Printer Friendly

The Liberal Critique of the Harm Principle.

As conceived by J.S. Mill, the harm principle holds that "[t]he only purpose for which power can be rightfully exercised over any member of a civilised community, against his will, is to prevent harm to others."(1) This essay concerns a narrower formulation of the principle, the claim that the criminal law should never punish conduct that does not harm unconsenting third parties. What I have to say about the harm principle as a limit on the justifiable reach of the criminal law likely holds, in spades, for any version of the principle that immunizes harmless conduct from civil or social sanctions, but I here advance no such claim. My concern is with the old and brilliant debate over morals legislation.(2)

The harm principle holds a central place in liberal theories of criminal law, and indeed in liberal political theory generally. Joel Feinberg goes so far as to say that "[l]iberalism, as I have understood it, is the view that" harm and offense to others exhaust "the relevant reasons for state coercion by means of the criminal law."(3)

My thesis holds that the harm principle is profoundly deficient as an expression of the limits that should restrict society's resort to its most coercive instrument. The harm principle's institutional deficiencies derive directly from describing the criminal law's proper limits in terms of the consequences of individual conduct which is to be immunized against punishment. Serious protection of individual liberty depends, I suggest, on limiting resort to the criminal sanction along procedural or jurisdictional lines, rather than according to the substance of private conduct.

The argument proceeds through four steps. First, I point out that, given the peculiar mission of constraining the popular impulse to punish, the harm principle should not be evaluated as a purely normative description of the criminal laws a just or good society would adopt. Instead, the harm principle should be evaluated as an appeal to individual actors to internalize a norm of tolerance, or as an external restraint to be respected even though not internalized. Second, I suggest that the harm principle is an unlikely candidate for an internalized norm of toleration. This is because the justifications for the harm principle run either too deep or too shallow. By that I mean that perfectionist arguments stressing the value of individual choice and the risk of official error fall short of categorically excluding punishment of harmless wrongdoing, while anti-perfectionist rights theories depend on premises that will never persuade a broad consensus in a free society.

Third, I suggest that the harm principle cannot fulfill the role of an external constraint on political choice. The improbability of justifying the harm principle as an internal norm of course makes it unlikely that any polity will adopt the principle as a constitutional limit on that polity's discretion to punish harmless wrongdoing. But even if the harm principle were realized as a constitutional constraint on state power to punish, the application of an immunity for private conduct based on its apprehended consequences would not provide a serious limitation on the actual practice of punishment. The idea of harm is too vague, too dependent on baseline assessments of private rights, too open to long chains of causal speculation, and too catastrophic in its categorical judgments to give liberty much practical protection.

Liberals should turn instead to content-independent limits on penal legislation. At the legislative level, I suggest a supermajority requirement for penal statutes, and a sunset provision terminating all such enactments unless readopted once a decade. At the judicial level, I propose investing constitutional courts with the power to insist upon even-handed enforcement of the criminal law. These approaches require some constitutional definition of criminal, as distinct from civil, sanctions. They do not, however, require any assessment of the consequences of the private behavior forbidden by a particular statute.

Part IV describes these proposals and explains that however esoteric they might sound, they have two dispositive advantages over the harm principle as expressions of a liberal theory of criminal law. They might actually be adopted, and they might actually promote liberty.

I. The Constitutional Mission of the Harm Principle

In an important paper, Chris Wonnell pointed out that much political theory can be understood as either pure normative philosophy or as applied political philosophy.(4) According to Wonnell, pure normative philosophy "seeks to identify the types of arguments which, if true, would morally justify particular types of legal rules and political actions."(5) By contrast, applied political philosophy "seeks to identify a set of values that would produce morally appropriate results if the values identified by the philosophy were sincerely held by judges, legislators, and other political actors."(6)

Liberal defense of the harm principle might thus be understood in one of two ways. On the applied political philosophy account, the harm principle states a norm of tolerance to be internalized by the polity, especially by its legislators. On the pure normative philosophy account, the harm principle is an external constraint that the polity should accept even if a majority of the polity disagrees with that constraint considered on the basis of first-order reasons.

Given the mission assigned the harm principle by liberals, these are the only fair ways of characterizing the principle. The harm principle is regulative; its objective is to limit society's propensity to punish. Justifying the harm principle by a philosophically compelling moral argument may be interesting, but it can be at most the first step in a liberal program ultimately concerned with the intensely practical problem of overcriminalization. The harm principle has a distinctly constitutional mission, aimed at constraining the ordinary processes of democratic legislation, either by changing the hearts and minds of the citizenry or by the force of constitutional law. Yet liberal theorists have done remarkably little--I am inclined to say nothing--by way of constructing institutional counterparts to the harm principle.

The harm principle might succeed in one of two ways. It might, over time, become accepted by an enduring majority of the citizens as an internal norm of tolerance. In the alternative, it might be adopted as an external constraint on legislative discretion, that is, as a constitutional provision, adopted by a temporary majority as a precommitment strategy to prevent future majorities from violating the principle.(7) In familiar legal parlance, proponents of the harm principle might hope to see it become part of either the written or the unwritten constitution of their society.

