Note: Subscriptions to the FAS Drug Policy Analysis Bulletin may be requested by sending email to fas@fas.org
The most striking characteristics of the US response to illicit drugs in the last decade have been its scale and its focus on punishment. In terms of its political support, this punitiveness is reflexive and ideological, but those who would argue against it are handicapped by the fact that there is almost no research on its effects.
Drug control is a $30-35 billion government program in the mid-1990s; more than half is spent by state and local governments. The focus on punishment is reflected in the allocation of that money. Three quarters of the total is spent on apprehending and punishing drug dealers and users; treatment receives two thirds of the remainder. State and local governments are even more enforcement oriented than the federal government.
The total punishment imposed for drug law violations has increased massively since 1981, when the concern with cocaine became prominent. In particular, the number of commitments to state and federal prison have risen approximately ten fold over the same time period, vastly more than any measure of drug use or sales. By 1992, there were approximately 390,000 people in prison or jail serving time for selling or using drugs (see Table 1).
Table 1: Trends in Drug Enforcement, 1980-1994
| 1980 | 1985 | 1990 | 1994 | |
| Drug
Arrests |
581,000 | 811,000 | 1,090,000 | 1,350,000 |
| Heroin and cocaine only | 70,000
(12%) |
240,000
(30%) |
590,000
(54%) |
635,000
(47%) |
| Distribution only | 104,000
(18%) |
192,000
(30%) |
345,000
(31%) |
370,000
(27%) |
| Inmates
[Total] |
31,000 | 68,000 | 291,000 | 392,000 |
| Local Jails | 7,000 | 19,000 | 111,000 | 137,000 |
| State Prisons | 19,000 | 39,000 | 149,000 | 202,000 |
| Federal Prison | 4,900 | 9,500 | 30,500 | 51, 800 |
Sources: Uniform Crime Reports, Correctional Population in the United
States
Note: Local jails population figures are author's estimate. Inmate totals cover drug-defined offenses only and exclude probation/parole revocations.
At the state level, many persons are being imprisoned
for drug possession felonies. In 1992 50,000 were
sentenced to state prison for non-distribution offenses,
mostly simple possession; some unknown share of these
are convictions of drug distributors when the prosecutor
could only find evidence of possession or the case was
bargained down from distribution to possession. Crude
calculations at the national level suggest that a typical
heroin or cocaine dealer faces about a 10 to 15 percent
chance of being incarcerated in the course of a year.
The Product of Punishment
Put aside the divisiveness and intrusiveness of this heavily punitive orientation. What has it accomplished in terms of reducing drug-use and related problems? The results are discouraging.
Toughness should raise prices, make drugs less accessible and reinforce messages that drugs are disapproved of and harmful, leading to less drug use and fewer drug-related problems. At the broadest level, all of this has been achieved: illegal drugs are remarkably expensive, not universally accessible, and generally feared. They have been that way for two decades, going back to well before the massive increase in enforcement effort. Marijuana is a cultivated weed like tobacco, but whereas a cigarette costs, even with excise taxes, hardly 10 cents, an equivalent amount of marijuana costs $5 or more. Cocaine, a mass-processed agricultural good, is vastly more expensive than gold [$1,500 vs. $350 per ounce].
However, notwithstanding sharply increased stringency of enforcement, prices are declining, many of the young see drugs as quite easy to get, and the fear of the mostly widely used drug (marijuana) is falling. Cocaine and heroin prices have fallen steadily since 1981; by 1995, after adjusting for inflation, they were only about one third of their 1981 levels. For marijuana, prices rose steadily and substantially from 1981 to 1992 and then fell in the next four years back to their 1981 level, adjusted for inflation. (Neither the apparent increase nor the apparent decrease has been adjusted to reflect the unmeasured variation in potency.) Jon Caulkins has found that crack cocaine, singled out for tough sentencing, both at the national level and in some major states (e.g. California, where possession of small quantities is subject to mandatory state prison sentence) is no more expensive at the retail level than powder cocaine in terms of price per pure milligram of cocaine. Nor has additional enforcement reduced availability. Between 80 and 90 percent of high school seniors report that marijuana is available to them, a figure that has been stable for two decades. The comparable figure for cocaine rose through the 1980s and began to decline after 1989; by 1995 it was just slightly below its 1985 levels.
