Abstract
Introduction:
Young adults experiencing homelessness are vulnerable to firearm violence. This study aims to explore the correlates of firearm violence involvement among this vulnerable population, which may inform firearm violence reduction intervention development.
Methods:
Between 2016 and 2017, young adults experiencing homelessness aged 18–26 years (N=1,426) were recruited in 7 U.S. cities. Respondents completed a self-administered computer-assisted anonymous survey regarding their homeless experiences and violence involvement. Separate multivariate logistic regression models were conducted in 2020 to explore the correlates of direct firearm violence victimization, witness of firearm violence, and firearm violence perpetration.
Results:
A high proportion of young adults experiencing homelessness were involved in firearm violence (witnessing firearm violence: 40%; direct firearm violence victimization: 28%; perpetration: 18%). Stressful experiences, such as childhood trauma and street victimization, were associated with greater odds of firearm violence involvement. Black (OR=2.4, p<0.001) and Latinx (OR=2.0, p<0.05) young adults had greater odds of experiencing direct firearm violence victimization than White young adults. Black (OR=2.0, p<0.01) and Latinx (OR=2.4, p<0.001) young adults were also at greater risk of witnessing firearm violence. Young adults with mental illness had greater odds of being directly victimized by firearm violence (OR=1.7, p<0.01).
Conclusions:
Given the inter-related nature of firearm violence involvement and given that risk factors for violence are often embedded in social and structural contexts, multipronged community-based approaches to prevent firearm violence among young adults experiencing homelessness are necessary. Targeted efforts may be indicated to attenuate the risk and promote resilience among subgroups of young adults experiencing homelessness who are disproportionately affected by firearm violence.
INTRODUCTION
There are more than 3.5 million young adults experiencing homelessness (YAEH) in the U.S.1 General strain theory2 highlights the critical role of stressful experiences, such as trauma, in elevating individuals’ risk of violence, including violence perpetration and victimization. Stressful experiences (e.g., adverse childhood experiences and street victimization) are not uncommon among YAEH.3 All types of firearm violence involvement (i.e., direct firearm violence victimization, witnessing firearm violence, and firearm violence perpetration), are prevalent among YAEH.4 Although firearm violence involvement may lead to damaging health consequences,4 research examining firearm violence correlates seldom focuses on YAEH. Therefore, this study aims to explore correlates of firearm violence involvement using data from YAEH across 7 U.S. cities. The identification of correlates will inform future YAEH firearm violence reduction intervention development.
METHODS
Study Sample
A purposive sample of 1,426 YAEH were recruited through a multi-institutional research collaboration from homeless service agencies in 7 U.S. cities (Los Angeles, Denver, New York, Houston, Phoenix, St. Louis, and San Jose) between 2016 and 2017. Individuals had to be: (1) aged 18–26 years and (2) currently experiencing homelessness/unstable housing. After consenting to participate in the study, respondents completed a self-administered computer-assisted anonymous survey delivered via tablets to assess their violence engagement. Respondents received $10–$20 gift cards to local vendors. IRB approvals were received from each collaborating researcher’s institution. Detailed methodology can be found elsewhere.5
Measures
Direct firearm violence victimization (i.e., been shot at by someone with a firearm), witnessing firearm violence (i.e., seen someone being injured or killed by a firearm), and firearm violence perpetration (i.e., shot a firearm at another individual) were all dichotomous variables and reflected lifetime involvement in the corresponding type of violence.
Informed by general strain theory2 and previous literature,6 independent variables included in this study were homelessness duration, foster care or justice system engagement, adverse childhood experiences, street victimization experiences, mental illness diagnosis, gang affiliation, and firearm access. Demographics (i.e., study cities, age, race and ethnicity, gender, sexual orientation, and education attainment) served as controls. Table 1 covers variable coding details.
Table 1.
