Abstract
Objectives. To estimate the prevalence of personal and household victimizations among transgender people in the United States.
Methods. We analyzed pooled 2017 and 2018 data from the National Crime Victimization Survey, the first nationally representative sample that allows identification of transgender respondents.
Results. Transgender people experienced 86.2 victimizations per 1000 persons compared with cisgender people’s 21.7 per 1000 persons (odds ratio [OR] = 4.24; 90% confidence interval [CI] = 1.49, 7.00). Households that had a transgender person had higher rates of property victimization (214.1 per 1000 households) than households with only cisgender people (108 per 1000 households; OR = 2.25; 90% CI = 1.19, 3.31). Transgender victims whose sex assigned at birth was male were more likely to perceive their victimization as a hate crime than cisgender victims whose sex assigned at birth was male. There were no disparities in reporting victimizations to authorities: only about half of the victimizations of both transgender and cisgender people were reported.
Conclusions. Public policy and administration need to consider the unique vulnerabilities transgender people routinely encounter, resulting in disparities in criminal victimization.
Anecdotal data and small-scale studies suggest that transgender populations are at a heightened risk of criminal victimization,1 which is defined as any action by others that violate laws affecting oneself or one’s property. However, outside of hate crime statistics, national data addressing this issue have been limited.2 Beginning in 2016, the National Crime Victimization Survey (NCVS)—the nation’s primary source of nonfatal criminal victimization statistics—began documenting the sexual orientation and gender identity of respondents.3 NCVS 2017 data showed significant disparities in victimization rates between lesbian, gay, bisexual, and transgender (LGBT) people and cisgender heterosexual people.4 However, small sample sizes prohibited analyses of LGBT subgroups (e.g., bisexual women or transgender people). By pooling 2 years of data, we report what are to our knowledge the first prevalence estimates of victimization among transgender adults in the United States from a nationally representative sample.
METHODS
The NCVS is administered to a nationally representative, longitudinal sample of individuals aged 12 years or older within households in the United States. The survey collects incident-level data about experiences with victimization both reported and not reported to police. The current analysis used pooled 2017 and 2018 NCVS data for a total sample of 296 563 households and 482 469 individuals.5,6 More information about the NCVS is available through Bureau of Justice Statistics (BJS) publications.7
Measures
In July 2016, the BJS began identifying transgender people among NCVS respondents aged 16 years or older.3 Gender identity was measured with 2 questions: sex assigned at birth (male, female, and don’t know) and current gender identity (male, female, transgender, or none of these). Respondents are categorized as transgender if they identified as transgender or their current gender identity was male or female and was different from their assigned sex at birth. Respondents are categorized as cisgender if their current gender identity matched their assigned sex at birth. In the 2017–2018 NCVS, about 0.10% (n = 420) people were thus classified as transgender and 99.9% (n = 435 061) were cisgender. This prevalence estimate is consistent with other government-sponsored surveys.3 This includes respondents who were categorized as transgender men if their sex assigned at birth was female and they had a current gender identity that was male or transgender (n = 181) and categorized as transgender women if their sex assigned at birth was male and they had a current gender identity that was female or transgender (n = 188). In addition, some respondents (n = 51) indicated they were transgender but refused to answer the “sex assigned at birth” question and were not categorized as transgender men or women; these respondents are included in overall analyses. We recognize that the terms “transgender,” “transgender men,” and “transgender women” may not be how respondents identify themselves, and we use these categorizations solely for analytic purposes and to clearly communicate findings. NCVS data do not allow for assessment of gender nonbinary identities.
The NCVS documents numerous types of crime, which are broadly categorized as either personal or property victimizations. Victims were asked if the victimization was reported to the police, either by the victim or by others (e.g., witnesses or other victims). For each incident, victims indicated whether they thought the incident was motivated by prejudice or bigotry against their characteristics or religious beliefs. Respondents reported their age, race or ethnicity, educational attainment, marital status, household income, and urbanicity of residence.