Drug use is estimated to be half as prevalent as in the
early 1980s, but its prevalence in the general population
has been roughly constant for the past five years. The
most proximate cause of the decline seems to be a shift
in attitudes concerning the risks of use of specific drugs.
Though enforcement might influence those perceptions,
there is no correlation between crude measures of
toughness and those perceptions. The number of people
dependent on cocaine and heroin has been fairly stable
over a long period of time, though an increasing share are
being locked up.
The Lack of Enforcement Research
Clearly, a central issue for American drug policy today is assessing the consequences of a highly punitive approach, since that is the policy the nation has chosen. Yet the vast bulk of drug policy related research is done on the effects of treatment and prevention, which account for no more than 25 percent of this nation's public expenditures on drug control. That may be the result of the dedication to punishment; any other program has to justify itself against the suspicion that it is kind to criminals (treatment) or too diffuse (prevention).
The National Institute on Drug Abuse has a research budget of $450 million; much of that goes to the development of a better understanding of the biomedical aspects of addiction. Research on drug enforcement has to fight for its share of the National Institute of Justice's paltry $30 million annual budget. Adding dribs and drabs from elsewhere, twenty million dollars is certainly far too generous an estimate of the funding for research related to drug enforcement. The federal enforcement agencies (e.g. DEA and FBI) neither conduct nor sponsor any research themselves, notwithstanding federal program expenditures of about $10 billion, and their internal capacities to gather and report such basic data as prices and quantities are seriously deficient. Foundations fund treatment and prevention research but eschew the enforcement issue.
Why there is so little research on drug enforcement?
James Q. Wilson has noted the lack of research on crime
control generally. In addition there is a curious confluence
of liberal and conservative interests. Those who support
tough drug enforcement see little to gain from evaluation;
Peter Rossi's "If you don't like a program, evaluate it" is a
highly relevant cliche here. Those who dislike punitiveness find the whole effort distasteful enough that they
simply want nothing to do with it, preferring to focus on the
programs in which they believe, namely prevention and
treatment.
Policy Reform and Drug Enforcement Research
The lack of enforcement research enormously handicaps the efforts to bring about major change in drug policy. Claims that too much is spent on punishment simply have no empirical support, relying instead on general notions of justice and impressions of effectiveness.
At a minimum, it would be useful to say whether longer prison sentences, more seizures or more intensive money laundering investigations can increase prices or reduce availability and what affect these changes would have on drug use by current and prospective users, and on drug related problems. There is not a single empirical study that attempts to answer that question. The closest one gets is a paper of 25 years ago which found that higher prices for heroin increased property crimes in Detroit. Lately, some papers have estimated the price elasticity of demand for various drugs and various populations, but that is just a baby step.
The failure of cocaine and heroin prices to rise with tougher enforcement is a major analytic and policy puzzle. Declining demand; reduced labor market opportunities for aging drug user/sellers; a decline in violence engendered by few new entrants and lower profit margins; and the locking up of criminal users are just some of the possible factors contributing to this. None have been subject to systematic examination.
This is a complicated and long-term research agenda. With $25 billion, more or less, being spent to catch and punish drug traffickers, and no end to the drug problem in sight, this research is a critical element for anyone interested in improving the effectiveness, as well as the humaneness, of our drug control efforts.
Thoughts on the Medical Cannabis Issue
Spurred by the passage of the medical marijuana propositions in Arizona and California, the Office of National Drug Control Policy has commissioned the Institute of Medicine to review the literature on the possible medical utility of whole cannabis (as opposed to the currently available THC pills) and to consider what future research, if any, should be conducted. At the same time, the Director of the National Institutes of Health convened a public review of NIH's research options in the area.
Public discussion to date has largely ignored two
important points: the potential for substantial benefits from
clinical trials even if their results are negative, and the
importance of developing procedures for inhaled
administration of cannabis vapors that do not involve
burning the plant material and breathing the smoke.
