, were kinship or pseudo family ties created to meet the emotional and relational deprivations of incarceration. She concluded that prison pseudo families and their corresponding social roles functioned as cultural substitutes for pre-prison families and gender-based value systems imported into prison. Of particular importance for the broader social structure were “matricentric families of varying sizes” (Giallombardo, 1966:163), led by pseudo mothers with significant power and influence (see also Dillavou et al., 2022). Giallombardo's (1966) emphasis on pseudo families and Ward and Kassebaum's (1964, 1965) focus on prison sexual relationships dominated research of women's prison social systems for decades to follow.

In the latter half of the twentieth century, prison research generally declined as America retreated from rehabilitative ideals and prison scholars were often denied embedded prison access (Simon, 2000; Wacquant, 2002).¹ Even so, carceral studies in the U.S. continued to estimate high prevalences of same-sex sexual behaviors among incarcerated women compared to the general population, with self-reports ranging from 30% to 70% (Greer, 2000; Heffernan, 1972; Koscheski and Hensley, 2001; Meyer et al., 2017; Owen, 1998).² Simultaneously, some scholars argued that the prevalence of sexual behaviors in women's prisons had declined over time as increasingly punitive prison regimes eroded inmate personal trust (Greer, 2000; Kruttschnitt and Gartner, 2005). Moreover, several feminist criminologists critiqued the focus on stereotypical same-sex behaviors as too narrow. They looked to the narratives of incarcerated women to demonstrate the complexity of women's pathways to prison and adaptations to confinement, as well as the varying forms and purposes of prison relationships, such as situational versus pre-prison homosexuality, consensual versus coercive sexual activities, same-sex relationships with other incarcerated women versus with correctional officers, and gender identities (Bosworth, 1999; Carr et al., 2020; Kunzel, 2002; Owen et al., 2017).

Compared to the relatively extensive research on the prevalence of same-sex sexual behaviors in women's prisons (actual and perceived), attitudes toward such behaviors have received less empirical scrutiny. When researching same-sex relationships in prisons, Richmond (1978:52) argued, “sociologists too often stress actual [same-sex] behavior rather than searching for more difficult, but no less significant, evidence of overt and covert attitudes to behavior.” Spurred by societal changes in perceptions of same-sex relationships and research surrounding the Prison Rape Elimination Act of 2003, U.S. scholars have begun exploring incarcerated women's attitudes toward same-sex behaviors and relationships in prison. Hensley (2000) surveyed attitudes toward sexual behaviors among incarcerated men and women and found that being male, white race, heterosexual, and having a longer prison sentence were positively associated with homophobic attitudes. In a subsequent study of incarcerated women, Hensley et al. (2002) similarly found that younger women and those with same-sex experiences had less homophobic attitudes than peers. More recently, Blackburn et al. (2011) found that incarcerated women who were older, heterosexual, less educated, and previously sexually victimized were more likely to think prison sex was a “big deal” or desire that incarcerated women be segregated based upon their sexual orientation.

Findings that older and long-term incarcerated women generally hold more negative attitudes toward prison sex also hold implications for the informal social organization in contemporary women's prisons. Specifically, decades of hyper-incarceration in the United States have resulted in an aging prison population and contexts where older, long-term, women may be more likely to hold high status positions and, thus, exude more power and influence in peer networks (Kreager et al., 2021).³ If these women hold negative attitudes toward same-sex behaviors, then sexual minority women may be marginalized and stigmatized within the prison informal social system. Interestingly, such expectations are at odds with earlier sociological studies of women's prisons that presented same-sex sexual relationships as primary adaptations to incarceration's deprivations and thus as central, not peripheral, to women's prison society. As the association between age, sexual attitudes, and peer social status may vary across women's prison contexts, it would be beneficial to compare prisons, or prison units, with different age distributions. We accomplish this in the current study by comparing results from surveyed women in two different prison units.

Experiencing homophobic attitudes from older high-status peers may also negatively impact the health of sexual minority incarcerated women. Studies consistently document mental health disparities between sexual minority individuals and their heterosexual peers (Bostwick et al., 2014; Lee et al., 2016). In his minority stress model, Meyer (1995, 2003) posits that such disparities arise due to “prejudice events” where stigma, discrimination, and marginalization are experienced in interpersonal interactions, creating excess stress and potentially “internalized homophobia” where sexual minorities incorporate the discriminatory interactions into negative self-images. These disparities and stress could be greater in contexts where sexual minority women are further marginalized and lacking social support (Chard et al., 2015). Despite the disproportionately high rates of sexual minority individuals within the prison context, homophobia may add to already stressful situations and heighten negative health outcomes among sexual minority incarcerated women compared to heterosexual peers. Within the United States, certain religious faiths may be particularly strict in their views of homosexuality and criticize sexual minority peers. Specifically, Ellis (2020) found that devout Protestant incarcerated women were more likely than women who were non-religious or adherents of other faiths to condemn sexual minority peers as sinful and report same-sex behaviors as prison misconduct. We thus include a measure of protestant faith in our analyses. Finally, any negative health consequences associated with sexual minority status could be exacerbated in prison contexts dominated by older, long-term, incarcerated women as the latter are more likely than their peers to hold negative attitudes toward same-sex sexual behaviors. We test these varying expectations with survey and network data collected in two women's prison units: one “good behavior” unit with a large proportion of older, long-term, incarcerated women and the other a “general population” unit holding more younger and shorter-term women.

Current study

Participants

Our sample and data come from the Women's Prison Inmate Networks Study (WO-PINS), a project investigating prison informal social organization and the family reintegration process of incarcerated women in the U.S. state of Pennsylvania. WO-PINS researchers collected network, health, and sexuality survey data in the summer of 2018 from incarcerated residents in a “good behavior” unit and a “general population” unit of a Pennsylvania women's medium/maximum-security prison.⁴ The units were similarly designed, each holding a full capacity of 76 women in a single-story, open-bay format containing four bunk bed cubicles each. These two units resulted in a total of 152 potential study respondents.

