Banning it doesn't help

Emily Owens, an assistant professor in the Department of Policy Analysis and Management at Cornell, has a column in The New York Times about Prohibition, then and now:

Proponents of legalizing drugs offer many arguments in their favor, but perhaps the most ubiquitous, and most readily accepted, is that illegal markets drive up the murder rate. For proof, they say, just look at the 1920s: as television shows like HBO’s Boardwalk Empire and Ken Burns’ new documentary on Prohibition seem to attest, the nationwide ban on liquor sales led to an explosion of criminal violence. Surely the same is true today regarding illegal drugs.
The legendary lawlessness of Prohibition makes for great entertainment. And it may be true that some specific crimes, like gang violence, did increase. But the conclusion that it caused a huge increase in the homicide rate is almost certainly wrong. At first glance, the numbers do seem to support the conventional wisdom. According to Census mortality statistics, the national homicide rate jumped forty percent under Prohibition; after it was repealed in 1933, murder rates plummeted. It’s only natural to conclude that criminalizing the market for alcohol caused a huge increase in crime. But there are three major holes in this conclusion. For one thing, the Census data on homicides is not actually from a complete census. In 1900, the Census counted deaths only in New England, Indiana, and Michigan. New states were added almost every year until 1931. The states that were added later, like Nevada and Texas, tended to have higher homicide rates than states added earlier. In other words, part of that forty percent increase in the “national” rate is due to the inclusion of more violent states in the statistics, not an actual increase in murders.
Furthermore, Prohibition wasn’t a sudden thing. The 1919 passage of the Eighteenth Amendment culminated a seventy-year movement that had gradually shut down liquor sales nationwide. By the time the federal government got involved, it was already illegal to sell alcohol in 32 states, home to roughly a quarter of the population covered by the Census. Nor did repeal make all states go “wet”: it remained illegal to buy or sell alcohol in some states for years and Mississippi did not legalize it until 1966.
This sets up a convenient comparison: if illegal markets really do drive crime, we would expect homicide rates to increase earlier in states with early dry laws, and decrease later in states that stayed “wet”. And yet when we compare state homicide rates in “dry” and “wet” years, the forty percent jump in homicide rates drops to 26 percent.
Even that 26 percent is suspect. Perhaps the most important hole in the conventional wisdom is that it overlooks the social context of Prohibition. The early twentieth century was a time of unprecedented urbanization. Americans, especially black Southerners, were moving into cities at the same time as immigrants from Europe and China. Then, less than a decade into Prohibition, the Great Depression plunged us into poverty. Such upheaval alone could have turned city streets into war zones.
Standard statistical analysis allows us to disentangle the impact of dry laws from changes in urbanization, immigration, race, and age distribution, and even the ameliorating effect of the New Deal. The hard data do not show a strong relationship between criminalizing the market for alcohol and homicide rates once these other factors are taken into account. In fact, depending on the model, the actual effect of going dry ranges from a five percent increase to a thirteen percent decrease in state homicide rates, with margins of error of four percentage points.
How could dry laws have reduced crime? By making alcohol harder to come by, dry laws most likely reduced drinking. And researchers almost all agree that alcohol is associated with high levels of psychopharmacological violence; that is, conflicts that escalate because one or more parties are intoxicated.
Of course, the news media presented a drastically different picture. Gang shootouts made for compelling headlines, and these might have increased when states went dry. But this wasn’t enough to outweigh a reduction in more mundane, alcohol-fueled violence.
The best policies seemed to be lenient dry laws, which, before federal Prohibition, allowed very limited access to legal alcohol markets by, for example, letting consumers import alcohol from a wet state. That reduced psychopharmacological violence, but it also reduced the need for illegal markets. In such states, homicide rates were about ten percent lower than one would expect, based on the other social changes taking place.
There may be other reasons to legalize drugs. But when put to the test of modern policy analysis, the American experience of alcohol prohibition provides no compelling evidence that legalizing modern drug markets would reduce violent crime.

Rico says the violence in the drug trade has more to do with money than drugs...