America's longest-running metaphorical war, a campaign against a hidden and even less well defined enemy than terrorism, is the war on drugs. This one also has its insidious domestic threats, its overseas campaign of interdiction and extermination, its potential to foster guerrilla wars and destabilize governments. It too has been supported with little dissent from a Congress where few dare to question the prevailing orthodoxy.

Of course the analogy is misleading. There are huge differences between the threat of drugs and the threat of terrorism, whose very object is the slaughter of innocents. But to point out that obvious distinction is also to underline the excesses of a campaign whose cost in lives, privacy, social damage and political instability easily exceeds the more than $25 billion in tax money that the nation now spends on it every year. More than half of all those sent to federal prison are drug offenders.

Nonetheless, the most significant challenge to that orthodoxy so far–most of it from intellectual and social elites–is a free-market libertarianism that's as ideological and unrealistic, both as politics and policy, as the case for an all-out war. So the issue tends to be vastly oversimplified: the zero-tolerance absolutism of former US drug czar William Bennett versus the libertarian, free-market absolutism of economist Milton Friedman; prohibition with long prison terms even for simple possession versus decriminalization, including, at the margins, regulated commercial sale.

Robert J. MacCoun, a professor of law and public policy at Berkeley, and Peter Reuter, a professor of public policy at the University of Maryland, are certain that there is a third, and better, set of alternatives–more rational, based on experience, less sure of itself–that can thread its way, almost on a case-by-case basis, between the ideological poles and out of the morass in which US drug policy has been stuck.

In part that third way requires doing more–in needle exchanges, safe-use campaigns targeted at addicts and a whole range of non-drug policy issues like better welfare and healthcare. In part it means doing less–particularly through selective, targeted enforcement of prohibitions, shorter criminal sentences and fewer encroachments on civil liberties. MacCoun and Reuter make a sharp distinction between decriminalization and what they call depenalization, which differs from conventional prohibition not in restricting access but in limiting the severity of the penalties, particularly by replacing criminal with civil penalties. (In the case of cocaine, which they regard as far too destructive, they don't favor depenalization but only a reduction of the severe and unequal criminal sentences the United States imposes even for possession.) Nor do they support anything that would lead to commercialization even of soft drugs like marijuana, which they feel would bring–and, in places like the Netherlands, has brought–expanded use.

But their preference, often implicit, nonetheless follows a general European model that seeks overall harm reduction rather than merely a reduction in the prevalence of use, as US policy now does. They acknowledge that total harm reduction–mitigating the overall social costs not only of drug use but of prohibition and the criminal behavior associated with it–is not always an easy calculation. Among other things, calculations need to include measures of total consumption–reduction in heavy use–not just in the number of users. But it's certainly more realistic than measuring the success of policy simply by how many fewer people regularly use some illegal substance.

The implicit, and occasionally explicit, policy preferences in Drug War Heresies seem almost an afterthought next to the huge amount of data that forms the core of this book and that sheds light on almost every aspect of this issue. MacCoun and Reuter have surveyed and analyzed hundreds of studies, past and present, in this country, Europe and Australia, not just on drug policies but on experience with a range of issues that have parallels to this one–alcohol prohibition in the United States, tobacco regulation, legalized gambling, the enforcement of laws against prostitution. The real objective of the book is to document the complexity of drug policies, their often unintended consequences and, more fundamental, the lack of scientific foundation for so much of US policy. The analysis of these data, dispassionately presented in all their complexity, makes this an enormously important book. This is especially true because drug policy is a field where tendentiousness prevails, with the exception of a very few other works, like Mark Kleiman's Against Excess: Drug Policy for Results (1992).