Although the two units shared a physical layout and dayroom, they differed in administrative policies and residents’ demographic characteristics. To qualify for the good behavior unit, residents must have been misconduct-free for 12 months prior to entrance and remain so during their residence. In return, residents in the good behavior unit were permitted greater privileges (e.g., more freedom of movement on the unit, longer use of cable television and telephone time) than residents in general population units. A consequence of the good behavior unit's qualification criteria is that this unit holds significantly older and longer-term residents than the general population unit (see Table 1 and Measures section below).

Table 1.

Descriptive statistics by prison unit.

	Good behavior		Means	General population
	Mean/%	Std. dev.	test	Mean/%	Std. dev.
Perceived sex prevalence in prison	4.80	1.08		4.66	1.19
Sexual minority	29%			44%
Acceptance of sex on unit	2.67	1.14	*	3.13	1.20
Equal rights for sexual minority	95%			95%
Mental health index	3.17	0.62		3.09	0.70
Health change since in prison	1.59	0.73	***	2.09	0.67
Friendship indegree	6.24	3.60		4.98	3.81
Power/influence indegree	1.94	5.00		2.11	3.42
Age	44.02	11.11	**	38.06	9.62
25 and younger	6%			7%
26-35	13%			22%
36-50	23%			22%
51 and older	23%			7%
Days on unit (100-day units)	5.39	7.77	**	1.61	4.65
Offense of conviction ^a
Property	14%			31%
Drug	22%			33%
Violence	43%			24%
Other	21%			13%
Prison pseudo mother	30%		*	11%
Years in school	11.90	1.51		11.45	1.26
White	67%		*	84%
Protestant	41%			51%
Observations	n = 63			n = 55

* p < 0.05, ** p < 0.01, *** p < 0.001.

T-test for continuous and ordered variables; Pearson chi-square test for binary variables.

Kruttschnitt and Gartner (2005:59) reported offense categories for the 832 women incarcerated at Frontera prison in 1963: 49% property offense (forgery, theft, and burglary), 25% drug offense, 21% violent offense (assault, robbery, and murder/manslaughter), and 5% other offense.

Upon receiving Institutional Review Board approval, the WO-PINS principal investigator visited the sampled units two weeks prior to data collection to describe the study's purpose and encourage study participation. Participation was voluntary and not shared with prison staff. Following state statute, no monetary incentives were provided for participation (Smoyer et al., 2009). Graduate researchers completed the informed consent process and administered the capacity to consent questionnaires (University of California, San Diego Brief Assessment of Capacity to Consent, Jeste et al. 2007) as well as surveys to participants in the units’ shared dayroom. Researchers administered computer-assisted questionnaires, sat side-by-side with respondents in a site outside of hearing range, read questions aloud, and typed participant responses on internet-deactivated laptop computers. This arrangement, agreed to by prison administrators, was thought to balance respondent confidentiality, trust, and comprehension with survey efficiency and correctional staff visibility (i.e., researcher safety).

Of the 152 women who lived on the two units, 121 consented to participate in the study. However, two respondents were excluded from the study as they did not meet the capacity to consent threshold, resulting in 119 eligible surveys (78% response rate). For the current analysis, one general population respondent did not complete the sexual perceptions and identity survey items, leaving 118 (63 [83% of unit] good behavior and 55 [72% of unit] general population) responses for our statistical sample. A logit regression comparing participants to non-participants did not show significant differences in basic characteristics obtained through administrative records except that participants were more likely to be white compared to non-participants.⁵

Measures

Table 1 presents descriptive statistics for our analytic sample, by unit, and tests for between-unit statistical differences.

Perceived Prevalence of Sexual Behavior in Prison. The WO-PINS survey replicated Ward and Kassebaum's (1965:92) question regarding the prevalence of prison same-sex relationships, “A rough estimate of the number of women who have sexual affairs at one time or another with other women in prison would be…” Responses were on a six-point ordinal scale, ranging from [1] “5 percent of less” to [6] “90 percent or more.” There were no statistical differences in means between the two WO-PINS units. On average, WO-PINS respondents perceived more than 50% of women in the prison had engaged in sexual behavior.

Sexual Minority Status. Respondents were asked “Who are you sexually attracted to?” with possible responses of [1] “Only men,” [2] “Only women,” [3] “Both men and women,” or [4] “Neither men nor women.” All participants expressed some sexual attraction (i.e., no one answered [4]). We recoded [2] and [3] into a “sexual minority” indicator. The prevalence of sexual minority status was marginally greater (p < 0.10) in the general population unit (44%) than the good behavior unit (29%).

We measure sexual attitudes with two items: acceptance of sexual behavior on the unit and whether respondents agree that sexual minorities should have the same rights and opportunities as heterosexual individuals. The former measures sexual behavior attitudes specific to the respondents’ prison context, whereas the latter measures broad attitudes toward sexual equality.

Acceptance of Sexual Behavior on Unit. WO-PINS participants were asked, “In general, how acceptable do you think it is for women in your unit to have sex with other women?” Responses were on a five-point Likert scale, ranging from [1] “Very acceptable” to [5] “Very unacceptable.” We reverse coded the scale so that higher numbers on the measure reflect greater acceptance. There was a statistically significant difference between the two unit means (p < 0.05): on average, women in the general population unit held neutral views toward same-sex sexual behavior on their unit (µ = 3.13), whereas women in the good behavior unit had less favorable attitudes (µ = 2.67) of the same behavior on their unit.

Equal Rights for Sexual Minorities. This measure indicates an affirmative response to the question “Do you think that gay and lesbian individuals should have the same rights and opportunities as straight people?” Respondents in both units overwhelmingly supported equal rights for sexual minorities (95% responded “yes”).

