“Soft” policing at hot spots—do police community support officers work? A randomized controlled trial

Objectives

To determine whether crime-reduction effects of increased police patrols in hot spots are dependent on the “hard” threat of immediate physical arrest, or whether “soft” patrols by civilian (but uniformed) police staff with few arrest powers and no weapons can also reduce crime. We also sought to assess whether the number of discrete patrol visits to a hot spot was more or less important than the total minutes of police presence across all visits, and whether effects based on counts of crime would be consistent with effects on a Crime Harm Index outcome.

Methods

We randomly assigned 72 hot spots into 34 treatment units and 38 controls. Treatment consisted of increases in foot patrol by uniformed, unarmed, Police Community Support Officers (PCSOs) who carry no weapons and hold few arrest powers beyond those of ordinary citizens. GPS-trackers on every PCSO and Constable in the city yielded precise measurements of all patrol time in all hot spots. Standardized mean differences (Cohen’s d), OLS regression model, and Weighted Displacement Quotient are used to assess main effects, to model the interaction effect of GPS data with treatment, and to measure the diffusion-of-benefits of the intervention, respectively. Outcomes included counts of incidents as well as the Cambridge Crime Harm Index.

Results

As intended, patrol visits and minutes by Police Constables were equal across the treatment and control groups. The sole difference in policing between the treatments groups was in visits to the hot spots by PCSOs, in both the mean daily frequency of discrete visits (T?=?4.65, C?=?2.66; p?≤?.001) and total minutes across all visits (T?=?37.41, C?=?15.92; p?≤?.001), approximately two more ten-minute visits per day in treatment than in control. Main effect estimates suggest 39 % less crime by difference-in-difference analysis of reported crimes compared to control conditions, and 20 % reductions in emergency calls-for-service compared to controls. Crime in surrounding areas showed a diffusion of benefits rather than displacement for treatment hot spots compared to controls. A “Reiss’s Reward” effect was observed, with more proactive patrols predicting less crime across treatment hot spots, while more reactive PCSO time predicted more crime across control hot spots. Crime Harm Index estimates of the seriousness value of crime prevented ranged from 85 to 360 potential days of imprisonment in each treatment group hot spot (relative to controls) by a mean difference of 21 more minutes of PCSO patrol per day, for a potential return on investment of up to 26 to 1.

Conclusions

A crime reduction effect of extra patrols in hot spots is not conditional on “hard” police power. Even small differences in foot patrols showing the “soft power” of unarmed paraprofessionals, holding constant vehicular patrols by Police Constables, were causally linked to both lower counts of crimes and a substantially lower crime harm index score. Correlational evidence within the treatment group suggests that greater frequency of discrete PCSO visits may yield more crime reduction benefit than greater duration of those visits, but RCTs are needed for better evidence on this crucial issue.

     

Testing for Analysts’ Bias in Crime Prevention Experiments: Can We Accept Eisner’s One-tailed Test?

Eisner (Journal of Experimental Criminology, this issue, 2009) suggests that developer-led evaluations often make programs look better than independent evaluations do because the former suffer systematic biases in favor of prevention success. Yet, his proposed remedies suffer their own systematic bias, constituting a ‘one-tailed’ test of bias in only one direction. In this response we suggest that a more objective assessment of ‘analysts’ effects’ requires a ‘two-tailed’ test of bias, in which reviewers would measure indications of bias both for and against success in evaluations reported by both developers and independent evaluators. After exploring the full complexity of the distinction between developers and evaluators, we report on one case in which independent evaluations were more favorable than those of developers. We then suggest possible indicators of analysts’ biases against finding success that may characterize the work of developers who “bend over backwards” to find harm in their programs, and of independent evaluators who may seek to “get a scalp” of a developer or a program.

Detection of amitriptyline, citalopram, and metabolites in porcine bones following extended outdoor decomposition

Skeletal remains of a domestic pig were assessed for relative distribution of amitriptyline, citalopram, and metabolites. Following acute exposure and outdoor decomposition for 2 years, drugs and metabolites were analyzed in 13 different bones. Bones were pulverized following a simple wash procedure, and drugs were extracted by passive incubation in methanol, followed by solid-phase extraction. Samples were analyzed by ultra-high performance liquid chromatography (UHPLC) and confirmed with gas chromatography-mass spectrometry. The Kruskall-Wallis test showed that bone type was a main effect with respect to drug level for all analytes, with levels varying from 33- to 166-fold. Ratios of levels of drug to that of the corresponding metabolite were less variable, varying roughly one- to eightfold. This suggests limitations in the interpretive value of drug measurements in bone and that relative levels of drug and metabolites should be further investigated in terms of forensic value.