This conversation about drug legalization was long and wide ranging. One possible view is that gangs exist wherever there is poverty; if it isn't drugs, it's sugar or milk or whatever they can control. The essential ingredient for gangs, in this view, is a large supply of young men with few alternatives in the legal economy.

In the middle was Tyler, arguing that legalizing drugs would reduce crime, but not by that much, because the gangs would persist.

On the other side was me, arguing that the crime reduction benefits should be large.

I've been thinking a lot about this over the last few days. It still seems to me that gangs are hard to support when there is good policing. Gangs flourish in places like Rio because the police force is corrupt and doesn't care about the favela inhabitants. They flourish in drugs and prostitution because contracts are not legally enforceable--if you can't sue to get the drugs you're owed, you need to use violence. Since there is safety in numbers, you get a gang.

As it happens, I'm reading The Bottom Billion, Paul Collier's excellent book on poverty traps in the developing world. As you can imagine, it has something to say on the subject of lawless bands of young men preying on the populace. A lot of it backs up the first position. "Civil war," Collier says, is much more likely to break out in low-income countries: halve the starting income of the country and you double the risk of civil war."

He expands:

. . . according to psychologists, on average about 3 percent of any population have psychopathic tendencies, so you can be sure that some of those in the recruitment line will be psychopaths. Others will be attracted by the prospect of power and riches, however unlikely; if the reality of daily existence is otherwise awful, the chances of success to not have to be very high to be alluring. Even a small chance of the good life as a successful rebel becomes worth taking, despite the high risk of death, because the prospect of death is not so much worse than the prospect of life in poverty.

Later he notes:

So what characteristics did make people more likely to engage in political violence? Well, the three big ones were being young, being uneducated, and being without dependents. Try as one might, itis difficult to reconcile these characteristics of recruitment with an image of a vanguard of fighters for social justice.

And even if you nominally resolve the source of the conflict, that doesn't necessarily end the violence:

Once a war has begun, the economic damage undoes the growth achieved during peace. Worse, even aside from this economic damage, the risk of futher war explodes upward. Civil war leaves a legacy of organized killing that is hard to live down. Violence and extortion have proved profitable for the perpetrators. Killing is the only way they know to earn a living. And what else to do with all those guns? . . . the emerging pattern seems to be that guns become cheap during conflict because so many get imported through official and semiofficial channels that a proportion of them leak onto the informal market. The legacy of conflict is cheap Kalishnikovs . . .

The end of the political fighting ushers in a boom in homicides. Presumably, this is part of a wider surge in violent crime.

So at least in the early stages, this seems to indicate that legalizing drugs wouldn't reduce crime too much; indeed, by disrupting a somewhat stable market, it might increase crime.

But over the long run, I think Collier's evidence supports my position. At the same time that he tells us that rebels are often attracted to money and power, he adds:

The key point of Weinstein's research is that in the presence of natural resource wealth--oil, diamonds, or perhaps drugs--there are credible prospects of riches, so that some of the young men in the queue to join will be motivated by these prospects . . .

Slightly later on, he says:

And where are the violent groups most likely to form? One might think it would be in the districts that are most deprived of social amenities, for that is supposedly what it is all about--oil wealth being stolen by the oil companies and the federal government instead of being used for the benefit of local communities. But Aderoju found that . . . there was no relationship between the social amenities that a district possessed and its propensity to political violence. Instead, the violence occurred in the districts with oil wells.

This suggests that an opportunity for economic rents--like, say, a line on a way to import a highly illegal substance--at the very least increases the supply of potential gang members, and gang violence.

There's also the fact that we seem to be exporting our drug-related violence elsewhere; according to the book, 95% of the world's supply of hard drugs are produced in conflict countries. Obviously, this is in part because conflict generates a zone outside of government control--but given the observation that violence clusters in areas where there are economic rents, it seems plausible to say that profits from the drug trade also increase the incentives for violence.

So I stand by my conclusion that the social benefits of legalization would be large, both here and abroad. It would not be without cost--I find it hard to believe that you wouldn't see more drug addicts if we legalized them. But the spectacular violence associated with vice trades, and the glamor drug rents lend to criminality, seem to outweigh the social cost of more drug abuse.

Of course, this does raise the question of what all those young men will do with themselves.

Update One participant emails:

On the one hand you argued that the drug wars implied huge wasted rents leading to crime. On the other you cite Levitt etc on the low returns to drug dealing. These two positions cannot be reconciled. If there are mostly winner take all rents in drug dealing, and the average returns are small, then the artificial rents due to drug control cannot, ipso facto, be large. Thus, one has to ask, if small rents with winner-take-all markets are sufficient to generate this huge amount of crime, then (probably) smaller rents due to legalization (and having to find substitutes) should still be sufficient to generate gang problems in the absence of a cure to the policing problems in poor areas with dysfunctional groups.

My response is that it all depends on the relative opportunities--absent the drug rents, would other rents be high enough to attract so many people into the tournament? Or would the legal sector become relatively more attractive?

That's an empirical question that I don't think we can answer without legalizing drugs and waiting fifteen years. But I suspect that there is a tipping point--that the gangs are supported in part by the fact that so many young men in their neighborhoods are criminals, which creates a culture hospitable to criminals. If you start moving more young men into legal work, you may hit a tipping point where criminality becomes stigmatized, and the social institutions that support it collapse. I'm not saying that there would be no criminals, obviously, but that the dominance of crime in some areas would cease.

Still, this is a very good point that I'm still pondering.

We want to hear what you think about this article. Submit a letter to the editor or write to