Study Estimates Roadside Drug Tests Result in 30,000 Wrongful Arrests Every Year
Despite the well-known problems with the kits, they're used in half of the roughly 1.5 million drug arrests in this country every year.

Roughly 30,000 people every year may be getting wrongfully arrested and jailed because of police departments' widespread use of unreliable roadside field tests for drugs, according to a study released today by the Quattrone Center for the Fair Administration of Justice at the University of Pennsylvania.
The study, which bills itself as the first comprehensive analysis of the use of presumptive drug field test kits by law enforcement agencies in the country, estimates that each year approximately 773,000 drug-related arrests involve the use of such test kits. That's half of the roughly 1.5 million annual drug arrests in the U.S.
These inexpensive field tests use color reactions to indicate the presence of compounds found in certain drugs. However, the well-documented problem is that the compounds these kits test for are not exclusive to illicit drugs and are, in fact, found in dozens of legal substances. Over the years, officers have arrested and jailed innocent people after drug field kits returned presumptive positive results on bird poop, donut glaze, cotton candy, and sand from inside a stress ball. The Jacksonville Sheriff's Office in Florida stopped using test kits this September after discovering that several common over-the-counter cold medications returned false positives for cocaine. A 2018 investigation by a Georgia news station found that one brand of test kit produced 145 false positives in the state in one year.
But no one has ever tried to quantify exactly how many innocent people are jailed because of these tests, until now.
Although the true error rate of these kits is not known, the Quattrone Center estimates, based on the incomplete data it could glean from state drug labs and other sources, that as many as 30,000 innocent people a year may be wrongly arrested for drug possession based on their results, making these tests "one of the largest, if not the largest, known contributing factor to wrongful arrests and convictions in the United States."
"Presumptive field drug test kits are known to produce 'false positive' errors and were never designed or intended to provide conclusive evidence of the presence of drugs," Ross Miller, Quattrone Center assistant director and lead author of the report, said in a press release. "But in our criminal legal system, where plea bargaining is the norm and actual fact-finding by trial is exceedingly rare, these error-prone tests have become de facto determinants of guilt in a substantial share of criminal cases in the United States and, as a result, a significant cause of wrongful convictions."
The Quattrone Center report also found that police departments and local prosecutors' offices often had few policies in place to safeguard against wrongful arrests and convictions based on faulty field tests.
The center's survey of prosecutor offices found that 89 percent of them allow guilty pleas to be entered without confirmatory testing of test kit results, and nearly 70 percent of drug labs in the U.S. reported that they were not asked to review samples when a plea agreement has been reached.
The danger of innocent people confessing to drug possession—often under threat of more serious charges from prosecutors—is not hypothetical. A 2016 ProPublica/New York Times investigation found that 212 people pleaded guilty between January 2004 and June 2015 to drug possession based on Houston Police Department field tests that were later invalidated by crime labs.
Reason reported in 2021 on how these tests are also used extensively in prison systems across the country to punish inmates, despite clear warnings from the manufacturers that the results should be confirmed by outside labs.
In 2020, the New York Department of Corrections and Community Supervision suspended the use of NARK II test kits for contraband because of reliability concerns. A New York State Inspector General report released last November found that reliance on unverified test kits resulted in 2,000 innocent inmates being unfairly punished—including being thrown in solitary confinement, having visitation rights suspended, and parole hearings canceled.
In 2021, a Massachusetts judge ordered the state prison system to stop using similar field tests, finding that they were "highly unreliable" and "only marginally better than a coin-flip." That suit followed claims by over a dozen Massachusetts attorneys who said they were falsely accused of sending drugs to their incarcerated clients.
And last September, a class-action lawsuit was filed against the Washington State prison system for similarly disciplining inmates based on unverified test kit results.
Among the Quattrone Center's recommendations for reducing the risk of wrongful arrests and convictions are limiting the use of colorimetric tests, requiring confirmatory testing when a guilty plea is accepted, and requiring regular blind audits of cases involving test kits to establish reliable false-positive rates.
"Every year, tens of thousands of innocent Americans are arrested on the basis of $2.00 roadside drug test kits that are known to give false positives. Now, this landmark study by the Quattrone Center demonstrates the scope of the harm done by these inaccurate test kits, including the disproportionate impact to African Americans," Des Walsh, founder of the Roadside Drug Test Innocence Alliance, said in a press release. "Based on this study, we look forward to working with law enforcement and other interested parties to implement policies and adopt better testing techniques to substantially reduce the tragic number of innocent people arrested and convicted because of these false tests."