Analysis
We conducted analyses for transgender and cisgender people separately and, within these groups, by current gender. After summarizing demographic characteristics, we estimated rates of personal victimizations per 1000 persons, rates of property victimizations per 1000 households, the percentage of victimizations reported to police, and the percentage of victimizations perceived as hate crimes.8 We documented property victimizations at the household level. We defined a household as a transgender household if at least 1 member of the household was transgender. We applied the same categorization by current gender, and these households were not mutually exclusive. We defined cisgender households as households in which there were no transgender people. For point and standard error estimations, we used NCVS complex design variables and weights, which account for the address-based cluster sampling, longitudinal design with repeated interviews, and multiple interviews per household. Prior to computing estimates for this article, we used the analytic approach and weights to produce estimates that replicated those generated by the BJS. We estimated standard errors using Taylor series linearization.9 We report unadjusted odds ratios with 90% confidence intervals as measures of association, and we report differences (Δ) between estimated rates and percentages and associated 2-tailed P values from use of the t test.
RESULTS
Transgender people were approximately evenly distributed by their sex assigned at birth, but those who self-identified as transgender were more likely than cisgender people to refuse to answer the “sex assigned at birth” question. Compared with cisgender people, transgender people had similar racial and ethnic and educational distributions but were younger and more likely to have never been married. Compared with cisgender people, transgender people were more likely to reside in urban locations and in households earning less income (Table A, available as a supplement to the online version of this article at http://www.ajph.org).
Transgender people experienced violence at a rate of 86.2 victimizations per 1000 persons compared with 21.7 per 1000 persons among cisgender people (Figure 1a; odds ratio [OR] = 4.24; 90% confidence interval [CI] = 1.49, 7.00). These differences remained for men and women. Transgender women and men had higher rates of violent victimization (86.1 and 107.5 per 1000 persons, respectively) than did cisgender women (23.7 per 1000 persons; OR = 3.88; 90% CI = 0, 8.55) and cisgender men (19.8 per 1000 persons; OR = 5.98, 90% CI = 2.09, 9.87), but there were no differences between transgender men and women (Δ = 21.4; SE = 68.7; P = .76).
FIGURE 1—
Unadjusted (a) Personal and (b) Household Victimization Rates Among Transgender and Cisgender People in the United States: National Crime Victimization Survey, 2017–2018
Note. cis = cisgender; trans = transgender. A transgender household is a residence with at least 1 person who is transgender, and a cisgender household is a residence with no transgender person. Thick black lines with caps represent 90% confidence intervals; thin black lines represent 95% confidence intervals.
aBecause property victimizations occur at a household level, dividing by gender for cisgender households produces two estimates that are nearly the same. Therefore, only total cisgender household values are shown.
Transgender households had higher rates of property victimization (214.1 per 1000 households) than cisgender households (108 per 1000 households; OR = 2.25; 90% CI = 1.19, 3.31; Figure 1b). These differences were consistent across genders.
Overall, there was a large, but not statistically significant, difference in the percentage of violent victimizations against transgender and cisgender people that were perceived to be hate crimes (19% vs 9%; Δ = 9.8; SE = 6.2; P = .12). Between transgender and cisgender women, there was a large and statistically significant difference in the percentage of violent victimizations believed to be hate motivated (28% vs 9%; Δ = 18.4; SE = 7.7; P = .02).
Approximately half of all violent victimizations were reported to police, with no differences between transgender and cisgender persons (51% vs 47%; Δ = 4.6; SE = 13.3; P = .73) or between transgender and cisgender women (49% vs 44%; Δ = 4.7; SE = 17.2; P = .79) and men (53% vs 50%; Δ = 3.1; SE = 19.0; P = .87).
More transgender than cisgender people believed property victimizations to be hate crimes, but these were imprecise estimates with large standard errors (4% vs 1%; Δ = 2.9; SE = 3.5; P = .40).
Approximately one third of property victimizations were reported to the police, but reporting by transgender and cisgender people was similar (35% vs 27%; Δ = 8.1; SE = 8.8; P = .35). This pattern did not differ by gender (transgender vs cisgender women: 39% vs 35%; Δ = 3.6; SE = 15.5; P = .82; transgender vs cisgender men: 21% vs 36%; Δ = −14.8; SE = 11.7; P = .21).
DISCUSSION
To our knowledge, this is the first study using a nationally representative sample to examine the victimization of transgender adults in the United States. Our findings evidence the disproportionate rate of transgender people’s victimization. Rates of victimization did not differ between transgender women and men. Reporting to police was low and similar to the cisgender rate and to findings from the 2015 United States Transgender Survey.10
Although some attention has been given to homicides of transgender women of color in the media, little attention has been given to the crimes reported here and the fact that victimization levels are similar among transgender women and men. We found that 1 in 4 victimizations against transgender women were perceived to be hate crimes.