Mental Health Index. The survey included an abbreviated Center for Epidemiological Studies Depression Scale (CES-D) (Radloff, 1977). Seven items from the CES-D scale (e.g., “I felt sad,” “I could not get going,” “My sleep was restless,” etc.) were used to assess WO-PINS respondents’ mental health (Grzywacz et al. 2016). Responses were on a four-point ordinal scale, ranging from [1] “rarely or none (< 1 day)” to [4] “most or all of the time (5-7 days)” referencing the prior week. Higher scores indicate poor mental health. There was no mean difference between the two units on this measure: µ = [3] “Occasionally or a moderate amount of time (3-4 days).”

Health Change in Prison. Respondents were asked, “Compared to three months before your current prison stay, how is your health different?” The response options were given on a three-point ordinal scale, ranging from [1] “My overall health is worse” to [3] “My overall health has improved.” On average, respondents in the good behavior unit felt their health got worse in prison ([1]), and respondents in the general population unit felt no change in their health since being incarcerated ([2]). Respondents in the good behavior reported significantly greater health declines than respondents in the general population unit (p < 0.001).

We measure peer social status/marginalization with two network measures: peer friendship nominations and peer power/influence nominations. The former identifies those more or less liked by peers, whereas the latter measures those who have more or less social status on the prison units.

Friendship Nominations. As incarcerated individuals are often cynical of “friendships” in prison (Crewe, 2009), we asked respondents to nominate “Who do you get along with most?” using up-to-date unit rosters. Participants were permitted an unlimited number of peer nominations. Received nominations were summed for each respondent where higher values indicate stronger peer acceptance on the unit. Friendship indegree (i.e., the number of friendship nominations received from peers) was marginally greater (p < 0.10) in the good behavior unit (µ = 6.24) compared to the general population unit (µ = 4.98).

Power/Influence Nominations. Like the friendship nomination process, respondents were asked to nominate “Unit residents you feel are the most powerful and influential?” from up-to-date unit rosters. This measure captures a woman's informal social status on each prison unit from the perspective of peers (Kreager et al., 2021). Participants were permitted to nominate as many peers as they wished. Received nominations were summed for each respondent. Women receiving more nominations held more powerful and influential positions on the unit. Average power/influence indegree was similar across the two units (µ = 2). Both the friendship and power/influence nomination data are outcomes in network analyses described below.

Age. This variable is derived from administrative data and measured in years. On average, women on the good behavior unit (µ = 44 years) were significantly older than women on the general population unit (µ = 38 years) (p < 0.01). This unit difference is most apparent at the older age range (51 years and older), where there are more than three times the women in this category residing in the good behavior unit compared to the general population unit (23% vs. 7%).

Days on Unit. Administrative records documented the number of days respondents resided on their respective unit on the day they completed the study survey. To increase interpretability of model estimates, we recoded the variable's metric to 100-day units. On average, women resided on the good behavior unit (µ = 5.39 [539 days]) significantly longer than women resided on the general population unit (µ = 1.60 [160 days]) (p < 0.01), a pattern largely explained by a difference in conviction categories. Women on the good behavior unit were almost twice as likely as women in the general population unit (43% vs. 24%) to be convicted for a violent offense.

Prison Pseudo Mother. Respondents were asked to nominate (including self-nominations) prison pseudo mothers on the unit. A total of 25 women (30% of the good behavior unit and 11% of the general population unit) were identified as prison pseudo mothers in the analytical sample. The difference between the two units on this measure is statistically significant (p < 0.05).

Years in School. Educational attainment is measured as the years of schooling a woman attained prior to incarceration. On average, women in the general population unit received 11 years of education, and women in the good behavior unit received 12 years of education. No statistical difference appears across the units.

White. Race comes from administrative records and indicates white race compared to non-white categories. More respondents in the general population unit (84%) were identified as white compared to those in the good behavior unit (67%). This difference is statistically significant (p < 0.05).

Protestant. Surveyed women were asked about their religious affiliation. This binary measure distinguishes women identifying as Protestant compared to other denominations/religions or non-religious categories (including other Christian branches, other religions, and atheists). Research suggests Protestants, in particular, are likely to negatively evaluate sexual minorities and non-heteronormative sexual behaviors for moral reasons (Ellis, 2020; Kerley et al., 2005).

Analytic strategy

We begin our analyses by comparing the perceived prevalence of same-sex sexual behavior in prison among our 118 respondents to the 263 participants in Ward and Kassebaum's (1965) study of Frontera prison. This replication connects a classic study to a contemporary sample and provides a starting point for the current analyses. We then present bivariate correlations between sexuality, sexual attitudes, health, and social relationships in the two surveyed prison units. These estimates provide preliminary insight into the associations between sexual minority status, health, and informal organization within a contemporary women's prison.

Next, we examine women's friendship ties and social status (i.e., friendships and power/influence nominations) in a prison unit conditioned on their sexuality and other personal traits likely correlated with social status (e.g., pseudo motherhood, age, and time on the unit). Our approach using social network analysis is threefold. First, we present figures of the friendship networks and power/influence networks by unit to visualize sexual minority women's positions in these two social structures. Next, we formally test if sexual minority women are segregated from others in the friendship networks by estimating modularity scores for that attribute in each network. Modularity is “[the] fraction of the edges in the network that connect vertices of the same type (i.e., within-community edges) minus the expected value of the same quantity divisions but random connections” (Newman and Girvan, 2004:7). Since the measure is chance-corrected, the modularity value will be zero when partitioning is no better than random. Values greater than zero demonstrate community structure based on the characteristic that exceeds what would be expected by chance. This measure has been used in criminology to measure a variety of topics, such as gang cohesion (e.g., Ouellet et al., 2019) and prison cultural schema (Young et al., 2023). In our case, we can use sexual minorities as a way of partitioning the friendship and status networks. Finally, we use exponential random graph models (ERGMs) to examine sexuality, social status, and attitudes toward on-unit sexual behavior. Social network analyses, especially ERGMs, are appropriate and beneficial for our study because of their ability to operationalize the concepts of interest (e.g., social exclusion, status) as network configurations, account for endogenous network processes, and test hypotheses using statistical inference. In particular, based on prior literature, we predict that (1) sexual minority incarcerated women have fewer prison friendships and less power/influence, (2) these associations will be stronger in the good behavior unit compared to the general population unit, and (3) attitudes toward (same-sex) sexual behavior will be more negative on the good behavior vs. general population unit.

