Yglesias bloviates a bit today about drug policy and decides that:
One plausible story is that marijuana would substitute for alcohol use, and since marijuana has fewer health risks than booze that would militate in favor of policies designed to make pot at least as easy to get as beer. Another also possible story is that legal marijuana will complement drinking, and therefore exacerbate existing alcohol-related problems. That would indicate that we should try to make pot less widely available than beer currently is (though still more available than under the status quo) along with higher taxes on alcohol.Like much drug policy writing, this idea suffers from several fairly coarse assumptions but most importantly the failure to make a distinction between overall drug use and "problem use" that carries a significant social cost. It's as if people were trying to reduce car accidents by making it harder to get a car. Yes, it may have some effect, but you're sort of missing the point.
Efforts to prevent the population at large from smoking pot (or really, doing any illicit drug) are doomed to be very ineffective. If you commit a burglary your chances of getting caught are something like four hundred times higher than if you consume illegal drugs (and your chances of getting caught for committing a burglary are not that high.)
There's good news there, though, because the really harmful problem use in the US is drug use by people who commit crimes. For many years it was believed that drug use actually turned people into criminals, but this was studied in the 1980's (check your university library for the excellent work of James Inciardi among others) and found to be incorrect.
It is actually very easy to lower the drug use of convicted criminals by drug testing them several times a week and imposing a mildly painful penalty like a small amount of community service. Those who are addicted and truly cannot stop will have to be provided with treatment (or, if you prefer, stiffer penalties like incarceration), but it is known that most drug users are very responsive to the costs - both monetary and other costs - of their drug use and if those costs go up significantly and predictably, they will use less.
This approach is known as "Coerced Abstinence" and with a google search you can read all kinds of stuff about it including a lot of scholarly research. It has been discussed since at least the late 1990's and has been implemented many places and its effects studied.
Long story short, it works very well, is cheap, and reduces the strain on the justice system. It also lowers drug use among criminals, which is a population whom EVERYONE, from prohibitionists to outright legalizers, agrees should not be using mind-altering chemicals recreationally.
This is not just an approach we should be trying. It is THE approach we should be trying. It makes sense. It works. It doesn't cost billions of dollars. What's the holdup?