Needless to say, Drug War Heresies is hardly an easy read, much less an easy book to summarize. Nor will either side in this fight find it entirely to its liking. It leaves the standard slogans and clichés–that better policy research on things like marijuana, for example, would send the wrong message–lying in the dust. Excepting only Sweden, most of Europe, as MacCoun and Reuter make clear, is moving away from the punitive model, only rarely toward legalization and more commonly toward a far more realistic, nuanced, "harm reduction" approach not stuck in the puritanical mode that so much of US policy finds itself in. (And even the Swedes, who reject methadone maintenance and needle exchange, provide well-supported treatment and social services to addicts.) That hardly means that policies in the Netherlands, Britain, Spain and Italy, all of which they examine, are beyond question. All of them, as MacCoun said in a recent talk, are flawed in one way or another. But Europe is a rich source of lessons.

At the core of those lessons is the question of trade-offs: How much do the substantial reductions in crime and criminal justice costs (including the cost of police corruption) resulting from any loosening of criminal penalties, plus the benefits of safe-use programs, offset the costs imposed on families, individuals and neighborhoods from increased drug consumption?

And that, in turn, depends again and again on individual circumstances–on the details of the policy and the surrounding culture. The Dutch, for example, appear to have successfully depenalized marijuana possession without terribly significant increases in use, thereby reducing both the costs of incarcerating marijuana users and the associated human costs. In the mid-1980s, when passive depenalization–essentially, nonenforcement of laws against personal possession–became de facto decriminalization, marijuana became commercially available in coffee shops and use did drift up. But even that increase didn't drive up the use of hard drugs or increase drug-related crime. Other than producing an increase in patients seeking treatment for marijuana-related problems and occasional complaints from neighboring retailers about certain coffee shops, say MacCoun and Reuter, "we are unable to document any significant social harms accompanying increased cannabis use."

MacCoun and Reuter make clear that at times harm reduction can go badly wrong. After years of chasing an active heroin scene around its neighborhoods, Zurich established its so-called Needle Park (Platzspitz), thereby concentrating heroin users in one park near the main railroad station, in an effort to minimize petty crime and neighborhood nuisances, and to create a central location for providing health services to addicts. The experiment failed: It drew heroin users from far and wide, and turned the place into a "Hieronymous Bosch vision of a drug hell," which in turn was cited by prohibitionists everywhere as evidence that such ventures never work. But there were also gains: from AIDS outreach, which appears to have driven down HIV-positive rates, and from the efficient handling of medical emergencies. And while there were some notorious gang-related murders, crime rates were surprisingly low. Switzerland had a serious heroin problem before Platzspitz was created, but there is no evidence that overall use of heroin in the country increased as a result of it.

The book's general read of the overall European experience is that it has a lot to tell us about what is feasible. "The Dutch have shown that harm reduction can be used as a principle to guide decisions consistently; [it has] some successes to show and no disasters to hide. Italy has removed criminal sanctions for possession of small quantities of cocaine and heroin without experiencing much greater problems than their neighbors." Swiss trials (begun following the Platzspitz failure) "show that heroin maintenance programs can operate in an orderly and systematic fashion for the benefit of a substantial fraction of the clients." The authors also point out that American experience with the enforcement of prostitution laws indicates that the harms that theoretically follow from vice prohibition can be mitigated–though not eliminated–by selective enforcement. Indeed, despite America's moralistic views about prostitution and adultery, policing of prostitution has much in common with the discretionary policing of drug use in many European cities.

Conversely, however, MacCoun and Reuter also caution against excessive enthusiasm for the contention that regulatory policies are inevitably an improvement over outright prohibition. Recent US experiences with alcohol and tobacco illustrate the power of commercial marketing and the difficulty of maintaining or tightening regulatory controls in the face of that power. The evidence for both of those licit substances shows quite clearly that while "prohibition may cause considerable harm, eliminating prohibition does not mean eliminating drug-related harm." Put briefly, they contend that contrary to the libertarian enthusiasm for such a course, the alcohol and tobacco model has to be approached with a lot of caution. In the case of tobacco, for example, restrictions on promotion, product regulation and taxation have all been greatly attenuated by the industry's strategic use of political contributions and reframing of legal issues (e.g., making promotion of a dangerous product a free-speech issue).

