Associations between Police Harassment and HIV Vulnerabilities among Men Who Have Sex with Men and Transgender Women in Jamaica.

Background

The criminalization of same-sex practices constrains HIV prevention for gay, bisexual, and other men who have sex with men (MSM) and, in part due to the conflation of gender and sexuality, transgender women. Criminalization is a structural driver of HIV that indirectly influences HIV vulnerability through multiple pathways: decreased funding for HIV prevention, treatment, and care programs tailored for MSM and transgender women; increased fear of seeking health care; denial of services due to stigma; social and familial exclusion that may contribute to elevated rates of homelessness; employment and housing discrimination that elevate economic insecurity and increase survival sex work; and a lack of human rights protection that increases exposure to violence from community members and the police. Criminalization may result in enacted stigma, such as overt forms of social exclusion and violence, and perceived stigma, whereby people experience fear and concerns of rejection and negative treatment by others because of actual or perceived sexual or gender minority identity.

There is scant evidence directly linking human rights violations of MSM and transgender women to HIV vulnerabilities in middle-income contexts where same-sex practices are criminalized. MSM in Jamaica have the highest HIV rates in the Caribbean, estimated between 14% and 31%. A recent study of transgender women in Jamaica reported an HIV prevalence of 25% among this group and reported that HIV infection was associated with violence. Qualitative studies have highlighted that violence targeting sexually and gender diverse people in Jamaica compromises their human rights and well-being.

The criminalization of same-sex practices in Jamaica dates back to 1864, during British colonial rule, with article 76 of the Offences Against the Person Act, which states that “buggery” (anal intercourse) is punishable by up to 10 years of imprisonment with possible hard labor. Under this provision, MSM and transgender women who are mislabeled as male, a concept known as misgendering, can also receive up to two years of imprisonment with possible hard labor if convicted of “being a male person who is party to the commission of any act of gross indecency with another male person.” Advocates suggest that arrest and prosecution are rare; instead, the law is used to justify other human rights violations, such as discrimination in employment, health, and housing, as well as violence. Human rights violations are not easily challenged given that sexual orientation and gender identity are not protected under Jamaica’s Charter of Fundamental Rights and Freedoms. A 2014 study by Human Rights Watch interviewed LGBT community members in Jamaica (n=71) and found that more than half had been victims of homophobic or transphobic violence. Over one-third had reported crimes to the police, who took formal statements in eight cases, resulting in only four arrests.

Some studies have begun to describe the impact of the criminalization of same-sex practices and homosexuality and, to a lesser extent, police harassment on HIV vulnerability among MSM and transgender women. A quantitative study conducted by Sonya Arreola et al. among MSM (n=3,340) from 115 countries found that lower levels of access to HIV prevention, testing, and treatment were associated with criminalization based on sexual orientation and gender identity and expression. In Nigeria, Sheree Schwartz et al. found that fear of seeking and avoidance of health care were higher for MSM after the country’s implementation of the Same Sex Marriage Prohibition Act. In Jamaica, current or previous incarceration due to being transgender was associated with substantially reduced odds of HIV testing among transgender women. Similarly, ever having been in jail was associated with increased odds of HIV infection among MSM in Jamaica. And in India, transgender women sex workers report experiencing such relentless police harassment that they are often forced to relocate and work in unfamiliar settings, decreasing their choice of clients and safety, which in turn increases their HIV vulnerability.

Utilizing Jamaica as a case study, this essay examines factors associated with police harassment targeting MSM and transgender women. We aim to demonstrate how police harassment in contexts where consensual same-sex sexual relations are criminalized shapes HIV vulnerabilities and operates as a social driver of HIV for MSM and transgender women.

Methodology

We conducted a cross-sectional study with gay, bisexual, and other MSM, as well as transgender women, in Kingston, Ocho Rios, and Montego Bay in 2015 to examine social drivers and protective factors influencing HIV and STI vulnerability among sexual and gender minorities in Jamaica. Participants were recruited using a chain referral sampling method by peer research assistants and hired and trained staff who self-identified as gay, bisexual, or other sexual or gender minorities. All participants completed a tablet-based survey with some overlapping and some unique questions tailored to their experience as MSM or transgender women. Written informed consent was sought at the time of the interview. The Research Ethics Board at the University of Toronto in Canada and the University of the West Indies, Mona Campus, in Kingston, Jamaica, provided approval for the study (Protocol #: 30130-UT; ECP 27, 13/14 UWI). Detailed descriptions of the methods and measures are documented elsewhere.

