Document View

The case against drug legalization rests on the empirical premise that drug use would increase in the absence of drug control laws. De Marneffe discusses the benefits and drawbacks of heroin legalization. Although it could bring momentary pleasure and relief, habitual heroin use could further dim the life prospects of young people in poor communities, and would result in drug abuse.

The case against drug legalization rests on the empirical premise that drug use would increase in the absence of drug control laws. Everything I know about human nature and the effects of drugs supports this. People use drugs because they are pleasurable, and because they are an effective antidote to anxiety, frustration, and feelings of inadequacy. Were drugs legal, they would be socially destigmatized and they would become easier to acquire, cheaper to purchase, and safer to use. Given the genuine psychological benefits of drug use, we can be sure that it would increase were drugs legalized.

The predictable increase in drug use alone, however, does not justify drug laws, nor does it provide a convincing argument against legalization. For one thing, drinking might decrease as a result of increased drug use, which would be a good effect if drinking is more harmful. Even if increased drug use would not have this indirect benefit, drug laws impose substantial burdens. They impose risks of criminal liability on those tempted to deal drugs, especially young men in poor urban neighborhoods; they create incentives for gang membership and gang violence; they cost money in law enforcement; they result in outlaw groups abroad having more money and influence than they otherwise would, with dangerous political consequences; they increase the risk of overdose and other health risks commonly associated with drug use; and, of course, they make an enjoyable recreational activity illegal. Given these burdens, the benefits of keeping drug use down must be substantial in order to justify drug control. What are these benefits? This depends on the drug in question. Here I focus on heroin.1

The strongest argument against heroin legalization, in my opinion, is its predictable effect on the life prospects of young people in poor communities. A person who is illeducated, whose skills are not in demand, who does not feel admired or respected by society, who has no clear path to social success or financial security, is likely to feel self-doubt and frustration in large measure. Such feelings can be relieved by heroin. Given the intense pleasure and temporary relief that heroin provides, many of those who try it in any community, rich or poor, are likely to use it again and some are likely to develop a habit. Why would this be bad? Habitual heroin use typically lowers a person's expectations of himself and decreases his concern with what others expect of him; it typically weakens a person's motivation to accomplish things and to meet his responsibilities to others. A general increase in the availability of heroin would therefore increase the risk to children of inadequate parenting, and would make it more likely that young persons disposed to drug abuse will not do what they need to do in order to get a decent education, develop good work habits, develop relationships with mentors, and so on.

By "drug abuse" I mean drug use that has negative consequences for the user or others that are sufficiently bad to make this use either imprudent or irresponsible. Poor people are not uniquely susceptible to drug abuse. Life is difficult, especially in adolescence. Heroin brings pleasure and relief. We can therefore predict that adolescents in every economic group would begin using it regularly if it were legalized. Some, perhaps most, would "mature out" of a habit, but drug abuse during crucial stages of a person's mental and emotional development, whether by oneself or one's parents, will typically have lasting consequences that are difficult to correct. The greater availability of heroin might distract some young people from problem drinking, but the impact of heroin abuse on a person's development is arguably more severe because the degree of indifference it produces toward tasks and mentors is arguably greater. If it is, then the impact of legalization on poor communities is especially worrisome because an increase in drug abuse there would further clog already narrow opportunities for achievement and success.2 Legalization would no doubt pose less of a threat were wealth distributed more equally, but this is no argument for legalization until it is.

Perhaps I am wrong. Perhaps the legalization of heroin would not result in an increase in drug abuse. Perhaps an increase in drug abuse would not have a disproportionately negative impact on the poor. I am convinced, however, that these effects would occur. So I wish to consider the justifiability of heroin laws on these assumptions in order to make a philosophical point about the morality of drug control. I argue that if drug laws are morally objectionable, this is not because of some general principle of political morality such as Mill's principle of anti-paternalism3 or Dworkin's principle of neutrality;4 it is because the burdens they impose on individuals outweigh the benefits they bring to them as a matter of fact.

A coercive government policy violates a person's moral rights, in my view, if and only if it violates a principle I call the burdens principle. This principle prohibits the government from adopting a coercive policy that burdens someone unless (a) someone would bear a burden in the absence of this policy that is at least nearly as great, and (b) there is sufficient epistemic reason for government officials to believe that someone would bear this burden in the absence of this policy. To say that a person is burdened by a policy is to say there is good reason for her to prefer her situation without this policy, that her situation is objectively better in some way without the policy. To say a person is burdened by the absence of a policy is to say there is good reason for her to prefer her situation with this policy in place, that it objectively improves her situation in some way.5 A coercive government policy violates a person's moral rights, then, if the reasons for her to prefer her situation in the absence of this policy have substantially greater objective moral weight than any reason a person has to prefer her situation under this policy.

