By German Lopez

Republicans in Congress want to make it easier for states to drug test the unemployed before they collect government benefits.

This is no longer just a wish. Over the past few weeks, the Republican-controlled House and Senate sent a bill to President Donald Trump’s desk—which he’s expected to sign—that would undo an Obama-era regulation that limited drug testing for federal unemployment insurance.

A 2012 law let states drug test and deny benefits to unemployment beneficiaries who work in fields that “regularly conduct drug testing.” But the Department of Labor under President Barack Obama took a narrow interpretation of the law, enacting a rule that made it so only people in fields in which employers randomly drug test after someone is hired can be drug tested for unemployment insurance. That effectively limited how many unemployment beneficiaries states could drug test—by blocking them from enforcing drug screenings for beneficiaries who work in occupations that only drug test as part of the job application process.

The new bill passed by Congress—almost entirely along partisan lines—overturns the Obama-era rule, giving Trump’s Department of Labor a chance to enact a broader regulation that allows for more drug testing.

But in doing so, the Department of Labor would get into a debate about just how far states should go in gating government benefits, if this is a practical policy to keep in place, or whether this kind of policy could actually lead to more drug use by removing one of the few financial safety nets the unemployed rely on to lead healthy lives.

What the latest bill does

House Speaker Paul Ryan trumpeted the bill over the weekend on Twitter, saying it “allows states to have drug testing to receive federal unemployment benefits.”

But this isn’t quite right. The bill only repeals a regulation enacted by the Obama administration. It’s now on the Trump administration to enact a new regulation that would actually let states drug test as a requirement to getting jobless benefits, which governments temporarily give (typically for up to half a year) after someone loses a job in order to provide some financial stability.

The Obama regulation took a narrow view of the 2012 Middle Class Tax Relief and Job Creation Act. The administration decided that when the law allowed drug testing for unemployed workers in fields that “regularly conduct drug testing,” that only applied to fields in which there’s random drug testing after someone is hired.

“Those kinds of jobs are typically situations where either federal or state law [says] you are subject to random testing,” said George Wentworth, senior counsel at the National Employment Law Project, an advocacy group focused on jobs. Wentworth cited, as one example, the transportation industry: “Pretty much everyone who’s driving some sort of vehicle of a certain size is going to be subject to drug testing protocols.”

Republicans in Congress, however, have argued for a much broader interpretation of the 2012 law—one that requires drug testing for jobless benefit applicants in any field in which an employer will do a drug test as part of the job application process. That would allow drug testing for unemployment benefits in many more occupations, Wentworth said.

But to actually put that interpretation into effect, the Trump administration now has to pass a regulation that does so, or Congress has to pass a bill clarifying the 2012 law to do what it wants. The latter would be particularly tricky with Democratic opposition—since, unlike a regulatory repeal that’s done through the Congressional Review Act, a bill would need to get 60 votes in the Senate to overcome the filibuster.

So it’s now likely on the Trump administration, should it sign the new bill into law, to follow through and enact the regulation that congressional Republicans want. Whether the Trump administration will do so remains up in the air.

The bill is meant to stop government money from going to drugs

Over the past few years, a growing number of states have moved to tie government benefits to drug tests. At least 15 states—by and large dominated by Republicans—have passed laws tying public assistance to drug testing or screening, according to the National Conference for State Legislatures. It’s a natural extension of those laws, as well as Republican-led efforts to limit and restrict the welfare state, to attach drug testing requirements to unemployment benefits.

The philosophical argument for these laws is simple: The government shouldn’t be in the business of helping people sustain a drug habit, so the government should make it difficult for people who receive government benefits to buy and use drugs.

Practically, the hope is this policy will do two things: It will save money by keeping drug users off government program rolls, and it will deter people from using drugs so they can obtain government benefits.

Tennessee’s legislature, for example, argued that “persons who are not under the pernicious influence of illegal drugs [are] less disruptive of the social fabric, persons and neighborhoods around them are safer as well” and that “tax dollars should go to persons who are trying to better themselves rather than to persons who violate our state and national laws and support a network of illicit purveyors of misery and disappointment.”

Just how states enforce these laws varies. Some do a screening based on suspicion of individual government benefit applicants, followed by an actual drug test if needed. Others may require a full drug screening for everyone. And a few—notably, Florida—have tried to require a full drug test for every single applicant, but these kinds of laws have so far been struck down in court.

Regardless of how states do it, it’s not at all clear that these laws actually accomplish what they seek to do.

When states have tried drug testing for government benefits, it hasn’t worked out well

Although there haven’t been many rigorous studies looking at how these laws play out, there has been some work analyzing whether they save money. That evidence suggests drug screenings for government benefits don’t save much, if any, money and may — perhaps surprisingly — actually cost states more money than doing nothing.

One investigation by Bryce Covert and Josh Israel for ThinkProgress summarized the aftermath after seven states attempted to tie welfare benefits to drug testing:

According to state data gathered by ThinkProgress, the seven states with existing programs   Arizona, Kansas, Mississippi, Missouri, Oklahoma, Tennessee, and Utah—are spending hundreds of thousands of dollars to ferret out very few drug users. The statistics show that applicants actually test positive at a lower rate than the drug use of the general population. The national drug use rate is 9.4 percent. In these states, however, the rate of positive drug tests to total welfare applicants ranges from 0.002 percent to 8.3 percent, but all except one have a rate below 1 percent. Meanwhile, they’ve collectively spent nearly $1 million on the effort, and millions more may have to be spent in coming years.

In short, because there seem to be so few drug users actually applying to government programs, the hundreds of thousands spent on drug screenings may cost more than the potential savings.