ERGMs statistically evaluate an observed network by modeling the probability of a tie in a network as a function of network, actor, and dyad characteristics (for a review, see Duxbury, 2023). The ERGM framework allows us to estimate the likelihood of a friendship tie or a power/influence tie based on actor attributes (e.g., age, attitudes toward sexuality) while accounting for more complex dependencies that generate network structure (e.g., reciprocity, transitivity).⁶ Friendship and power/influence are modeled through a series of sender/outdegree, receiver/indegree, and homophily effects that evaluate the likelihood of a tie from person i to person j (i→j) conditioned upon i's and/or j's value on a given attribute. Homophily is measured using a match effect to indicate whether i and j are identical on a categorical attribute (yes = 1, no = 0) or an absolute difference effect to indicate whether i and j are more similar on a continuous attribute (smaller differences indicate greater similarity). In the model of power/influence, we control for the friendship network by including an edge covariate effect, representing whether the probability of an i→j power/influence nomination is dependent on the presence of an i→j tie in the friendship network. We also include an edges term to control for the baseline likelihood of a tie in the network, a mutuality term to capture the tendency for ties to be reciprocated, a twopath covariate that counts the number of two-paths in the network between i and j (i.e., i→k→j) in which there is also a tie from k to i, a geometrically weighted indegree (GWINDEGREE) term that takes into account the indegree distribution, and the geometrically weighted edge-wise shared partner (GWESP) term that models the tendency toward triadic closure. As shown in Table 1, the amount of nomination activity in the two power/influence networks is relatively infrequent, with each respondent receiving two nominations, on average. This, coupled with modest sample sizes, likely reduces statistical power and increases risk of Type II errors in these analyses. Simultaneously, however, these factors boost confidence in any observed statistically significant estimates.⁷

Results

Table 2 presents WO-PINS survey responses for the perceived prevalence of sexual behavior in prison, alongside the distribution of responses from Ward and Kassebaum's (1965:92) incarcerated sample. The WO-PINS participants had substantially higher estimates of same-sex sexual prevalence compared to the Frontera sample collected over a half-century ago. Both the mode (70% vs. 30%) and mean (4.72 vs. 3.51) were greater for WO-PINS participants compared to the Frontera sample. Indeed, on average, the WO-PINS women perceived that over half of incarcerated women engaged in same-sex sexual behaviors (compared to 30% in Ward and Kassebaum's [1965] sample). This result suggests that perceptions of such behaviors have certainly not declined in a contemporary prison setting despite the systemically punitive atmosphere that accompanied an era of hyper-incarceration in the United States. Indeed, more liberal societal attitudes towards same-sex relationships (Kranjac and Wagmiller, 2022) may help to explain increased perceptions of same-sex sexual behaviors in women's prison over time.

Table 2.

Estimates of same-sex sexual prevalence in two prisons.

Estimate	WO-PINS (2018)	Ward and Kassebaum (1965)
[1] 5 percent or less	2%	12%
[2] 15 percent	3%	12%
[3] 30 percent	10%	25%
[4] 50 percent	19%	22%
[5] 70 percent	41%	22%
[6] 90 percent or more	25%	7%
Observations	113^a	263
Mean ([1]−[6] range)	4.72	3.51

Five women in the analytical sample did not answer this question.

Comparing sexuality measures across the two units (see Table 1), we do not find significant attitudinal difference towards homosexuality. In fact, women in both units overwhelmingly (95%) supported equal rights for sexual minority women. We also do not find significant unit differences in the proportions of women identifying as sexual minorities. However, there was a significant unit difference in the acceptance of sexual behavior on the unit: women in the good behavior unit were significantly less likely to accept such behavior than the women in the general population unit. We explore further this unit difference with social network analyses below.

Table 3 presents bivariate correlations for our covariates. Women who are older, have been incarcerated longer, and are perceived as a prison pseudo mother are significantly (p < 0.05) more likely to reside in the good behavior unit. They are also more likely to receive friendship nominations and be considered powerful or influential by their unit peers. Additionally, these women are more likely to be heterosexual and less likely to accept on-unit sexual behavior. In contrast, sexual minority women are more likely to be younger than their peers and more tolerant of unit sexual behavior. There are no significant correlations between sexuality and the self-reported health measures, suggesting that sexual minority women's health was no different, and not perceived as worsening during incarceration, compared to their heterosexual peers. Similarly, neither friendship indegree nor power/influence indegree is correlated with sexuality, suggesting that sexual minority women were just as likely as their peers to be nominated as a friend or person of status.

Table 3.

Bivariate correlations.

		(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)
(1)	General population unit
(2)	Sexual minority	0.25
(3)	Acceptance of sex on unit	0.20	0.17
(4)	Mental health index	−0.05	−0.14	−0.19
(5)	Health change since in prison	0.33	−0.02	−0.03	0.19
(6)	Friendship indegree	−0.17	−0.06	−0.28	0.23	−0.14
(7)	Power/influence indegree	0.02	0.14	−0.11	0.06	−0.09	0.47
(8)	Age (logged)	−0.27	−0.27	−0.28	0.07	−0.27	0.16	0.14
(9)	Days on unit (100-day units)	−0.28	0.06	−0.19	−0.07	−0.21	0.38	0.30	0.33
(10)	Prison pseudo mother	−0.42	−0.41	−0.32	0.13	−0.12	0.35	0.28	0.43	0.19
(11)	Years in school	−0.16	0.00	0.04	0.02	−0.26	0.02	0.02	0.18	0.04	−0.03
(12)	White	0.33	−0.09	−0.03	−0.03	0.11	−0.03	−0.09	−0.10	−0.18	0.03	0.11
(13)	Protestant	0.16	−0.03	−0.06	−0.13	−0.03	0.00	−0.10	−0.03	0.09	0.01	0.09	−0.24

Bold: p < 0.05.