Descriptive Statistics of Young Adults Experiencing Homelessness (YAEH; N=1,426)a
| Variable | Descriptive statistics | |
|---|---|---|
| N (%) | Mean (SD) | |
| Violence involvement variables | ||
| Firearm violence perpetration | 247 (17) | |
| Direct firearm violence victimization | 388 (27) | |
| Witness of firearm violence | 560 (39) | |
| Demographic control variables | ||
| Study cities | ||
| Los Angeles | 215 (15) | |
| Denver | 208 (15) | |
| Houston | 202 (14) | |
| Phoenix | 198 (14) | |
| New York | 208 (15) | |
| San Jose | 197 (14) | |
| St. Louis | 198 (14) | |
| Age, years | 20.9 (2.1) | |
| Gender identity | ||
| Cisgender male | 833 (59) | |
| Cisgender female | 482 (34) | |
| Gender minority | 101 (7) | |
| Sexual orientation | ||
| Heterosexual | 1,018 (72) | |
| Sexual minorities | 404 (28) | |
| Race/Ethnicity | ||
| White | 270 (19) | |
| Black | 531 (37) | |
| Latinx | 247 (17) | |
| Multiracial or Others | 376 (26) | |
| Education | ||
| High school or GED | 986 (69) | |
| Independent variables | ||
| Duration of homelessness | ||
| Homelessness at least 2 years | 436 (31) | |
| System engagement | ||
| Foster care system engagement | 553 (39) | |
| Correctional system engagement | 846 (60) | |
| Adverse childhood experiences | 4.6 (3.0) | |
| Street victimization | 1,072 (75) | |
| Mental illness diagnosis | 848 (60) | |
| Gang affiliation | 274 (19) | |
| Firearm access | 470 (33) | |
Independent and sociodemographic control variables are coded as follows. Independent variables focused on duration of homelessness, system engagement, trauma exposure, mental health status, gang affiliation and firearm access. Duration of homelessness is a dichotomous variable, with 1=lifetime homeless duration of at least 2 years. System engagement variables included lifetime involvement in the justice system (dichotomous; 1=history of arrest or incarceration) and foster care system (dichotomous; 1=history in foster care system). Adverse childhood experiences (continuous; the higher the score the more adverse childhood experiences growing up; Felitti et al., 1998). Street victimization (dichotomous; 1=had been robbed, threatened, assaulted, or witnessed victimization during homeless). Mental illness diagnosis is a dichotomous variable with 1=self-reported mental illness diagnosis. Lifetime gang affiliation is a dichotomous variable with 1=history of gang affiliation. Firearm access is a dichotomous variable, with 1=self-reported having access to a firearm is wanted. Demographic controls covered study cities (nominal), age (continuous), race and ethnicity (nominal; 1=White, 2=Black, 3=Latinx, and 4=Multiracial or Other), gender identity (nominal; 1=cisgender male, 2=cisgender female, and 3=gender minority, including people who identify as transgender or gender nonconforming), sexual orientation (dichotomous; 1=heterosexual), and education (dichotomous; 1=at least had high school or GED degree). Only independent variables found to be significant in univariate analysis (p<0.05) were included in the final multivariate models. Sociodemographic controls were included in all multivariate models.
Statistical Analysis
Logistic regression models were conducted in 2020 to explore correlates of 3 different types of firearm violence involvement separately. Bivariate analyses were conducted first. Multivariate logistic regression models were then conducted, including independent variables significant in bivariate analyses (p<0.05) and all demographic controls. For better interpretability, each multivariate model excluded the other 2 types of firearm involvement variables (though significant associations were observed between all firearm involvement variables in bivariate analyses).
RESULTS
Table 1 depicts respondent characteristics, and Table 2 illustrates the multivariate analysis results. Forty percent of YAEH had witnessed firearm violence and 28% had been directly victimized by firearm violence. Approximately 18% disclosed perpetrating firearm violence. Compared with White YAEH, racial minority YAEH were more likely to witness and experience direct firearm violence. Cisgender female YAEH were less likely to perpetrate firearm violence and were at lower risk for experiencing direct firearm violence victimization. Longer duration of homelessness was associated with higher risk of witnessing firearm violence. Street victimization was associated with elevated risk of firearm violence perpetration and witnessing firearm violence, while adverse childhood experiences were positively associated with all 3 types of firearm violence involvement. Having been diagnosed with a mental illness was associated with an elevated risk of direct firearm violence victimization. Gang affiliation and having easy firearm access (i.e., know how to access a firearm if needed) were both associated with elevated risk of firearm violence perpetration, direct victimization, and witnessing firearm violence. The authors also conducted an auxiliary analysis with firearm violence involvement variables entered in the multivariable models. The results remained comparable, with all violence involvement variables significantly associated with one another.
Table 2.