A natural objection is that this account of rights rules out all criminal laws since the burden imposed on those who are caught and punished for breaking a criminal law is generally worse than the worst burden someone would bear in the absence of this law. The burden of being imprisoned seems worse, for example, than a greater risk of being robbed. However, the situation relevant to assessing whether a criminal law violates the burdens principle is not the situation of the person who has been caught or who is already in prison; it is the situation of the person contemplating doing something illegal. With respect to laws against theft, the question is whether the reasons we have to want our possessions made more secure by law are at least as weighty as the reasons a person has to want the legal freedom to take from others whatever he wants when he wants it. In a society in which no one must steal to survive, this is true, and laws that prohibit theft therefore satisfy the burdens principle.

In my opinion, current heroin laws violate our moral rights because they license penalties for early offenses, and this makes the risk of imprisonment too great for those tempted to buy or sell heroin. Furthermore, some penalties are draconian and unfair because disproportionate. Current policy, however, is not the only form that drug control might take. Penalties for manufacture, sale, and possession of heroin might be gradual and proportionate: no one serves prison time for first or second offenses, provided he meets the conditions of probation, and prison terms after multiple offenses remain proportionate in length to the potential harm involved.6 Suppose a system of penalties like this one, if reliably enforced and effectively targeted at assets, would reduce drug consumption roughly as much as current policy does. This is not utterly fanciful, since criminologists take seriously the possibility that softer penalties-reliably enforced-are more effective than harsh penalties enforced erratically.7 If current laws reduce heroin abuse substantially, then, I think, a moderate system of penalties like this one might better satisfy the burdens principle and so respect our moral rights.8

Of course it is commonly objected that drug laws "don't work," in which case an alternative policy that works just as well will not work either. The "don't work" objection, however, might mean at least three different things. It might mean that drug laws do not eradicate drug use. It might mean that they do not decrease drug use at all. Or it might mean that they do not decrease it enough to justify the burdens they impose. Certainly drug laws do not eradicate drug use. Neither do property laws eradicate theft. This is no argument for their abolition. On the other hand, no informed student of drug policy supposes that drug laws do nothing to lower drug use. Since they make drugs scarcer, more expensive, more dangerous to use, and reinforce social norms against use, they predictably reduce consumption, a prediction supported by the best available evidence.9 The "don't work" objection should therefore be taken seriously only if it means drug laws do not reduce consumption enough to justify the burdens they impose. This may be true, but if the alternative policy I have sketched would reduce heroin abuse in poor communities roughly as much as current policy does, and current policy does reduce this drug abuse substantially,10 then I think it is not true.

The burden of criminal liability is not the only burden imposed by drug laws. Among other things, they make an enjoyable recreational activity more difficult, costly, and dangerous. A complete defense of any drug law must consider the greatest total burden it imposes on any individual and explain why this burden is justified by its greatest benefit to someone. My aim here is not to provide a conclusive defense of drug laws. It is to explain how they might be justified as compatible with respect for our moral rights. Drug laws are compatible with our rights, I maintain, if they satisfy the burdens principle. But some may think that drug laws are wrong because they are paternalistic or fail to be neutral toward different conceptions of a good life. Thus I should explain why I think a policy that satisfies the burdens principle is not open to criticism on these grounds.

I have given two arguments in favor of heroin laws. The first is that they would reduce the risk to children of inadequate parenting by reducing the likelihood that their parents will abuse heroin. The second is that heroin laws would reduce the risk to young people of developing a habit that undermines their own opportunities for future achievement and well-being.11 The first justification is not paternalistic in the way liberals have traditionally opposed, since it limits the liberty of adults to prevent harm to children. The second justification, however, may seem objectionably paternalistic in aiming to deter young adults from heroin abuse that undermines their own life prospects. Heroin addicts commonly oppose drug legalization when asked about it.12 This is noteworthy since the paternalism Mill opposes in On Liberty consists in policies that limit a person's liberty for his own good against his will,13 and if a person supports a coercive policy, then the coercion it involves is not against his will in the relevant sense. But I will assume for the sake of argument that the policy I am considering can be fully justified only as benefiting some young adults who oppose it, and that it is therefore paternalistic in the way Mill opposed.

