Evidence-Based Policing

Targeting the Most Harmful Co-Offenders in Denmark: a Social Network Analysis Approach Abstract Research Question Is there a 'power few' individuals in Denmark who, through consistent co-offending, produce the highest frequency of crimes and the most harm to society amongst all co-offenders? Data We analysed official statistics from the Police Crime Case Management System in Denmark on all 437,717 charges for violations of the Danish Criminal Code, the Illegal Substances Act and the Weapons Act, in which co-offender relationships were identified from 2007 to 2017, equal to 28% of the national total of all 1,554,943 such charges filed against both solo offenders and co-offenders in that time period. Methods We cross-referenced charging records with crime harm values taken from the Danish Crime Harm Index to measure the severity of all offence types charged. A social network analysis (SNA) algorithm was applied to the data to test for centrality and identify key co-offenders. Findings While 7.5% of the co-offending population accounted for 50% of crime volume, only 3.6% of the co-offenders accounted for 50% of total crime harm. The latter made up just 1.2% of the overall offender population in Denmark, but contributed 24% of overall harm. Social network analysis of how central that power few was in relation to other co-offenders suggests an even smaller cohort of co-offenders—the 'power few of the power few'—who are disproportionality more connected to other co-offenders. Conclusions The 'power few' phenomenon exists in co-offender networks, with a pronounced concentration of harm caused by a small number of co-offenders. The evidence suggests that targeting co-offenders based on social network analysis can enhance the harm potentially reduced by both investigations and crime prevention strategies.

Evidence vs. Professional Judgment in Ranking "Power Few" Crime Targets: a Comparative Analysis Abstract Research question How accurately can local police officers use professional judgement to identify the highest-crime street locations and offenders with the most crime and harm, in comparison to an evidence-based rank-ordering of all possible locations and names derived from police force records? Data A face-to-face survey was conducted in groups with a purposive convenience sample of 123 operational police officers to ask their professional judgement for selecting the ten most crime-prone streets and suspected offenders in their command areas. Separate rankings by crime harm were also requested. Cambridgeshire Constabulary crime and confirmed suspect reports were analysed to create the same lists the officers were asked to provide. Methods The study compared results of surveys of police officers asked to name the top 10 streets and offenders for volume and harm of crimes committed in each policing area to the top ten lists generated by comprehensive and systematic analysis of reported crimes. Findings The top ten lists generated by officers were highly inaccurate compared to the lists produced by comprehensive analysis of crime and charging records. Officers surveyed were 91% inaccurate in naming the most prolific suspected offenders in their areas and 95% inaccurate in naming the most harmful suspected offenders. Officers were slightly less inaccurate in naming the streets in their areas with the highest frequency of crimes (77% incorrect) and the greatest severity of crimes (74% incorrect). Officers in urban areas (N = 42) were substantially more accurate than officers working in semi-rural areas (N = 30) in identifying streets with the highest crime frequency (Cohen's d = 0.9; p = .00) and highest total harm (Cohen's d = 1.3; p = .00), but urban officers still failed to name about two-thirds of the most harmful streets. Conclusions Police officers can benefit from evidence-based targeting analysis to help them decide where their proactive and preventive work can be deployed with the greatest benefit.