Correlations based on variable distributions (i.e., Pearson, Tetrachoric, Polychoric, Spearman, Kendall Tau, Point Biserial).

To further investigate connections between sexuality, sexual attitudes, and unit social organization, we next examine visualizations of each network and then interpret ERGMs modeling the probability of a tie in the unit-level friendship and power/influence networks.

Figure 1 illustrates sexual minority women's positions in the units’ friendship networks. Visually, sexual minority women (shaded nodes) do not appear marginalized or highly clustered in either unit's friendship network structure. To test this more formally, the modularity scores in both the good behavior unit (−0.013) and general population unit (0.021) are very close to zero, suggesting that the structural isolation or clustering of sexual minority women are no greater than found by chance.

Figure 1.

Friendship (ties) and sexual minority status (red/shaded nodes).

To further examine sexual minority status and friendship network structure, Table 4 lists ERGM estimates predicting friendship ties. In the good behavior unit, we see that sexual minority women are more likely to send friendship nominations to peers. We also find that sexual minority women are no less likely to receive friendship ties compared to their heterosexual counterparts.⁸ Additionally, women who experience a change in health are less likely to receive friendship ties, women who are older and white are more likely to send friendship ties, and women nominated as a prison mother are less likely to send friendship ties. In terms of homophily, women who are similar regarding experiencing a health change and years in school are more likely to form a friendship tie.

Table 4.

Exponential random graph models for friendship network.

	Good behavior		General population
	Estimate	Standard error	Estimate	Standard error
Edges	−6.222***	(1.421)	−3.567*	(1.615)
Mutual	1.428***	(0.210)	1.884***	(0.266)
Two-Path	−0.141***	(0.015)	−0.199***	(0.018)
GW indegree (decay = 1.50)	0.169	(0.327)	−0.533	(0.323)
GWESP (alpha = 0.75)	1.176***	(0.076)	1.205***	(0.085)
Sexual minority
Receiver	0.174	(0.118)	−0.002	(0.107)
Sender	0.235*	(0.111)	−0.013	(0.117)
Match	−0.049	(0.112)	0.082	(0.098)
Acceptance of sex on unit
Receiver	−0.069	(0.042)	0.086	(0.052)
Sender	0.074	(0.042)	−0.037	(0.054)
Match	−0.011	(0.045)	−0.103*	(0.050)
Mental health index
Receiver	0.149	(0.081)	0.111	(0.082)
Sender	−0.101	(0.078)	−0.056	(0.092)
Match	−0.092	(0.092)	0.063	(0.107)
Health change since in prison
Receiver	−0.177*	(0.086)	−0.060	(0.084)
Sender	−0.122	(0.083)	−0.034	(0.090)
Match	0.231*	(0.097)	−0.019	(0.085)
Age (logged)
Receiver	0.346	(0.240)	0.208	(0.231)
Sender	0.593**	(0.221)	−0.303	(0.263)
Match	−0.131	(0.235)	−0.022	(0.274)
Days on unit (100-day units)
Receiver	−0.004	(0.008)	0.082	(0.048)
Sender	0.005	(0.008)	0.096*	(0.047)
Match	0.012	(0.010)	−0.054	(0.049)
Prison pseudo mother
Receiver	0.008	(0.141)	0.190	(0.300)
Sender	−0.283*	(0.120)	−0.292	(0.295)
Match	−0.028	(0.125)	0.076	(0.292)
Years in school
Receiver	-.007	(0.032)	−0.062	(0.046)
Sender	−0.018	(0.031)	0.089	(0.052)
Match	−0.058*	(0.028)	0.022	(0.049)
White
Receiver	−0.153	(0.116)	0.577	(0.500)
Sender	0.322**	(0.106)	0.713	(0.506)
Match	0.160	(0.088)	−0.509	(0.503)
Protestant
Receiver	−0.129	(0.102)	0.067	(0.110)
Sender	0.104	(0.093)	−0.039	(0.117)
Match	0.122	(0.091)	−0.006	(0.117)
AIC	2078.95		1463.85
BIC	2298.40		1675.00
Observations	n = 63		n = 55

* p < 0.05, ** p < 0.01, *** p < 0.001.

Shifting to the general population unit, women who have been on the unit longer send more friendship ties and women who are similar in their attitudes toward sexuality are more likely to form a friendship tie. Again, sexual minority women do not appear to receive fewer friendship nominations than their peers in the general population unit. Together, the sexual minority estimates are consistent with Figure 1 and suggest that sexual minority women are not socially isolated in the friendship networks.

Figure 2 displays the positions of sexual minority women in the units’ power/influence networks. As in the friendship networks, sexual minority women do not appear to primarily occupy peripheral positions in the status network, nor do they cluster into groups within this social structure. The modularity values for sexual minority women in the status networks are 0.003 (good behavior unit) and −0.029 (general population unit), both very close to zero. These findings again suggest that sexual minority women are no more socially isolated or clustered in the power/influence network than their heterosexual peers.

Figure 2.

Power/influence (ties) and sexual minority status (red/shaded nodes).

Table 5 presents ERGM results predicting power/influence ties in the two units. First, we see that in both units having a friendship tie is a strong predictor of having a power/influence tie. Therefore, it is important to keep in mind that other model estimates are net of the effect of the friendship network. In the good behavior unit, we see that women who are more accepting of sexual behavior on the unit are less likely to receive power/influence ties. In other words, women who are powerful and influential in the good behavior unit are less accepting of sexual behavior on that unit. We do not observe this relationship in the general population unit.⁹ Additionally, women with more education are more likely to send power/influence ties. In the general population unit, we do not see any significant predictors of power/influence nominations, aside from the friendship nominations.

Table 5.

