Race and Officer Decision Making: Examining Differences in Arrest Outcomes between Black and White Officers
Robert Brown and James FrankJustice QuarterlyVol. 23No. 12006pp 96-126,
Brown and Frank (2006) contribute to the overall understanding of officer decision-making by presenting results from an observational study of Cincinnati Police Department officers. The data used in their analysis was collected between April 1997 and April 1998, during which observers witnessed 614 police-suspect encounters. In order to determine possible differences in arrest decision-making, their analysis was restricted to observations of police-citizen encounters where the citizen involved was considered to be a suspect at some point during the contact (Brown and Frank, 2006). Their primary research question addressed the possibility that influences on officer decision making vary by officer race. That is, are white and black officers influenced by the same factors when it comes to arrest decisions?
The authors originally intended to estimate and report Hierarchical Linear Models (HLM) to assess the nature of officer decision making. However, due to the small number of cases nested within each geographical area, concerns about the stability of the coefficients needed to be addressed. Brown and Frank (2006)thus estimated and reported a series of logistic regression models in order to assess influences of officer race on arrest decisions. Initially a model including all police-suspect encounters was estimated, producing results consistent with prior research. Specifically, the authors found that officers were more likely to arrest younger, non-white, male suspects and that quality of evidence, suspect demeanor and whether the suspect was intoxicated influenced arrest decisions (Brown and Frank, 2006). Although officer race was not included in the reported model, Brown and Frank (2006) footnote that white officers were more likely to make an arrest than black officers when they estimated an additional model (not reported) including all police-suspect encounters.
As previously mentioned, the nexus of this article was to determine if certain factors differentially affect black and white officers’decision-making processes. Consequently, Brown and Frank (2006) partitioned the data and estimated two additional models, one for black officers and one for white officers. Contrary to conventional wisdom, their analysis revealed that exogenous variables in the models differentially influenced black and white officers’ arrest decisions. When arrest decisions were examined separately,certain factors influenced black officers’ decision making but produced no such effect on white officers’ decisions to arrest. For example, black officers’ decisions were influenced by length of service, suspect race[1],number of bystanders and whether a crime was committed in the presence of officers, but none of these factors affected arrest decisions for white officers(Brown and Frank, 2006). Conversely, the authors found that white officers were influenced by the quantity of evidence and being dispatched to the scene whereas black officers were not.
While these differences between officers are important, it should be noted that only two correlates exhibited statistically different effects on officer decision making. Despite the fact that white officers were more likely to make an arrest in general, black officers were more likely to arrest black suspects yet less likely to arrest white suspects than their counterparts (Brown and Frank, 2006). The only other factor that differentially influenced black and white officers significantly was years of experience. The longer black officers worked with CPD the less likely they were to make an arrest. This was not the case for white officers.
The findings presented by Brown and Frank (2006) diverge from previous study results and suggest officer race does not have a null effect on arrest decisions. At least in the sample they examined, white and black officers were differentially influenced by certain factors surrounding the police-suspect encounters. Although their findings could be context specific and most previous studies indicate these factors have little influence over officer behavior,future research should consider these correlates to determine the true relevance of officer level characteristics on decision making.
Charles F. Klahm IV University of Cincinnati, USA
