Evaluation of Durham’s ShotSpotter Installation

ShotSpotter evaluation shows increased gunshot notifications, quicker police response times, and more arrests, but questions of accuracy and impact remain, and many focus group participants expressed skepticism

For total gunshot notifications, 54% had a ShotSpotter only notification, 15% had both a ShotSpotter notification and a 911 call, and 28% had a 911 call. For confirmed gunshots, 26% had a shotspotter alert, 34% had both a shotspotter alert and a 911 call, and 40% had a 911 call only. For serious incidents, Shotspotter alerted for 26 out of 52, and 911 calls were received for 50 out of 52. Median police response time dropped 1.2 minutes.

ShotSpotter more than doubled gunshot notifications in the Durham pilot area and led to a 1.2-minute reduction in median police response time, researchers were unable to draw conclusions from the data about its impact on gun violence according to a new report from the Wilson Center for Science and Justice at Duke Law.   

The ShotSpotter pilot was deployed from December 15, 2022 – December 14, 2023 in a three-square-mile area of Durham that have historically had comparatively high rates of gun violence. The report, Evaluation of Durham’s ShotSpotter Installation: Results of a 12-month Pilot Project, examines the performance of ShotSpotter in the pilot area and is authored by Philip Cook, Professor Emeritus of Public Policy and Economics, Duke University, and Adam Soliman, Assistant Professor of Economics, Clemson University. It delves into ShotSpotter’s effects on notifications and police deployment; productivity of investigations; and the costs to the city’s budget and police resources. 

The major findings include 

  • For total gunshot notifications in the pilot area (1447), 57% had only a ShotSpotter notification, 15% had both a 911 call and a ShotSpotter notification, and 28% had only a 911 call. 
  • For confirmed gunshots (282), 26% had only a ShotSpotter notification, 34% had both a 911 call and a ShotSpotter notification, and 40% had only a 911 call. 
  • There were 52 serious incidents in the pilot area where someone was wounded or killed. ShotSpotter alerted for 26 of those 52 and 911 calls were received for 50 of the 52. By design, ShotSpotter only detects gunfire occurring outdoors. However in eight incidents where ShotSpotter did not alert, it was due to human or system error. 
  • The ShotSpotter-only notifications more than doubled the total notifications and led to an additional 2.3 police deployments per day in the pilot area.   
  • All gunshot notifications in the city are a level 2 priority for police response, the same as response to household alarms.  ShotSpotter alerts increased the citywide level 2 deployments by about 2%.   
  • ShotSpotter alerts also supplemented 911 notifications by being quicker and providing the location coordinates of the sound.  The median response time of officers to the scene for 911 notifications dropped by 1.2 minutes in the pilot area compared to the rest of the city. 
  • Expanded coverage of gunshots due to ShotSpotter-only notifications resulted in 7 additional incidents that resulted in arrest (an increase of 32%), and 71 additional incidents that resulted in collection of shell casings and other evidence (almost doubling the total). 
  • For incidents that generated both 911 calls and ShotSpotter alerts, the likelihood of arrest or evidence collection did not increase. But there was one such incident in which the responding officers arrived in time to administer first aid to the victim and it is plausible, though uncertain that this contributed to saving a life. 
  • Researchers were unable to determine from the data if ShotSpotter had any impact on gun violence. 

“We are grateful to be able to provide City Council with clear data and analysis that we hope will aid their decision-making process as they consider whether to use ShotSpotter in the future,” said Cook. “While we do not offer a conclusion on whether the benefits of ShotSpotter exceed the costs, our analysis provides a good indication of specifically what Durham got for its money.”  

In addition to the performance evaluation report the Wilson Center has also published a report that looks at the perspectives and experiences of individuals who live in the pilot area: ShotSpotter in Durham: A Community Sentiment Evaluation.  

In this evaluation, authored by Pilar Kelly, Lead Researcher; Angie Weis Gammell, Policy Director; and Lindsay Bass-Patel, Policy Analyst, at the Wilson Center for Science and Justice at Duke Law, researchers conducted focus groups with 30 residents of the pilot area. While participants’ perspectives varied, most participants had not observed any change in police response to gun crime since ShotSpotter was implemented. Participants who expressed opposition to having ShotSpotter in their neighborhoods primarily cited lack of trust in City Council, SoundThinking, policing as an institution, and concerns about data privacy, rather than direct experiences with ShotSpotter.  

“Community members who participated in our focus groups often cited concerns about both the cost and opportunity costs of ShotSpotter, and shared ideas for how the money might be better spent. Another theme that arose was the need to feel heard and engaged by City Council and the Durham Police Department. These insights could help inform decision-making regarding the future use of ShotSpotter or other initiatives to address gun violence in Durham,” said policy director Angie Weis Gammell.  

Read the full performance evaluation report here and the full community sentiment evaluation here.