Media coverage insinuated that drug usage among welfare recipients was, contrary to Florida governor Rick Scott's claim that welfare recipients have higher rates of drug use than non-recipients do, quite low--perhaps lower than that of the broader population--as less than 3% of those who went ahead and had a test administered on themselves failed to pass it clean. The hole in this implication, of course, is that many of those who qualify for welfare but have recently used drugs aren't going to drop the cash to take a drug test they know they're going to fail. One could make the argument that the more surprising result is that 1 in 40 willingly paid to have a drug test administered on themselves for no reason. Maybe they thought the test would fail, or perhaps these are people on the bottom rungs of the societal ladder and are consequently not well known for high conscientiousness or foresight.
There's a little bankshot in Scott's plan, as it disincentivizes drug usage among the poor while simultaneously reducing the number of people who are eligible to receive benefits by putting another hurdle in their way.
Since then, I stumbled across the 2011 National Survey on Drug Use and Health, with free online access available through the SDA interface. Unfortunately, the closest the study gets to querying respondents on welfare recipiency is whether or not they (or their families, for minors) are eligible for Medicaid (20% of participants said they were, 80% said they were not--I suspect those potentially eligible but who've never actually used Medicaid are mostly unaware of their eligibility or unwilling to take advantage of it for whatever reason, so I'd guess we're doing more than just proxying for low income here), and that question can only be cross referenced with "have you ever used X?" inquiries, not "have you used X in the last month?" questions.
Thus the following attempt to get a feel for whether welfare recipients are more or less likely to use illicit drugs than non-recipients are is a pretty rough approximation, but we work with what we have. To give participants a chance to have used the various stuff if so inclined to do so, I excluded respondents under the age of 20.
The percentages who have ever smoked marijuana by whether they are eligible or ineligible for Medicaid:
Negligible differences between those eligible for Medicaid and those ineligible when it comes to weed and powdered cocaine, higher rates of having used crack, heroin, and meth among the eligible, and higher rates among the ineligible when it comes to having ever tripped on acid.
Crack is a black drug and meth is a white trash drug. Powdered cocaine and heroin are a little, uh, classier, acid is for the college-aged children of the middle class and the affluent who are on existential missions to find themselves, and weed is something people from all walks of life have used sometime in their adolescent and early adult years.
Parenthetically, if I'm able to use the data at hand to validate some of the stereotypes I'm asserting in the body of a post, I better do so. The percentages of respondents, by race, who've ever smoked crack:
And used meth:
GSS variables used: NEWRACE2, AGE2(9-17), MJEVER, COCEVER (fag!), CRKEVER, HEREVER (...better), LSD, METHDES
* I realize the usage figures for meth and LSD strain credulity, apparently being more experimented with than even marijuana. I'm unsure of why this is the case, but I doubled-checked the figures and that's what the results show, so take the meth data with a grain of your favorite substance. It's probably still worth noting the disparity between those eligible and ineligible for Medicaid, though.
++Addition++Dan shows how wildly off the meth (and by extension LSD) numbers are. The LSD and meth questions are grouped in under larger categories for hallucinogens and amphetamines, respectively, so it's probably "if you have ever used a hallucinogen, have you used LSD?", etc. Consequently, the racial distribution for meth use likely only applies to those who've used amphetamines of some kind. I'm fairly confident about the figures for the other drugs, though.