Our study is limited by relatively small sample sizes of transgender people, which accounts for large confidence intervals and limits our ability to assess victimization subtypes. We also could not investigate victimization at the intersection of gender identity, race and ethnicity, age, marital status, urbanicity, and other characteristics. Some of these characteristics may confound our findings, but others, such as household income, may be products of being transgender (e.g., employment discrimination) along a causal chain leading to criminal victimization. Future research, using multiple years of NCVS data, could unpack the type of hate crime and its severity, and consider potential confounders and mediators of victimization. There are also general limitations in the NCVS, such as the reliance on self-report.11
PUBLIC HEALTH IMPLICATIONS
The documentation of violence from population-based data should spur policymakers to enact “more effective and necessary policies at the local, state, and federal levels to protect people based on their gender identity and gender expression.”1(p170) This is particularly important because victimization is related to other measures of well-being—such as suicide rates—of gender-diverse populations.12
ACKNOWLEDGMENTS
We thank Jennifer Truman at the Bureau of Justice Statistics for assistance with variable recodes to produce estimates consistent with official reports.
CONFLICTS OF INTEREST
The authors have no conflicts of interests to declare.
HUMAN PARTICIPANT PROTECTION
We used publicly available unidentified data, which is exempt from human participation protection review.
REFERENCES
- 1.Stotzer RL. Violence against transgender people: a review of United States data. Aggress Violent Behav. 2009;14(3):170–179. doi: 10.1016/j.avb.2009.01.006. [DOI] [Google Scholar]
- 2.Woods JB. LGBT identity and crime. Calif Law Rev. 2017;105:667–734. [Google Scholar]
- 3.Truman JL, Morgan RE, Gilbert T, Vaghela P. Measuring sexual orientation and gender identity in the National Crime Victimization Survey. J Off Stat. 2019;35(4):835–858. doi: 10.2478/jos-2019-0035. [DOI] [Google Scholar]
- 4.Flores AR, Langton L, Meyer IH, Romero AP. Victimization rates and traits of sexual and gender minorities in the United States: results from the National Crime Victimization Survey, 2017. Sci Adv. 2020;6(40):eaba6910. doi: 10.1126/sciadv.aba6910. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.US Bureau of Justice Statistics. National Crime Victimization Survey, [United States], 2017. National Archive of Criminal Justice Data, distributor. March 29, 2020. Available at: https://www.icpsr.umich.edu/icpsrweb/NACJD/studies/36981. Accessed July 1, 2020.
- 6.US Bureau of Justice Statistics. National Crime Victimization Survey, Concatenated File, [United States], 1992–2018. Inter-University Consortium for Political and Social Research, distributor. March 5, 2020. [DOI]
- 7.US Bureau of Justice Statistics. National Crime Victimization Survey, [United States], 2018. National Archive of Criminal Justice Data, distributor. March 29, 2020. Available at: https://www.icpsr.umich.edu/icpsrweb/NACJD/studies/37297. Accessed July 1, 2020.
- 8.Morgan RE, Ouderkerk BA. Criminal Victimization, 2018. Washington, DC: Office of Justice Programs, Bureau of Justice Statistics; 2019. [Google Scholar]
- 9.Shook-Sa B, Couzens GL, Berzofsky M. RTI International; 2011. User’s guide to the National Crime Victimization Survey (NCVS) direct variance estimation. Available at: https://www.bjs.gov/content/pub/pdf/NCVS_Variance_User_Guide%2011.06.14.pdf. Accessed July 1, 2020. [Google Scholar]
- 10.James SE, Herman JL, Rankin S, Keisling M, Mottet L, Anafi M. The Report of the 2015 US Transgender Survey. Washington, DC: National Center for Transgender Equality; 2016. [Google Scholar]
- 11.National Research Council. Surveying Victims: Options for Conducting the National Crime Victimization Survey. Washington, DC: National Academies Press; 2018. [Google Scholar]
- 12.Goldblum P, Testa RJ, Hendricks ML, Bradford J, Bongar B. The relationship between gender-based victimization and suicide attempts in transgender people. Prof Psychol Res Pr. 2012;43(5):468–475. doi: 10.1037/a0029605. [DOI] [Google Scholar]