If the members of a democratic society believed in the harm principle, and applied it in good faith and with right reason, liberals would see no need for any constitutional limit on a legislative majority's power to impose criminal liabilities. Thus, the harm principle might be thought of as a principle liberals should persuade fellow citizens to endorse, trusting democracy itself to translate popular into official tolerance. J.S. Mill, I think, had this vision of the harm principle; he phrased the harm principle as a limit on the ends for which mankind, not governments, might coerce individuals,(8) and he was as concerned with private as with official repression of individuality.(9) Joel Feinberg aims to identify the moral limits on the criminal law that would be observed by an "ideal legislature in a democratic country."(10) Like Mill, Feinberg hopes to convince enough of his fellow citizens to endorse the harm principle that paternalistic or morals legislation would become generally accepted as beyond the pale.

That the harm principle takes the form of a categorical and general denial of legitimate state power to adopt certain laws--the precise form that many constitutional guarantees of human rights take--might therefore be a coincidence. The harm principle might be the slogan of those who support a particular political program, rather than a prototype of an actual constitutional provision. Philosophical defense of the harm principle, on this account, would have the end of persuading all citizens to internalize a norm of political tolerance for conduct regarded as immoral but not harmful to others.

On the other hand, the harm principle might be understood precisely as a prototypical constitutional provision (or even as a principle that could be read into existing human rights conventions or national constitutional provisions, such as the due process clauses in the United States constitution). The harm principle would then play the role of an external constraint on legislative majorities, which would approve morals legislation were it not for the bar of the (legally enforceable) harm principle. The harm principle would function as a collective precommitment decision, prohibiting the polity that adopted it from subsequently resorting to the enforcement of morals.

I shall argue that the harm principle is, from a liberal viewpoint, deficient as either a norm to be internalized or as an external constraint. But the arguments against the harm principle vary somewhat depending on which role the harm principle's proponents have in mind. Let us begin with considering the harm principle solely as an appeal to individuals.

II. Mission Implausible I: The Harm Principle as a Norm To Be Internalized

Liberals who put forth the harm principle as a conclusion individuals should endorse and act upon as citizens need to consider just what they hope to accomplish. True, if all individuals agreed with the harm principle (really agreed with it, as distinct from paying it lip service), then indeed liberty would be greatly advanced. An ingrained reluctance to meddle in the lives of fellow citizens prevents a great many political mistakes. Indeed, some nations manage very well with an entire constitution organized as a shared understanding rather than as a legal document. To take but one example, there are reasons to believe that many courts would not read the Constitution of the United States to forbid the enforcement of a fornication statute. Nonetheless, toler-ance of sex outside of marriage is so widely accepted by the citizenry that the adoption and enforcement of a fornication statute is practically unthinkable (even if some of the sodomy statutes technically cover some extramarital sex acts).

But liberals need to ask themselves: What prospect has the harm principle of ever commanding the kind of consensus among individuals that would yield the desired fruit? My own answer is that the prospects are bleak, not so much because of the reactionary Zeitgeist, but primarily because the justifications offered for the harm principle either fall short of actually justifying it (justifications that run too shallow) or because the justifications offered conflict with deep, even fundamental beliefs about politics and morals (justifications that run too deep). More particularly, perfectionist cases for the harm principle seem incapable of categorically excluding all morals legislation whatsoever. And anti-perfectionist cases for the harm principle feature the principle only as the last implication of deontological liberal premises that are at odds with much secular, and almost all religious, moral philosophy.

A. Perfectionist Cases for the Harm Principle: The Problem of Underjustification

Mill defended his sweeping version of the harm principle with two basic arguments. First, individuality is valuable both to the individual and to society, and liberty of action is necessary to promote individuality.(11) Second, collective decisions about private behavior are more likely wrong than right.(12)

There is surely great force to both arguments, but just as surely each falls short of justifying a categorical harm principle. To take individuality first, it seems highly unlikely that a single item of paternalistic or moralistic legislation will stunt the development of those regulated in any significant way. I noted, for instance, no sudden uniformity among the citizens of Illinois when the state legislature adopted a law punishing the failure to wear safety belts in automobiles. If individuality is a good, not a right, as Mill maintained,(13) it must compete against other goods such as safety (or for that matter moral fiber). All individuality requires is a substantial, not an absolute, scope of individual control. The marginal damage to self-development from a few paternalistic or moralistic laws is likely to be small. The payoff, in terms of welfare or virtue, might be very high. In short, the trouble with Mill's argument for liberty as a catalyst of diversity is that there is no linear relationship between official restraints and individuality of character.

Mill's second argument, his skepticism about the public's wisdom in censoring harmless conduct, is certainly rooted in long and bitter experience. But again the argument runs too shallow, for if a legislator, after giving due regard to her own fallibility, nonetheless concluded that despite a presumption to the contrary a particular paternalistic or moralistic law is justified, there is nothing in Mill's argument that should convince her of the contrary.