Forecasting Knife Homicide Risk from Prior Knife Assaults in 4835 Local Areas of London, 2016–2018 Abstract Research Question How accurately can all recorded locations of 97 knife homicides in one year be forecast across all 4835 Lower Layer Super Output Areas (LSOAs) of London, based only upon all 3506 known locations of nonfatal knife injury assaults in the preceding year? Data All recorded "knife crimes" in the Metropolitan Police Service (MPS) area of London in the financial year 2016–2017 (year 1) were manually reviewed to identify all 3506 reported locations of knife-enabled (KE) but nonfatal injuries, as distinct from other events digitally coded as "knife crimes", such as displaying, threatening with, or carrying knives. All KE homicides in 2017–2018 (year 2) were then added to the database. Methods Each KE injury assault in year 1 was classified for occurrence in one of London's 4835 LSOAs. The total N of such crimes within each LSOA was summed across all records to divide all LSOAs into seven categories of frequency of KE injury assaults in 2016/2017: 0, 1, 2, 3, 4, 5, and 6 or more. We then divided the N of LSOAs in each category in 2016/2017 (year 1) into the total N of knife homicides in 2017/2018 (year 2). False positives, false negatives, and percentage of homicides targeted were calculated for six different targeting plans. Findings Over two thirds (69%) of KE homicides in 2017/2018 (year 2) occurred in just 67 (1.4%) of all 4835 LSOAs, comprising 3.3% of the 2048 LSOAs that had had one or more of the 3506 KE injury assaults mapped in 2016/2017 (year 1). The proportion of LSOAS with a KE homicide in year 2 was higher where there were higher numbers of KE injury assaults per LSOA in year 1. Among LSOAs with zero assaults in year 1, only 1% had a KE homicide in year 2. Among LSOAs with six or more KE injury assaults in year 1, 15% had a KE homicide in year 2. While the risk of homicide was 1400% higher in the hottest spots of knife assaults than in the coolest spots, the absolute number of year two KE homicides in those hottest areas was only 6% of the total. All LSOAs with one or more year 1 KE assaults had three times as much KE homicide risk than LSOAs with no year 1 KE assaults. Targeting all 2048 areas that had one or more year 1 KE assaults would have yielded a true positive rate of over 3% and covered 69% of actual homicide locations. The true negative rate for 2781 areas that would not have been targeted based on having no year 1 KE injuries would have been 99%. Conclusions Predicting which local areas are most likely to suffer knife-enabled homicides, based only on recent nonfatal knife injuries, can pinpoint risks of homicide in local areas that are up to 1400% higher than in most local areas, offering a range of strategies for resource allocation.

Intimate Partner Homicide in Denmark 2007–2017: Tracking Potential Predictors of Fatal Violence Abstract Research Question What are the characteristics of intimate partner homicide in Denmark that may inform more accurate prediction and prevention of such crimes, and how is knowledge of those facts distributed across families, police and other agencies? Data All Danish police data on all 77 cases of intimate partner homicide (IPH) reported over 11 years (2007 through 2017) in 10 of the 12 Danish Police Districts were coded by the first author, comprising 75% of all known IPH cases in Denmark for those years and 100% of the cases in Districts that allowed access to their investigative files. Methods Potentially predictive variables were selected for coding based on similar studies recently completed in the UK and Australia, with comparisons made between these Danish results and the prior findings. Special emphasis was placed on which potential predictors had been known by some organisation or person, but never reported to police until after the homicides had occurred. Findings Characteristics of IPH cases in this study are remarkably similar to those found in the UK studies, starting with the nearly identical rates of IPH per 100,000. The Danish data show even greater prevalence of male offenders having discussed, threatened or attempted suicide prior to killing their intimate partner (52%) than in the all-England sample (40%). Danish IPH cases were even more likely (71%) than the Thames Valley cases (54%) to have had no prior recorded police contact as a couple before the IPH incident. In Denmark, parties other than the police had prior knowledge of domestic abuse in almost half (47%) of the murders, including public agencies (22% of IPH), but did not share that information with the police. Conclusions In Denmark as in the UK, police have no prior contact with most intimate couples suffering a homicide. While suicidal tendencies are highly prevalent in such cases, police have no legal framework for obtaining intelligence about such risk factors from health and social agencies. These findings show that in Denmark as elsewhere, creating a duty by other public agencies to share information might improve prediction and prevention of domestic homicides.

Implementing a Burglary Prevention Program with Evidence-Based Tracking: a Case Study Abstract Research Questions How was a system of tracking and feedback implemented to achieve a high percentage of treatment delivered as intended to residences targeted for super cocooning, and how successful was the tracking in ensuring implementation of the cocooning program? Data The case study of implementing a tracking system for notifying neighbours of burglary victims is based on intensive tracking of all 43 burglaries in a 2-month summer period in one small area of Greater Manchester. Method The system of tracking featured a written paper and pencil report on each residence visited, with entries audited against GPS records of where each reporting officer was located at the times and places listed in the written report. Body-worn video records were used to triangulate the written and GPS records. The policy tracked was the goal of face-to-face contact with residents of four houses on either side of each burgled residence, plus the initially burgled home, or nine visits per burglary. Results Of the potential 387 homes to visit after each burglary, only 266 occupied dwellings were identified, of which 230 (93%) were visited by Police Community Service Officers (PCSOs) assigned, 141 of them within 2 days after the initial burglary (61%). These visits and second visits resulted in 154 face-to-face discussions warning occupants about an elevated risk of burglary (58%). No contact or leaflet at all was delivered to 12.5% of all potential properties. The case study showed that the tracking and feedback was instrumental in delivering these outputs. Conclusion The time and effort needed for supervisors to track the delivery of an evidence-based, tested crime prevention programme does not exceed supervisory time available, making tracking systems of this kind a feasible element of police business as usual.

