Mark Kleiman, writing in this space about the marijuana legalization initiatives that passed last month in Washington and Colorado, endorses the “laboratory of democracy” rationale for state-level experimentation with marijuana legalization. He observes that if the federal government allows such policies to take effect—whether by statute or prosecutorial discretion—we will actually get some real evidence about whether legalization will increase usage rates, and whether there will be associated secondary effects.
The policy interest in Colorado and Washington goes beyond marijuana.
More generally, data from those states should give us insight into the general question of deterrence: does the threat of criminal prosecution deter people from engaging in criminal activity? And if so, how much does it deter people?
The policy justification for criminalization of drug possession and distribution has always been deterrence. The theory is that criminal penalties are necessary to deter people from getting involved with drugs. It is therefore vitally important to any assessment of our drug policies that we have some idea of how much criminalization actually deters people from getting involved with drugs.
This is because the one thing we know for certain is that it is enormously expensive to investigate, prosecute, and incarcerate drug offenders. If the rationale is that doing so deters others from drug use, then we the taxpayers are entitled to an accounting of how well it’s working.
The problem is that data is scarce.
The only way to measure the effect of laws on behavior is to observe behavior, and it’s not easy to obtain good data sets on underground economies like drug markets. So for the past two years my colleagues and I have been collecting and analyzing data on the wages paid by drug trafficking organizations to the couriers, or “mules,” who transport drugs across the U.S.-Mexico border into California. We get our data from the probable cause statements filed by the agents who interview the mules upon arrest.
Drug-mule sentences vary greatly depending on the type of drug and the size of the load: sentences for marijuana smuggling are shorter than those for cocaine smuggling, which are shorter than those for meth smuggling. We have data now from around 8,000 cases, and we can use it to answer the key question: do trafficking organizations have to pay more to get mules to transport loads that will result in longer sentences?
If they do, then the amount of money they have to pay to get mules to take on extra risk would be a very nice measurement of the deterrent effect of prison time.
All else equal, employers have to pay workers more to do work that comes with higher risk. The extra pay required to induce workers to take the job is called a “risk premium,” and the size of the risk premium is a good measure of the deterrent created by the risk. So we created a statistical model of sentencing risk and regressed risk against pay to see whether couriers get paid more for riskier loads.
We found that, sure enough, there is a risk premium in drug mule wages. The threat of imprisonment does create a deterrent to people taking work as drug mules. The problem is, it’s not very much of a deterrent. The risk premium is small, and declines as sentence lengths increase.
In California, for example, the market rate for a mule to bring across a small cocaine load with a likely sentence of two years is around $1000. For a medium-sized load with a likely sentence of three years, the market rate is around $2000. And for a large load with a likely sentence of four years, the market rate is $2200. In other words, the market values the first extra year of incarceration risk at $1000, but the second extra year at only $200. We see similar patterns for methamphetamine and marijuana.
And these loads are worth hundreds of thousands of dollars wholesale, so mule wages account for under one percent of load value. Why is the risk premium so small? We think it’s because detection rates—the risk of actually being caught and incarcerated—are low. Our best estimate for cocaine loads is a 5 percent detection rate
More problematic from a policy perspective is our finding that the deterrent effect is nonlinear. It decreases as sentence lengths increase. In other words, most of the deterrent effect of prison sentences is on the front end—the fear of being sent to prison at all—and it’s relatively cheap to compensate people for taking on the risk of additional years once they’ve agreed to commit a crime in the first place.
As the cocaine example illustrates, our data shows that it takes more money to compensate a mule for taking on the first year of sentencing risk than for the second year, and the amount continues to decline for the third and fourth years. The pay/risk curves for our California data flop over to the right, and flatten out completely at around four years of risk. Because the deterrent effect declines with each added increment of sentence length, shorter sentences applied to more mules would be a much more efficient deterrent than longer sentences applied to fewer mules.
But even an increased deterrent effect does not actually reduce the number of people working as drug mules: it just raises the market wage for drug mule work.
Our data shows that trafficking organizations can easily overcome the deterrent effect by simply paying mules a risk premium. The market-rate risk premium is low enough that traffickers can easily afford it without any significant impact on their profitability. In other words, the incarceration deterrent does not impede mule recruitment, and thus does not reduce the supply of incoming drugs: it simply raises the salaries of drug mules.
That’s not a debating slogan; it’s simply what the data shows.
The Colorado and Washington initiatives should encourage us to be honest about what our drug policies are and are not accomplishing. I want to live in a society with less drug use. Drug addiction ruins lives, and controlling the production and distribution of dangerous addictive drugs should remain a government priority.
But if our current criminal-justice approach is not actually reducing the supply of drugs, we should think about better ways to spend our money.
My suggestion: shorten drug-crime sentences and shift our law-enforcement resources from the back end—incarceration— to the front end: direct supply reduction (border interdiction) and demand reduction (addiction rehabilitation and prevention).
For conservatives, that means more dog teams on the border. Dog teams actually interdict loads and directly reduce supply.
For liberals, that means more rehab and prevention programs to help addicts recover and prevent others from getting addicted.
For an ideologically split country, I suggest some of both.
Caleb Mason is a lawyer at Miller Barondess LLP in Los Angeles. A former federal prosecutor and law professor, his research on border drug interdiction policy is ongoing with Professor David Bjerk of Claremont McKenna. He welcomes comments from readers.