Environmental Contexts and the Polyvictimization of Latino and African American Youth

Background and Purpose: Numerous studies have documented high rates of polyvictimization – the simultaneous exposure to multiple forms of violence – among low-income, minority youth.  Few studies have attempted to simultaneously control for the multiple environmental contexts associated with polyvictimization and the literature suffers from the inability to address geographic selection bias.  This study examines the following questions: (1) To what extent do low-income, Latino and African American youth experience polyvictimization during childhood?; (2) Does polyvictimization vary by ethnicity, gender, and the duration of exposure to environmental contexts?; (2) What environmental contexts make them most dangerous for youth in terms of their chances of becoming polyvictims and the timing of such victimization?

Methods: Data analyzed come from the Denver Child Study, a natural experiment that controls for parental geographic selection bias. This is a retrospective panel study of youth between the ages of 5 and 18 (N=800) who resided in randomly assigned public housing units in Denver (DHA) for at least two years during childhood and prior to any experience of polyvictimization. Normalized Accelerated Failure Time (AFT) models with frailties were estimated to predict time to first polyvictimization. Our empirical models include a wide array of environmental context indictors measuring neighborhood safety, social status, nativity/ethnic composition, physical environment, controlling for youth, caregiver and household characteristics.  Stratified models estimate differences in the influence of these environmental contexts by gender and ethnicity.

Results: Almost half (48%) of the youth had been polyvictims after moving into DHA; the average age of first polyvictimization was 7.9 years.  African American children were significantly more likely to be polyvictims than Latinos; females more than males. The strongest predictors of earlier polyvictimization were higher levels of residential instability (TmR=.935, p < .05) living in older neighborhoods (TmR=.897, p <.001) and higher property crime rates (TmR=.895, p < .01).  Higher proportions of affluent (TmR=1.288, p < .01) or foreign born neighbors (TmR=1.121, p < .01) were significant protective factors delaying the onset of polyvictimization for both Latino and African American children, perhaps underscoring heightened monitoring and supervision by adults in more affluent or immigrant communities that extend to all children in the neighborhood.  Polyvictimization was delayed in neighborhoods with higher violent crime rates (TmR=1.113, p < .01) and concentrated disadvantage (TmR=1.132, p < .01), again suggesting the role of parental buffering to mitigate these negative conditions.  

Implications: The findings help inform the longstanding debate about the aims and consequences of affordable housing policy for increasing opportunities and enhancing the life changes of low-income youth through the location of subsidized households in places that enhance positive environmental contexts. A greater understanding of the causal mechanisms which lead to polyvictimization is important in order to devise appropriate primary and secondary intervention strategies at the neighborhood level.