Most Welfare Recipients Don’t Use Drugs, So Why Do States Keep Drug-Testing Them?

By Johnny Magdaleno
Vice News

On Friday, Michigan Governor Rick Snyder signed into law suspicion-based drug-testing for welfare recipients in three of the state’s counties in late 2015. It allocates half a million dollars to the Michigan Department of Human Services to carry out the pilot program, and marks the second time the Wolverine State has made moves in favor of punishing drug use by those receiving government assistance.

Michigan’s 1999 law was struck down because it required drug testing for any and all assistance recipients—setting a precedent that similar laws are likely to dissolve at the federal level. Coincidentally, on the same day that Michigan legislators approved the 2015 pilot program, the 11th US Circuit Court of Appeals killed a Florida law requiring drug testing for all government assistance recipients after it was challenged in a 2011 case led by the American Civil Liberties Union (ACLU).

Yet in spite of these pronounced cases of federal resistance, states are increasingly putting their faith in welfare drug-testing legislation. Michigan is the 12th state to have enshrined such a law since 2011, though of those 12, only Georgia and Oklahoma pursued laws with conditions as stringent as those in Florida.

The public, it seems, is equally faithful. According to polls from 2011 and 2013, 53 percent and 51 percent of Americans favored drug tests for all government assistance recipients, respectively. The same year Florida enacted its now-defunct law, 71 percent of those surveyed in-state supported it. And a 2011 poll in Michigan revealed 79 percent of residents supported the state’s new pilot program back then, before it had passed.

So most Americans believe states should drug test individuals who benefit from TANF (Temporary Assistance for Needy Families, commonly known as welfare), SNAP—aka food stamps—and other programs. But is there any evidence that this is a good idea?

While most data on this subject has only been collected within a few months to a year after the laws were signed into action, the majority of states with drug testing regimes haven’t achieved significant reductions in government aid spending. Nor have they substantiated the assumption that assistance recipients are more likely to use drugs, which is often touted as the primary reason for pursuing such laws.

The federal government looked into these tests a few years ago, when their popularity began to climb. In its 2011 study, the Office of the Assistant Secretary for Planning and Evaluation (ASPE), which operates beneath the US Department of Health and Human Services, reviewed drug testing policies for government assistance recipients in a number of states that had already enacted them. It weighed legislative cost estimates for each state, and also looked at the 82 welfare drug testing bills proposed in 31 different states in 2010 and the first half of 2011.

The results were remarkable because of what the study didn’t find. It didn’t find cost reductions, noting “none of the legislative costs estimates [identified] estimated net savings as a result of the proposed drug testing programs.” It also didn’t find any evidence that required drug testing will deter government assistance recipients from using drugs and thereby increase their employability, noting that a pilot edition of Florida’s drug-testing program showed “there was very little difference in employment and earnings between [government assistance recipients] who tested positive [for drugs] and those who tested negative.”

On the state level, statistics echo these findings. In its first month, a drug testing law in Tennessee that went into effect on July 1 this year resulted in only one positive test out of 812 assistance applicants. In Utah, results from mid-2013 showed that since the state’s suspicion-based drug testing law was enacted in August 2012, only 12 out of 4,730 people tested positive.

And two days after Michigan’s pilot program went up and Florida’s law went down earlier this month, Mississippi released data showing that out of 3,656 people who applied for government assistance since the state’s suspicion-based drug law went into effect on August 1, only two tested positive—or .06 percent.

In the larger debate over whether these laws, both mandatory and suspicion-based, are legal, Utah and Mississippi are key for a number of reasons. When Utah initially announced its findings, it trumpeted the law as a success. Only 12 people tested positive for drug use, lawmakers noted, but around 250 ended up not going through with the screening process at all, and as a result they forfeited their chance to receive government assistance. That means about 262 people were cut off, and the state saved over $350,000 within the law’s first year.

None of the other states with drug-testing laws in place had previously reported a number quite like $350,000, let alone any significant net savings at all. Utah emerged as a hallmark, and gave proponents for drug-testing laws a model they could brandish when discussing whether these initiatives were worth putting money into.

That’s what happened in Mississippi. Nearly a year after Utah made public its results at the end of 2013, Mississippi enacted a suspicion-based drug testing law modeled after Utah’s, and one thing in particular they took from the western state was its preferred drug-screening questionnaire, the SASSI (Substance Abuse Subtle Screening Inventory). It’s a written test—mostly true or false questions—that allegedly determines if someone is likely to be a substance abuser. If determined as such, this gives the authorities “suspicion,” and with that suspicion, the state is constitutionally authorized to ask that someone to piss in a cup.

