The American Interest
Policy, Politics & Culture
Triage in the Drug War
Published on February 21, 2012

There’s really nothing wrong with existing U.S. drug policy, except that all but one of its underlying ideas is false.

The one true idea is that drug use can be a dangerous activity for drug-takers and others, both because people under the spell of intoxication partially lose control of their behavior and because people under the spell of addiction partially lose control of their drug-taking. It follows that a purely free-market approach to addictive intoxicants will not lead to good results; taxes and regulations are necessary, and prohibition may be justified. Whether prohibition is actually justified for any given drug depends on the risks of the drug, how deeply socially embedded its use has become, and the state’s capacity to enforce it. (In my view, the answer is “no” for alcohol, but “yes” for cocaine.)

But this simple truth is taken, by both U.S. domestic law and policy and by the international drug control regime, as implying a number of stark falsehoods:

  • That the addictive intoxicants we have chosen to prohibit are radically different in nature from alcohol, the addictive intoxicant we have chosen to permit;
  • That all use of the illicit drugs is “abuse,” pathological in origin and result, causing harm to users and others.
  • That all, or at least most, users of illicit drugs are addicts, or on their way to addiction.
  • That drug abuse (and the more severe diagnosis of drug dependency) is typically a “chronic, relapsing brain disease.”
  • That, because drug-dealers are often violent, and because users often engage in acquisitive crime (since prohibition has made drugs artificially expensive), in addition to intoxicated crime, it follows that enforcing the drug laws naturally tends to reduce non-drug crime. The opposite is more often the case; greater enforcement (unless specifically targeted at the most violent dealers) tends to create incentives for more violence, because violence is one way that drug dealers protect themselves from enforcement.

In addition, the “drug war” has created a set of governmental and non-governmental bodies devoted to the three recognized means of controlling drug abuse: law enforcement, prevention (persuading non-users not to start and casual users to quit), and treatment. Naturally, the people who work in these agencies want to believe that their activity is useful, and that larger budgets allowing them to do more of it would be beneficial. That creates an additional set of false beliefs, which Francis Bacon might have listed alongside the Idols he identifies as barriers to truth-seeking (the Idols of the Tribe, the Cave, the Marketplace, and the Theater) as “Idols of the Agency.” The belief that drug law enforcement reduces non-drug crime is such an Idol. So is the belief that drug law enforcement can substantially shrink the size of an established drug market. Enforcement may be able to prevent, or at least retard, the growth of a new market, but replacement and adaptation make anything resembling “victory in the drug war” chimerical. Similarly, prevention and treatment providers systematically overestimate the potential of their activities to shrink the underlying problem. Drug treatment, for example, can be enormously useful to individuals, but most people with drug problems won’t enter treatment or stick with it, even under nominal compulsion.

Thus the grand fallacy of drug policy is that we have strategies in place which, if done with somewhat greater skill or on a somewhat larger scale, could put a substantial dent in the problem.

The most blatant and most destructive incarnation of this fallacy is the belief that “source and transit countries”—lands, typically poor or middle-income, where illicit drugs are produced or trafficked on their way to rich consumer countries such as the United States—have the responsibility to reduce the flow of drugs, and that they can do so if properly motivated. The grim reality is that export prices are such a small fraction of import prices that changes at the source or in transit have negligible impacts on final consumption. It’s possible to move the traffic around; for instance, massive enforcement in the Caribbean in the early 1980s managed to force cannabis and cocaine traffickers to abandon the sea route from Columbia to Florida and the Gulf Coast and ship their products overland through Central America and Mexico. But in the end, the quantity of drugs sold will be determined by the behavior of buyers and sellers in the United States. The ritual of the U.S. government scolding and threatening the Mexican government to demand that it “do more” about drug dealing has a somewhat farcical quality. Or it would, if the consequences in Mexico weren’t so bloody.

