More adults are using weed in Washington and Colorado, but the kids are still alright. (AP Photo/David Zalubowski)

It's Christmas for drug policy data nerds! In late December to little fanfare -- not so much as a tweet! -- the Substance Abuse and Mental Health Services Administration released state-level numbers from the most recent National Survey on Drug Use and Health (NSDUH).

The national-level data from this study, released back in September, showed that drug and alcohol use rates among all Americans is essentially flat, while teen use is actually trending downward. These figures were borne out by more recent 2014 numbers from the federal government's other major drug study, the Monitoring the Future survey.

The new state-level numbers paint a more nuanced picture of drug use in the U.S. I'm going to focus on marijuana for today's post, since that's where most of the action's happening at the policy level. The first big takeaway? Colorado and Washington -- the states voting to legalize weed in 2012 -- saw big spikes in their overall marijuana use rates in 2012 and 2013.

For the state-level numbers, SAMSHA combines two years of data to ensure an adequate sample size. So, for instance, 2013 on the chart above actually reflects 2012 AND 2013. I've highlighted Colorado and Washington, as well as the U.S. total. You can mouseover the other lines to see rates in the other states.

The first important thing to point out is that since these numbers only go through 2013, they only reflect the period when Washington and Colorado had legalized the possession of weed, but had not yet set up their fully taxed and regulated marijuana markets -- that didn't happen until this past year. So in this context, the spike in use is even more noteworthy than it would be otherwise -- we can probably expect rates to climb even higher when the data reflect the effects of a fully legal marketplace.

On the other hand, plenty of other states saw their use rates climb by significant amounts too. Maine. Georgia. Maryland. Missouri. So Colorado and Washington are by no means alone in seeing increased marijuana use.

It's also significant that despite dramatic warnings from the attorneys general of Oklahoma and Nebraska, pot use in those states is basically flat over this period -- so at least from these numbers it's premature to claim that Colorado's weed is flowing into those states at significant rates.

In fact, Nebraska and Oklahoma can boast some of the lowest marijuana use rates in the nation. Another Colorado neighbor, Kansas, has the country's lowest rate. I mapped the current state-level use rates below.

Here's another important datapoint: the spike in use in Colorado and Washington happened almost exclusively among the 26-and-older crowd: use among teens aged 12-17 edged upward a little bit in those states, but this change was not statistically significant.

So on some level, what's happening in Colorado and Washington is exactly what you'd expect, and probably hope for, if you're a legalization proponent: increased use among the adults who can legally do so, but little change in use rates among teens.

It's instructive to compare these numbers to Alaska's, where weed has been de-facto legalized for personal use for almost 40 years. In that state, adult marijuana use is among the highest in the nation. But teen use rates are completely unremarkable, mirroring what we see so far in Colorado and Washington.

Again: these numbers are preliminary, and don't tell us anything about the effects of the retail marijuana markets being put in place in various states. But they're important, because they represent the baseline against which the impact of state-level pot policy changes will be measured.

Overall, I'd expect to see a continued rise in adult use in states that legalize weed. A big part of this will probably be the novelty factor: people who were previously discouraged from using marijuana due to its legal status may be tempted to give it a whirl when they can simply walk down the street and buy some at the store.

But weed isn't for everyone (see: Dowd, Maureen). It's reasonable to expect that many, if not most, new users may simply try it once or twice and decide it's not their thing. This seems to be what happened in Portugal, which decriminalized all drugs in 2000: use rates rose in the year or two after decriminalization, but have fallen since then. Marijuana legalization experiments in the U.S. may very well yield similar results.