Exponential random graph models for power/influence network

	Good behavior		General population
	Estimate	Standard error	Estimate	Standard error
Edges	−4.681	(2.957)	0.647	(3.159)
Mutual	N/A^a	N/A	0.901	(0.564)
Two-path	−0.256***	(0.065)	−0.094	(0.049)
GW indegree (Decay = 1.50)	−3.168***	(0.333)	−2.413***	(0.376)
GWESP (Alpha = 0.75)	1.076***	(0.216)	0.348	(0.188)
Friendship network
Edge covariate	2.225***	(0.215)	2.335***	(0.228)
Sexual minority
Receiver	−0.187	(0.182)	−0.085	(0.194)
Sender	−0.121	(0.270)	−0.223	(0.258)
Match	0.147	(0.269)	−0.075	(0.221)
Acceptance of sex on unit
Receiver	−0.169*	(0.068)	−0.049	(0.085)
Sender	−0.177	(0.108)	0.029	(0.114)
Match	−0.003	(0.113)	−0.099	(0.116)
Mental health index
Receiver	−0.119	(0.109)	0.000	(0.156)
Sender	0.027	(0.188)	0.163	(0.204)
Match	−0.101	(0.195)	−0.244	(0.223)
Health change since in prison
Receiver	0.016	(0.098)	−0.039	(0.149)
Sender	0.163	(0.158)	−0.054	(0.204)
Match	−0.217	(0.175)	−0.183	(0.182)
Age (logged)
Receiver	0.074	(0.322)	−0.529	(0.433)
Sender	0.230	(0.463)	−0.176	(0.570)
Match	0.737	(0.490)	−0.451	(0.550)
Days on unit (100-day units)
Receiver	−0.004	(0.015)	−0.015	(0.099)
Sender	−0.017	(0.020)	0.008	(0.100)
Match	−0.006	(0.022)	0.031	(0.103)
Prison pseudo mother
Receiver	0.018	(0.199)	0.175	(0.407)
Sender	−0.231	(0.277)	0.807	(0.520)
Match	0.036	(0.270)	0.095	(0.457)
Years in school
Receiver	0.035	(0.073)	−0.023	(0.074)
Sender	0.177*	(0.081)	−0.062	(0.100)
Match	−0.104	(0.072)	0.167	(0.086)
White
Receiver	−0.213	(0.158)	0.263	(0.551)
Sender	−0.296	(0.236)	0.053	(0.625)
Match	0.047	(0.235)	−0.309	(0.600)
Protestant
Receiver	−0.093	(0.148)	0.335	(0.206)
Sender	0.207	(0.229)	0.399	(0.256)
Match	0.131	(0.245)	−0.182	(0.216)
AIC	761.40		760.95
BIC	980.00		978.12
Observations	n = 63		n = 55

* p < 0.05, ** p < 0.01, *** p < 0.001

This term is not estimable because no mutual ties are observed in the network for the good behavior unit. It is excluded from the analyses.

Discussion

This study returns to a topic at the heart of women's carceral research: the association between sexuality and prison informal social organization. Against expectations, we find that, despite the systemically punitive atmosphere associated with the era of hyper-incarceration, incarcerated women in a contemporary sample perceive the prevalence of prison sexual behavior to be greater than Ward and Kassebaum's (1965) finding over a half-century ago. Indeed, most of the 118 women in our incarcerated sample believed that 70% or more of women incarcerated in their prison engaged in same-sex sexual behavior, compared to 30% in Ward and Kassebaum's sample. Due to definitional and social desirability challenges associated with measuring sexual behavior (e.g., sexual behavior violates prison policy and may threaten outside heterosexual marriages or relationships), it is difficult to determine how such perceptions correlate with actual rates of sexual behavior in contemporary women's prisons. However, the relatively high perceptions of prison sex prevalence observed in this study suggest that prison sexuality continues to be an important aspect of women's prison experiences.

Significant numbers of women in our sample identified as sexual minorities in the good behavior (29%) and general population (44%) units, suggesting a continued overrepresentation of incarcerated sexual minority women. However, these women were indistinguishable from their peers on measured health and social integration outcomes. Sexual minority status was not correlated with the mental and overall health outcomes measured in this study. Similarly, sexual minority women were not in peripheral positions in their units’ friendship networks. Net of other covariates and network properties, ERGM estimates found no differences in the friendship nominations received by sexual minority women and their heterosexual peers. We thus found little evidence that sexual minority women were stigmatized, socially segregated, or otherwise negatively impacted relative to other incarcerated women. Indeed, 95% of participants in both units reported that sexual minorities should be provided equal rights to those of their heterosexual peers.

The one area where there was observed unit-level differences was in attitudes toward sexual behaviors on the unit. Here, respondents in the good behavior unit, on average, held significantly more negative views toward sex on the unit than did women in the general population unit. Our correlational and social network analyses suggested that an explanation for this unit difference was that the good behavior unit held more older, long-term, and high-status women (equivalent to the “old head” concept presented in Kreager et al., 2017, 2021) than did the general population unit and that these women were more likely to hold negative views toward unit sexual behavior. In other words, there were greater numbers of older, long-term, high-status, and predominantly heterosexual women in the good behavior unit and these powerful women held more negative views toward same-sex sexual behaviors on their respective unit.

This finding begs the question of why these women would discourage sex on both units while also holding progressive views toward sexual equality and the lack of social segregation of sexual minority women in the units? The answer, we argue, lies in the social consequences of overt sexual behaviors for older women seeking to peacefully “do their time.” As Owen et al. (2017) observed, prison sexual relationships are commonly “messy” and accompanied by jealousy, coercion, or exploitation, all of which disturb the peace in prison. Indeed, scholars have argued that sexual relationships in female prisons can be accompanied by economic exploitation, and that such relationships tend to be unstable, fluid, and conflict-producing (Greer, 2000; Hensley et al., 2002). Women who “keep their head down” and want to “do time quietly” are likely to hold a negative attitude towards behaviors that are related to conflicts and rule violations. This would be particularly true for older, long-term, and high-status women in a good behavior unit who value its privileges and would not want them jeopardized by the conduct of others. In prison settings where “old heads” hold sway, we suggest, norms discouraging behaviors that draw the attention of authorities are more likely to be enforced. This argument is then the female equivalent of Sykes’ (1958) “right guy” in men's prisons, but instead of focusing on curbing aggression to increase unit solidarity and peace, the female “old head” seeks to reduce sexual activity on the unit for the same outcomes. We suggest future qualitative research further examine this hypothesis through the narratives of incarcerated women.

