The recent, laughably bad Centennial Institute report on marijuana legalization in Colorado talks a lot about the impact of cannabis consumption on high school students. In fact, lost productivity due to marijuana-related high school dropouts is the biggest component of the costs it erroneously attributes to legalization. Yet conspicuously missing from the report is any claim that legalization led to an increase in marijuana use by teenagers, a popular theme among pot prohibitionists. The omission is understandable in light of survey data showing that rates of marijuana use among middle and high school students in Colorado and Washington, the other state that legalized recreational use in 2012, have been essentially flat since then: These numbers come from the National Survey on Drug Use and Health (NSDUH), which generates state-specific data based on two-year averages to compensate for the relatively small samples at that level. According to the most recent numbers, the rate of past-month marijuana use among 12-to-17-year-olds was a bit lower in 2016-17, after three years of state-licensed sales (and four years of legal home cultivation in Colorado), than it was in 2011-12.
Results from the biennial, state-sponsored Healthy Kids Colorado Survey, in this chart limited to high school students (which helps explain why the rates are higher than the NSDUH numbers), are similar:
I have to say I am rather surprised by these numbers. I was never comfortable with assurances that legalization would not lead to more underage use. While it's true that licensed marijuana merchants, unlike black-market pot dealers, actually card their customers (and anyone who has bought marijuana in Colorado or Washington can testify that the stores take that responsibility seriously), it seemed to me that leakage from the adult market might very well make pot more readily available to teenagers, resulting in more cannabis consumption. So far it hasn't happened, maybe partly because legalization made marijuana less appealing to adolescents—a reversal of the "forbidden fruit" effect.