Paternalistic interference with a person's liberty is wrong in my view only when it violates the burdens principle with respect to that person. A policy violates the burdens principle with respect to a person only if the (objective) reasons for her to want the government not to limit her liberty decisively outweigh every (objective) reason she has to want the government to limit it.14 When this is not the case-when a person's reasons for wanting the government to limit her liberty in some way outweigh her reasons for wanting the government not to do so-paternalistic interference with her freedom does not violate the burdens principle, at least not with respect to her, and so does not violate her moral rights. My argument for heroin laws assumes that the reasons why some individuals want the manufacture, sale, and possession of heroin to be illegal do outweigh their reasons to want it to be legal. If so, then heroin laws do not violate their rights even if paternalistic.

A persistent objection is that if this kind of paternalism is permissible, there is no limit to the kind of paternalism the government may engage in. If the government is justified in prohibiting heroin because its habitual use is bad for the user, it must also be justified in prohibiting alcohol, tobacco, gambling, and fatty foods. Where will it all end? I will remain neutral here on the justifiability of alcohol prohibition,15 but I will concede that if the burdens principle permits the government to restrict the possession and sale of every product that provides an opportunity for self-destructive use, then it licenses too much government interference with our liberty. I should now explain, then, why I do not think the burden's principle implies this.

I believe a person's reasons for wanting the legal freedom to buy fatty foods decisively outweigh his reasons for wanting not to have this legal freedom, and that the burdens principle therefore prohibits paternalistic interference with a person's freedom to buy fatty foods. To be sure, there are good reasons for a person who is tempted to eat himself into an early grave to want to have greater difficulty in acquiring unhealthy foods. But since the length of a person's life is not the only factor in happiness, and eating, preparing, and sharing fatty foods is part of what makes life enjoyable, and a general legal discretion over what foods to buy is valuable in other ways, I believe that paternalistic interference with the freedom to buy fatty foods is not in fact justifiable. My argument for heroin prohibition would suggest the opposite if our reasons to want the legal freedom to buy heroin were as weighty as our reasons to want the legal freedom to buy fatty foods, and our reasons to want not to have the legal freedom to buy fatty foods were as weighty as our reasons to want not to have the legal freedom to buy heroin. But I think this claim about the relative weight of reasons is incorrect. My argument for drug prohibition is that drug legalization would further dim the life prospects of young people in poor communities. The main reason to prohibit the sale of fatty foods is to discourage behavior that tends to shorten a person's life. I think the reasons for a young person to want to be deterred from engaging in conduct that closes important opportunities to achieve the good things in life going forward have objectively greater moral weight than the reasons a person has to want to be protected from engaging in conduct that will shorten his life. This is because a personwill have already had most of the opportunities he will have to achieve and enjoy these good things by the time that eating fatty foods will have the effect of shortening his life. This is not to deny that obesity closes desirable opportunities, but these are not so numerous or important, it seems to me, as those closed by a failure to become educated and to develop self-discipline in adolescence and early adulthood, which I assume is a common consequence of habitual heroin use. It thus makes sense, in my view, to hold that while the burdens principle prohibits fatty food paternalism it permits paternalistic drug laws.

Here it is perhaps natural to ask: on what principle do I distinguish permissible from impermissible forms of paternalism? The formal answer is, on the basis of the burdens principle. But, some may wonder: on what more specific principle do I distinguish those forms of paternalism that violate the burdens principle from those that do not? The answer is, none. Judgments about whether a government policy violates the burdens principle rest on judgments about the relative weight of the relevant reasons. Judgments of this kind are not based on further principles; rather they are the basis on which more specific principles of liberty are justified. How, then, does one know whether a particular form of paternalism violates the burdens principle? By thinking carefully about the relevant reasons, checking for consistency and coherence, and discussing the matter with others. If this seems insufficient to ground a knowledge claim about our rights, it is important to recognize that no claim about our rights can be justified without relying on this process of reflection at some point. Thus while one might justify a rights claim by reference to whatever principle of liberty would be chosen in John Rawls's "original position,"16 the argument for this principle from the original position must itself rely on assumptions about the relative weight of individuals' interests in liberty and other goods that are justified by this same process of reflection.