Although our study makes notable contributions to the understanding of sexuality and social organization in today's women's prisons, there remain limitations that qualify our findings. Most notably, we lack detailed information on the relational contexts of sexual behaviors and related attitudes on the observed units. For example, we do not have measures of perceived social safety on the prison units, which, as suggested in prior literature, can influence women's openness in expressing their sexual identity, attitudes, and behavior (Corteen 2022). We also do not have measures of consensual sex or sexual coercion and how perceived coerciveness may impact sexual attitudes. Such behaviors and attitudes are notoriously difficult to measure with survey data and, absent additional contextual information, are open to the critique of reinforcing existing stereotypes (Owen et al., 2017). Indeed, the WO-PINS investigators recognized this possibility, along with the potential discomfort respondents might experience in answering questions about specific romantic and sexual relationships, prompting them to exclude such items from the prison survey questionnaire. Future qualitative work focused on this topic is better suited to understand the contexts and nuance of prison sex.

We also hesitate to generalize our findings to all women's prison settings. The relatively high response rates for each unit build confidence in the between-unit comparisons, but we remain uncertain how representative these units are to others inside and outside of Pennsylvania. We encourage future research to investigate the generalizability of the patterns we observe. In particular, trends in women's incarceration, sexual attitudes, religiosity, and health during imprisonment are likely to vary considerably across national, regional, and cultural contexts.

Finally, we also note that our relatively small sample size reduces statistical power to observe small statistical associations. The network analyses somewhat ameliorate this problem by using tie (i.e., nomination) activity, but even these estimates can have reduced precision in sparse networks. Future network studies with larger incarcerated samples would increase statistical power, provide more precise estimates, and reduce potential Type II errors.

Limitations aside, our results contribute to an important and longstanding literature using novel and rich data. Using network and survey data, we provide a new window into sexuality and informal social organization in contemporary women's prisons. Moreover, our findings suggest that incarcerated sexual minority women are not “doubly marginalized” in an overall punitive correctional context, but instead successfully integrate with their heterosexual peers. For these women, they are not more likely to experience discrimination from their peers inside than out. However, in situations where sexual behaviors are perceived as compromising unit social stability, such behaviors may be discouraged or punished by high status older peers, reducing overt sexual relationships and pushing sexual behaviors underground.

Footnotes

Acknowledgements

This research is supported by funding from the National Institute of Justice (2016-MU-MU-0011) and the Penn State Criminal Justice Research Center. The content is solely the responsibility of the authors and does not necessarily represent the official views of funders.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Institute of Justice, (grant number 2016-MU-MU-0011).

ORCID iD

Derek A Kreager

Notes

References

Blackburn

Fowler

Mullings

, et al. (2011) Too close for comfort: exploring gender differences in inmate attitudes toward homosexuality in prison. American Journal of Criminal Justice 36(1): 58–72.

Bostwick

Boyd

Hughes

, et al. (2014) Discrimination and mental health among lesbian, gay, and bisexual adults in the United States. The American Journal of Orthopsychiatry 84(1): 35–45.

Bosworth

(1999) Engendering Resistance: Agency and Power in Women’s Prisons. London, UK: Routledge.

Carlen

(1983) Women's Imprisonment: A Study in Social Control. London, UK: Routledge.

Carr

Serisier

McAlister

(2020) Sexual deviance in prison: queering identity and intimacy in prison research. Criminology and Criminal Justice 20(5): 551–563.

Chard

Finneran

Sullivan

, et al. (2015) Experiences of homophobia among gay and bisexual men: results from a cross-sectional study in seven countries. Culture, Health and Sexuality 17(10): 1174–1189.

Clemmer

(1940) The Prison Community. New York, NY: Christopher Publishing House.

Corteen

(2022) Lesbian safety talk: problematizing definitions and experiences of violence, sexuality and space. Sexualities 5(3): 259–280.

Crewe

(2009) The Prisoner Society: Power, Adaptation, and Social Life in an English Prison. Oxford: Oxford University Press.

10.

Dillavou

Kreager

Greenfelder

, et al. (Online First) Mothers inside and out? Pseudo-families and motherhood in a women’s prison. Crime and Delinquency: 001112872210838. https://doi.org/10.1177/00111287221083884.

11.

Dirsuweit

(1999) Carceral spaces in South Africa: a case study of institutional power, sexuality and transgression in a women’s prison. Geoforum; Journal of Physical, Human, and Regional Geosciences 30(1): 71–83.

12.

Duxbury

(2023) Longitudinal Network Models. Quantitative Applications in the Social Sciences. Thousand Oaks, CA: Sage Publications, Inc.

13.

Ellis

(2020) Redemption and reproach: religion and carceral control in action among women in prison. Criminology 58(4): 747–772.

14.

Ford

(1929) Homosexual practices of institutionalized females. The Journal of Abnormal and Social Psychology 23(4): 442–448.

15.

Giallombardo

(1966) Society of Women: A Study of a Women’s Prison. New York: John Wiley and Sons.

16.

Giallombardo

(1974) The Social World of Imprisoned Girls: A Comparative Study of Institutions for Juvenile Delinquents. New York: Wiley.

17.

Goffman

(1961) Asylums: Essays on the Social Situations of Mental Patients and Other Inmates. Oxford, England: Doubleday (Anchor).

18.

Greer

(2000) The changing nature of interpersonal relationships in a women’s prison. The Prison Journal 80(4): 442–468.

19.

Grzywacz

Hovey

Seligman

, et al. (2016) Evaluating short-form versions of the CES-D for measuring depressive symptoms among immigrants from Mexico. Hispanic Journal of Behavioral Sciences 28(3): 404–424.