Face-to-Face Restorative Justice Conferences for Intimate Partner Abuse: an Exploratory Study of Victim and Offender Views Abstract Research Question In what proportion of intimate partner abuse cases (in which offenders have made some level of formal admission of wrongdoing) would victims, offenders or both be willing to attend a face-to-face restorative justice conference, at least in principle? Data The study included offenders, victims and dyads in 75 low-harm cases of intimate partner abuse (IPA), defined as punishable by 10 or fewer days of Cambridge Crime Harm Index value for the offence. Eligible cases were limited to couples involving a male offender who had admitted responsibility for the harm and a female victim, both age 18 years or over, in cases finalised within 6 months of the incident and with no new incidents since the presenting offence. Of the 75 eligible cases of IPA for which contact was attempted with 150 people, 54 people agreed to grant interviews: 34 victims (45% of those successfully contacted) and 20 offenders (49% of successful contacts) completed interviews, including 13 cases in which both offender and victim in the dyad gave interviews. Method Structured phone interviews were conducted by police officers or staff within 10 months of case reporting. The key feasibility measure was a statement that the interviewee would be 'likely' to agree to participate in a face-to-face restorative justice conference (RJC) for crimes in an IPA relationship, based on a 5-point Likert scale. Findings There were 14 (41%) victims and five (25%) offenders who indicated they would likely or definitely be willing to participate in a RJC if one were offered to them now. Seven of the 13 dyads matched on some degree of willingness to participate between both parties. Victims and offenders with some reason for ongoing contact with the other party tended to be more interested in RJCs than those who had severed contact. Victims reporting lower current levels of fear of the offender and who self-reported fewer prior offences were more willing to participate in RJCs. Cases in which victims indicated willingness had more often resulted in a caution than a charge. Conclusions These findings demonstrated an appreciable level of interest in RJCs amongst victims and offenders in IPA cases with low harm. Based on the results of this study, a controlled trial could be considered offering this intervention to couples fitting the eligibility criteria of this study, with emphasis on couples found most likely to accept the offer.

Can Police Prioritize Highest Risks of Harm Among 6670 Children Exposed to Domestic Abuse? Abstract Research Question Can police prioritize children they encounter by risk of future harm involving those who witnessed domestic abuse? Data We tracked the prevalence and frequency of future harm involving all 6670 children who Kent Police (UK) found witnessing domestic abuse and referred to Social Services in 2012. A 3-year tracking period lasted exactly 1065 days after the date of each initial referral. Methods Four categories of events were tracked for each child: (1) further witnessing of domestic abuse, (2) a child being named as a suspected offender of a criminal offense, (3) a child being reported as a victim of a criminal offense, and (4) a child being reported as a missing person from the home. Findings Further harmful events were highly prevalent (48%), yet over 92% of those events were repeated reports of witnessing domestic abuse. Less than 3% of the cohort went on to become a victim of crime, 1.6% were reported missing, and 6% were named as offenders in a crime over the 3 years. Total events including new witnessing of DA were heavily concentrated in a "power few": under 10% were reported in 50% of the 8685 subsequent events over the next 3 years. Conclusions Tracking subsequent harm to children present at domestic abuse shows more than enough concentration of future events to justify selective targeting of resources for further efforts to protect these children, ideally by using advanced predictive analytics such as random forests modeling.

Tracking Procedural Justice in Stop and Search Encounters: Coding Evidence from Body-Worn Video Cameras Abstract Research Question To what extent do police stop-and-search practices, as captured on officers' body-worn video cameras, adhere to key dimensions of procedural justice theory, and do levels of adherence vary across the dimensions of (1) citizen participation and voice, (2) neutrality and explanation, (3) respect and dignity and (4) trustworthy motives? Data A random sample of 100 recorded stop-and-search encounters was selected from all 601 encounters recorded by Greater Manchester Police between 1st January and 31st August 2017, with an average duration of 12 minutes of combined video and audio with transcripts. Methods These records were coded entirely by the first author for the four main dimensions of procedural justice and details about the actors/participants involved. The dimensions were combined into an overall index score of procedural justice for each encounter. Findings Most stop-and-searches were characterised by a strong element of police allowing citizens to express voice, followed by police demonstrating respect and offering explanation. The lowest scores were given to "conveying trustworthy motives". A standardised metric of each element into a measure of 0–100 coded mean scores for participation/voice = 94, explanation/neutrality = 65, respect = 71 and trustworthy motives = 47. In this latter category, not one officer linked the purpose of the stop and search to the wider organisational purpose of protecting society and helping to keep people safe. Conclusions This evidence suggests the potential value of an ongoing tracking measure for police legitimacy which could be used as a supervisory and human development tool for operational officers, comparing individuals, units, areas and trends over time in objectively coded features of police behaviour towards citizens.