The Case for Research
A major impediment to such research in recent years has been the view of experimental biomedical researchers that smoking cannabis is not the next most promising therapy for treatment of glaucoma, the nausea accompanying chemotherapy, and other proposed indications. Even if this belief is correct, it should not be dispositive.
Assume for the sake of argument that the results of clinical trials were entirely negative. There would still be great value in providing convincing evidence that inhaling cannabis vapor does not help one or more of the conditions or side-effects for which its use has been suggested.
Some physicians, and some seriously ill people, believe that cannabis vapor is medically useful. Those beliefs, if they are false, are harmful. As long as they remain current, and contradicted by nothing but the equivocal statement that no good published research currently provides enough evidence for them to support FDA approval, some sick people will continue to go to dope dealers for their medicine, and law enforcement and prosecution agencies will be forced either to try to punish them or to wink at widespread violation of the law.
If the results are negative, publicizing them should help reduce the incidence of illicit quasi-medical cannabis use. If they are positive, then cannabis can be approved by the FDA (presumably as a Schedule II or III controlled substance), prescribed by physicians, manufactured by pharmaceutical companies, and dispensed by pharmacists. Either way, the market for illicit cannabis would shrink and the doubt about whether a given illicit grower might be acting out of compassion rather than greed would be resolved. The worst case, from this perspective, would be inconclusive findings; even then, it's hard to see how research could make the current situation worse.
It has been suggested that even running the trials would be unethical on human-subjects grounds, since the experimenters have little faith in the efficacy of the experimental drug; comparisons have been made to Laetrile. But since delta-9 THC, the principal active agent in cannabis, has already been approved as a medicine, and given the widespread anecdotal reports of effectiveness, it is hardly unreasonable to think that the inhaling a vapor containing THC and the other active agents present in the plant material might prove helpful, and in fact some physicians believe just that. That their belief does not represent the current professional consensus does not make it unethical to test that belief on subjects who have given informed consent; human-subjects protection should not be allowed to act as a cover for the censorship of research on political grounds.
Moreover, if the subjects are recruited from among those who otherwise would use marijuana illegally and largely unsupervised, then surely they would be safer doing the same thing legally and under clinical supervision. Viewed in this light, the human-subjects issues are much less problematic.
Without new clinical research, there is no escape from
the current situation, which forces officials to choose
between enforcing the law by punishing sick people and
their physicians and allowing its widespread, and
increasingly public, flouting. Sooner or later, the research
will have to be done, despite the lack of commercial
potential, the lack of scientific interest, and the fearsome
political cross-currents. Time to get on with it.
Alternatives to Smoking
Under current circumstances, the medical and quasi-medical use of whole cannabis usually takes the form of smoking "joints." This is cheering to those who see the medical issue as an entering wedge for the repeal of the prohibition on the non-medical use of cannabis, and frightening to those concerned about that very prospect.
It is also transparently, and unnecessarily, bad medicine. Developing alternatives to smoking for getting cannabis vapors into patients' lungs ought to be a research priority, especially since it could proceed without either using human subjects or committing to one or another potential indication for use.
Whole cannabis vapor has two potential advantages (aside from price) over the currently legal THC pill: that it contains a variety of active agents and that it comes in by lung rather than by mouth.
The mix of active agents in the whole plant material may be more effective or have fewer side-effects than pure delta-9 THC. One study [Zuardi et al. in Psycho-pharmacology 76:245-250 (1982)] lends some support to this hypothesis with respect to anxiety as a side-effect, by showing that a combination of THC and cannabidiol (another chemical present in cannabis) is less anxiety-inducing than THC alone. Of course, it might equally be true that the whole plant performs worse than THC in terms of effectiveness or side-effects; the question is entirely empirical, and most of the appropriate work has simply not been done.
Administration by inhalation is clearly superior to oral administration when the symptom being treated is nausea or lack of appetite. In addition the much more rapid action allows patients to "titrate" dosage, by taking a little, waiting a few minutes to see if adequate relief has been obtained, and, if not, taking a little more. This last effect is analogous to patient-controlled analgesia, which has been show to produce better pain relief and lower doses of narcotics than the older practice where the physician writes an order for so-and-so much morphine every four hours. Again, the clinical significance, if any, of the capacity to titrate remains to be studied.