Maybe Mill's harm principle, understood as pure normative philosophy, is right--the world would be better if the harm principle were observed categorically. But I see no prospect whatever of convincing a great majority of citizens in any country based on the arguments he made in On Liberty. What Mill thought of as examples of the general case for liberty--the liberty of thought and discussion(14)--have been accepted as beyond political restraint, not because of their connection to individual character but because of the conviction that speech and conscience have special value and may not be restricted even though they may cause harm. Sexual conduct is slowly being accepted as in the same category and for the same reason.

Legislators who fail to internalize the principle because the Millian justifications run too shallow are left without an internal check on the impulse to punish. One who quite reasonably treats the Millian justifications as prudential arguments for a categorical principle will then evaluate proposed legislation on an ad hoc basis, acknowledging that one should account for one's own fallibility and for the damage repression does to character. Of course, this might be a disaster for liberalism on the ground. Asking a legislature to adopt only a few morals laws is like asking a teenager to have just a few cigarettes. But in the absence of a compelling categorical argument of the sort Mill never delivered, the harm principle probably never will achieve the broad acceptance required for it to function as an internalized norm of tolerance.

Indeed, it is largely because utilitarianism seems logically consistent with gross violations of human rights, such as those that shocked the world during the middle portion of this century, that modern liberals turned away from consequentialism in favor of deontological rights theories. Now that deontological liberalism is on the run,(15) philosophers have devoted new attention to consequentialism. Joseph Raz, in particular, has tried to develop a version of liberalism that allows the state to promote the good, but which secures individual liberty because of the primacy of autonomy in the goodness of human life.

Raz makes some commonsense concessions about the value of autonomy. Autonomy is relative, not absolute,(16) and indeed depends on the constraints imposed by social forms.(17) And wicked choices are not converted into good ones simply by virtue of being autonomous.

Raz sounds a lot like a social conservative or a communitarian. His case for the harm principle is as follows:

[P]ursuit of the morally repugnant cannot be defended from coercive interference on the ground that being an autonomous choice endows it with any value. It does not (except in special circumstances where it is therapeutic or educational). And yet the harm principle is defensible in the light of the principle of autonomy for one simple reason. The means used, coercive interference, violates the autonomy of its victim. First, it violates the condition of independence and expresses a relation of domination and an attitude of disrespect for the coerced individual. Second, coercion by criminal penalties is a global and indiscriminate invasion of autonomy. Imprisoning a person prevents him from almost all autonomous pursuits. Other forms of coercion may be less severe, but they all invade autonomy, and they all, at least in this world, do it in a fairly indiscriminate way. That is, there is no practical way of ensuring that the coercion will restrict the victims' choice of repugnant options but will not interfere with their other choices. A moral theory which values autonomy highly can justify restricting the autonomy of one person for the sake of the greater autonomy of others or even of that person himself in the future. That is why it can justify coercion to prevent harm, for harm interferes with autonomy. But it will not tolerate coercion for other reasons. The availability of repugnant options, and even their free pursuit by individuals, does not detract from their autonomy. Undesirable as those conditions are they may not be curbed by coercion.(18)

This seems like an unlikely turn, because it denies that the evil of any self-regarding choice might be greater than the evil of official coercion.

I do not understand how Raz's premises can support a universal proscription of morals legislation. Raz seems to admit that many autonomous preferences for vice are bad and that they cannot be made good simply because they are autonomous. He must therefore argue for the priority of autonomy over all the values that might be offended by harmless private conduct. Why this competition should be resolved absolutely in favor of autonomy is mysterious.

For example, a misdemeanor statute curbing the distribution of pornography, say, does not look like the kind of "global and indiscriminate" invasion of autonomy he condemns. The concession that autonomy is relative and contingent on social forms makes this point all the stronger. The statute does not threaten consumers with anything. It denies them only the "social form" of masturbating with visual aids. It denies sellers of pornography only a single narrow role even within the entertainment sector of the economy.

Perhaps this explains why Raz has tried to reinforce his defense of the harm principle by invoking the idea of trust between citizens and their government.(19) Why trust deserves categorical priority over the ends of morals legislation seems as cogent a question as why autonomy deserves this priority. In particular, his claim that morals legislation expresses disrespect for the regulated citizens, and that this disrespect undermines necessary trust between citizens and their government, seems to contradict his comfortable acceptance of physical, as distinct from moral, paternalism.

Social conservatives find Raz quite simpatico.(20) They rightly realize that his defense of the harm principle is not a necessary implication of his perfectionist premises. For legislators might well place great value on autonomy and yet conclude that it does not always deserve priority over the moral well-being of citizens.(21) Raz's position, like Mill's, is underjustified. As Wolfe notes, "While some of [Raz's] arguments suggest useful prudential limits to moral paternalism, they do not exclude moral paternalism as a matter of principle."(22)

As strategies for persuading a polity to internalize a norm of tolerance, perfectionist defenses of the harm principle seem unpromising. They begin with the idea that the purpose of politics is to promote the good life, and while liberty doubtless promotes the good for the most part, the harm principle depends on the improbable claim that liberty always promotes the good absent harm. A few philosophers may cling to that conclusion, but common sense is against it, and it is common sense to which the harm principle must appeal if it is to function as an internalized norm of tolerance.

