REACTION ESSAY
TOWARD (MORE NEARLY) OPTIMAL SENTENCING FOR DRUG OFFENDERS MARK A.R. KLEIMAN UCLA A sentencing scheme for drug-law violations has direct effects on which dealers go to prison and for how long, and indirect effects on the behavior of drug-market participants as they adjust their activity patterns to accommodate to the legal threats they face. As the sentencing pattern varies, so will the incapacitative and deterrent impacts of drug law enforcement on the size and operations of the illicit drug markets. In addition, sentencing has a second set of indirect effects: on the behavior of investigators, prosecutors, and defendants (who are also potential witnesses against other defendants). Sentencing is one way prosecutors keep score. A long sentence reflects a judgment that the case was an important one; a short sentence reflects the opposite judgment. Prosecutors are no fonder than other people of handling unimportant matters. Prosecution is one way investigators keep score. When investigators learn that a particular type of case rarely if ever leads to prosecution, or leads to prosecution only on minor charges, they look for better things to do. (For example, the short sentences provided at the federal level for illegal sales of firearms help explain the lack of gun-trafficking cases.) For defendants, sentencing is more than a scorekeeping device. They have a direct and lively interest in minimizing the amount of time they will spend behind bars. Quickly agreeing to plead guilty rather than insisting on going to trial is one way to do so. “Cooperating” by providing testimony against others is a second and more potent means to the same end. A cynic might even say that the threat of very long prison terms serves, in the contemporary justice system in the United States, many of the purposes that judicial torture served in earlier systems: encouraging confession and penitence, and inducing witnesses to tell what they remember, and in some instances even to tell what they do not remember. In most actual drug-sentencing schemes in the United States, of which the Federal scheme is typical, sentencing is primarily based on the identity and quantity of the drug involved, with adjustments for such factors as the role of the individual in the transaction, the presence of firearms, and dealing within designated school zones. At the Federal level, and in some
VOLUME 3
NUMBER 3
2004
PP 435-440
436
KLEIMAN
states, it is also the case that the absolute level of sentencing for drug offenses tends to be fairly high, with multiyear sentences not uncommon. That approach produces the results reported by Sevigny and Caulkins: a large population of incarcerated dealers, some of them rather minimal participants in the markets, not very many of them either notably violent or notably high-ranking in the illicit traffic. Would some alternative sentencing system work better, measured against some set of objectives? In considering an optimal sentencing strategy for drug offenses (a project that necessarily involves significant abstraction from the complicated process in which enforcement agencies, prosecutors, judges, and legislators among them determine who goes to prison and for how long), there are two questions to answer: 1. How many people should be in prison for violating the drug laws (as an absolute number or as a proportion of the total population behind bars)? 2. How should that (hypothetical) allocation of prison space to drug law enforcement be allocated in turn among the persons being arrested, prosecuted, and sentenced for those offenses? Doing that optimization would require a more careful elucidation of our objectives in incarcerating drug offenders than is usually carried out in the debates on the subject. Why do we lock up (so many) drug dealers? There are three obvious answers, two of them based on the consequences of incarceration and the third on considerations of deservingness. The most straightforward answer to the question “Why imprison drug dealers?” is “to prevent drug abuse by preventing drug dealing,” via some combination of deterrence and incapacitation. The underlying purpose of the drug laws is to reduce drug abuse and the bad consequences that flow from drug abuse to consumers and others: physical and psychological health damage from the drugs themselves; accidents, crimes, and derelictions of responsibility as a result of intoxication; and addiction. Insofar as raising the prices of illicit drugs is a goal of sentencing, then imprisoning high-earning participants in the drug markets should, other things equal, be more cost-effective than imprisoning lower level participants, simply because high-level participants are likely to place a higher subjective monetary valuation on a spell behind bars. If we accept the idea that earnings is one measure of culpability, then just-deserts considerations will tend to track the decisions that would be made on the basis of the market-restriction objective. However, the logic of replacement-deterring or incapacitating one
REACTION ESSAY
437