Mexican drug violence has been rising at least since 2004, and the rise seems to have been accelerated rather than slowed by the Mexican government’s decision—announced to wild applause in Washington—to go to war against the major drug trafficking organizations (DTOs). Right now, drug-related homicides in Mexico are reckoned to number about 1,000 per month, giving Mexico an overall homicide rate about three times that of the United States, which in turn is about five times the average for developed countries. The gangs are sufficiently well-armed, and the Mexican law enforcement agencies sufficiently demoralized (of the 40,000 drug homicides since President Calderón took office, fewer than three dozen have led to homicide convictions) that the Mexican army has been called into action. But the best that can be claimed for the results so far is that the bloodshed may have reached its peak: The rate of increase in violence seems to have begun to decline. When the good news is in the second derivative, fundamental policy changes are called for.

Some of those changes could come from the U.S. side. The HOPE probation-enforcement program scraps the ineffective threat of long-term incarceration for swift and certain sanctions for drug use by offenders under community supervision. It has demonstrated the capacity to reduce their drug use by at least three-quarters. The majority of the long-term, high-volume drug users who account for most of the illicit markets are also criminally active and frequently arrested. Carried out on a national scale—which it could be, since the program more than pays for itself in reduced incarceration—HOPE could shrink the total drug markets by as much as 50 percent. But that means changing the operations of fifty state parole agencies and 3,000 county parole agencies. Even a target of 50 percent coverage five years from now would reflect a degree of optimism bordering on the fanciful. Mexico doesn’t have five years to wait.

The same applies to the long-overdue improvements in the pay and training of Mexican police, and the capacity of Mexican courts to process serious crimes. Such efforts have no real promise of showing significant results before another 40,000 corpses have piled up.

In the short and middle term, any relief will have to come from changes in enforcement practices to make violence a source of competitive disadvantage rather than competitive advantage for drug-dealing organizations. Enforcement efforts should be designed to influence the way the drug business is carried out rather than attempt, futilely, to shrink the volume of the trade.

Drug law enforcement can’t stop the flow of drugs in the face of the pressure created by drug demand, but it remains the most important cost and risk of illicit drug dealing. An organization singled out for enforcement attention becomes uncompetitive with its rivals. That is, we can’t dismantle the industry by dismantling all the firms at once, but we can put any selected firm out of business.

The great example of success here is the experience of the Mafia in the U.S. cocaine trade. In 1980, the Mafia families remained the most powerful set of criminal enterprises in the country, and the cocaine trade was perhaps the most lucrative criminal market in the history of the world. Law enforcement spectacularly failed to prevent the cocaine epidemic. But the Mafia never got more than a taste of that huge revenue stream. Why not? Because by then the Organized Crime Strike Forces and the FBI had put so much pressure on the Mafia that selling to them, or buying from them, would have been an act of folly on the part of any drug-dealing operation. The dons simply couldn’t compete in the cocaine market.

Given the weakness of Mexican law enforcement, there is no way to put comparable pressure on any of the big DTOs. But there is no such weakness on the U.S. side, and the Mexican groups make most of their money selling to U.S. drug distributors. That connection is, potentially, their vulnerability.

It is claimed—with how much accuracy I don’t know—that U.S. authorities can trace the connection between specific Mexican DTOs and the major middlemen and retail distribution networks in the United States that buy their drugs. Assuming that’s true, or nearly true, imagine that the full weight of U.S. drug law enforcement, which keeps half a million dealers in prison at any one time, landed just on the middlemen and distribution networks who were the customers of one of the big Mexican DTOs: the Gulfo organization, Los Templarios, or Los Zetas, or whichever organization the Mexican government identified as the most violent.

That wouldn’t change the overall volume of the drug business. But the target Mexican group would quickly find itself without customers, as some were dismantled by enforcement and the others scrambled to find new sources to get themselves out of the enforcement crosshairs.

If the original target selection had been well-made, destroying the most violent organization would be expected to reduce the overall level of violence in Mexico. But that’s only part of the payoff. If, in the wake of the dismantlement of one big DTO, the Mexican government announced that it was going to pick a second target, and would choose the group that carried out the most violence over the next three months (or simply the next group to carry out a spectacular outrage), it’s possible to imagine a “race to the bottom” in violence levels, as each of the remaining groups sought to be less violent than at least one other group.