B. Anti-Perfectionist Theories: The Overjustification Problem

Deontological liberal theories that assert the priority of the right over the good do not suffer the logical insufficiency of perfectionist cases for the harm principle. If the right is prior to the good, and if the right requires freedom from legal coercion absent harm, then the harm principle would indeed be categorically justified. The trouble with deontological liberalism, from the standpoint of achieving an internal norm of tolerance based on the harm principle, is not that the deontological justification runs too shallow, but that it runs too deep.

The claim that the government ought to be neutral about what gives value to life,(23) or that it should never burden one person solely to benefit others,(24) is simply too deeply contested to have much hope of ever grounding a consensus in favor of the harm principle. As John Rawls himself argues in Political Liberalism, precisely because of its openness the liberal society will foster a plurality of inconsistent comprehensive moral views.(25) Any enduring constitutional norms must be supported by an "overlapping consensus," that is, the norms must be consistent with many different comprehensive moral views and therefore not dependent on any one.(26) A political conception of liberalism, Rawls asserts, is just such a set of norms.(27)

I agree with all of this except the last. Principles of justice will be as subject to disagreement as judgements about religion and morality.(28) One can debate the significance of the prospect of perpetual disagreement about justice, but denying the fact of it is unreasonable. Thus any number of rights-based liberal theories can yield the harm principle as an external constraint liberals should persuade a skeptical society to fasten upon itself. But rights-based theories, conflicting as they do with most religion and with consequentialist moral theories, have only dismal prospects of achieving enduring voluntary assent from a free and diverse polity.

If John Rawls cannot convince his Harvard colleague Michael Sandel,(29) and if Ronald Dworkin cannot persuade his Oxbridge colleague Joseph Raz,(30) and if Bruce Ackerman cannot convince his erstwhile Yale colleague Robert Bork,(31) what are the chances for convincing that vast spectrum of society outside the academy that the right is prior to the good and that the right encompasses the harm principle?

The concessions liberals themselves are willing to make about the harm principle afford another telling indicator of the principle's deficiencies as a prospective norm to be internalized. Consequentialists like H.L.A. Hart(32) and Joseph Raz(33) agree that preventing harm to the actor can justify criminal penalties. Joel Feinberg would impose criminal penalties for staging consensual gladiatorial contests or for deliberately conceiving a defective baby;(34) he struggles mightily to explain away laws that require motorcyclists to wear helmets.(35) If liberty in particular cases depends on convincing the polity to accept the harm principle generally, liberty's prospects should disturb liberals of all descriptions.

III. Mission Implausible II: The Harm Principle as External Constraint

The fact of deep disagreement about political morality makes it unlikely that the harm principle could ever be sold to a supermajority of the polity in a liberal state. Nonetheless, that prospect is more likely than the internalization scenario. Perfectionist and anti-perfectionist liberals both might endorse the harm principle as a political precommitment strategy the polity should adopt. If, however, the harm principle were adopted as a constitutional provision without being internalized, the principle would be largely ineffectual in protecting liberty. At least three distinct liabilities attend the application of the harm principle as an external constraint on democratic politics.

First, the concept of harm is vague, vague enough that proponents of morals laws could frequently point to some immediate consequence of private vice that can plausibly be characterized as harm.(36) Second, even if a narrow understanding of harm could be counted on, seemingly private behavior very often initiates a causal sequence that ends in harm, albeit the chain may be long and speculative. Third, because the harm principle takes a categorical form, the principle allows the imposition of criminal liability upon a showing of any harm, however slight. The principle says nothing about how to balance benefits against harms. The second and third application problems together mean that the harm principle would be at best a feeble check on a majority's inclination to punish nonconformity as vice.

The constitutional mission of the harm principle needs to be kept in mind. Conceived as an external constraint, the harm principle is likely to be interpreted and applied by skeptical or actively hostile decisionmakers. Even if liberal theorists can articulate a version of the harm principle that effectively regulates the political process, such a version of the harm principle is highly unlikely to be adopted by legislators or constitutional courts.

A. The Concept of Harm Is Insufficiently Determinate

The vagueness of "harm" is widely recognized. Joel Feinberg has devoted the greatest care to explicating the concept. According to Feinberg, harm "refers to those states of set-back interests that are the consequences of wrongful acts or omissions by others."(37) Feinberg refines this definition of harm with mediating maxims, including volenti non fit injuria and de minimis non curat lex. Even as specified by Feinberg, the harm criterion is insufficiently determinate on three levels.