dealer opens up a market niche for another dealer-suggests that deterrence and incapacitation have limited capacity to shrink the drug markets, especially once a drug achieves mass-market penetration (Kleiman, 1997). The experience of the past quarter-century, which has seen the inflationadjusted prices of cocaine and heroin prices fall more than 80% in the face of an expansion of more than tenfold of the number of dealers in prison, bears out that concern. The idea that incarceration levels are important determinants of drug prices (Reuter and Kleiman, 1986) is hard to reconcile with that history. If incarceration cannot substantially raise prices or limit their physical availability, then it cannot reduce drug consumption, which means that it cannot reduce the side-effects of drug consumption in terms of health damage, accidents, and crimes. The second consequentialist reason for incarcerating drug dealers is to prevent non-drug crime-especially violent crime, either violence incident to the drug markets themselves or violence that the incarcerated dealers would otherwise commit-and to protect neighborhoods from the disorder resulting from dealing activity. Various “enhancements” in the federal drug sentencing system-for example, for using a firearm-reflect an attempt to use drug sentencing to reduce non-drug crime. But that consideration is, in practice, distinctly a secondary one. The third objective of a sentencing scheme should be to punish offenders proportionately to the wrongfulness of their actions, somehow defined. Those who engage in crime on a large scale, those who profit greatly from their crimes, those whose criminal activity includes directing the criminal activity of others, those who commit violence in addition to carrying out illicit transactions, those who corrupt or intimidate the enforcement process, and those whose transactions victimize especially vulnerable groups (such as children) are generally taken to be more culpable than those whose actions do not have those features. This, too, is reflected in actual drug sentencing systems, but again as a set of adjustments to a base sentence based primarily on the type and quantity of the drugs involved. (Quantity may also reflect culpability, but it does so very imperfectly because a minor participant in a large transaction can wind up taking a heavy fall.) How do the characteristics of drug offenders as, they present themselves for sentencing (or for prosecution, or for enforcement targeting) map onto those objectives? That is the question the myth of the kingpin is designed to answer, or avoid answering. The kingpin-the violent head of a major trafficking organization, as played by A1 Pacino-is imagined to be both the most culpable offender and the offender whose incarceration will best serve the goals of reducing drug abuse and reducing the violence incident to the
438
KLEIMAN
drug markets. On this account, the more closely a given defendant resembles a “kingpin,” the longer his term should be. Underlying the “kingpin” idea is the notion that, although all other drug market participants are replaceable, the talents required to be a drug kingpin are so scarce that, if we imprison enough of them, the role will start to go unfilled and therefore the drug markets will shrink. But given the financial rewards of large-scale drug dealing, it seems implausible that the job of “kingpin” will stay vacant for long. The fluid nature of the drug markets casts doubt on the significance of organizational hierarchies, and still more on the significance of the identities of the individuals who head them. In the licit markets, no one would expect that imprisoning the CEO of General Motors would shrink the volume of automobiles sold; it is not obvious that imprisoning the current cocaine kingpin, whether in some source or transit country or in a local market, will have much more impact on the volume of cocaine consumed. If drug sentencing cannot do much to reduce the quantity of illicit drugs consumed, then it ought to aim at some combination of punishing those who deserve it and reducing the side-effects of dealing. But drug sentences continue to depend largely on the identity and quantity of the drug involved. That does not seem to track very closely our intuitions of desert, and it does not obviously create the incentives we would like to see operating on the illicit drug markets. Although replacement effects imply that the volume of drugs sold is largely independent of the sentencing scheme, there is no reason to think that the level of violence surrounding the markets, or the extent to which dealers employ juveniles as apprentices, is similarly invariant. Locking up a violent dealer does not in any obvious way increase the incentive for violence among other dealers; on the contrary, insofar as the capacity for violence has defensive as well as offensive uses, reducing the level of violence in a market by removing its most violent participants ought to be expected to cause the other participants to invest less, rather than more, in acquiring and demonstrating the capacity and willingness to inflict injury. That reasoning suggests that an optimal sentencing system would put more weight than existing systems do on violent (or otherwise noxious) conduct, and correspondingly less weight on the volume of drugs sold. Moreover, as replacement is less of a problem for non-drug crimes than for drug dealing, and as on average incarcerated drug dealers have lower rates of non-drug crime than do persons incarcerated for non-drug offenses (Cohen, 1992), a possible implication of this analysis is that a shift of prison cells from drug dealers to non-drug offenders might increase the overall crime-reduction effect of any given level of imprisonment. That would suggest some shortening of sentences for nonviolent drug dealers and some reduction in the number of such dealers sent to prison at all.
