Exploration of Arrest Activity Among Homeless Young Adults In Four U.S. Cities

Purpose: Prior research suggests that chronically homeless young adults, who are estranged from conventional institutions and rely on support from peers embedded in the street economy, engage in criminal activity. Studies to date have not examined both direct and indirect relationships among predictors of homeless young adults' criminal activity and have been limited to samples from one or two cities or several cities within one region. Using a sample drawn from four cities across disparate regions of the U.S., this study identifies pathways from length of time homeless to arrest activity by exploring transience, substance-use disorder, and survival strategies as potential mediating factors. The study hypotheses were that greater arrest activity would be reported by homeless young adults who: 1) are male; 2) have been homeless longer; 3) are more transient; 4) meet criteria for a substance-use disorder; and 5) use survival strategies to earn income. Length of time homeless was also expected to indirectly predict arrest activity, as mediated through transience, substance-use disorder and survival strategies.

Methods: Using convenience sampling, 188 homeless young adults from Los Angeles (n=50), Denver (n=50), New Orleans (n=50) and St. Louis (n=38) were recruited from shelters and drop-in centers using similar methods. Primary inclusion criteria included being 18-24 years of age, having spent at least 2 weeks away from home in the month before the interview, and signing informed consent. Data collection consisted of a semi-structured retrospective interview to examine participants' homelessness history, transience, DSM-IV criteria for substance-use disorders using the Mini-International Neuropsychiatric Interview, survival strategies and arrest activity. A theoretical model was developed from the estrangement and homeless-youth literature and tested via observed-variable path analysis. The model included five independent variables (gender, length of time homeless, transience, substance-use disorder and survival strategies), the dependent variable (arrest activity), and two mediation pathways. Mplus software was utilized to determine the significance of direct effects and the overall mediation effects using the Sobel test.

Results: The final model represented an excellent fit to the data (X2=9.21, df=8, p=0.33, CFI=0.99, RMSEA=0.028). Greater time homeless, abusing drugs, utilizing survival strategies, and being male were significantly associated with arrest activity. Collectively, gender, length of time homeless, transience, substance-use disorder and survival strategies explained 23% of the variance in arrest activity. The overall mediation effect from length of time homeless to arrest activity was 0.019 (p=0.048) and accounted for 11% of the total effect from length of time homeless to arrest activity.

Implications: Path-analysis results underscore the importance of simultaneously modeling criminal risk factors. The findings that homeless young adults may experience a process toward criminal behavior suggest several points of intervention that may help prevent future criminal activity, including establishing homeless young adults in stable housing, providing substance-use prevention and intervention, and engaging them in employment opportunities. While criminal behavior and arrests appear common among homeless young people, development of services to connect them to supportive adults and institutions is likely to reduce risks to this population and to society as a whole.