Multivariate Logistic Regression Results of Firearm Violence Involvement Correlates Among Young Adults Experiencing Homelessness (YAEH; N=1,426)a,b
| Variable | Firearm violence perpetration | Direct firearm violence victimization | Witness of firearm violence |
|---|---|---|---|
| OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| Demographic control variables | |||
| Study cities (ref: Los Angeles) | |||
| Denver | 0.9 (0.5, 1.8) | 1.1 (0.7, 1.9) | 1.0 (0.6, 1.6) |
| Houston | 1.3 (0.7, 2.5) | 1.1 (0.6, 1.8) | 0.9 (0.6, 1.5) |
| Phoenix | 1.1 (0.5, 2.2) | 0.8 (0.4, 1.4) | 0.8 (0.5, 1.4) |
| New York | 1.5 (0.8, 2.8) | 1.2 (0.7, 2.0) | 0.8 (0.5, 1.3) |
| San Jose | 0.7 (0.3, 1.5) | 0.5* (0.3, 0.9) | 0.4** (0.3, 0.7) |
| St. Louis | 1.2 (0.6, 2.4) | 1.2 (0.7, 2.1) | 0.9 (0.5, 1.5) |
| Age | 1.1 (0.9, 1.2) | 1.0 (0.9, 1.1) | 1.0 (0.9, 1.1) |
| Gender identity (ref: cisgender male) | |||
| Cisgender female | 0.3*** (0.2, 0.5) | 0.3*** (0.2, 0.4) | 0.8 (0.6, 1.1) |
| Gender minority | 0.9 (0.4, 2.0) | 0.5 (0.3, 1.0) | 0.6 (0.3, 1.1) |
| Sexual orientation | |||
| Heterosexual | 1.1 (0.7, 1.7) | 1.4 (0.9, 2.0) | 1.4 (0.9, 1.9) |
| Race/Ethnicity (ref: White) | |||
| Black | 1.5 (0.9, 2.6) | 2.4*** (1.5, 3.8) | 2.0** (1.3, 2.9) |
| Latinx | 1.0 (0.5, 1.9) | 2.0* (1.2, 3.5) | 2.4*** (1.5, 3.7) |
| Multiracial or Others | 1.2 (0.7, 2.1) | 2.5*** (1.6, 4.0) | 1.7** (1.2, 2.5) |
| Education | |||
| High school or GED | 0.9 (0.6, 1.3) | 0.7 (0.5, 1.1) | 1.0 (0.7, 1.3) |
| Independent variables | |||
| Duration of homelessness | |||
| Homelessness at least 2 years | 0.8 (0.5, 1.2) | 1.1 (0.8, 1.5) | 1.6** (1.2, 2.2) |
| System engagement | |||
| Foster care system engagement | 1.4 (0.9, 2.0) | 1.1 (0.8, 1.5) | - |
| Correctional system engagement | 1.3 (0.9, 2.1) | 1.2 (0.9, 1.7) | 1.1 (0.8, 1.5) |
| Adverse childhood experiences | 1.1* (1.0, 1.2) | 1.2** (1.1, 1.2) | 1.1*** (1.1, 1.2) |
| Street victimization | 2.1** (1.2, 3.8) | 1.2 (0.8, 1.7) | 1.6** (1.2, 2.3) |
| Mental illness diagnosis | 1.5 (0.9, 2.2) | 1.7** (1.2, 2.4) | 1.3 (0.9, 1.7) |
| Gang affiliation | 6.3*** (4.3, 9.1) | 3.5*** (2.5, 4.9) | 2.7*** (1.9, 3.8) |
| Firearm access | 4.3*** (2.9, 6.2) | 2.8*** (2.1, 3.8) | 3.3*** (2.5, 4.3) |
Only independent variables found to be significant in univariate analysis (p<0.05) were included in the final multivariate models. Demographic controls were included in all multivariate models.
Boldface indicates statistical significance (*p<0.05; **p<0.01; ***p<0.001).
DISCUSSION
The current study is the first known research to investigate correlates of various types of firearm violence among YAEH. A key observation was that different violence experiences were associated with one another. Specifically, negative life experiences, including adverse childhood events and street victimization, were positively correlated with witnessing firearm violence and firearm perpetration. Moreover, bivariate and auxiliary analyses found strong links among direct firearm victimization, witnessing violence, and perpetration. This falls in line with previous theoretical and empirical work related to criminal offending and victimization. In particular, general strain theory argues that experienced, vicarious, and anticipated strains contribute to both victimization and perpetration.2 Taking into account that YAEH face innumerable strains (e.g., traumatic experiences) that often precipitate and persist during homelessness,3 high rates of firearm violence are not unexpected. Yet, it is no less discouraging that YAEH are frequently stuck in webs of firearm violence that may be especially traumatic and lethal.