For this analysis, the main outcome “ever experiencing police harassment” was measured by asking, “How often have you been harassed by police for being gay or bisexual (for gay, bisexual and MSM) or for being trans (for transgender women),” dichotomized as “never” and “ever” (for those who reported sometimes, many times, or always).

Individual- and historical-level factors measured included age (continuous, years), education (less than high school versus high school or higher), monthly income (continuous, USD), HIV status (positive versus negative), and depression (continuous, measured using the Patient Health Questionnaire-2).

Interpersonal level factors measured included social support (continuous, measured using a brief social support sub-scale to assess unmet social support needs), consistent condom use (dichotomous, yes versus no; participants were coded as practicing “consistent condom use” if there was parity in the number of times participants reported having sex and using condoms), relationship status (categorical: in relationships/casual dating, no partner, concurrent partners), safer sex self-efficacy (continuous, using a scale for negotiating safer sex), and physical violence (dichotomous, ever versus never).

Structural and environmental factors measured included any sex work in the past 12 months, food insecurity (dichotomous, yes versus no; participants were coded as “food insecure” if they reported at least one occurrence of going to bed hungry in a week), unstable housing (dichotomous, yes versus no; participants were coded as having unstable housing if they usually slept outside, in a shelter, or at a friend’s or relative’s house), personal experiences of perceived sexual stigma (continuous, five-item scale measuring awareness of negative social and community norms about MSM; for example, “How often have you heard that gay or bisexual men are not normal?”; Cronbach’s alpha = 0.73; range 7–35), personal experiences of enacted sexual stigma (continuous, seven-item scale measuring acts of discrimination, violence, and mistreatment based on sexual orientation; for example, “How often have you been hit or beaten up for being gay or bisexual?”; Cronbach’s alpha = 0.88; range 7–49); personal experiences of perceived transgender stigma (continuous, five-item scale measuring awareness of negative social and community norms about transgender persons; for example, “How often have you heard that transgender people are not normal?”; Cronbach’s alpha = 0.77; range 7–28), and personal experiences of enacted transgender stigma (continuous, seven-item scale measuring acts of discrimination, violence, and mistreatment based on transgender identity, for example, “How often have you been hit or beaten up for being transgender?”; Cronbach’s alpha = 0.61; range 5–20). We also assessed whether participants had experienced any barriers to health care access (dichotomous, yes versus no) and had regular access to a health care provider (dichotomous, yes versus no), and we measured participants’ empowerment scores (continuous, measured using the Growth and Empowerment Measure).

We used quantitative analysis methods—specifically logistic regression—to estimate the unadjusted and adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for the odds of ever experiencing police harassment among (1) MSM and (2) transgender women. Variables that were statistically significant, indicated with a p-value of <0.05, or theoretically important in determining HIV vulnerability were considered for inclusion in the full multivariable model. A manual backward stepwise approach was used, whereby variables with lower strength of association were systematically removed from the model so that the final model included only those variables most significantly associated with ever experiencing police harassment. Tables 2 and 3 show two-sided p-values and unadjusted and adjusted odds ratios with 95% confidence intervals for those factors significantly associated with the outcome for MSM and transgender women, respectively. All statistical analyses were conducted using SAS software version 9.3 (SAS Institute, Cary, NC, USA) or SPSS version 24 (SPSS, Chicago, USA).

Table 2

Bivariable and multivariable analyses of factors associated with police harassment among men who have sex with men in Jamaica (n=556)

Characteristic	Unadjusted OR (95% CI)	Adjusted OR (95% CI)
Individual-level factors
Education, less than high school	2.73 (1.64, 4.53)**
Monthly income	0.89 (0.80, 0.99)^+*
HIV positive	1.96 (1.12, 3.44)*	1.85 (1.01, 3.38)*
Sex work in the last 12 months	4.05 (2.67, 6.15)***	2.47 (1.54, 3.96)**
Interpersonal-level factors
Relationship status
Concurrent partnerships (versus in a relationship)	5.68 (2.68, 12.04)***
Social support score^	1.06 (1.03, 1.09)^**
Consistent condom use	1.74 (1.06, 2.87)*
Safer sex self-efficacy score^	0.90 (0.83, 0.97)^**
Structural- and environmental-level factors
Food insecurity	3.47 (2.25, 5.35)***	2.44 (1.51, 3.94)**
Unstable housing	2.23 (1.46, 3.40)**
Currently unemployed	1.85 (1.21, 2.85)**
Perceived sexual stigma score^	1.33 (1.24, 1.44)^***
Enacted sexual stigma score^	1.54 (1.42, 1.66)^***
Empowerment score^	0.95 (0.93, 0.98)^**
Experienced 1 or more barriers to health care access	1.76 (1.17, 2.64)**
Does not have a regular health care provider	1.99 (1.30, 3.05)**	1.66 (1.02, 2.71)*

per 1-unit increase

per 100 USD increase

p<0.05

p<0.001

p<0.0001

Table 3

Bivariable and multivariable analyses of factors associated with police harassment among transgender women in Jamaica (n=137)

