In recent months, the Washington Post has published many articles giving favorable coverage to marijuana legalization. But as news about marijuana—particularly the steady stream of scientific research on marijuana’s negative effects—undercuts the argument that marijuana is benign, the Post has had to take liberties to continue propping up the pro-legalization position.
Whether to garner more clicks or to support a pro-marijuana agenda, the Post has distorted basic facts to advance a message not found in the original research on which they report. The Post's May 29, 2016, story, “No, Legal Weed is Not Dumbing Down the Nation's Teens” presents just such a troubling case.
The headline is quite clear, and the reader can be forgiven for inferring that the study involved intellectual performance, and further, for assuming that legalization of marijuana was somehow implicated in this surprising result. Neither assumption is true.
The actual study found a decline in marijuana use disorders (not intelligence), based on analysis of 12-17 year olds who reported using marijuana at least once in the past year in the National Survey on Drug Use and Health (NSDUH). The period covered was 2002-2013 (even though 2014 results are now available, which if analyzed, would represent the first year when legal marijuana could have produced an impact).
But before you express relief over the decline, the study also found that the rate of marijuana disorders (meaning such things as drug dependency, or anti-social conduct, such as fighting or stealing) is “quite high … higher than all other illicit substance disorders, combined.” Nevertheless, the study attributes the decline to two factors: a claimed drop in overall prevalence of cannabis use, as well as a decline in the probability of developing a disorder, among those who reported marijuana use during the past year.
Immediately, one can see potential problems with this type of analysis. A stumbling block is that marijuana prevalence among youth indeed appears down for that overall time period, but implications are confounded by the shape of that decline.
Looking at the NSDUH category of “past month” marijuana use—which captures the more policy-relevant measure of adolescents who use it regularly—we first see a steep decline in marijuana use between 2002 and 2008 of 18 percent.
But this decline was followed by a 10-percent increase in use from 2009 until 2014.
Hence, the final outcome is in fact “down” from 2002, but only because the rise since 2009 has yet to eclipse the high point from which measurement began. An honest look at the push for marijuana legalization would acknowledge that fact, and would note that the period of decline occurred when national policy was opposed to legal marijuana, whereas the rise in prevalence has occurred when legal marijuana has been encouraged.
Regardless of the impact, one cannot posit for the same factor (i.e. “legal weed”) an effect of both driving down marijuana prevalence and also, later, driving it back up. The Post ignores this dynamic entirely, notwithstanding an editorial attached to the very journal article they are reporting on cautions them specifically about this “non-linearity” in the data.
But this same editorial in the Journal of the American Academy of Child and Adolescent Psychiatry creates even greater problems for the Post, further negating its conclusions in fatal ways.
First, in relation to the disorders, the Post posits that if more kids were smoking dope, “by now we’d be seeing … more problematic use, such as addiction and dependency… but in fact the exact opposite has happened.” Hence, their opening to suggest that perhaps legalized weed has helped. This is wrong. The study is careful to note that the disorder decrease is not centered on those with drug dependency, but critically on “conduct disorders.”
That is, those not suffering conduct problems experienced no decrease. This point is magnified in the editorial, which further posits the reasonable hypothesis that marijuana use has widened to a different population, Moreover, the results of other studies seem to contradict the notion of a decline in disorders.
Recent research from the National Epidemiological Survey on Alcohol and Related Disorders reported that not only did marijuana prevalence “more than double” among adults between 2001 and 2013, but “marijuana disorders” (for them, “abuse and dependency”) also doubled. The Post showed no interest in covering these compelling, but contrary results.
So what is the hypothesized explanation for the newest findings, whereby youth are going in one direction, but adults in the opposite? In prior years, marijuana use was found primarily among youth with co-occurring “conduct disorders.” But now that use is “acceptable,” youth who are not susceptible to conduct disorders are also using. That would affect the marijuana disorder probability finding, since fewer users, as a proportion, are now involved in “antisocial problems.”
This result would appear as a “decline” of the rate of disorders for youth, when in reality it would reflect a weakening of the association between use and negative conduct in the wider pool of users.
What are we to make of the Post's effort? It could be simple sloppy journalism, were it not for additional factors. First, the Post and other outlets routinely run articles that serve as defense briefs for legal marijuana. This story suspiciously fits the pattern, especially when the advocacy becomes blatant, as in this paragraph: “As a number of states consider legalization… opponents are already asking, ‘But what about the children?' If (this) research… is any indication, the kids will be just fine.”
But the clincher, again, comes from the editorial in the same journal, where we are explicitly warned: “(T)his article also could be interpreted by some as ‘proof' that marijuana liberalization laws have no effect on adolescent marijuana use. No such inference is warranted.”
As no mention nor link to the editorial is ever included in the Post's reporting, this admonition makes the Post story practically pernicious since journalists were explicitly warned not to disseminate the very false conclusions the Post does.