A study published in Lancet last week raised some eyebrows for describing a strong link between frequent marijuana use and a variety of negative outcomes, particularly lower high school graduation rates. As I wrote at the time (and as plenty of other observers have pointed out), while the study provides compelling evidence of a link, the direction of causality isn't clear.
But that hasn't stopped opponents of marijuana legalization from using the findings to argue that increased marijuana use will lead directly to lower graduation rates. This week, for instance, the Washington Post's editorial board cited the findings as part of an argument against D.C.'s November ballot measure that would legalize marijuana. An Australian medical news outlet headlined its story on the findings with "Cannabis use catastrophic for young brains."
Given that we'll see a lot more rhetoric like this as more states consider liberalizing their marijuana laws, it's important that we put the Lancet findings in context. Wouldn't it be nice, for instance, if we could look at teen marijuana use rates over time and see how they interact with high school graduation rates? Or if we could look at different regions with different prevalences of marijuana use, and see if their graduation rates differed significantly?
As a matter of fact, we can.
The chart below plots monthly marijuana use rates among U.S. high school seniors, from the Monitoring the Future study, against public high school graduation rates from the National Center on Education Statistics. You'll notice that from 2003 to 2006, marijuana use declined slightly while graduation rates held steady. But from 2006 to 2012, the rates of monthly marijuana use and high school graduation increased in tandem.
From 2006 to 2012, monthly marijuana use among high school seniors increased by more than 4 percentage points*, from 18.3 percent to 22.9 percent. If indeed marijuana use were the educational catastrophe that opponents predict, you'd expect to see downward pressure on national graduation rates as more kids took up the habit. But in actuality, the opposite happened: over the same period, as kids were smoking more, graduation rates jumped 8 percentage points.
This should not be at all construed to imply that increasing rates of marijuana use are somehow causing higher graduation rates. Correlation doesn't equal causation. And these numbers don't even constitute an argument against the Lancet study findings - it's perfectly plausible that any negative consequences of marijuana use are too small to show up in a simple national trendline like this.
But it's a useful corrective against the facile notion that "more weed = less graduation." In reality, there are a whole host of factors that influence graduation rates, from income to demographics and beyond. Marijuana use may indeed exhibit some pull on the graduation numbers at the national level. But this effect, if it exists, is likely dwarfed by all the other factors at play.
I also looked at the relationship between state-level marijuana use rates, as measured in the National Survey of Drug Use and Health, and state-level graduation rates. There was zero correlation. For instance, Vermont teens were about twice as likely as Utah teens to use marijuana in 2009. But Vermont's graduation rate was about 12 percentage points higher than Utah's that year.
Again, we can't make any positive assertions based on a simple correlation like this. But it does provide evidence against the argument that places with higher rates of teen marijuana use will necessarily see lower graduation rates as a result.
In sum, none of this is to say that Lancet findings are incorrect. It seems eminently reasonable to assume that, at the individual level, people who use drugs heavily in their youth are setting themselves up for a bad time down the road. But those findings absolutely should not be translated into simple policy prescriptions that say "if you legalize marijuana, fewer kids are going to graduate." The reality is considerably more complicated than that.
*Note that these figures are for high school seniors only, which partially explains whey they're higher than the numbers on teen drug use released as part of the National Survey on Drug Use and Health this week. The two surveys use different methodologies, with the MTF data usually showing higher incidences of drug use than NSDUH.