Other analyses have reached similar conclusions. The Center for Law and Social Policy (CLASP), an anti-poverty organization, regularly puts out a report that summarizes how these programs can lead to higher costs than savings.

Elizabeth Lower-Basch, director of income and work supports at CLASP, walked me through some of the math. She pointed out that the maximum grant through the Temporary Assistance for Needy Families (TANF) program for a family of three in 2016 was $432 a month in the median state, and that 62.9 percent of families on TANF remained on the program for 12 months or less.

“So even if you assume that the family is receiving the maximum benefits (e.g., no other income),” she told me, “and that the whole family is denied benefits due to a positive drug test (in theory the children are supposed to remain eligible, but it’s not clear that this really happens in most states), and that families would have stayed on for 12 months, you only save a maximum of $5,184 per positive test.”

Using Missouri as an example, here’s how that plays out in the real world: The state budgeted $336,297 in 2014 for its testing program. But if you assume that the 48 people who tested positive—out of the 38,970 screened applicants—were denied benefits and led to the savings outlined by Lower-Basch, that would produce less than $250,000 in savings. The state, then, is spending $80,000 more than it would if it hadn’t done any drug screening at all. (If anything, this is generous to Missouri’s drug screening law, since Missouri is below the median state for monthly TANF benefits, so the savings would likely be lower.)

Of course, this isn’t a rigorous scientific analysis, and some states may be able to carry out drug screening at cheaper rates than Missouri. But the fact that the budget math is close at all suggests that drug screening may not lead to the savings that supporters assumed it would.

Supporters, however, argue that just looking at the math in this way is deceptive, because it doesn’t consider the people who were deterred from applying in the first place or purposely avoided a drug screening after applying—and therefore didn’t get benefits—due to the drug testing requirements.

In Missouri, for example, more than 700 applicants in 2014 refused to complete a drug test and were denied benefits as a result. Using Lower-Basch’s math, that’s up to $3.6 million in savings for the state, which would be well over the $336,297 that the state budgeted for the program. But that presumes all of those 700-plus applicants declined the drug test because they thought they would fail, which we just don’t know.

Even if that’s true, there’s a case that drug testing is particularly unnecessary for unemployment benefits. Unemployment benefits are, essentially, insurance that’s supposed to tide someone over after losing a job as they look for new work. But if a person is looking for work, they might try to stay drug-free anyway, because many employers already test for illegal drugs as part of the job application process. So the screening requirement just adds another test on top of the one that an employer would probably already conduct, adding unnecessary costs to the taxpayers who now have to pay for the extra screening.

“It’s just redundant,” Wentworth of the National Employment Law Project said.

The mentality behind the bill fundamentally misunderstands drug use and addiction

Beyond the cost concerns, there’s a question about what this means for how the US approaches drug use in general: Since addiction is a disease, does it really make sense to punish addicts if they apply for unemployment benefits? Would someone even think to apply the same approach to other illnesses, like heart disease or diabetes?

Wentworth, however, said that states will sometimes use physical and mental conditions to deny unemployment benefits. For example, if someone is unable to walk but wants to work in construction, the state would deem that person unable to work and, therefore, possibly deny benefits on those grounds. (Although that person would likely be able to obtain disability benefits.)

Similarly, a person whose addiction led them to get caught getting high on the job and fired as a result, for instance, wouldn’t qualify for unemployment benefits today, drug test or not.

But by requiring a blanket screening, states would remove these kinds of individualized assessments to instead ban all drug users, including those who might otherwise function just fine in their day-to-day lives and work despite their substance use. It’s telling people they will be denied financial security just because they suffer from a specific disease. And it’s allowing the government to directly peer into people’s personal lives so they can get benefits they may be legally entitled to, further posing potential constitutional problems about privacy. (Wentworth, for one, described the idea of drug testing for unemployment insurance as “an unconstitutional solution in search of a problem.”)

States justify this by pointing to evidence that shows the unemployed have higher rates of drug use, which they argue shows the need for extra steps to deter drug use among the jobless. As a 2013 analysis from the Federal Reserve Bank of St. Louis found, “Illegal drug use was 18 percent for the unemployed, followed by 10 percent for part-time workers, 8 percent for full-time workers and less than 6 percent for those in the ‘other’ category, which includes retirees.”

But as the St. Louis Fed pointed out, this disparity might be a result of the exact problems that unemployment insurance is trying to alleviate. A 2011 review of the research found, for instance, that unemployment increases the risk of a drug use disorder and the risk of relapse after drug treatment. A 2013 study looking at alcohol misuse in particular similarly concluded that “job loss during the past year leads to a corresponding increase in average daily [alcohol] consumption, binge-drinking days, and alcohol abuse/dependence diagnoses, possibly due to factors such as mental strain, financial pressure, and shame.”

This makes sense: People are simply more likely to resort to drugs if they feel a great sense of despair. Unemployment can lead to that sense of despair, especially if it leads to significant financial uncertainty and instability. (Similarly, unemployment can lead to greater risk of other diseases impacted by behavioral or environmental factors, such as heart disease.) And without a job or income, it’s much harder—even impossible—to pay for treatment to stop using drugs.

Unemployment benefits are supposed to combat some of that despair by providing a measure of financial security. Without that security, though, someone is more likely to face economic hardship, and that economic hardship may, based on the studies above, increase the risk that someone will resort to illegal drugs.

In this way, state-mandated drug screenings could backfire: By eliminating people’s financial safety net and pushing them into even worse situations, they might become more likely to use drugs. The policy that’s meant to keep people drug-free, then, may in fact do the opposite.