Another important observation was that YAEH of color were at higher risk for direct firearm victimization and witnessing firearm violence, but not perpetration, than their White counterparts. This pattern of findings runs counter to racially or ethnically biased misconceptions, but aligns with studies of adolescent violence7 and research showing that people of color are more likely to know someone who has been victimized by firearm violence than their White counterparts.8 Similarly, the present study found that people with a history of mental health diagnoses experienced higher rates of firearm victimization. Again, this is contrary to mistaken beliefs about mental illness as a major cause for violence, but consistent with meta-analytic work that highlights elevated risk for victimization among people with psychiatric disorders.9 Together, these results demonstrate that YAEH who often face added marginalization for other aspects of their identities are particularly vulnerable to firearm violence. Finally, YAEH’s gang affiliation and access to firearms may provide them with a sense of support and perceived safety,10 but may also increase their risk of both firearm violence perpetration and victimization.
Limitations
Despite being the first known research focusing on understanding firearm violence involvement among YAEH from a wide geographical area, the study’s cross-sectional design prevents the authors from ascertaining causal relationships. For example, it is possible that firearm violence victimization experiences may lead to YAEH’s adverse mental health outcomes. It is also possible that YAEH’s firearm violence exposure may proceed their homelessness experiences. Furthermore, not all forms of firearm violence were included in this study owing to constraints on the collected data (e.g., hearing about firearm violence). Future research should focus on further understanding indirect firearm violence among YAEH. Finally, this study did not examine social determinants (e.g., community violence) of YAEH’s firearm violence involvement. This is a critical consideration given that YAEH are situated in complicated macrostructural contexts (e.g., poverty, structural discrimination) that make them vulnerable to violence. Therefore, research adopting a socioecological framework is needed to gain a comprehensive understanding of firearm violence among YAEH.
CONCLUSIONS
Collectively, these findings have important public health implications for reducing firearm violence among YAEH. Given the interrelated nature of firearm violence involvement and previous work demonstrating that risk factors for violence are often embedded in social and structural contexts,10 multipronged approaches to reduce firearm violence are necessary. For instance, Safe Streets and Save Our Streets are violence prevention initiatives that involve media campaigns, street outreach, and coalition building.11 Targeted efforts may also be indicated to attenuate risk and promote resiliency among subgroups of YAEH who are disproportionately burdened by firearm violence. For example, adapting community-based, culturally tailored, evidence-based violence reduction interventions, such as the Cure Violence program,12 may be promising to reduce firearm violence faced by YAEH. Additionally, although research focused on trauma-informed interventions for YAEH remains underdeveloped, aggressively preventing and treating trauma in homelessness contexts may help to disrupt cycles of firearm violence. Indeed, violence perpetration is commonly motivated by self-protection rooted in victimization experiences among YAEH.13
ACKNOWLEDGMENTS
We greatly appreciate young adults for their time and efforts participating in this study. We also thank all service providers collaborating with us in this project. This research received support from the Greater Houston Community Foundation Funders Together to End Homelessness (Diane Santa Maria and Sarah Narendorf), F31MH108446 (Robin Petering), and Arizona State University Institute for Social Science Research (Kristin Ferguson).
The research presented in this paper is that of the authors and does not reflect the official policy of NIH, Foundation Funders Together to End Homelessness, and Arizona State University Institute for Social Science Research.
The IRB approvals were received from each study site. IRB Numbers are listed below: St. Louis site IRB#: 2006814; Los Angeles and San Jose IRB#: UP-16-00345; Denver IRB#: 913077-2; New York IRB#: 2016-1220; Houston IRB#: HSC-SN-16-0336 and 16378-01; Phoenix IRB#: STUDY00003633.
H-TH was the Principal Investigator at the St. Louis site. He was responsible for participant recruitment and data collection in the St. Louis area. H-TH formulated the major concept of the current study with assistance from AF. H-TH also drafted, finalized, and submitted the manuscript. AF helped H-TH to conceptualize the manuscript and interpret the findings. AF also helped drafting the discussion section. RP was the Principal Investigator at the Los Angeles and St. Jose sites, who was responsible in participant recruitment and data collection in the Los Angeles area and in San Jose. RP provided critical insights on community-based violence reduction intervention literature. AB-A and K Bender were the Principal Investigators at the Denver site. They both were responsible in participant recruitment and data collection in Denver. AB-A provided insights in risk engagement among young adults experiencing homelessness, while K Bender provided insights on trauma exposure and violence engagement. K Bedell was a doctoral research assistant who conducted literature review for this study and drafted the introduction section of the paper. KMF was the Principal Investigator at the Phoenix site, who was responsible in participant recruitment and data collection in the Phoenix area. KMF provided insights regarding violence victimization under the context of homelessness and its prolonged developmental consequences. SCN and DSM were both Principal Investigators at the Houston site, who were responsible in participant recruitment and data collection in the Houston area. SCN provided insights regarding mental health and violence involvement, while DSM provided feedback regarding violence exposure and risk engagement consequences. JS was the Principal Investigator at the New York site, who was responsible for participant recruitment and data collection in the New York area. JS provided critical insights on violence exposure disparities based on gender identification and sexual orientation. Finally, ER assisted with data analysis and finding interpretation for this study. All authors listed in this manuscript have reviewed the manuscript thoroughly and provided constructive feedback and edits. The corresponding author, H-TH, finalized the manuscript, with approval from all authors.