Editor's Note: As of February 29, 2024, commenting privileges on reason.com posts are limited to Reason Plus subscribers. Past commenters are grandfathered in for a temporary period. Subscribe here to preserve your ability to comment. Your Reason Plus subscription also gives you an ad-free version of reason.com, along with full access to the digital edition and archives of Reason magazine. We request that comments be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of reason.com or Reason Foundation. We reserve the right to delete any comment and ban commenters for any reason at any time. Comments may only be edited within 5 minutes of posting. Report abuses.
Please
to post comments
If it's used 750,000 times and fails 30,000 times, that's a failure rate of 4%. Or a success / accuracy rate of 96%.
This is cherry-picking from the above article, and misses most of the entire point! Your 30,000 failures only includes those exonerated by better tests or other evidence, and totally ignores people too stupid or too poor or too abused-by-Government-Almighty, to challenge these HIGHLY fallible results!
How about cherry-picking, instead, the below out-take from above...
"In 2021, a Massachusetts judge ordered the state prison system to stop using similar field tests, finding that they were "highly unreliable" and "only marginally better than a coin-flip.""
WHY do authoritarians LOVE-LOVE-LOVE all of this PUNISHMENT-PUNISHMENT-PUNISHMENT dished out by a Government Almighty run amok?
I get paid more than $120 to $130 every hour for working on the web. I found out about this Qs activity 3 months prior and subsequent to joining this I have earned effectively $15k from this without having internet working abilities Copy underneath site to..
Check It—>>> http://Www.Smartcareer1.com
Authoritarian types don't care about facts or actual guilt. These are the same people who cheer when a drug addict is gunned down by the police in the street, bypassing the entire justice system and giving police absolute power judge, jury and executioner.
No, it’s “used” 750,000 times, not “gets a positive result” 750,000 times. The error rate is the ratio of false positives to total positives, not to total uses. (Okay, a better error rate will consider both the false positives and the false negatives but that math is probably beyond a general-audience article.)
Fortunately, the article links directly to the detailed study where they do lay out their math (along with all the disclaimers and qualifiers). Their conclusion is that the failure rate is closer to 15% and could be as high as 38%.
The article does also cite to a judge who had access to a great deal more detailed information than is presented here and who concluded that the tests he was being asked to evaluate were “only marginally better than a coin-flip.”
Bottom line – do I know what the accuracy rate is? Not yet. But it’s definitely not 96%.
From the study:
I guess that would be one way to explain how the number of bad arrests could be so low while the percentage of bad tests is so high...
Could be but that's a single opinion offered without apparent validation. In my opinion, the more plausible explanation is in the discussion of the coercive effect of plea bargains.
You're correct about that not being the measure of the test's accuracy... but it *does* imply that the police are actually not that bad at mostly testing samples which do, in fact, contain drugs. Think about it: if the false positive rate is 15% to 38% but only 4% of arrests are bad, it means the police must be testing samples which *do* actually have drugs about 73% - 89% of the time.
This stance of yours totally ignores people too stupid or too poor or too abused-by-Government-Almighty, to challenge these HIGHLY fallible results! Like, people who are already in jail! Also, have you never heard of "plead guilty or else we will keep you in jail for up to 10 years"... Yes, this happens... "And we will PILE ON MORE CHARGES for your arrogance, in demanding a jury trial!"?
How much "justice" do you want? How much power, fame, glory, and especially... $$$MONEY$$$ do you have? None, you say? Then these piss-poor, SHITTY tests will be PLENTY to screw you with!!!
https://www.themarshallproject.org/2017/06/12/the-man-who-spent-35-years-in-prison-without-a-trial
The Man Who Spent 35 Years in Prison Without a Trial
The Jerry Hartfield case is an extraordinary tale of justice delayed and denied.
Where did you pull your "4% of arrests are bad" metric? That's not in either the article above or the linked study - at least, not that I can find.
If there are ~750K arrests using these tests and ~30K were bad arrests from bad tests, that means ~720K of them were correct, right?