Now, there are at least five unsupported assumptions behind this strategy:

  • That the organizations can control the violence of their members;
  • That the Mexican government can develop a process that would accurately target the most violent group;
  • That the U.S. government knows enough about trafficking patterns to be able to attack specifically the distributors of the target group’s drugs;
  • That the DTOs wouldn’t, or couldn’t, strike back with random violence against officials and citizens in Mexico and the United States; and
  • That the result would not be a fragmentation leading to still higher violence levels, at least in the short run.

But at least the first three of those assumptions could be confirmed, or refuted, by facts currently available or obtainable by the two governments. If they seemed to be correct, the next step would be developing a detailed strategy entailing the targeting process, the announcement process, and the enforcement follow-through, along with a monitoring system to ensure that the program is being carried out as planned.

Even after all that staff work, the strategy would remain a difficult, high-risk venture. But it’s possible to imagine that it would work, and work relatively quickly, to reduce the homicide rate in Mexico. If anyone else has a proposal that can reasonably make the same claim, I’d like to hear it.


Planning Process

Data Gathering:

  • Mechanisms of violence.
  • To what extent are killings centrally ordered or controlled?
  • What are the group-to-group variations in the degree of centralization?
  • How is violence currently advantageous for the individuals and groups that use it?
Flows of drugs, arms, and money:
  • Can DEA actually assign to each substantial U.S. drug distribution network one or more sources in Mexico?
  • How “networked” is the system:  does a single organization have multiple sources?
  • How mobile are individuals and subunits from one major organization to another?
  • What happens when a major organization goes down (Beltran Leyva, for example)? How do trading patterns change? Does violence go up or down?
Governmental capacities (Mexico and the United States):
  • Can the authorities convincingly attribute most killings or other violent gestures to specific organizations and sub-units?
  • How accurately can individuals in Mexico and drug distribution networks in the United States be linked to specific Mexican DTOs?
  • How much pressure could the combined efforts of the two governments put on any one designated organization?
  • How great a competitive disadvantage would that level of enforcement pressure create?
A Scoring System for Violence
Scoring elements:
  • Number of killings
  • Social roles of victims (other gang members, ordinary citizens, police, journalists, community leaders, elected or appointed public officials
  • Terroristic actions and threats
  • Extortion
  • Kidnapping
  • Systematic intimidation of public and civil-society institutions
  • Systematic corruption
A weighting system to combine those elements into a single “score” for each organization:
  • Assigning relative importance to different elements
  • Dividing by revenues or drug flows to measure “violence-intensity”
Process Considerations
  • Scoring system development should be public and transparent.
  • Data elements must be transparently verifiable and minimally subject to “false flag” operations.
  • Scoring system must be announced and published, and some time-frame set out (weeks, not months) for the determination of the “winning” organization, with a pre-commitment to use U.S. and Mexican enforcement preferentially against that group until it goes out of existence. (Reduced violence is harder to observe and more transient, and thus threatening only to concentrate until that group reduces violence is a much less potent threat.
  • The United States must announce in advance its willingness to abide by the Mexican determination and to focus enforcement attention on the winning group and its customers. The effective declaration of war against Los Zetas in the transnational crime strategy is problematic.
  • Ideally, the management of the scoring system and the declaration of “the most dangerous organization” should be delegated to a body of respected experts not limited to public officials. Would anyone be willing to serve on that body, given the personal risks?
  • The “winner” must be clearly announced, and enforcement progress against that group publicly updated so that “the fight against Los Zetas” or “against Golfo” becomes a public event with the government’s prestige invested in speedy victory.
  • A new selection process begins when victory is declared against the first target.
Operational Planning
  • Identify responsible agencies (Mexican and U.S.)
  • Identify resource needs: personnel; institutions (courts, prisons); weaponry; financial; authority
  • Develop internal accountability systems to ensure fidelity to the planned concentration on a single target
  • Develop external reporting to make progress transparent to officials and citizens.



Mark A.R. Kleiman is a professor of public policy at the UCLA Luskin School of Public Affairs and a visiting fellow at the National Institute of Justice. The opinions here are his own, not those of the U.S. Department of Justice. He is the author (with Jon Caulkins and Angela Hawken) of Drugs and Drug Policy (Oxford University Press) and (with them and Beau Kilmer) of Marijuana Legalization (forthcoming in June, also from Oxford).