Despite the wealth of research available to help guide drug policy, the tests and calculations–essentially on the harm-reduction principle–MacCoun and Reuter are under no illusion that there's any specific formula by which to evaluate reform proposals. In the end, value judgments still have to be made, weights attached to each element of harm. Politically, moreover, the burden of proof is still on reformers to show why their proposals are preferable to the status quo, no matter how dismal it is. And that's often complex. Much easier, unfortunately, are the simplistic warnings put out by government prohibitionists that any experiment–say, with safe-use programs or even good medical studies on the safety and efficacy of marijuana in reducing the nausea associated with chemotherapy or the loss of appetite of AIDS patients–would "send the wrong signal."

MacCoun and Reuter may overestimate the political obstacles blocking the kind of reform that they clearly seem to prefer. A call for "nonzero tolerance," they write, is tantamount to treason in some circles; but such a call might encourage more humane, less intrusive, less damaging ways of coping with drugs and their harms. They cite the passage of the first initiatives, in California and Arizona in 1996, permitting the medical use of marijuana, which they call "at best sloppy," because those ask doctors to make decisions without adequate scientific evidence. But their book apparently went to press before the wave of recent ballot measures and state laws: medical marijuana initiatives in six or seven other states, state laws liberalizing sterile syringe access and reducing prison terms for drug possession, and California's Proposition 36, passed in the fall of 2000, which requires all those convicted of simple drug possession or drug use to be sent to treatment rather than prison. All suggest that, at least before the terrorist attacks of September 11, the public may have been in a far more tolerant and reformist mood than its elected leaders.

Still, the authors are right that despite polls showing that Americans believe the drug war has been a failure, it's a political standard, not a philosophical or analytic one, that reformers have to meet. And that standard is quite protective of the status quo. The combination of high uncertainty about the outcome of any change; the partial irreversibility of any bad outcomes; and a pervasive tendency for decision-makers to favor the status quo pose steep barriers for reformers. Despite the high number of Americans incarcerated for nothing more than marijuana offenses–an affront to a liberal society's belief in the benevolence of government–reactions to existing policies have not been strong enough for politicians to risk any real reforms. A punitive stasis prevails.

Yet even in the face of such passive resistance, Drug War Heresies should pose a formidable challenge, not necessarily to cause pursuit of the policies and trial programs that MacCoun and Reuter seem to favor–maintenance, reducing the penalties for use of marijuana, more judicious drug law enforcement–but to pay attention to the data, end the misrepresentation of information where it exists and go after it where fear has repressed even research, especially in assessing the consequences and efficacy of existing policies.

More fundamentally, the book may also introduce policy-makers to the relatively novel thought that prevalence reduction and use reduction are not the same. While cocaine prevalence has gone down, they point out, "total cocaine consumption and its related harms have remained relatively stable." It may also make clearer that harm reduction is not simply a flag flown by closet libertarians who are philosophically opposed to all prohibitive drug laws.

At the same time, Drug War Heresies leaves no doubt about the limits of policy–and on that score it's important for a lot of other fields. It's doubtful, as the authors say, that a complete solution to the US drug problem exists. The major differences between the American and European illicit drug situations, they suggest, may be rooted as much in broader societal differences, in the peculiarities of geography or in other policies–in lack of healthcare or unequal income distribution–as in drug law per se and its enforcement. That's particularly true of treatment programs, which, even under the best of circumstances, will only be partially successful. But that hardly eliminates the need for reform, in reducing the severity of sentences and the intrusiveness of drug law enforcement, and shifting to more selective, targeted enforcement.

Such a course, MacCoun and Reuter acknowledge, reflects only their opinion. But they leave little doubt that the evidence indicating a need for major reform has both an empirical and an ethical basis. "To scorn discussion and analysis of such major changes, in light of the extraordinary problems associated with current policies, is frivolous and uncaring." For many reasons this book isn't easy; but for anybody seriously and earnestly concerned about drug policy, it is likely to become indispensable.