REACTION ESSAY
439
It should be noted, however, that sentencing in practice tracks written sentencing systems only imperfectly; the real system may be closer than the sentencing guidelines to the approach suggested by the above analysis. Because, especially at the Federal level, the base sentences for largescale drug dealing are so long, investigators may not bother to gather, or prosecutors to offer at trial, the evidence that would be required to trigger some of the possible sentence enhancements. Still, in practice, some investigators and prosecutors make a point of trying to “nail” the more violent dealers. Gang-violence crackdowns on the model of Boston’s Operation Cease-Fire rely on the ability of willingness of investigators to “take out” members of designated gangs by developing drug charges against them (Kennedy, 1997). One result of this pattern is that the violence known to-or at least believed in by-investigators and prosecutors may never be documented in court and thus never appear in the criminal history of the defendant. He may therefore show up statistically as a “nonviolent” drug offender. (That point emphasizes the importance of using self-report data, as Sevigny and Caulkins do, to supplement official records.) Still, it seems likely that changing formal sentencing policies might help increase the average propensity to violence of the population of incarcerated dealers and thus the violence-reduction effects of the drug law enforcement effort. But doing so would have costs and other disadvantages as well as benefits. Proving drug and weight requires no more than a lab report and a witness to tie the defendant to the transaction. (If the drugs are found in the defendant’s possession, no witness other than the arresting officer is needed.) Proving that the dealer is a perpetrator of violence, or the leader of an organization that employs juveniles, is much harder. So the aggregate number of cases would likely fall as a result of the proposed shift in sentencing patterns. As case counts are used as workload measures by investigative units and by the budget processes that fund them, that reduction in “productivity” represents a potentially serious problem. Case production would also fall for another reason. Shorter sentences for low-level nonviolent dealers mean less pressure on those dealers to provide testimony against more violent and higher level dealers (and less pressure to accept a plea bargain rather than forcing the issue of guilt to trial). Insofar as prosecutors and investigators are currently (outside the existing sentencing schemes) targeting violent dealers for long incarceration, and insofar as proving violence in a courtroom is harder than knowing about it, some violent dealers who now go to prison for long terms
440
KLEIMAN
would face shorter terms under a sentencing system with an explicit focus on violence. Still, the fact that the existing system can be twisted to good purpose does not imply that, as it stands, it is anything close to an optimal system. On the evidence, it probably is not.
REFERENCES Cohen, Jacqueline The Incapacitation Effects of Incarcerating Drug Offenders. NIJ 139737. 1992 Kennedy, David Pulling levers: Chronic offenders, high-crime settings, and a theory of pre1997 vention. Valparaiso University Law Review. Kleiman, Mark A.R. The problem of replacement and the logic of drug law enforcement. Drug 1997 Policy Analysis Bulletin Issue Number 3. Reuter, Peter, and Mark Kleiman Risks and prices: An economic analysis of drug enforcement. In Michael 1986 Tonry and Norval Morris (eds.), Crime and Justice: An Annual Review of Research, Vol. 7. Chicago, Ill.: The University of Chicago.
Mark A.R. Kleiman is Professor of Policy Studies and Director of the Drug Policy Analysis Program at UCLA. His teaching and research cover drug policy, crime control policy, and methods of policy analysis. Before entering academic life, he worked on Capitol Hill, as Special Assistant to Edwin Land at Polaroid, as deputy director of management and budget for the City of Boston, and as Director of Policy and Management Analysis for the Criminal Division of the US. Department of Justice. He is also the editor of the Drug Policy Analysis Bulletin and the Chairman of BOTEC Analysis Corporation, which provides policy advice to governments at all levels on drugs, crime, and health. His books include Marijuana: Costs of Abuse, Costs of Control and Against Excess: Drug Policy for Results. He is currently at work on Getting Deterrence Right: Crime Control Policy in the Light of Game Theory and Behavioral Economics. Before moving to UCLA, he taught at Harvard’s John F. Kennedy School of Government, where he received his master’s and Ph.D. degrees in public policy. His website is http:// markarkleiman.com.