No financial disclosures were reported by the authors of this paper.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- 1.Morton MH, Dworsky A, Matjasko JL, et al. Prevalence and correlates of youth homelessness in the United States. J Adolesc Health. 2018;62(1):14‒21. 10.1016/j.jadohealth.2017.10.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Zavala E, Spohn RE. The role of vicarious and anticipated strain on the overlap of violent perpetration and victimization: a test of general strain theory. Am J Crim Justice. 2013;38(1):119‒140. 10.1007/s12103-012-9163-5. [DOI] [Google Scholar]
- 3.Bender KA, Thompson SJ, Ferguson KM, Yoder JR, Kern L. Trauma among street-involved youth. J Emot Behav Disord. 2014;22(1):53‒64. 10.1177/1063426613476093. [DOI] [Google Scholar]
- 4.Hsu HT, Fulginiti A, Petering R, et al. Firearm violence exposure and suicidal ideation among young adults experiencing homelessness. J Adolesc Health. 2020;67(2):286‒289. 10.1016/j.jadohealth.2020.02.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Santa Maria DM, Flash C, Narendorf S, et al. Knowledge and attitudes about PrEP and nPEP among a 7-city sample of homeless young adults. J Adolesc Health. 2018;62(2 suppl):S17. 10.1016/j.jadohealth.2017.11.035. [DOI] [Google Scholar]
- 6.Sanchez C, Jaguan D, Shaikh S, McKenney M, Elkbuli A. A systematic review of the causes and prevention strategies in reducing gun violence in the United States. Am J Emerg Med. 2020;38(10):2169‒2178. 10.1016/j.ajem.2020.06.062. [DOI] [PubMed] [Google Scholar]
- 7.Rojas-Gaona CE, Hong JS, Peguero AA. The significance of race/ethnicity in adolescent violence: a decade of review, 2005–2015. J Crim Justice. 2016;46:137‒147. 10.1016/j.jcrimjus.2016.05.001. [DOI] [Google Scholar]
- 8.Kalesan B, Weinberg J, Galea S. Gun violence in Americans’ social network during their lifetime. J Prev Med. 2016;93:53‒56. 10.1016/j.ypmed.2016.09.025. [DOI] [PubMed] [Google Scholar]
- 9.de Vries B, van Busschbach JT, van der Stouwe EC, et al. Prevalence rate and risk factors of victimization in adult patients with a psychotic disorder: a systematic review and meta-analysis. Schizophr Bull. 2019;45(1):114‒126. 10.1093/schbul/sby020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Turanovic JJ, Pratt TC, Piquero AR. Structural constraints, risky lifestyles, and repeat victimization. J Quant Criminol. 2018;34(1):251‒274. 10.1007/s10940-016-9334-5. [DOI] [Google Scholar]
- 11.Petrosino A, Campie P, Pace J, et al. Cross-sector, multi-agency interventions to address urban youth firearms violence: a rapid evidence assessment. Aggress Violent Behav. 2015;22:87‒96. 10.1016/j.avb.2015.04.001. [DOI] [Google Scholar]
- 12.Butts JA, Roman CG, Bostwick L, Porter JR. Cure violence: a public health model to reduce gun violence. Annu Rev Public Health. 2015;36:39‒53. 10.1146/annurev-publhealth-031914-122509. [DOI] [PubMed] [Google Scholar]
- 13.Heerde JA, Pallotta-Chiarolli M. “I’d rather injure somebody else than get injured”: an introduction to the study of exposure to physical violence among young people experiencing homelessness. J Youth Stud. 2020;23(4):406‒429. 10.1080/13676261.2019.1610558. [DOI] [Google Scholar]