Statistically speaking, Mexico is a relatively safe place with 12 murders per 100,000 inhabitants in 2009. The trouble is that the violence is concentrated, and according to one economist I talked with here, that's because the drug-trafficking business is structured much like Colombia's was in the 1980s and '90s.

Powerful monopoly suppliers need to control key zones so they can guarantee an army of contract employees. These "ants" carry the drugs over the U.S. border at a limited number of strategic points in small shipments. Without mafia-style terror, the cartel's domination along the route cannot be maintained.

Mexican law enforcement has been courageous in trying to confront these monopolies, but firepower has not done the job. That's because this is an economic problem. Lower levels of violence in the U.S., despite widespread availability of drugs, and an improved picture in Colombia, where cocaine still flows, are best explained by competition and the smaller scale of the operators. It wasn't always that way in Colombia. In Mexico it could also change.

To help Mexico deal with this "antitrust" problem, the U.S. has to recognize that competition in the narcotics sector is preferable to the monopolistic syndicates that threaten the state and could move north. But this would require greater flexibility from U.S. drug warriors.

Some progress may be in the making on marijuana, and Mexicans will be watching the California ballot initiative that asks the electorate to approve the legalization of the ubiquitous weed. It is far from clear that Proposition 19, as it is known, will pass. The combination of conservatives who fear that legalization would transform us into a hash-happy heap of hippies, drug warriors who make a living off of the criminalization of pot smoking, and gangsters whose profits are tied up in prohibition could be enough to defeat it by a narrow margin.

Nevertheless, the competitiveness of the "yes" vote on this proposition suggests that attitudes toward "grass" have generally softened, and that many Americans would prefer the business be run legally. For sure, the U.S. market is robust, and "medical marijuana" looks like a way of legalizing without admitting to it. There is also the fact that the stuff seems to move around the country quite easily, demonstrating some tolerance on the part of U.S. law enforcement for the retail sector that distributes it.

More competition in marijuana production and distribution in the U.S. would help beleaguered Mexico. As it stands now, the gangsters have good reason to pull out all the stops to get their marijuana across the border where the market is large, barriers to distribution are low and prohibition adds value. Profit margins are not huge but the sales volume is there.

Mexican officials estimate that the marijuana business makes up more than half of the Mexican cartels' income. Legalizing grass in the U.S. would mean increased competition for Mexican exporters and lower profit margins, thereby depriving the monopolies of important income.

The bigger problem for Mexico is U.S. cocaine demand. Here there seems to be at least some recognition among drug warriors of what hasn't worked. Wrote former Drug Enforcement Administrator Robert Bonner in a recent issue of Foreign Affairs magazine: "The goal must be clear. In Colombia, the objective was to destroy the Cali and Medellin cartels—not to prevent drugs from being smuggled into the United States or to end their consumption."

This is risible. The entire raison d'être of the last 40 years of U.S. drug policy abroad has been to stop supply in order to reduce demand in the U.S. Of course when this plan backfired and Colombian cartels grew more powerful, American and Colombian authorities had to adjust. But their war was predicated on the belief that interdiction of supply could diminish U.S. drug consumption.

If Mr. Bonner is now backing away from that argument, it can only be because he is looking at the numbers. Andean cocaine production in 2008 was down only 8% since 1999, and even that might be explained by a shift in preferences in the U.S.

Analysts and policy makers agree that a crackdown on Caribbean narco-routes has driven the business through Mexico, though it hasn't reduced U.S. drug use. The economist I talked to argued further that if cocaine moved more easily through the Caribbean as it once did and the Mexican border were more porous, it would be harder for a big cartel to monopolize the traffic, even through violence.

It's an interesting theory and of course runs totally counter to the direction of U.S. policy. But if that policy is proven wrong, it wouldn't be the first time in the long history of the drug war.