CERC assisted the Institute for Municipal and Regional Policy (IMRP) at Central Connecticut State University with their testing for the potential of racial profiling by law enforcement officers in Connecticut. CERC was requested to conduct statistical analyses of the approximately 620,000 traffic stops submitted by Connecticut’s 117 law enforcement agencies.
The data, collected and submitted as mandated by the Alvin W. Penn Racial Profiling Prohibition Act, was analyzed and reported by IMRP and CERC research economists to the Connecticut General Assembly’s Public Safety and Security Committee on April 7, 2015.
CERC designed a methodology that consisted of the statewide application of various tests including: Grogger and Ridgeway’s Veil of Darkness, Knowles Persico and Todd’s Hit Rate, and Ritter’s Solar-powered Model of Stops and Searches. CERC’s work underscored the importance of applying rigorous controls and quality data when using statistical analyses to address a complicated issue like the potential for racial profiling. The analyses showed that:
- There was a statistically significant disparity in the numbers of minorities stopped during periods of daylight as compared to the numbers stopped during periods of darkness. Assuming that motorist race and ethnicity is visible to police officers during daylight hours and that the analyses adequately controlled for possible differences in the risk-set of drivers on the road, this finding indicates that some departments and/or officers could be engaged in racial profiling.
- The analyses took a closer look into which departments were affecting the statistical disparity and found that a large portion of the statewide variation was driven by police departments in Granby, Groton Town and Waterbury as well as in State Police Troops C (Tolland) and H (Hartford).
- CERC’s data analysis was critical to the validation of results. It helped to discard invalid assumptions such as: that the driving population is the same as the resident population, that all residents drive equally, and that there is no seasonal or circadian difference in driving patterns. CERC presented the methods and results to the Racial Profiling Prohibition Advisory Board and Data Working Group. CERC also provided technical summaries in a variety of public forums and meetings.
The raw data has been made publicly available on CTdata.org, the website of the Connecticut Data Collaborative, and the data will be updated on a quarterly basis. Links to the April 2015 report and the raw data can be found online at http://s429795233.onlinehome.us/reports/.