Why ShotSpotter Case Clearance Depends on More Than Tech

Estimated reading time: 11 minutes

Here’s the pitch every city council has heard at least once: put sensors on the rooftops to catch gunfire the instant it happens, put cameras on the corners to catch what happens next, and case clearance rates should climb. Sounds like a no-brainer. Chicago bet on exactly that combination for the better part of two decades, spending tens of millions of dollars a year on it. Then, amid mounting public controversy, on September 22, 2024, the city shut its gunshot sensors off for good.

So did the bet pay off? It started small. In 2005, Chicago began installing a citywide bank of surveillance cameras, eventually building out to nearly 6,000 unique camera deployments. In 2012, the city piloted ShotSpotter, the acoustic gunshot detection system, in small pockets of two police districts. By 2017, it covered more than 60% of Chicago’s land area, more than 136 square miles, at a price tag between $8.8 and $12.3 million a year.

For crime prevention, the early evidence looked decent, at least for the cameras. Prior research had already shown Chicago’s camera network reduced crime in the neighborhoods where it was installed, though when CCTV research does turn up a case-clearance benefit elsewhere, it tends to be for theft and disorder, not gun violence. GDT’s own record was messier: no measurable drop in fatal shootings, nonfatal shootings, or gun crime overall, just a real increase in guns recovered from scenes.

Did they end the contract too soon? Maybe GDT isn’t much of a crime prevention tool on its own, the thinking went, but pair it with cameras that catch the aftermath on video, and surely detectives get what they need to close the case after the fact. Officers get notified fast, nearby footage catches the shooter fleeing, cases get solved. Or do they?

So, here’s the question this study set out to answer: forget crime prevention for a second. Did any of this technology actually help detectives close shooting cases? And since GDT and CCTV are frequently deployed in the same footprint, does having both do something that having just one doesn’t?

The Study

Eric Piza, Nathan Connealy, Rachael Arietti, George Mohler, and Jeremy Carter pulled every fatal shooting (6,068 of them) and nonfatal shooting (64,884) recorded in Chicago between 2005 and 2019, then built three separate treatment conditions to test: whether the incident happened inside an active GDT district after the sensors went live, whether it happened within 500 feet of a working camera, and whether it happened inside both a GDT zone and camera range at the same time (the “joint effect” condition).

Here’s what makes this study more impactful than a lot of what’s come before it: instead of eyeballing which neighborhoods “look similar” and calling it a comparison group (the fuzzy matching approach that’s dominated this line of research for years), the authors used a technique called entropy balancing. Think of it as building a control group by hand out of every incident that didn’t get the technology, then reweighting those cases so that, statistically, they line up with the treatment group on 18 to 20 different factors: prior shootings and arrests on that block, whether it happened at night or on a weekend, neighborhood poverty and demographics, population density, even the weather that day. Only a handful of prior CCTV or GDT studies have used this level of rigor, and most of those studies only covered 18 months or less of data. This one covers 15 years.

Once the matching was done, the researchers ran logistic regression models (essentially, “does having this technology change the odds a case gets cleared”) for each of the three treatment conditions, separately for fatal and nonfatal shootings. That’s six models total, all controlling for whether the incident happened in a district with an active SDSC.

What They Found on ShotSpotter and CCTV Case Clearance

Here’s what the numbers actually showed:

• GDT alone: null. Fatal shootings exposed to GDT were cleared in 25.9% of cases, compared to 39.2% for the matched control group. Nonfatal was similar (25.1% vs. 30.6%). Once the researchers ran the full model accounting for other differences between the two groups, that gap wasn’t statistically significant, meaning it could plausibly have shown up by chance rather than reflecting a real effect of GDT. Either way, it certainly wasn’t a point in GDT’s favor.
• CCTV alone: also null. For fatal shootings, camera-adjacent cases and their controls cleared at almost identical rates (37.5% vs. 36.9%). For nonfatal shootings, camera cases actually cleared slightly less often than controls. No significant effect either way.
• GDT and CCTV together: still null, for both fatal and nonfatal shootings, across all the joint models. There was a slightly higher odds ratio for fatal shootings when both were present (1.16), but the sample was small (only 211 treated cases), and the result could easily have landed anywhere in a wide range just by chance. The authors are upfront that this isn’t something to hang a policy decision on.
• SDSCs, included only as a control variable, quietly stole the show. Nonfatal shootings in a district with an active SDSC had 19% higher odds of clearance in the GDT model and 24% higher odds in the CCTV model. Chicago’s SDSCs, rolled out starting in 2017, pair the same camera and sensor feeds with dedicated staff, crime analysts, and a mandate to actually work the leads those technologies generate.

Why would the sensors and cameras come up empty while the staffed centers using those same feeds come up positive? The authors’ read, and it lines up with plenty of prior research on GDT and CCTV: the technology can tell you a shooting happened and roughly where. What it can’t do is process ballistics, canvass for witnesses, chase down the plate on the car recorded three blocks away, or keep a detective’s caseload from being buried under the next call. GDT has been shown repeatedly to increase gun and shell casing recoveries at scenes, and that’s genuinely useful physical evidence. But evidence sitting in an inventory bag isn’t the same as evidence that gets analyzed, matched, and turned into an arrest. A patrol officer responding to a GDT alert is still just a patrol officer responding to a call. Nothing about the sensor changes what happens after the officer clears the scene.

