Last reviewed: September 19, 2026
The honest answer is:
Sometimes, for some purposes.
That may be less satisfying than either side’s easiest slogan.
Automated license plate readers can scan vastly more plates than officers can check manually. They can generate stolen-vehicle alerts, help locate vehicles and provide investigators with useful historical observations.
There is also newer evidence that fixed Flock networks may reduce motor-vehicle theft in some jurisdictions.
But those findings do not establish that ALPR networks broadly reduce crime everywhere, that every deployment is cost-effective, or that every claimed investigative improvement was caused by the cameras.
And even strong evidence of effectiveness would not, by itself, answer whether mass vehicle-location collection is justified.
The evidence answers three different questions.
Case evidence: ALPR observations have helped locate wanted or stolen vehicles and develop leads in individual investigations.
Outcome research: Experimental and observational studies ask whether deployments changed theft, recovery, clearance or other measurable outcomes across many cases.
What remains uncertain: A successful case does not prove broad crime reduction, and a national study does not establish that a particular Metro Detroit deployment produced the same result locally.
Keeping those categories separate lets residents acknowledge genuine benefits without converting every success story into proof that the whole network is necessary.
Efficiency is not the same thing as effectiveness.
This distinction matters.
An ALPR is clearly more efficient at checking plates than an officer entering them manually.
That does not automatically answer:
- Does crime fall?
- Are more cases solved?
- Are stolen vehicles recovered materially faster overall?
- Is the local system worth what it costs?
- Would narrower investigative methods produce enough of the same benefit?
- Does the benefit justify routinely collecting searchable location information about ordinary traffic, including people not suspected of wrongdoing?
Those are different questions.
Early randomized research found clear operational benefits.
A National Institute of Justice-sponsored randomized experiment in Mesa, Arizona found that LPR-equipped officers checked roughly eight to ten times as many plates as manual checking, generated substantially more stolen-vehicle hits and recovered about twice as many stolen vehicles.
That is a real benefit.
But the study did not find evidence that LPR patrol itself reduced auto theft. In one phase, specialized officers performing manual plate checks showed a short-term auto-theft reduction that the LPR intervention did not.
The useful conclusion is not that the technology failed.
It is that:
finding more stolen vehicles and preventing more vehicle theft are different outcomes.
Other experimental research found little broader deterrent effect.
Another NIJ-supported study compared police hot-spot patrols with and without LPR equipment.
LPR use increased stolen-vehicle recoveries, but did not increase arrests and did not measurably reduce the likelihood, timing or seriousness of subsequent crime or disorder calls in the tested locations.
The researchers described the evidence for general patrol crime-prevention effects as limited.
That is another reason to distinguish:
the tool can find useful things
from:
the tool broadly prevents crime.
A 2025 evidence review found the research base surprisingly thin.
In December 2025, researchers Cynthia Lum, Christopher Koper, Hyunji Lee, Daniel Nagin and Lawrence Sherman published an open-access review focused on ALPR effectiveness and cost-effectiveness.
They concluded that ALPRs are plainly more efficient than manual plate checking, while the broader evidence of effectiveness remained much less settled.
They also noted that only a small number of methodologically rigorous studies had evaluated intended crime-control outcomes.
In a sample of investigations where ALPR use was mentioned, the review reports that the technology did not contribute meaningfully to case resolution in almost three-quarters of the sampled events.
That does not mean ALPR never matters.
It means buying the technology and achieving meaningful public-safety outcomes are not automatically the same thing.
Then came an important 2026 study of fixed Flock networks.
In August 2026, criminologists Scott M. Mourtgos and Ian T. Adams released a working paper examining fixed Flock deployments across 216 U.S. law-enforcement agencies using Flock deployment records and federal crime-reporting data.
The study estimated that motor-vehicle theft fell by 11% after deployment in its primary model.
It also estimated an increase in motor-vehicle-theft arrest clearance.
Those are meaningful findings in favor of the technology and deserve to be taken seriously.
The 11% result does not mean every city gets an 11% reduction.
The study’s own analysis shows that the result is not best understood as a uniform effect across every adopting agency.
The treated agencies were generally larger and carried substantially more motor-vehicle theft than the comparison agencies. The authors’ preferred weighting gives greater influence to jurisdictions with more pre-deployment theft.
Alternative ways of weighting agencies produced much weaker or non-significant estimates.
So the responsible summary is:
Fixed Flock deployment was associated with a meaningful reduction in vehicle theft in the study’s primary analysis, with the decline concentrated in jurisdictions carrying more of the theft burden.
That is more precise than:
Flock reduces vehicle theft 11% everywhere.
The clearance finding needs context too.
The working paper estimated roughly a 15.9% increase in motor-vehicle-theft arrest clearance after deployment.
