A new Federal Reserve study by Jason P. Brown, Elior D. Cohen, and Alison Felix looks at the effects of marijuana legalization.  Here is the abstract:

We analyze the effects of legalizing marijuana for recreational use on state economic and social outcomes using difference-in-differences estimation robust to staggered timing and heterogeneity of treatment. We find moderate economic gains accompanied by some social costs. Post-legalization, average state income per capita grew by 3 percent, house prices by 6 percent, and population by 2 percent. However, substance use disorders, chronic homelessness, and arrests increased by 17, 35, and 13 percent, respectively. Early legalizing states experienced larger economic gains yet similar social costs, implying a potential first-mover advantage.

Tyler Cowen discusses this study in Bloomberg:

The researchers used appropriate statistical controls, but there is some question about causation vs. correlation. At the very least, it seems highly likely that state GDP went up: A state with legal marijuana can sell it, including to users in other states. Selling marijuana is a new business, and like any new business, it boosts the local economy.

Due to the replication crisis in the sciences, it’s sensible to remain cautious about this sort of research.  But in this post, however, I’ll assume their findings are accurate.

Let’s start with the fact that the estimated gains in income are huge.  To a non-economist, 3% may not sound very large, but it is.  The US defense budget is roughly 3% of GDP, and you rarely see people describe defense spending as small.  In contrast, the legal marijuana industry is tiny, well below 0.2% of GDP in California.  Therefore, this large a rise in income cannot plausibly be attributed to the direct effect of adding legal pot to a state’s GDP.  Instead, marijuana legalization seems to have produced some strong positive externalities—some combination of making workers more productive and adding to the number of workers.  If true, that’s a finding that we should be “shouting from the rooftops”.

Tyler has mixed views on pot legalization, and in his Bloomberg piece he mostly emphasizes the negative:

It would be hard to use this latest research paper to persuade people that additional drugs should be legalized as well. And I would not be surprised if some governments decided to end their experiments with marijuana legalization. Unless you are a responsible user, how exactly does it make you better off? Looking only at the practical issues, what is the case for legalization?

Well, the study says it leads to higher incomes.  Yes, that seems unlikely.  But then why cite the study?

The strongest argument for pot legalization is that it is cruel to send people to prison for selling or consuming pot.  After legalization in California, the number of people imprisoned for marijuana offenses fell dramatically.  On the other hand, the black market has not gone away, and thus the criminal justice benefits have been far less than they should have been.  That’s partly because pot remains illegal at the federal level, and this substantially increases the cost of doing business.  In addition, states have adopted legalization in such a way as to encourage the continuance of a black market.  There’s nothing special about marijuana that would make it more susceptible to a black market than are toasters or tee shirts.  The black market is almost entirely caused by burdensome regulation.  (Contrary to popular wisdom, taxes are not the main problem.)  The government may wish to restrict sales to people below a certain age and ban driving while under the influence.  Otherwise, it’s not obvious why there should be any regulation of pot production and distribution.

So what would we expect from complete pot legalization?  Here are my guesstimates:

 1. Some increase in a state’s population, but probably less than 2%.

2. No significant change in per capita productivity or income.

3. Some increase in both total usage and problem usage.

4. A substantial decrease in crime and punishment, much bigger than what we have observed thus far.  The black market would be almost completely ended, except for resale to underage teens.  (We’d have an even smaller black market than for cigarettes, which have higher taxes than pot.)

(Note that for points #1 and #2 I am actually more pessimistic than the Fed study.  I believe they overstate the economic gains.)

It is interesting to compare this list to the effects of alcohol legalization.  I suspect that alcohol has a much more negative impact on productivity than does pot.  It also seems likely that there is more problem usage of alcohol than pot, and that the health costs are greater.

If society were serious about banning “bad things”, it might make more sense to start by banning alcohol.  Of course that experiment was tried, and the effects were roughly consistent with the pros and cons discussed above.  Banning alcohol reduced both consumption and problem consumption, and led to a big increase in crime and punishment.  The latter is a clear negative from prohibition, whereas the former is ambiguous.  Many people enjoy consuming alcohol, while heavier users suffer from some pretty severe consequences. I suspect that both the gains and losses from pot consumption are a bit lower than for alcohol.

When I look at proposals to ban products such as alcohol, tobacco and pot, I see one massive negative consequence (more crime and punishment), and then some other effects that are hard to judge.  In 1933, the US public rejected alcohol prohibition, and is now beginning to develop the same view of pot prohibition—the policy has uncertain gains and huge losses.

PS.  When I say “complete” pot legalization, I mean legalization at the federal level combined with state laws that are not more burdensome than the laws for selling beer.