CCTV tells a similar story from prior research. Cameras have moved the needle on clearance in railway systems (Useful footage more than doubled clearance rates for metal theft on the British Transport Police’s network from roughly 20% to about 50%) and for lower-level offenses like theft and disorder (14% higher clearance at Milwaukee intersections, almost entirely driven by disorderly conduct and loitering cases, not violent crime). Gun violence is a different animal. It’s chaotic, often has few cooperative witnesses, and a camera 500 feet away catching a few seconds of a shooter fleeing doesn’t hand a detective a name.

What This Means For Your Agency

None of this means rip out your cameras. It means being honest about what a technology purchase is actually buying you.

If your agency has GDT, this is a good moment to take a hard look at it. Ask what happens after the alert fires. Is there a defined follow-up protocol, or does it depend on whoever happens to be free that shift? This study, echoing the Kansas City findings from largely the same research team, suggests the sensor by itself isn’t moving clearance rates. If the alert isn’t triggering a specific investigative response, that’s worth naming out loud in your next budget conversation, not glossing over.

Same goes for CCTV, with a caveat. A 2025 research synthesis of CCTV crime-prevention studies (also by Piza) found cameras produce a real, if modest, overall crime reduction, about 13% across the pooled evidence, but no significant effect on violent crime specifically. What separated the systems that worked from the ones that didn’t wasn’t the camera. It was what else was running alongside it: setups paired with active proactive policing saw real crime drops, while systems paired only with other passive technology, like license plate readers or gunshot detectors, didn’t move the needle at all, the same pattern this Chicago study just found with GDT and CCTV together. So don’t assume a camera system is pulling its weight on shootings just because it reduces crime somewhere in the city. If your system is justified on cutting property or drug crime, or it’s paired with active proactive policing, that’s on firmer ground. If it’s justified on “it’ll cut down shootings” on its own, both this study and the broader literature say pump the brakes. IAPSC EBSP-25-01 Video Surveillance (Piza)

Consider whether your agency has anything resembling Chicago’s SDSC model, even at a smaller scale. You don’t need a $10 million build-out to test the underlying idea: pairing existing technology (camera feeds, GDT alerts, license plate reader hits, whatever you’ve already paid for) with a dedicated person or small team whose job is to actually work those leads, rather than layering it onto patrol’s plate on top of everything else. A 2024 companion study on Chicago’s SDSCs found they improved clearance for overall violent and property crime. That’s not a coincidence. It’s staffing and workflow doing what the sensors alone couldn’t.

Here is a guide to a deeper look on Chicago’s SDSC:

Real-Time Crime Centers Can Help Reduce Crime

District-level strategic decision support centers used by Chicago police enabled novel responses to crime problems that were previously impossible. Crime reductions varied between 3 and 17 percent. However, there are risks to the long-term support of the centers.

Budget for the follow-up, not just the install. The authors note Chicago’s GDT system ran $8.8 to $12.3 million a year, while a comparable slice of the city’s camera network cost roughly $571,000 annually in maintenance. Prior research found the camera system returned $2.81 to $4 in crime reduction benefit for every dollar spent. Nobody has run that same math for case clearance from GDT, because there isn’t a positive number to run it on. If you’re the one signing off on next year’s technology budget, that comparison is worth having in your back pocket.

Final Thoughts

Chicago spent close to two decades and tens of millions of dollars betting that watching more and hearing more would translate into solving more. The watching part paid off, at least for preventing crime. The hearing part never quite delivered on its own terms. What this study actually demonstrates is narrower and, honestly, more useful than either takeaway: technology generates data, and data only becomes an investigation when someone with the time, training, and mandate picks it up and runs with it. Chicago’s SDSCs are the clearest evidence in this study that the fix was never really about buying the next sensor. It was about building the structure to use the ones you already had. Before your agency signs the next technology contract, ask the question this study keeps circling back to: who’s going to work the lead once it comes in, and do they actually have the time to do it?

Case clearance is only half the story. If you want the crime-prevention verdict too, we’ve already run the numbers:

CCTV and Crime: Insights from 11 Years of Research

A decade-long study shows how CCTV impacts crime over time, revealing when it works, when it fades, and why monitoring matters.

Gunshot Detection: What the Data Really Shows

Gunshot detection promises faster response and more evidence, but does it actually reduce gun violence? See what the data from Kansas City reveals.

Enjoyed this post?

Get more research translated into plain language delivered straight to your inbox.

Subscribe below and let me know which topics you want to hear more about. It only takes a minute.

If you found this helpful, share it with your team or on your socials. It helps more people make sense of research that matters.

Carlee Ruiz

Carlee Ruiz is the founder of PoliceResearchHub, a platform that turns academic research into practical insights for law enforcement. She brings seven years of experience as a crime analyst and justice researcher, with a focus on evidence based policing, gun violence prevention, and public safety strategy. Carlee currently works as a researcher at National Policing Institute, where she supports evaluations of law enforcement programs and criminal justice initiatives across the country. Through PoliceResearchHub, she aims to make research easier to understand and more useful for the people doing the work on the ground.