If someone's reasons for wanting the government to prohibit him from buying fatty foods have at least as much moral weight as any reason he has to want to be free to do so, then, it is true, the burdens principle does permit this form of paternalism, assuming it could be adopted and enforced at a tolerable cost to others. Some may conclude the burdens principle permits too much paternalistic interference and it is therefore a poor account of our rights. But on this assumption about the relative weight of reasons, it is not possible to identify anything wrong with this form of paternalism. If no one's reasons to want the government to reject a policy have greater moral weight than anyone's reasons to want it accepted, then no one has a decisive complaint against this policy, and if no one has a decisive complaint against a policy, it does not violate anyone's moral rights. It is thus only on the assumption that a form of paternalism does violate the burdens principle-whether it limits our freedom to buy fatty foods, cigarettes, or chances to win at the casino-that it makes sense to think it is wrong.

I have just explained why I do not think the paternalistic character of drug laws provides a decisive reason for objecting to them. Another general principle of political morality that drug laws might be thought to violate is the liberal principle of neutrality toward conceptions of a good life. Drug laws may seem to violate this principle because they make it more difficult to pursue certain conceptions of the good-those conceptions of good that endorse recreational drug use. But the principle of neutrality cannot be understood to rule out policies that make some conceptions of a good life more difficult to pursue as a matter of fact since every defensible system of laws will make some conceptions of the good more difficult to pursue: those that require a system of slavery, for example. The principle of neutrality therefore must be interpreted to rule out only policies that cannot be fully justified by neutral reasons. Any policy that satisfies the burdens principle can be justified by neutral reasons, I would argue. So heroin laws do not violate the principle of neutrality, properly understood, if they can be justified in the way I propose.

Some advocates of neutrality may, of course, oppose all forms of paternalism, but it is important to make a distinction here between those kinds of paternalism that clearly violate the principle of neutrality as intuitively understood and those that do not. Among the paternalistic reasons that seem clearly non-neutral are "perfectionist" reasons: the reason to prohibit an activity or pursuit because it is intrinsically bad, for example, or degrading, ignoble, or worthless. Not all paternalistic reasons are perfectionist in this way, and some that are not seem relatively neutral, like the paternalistic reason of safety to require motorists to wear seatbelts. The argument I have given for heroin laws is paternalistic but not perfectionist. It holds that drug laws are justified by the negative effects the easy availability of drugs will have on the life prospects of the least advantaged. This is a neutral reason in my view because improving the life prospects of the least advantaged is a neutral goal. It is a "political value" in Rawls's sense, one that might be appealed to in defending drug control while fully adhering to his idea of public reason.17

I have just explained why I do not think heroin laws-justified in the way I have proposed-are wrong in being paternalistic or non-neutral. Some may think they are still wrong in sacrificing the liberty of the many for the benefit of the few. My argument does not suppose heroin laws are to everyone's benefit. It does not suppose that most people who use heroin are likely to abuse it. Nor does it suppose that heroin poses such a serious threat to a person's autonomy, by short-circuiting his capacities for rational self-governance and thereby turning him into a zombie, that the protection of everyone's autonomy requires its prohibition. Indeed, my argument for heroin laws allows that they make some people worse off: those recreational users for whom pleasure is a good reason to use it, who would not abuse it if they did use it, but for whom the price, difficulty, and danger of its use is increased sharply by its illegality. For this reason it would be inaccurate to criticize my argument as involving an illegitimate generalization from worst-case scenarios. To the contrary, it supposes that the genuine risk to some individuals due to the substantial losses that would accompany an increase in heroin use suffices to justify limiting the liberty of everyone. But this may seem to make my argument even worse. I should therefore explain why I do not think it does.

The burdens principle is individualistic in nature, not aggregative. It requires that we evaluate policies by making a one-to-one comparison of the burdens on individuals. Only if the burden some individual would bear in the absence of a policy is great enough to justify the worst burden this policy would impose on anyone can this policy be justified. Aggregative considerations properly come into play only when the worst burdens that individuals would bear with and without this policy are roughly equivalent, and the numbers must be appealed to in order to break a tie.

Individualism cuts in two ways against aggregative arguments. If individualism prohibits enslaving a minority to benefit the majority, it also permits restraining the majority in other ways to benefit the minority, at least in principle. The mere fact that a policy burdens more people than it benefits is no objection; a policy is morally objectionable only if the burden it imposes on the many is at least nearly as great as the benefit it provides the few. Anti-discrimination laws, for example, limit the liberty of more people than they directly benefit, but this does not mean that they are wrong. They are permissible because the burden of being discriminated against is substantially worse than the consequent burden of having less associational freedom.

