Contextual Exploration of Neighborhoods and Workplace Violence

Three studies were undertaken to address workplace violence literature gaps related to classification of occupational settings at high-risk for homicide and socioeconomic predictors of workplace violence and robbery risks. The first study used previously collected data on workplace homicide to investigate the designation of workplaces as high-risk informed by workplace characteristics. Cluster analysis was used to group the workplaces according to nine a priori selected workplace characteristics potentially predictive of violence. Cluster solutions were then entered as predictor variables into logistic regression to identify subsets of industries associated with increased odds of worker homicide. Besides identifying high-risk industries in retail and service, consistent with the literature, manufacturing industries were also identified. The absence of gas stations among identified high-risk industries was inconsistent with a priori expectations. The second study used a cross-sectional design to identify associations with Census block group (BG) socioeconomic factors related to the proportion of high-risk industry workplaces (HRIWP). Factor analysis was used to summarize constructs for poverty/deprivation, human/economic capital, and transience/instability. Increasing levels of poverty/deprivation and transience/instability were associated with a higher proportion of HRIWP (odds ratios [OR]=1.98, 95% confidence interval [CI]: 1.47, 2.67 for 4th vs. 1st quartile of poverty/deprivation and OR=1.77, 95% CI: 1.31, 2.37 for 4th vs. 1st quartile of transience/instability, comparing >15% HRIWP to no HRIWP). Contrastingly, high human/economic capital levels were associated with a lower proportion of HRIWP (OR=0.65, 95% CI: 0.49, 0.87 for 4th vs. 1st quartile comparing >15% HRIWP to no HRIWP). The third study examined whether the same BG-level constructs developed in the second study (poverty/deprivation, human/economic capital and transience/instability) were related to workplace robbery risk. A case-control design with pre-existing data was used. Independent associations of poverty/deprivation (OR=1.22, 95% CI: 0.81, 1.86) and transience/instability (OR=1.44, 95% CI=0.96, 2.18) with workplace robbery were consistent with a priori expectations and persisted after adjustment for workplace level factors and crime. Social factors were found to be associated with presence of HRIWP and workplace robbery. This research suggests that development of effective workplace violence prevention controls should recognize and understand social influences on violence and crime in the workplace