As a summer surge in national gun violence rampages into the fall, the “defund the police” debate that now includes a movement to abolish the tools used by police has been playing out nationally, in Buffalo, and in the pages of this newspaper.
This debate has drawn the company I lead, ShotSpotter, into the crossfire. This controversy has been counterproductive due to misconceptions, imperfect information and intentional obfuscation. The costly byproduct has been an unfortunate distraction from a much more important conversation of how communities and police need to work together to enhance public safety for all.
Founded 25 years ago, ShotSpotter’s core purpose is to make communities safer, help reduce gun violence and, ultimately, save lives. Specifically, our acoustic gunshot detection service helps police departments quickly and precisely respond to gunfire within a minute of a sound of a gunshot, so that officers can get to shooting locations as quickly as possible to attend to victims, gather evidence and engage with residents in order to best serve and protect those communities suffering from persistent criminal gunfire. Our mission is particularly critical considering a finding by a Brookings Institution study that found approximately 88% of gunshots go unreported using the traditional 911 system.
In addition to providing instantaneous alerts, we are requested to prepare forensic reports to submit to courts in the context of a criminal case proceeding. These reports are prepared by experts who commit hours to harnessing all the acoustic and location data captured by our state-of-the-art sensors in coverage areas. The technology is a focused tool – highly accurate and unbiased in delivering evidence of a gunshot incident, including recorded sound of gunfire and the time and location of a shooting. This highly objective and factual evidence is commonly used by courts and can be examined and tested by both the prosecution and defense. ShotSpotter evidence on its own has never been nor could never be responsible for the charging, arrest or conviction of anyone accused of a crime. These forensic reports have been accepted by courts in more than 200 cases in 20 states, meeting a variety of legal standards and challenges to its admissibility.
Critics have purposefully distorted how our system works, claiming our algorithms target people. This is simply not the case. Our software uses two basic algorithms. The first is used to detect and locate impulsive acoustic events, which may or may not be gunshots. Our location method gathers time stamps of impulses from multiple sensors used to mathematically compute a specific location. This process is based on techniques that date back more than 100 years (although, we feel that we have perfected this over the years). We have published a detailed description of this method.
The second algorithm looks at all of the incidents with the sole purpose of dismissing those incidents that are not gunfire. This is done before they ever reach one of our human reviewer experts. Dismissing this noise and thereby reducing the inbound workload helps our human reviewers more effectively perform their crucial analysis on the remaining non-dismissed gunfire. Our reviewers classify out the remaining noise and make the final determination on what to publish as actual gunfire. Far from being the kind of secret or sinister algorithm that has given rise to legitimate scrutiny for racial bias, unfairness or infringement on privacy, ours focuses on sounds, exclusively dismissing those sounds that are not gunshots. And again, nothing in our entire process including human review can identify a shooter or anything else beyond the sound, the number of rounds and potentially the type of weapon.
And our technology, human review and publishing process is 97% accurate based on reporting from our customers. This figure has been independently verified in a ShotSpotter-commissioned inquiry by data science firm Edgeworth Analytics. So-called “reports” or “analyses” by critics use cherry-picked data and ignore the fact that much more data is needed to tell the full story of a gunshot incident than what officers find (or don’t find) at the scene. Simply put, absence of evidence does not mean a gunshot did not happen, and it is often the case that more facts come to light in the days after an incident.
Lastly, the most troubling aspect of the misplaced criticism of ShotSpotter ignores the many successes agencies have had utilizing our technology. For example, in Oakland, Calif., 101 gunshot wound victims were found and aided by police due to ShotSpotter alerts when no one called 911. Pittsburgh, Pa., reported a 36% drop in homicides year-on-year in 2018 and 2019, continuing a downward trend over five years. In 2019, Greenville, N.C., experienced a 29% reduction in gun violence injuries and 20% reduction in homicides in the first year of ShotSpotter deployment. In West Palm Beach, Fla., there has been a 60% reduction in homicides and other gun-related injuries after the installation of ShotSpotter in 2018, while Miami had a 35% decrease in homicides between 2014 and 2017 after adopting ShotSpotter. In April of this year, Chicago police were alerted to the scene of a shooting by ShotSpotter and were able to save a 13-year-old boy’s life. Without ShotSpotter, police may have never found him. In August, ShotSpotter helped police investigate shootings in Columbia, Pittsburgh and Toledo, and was cited as a valuable tool for police in three recent Youngstown, Ohio, shootings.
Critics ignore or downplay these success stories because they would rather focus instead on the over-policing narrative – charging that ShotSpotter alerts cause police to respond in ways that increase hostile interactions with the community. There is zero data supporting that claim and, in fact, since the beginning of 2020, ShotSpotter alerts have resulted in an “officer involved shooting” in only 0.08% of incidents, many of which the officer was shot at and did not discharge his weapon. The principal question should be, when so few gunshots are reported to 911, why shouldn’t areas where gun violence occur receive a prompt police response? Police that are meant to protect and serve simply cannot afford to ignore and not respond to criminal gunfire. This is not an effective strategy.
I understand that we are living in a moment in which society is engaging in a bracing, honest reflection on issues of how to make policing more efficient, effective and equitable. As a company, we are grateful be able to contribute technology innovation that advances policing and the public safety outcomes that can be co-produced through more precise and actionable data, specifically focused on the seemingly intractable issue of gun violence.
Ralph A. Clark is the president and CEO of ShotSpotter.