Consider in this connection how the burdens principle bears on the common objection that drug laws are costly. Suppose that heroin laws cost each of us an average of $50 a year in police, court, prison, and military expenditures, and foreign aid. Whether this is a decisive objection depends on how this cost is distributed and on whether the policy improves the situation of someone enough to justify this cost to everyone. To assess this from an individualistic perspective, one asks whether the benefit to someone is great enough to justify the loss to each of us of whatever money an equitable tax would impose on us, in addition to whatever other burdens this policy imposes. We do not aggregate here in comparing benefits and burdens; we compare the total burden on each individual one to one. If the reduced risk to someone of losses associated with drug abuse is worth at least as much to her as the amount that any individual is required to pay, then this expense is compatible with the burdens principle and so with the individualism essential to taking rights seriously, provided that it does not make the total burden on anyone too great.

One might still argue on purely individualistic grounds that drug laws unjustifiably deprive some individuals of the opportunity to enjoy heroin legally. The enjoyment of life is a good thing, and moderate drug use promotes it. But I think the government is justified in restricting this particular source of enjoyment, given the existence of so many other sources, if doing so is necessary to secure goods of greater moral importance. Successful emotional and intellectual development in adolescence and early adulthood are goods of greater moral importance. So if prohibition would substantially reduce the risk to someone of losing these goods, this can justify the consequent loss of this legal opportunity for enjoyment.

My argument against heroin legalization rests on a number of empirical assumptions. First, that legalization would result in an increase in heroin abuse with a disproportionate negative impact on the opportunities of individuals in poor communities. Second, that heroin use could be kept at its current levels by laws reliably enforced by a proportionate and fair system of penalties that targets assets effectively. Third, that the financial cost of enforcing these laws could be equitably distributed through a progressive tax without substantially reducing anyone's standard of living. Many other empirical assumptions must be added to provide a conclusive case against heroin legalization. It must be assumed that the risks of gang violence can be kept at acceptable levels by effective neighborhood policing without drug legalization. It must be assumed that the health risks created by drug control can be reduced to acceptable levels by the distribution of public information and sanitary needles. It must be assumed that drug control can be effectively enforced without destabilizing governments in other countries and without greatly increasing the risk of terrorist violence here and elsewhere. And these assumptions are only the beginning.

The fact-dependence of this argument naturally raises the question of burden of proof: how much evidence must one have before being morally justified in supporting drug prohibition? I suppose a person is justified in supporting a coercive policy for a reason only if there is sufficient epistemic justification for her to accept the proposition that provides this reason. Whether this is true of the empirical assumptions I have made here is open to debate. What I have done is sketch an individualistic framework within which drug laws might be justified, consistent with taking our rights seriously. If this framework is sound, and my empirical assumptions are correct, and we are justified in believing them, the government may therefore prohibit heroin without violating anyone's moral rights.

[Footnote]

NOTES

[I thank Andrew Koppelman, Daniel Shapiro, and Leif Wenar for comments on drafts.]

1 I intend my argument here to apply also to morphine, heroin's less popular sibling, with the exception mentioned in note 8, below.

2 Acceptance of this argument, or something like it, provides the most plausible explanation, in my view, of why leaders of poor urban communities typically support drug laws, a fact that casts doubt on the charge that such laws are "racist."

3 See J. S. Mill, On Liberty, ed. Elizabeth Rapaport, Indianapolis: Hackett Publishing Co., 1978, p. 9.

4 See Ronald Dworkin, "Liberalism," A Matter of Principle, Cambridge, MA: Harvard University Press, 1985, p. 191.

5 Some may find it unnatural to say that a person is burdened by the absence of a coercive policy, but since I think that individuals would be burdened by the repeal of some coercive policies, like laws against theft, I think it makes sense to say that individuals can be burdened by the absence of a policy.