Characteristic	Unadjusted OR (95% CI)	Adjusted OR (95% CI)a
Individual-level factors
Depression	1.23 (1.01, 1.50)*
HIV positive	2.44 (1.01, 5.86)*	3.11 (1.06, 9.12)*
Sex work in the last 12 months	2.61 (1.30, 5.25)**
Interpersonal-level factors
Social support score^	1.09 (1.03, 1.15)**
Physical abuse	2.24 (1.12, 4.48)*
Structural- and environmental-level factors
Food insecurity	2.47 (1.20, 5.05)*
Unstable housing	2.30 (1.14, 4.64)*
Perceived transgender stigma score^	1.19 (1.06, 1.33)**
Enacted transgender stigma score^	1.46 (1.28, 1.66)***	1.32 (1.15, 1.52)***

per 1-unit increase

p<0.05

p<0.001

p<0.0001

Controlling for education and income

Study results

Participant characteristics (Table 1)

This sample of young MSM (n=556; median age 24, IQR: 22–28) and transgender women (n=137; median age 24, IQR: 15–44) was characterized by extreme economic insecurity, poor health, and high rates of police harassment. Specifically, almost half of MSM and over half of transgender women reported food insecurity, and one-third and one-half of MSM and transgender women reported unstable housing, respectively. While 13.5% of MSM were HIV positive, over one-quarter of transgender women were HIV positive. One-fifth (n=124, 22.3%) of MSM reported having experienced police harassment due to their sexual orientation, and 60 (43.8%) transgender women reported having experienced police harassment due to their gender identity. Among transgender women, 11.8% reported being incarcerated one to three times, and 4.4% reported being incarcerated four to six times, due to being transgender.

Factors associated with police harassment among men who have sex with men in Jamaica (Table 2)

In unadjusted bivariable analyses with MSM, having less than a high school education, being HIV positive, reporting any sex work in the past 12 months, being in a concurrent partnership versus in a relationship, having a higher need for social support, having consistent condom use, experiencing food insecurity, having unstable housing, being currently unemployed, experiencing perceived and enacted sexual stigma, experiencing one or more barriers to health care access, and not having a regular health care provider were all associated with increased odds of experiencing police harassment due to one’s sexual orientation. A higher monthly income, higher safer sex self-efficacy, and higher empowerment were associated with lower odds of experiencing police harassment. In the final multivariable model, the adjusted odds of experiencing police harassment were higher for those who were HIV positive (adjusted OR: 1.85, 95% CI: 1.01, 3.38), who reported undertaking sex work in the past 12 months (adjusted OR: 2.47, 95% CI: 1.54, 3.96), who were food insecure (adjusted OR: 2.44, 95% CI: 1.51, 3.94), and who did not have a regular health care provider (adjusted OR: 1.66, 95% CI: 1.02, 2.71).

Factors associated with police harassment among transgender women in Jamaica (Table 3)

In unadjusted bivariable analyses with transgender participants, the factors of depression, HIV-positive serostatus, any sex work in the last 12 months, a higher need for social support, ever experiencing physical abuse, food insecurity, unstable housing, and perceived or enacted transgender stigma were all associated with increased odds of experiencing police harassment due to one’s transgender identity. In the final multivariable model, the adjusted odds of experiencing police harassment were higher for those who were HIV positive (adjusted OR: 3.11, 95% CI: 1.06, 9.12) and reported higher levels of enacted transgender stigma (adjusted OR: 1.68, 95% CI: 1.26, 2.07, per one unit increase in enacted transgender stigma score).

Implications

Our study highlights widespread police harassment among MSM (22%) and transgender women (43%), an indicator of human rights violations. In contexts where consensual same-sex sexual relationships and practices are criminalized, it is likely that MSM and transgender women have little to no recourse to justice when police are perpetrating violence. In multivariable analyses, we found clear linkages between police harassment and HIV vulnerabilities: HIV-positive MSM and transgender women were more likely to report police harassment than HIV-negative peers. While this comparison has not been documented elsewhere, among MSM HIV-prevention outreach workers in India, 85% reported harassment by the police, suggesting the potential targeting of MSM due to their association with HIV or HIV-related work.