At the level of interests, Feinberg excludes trivial injuries and moral harm. Yet he would allow criminal punishment for conduct that offends, but does not harm, others, subject to the balancing test of traditional nuisance law.(38) He writes with sympathy for those who are subjected to "profound offense," but so long as the offending conduct is not involuntarily witnessed he would not allow criminal punishment.(39)

Imagine an Israeli court asked to decide whether forbidding the operation of businesses on the Sabbath violates a constitutional provision forbidding criminal penalties in the absence of "unconsented harm"; or a Saudi court asked to apply such a constitutional provision to laws prohibiting alcoholic beverages; or a California court asked to apply such a provision to a statute prohibiting the incestuous union of two siblings. I think it highly likely that courts in such cases would equate "profound offense" or "moral harm" with "harm" in the constitutional provision. Such courts might reason that mere knowledge that the forbidden practice is going on "harms" the right-thinking members of the community. The basic liberal answer to the "bare knowledge problem" is that if such knowledge constitutes harm then nothing would be left of liberty(40) As Feinberg notes, this is a convincing reply from liberal premises.(41) Conceived of as an external constraint, however, the harm principle must be tested on the assumption that those applying it will not have internalized the liberal position.

The harm principle is also vague at the level of the mediating maxims. As indicated, officials likely to apply the harm principle may well regard "profound offense" or "moral harm" as anything but de minimis. The volenti maxim is also open to manipulation. It is easy to imagine judges and legislators inferring failures of rational consent solely on the basis of their assessment that forbidden conduct is immoral or self-destructive.(42) Could a sane person really consent to, say, drug abuse? Given that most of the familiar vices are self-destructive to at least a degree, there would be a great deal of room for repressive elements to find failures of consent when all that is really present is a choice with which they disagree.

Third, the harm principle must incorporate some idea that injuries not wrongfully inflicted do not qualify as harm. Otherwise, successful economic performance would "harm" competitors. Yet any theory of rights presupposes some baseline against which to compare alternative states of affairs.(43) The criminal law itself can change these baselines by giving individuals a right to immunity from conduct they would not otherwise have enjoyed.

For example, Robert George makes the clever argument that punishing harmless wrongdoing is consistent with retributivism because even in the absence of a particular victim, crimes necessarily violate the moral obligation to obey the law.(44) Thus the criminal law can create rights against injuries. I am not persuaded by the argument,(45) but such claims are exactly the sort of argument that any external constraint on legislation must be able to resist. There is nothing in the harm principle that excludes the use of the criminal code itself to create (or to destroy) rights against private conduct. If the existence of the criminal provision itself establishes baseline rights against private conduct, the harm principle would not exclude, for example, competitive injury or the failure to confer benefits.

What goes for rights goes for setbacks to interests. If the state by law provides that it will pay for smoking-related illness, then the manufacture of cigarettes injures the state and its taxpayers. One might reply that all this justifies are special rules of insurance for risky behavior.(46) Leaving aside the problems of monitoring special insurance obligations, the choice of baselines is not given by the harm principle. If there is an extant compulsory insurance scheme, every self-inflicted but insured physical injury imposes an unconsented injury on the taxpaying public.

B. The Harm Principle Invites Speculative Attributions of Harm

As Richard Epstein has noted, the idea of harm has greatly expanded over the past century.(47) Some of the expansion is due to changes in normative thinking, but much of it reflects the simple fact that our understanding of causation is broader than it used to be. Because our understanding of causal sequences has been extended by modern science, both physical science and social science, we are open to causal hypothesis that would never have occurred to legislators of the nineteenth century.

To take some pretty obvious examples, pornography is alleged to cause rape(48) and guns are alleged to cause murder.(49) Given the tendency to believe the pleasant rather than the true, such causal linkages will seem very attractive to legislators and judges subjected to the external constraint of the harm principle. For a droll but instructive example, consider the official explanation of why marijuana must remain illegal which characterizes marijuana as a "gateway drug."(50) That is, marijuana use, although harmless to the user, causes the user to progress to the use of drugs that are harmful to the user, such as cocaine and heroin. The vast majority of heroin users previously used marijuana. Q.E.D.

Of course, the vast majority of marijuana users first tried beer, and, I suppose, cola. Once a connection between two behaviors is hypothesized, it becomes very hard to disprove. The harm principle's application is contingent on the facts, and the facts are likely to be found adverse to liberty. For example, some state courts have upheld motorcycle-helmet laws on the ground that these laws protect other motorists from being injured in accidents caused when a flying rock strikes an unhelmeted cyclist in the head.(51) A little causal creativity can go a long way toward eviscerating the harm principle.

C. The Harm Principle Is Overinclusive

Even if mediated as liberals would wish, the harm principle leaves too much scope for the imposition of criminal penalties. Most of the conduct legislators might be tempted to criminalize really does cause at least some genuine harm to unconsenting parties. Gay sex in New York in 1980 kills children in Los Angeles in 1997. Drugs, alcohol included, kill people in highway accidents.(52) Guns kill people when children mishandle them.(53) Gambling causes some people to turn to crime to support their compulsion.(54) Second-hand smoke is apparently genuinely harmful.(55) Self-induced health problems, due to smoking, diet, sexual practices, and so on cost third parties billions of dollars.(56)

Indeed, even seemingly innocent pastimes cause unconsented harm. The fertilizer runoff from golf courses is a serious environmental problem,(57) leaving aside bystanders who are injured during play by errant shots.(58) Swimming pools lure children to death by drowning.(59) Think, for a moment, about all the unconsented harm that might be prevent by banning automobiles.