20.

Handcock

Hunter

Butts

, et al. (2003) statnet: Software tools for the Statistical Modeling of Network Data. https://statnet.org/.

21.

Heffernan

(1972) Making It in Prison: The Square, the Cool, and the Life. Hoboken, NJ: Wiley-Interscience.

22.

Hensley

(2000) Attitudes toward homosexuality in a male and female prison: an exploratory study. The Prison Journal 80(4): 434–441.

23.

Hensley

Wright

Koscheski

, et al. (2002) Examining the relationship between female inmate homosexual behavior and attitudes toward homosexuality and homosexuals. International Journal of Sexuality and Gender Studies 7(4): 293–306.

24.

Howe

(1994) Punish and Critique: Towards a Feminist Analysis of Penality. London, UK: Routledge.

25.

Hunter

Handcock

Butts

, et al. (2008) ERGM: a package to fit, simulate and diagnose exponential-family models for networks. Journal of Statistical Software 24(3): nihpa54860.

26.

Irwin

Cressey

(1962) Thieves, convict, and the inmate culture. Social Problems 10: 142–155.

27.

Jeste

Palmer

Appelbaum

, et al. (2007) A new brief instrument for assessing decisional capacity for clinical research. Archives of General Psychiatry 64(8): 966–974.

28.

Jones

(2021) Visualizing the Unequal Treatment of LGBTQ People in the Criminal Justice System. The Prison Policy Initiative. Available at: https://www.prisonpolicy.org/blog/2021/03/02/lgbtq/.

29.

Kerley

Matthews

Blanchard

(2005) Religiosity, religious participation, and negative prison behaviors. Journal for the Scientific Study of Religion 44(4): 443–457.

30.

Koscheski

Hensley

(2001) Inmate homosexual behavior in a southern female correctional facility. American Journal of Criminal Justice 25: 269–277.

31.

Kranjac

Wagmiller

(2022) Attitudinal change, cohort replacement, and the liberalization of attitudes about same-sex relationships, 1973–2018. Sociological Perspectives 65(2): 262–277.

32.

Kreager

Young

JTN

Haynie

, et al. (2017) Where “old heads” prevail: inmate hierarchy in a men’s prison unit. American Sociological Review 82(4): 685–718.

33.

Kreager

Young

JTN

Haynie

, et al. (2021) In the eye of the beholder: meaning and structure of informal Status in women’s and men’s prisons. Criminology 59(1): 42–72.

34.

Krivitsky

Hunter

Morris

, et al. (2023) ERGM 4: new features for analyzing exponential-family random graph models. Journal of Statistical Software 105: 1–44.

35.

Kruttschnitt

Gartner

(2005) Marking Time in the Golden State. Cambridge, UK: Cambridge University Press.

36.

Kunzel

(2002) Situating sex: prison sexual culture in the mid-twentieth-century United States. GLQ: A Journal of Lesbian and Gay Studies 8(3): 253–270.

37.

Lee

Gamarel

Bryant

, et al. (2016) Discrimination, mental health, and substance use disorders among sexual minority populations. LGBT Health 3(4): 258–265.

38.

Meyer

(1995) Minority stress and mental health in gay men. Journal of Health and Social Behavior 36(1): 38–56.

39.

Meyer

(2003) Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: conceptual issues and research evidence. Psychological Bulletin 129(5): 674–697.

40.

Meyer

Flores

Stemple

, et al. (2017) Incarceration rates and traits of sexual minorities in the United States: national inmate survey, 2011-2012. American Journal of Public Health 107: 267–273.

41.

Newman

MEJ

Girvan

(2004) Finding and evaluating community structure in networks. Physical Review E 69(2): 026113.

42.

Ouellet

Bouchard

Charette

(2019) One gang dies, another gains? The network dynamics of criminal group persistence. Criminology 57(1): 5–33.

43.

Owen

(1998) In the Mix: Struggle and Survival in a Women’s Prison. Albany, NY: SUNY Press.

44.

Owen

Wells

Pollock

(2017) In Search of Safety: Confronting Inequality in Women’s Imprisonment. Oakland, CA: University of California Press.

45.

Propper

(1982) Make-Believe families and homosexuality among imprisoned girls. Criminology 20(1): 127–138.

46.

Radloff

(1977) The CES-D scale: a self-report depression scale for research in the general population. Applied Psychological Measurement 1(3): 385–401.

47.

Richmond

(1978) Fear of homosexuality and modes of rationalisation in male prisons. Australian and New Zealand Journal of Sociology 14(1): 51–57.

48.

Simon

(2000) The society of ‘Captives’ in the era of hyper-incarceration. Theoretical Criminology 4(3): 285–308.

49.

Smoyer

Blankenship

Belt

(2009) Compensation for incarcerated research participants: diverse state policies suggest a new research agenda. American Journal of Public Health 99(10): 1746–1752.

50.

Sykes

(1958) The Society of Captives: A Study of A Maximum Security Prison. Princeton, NJ: Princeton University Press.

51.

Wacquant

(2002) The curious eclipse of prison ethnography in the age of mass incarceration. Ethnography 3(4): 371–397.

52.

Ward

Kassebaum

(1964) Homosexuality: a mode of adaptation in a prison for women. Social Problems 12(2): 159–177.

53.

Ward

Kassebaum

(1965) Women’s Prison: Sex and Social Structure. Hawthorne, CA: Aldine Pub. Co.

54.

Worrall

(1990) Offending Women: Female Lawbreakers and the Criminal Justice System. London, UK: Routledge.

55.

Young

JTN

Meyers

(2023) What is “prison culture”? developing a theoretical and methodological foundation for understanding cultural schema in prison. Criminology 61(3): 421–448.

Sexuality,social organization,and health in a women's prison

Abstract

Keywords

Introduction

Literature review

Current study

Participants

Measures

Analytic strategy

Results

Discussion

Footnotes

Acknowledgements

Declaration of conflicting interests

Funding

ORCID iD

Notes

References