Among MSM participants, those who were engaged in sex work, were food insecure, and lacked a health care provider were more likely to report police harassment; and among transgender women, police harassment was associated with enacted transgender stigma. This evidence points to the need for an intersectional approach to understanding the impacts of police harassment among MSM and transgender women who experience marginalization on the basis of multiple, intersecting identities and experiences: sex work, poverty, and transgender stigma. Studies on sex workers globally have highlighted the negative impacts of sex work criminalization on their human rights, well-being, and access to HIV prevention tools. According to Kate Shannon et al., the decriminalization of sex work would avert 33% to 46% of HIV infections in the next decade and would increase access to health care and respect for human rights. Similarly, it is suggested that behavioral interventions to mitigate HIV vulnerability for transgender women sex workers be coupled with structural changes (for example, economic and community empowerment, the provision of culturally competent health services, and a protective legal and social environment that upholds their human rights).

Our study’s limitations include a cross-sectional design that precludes understanding of causality, self-reporting measures that are subject to recall and social desirability bias, and the use of only one measure of police harassment. Our study would have been further strengthened by asking MSM about their incarceration history. Despite these limitations, our analyses provide quantitative evidence for HIV vulnerabilities associated with police harassment in Jamaica among key populations: MSM and transgender women. The negative effects of criminalization and subsequent police violence compromise efforts to reduce HIV transmission among key populations and reduce the likelihood of reaching goals of engaging people living with HIV in Jamaica in the HIV care cascade. Future studies could use a longitudinal design to better understand the directionality of the relationships between police violence and HIV infection, to identify potential mediators, and to answer key questions. For instance, does police harassment contribute to reduced access to health care and HIV prevention services, and in turn increase vulnerability to HIV acquisition among MSM and transgender women in Jamaica? Are HIV infection and police violence both associated with a third variable (such as community-level stigma or poverty)? Future research could further explore the complexity of the relationships between police harassment and HIV vulnerabilities among these key populations in Jamaica and elsewhere.

Police harassment among HIV-positive MSM and transgender women in Jamaica has clear implications for the protection of human rights in order to ensure access to the HIV care cascade. There has been a call to action to increase research on effective strategies for collaboratively engaging the police in addressing discrimination, stigma, and HIV risk. These programs may involve components such as trainings that integrate information on the importance of police engagement in HIV prevention efforts and police collaboration with affected communities on human rights and harm reduction; peer advocacy and education; and, strategies that bring the police together with communities in non-conflict settings. However, as Andrew Scheibe et al. suggest in their study describing attempts to implement interventions to improve the relationship between the police and key populations in South Africa, without buy-in from the police or society more broadly, such interventions remain small in scale or unimplemented altogether. Future interventions in Jamaica may include joint discussions between the police and communities and the development of shared language that seeks to shift negative interactions between the police and key populations. Interventions that address stigmatizing social attitudes, legal protections to increase access to health and social services, and strategies to strengthen relationships between the police and MSM and transgender women may help reduce HIV vulnerabilities and promote human rights for MSM and transgender women in Jamaica.

Table 1

Participant characteristics

Characteristic	MSM		Transgender women
	n=556	Missing	n=137	Missing
Age, years (median, IQR)	24 (22–28)	14	24 (15–44)	7
Has at least a high school diploma (n, %)	478 (86.0)		109 (80.7)	2
Monthly income in USD (median, IQR)	144 (10–280)	20	123.45 (0–2469)	23
Is HIV positive (n, %)	67 (13.5)	58	26 (25.24)	34
Relationship status (n, %)		2		1
Is in a relationship or casually dating	383 (69.1)		79 (58.1)
Does not have a partner	133 (24.9)		31 (22.8)
Is in concurrent partnerships	33 (6.0)		26 (19.1)
Has experienced physical violence (n, %)	338 (61.3)	5	62 (45.93)	1
Has undertaken sex work in past 12 months (n, %)	182 (32.7)		71 (51.82)
Is food insecure (n, %)	266 (47.9)	1	82 (59.9)
Has unstable housing conditions (n, %)	175 (32.8)	23	71 (51.8)	3
Does not have access to a regular health care provider (n, %)	235 (42.3)		95 (69.34)
Has experienced police harassment due to sexual orientation or gender identity (n, %)	124 (22.3)		60 (43.8)
Has been incarcerated seemingly as a result of transgender identity (n, %)				10
1 to 3 times	-	-	15 (11.8)
4 to 6 times	-	-	6 (4.4)