Governments Almighty NEVER makes mistakes? Jesus Christ was killed for perfectly good reasons, Adolf Hitler took power for perfectly good reasons, and OJ Simpson got off for murder, for perfectly good reasons, right, right-wing wrong-nut?
Yeah, the police are decent at identifying drugs. Which is another reason why these tests are unnecessary (even if you think the WOD is a good idea).
This is why bird poop, donut glaze, cotton candy, and sand from inside a stress ball should be enough to convict you, with or without a drug test! TRUST in the police, just TRUST in them!
Since when? Cops think everything smells like weed. They profile the person and then claim to smell weed for an excuse to disassemble the car and run everything through the test kits until they get an excuse to lock some innocent guy up.
You don't even need to do drugs to wind up on the plea bargain express and find yourself pleading guilty to some BS lesser felony to avoid the "intent to distribute ketamine" charge dismissed. Then you get time served, some probation and you never get to own a gun again. All so some cops can claim a drug war victory.
Hell, COVID tests aren't 96% accurate.
4% failure rate is far too high. That doesn't pass the "reasonable doubt" test which is a pesky legal requirement that cops hate. You seem fine with ruining the lives of 30,000 innocent people a year, but those of us who aren't worshippers of today's out-of-control cops disagree.
Who could possibly have imagined that the initiation of deadly coercion based on mystical superstition and blind hatred could possibly result in anything untoward? I'm shocked, SHOCKED to see how faith-based asset-forfeiture looting persists 13 years after causing the Republican Crash and Great Recession of 2008-2010. (Psst, Congressman, your graft.)
On the other hand, eliminating these tests might not have the outcome you desire...
At least it will stop them testing random trash from the floor mat when they really want to arrest someone.
Of course you are right and police never needed these tests to arrest someone found with a baggie of white powder (or whatever) in their pocket.
Cops never have to try too hard to find reasons to arrest people. They don't even have to drop the baggie. They can just say they found it and get the lab geeks to write up a report that the substance was super meth laced with fentanol, then provide a small baggie with some powdered sugar to show the judge.
Follow the Junk Science.
Yeh, wonder if they had Aaron Brown look through this study and do an assessment of how good it is. I would assume likely not, because this study fits with their preferences.
officers have arrested and jailed innocent people after drug field kits returned presumptive positive results on bird poop, donut glaze, cotton candy, and sand from inside a stress ball.
Who says they are 'innocent'? Drug dealers who are selling bird poop, donut glaze, cotton candy, or stress ball sand are the worst sort of fraudsters. Not only are they deceiving the buyer, they are defaming the truly innocent who understand that one NEVER transports dangerous stuff like bird poop or cotton candy in a motor vehicle on a public road.
What. The. Literal. Fuck.
Roadside covid tests for those not wearing a mask.
"study" and "estimates" in the title?
Oh, wait, C.J. again.
This is the kind of article that I come to Reason to read.
I just wish that they would investigate breathalyzers the same way that they are investigating these kits.
So don't consent to the test and don't confess to the crime. Why is this so difficult for you junkies?
Yeah, it's terrible that people without the resources to fight effectively should take a plea rather than risk years in prison. Totally their fault, not the fault of a system that stacks the deck against them.
Seriously, fuck all the way off. It's easy to stand up for principal when it's not your ass on the line.
You don't have to take a plea. We're talking about people who haven't even been arrested for anything yet. What are you talking about?
I promise you are not going to get "years in prison" because a cop suspects drugs, and you refused him any roadside evidence to confirm it. Might lose your driver's license, but your chance of being convicted anything are nil. Because they don't have any evidence since you refused it to them.
They can't force these tests on you dude. Maybe if they see a bunch of weed or cocaine in your passenger seat; maybe if you just killed a minivan full of kids because you were clearly stoned out of your gourd - but that's clearly not what this article is about.
Roadside tests, even if 100% accurate, cannot possibly determine impairment. This is because the test readouts are simply positive or negative. Basic pharmacology tells us that the presence of a drug is irrelevant; it is the amount (concentration) of the drug compared to a known standard of impairment (as is the case with the Breathalyzer test). The test says nothing about the level of the drug (one opioid pill to treat a tooth extraction the day before will give the same answer as 10 pills consumed an hour before driving) or the time since it was taken. Even if it did there are no standards for what constitutes impairment with different drugs. The use of these tests to prosecute drivers is unfair and not scientifically valid.