Tracking the Use of Exclusion Zone Orders in Denmark: Individual and Place-Based Crime Trends Before and After 161 Individual Orders Abstract Research Question To what extent does citizen-reported crime in 500-meter square areas in Denmark, and arrests of individuals legally excluded from those areas for intimidating behavior, decline in the 3-month time periods in which their Exclusion Zone Orders (EZOs) are in effect, compared to the most recent 3-month period prior to the EZO? Data Individual-level data on 2441 arrests of the 161 offenders who were the subject of all 218 EZOs granted in Denmark from 2009 through 2016 were collected from the Danish National Police Database, as well as place-based reported crime data for the location of each of the busiest 7 (out of 41) of the Exclusion Zones (EZs) and matched control areas. Methods The two units of analysis are individuals targeted by the EZOs and location of the EZs. A descriptive time-series analysis compared offending before and individuals received EZOs. Standardized mean differences were calculated between seven locations which received a high rate of EZOs and their matched control sites, all distinguishing police-reported crimes from those reported by victims or witnesses. Findings A moderate decrease in the number of citizen-reported offenses was observed in six out of seven targeted zones when the EZOs were in effect. There was no indication of a displacement effect to offenses outside the EZ. Conclusions The positive results of this tracking study suggest that more rigorous testing the use of EZOs across Denmark could demonstrate that EZOs can cause reductions in crime and harm.

'Victim-Offenders': a Third Category in Police Targeting of Harm Reduction Abstract Research Question To what extent do people classified by police as 'offenders' and those they classify as 'victims' overlap, thus creating a third category of individual targets for crime and harm reduction called 'victim-offenders'? Data Leicestershire police records in the 3-year period 2014–2016 were compiled on all 117,766 individuals who were identified as either victims or offenders in all 159,702 crimes in which one or more individual victims were named. A 730-day-at-risk tracking period after individuals' initial appearance in the data set was examined for the tracked cohort of 42,916 individuals who entered the data set in 2014 only and were ultimately identified in 62,336 individual-victim crimes. Methods Each individual who entered the data set from 2014 to 2016 was classified as (only) a victim, (only) an offender, or (both) victim and offender in separate events. We designate the latter category as victim-offenders. The total count of criminal events and total recommended days of imprisonment value of the Cambridge Crime Harm Index (CHI) were calculated for each individual, both for crimes committed and victimizations experienced. Individuals who entered the data set in 2014 were tracked for 730 days from their first recorded offence or victimization. All individuals were ranked, by both crime counts and CHI values, from highest to lowest. Findings From 2014 to 2016, victims comprised 89.9% of the 117,766 individuals ever appearing as victims or offenders, offenders comprised 7.9% and victim-offenders were 3.2%. In 2014, all 38,318 individual victims and 4598 offenders were placed into the 730-day tracking period in which 1825 individuals became victim-offenders, or 4.2%. Initial offenders were seven times as likely to become victims (17.9%) as initial victims were to become offenders (2.6%). Victim-offenders had 74.5% higher average harm scores and 68% higher average crime counts than the overall 2014 entry cohort, with 65% of victim-offenders' harm a result of their victimisation. Only 417 'high harm-high volume' individuals experience or cause the most crime harm as well as the most crime, of whom 49.9% (208/417) are victim-offenders, 33.3% (139/417) victims and 16.7% (70/417) offenders. Conclusion Opportunities to reduce crime and harm through targeting police and partnership resources on the 'power few' (Sherman, Journal of Experimental Criminology, 3(4), 299–321, 2007) can be enhanced by identification of victim-offenders and testing prevention strategies appropriate to this third category of people.

Alexandros Sfakianakis
Anapafseos 5 . Agios Nikolaos
Crete.Greece.72100
2841026182

6948891480

alsfakia@gmail.com