Cops on Dots Doing What? The Differential Effects of Law Enforcement-Oriented Actions in Hot Spots

Statement of the Problem: Researchers have not directly measured or compared the effectiveness of different policing tactics implemented in crime hot spots. This restricts the literature's ability to recommend which tactics police should use during hot spots policing (Braga, 2001, p. 121; Telep & Weisburd, 2012, p. 333; Weisburd & Braga, 2006, p. 234). Therefore, this study will address four research questions. (1) Do four police law enforcement-oriented actions reduce violent crime in hot spots? The four law enforcement-oriented actions include: (a) violent crime arrests, (b) disorder crime arrests, (c) traffic enforcement, and (d) stop, question, and frisk. (2) Are any of these four law enforcement-oriented actions more effective than the others? (3) When police commanders allocate resources in crime hot spots, what do police commanders think they are doing? (4) What are police commanders' rationales for what they do in crime hot spots? Answering these questions will provide police departments with evidence to improve hot spots policing. Research Design and Methods: Data (2009 - 2013) from the Philadelphia Police Department (PPD) (Philadelphia, PA USA) will be used to answer research questions one and two. Hot spots will be defined as all street blocks and intersections experiencing five or more violent crimes during 2008. Research question one will be addressed with a repeated measures multilevel model estimating the effect of total enforcement actions on monthly violent crime counts in violent crime hot spots. Research question two will be addressed with a repeated measures multilevel model stimating the effects of violent crime arrests, disorder crime arrests, traffic enforcement, and stop, question, and frisk on monthly violent crime counts in violent crime hot spots. Additionally, Wald tests will be used to statistically compare the magnitudes of the four law enforcement-oriented actions' effects from model two. If any of the four effects are found to be statistically equivalent then their coefficients will be constrained to be equal in a subsequent model and Likelihood-Ratio tests will be used to compare the fit of the constrained and unconstrained models. Research questions three and four will be addressed using non-participant observations of the PPD's crime strategy meetings and interviews with executive and district police commanders responsible for implementing hot spots policing (N = 15).

Products: The findings will be reported in a full-length dissertation, a summary report, peer reviewed articles, and presentations at professional meetings.
ca/ncf