Almost any activity can be made harmful by making it criminal. Suppose the government banned coffee, or blue jeans, or chess. Soon enough underground coffee bars, denim markets, and chess clubs would come into existence. Their operators would bribe the police and feud with each other. They would need resort to private violence to enforce their contracts. Soon coffee would be adulterated with God knows what while underground denim factories exploited workers in unsafe conditions. Chess addicts would turn to crime to pay the high fees charged by underground club operators. Logically, the harm principle should be applied ex ante the decision to punish, not ex post. But in practice there is a strong temptation to perform the analysis the other way around, especially when the issue is the repeal, rather than the adoption, of criminal penalties.

The harm principle operates catastrophically; conduct is either harmless and therefore immune from punishment or harmful and thus fair game. By itself the principle gives no weight to the benefits of harmful conduct, leaving that balance entirely to legislative discretion. Not surprisingly, legislatures do not apply the Kaldor-Hicks standard.(60) They are, rather, quick to see the benefits of golf and automobiles and blind to the benefits of drug use or prostitution.

I cannot think of a traditional vice law that could not be defended by at least a colorable argument that the law prevents unconsented harm. Such harms of course do not motivate morals legislation, but they would be available to justify morals legislation in a regime subject to the external constraint of the harm principle.

D. The Example of Pornography

Imagine a society subject to a constitutional provision adopting the harm principle but without a specific guarantee of free expression. Its legislature approves a statute prohibiting pornography, and the statute is challenged in the constitutional court. The judges are legal moralists in sympathy with the statute, but try in good faith to apply the harm principle.

Their opinion could take any of several approaches. First, they might simply say that the legislature has apparently accepted the claim that pornography causes rape and that questions of fact are for the legislature to resolve. Second, they might conclude that the existence of pornography harms women and moralists, either because mere knowledge of the profoundly offensive constitutes harm or because pornography expresses a social order of oppression. Third, they might reason that pornography stimulates sexual activity and that this causes the spread of disease. Fourth, they might conclude that pornography consumption is the product of some mental aberration that vitiates the consumer's consent to the purchase. Fifth, they might say that the production of pornography is typically harmful to the performers, whose consent is often obtained by force or fraud. Evidence for this conclusion might well be drawn from post-criminalization experience.

None of these arguments is frivolous. The chance that a court that had not internalized the harm principle would refuse to bite on any of them seem to me very low. In short, the harm principle is a pretty unpromising way of promoting liberty.

It might be objected that the courts in the United States declared an anti-pornography ordinance unconstitutional. It should be remembered, however, that Judge Easterbrook's celebrated opinion in the Hudnut case "accept[ed] the premises of this legislation."(61) Easterbrook wrote for the Seventh Circuit that "[d]epictions of subordination, tend to perpetuate subordination. The subordinate status of women in turn leads to affront and lower pay at work, insult and injury at home, battery and rape on the streets."(62) If the issue in Hudnut had been whether the ordinance violated the harm principle, the city, not the booksellers, would have won the case.

The First Amendment defines a protected species of private conduct independently of that conduct's consequences. So far as the First Amendment is concerned, speech cannot be suppressed even when it causes harm, short of catastrophic harm the government has a compelling interest in preventing. The First Amendment immunity for free expression might succeed because it has become an internalized norm, or because judges can be trusted to enforce external constraints articulated independently of consequences or variable baselines. First Amendment experience does not afford much encouragement about a constitutional harm principle.

We might contrast the robust state of First Amendment liberties, even in a conservative political period, with the decline of Fourth Amendment security from "unreasonable searches and seizures." Any constitutional provision that incorporates morality by reference very nearly reduces to a procedural check on government authority; in effect, legislation authorizing unreasonable searches and seizures has to pass one additional committee (the courts) to become effective. This is not an entirely trivial check, but neither is it particularly well-suited to protect civil liberty. In a conservative political culture, the judges will take (and indeed have taken) narrow views of what police tactics amount to searches or seizures,(63) and broad views of what makes searches or seizures reasonable.(64)

Borrowing from the first amendment, we might attempt to draft constitutional provisions that protect definite types of private conduct. For example, a constitutional amendment might provide that the government may not "abridge the right to engage in sexual behavior with consenting adults in the privacy of the home." As harm principle proponents have recognized, however, the impulse to punish ranges widely against many types of conduct, depending on the passions of the day. Constitutional protection for specific types of conduct deserves liberal support, but some more general response to the overcriminalization problem seems in order.

E. The Potential Disadvantages of a Constitutional Harm Principle

What I have said so far arguably establishes no more than that the harm principle, by itself, would be inadequate to protect individual liberty as fully as liberals might desire. The vagueness of the harm principle, however, would not only permit judges to uphold most criminal laws. It would also permit judges to strike down many criminal laws that elected legislatures might deem wise and that liberal principles do not condemn.

