Duke study casts doubt on effectiveness of ShotSpotter in Fayetteville 

An independent evaluation of Fayetteville’s use of the automated gunshot detection technology ShotSpotter will test some Fayetteville City Council members’ firm faith in the controversial program.   

The study found that, at most, 24% of ShotSpotter alerts, when not in conjunction with a 911 call, could be linked to confirmed gunfire. While ShotSpotter did expedite first responders’ arrival times, the evaluation was unable to provide conclusive evidence of the technology providing measurable cost-saving benefits to the city or improved outcomes for gun violence victims. 

The study, conducted by the Wilson Center for Science and Justice at Duke Law, is the result of the Fayetteville City Council’s decision last September to renew a one-year contract with SoundThinking, the company that offers the technology. When the council voted to approve the $210,000 contract, they also asked for an independent evaluation of the data to gauge its effectiveness at reducing gun violence. 

That report was made public on Wednesday, just over a week after activists held a demonstration at city hall, urging the Fayetteville City Council to divest from ShotSpotter amid its impending contract renewal decision as the one-year mark approaches.

The Duke Law researchers could not come to a conclusion about whether ShotSpotter reduced gun violence in Fayetteville, a claim that several city council members have made. But researchers did determine that the Fayetteville Police Department’s use of ShotSpotter resulted in faster dispatch times, with responding officers arriving at the scene over two minutes faster than when only a 911 call was made. 

“Despite the high volume of ShotSpotter alerts, we do not have evidence that ShotSpotter-only notifications significantly improve police productivity or outcomes without corroborating 911 calls,” the study states. “Alerts confirmed by both ShotSpotter and 911 calls produce more evidence collection, victim identification, and arrests than ShotSpotter-only alerts. ShotSpotter alerts alone, when not accompanied by a 911 call, however, have low yields.” 

Fayetteville City Council members also discussed the report in closed-door meetings on Aug. 13. City communications director Loren Bymer told CityView that public access was restricted because no more than five city council members were present. The majority of members of a public body are required to be there for a meeting to be open to the public, as per state laws governing open meetings. 

At Monday’s meeting, city council members briefly discussed the study, when Police Chief Roberto Bryan presented some slides on ShotSpotter during his quarterly report. Bryan was asked by council members about discrepancies between the Duke Law study and his report, which showed Fayetteville police connected 250 alerts to 239 bullet casings, but it did not specify the percentage of bullet casings found in each alert. In contrast, the Duke study reported 16.5% of ShotSpotter-only alerts resulted in shell casings recovered, or 127 alerts out of 771. 

“But I didn’t see that in our report. Is there a reason?” Council Member D.J. Haire asked. 

“Sir, I can’t tell you why they weren’t able to pull this information,” Bryan replied.  

Council Member Mario Benavente asked what percentage of the 250 alerts had corroborating evidence of gunfire. Bryan said he could not provide the number.

Key findings of Fayetteville study 

ShotSpotter works by placing acoustic sensors in areas with high gun violence; officers are then alerted of the location, and respond accordingly. In Fayetteville, ShotSpotter is deployed in three one-mile radius zones: around Cliffdale and Reilly Roads in west Fayetteville, in the Murchison Road corridor, and in the Massey Hill neighborhood. According to the Duke Law study, these areas account for about 3% of the city’s total area, but on average 11% of the city’s gunshot-related incidents occurred in ShotSpotter zones. 

The study examined gun violence data from two periods: before ShotSpotter installation, from March 1, 2022 to Sept. 25, 2023, and after ShotSpotter installation, from Sept. 26, 2023 to March 31, 2025. The researchers used a variety of datasets to conduct their findings, including firearm-related 911 calls (7,625), the police department’s ShotSpotter tracking system (975 records), and the police department’s open crime data portal (1,166 records).   

While less than a quarter of ShotSpotter-only alerts could confirm gunfire, the number of ShotSpotter notifications that corresponded with a 911 call and were tied to confirmed gunfire was far higher, with approximately 67% of those alerts leading to evidence of gunfire. The researchers were unable to compare the data from 911 calls alone, a key metric they said could have shed light on the usefulness and effectiveness of the program. 

“Due to the nature of the data available for this evaluation and the nature of the evaluation itself, we cannot directly address several questions of potential interest, for instance, whether the implementation of ShotSpotter has reduced overall levels of gun violence in Fayetteville and how ShotSpotter has affected policing resources in Fayetteville,” the study states. 

Because of the data limitations, the authors used various parameters as evidence to determine if a ShotSpotter alert corresponded with a confirmed shots-fired incident. These included arrests, shell casings, guns, property damage, witnesses and victims. 

ShotSpotter criticism 

While SoundThinking, the company that operates ShotSpotter, has asserted the technology can detect a gunshot with 97% accuracy, research studies have not been able to verify that claim. The technology has faced widespread criticism — both locally and nationally — over its effectiveness and potential for racial bias. 

“It has been well established in examinations of ShotSpotter implementations across numerous jurisdictions that the number of ShotSpotter alerts is much higher than the number of 911 calls for service about gunshots, and that the majority of ShotSpotter alerts are unable to be confirmed,” the Fayetteville study states. 

For example, in a study from Chicago’s Office of the Inspector General of Chicago’s use of the technology, evidence of gunfire was found in less than 10% of cases. A 2024 audit from New York’s comptroller found that in June 2023, 82% of ShotSpotter alerts could not be confirmed when officers arrived at the scene.

“While ShotSpotter provided more alerts about potential gunfire than 911 calls alone and enabled faster response times, it remains unclear to what extent these increased alerts represent false positives,” the Fayetteville study concludes. “The value of increased alerts and faster responses, including if some portion of them are false positive alerts, must be weighed against budgetary and opportunity costs.” 

The authors of the study will present their findings to the Fayetteville City Council at its work session Tuesday. The city council will then decide whether to move forward with the tech, or follow the path of other cities, like Durham and Winston-Salem, that have cancelled their contracts.

Government accountability reporter Evey Weisblat can be reached at eweisblat@cityviewnc.com or 216-527-3608. 

Did you find this story useful or interesting? It was made possible by donations from readers like you to the News Foundation of Greater Fayetteville, a 501(c)(3) charitable organization committed to an informed democracy in Fayetteville and Cumberland County.

Please consider making a tax-deductible donation so CityView can bring you more news and information like this.