Predictive policing in hospitality and tourism venues — The case of Orlando

Abstract

The association between crime and tourism has been studied for more than 40 years. Most of the studies, however, have focused on differentiating crimes against tourists from crimes against locals despite the high correlation between the two. To date, there are only a few studies that have pointed out to the role of location and time in tourism-crime analysis despite the existence of theoretical frameworks such as routine activity, hot spot, and rational choice that validate the role of temporospatial analysis in tourism/hospitality crimes. Furthermore, prior literature has only used the predictive policing model in relation to police-criminal activities. This study, however, claims that by using the principles of the predictive policing model in conjunction with the community policing model, benefit can be derived from active public participation in preventing/disrupting crimes that have temporospatial patterns. In order to address the gap and achieve the purpose of the study, 160,947 structured observations of Orlando police public records from 2009 to 2015, types and locations of crimes, decision tree models of classification and regression (CART or CRT), and chi-square automatic interaction detection (CHAID) were employed. The results confirm that crimes at recreation/tourism and hospitality venues have a clear temporospatial pattern and, as such, they could potentially be intervened in and reduced with active participation of the public.