Being intoxicated at work is a bad idea. So is showing up for work unfit to work for other reasons: for example, being under-slept. The fixation on the “drug-free workplace” has less to do with safety than with the culture wars; otherwise, more firms would test for alcohol, especially after lunch.
If impairment at work were the issue, firms would be better advised to use impairment tests rather than testing employees’ urine, because urine testing tells you whether a person used a given drug within the past 72 hours (longer for cannabis) and not whether the person is currently under the influence. (The cheek-swab test, which can also cover alcohol, is more specific to recent use, but that’s not the currently approved technology for workplace use.)
With the non-medical use of prescription drugs a rising problem, some companies have begun to test – and fire – workers for using such drugs (notably opioids such as oxycodone and hydrocodone and the benzodiazepine anti-anxiety drugs such as Valium and Xanax). At at least one company, seemingly run by mental and moral defectives, workers were fired, without warning, for using drugs according to medical direction.
Of course, someone can be just as zonked on a prescription drug as on a non-prescription drug, but again, the companies aren’t measuring intoxication, merely recent drug use. When it came to the illicit drugs, the excuse for the invasion of privacy involved in having an employer meddle in employees’ off-the-job activities was that taking those drugs involved breaking the law. But extending the rule to prescription drugs gives employers a convenient way of getting rid of employees with chronic medical conditions.
What’s shocking is the thinness of the evidentiary basis underlying the whole exercise. None of the studies purporting to show that testing positive predicts poor work performance has anything like adequate statistical controls.
There’s a simple way to find out whether workplace drug-testing is actually relevant to safety and work performance. What the employer gets back from the lab is a “positive” (drug metabolites present) or a “negative” (metabolites absent). But the testing equipment actually produces a quantitative measurement, converted to “positive” or “negative” by using some fairly arbitrary cut-off value. Test just above that value, you’re fired; test just below it and your employer never even knows.
So here’s the study I’d like to see someone do: the military, for example, which has an aggressive drug-testing policy. Draw a sample of people who tested “negative” and get the quantitative scores. If off-the-job drug use is really a risk factor for accidents, absenteeism, and poor work performance, then there should be an observable gradient, with people testing zero having better subsequent performance than people testing just below the cutoff.
The fact that none of “drug-free workplace” advocates, and none of the testing labs, have published such as study might suggest to the suspicious-minded that they might not like the results.