If the courts are allowed to second-guess legislative factual determinations about harm, criminal provisions regulating pollution, gun possession, smoking in public, and consumer product safety would be vulnerable to Lochner-style judicial vetoes. Constitutionalizing the harm principle is dubious not because the harm principle is wrong. On my view the harm principle, fairly applied, would prevent wise legislation very rarely, but quite regularly would prevent unwise legislation. While there is little reason to hope for success, liberals should therefore continue to pursue the internalization of the harm principle by the polity. As an external constraint, however, the harm principle could succeed only by authorizing the courts to strike down legislation inconsistent with the principle. Any such constitutional harm principle remains dubious, because of the likelihood that such a provision would be unfairly applied, not just to tolerate oppressive morals legislation but to strike down other laws quite properly within legislative authority.

IV. After the Harm Principle

The harm principle's deficiencies all derive from the principle's categorical focus on the consequences of private conduct. The effort to define a class of conduct that is immune from punishment collides with deeply held moral and religious views, and invites ingenious arguments about the likely consequences of despised behavior. Liberals who are serious about constraining society's punitive impulses must start a new search for appropriate principles of applied political philosophy. Even if the harm principle describes the ideal limits on the criminal law, other institutional arrangements will be necessary to realize those limits.

The harm principle's shortcomings suggest some criteria for alternative institutional arrangements. Liberals surely should continue to persuade our fellow-citizens that morals legislation is unwise or unjust, but the depth of our disagreement with conservatives and communitarians gives good reason for focusing on external constraints rather than on internalized norms. Liberals should work to persuade the polity to adopt constitutional limits on criminal punishment that (a) might plausibly win the support of an overlapping consensus of citizens who entertain diverse comprehensive moral or religious views and (b) that might effectively defeat misguided morals legislation in practice. I shall argue that constitutional limits articulated in content-independent terms--rules about who decides what shall be criminal, or about how the criminal law must be enforced, as distinct from rules about the substance of immunized conduct--might fulfill these criteria.

First, I suggest a constitutional provision that forbids imposing penal liability not authorized by legislation which has been approved by two-thirds of the legislature within ten years of the charged conduct. Second, I suggest a constitutional requirement of even-handed enforcement of the criminal law. I will discuss each of these proposals in turn.

A. The Contemporary Supermajority Requirement

At present criminal statutes are not subject to any special constitutional procedural requirements, as are, for instance, taxes,(65) impeachments,(66) and treaties.(67) As a result criminal prohibitions can survive long after they have lost the support of the polity. And new criminal liabilities can be added by a bare majority in the heat of popular passion.

Adopting a contemporary supermajority requirement for penal legislation would go a long way toward actually securing the harm principle. Morals legislation that attracts the opposition of a committed minority of citizens--sodomy laws and at least some drug laws come to mind--would have to be repeatedly affirmed by two-thirds of the legislature. The expiration of the ten-year time limit would give opponents of such laws repeated opportunities to block reenactment. The core criminal code provisions on murder, rape, home invasion, and theft would easily be readopted, perhaps with some beneficial fine-tuning. But borderline criminal legislation of any description would face an uphill fight.

While the harm principle is silent about when harmful private conduct should nonetheless escape punishment on account of its benefits, the proposed constitutional regime tilts the scales against criminalization. Farfetched arguments about causation or the nature of harm could overcome the harm principle. By contrast, the proposed rule is as easy to apply as the supermajority requirement for treaties, impeachments, and constitutional amendments.

The proposed regime thus achieves the effective restraint of the impulse to punish, an achievement beyond the harm principle. Could such a regime ever win the support of an overlapping consensus--the sort of consensus that would be required to adopt a constitutional amendment? We do not know because liberals have not spent much energy on behalf of such proposals. But there are reasons to think that such a jurisdictional approach might win the support of an overlapping consensus.

We should not forget the very powerful perfectionist cases for the harm principle. Mill captured two profound truths about coercive moral engineering. The engineers are often wrong, and coercive means stunt personal development. Raz too is surely right to defend autonomy as a great value, even if his conversion of the good of autonomy into a moral trump savors of all the mysteries of deontology.

The force of these considerations has never been lost on legal moralists. Thomas Aquinas,(68) James Fitzjames Stephen,(69) Patrick Devlin,(70) and Robert George(71) all have agreed that the criminal sanction should not be resorted to lightly. The good life is plural; the legislature may be wrong; autonomy is worth something, a great deal to many; black markets breed unsavory side effects. None of these writers would be committed by their published work to oppose a contemporary supermajority requirement.

Indeed, for communitarians such as Michael Sandel(72) the proposed requirement has much to commend it. For by forcing a debate about borderline penal statutes every ten years, the proposed regime in effect mandates collective deliberation about the criminal law. Should any person be punished as a criminal for violating a norm that a substantial minority of the polity, after sustained debate, rejects? Much in the tradition of civic republicanism suggests a negative answer to that question. From the standpoint of democratic theory, it is difficult to explain why a bare majority can authorize prison for private conduct while a two-thirds majority is required to ratify a treaty setting tariffs on winter wheat or to cashier a bankruptcy judge for taking bribes.

