Studies Show Flock’s ALPRs Reduce Crime… So Long As Flock Controls The Inputs And The Methodology

Overhype

from the our-stats-say-otherwise dept

404 Media, the new news outlet formed by excellent journalists kicked to the curb by the successive brunchlord regimes overseeing Vice Media, continues to stake its claim as the best investigative media outlet in the country.

Their latest effort digs deep into dubious claims made by up-and-coming surveillance tech company, Flock. Flock made its entry into the market by pitching its automatic license plate readers to the biggest assholes in the private sector: homeowners associations and gated communities.

Pitching to private individuals and firms made it easy to dodge constitutional concerns. And it made it very easy for their customers to regulate traffic in their neighborhoods by engaging in always-on surveillance of everyone traveling “their” streets.

Flock then moved on to the next biggest set of assholes in the nation: law enforcement. Citing its success in the private surveillance market, Flock started partnering with cops shops to place ALPR cameras wherever law enforcement felt they might be useful. Generally speaking, this worked out the way it has for any surveillance tech product pitched to cops: neighborhoods with large minority presences were the ones immediately blanketed by this tech.

Flock is a business. Therefore, it makes sense it might play a little fast and loose with facts to secure sales. But once you become a preferred government contractor, you’re expected to adhere to the truth a bit better because it’s the public’s money that’s being spent.

No problem, said Flock. Here’s a study that says our ALPRs are helping drive crime down.

• In a typical agency, one additional Flock Safety License Plate Recognition (LPR) camera per sworn officer correlates with a 9.1% increase in clearance rate.
• 20 additional Flock customers within 50 kilometers of the original agency leads to a 1% increase in clearance rates.
• Broad access to Flock technology within an agency leads to improved case clearance outcomes. 

That’s what Flock says its tech can do. To buttress its claims, Flock claims the study is based on input from “independent criminology research experts” from Texas Christian University (TCU) and the Tyler campus of the University of Texas.

Sounds great. But it isn’t. As Jason Koebler’s extensive debunking for 404 Media notes, Flock’s claims were immediately criticized by others in the criminology field, which apparently only extends Flock’s streak of dubious law enforcement effectiveness claims.

Flock’s most recent research has been widely criticized by academics in reporting by Forbes, and the validity of its case studies more broadly has been questioned over the years. 

The most outrageous claim is that Flock’s ALPRs are “instrumental in solving 10 percent of reported crime in America.” Whoa if true but it definitely isn’t. And Flock’s willingness to massage stats and reject findings it doesn’t care for has resulted in TCU’s Johnny Nhan — one of the TCU researchers involved in the latest Flock study — to publicly state he would have handled things very differently if he had known how Flock was abusing his research and preventing him from finding facts that didn’t support Flock’s marketing narrative.

Communications between Flock, researchers, and law enforcement agencies obtained by 404 Media — along with Nhan’s own statements — make it clear Flock chose to go to press as soon as it had the data it wanted, even if the data did not meaningfully depict Flock’s contribution to public safety.

In an email exchange with 404 Media, Nhan said that he and Helfers were brought into the study late in the process, that he has concerns with how the research was framed, and that he and Helfers are working on future research with Flock that is more qualitative in nature and will focus on case studies rather than quantitative analysis.

“Dr. Helfers and I are working on a paper that is still in the development stages and is still evolving as we’re looking at what data is available to us,” Nhan said. “As of right now, the plan is a peer-reviewed paper that looks at the uses of flock ALPR technologies, how agencies are ensuring privacy, and what policies they have in place for that. The information that is collected by the police departments are too varied and incomplete for us to do any type of meaningful statistical analysis on them so we’re pivoting to interviews with a sample of agencies.”

Nhan was told by Flock that the data collected so far was only meant to be a “starting point” for further research. But it liked what it saw enough to push forward with press releases selectively quoting Nhan and his research — claims that included its supposed “ten percent” contribution to the nation’s case clearance rates.

404 Media’s Koebler didn’t find any of this believable and went searching for background info. Flock is a private company but it’s in contact with plenty of public entities, including the two universities whose services it engaged to give its sales pitch a science-y backdrop. Most of these requests were denied, mainly due to Flock’s interference in the public records process.

But Koebler did come across some interesting communications, which suggest even researchers have been compromised by Flock’s insistence the research its funding deliver the results it’s seeking.

[T]he records I did get back show that Flock has been recommending which specific police departments Nhan and Helfers should talk to for a future research paper, and an email Nhan sent to one police department states that he would like to see data that shows a “big swing” in the data from before Flock’s adoption to after Flock’s adoption. 

While it’s great to see Nhan distancing himself from this study, it also shows researchers feel at least some pressure to deliver the results Flock clearly wants. Other emails show Flock — via Andrea Korb (Flock’s director of policy) — informing the researchers that the company will provide a list of agencies they can approach for data. That included Flock steering researchers towards small towns with already-low crime rates, where any reduction would produce the “big swing” so clearly sought by the company for its “research.”

The entire article is worth a read. The communications Koebler managed to secure show a company willing to toss a lab coat on top of sensationalist stats gleaned from its preferred law enforcement agencies in order to sell its products to other law enforcement agencies. It also shows the company is willing to place itself between researchers and their research, as well as between public entities and the public by crafting contracts that give it the “right” to redact information and withhold documents requested by records requesters.

If Flock’s tech was really that shit hot, it wouldn’t need to do this. But it’s apparently no better than anything else on the market. Not that Flock is the first — or only — surveillance tech firm to engage in this sort of chicanery. Just because you read it at a site created for the sole purpose of reprinting press releases doesn’t mean it’s true. If only law enforcement agencies were willing to perform the sort of due diligence being done by this very small journalistic outfit before spending public money, we might all be a whole lot better off.

Filed Under: alpr, crime, license plate readers, studies
Companies: flock