It would be interesting to test white collar workers.
1) Contract language that says your workforce has to be drug free (government sourced contracts doubly so)
2) Insurance. If you have a fork lift driver run into a rack of transmissions (or drive your work truck into a building) your insurance is going to demand a urine test. If the urine test comes back unclean they will deny your claim.
Look, you got one task, one task only, it's not even hard to figure out, do not, repeat, do not, get high the week or two before your pee test. That simple. There are human bipeds burning valuable oxygen right now, who can't follow a test that elaborate and complicated. They are literally the kind of people where if you told them not to lick a circuit breaker, would turn around and an hour later electrocute themselves and probably a coworker or two by licking a circuit breaker on the jobsite. Whats the minimum IQ and discipline level necessary to pass a pee test? Not much, but there are failures out there walking around...
Its interesting that as far as I know this is the first strictly chemistry based IQ test. AFAIK there is no "pee in a bottle to determine if you can fizzbuzz" test for code monkey work. No chemical marker that can identify if you know the modulus operator... not yet...
There's also the possibility that giving up an addiction is not an intellectual activity (if you think about it a little!)
But the process of getting addicted certainly involves making some choices poorly.
Addiction is continued use in the face of adverse consequences. [1] You may have a physical dependence on coffee, but it doesn't sound like you're experiencing negative consequences in your life from its use.
[1] https://www.addictioninfo.org/articles/2216/1/Addiction-Do-Y...
In case someone is wondering, hair follicle tests are apparently problematic [1]; had to look that up because I remembered correctly that follicle tests can show use over the long-term, but didn't know some of the intrinsic challenges with that method. Come to think of it, I've never been asked to submit a hair sample when I've had to pass FBI interviews and drug tests to gain admittance to extremely-sensitive data centers at some of my clients in the past.
The "week or two" needs contextual guidance. An article on Wikipedia [2] gives a good run down of the detection periods. I would imagine detecting pot is the most common request, so that's about 100 days on the conservative side for a heavy (daily?) user, but apparently three weeks is fine for more casual users (what, once a week?).
[1] http://www.fleetowner.com/driver-management-resource-center/...
[2] https://en.wikipedia.org/wiki/Drug_test#Detection_periods
Have you ever seen a lathe accident, for example?
You can't spot test for being high on marijuana or pills like you can for drunkenness, so you rely on the urine test to keep everyone on the floor safe from being crushed in a press by a stoned co-worker (or themselves.)
You have two job applicants. The only thing you know about them is their resume (nicely embellished) and how they performed at their interview (nicely rehearsed). One applicant passes a drug test, one does not.
Can you really say that drug test outcomes don't change the risk assessment of each candidate?
We can have a conversation about whether employers ought to have access to that information, the same way we can say that employers ought not to have access to a candidate's medical history or prior salary figure, because society is better served when employers don't have access to these prejudicial sources of data. But that is a different question than whether the information is useful at all in hiring decisions, which of course it is.
[1] https://www.nytimes.com/2017/07/24/business/economy/drug-tes...
Remember, new hires are not guaranteed to add value - even if they worked for free, there is always a risk that some will reduce value. If a new hire's inability to show up on time holds up a production line, or a new hire's irresponsibility damages equipment, then you're better off not making the hire, even if you desire, in the abstract, to hire more people and expand your business.
Though now that I think about it I have technically worked at two companies that didn't drug test: my own, and a small 3-man startup I did some work for.
It's worth mentioning that I live in a red state. I suspect blue states have cultures that are a little more lax on drug testing, though I have no real data to back that up.