But neither class of potential advantages requires burning the plant material in order to volatilize the active agents. If the heat required for vaporization is supplied externally, say by a resistance-heating element, a flame, or a jet of hot air, THC and its chemical cousins can be vaporized and inhaled without combustion, and thus without the cloud of noxious gasses and respirable particulate matter that constitutes most of the smoke coming out of a joint. The most obvious health risk of chronic marijuana smoking is respiratory damage. But the cannabinoids themselves do not damage the throat, bronchi, or lungs; the smoke does.
Likewise, it is the smoke that is carcinogenic; for some potential medical users of cannabis that hardly matters, either because the course of treatment is too brief to present any substantial risk or because their life expectancies are too short for cancer to be a real issue. But glaucoma patients have normal life expectancies; due to the protease inhibitors, that may soon be true, or almost true, of persons infected with HIV.
It used to be believed by some that water filtration could remove the toxins from cannabis smoke. MAPS, an advocacy group, actually received permission to find out. To the disappointment of the sponsors, the result was negative; the water turned out to take out as much, proportionately, of the active agents as it did of the toxins. (To their credit, they published the results.)
Thus a potentially important research activity is to develop and test non-combustive means of volatilizing the active agents in cannabis for inhalation. The project shouldn't be horribly challenging technically, and can proceed quite independently of studies about the utility of the plant for any specific condition or symptom. It might not even need to involve actual administration to subjects; the topic to be studied is merely the chemical content and temperature of the vapors that emerge from various pieces of equipment.
Splitting the question of the medical uses of whole cannabis from the practice of smoking joints would be useful, both practically and symbolically, to those with an interest in de-coupling the question of medical cannabis from the question of marijuana legalization.
Treatment Works...Well Enough
One reason that drug treatment fares poorly in budgetary terms and gets such a small share of health expenditures is that it does not offer much by way of "cure," as antibiotics do for infections or surgery for cataracts. It also lacks the glamour of exciting breakthroughs in technology, even though there is a rapid increase in understanding of the basic brain chemistry of addiction. Most of those who enter treatment will drop out of their first treatment episode and will do so again a number of times in the following ten years; they will also get very much the same kind of treatment they would have gotten twenty years ago.
Nonetheless, treatment can unambiguously lead to reductions in the social cost of addiction vastly greater than the cost of providing the services. The principal costs to society from drug addiction arise from violent and property crime. While in treatment, the drug-dependent consume a much smaller quantity of drugs; reductions of 80 percent are reported in many studies, though most patients are not abstinent, even while in treatment. But the reduction in drug use, and the lowered expense of that consumption, is sufficient to dramatically reduce crime rates. Even though these reductions are frequently short-lived, treatment's ratio of benefits to costs is very impressive; the studies that support those conclusions are admittedly less impressive, plagued as they are by lack of control groups and selection problems.
Most studies find only modest improvement in labor market performance; a 35 year old heroin addict who has been drug dependent for 10 years, one third of which time he spent in prison (all figures close to the median levels) is not an attractive employee. He may well end up on some form of public assistance, perhaps even Supplemental Security Income, the federal program for those whose poverty arises from a disability. Congress has eliminated drug and alcohol dependence as an eligible disability, but these clients often have other serious disorders that qualify them anyway. Nonetheless, the reductions in crime rates alone, putting aside the large reductions in HIV and other STD risks, make even not very successful treatment a genuine public bargain.
Evaluating treatment programs in terms of their contribution to reducing social damage, rather than the one-year abstinence rates of their alumni, would create incentives for providers to seek out high-social-cost, hard-to-treat clients rather than low-social-cost, easy-to-treat clients. It might also create incentives for programs to address the whole range of risky and harm-causing behaviors rather than concentrating on drug use alone.
None of this is new. It has not been politically persuasive, partly because there is little public support for what is perceived as "being nice to drug addicts." Finding ways of overcoming the gap between the analytically supportable and the politically feasible represents a major challenge. --Peter Reuter