RALEIGH – When during a recent speech House Speaker Thom Tillis endorsed the idea of drug testing for North Carolina welfare recipients, he set off a raging controversy that shows no signs of abating anytime soon.

The Mecklenburg Republican’s most-controversial suggestion wasn’t really about welfare families, actually, but about state employees. In response to an audience question, Tillis opened the door to random drug testing for many or all state employees. He should immediately close that door – the idea is likely to be neither cost-effective nor consistent with privacy concerns.

More deserving of consideration is the prospect of North Carolina emulating Florida and other states that already impose drug tests for welfare. The federal courts have upheld the legitimacy of other state conditions, and drug testing is explicitly authorized by federal law. But a decade ago, a broad drug-test policy in Michigan was struck down.

Florida’s policy – which requires new applicants for cash welfare to take a state-approved drug test and then reimburses recipients who test negative – is currently being challenged by the American Civil Liberties Union.

To say that North Carolina may have legal sanction to test welfare recipients for drug use is not necessarily to say the state should impose a drug test. Given the existence of Temporary Assistance for Needy Families, the main cash-welfare program, it is in the interest of taxpayers that the program be administered in the way most likely to minimize both dependency and cost.

All other things being equal, it is folly for the government to give cash assistance to those who regularly use mind-altering drugs. Whatever you think of the wisdom and efficacy of drug prohibition, surely you don’t think it wise or efficacious to force taxpayers to finance the drug habits of welfare recipients.

But for a drug-testing rule to make sense, the benefits must exceed the costs. I’m inclined to think that it might, based on Florida’s early experience, but I also think the debate is plagued by misstatements.

Liberal critics argue that Florida’s policy is a waste of time and money because very few would-be welfare recipients have failed their drug tests. That’s not the relevant statistic. If you lived in Florida, used drugs regularly, and learned that a drug test was required to apply for cash welfare, would you submit your application anyway and hope to get lucky? No. You might try to game the system by submitting results from a fake or disreputable lab. Or you wouldn’t apply in the first place.

Tarren Bragdon, head of Florida’s Foundation for Government Accountability, has run the numbers for the first quarter of testing. While only 32 applicants were denied benefits because they tested positive for drugs, many others were denied benefits because they tried to submit results from an unapproved lab. Indeed, about a fifth of Florida applicants were denied cash welfare during the first quarter because of some violation of the drug-testing rule.

More importantly, there was a very large decrease in the number of first-time welfare recipients in Florida during the first quarter. In September alone, the number of recipients in the relevant category came in 62 percent lower than the number of recipients Florida enrolled in September 2010. First-time recipients also dropped for the quarter as a whole, though by a smaller percentage.

Because participation in other welfare programs not subject to the testing requirement, such as Food Stamps and Medicaid, did not substantially change, it is likely that the drop in cash-welfare caseloads was due primarily to the prospect of drug testing, not to some unrelated trend.

Now, before fiscal conservatives get their hopes up too much, don’t expect such a requirement to save the state a great deal of money. If you assume Florida’s early experience will persist through the fiscal year, and further assume that North Carolina’s experience would be similar, the net savings (after subtracting the cost of the drug tests) would be around $5 million.

I think we should wait, gather more data from Florida and other states, and then make a rational calculation. Then again, I’m just nerdy and naïve enough to think rational calculation and public policy go together.

Hood is president of the John Locke Foundation.