SASSI testing kits are a product sold by a company that operates beneath the same name. It was developed at the SASSI Institute in Springfield, Indiana in 1985, but has gone under three revisions since then—most recently in 1997. It is commonly used in hospitals and drug treatment programs, as well as prison and jail intake programs. But because government assistance drug testing laws are relatively new, SASSI’s accuracy in relation to them has yet to undergo thorough study.

Regardless, SASSI’s website claims that SASSI-3, its flagship exam, has an incredible rate of prediction. According to a SASSI Institute-led study on SASSI, entitled “Estimates of the Reliability and Criterion Validity of the Adult SASSI-3,” the test is able to determine “substance-abusing and substance-dependent respondents from those without a substance abuse disorder” with 94 percent accuracy.

Leaving aside the scientifically dubious fact that this study was conducted in part by SASSI developer Franklin Glenn Miller, the data coming out of Utah and Mississippi raised serious questions about this claim.

From August 2012 to July 2013, Utah issued SASSI to nearly 5,000 government assistance applicants. Of those, 466 were identified by the test as having a high likelihood of being a drug user. Yet of the 466, as mentioned above, only 12 tested positive for drugs. SASSI had 2.5 percent accuracy in this case.

Four months into Mississippi’s new law, there were 3,656 applications for TANF. Of those, 38 were deemed likely substance abusers by their SASSI test results. And of those, also mentioned above, only 2 tested positive. That’s 5 percent accuracy in this case.

(I called SASSI to speak with a representative about these numbers. Maybe I was missing something; maybe I had to compare data in a certain, SASSI-certififed way. But as of publishing, I’ve yet to receive a response.)

Jason Williamson, a staff attorney for the ACLU, spoke against SASSI earlier this year during a public hearing held by Mississippi’s Department of Human Services. This was before the state enacted its drug testing law, when it was considering whether or not it should nominate SASSI as its go-to screening technique.

His speech outlined three separate studies by researchers from 2002, 2009 and 2011. Each urges caution towards professional reliance on SASSI.

Because SASSI has proven ineffective, and because evidence overwhelmingly shows that hardly any welfare recipients use drugs—and because Florida’s mandatory drug-testing law was just struck down at the federal level—I wondered if suspicion-based drug testing laws now more likely to get struck down, too.

“The short answer is no,” Williamson told me. Due to the America’s major welfare reform act signed by Bill Clinton in 1996, states have the right to drug test beneficiaries of government programs as long as the state can prove suspicion. “But the more relevant question is: Have the states figured out a reliable way to identify drug users? This issue is tantamount to the suspicion-based drug testing programs.”

Williamson also sent me the text for a drug-testing bill Georgia introduced last year. Instead of relying solely on a questionnaire, the document says, case workers could suggest applicants for drug testing based on, though not exclusively, their “demeanor,” or “missed appointments.”

“Things that can obviously be explained by an infinite number of reasons,” Williamson says. (The bill was suspended earlier this year when the US Department of Agriculture ruled that drug testing SNAP recipients—one of the law’s provisions—was illegal because food stamps are a federal program, outside the jurisidiction of individual states)

This brought me back to Utah’s excitement over the money saved when 250 people forfeited their assistance because they never showed up to take a screening exam. I called Utah’s Department of Workforce Services, which administers both SASSI tests and government aid, to ask about this number. They said the state had not looked into why these people missed their appointments.

Were they all drug users, fearful of getting called out? Likely not, as all these statistics show. And getting a “positive” in Utah doesn’t guarantee that your benefits get cut off, though it does require that you begin a drug treatment program.

“The applicants have to pay upfront for the tests themselves,” Williamson said, speculating why assistance applicants might not make it to their prescreening session. “There could also be transportation issues, because tests are administered by different authorized facilities. Or maybe these people just got a job and no longer needed the assistance.”

A savings of $350,000 is not insignificant, and it’s understandable why states flirting with drug-testing laws would turn towards Utah while blueprinting their own. But since that amount was produced thanks to unknown, un-researched factors that led 250 individuals to not show up on testing day, suffice it to say there’s no guarantee that following Utah’s lead will produce comparable results.

And if states digest this fact and still fail to consider the way these laws help encourage unfounded stereotypes of people on welfare, it’s fair to wonder if lawmakers are just using Americans in need as a disposable chip to satisfy the assumptions of their constituents.