But the authors also identify movement in the clearance measure before deployment and conduct sensitivity analysis around those pre-trends.
That makes the clearance result less straightforward than a simple headline percentage suggests.
The finding is worth considering.
It should not be presented without the caveat.
Faster recovery was a much smaller result.
Among vehicles recorded as recovered, the 2026 study found the median recovery lag was about 0.28 days shorter, roughly seven hours.
But measures that kept all stolen vehicles in the analysis did not show a measurable overall improvement in recovery timing.
The authors therefore describe the recovery evidence as a modest change within the recovered subset, not a population-wide acceleration in recovery.
That is another good example of why reading beyond a headline matters.
The 2026 study is a working paper.
That does not make it invalid.
Working papers are a normal part of academic research and allow methods and findings to be examined before or during journal review.
But its status should be described accurately.
It is also observational rather than a randomized trial. The authors discuss limitations including non-random adoption, possible use of other ALPR systems in comparison agencies and spillover effects from data sharing.
Those limitations do not erase the findings.
They tell us how cautiously to generalize them.
Vendor case studies answer a different question.
Flock and police departments publish many examples where ALPR information contributed to:
- locating stolen vehicles;
- developing suspects;
- finding wanted vehicles;
- recovering property;
- or advancing serious investigations.
Those examples can demonstrate that the system is useful in particular cases.
They do not necessarily tell us:
- how often it is decisive;
- how often searches produce nothing useful;
- what would have happened without the technology;
- whether another method would have found the same lead;
- or whether crime fell because the cameras were installed.
Anecdotes and causal research answer different questions.
“Assisted” can mean a lot of things.
Watch the language used in effectiveness claims.
A technology may have:
assisted
supported
contributed to
generated a lead in
or:
helped solve
an investigation.
Sometimes it was the decisive lead.
Sometimes it was one piece among witnesses, security footage, phone records, physical evidence, public tips and ordinary detective work.
When evaluating renewal, officials should ask how substantial the contribution actually was.
Local results matter.
A national study tells us what may happen across many agencies.
It does not tell us whether a specific Metro Detroit deployment is producing enough benefit to justify its cost and surveillance footprint.
Local officials can ask:
- How many searches were performed?
- How many useful alerts resulted?
- How many stolen vehicles were recovered?
- How many investigations were materially advanced?
- How many alerts were wrong or required correction?
- What did the program cost?
- What problem was it originally purchased to solve?
- Did that local outcome actually improve?
Those are normal oversight questions.
Do not confuse correlation with causation.
If crime falls after cameras are installed, the cameras may have helped.
Other things may have changed too:
- police staffing;
- vehicle security technology;
- prosecution practices;
- regional crime trends;
- major arrests;
- insurance or manufacturer interventions;
- or other policing initiatives.
That is why rigorous studies use comparison groups and statistical methods rather than simply placing a before-and-after number on a slide.
The reverse is true too.
If crime rises after cameras are installed, that alone does not prove the cameras failed.
Effectiveness is not the same thing as justification.
This is the most important distinction on the page.
Suppose future research confirms that fixed ALPR networks reduce vehicle theft substantially in certain communities.
That strengthens the argument for using them.
It still leaves other questions:
- How much location information is collected about uninvolved people?
- How broadly can it be searched and shared?
- What mistakes occur?
- What misuse is possible?
- What does the system cost?
- What happens as the software becomes more capable?
- Can narrower tools achieve enough of the same benefit?
- Is the surveillance power proportionate to the public-safety gain?
Useful technology can still be bad policy.
Society makes that kind of tradeoff all the time.
Our argument does not depend on claiming the cameras are useless.
If our case depended on:
Flock never helps solve anything.
then one strong study or one serious case would undermine the whole campaign.
That is not our position.
A fair summary of the evidence is:
ALPRs are highly efficient at scanning plates.
They can improve stolen-vehicle detection and produce useful investigative leads.
Older rigorous studies found limited evidence for broader crime deterrence in the deployments they tested.
A 2025 evidence review found the rigorous research base remained small and broader effectiveness difficult to establish.
A 2026 multi-agency working paper provides meaningful evidence that fixed Flock networks may reduce motor-vehicle theft, particularly where theft burden is substantial.
And:
None of those findings answers the privacy and mass-surveillance question by itself.
Our conclusion remains the same.
We want serious crimes investigated.
We want stolen vehicles recovered.
We want evidence-based public safety.
We also believe government should not routinely build searchable movement records on entire communities to achieve those goals.
The technology can be useful.
We do not believe the demonstrated benefit justifies the mass-tracking infrastructure.
Public Safety Without Mass Tracking →
For Police & Public-Safety Professionals →