6 In allowing penalties for possession I do not directly disagree with the claim that use should not be criminalized, which Douglas Husak argues ("Four Points About Drug Decriminalization," Criminal Justice Ethics, 22 [this issue, 2003], pp. 3-11). The goal of the drug prohibition I defend here is not to reduce use per se, but to reduce certain kinds of abuse by reducing drug availability and by making drug use more burdensome legally and socially. If drug abuse would be substantially reduced in these ways, by a gentle system of penalties for possession combined with stiffer penalties for manufacture and sale, I would defend penalties for possession on this ground. By a "gentle" system of penalties, I mean one that gives everyone convicted the probationary opportunity to avoid imprisonment by accepting some form of treatment, performing community service, or both. For some considerations that weigh against legalizing possession completely, see James Q. Wilson, "Heroin," Thinking About Crime, New York: Basic Books, 1975, pp. 144-52.

7 See, for example, James Q. Wilson & Richard J. Herrnstein, Crime and Human Nature, New York: Simon and Schuster, 1985, p. 62.

8 Another objection to current policy is that it makes illegal the medical use of heroin for the treatment of pain. This objection is compelling only if heroin is more effective than morphine. If not, then the current policy of permitting the medical manufacture, sale, and possession of morphine suffices to meet the medical needs of patients. If heroin is more effective, then it should be treated by the law in the same way morphine is now. For an expression of doubt about the greater effectiveness of heroin, see John Kaplan, The Hardest Drug, Chicago: The University of Chicago Press, 1983, p. 6.

9 Three facts are commonly cited as evidence that heroin use varies in proportion to availability, which, it is assumed, drug prohibition reduces to some degree. First, the percentage of American troops who used heroin in Vietnam, where it was easily available, is much higher than the percentage of the same group who used the drug before and after their service there. Second, opiate use is much higher among medical professionals, who have much greater access to opiates, than it is in the general population. Third, alcohol consumption fell substantially during Prohibition. All three arguments are endorsed by Erich Goode, Drugs in American Society, 4th ed., New York: McGraw Hill, Inc., 1993, pp. 375-76. See also Kaplan, The Hardest Drug, p. 113, for an endorsement of the first two arguments.

10 It is sometimes said that since heroin is already available in poor urban communities, drug laws do little to reduce drug consumption there, but this reasoning is flawed in at least three ways. First, while heroin is relatively available in some poor urban communities, it is not nearly as safe, plentiful, and affordable as it would be if sold legally along with liquor or candy at the corner store (Kaplan, The Hardest Drug, pp. 123-24). Second, although heroin is now quite available in some poor urban communities, it is much less available in others (id., p. 88). Third, even in those poor communities in which heroin is now quite available, the expense, unreliability, danger, and inconvenience of supporting a habit remain a strong incentive to stop using it (id., p. 125).

11 Note that in both cases the relevant burden is the greater risk of loss under a policy of legalization, and not the actual loss that results. If the relevant comparison were of the actual losses from prohibition and legalization, then drug laws would surely violate the burdens principle because, to make just one comparison, death from drug-related gang violence is worse than inadequate parenting.

12 Wilson, "Heroin," p. 135.

13 Mill, On Liberty, p. 9.

14 Note that I do not say that the burdens principle is violated if the reasons someone has to want the government not to limit her liberty outweigh every reason she has to want it to do so. I think this claim about reasons holds for many recreational drug users, but the burdens principle is violated only if their reasons to want drugs to be legal decisively outweigh everyone's reasons to want it to be illegal, a claim that I believe to be false.

15 Perhaps it is worth observing here that while alcohol prohibition might be permitted by the burdens principle, the argument I have given for heroin laws does not alone warrant the conclusion that it is. This is because the legal prohibition of a drug is likely to be far more effective in reducing abuse where its use is already generally believed "wrong." This is now the case with heroin but not with alcohol. Although the (greater) deterrent benefits of reducing heroin abuse might now be sufficient to justify continuing heroin prohibition, the (lesser) benefits of reducing problem drinking might not be sufficient to justify a return to alcohol prohibition, even if drinking is equally harmful. The reason I remain neutral on the justifiability of alcohol prohibition is that I suspect that problem drinking is more harmful on average than heroin abuse is, especially to others.

16 See John Rawls, A Theory of Justice, Cambridge, MA: Harvard University Press, 1971, pp. 11-15.

17 See John Rawls, Political Liberalism, New York: Columbia University Press, 1993, pp. 223-24. For another argument for drug prohibition that would be supported by Rawls's liberal political values on different empirical assumptions, see Samuel Freeman, "Liberalism, Inalienability, and Rights of Drug Use," in Pablo De Greiff, ed., Drugs and the Limits of Liberalism, Ithaca: Cornell University Press, 1999, esp. pp.128-30.