Study
The 2013 study by Azrael, Braga, and O’Brien of the Milwaukee Homicide Review Commission (MHRC) consisted of three principal components: 1) a formative evaluation, 2) a process evaluation, and 3) an impact evaluation (the focal point of this review is the impact evaluation). The impact evaluation included a statistical analysis of what impact the implemented violence prevention strategies had on crime reduction in the treatment, districts relative to control districts. The goal of the evaluation was to assess whether the MHRC made a measurable short-term impact on homicides in treatment districts relative to control districts.
The evaluation used a quasi-experimental design and involved matching the Milwaukee Police Department (MPD) districts into like pairs based on homicide counts, nonfatal shooting counts, and sociodemographic characteristics of the populations in the districts. After the districts were matched based both on their demographic and crime profiles and on the expert opinion of police personnel, one district from each pair was randomly allocated to the MHRC treatment group. This resulted in three districts (Districts 2 and 6 on the south side of the city and District 5 on the north side) serving as the treatment districts, and four districts (1, 3, 4, and 7) serving as the control districts (District 1 experienced no homicides or nonfatal shootings in the period used to match districts, thus it was excluded from consideration for the MHRC intervention and designated as part of the control group).
The MHRC intervention was implemented in August 2005. Time series of monthly counts of homicides between January 1999 and December 2006 in the treatment and control districts were examined to determine whether the MHRC was associated with any reductions in homicides. Regression analyses (which controlled for secular trends, seasonal variations, population changes, and violent crime rate trends) were used to estimate changes in the monthly counts of homicides in the treatment and control districts. Data on homicide counts came from the MPD.
From May 2005 through December 2007, the MHRC reviewed 173 homicides and 99 nonfatal shootings. Eighty percent of violent crimes were committed with firearms, and the majority of victims were Black males (65 percent), as were the suspects (80 percent). More than 60 percent of homicides resulted from arguments, each of these either an ongoing issue or a random dispute involving members of a community or gang. Thirty-nine percent of crimes were precipitated by another crime, and 23 percent were associated with a robbery incident. Fewer than 20 percent involved gang members.
The study authors noted that because the necessary pretest and posttest process data was not collected, it was not possible to shed light on the specific mechanisms responsible for any observed reductions in homicide associated with the MHRC interventions. The impact evaluation was designed to measure the value added by the MHRC process, but the authors did not attempt to parse out the varying effects of the specific initiatives that were implemented.