The Baltimore Crime Blog reported that “a Department of Justice survey of multiple studies has concluded that license plate readers have no appreciable impact on crime rates.”

A check of The Office of Justice Programs’ CrimeSolutions.gov showed a rating for “License Plate Recognition Technology” of “No Effects – More than one study.” According to the site, this rating is used where “Programs or practices have strong evidence indicating that they had no effects or had harmful effects.”

In a middle-school book-report fed to policy makers for consumption, the people at CrimeSolutions.gov regurgitated various facts mentioned in the 2 studies they relied on to make that sweeping conclusion, but didn’t actually conduct any analysis to see if what they were reading actually stood for the propositions they were citing. (Ok, there were some big words in there… it was more like a poorly executed high-school book report)

The review relied entirely on 2 studies, both of which confirmed that LPRs are effective in identifying stolen vehicles and stolen registration plates, with collateral benefits related to identification of wanted persons and persons on various watch-lists (gang, terrorist, sex-offender).

Neither study demonstrated an impact on the crime RATE, but given the limited scope of the studies and the inherently mobile nature of auto-theft, this shouldn’t have been an expected outcome.

What would deter would-be criminals from stealing cars when study 1 consisted of 4 officers tasked with traveling through 30 areas in 2 cities (Alexandria and Fairfax) over the course of 30 days? The size of the trial was woefully inadequate to test deterrence, there was no public outreach or education on what was happening, there was only limited intervention (no word on whether apprehended car thieves struck again after being caught by the LPRs), and… IT WAS CONDUCTED RIGHT AFTER SNOWMAGEDDON 2010!

A “scientific” auto-theft experiment wouldn’t be affected by a record-shattering blizzard, would it? So no need to account for impact or postpone the study… or do anything other than drop a footnote that the largest snowstorm in recent memory struck the target area immediately prior to beginning the experiment. So little attention was paid to the impact of the blizzard that I’m surprised study 1 didn’t conclude that “deployment of LPRs results in 15-20 foot snowpiles in area parking lots.”

And the second study, while conducted in a more comprehensive manner, employed only 4 LPR vehicles over nearly 60 miles of road in Arizona.

License-Plate Readers (LPRs) are basically cameras attached to either vehicles or fixed positions that scan license plates (tags) and run the plates through a system to determine if the vehicle is stolen, registered owner is suspended or has a warrant, etc. When that data is stored, it can also be used to track travel of a particular vehicle based on where/when its tag was scanned.

CrimeSolutions.gov self-reports that it “uses rigorous research to inform practitioners and policy makers about what works in criminal justice, juvenile justice, and crime victim services.”

So what “strong evidence” did CrimeSolutions.gov and their “rigorous research” use to “inform practitioners and policy makers about what works in criminal justice”?

Why should the mayors and police commissioners of the nation’s cities abandon the LPR program?

Because of two studies:
– The first study consisted ENTIRELY of 4 officers in 2 different cities (for those not mathematically inclined, that’s 2 officers per entire city) that rode around with LPRs in 30 different areas scattered across the cities “for 30 consecutive days when their schedules permitted.” Based on this 4 officers/2 cities/30 areas/30 days trial, they declared that deployment of LPRs “in autotheft hot spots does not appear to result in a reduction of crime generally or autotheft specifically.” (For purposes of this post, let’s pretend that “autotheft” is a word)

Other issues with this “trial”? Well, 1 of the officers never shut off the LPR and so average data for the city of Alexandria was “calculated using only one officer’s reported numbers.” Oh yeah, AND IT WAS IN THE MIDDLE OF A BLIZZARD.

CrimeSolutions.gov somehow failed to reveal this small fact, despite their “rigorous research,” and GMU for some reason decided to continue with the study anyway (noting the blizzard in a footnote, because if there’s anything that has an insignificant relationship with auto-theft, it’s a “SNOWMAGEDDON!”). (As if that wasn’t enough of a joke, the study “was tasked by SPAWAR,” which may be coming to The History Channel soon to premier after Pawn Stars)

But even this ridiculous study, in the aftermath of a ridiculous blizzard that staggered the East Coast and left hundreds of thousands without power, showed that the LPRs were STILL EFFECTIVE. Study 1 reported for Fairfax: three stolen vehicles found, one “lost vehicle,” (no idea) and one set of stolen plates along with 14 “accepted” hits for terrorist/gang/sex-offender watch-lists. And for Alexandria: 4 stolen vehicles, 2 stolen tags, 4 already-recovered/located vehicles, and 4 terrorist watch-list hits. But Study 2 reminded the reader “not to focus on those successes because “this experiment focuses on measuring the impact on LPR’s ability to deter crime” (which may be like ignoring the impressive cleaning power of Clorox because it has no deterrent effect on the mess your husband makes when he cooks chili).

Ok, well the other study must have torn LPRs apart, right?

– The second study was significantly more comprehensive, was conducted over nearly a year, and… oh yeah… “Experimental results showed that LPR use considerably enhanced the productivity of the auto theft unit in checking license plates, detecting stolen vehicles and plates, apprehending auto thieves, and recovering stolen vehicles.”

“What,” you say? “That completely contradicts the conclusion of CrimeSolutions.gov!” Yes. Yes it does. And that 48-week study wasn’t conducted in a blizzard.

For some reason, CrimeSolutions.gov downplayed the improved performance of the LPR in both studies. The 2nd study noted that “[c]ombining results across both phases, the use of LPRs resulted in 8 to 10 times more plates checked, nearly 3 times as many “hits” for stolen vehicles, and twice as many vehicle recoveries. Further, all hits for stolen plates, all arrests for stolen vehicles or plates, and all recoveries of occupied vehicles were attributable to use of the LPRs.”

Unfortunately, CrimeSolutions.gov seems to have spent all of their energy “rigorously” researching and were all tuckered out when it came time to conduct an actual analysis of the studies, resulting in a report to “practitioners and policy makers” that LPRs have “strong evidence indicating that they had no effects or had harmful effects” in the face of evidence to the contrary.

Nothing in either study suggested that LPRs were a panacea, but maybe a fed shutdown would give the folks over at CrimeSolutions.gov a chance to actually read and think about the studies they’re analyzing before they go handing out ill-considered advice.

Is “no advice is better than bad advice” a saying?