The liberal in me has never taken the caution about punishment piously expressed by communitarians and conservatives quite at face value. I confess to suspecting that these expressed cautions are simply rhetorical tropes offered by those who know very well what the yahoos will do with the criminal law if left unconstrained. Illuminatingly, legal moralists like Stephen and George duly note that legislators make mistakes, that the criminal law is a blunt object, that the good life is plural, and so on. For all of that, these writers fail to identify a single criminal prohibition that ought to be repealed. Surely, if the ritual cautions about vice laws are correct, it is fair to assume that some of the existing ones are major errors, even if we should also be willing to at least consider some morals legislation not now on the books.

Given the categorical character of the harm principle, conservatives and communitarians have been let oft too easily. They can oppose the harm principle by talking about consensual gladiatorial contests and misconceived babies, knowing that the indefensible laws against recreational drug use and unconventional sex can be expunged only by heroic political effort.

The proposed constitutional regime invites communitarians and social conservatives to show their good faith. If their quarrel with the harm principle is really about categorically excluding the criminal law from the available means of promoting the good, they should be willing to sign on to the contemporary supermajority requirement. If by contrast they oppose the harm principle precisely because the harm principle would prevent the oppressive laws they tacitly favor, the proposed constitutional rule would force them to declare their intentions.

In this country a constitutional supermajority requirement could of course be institutionalized through the process of constitutional amendment, including the amendment of state constitutions. Whether legislative rules can establish binding supermajority requirements is a debated question,(73) but what seems clear is that legislative rules adopting supermajority requirements must themselves be subject to repeal by simple majorities of the legislative chamber that adopted the rules.(74) A supermajority requirement adopted as a matter of house rules would be easier to realize than one adopted by constitutional amendment, but its checking value would be correspondingly weaker.

Because the states retain primary responsibility for defining and prosecuting criminal offenses, the political prospects for a constitutional amendment are less bleak than might be supposed. Liberals would need to devote much time and effort to educating the polity about the overcriminalization problem. If liberals succeeded in persuading the polity of a state that the criminal sanction should be subject to special limits, however, state constitutions are generally easier to amend than their federal counterpart. The prospect is not beyond the vision of those who take long views.

Practically speaking, a supermajority requirement presents some important technical questions. First, applying the supermajority requirement would require a definition of criminal, as opposed to civil, sanctions. The constitutional criminal-procedure provisions, however, require this same inquiry.(75) I have elsewhere argued that the appropriate inquiry in the criminal procedure cases identifies corporal restraint or catastrophic financial liabilities for conduct that is morally condemned as the hallmark of the criminal sanction.(76) Other approaches might be taken, but the important point is that the supermajority requirement would not require courts to develop the meaning of a new constitutional category.

A great many purely civil laws may determine whether particular conduct violates a criminal statute. For example, the law of property may determine whether an individual who asports former marital property after a divorce is guilty of theft or engaged in self-help. A supermajority requirement should apply only to statutes that define actual offenses, and not to other laws that may be incorporated by reference, implicitly or explicitly, into the criminal code. If this produces anomalies or injustices in particular cases, as when a change in the law of intellectual property exposes millions of computer users to potential liability for theft, the ten-year sunset provision would ensure that the legislature would be compelled to address the issue within the context of re-enacting the theft statute.

As with any legislation, criminal penalties might be packaged in different ways, making them more or less politically acceptable. If the murder statute were in danger of expiring because of the sunset provision, a legislative majority might include controversial criminal offenses in the same code as the homicide statutes, forcing opponents of, say, prostitution laws to vote either to outlaw prostitution or to legalize murder. This type of blackmail, however, is double-edged. If the opponents of the prostitution provision stand fast, the voters may blame the proponents of the prostitution statute for refusing to sever the homicide and prostitution provisions.

The risk of blackmail-by-rider may still seem serious enough to justify exempting the core criminal offenses from the supermajority and sunset provisions. This turn, however, presents more problems than it solves. Even centuries-old criminal prohibitions like those against theft and rape can change their content dramatically as terms like "property of another" or "consent" are modified by experience or by changes in the civil law.

A constitutional amendment regulating legislative processes for passing criminal laws might address the packaging problem directly. If the amendment provided that no person may be punished for a crime unless two thirds of the legislators had voted to approve "the definition of that offense," legislatures would be required by the constitutional provision to vote on each statutory offense independently. Given that most of the progress in the substantive criminal law in this century resulted from codification projects, this turn too is troubling.

In the end it seems best to leave legislatures free to package the criminal offenses together. Obviously, liberal legislators will be less willing to vote against a code that includes victimless crimes along with murder, rape, and robbery. Just as obviously, the liberal members gain some leverage from the prospect of defeating the re-enactment of the entire criminal code. For example, liberal members could introduce a bill that re-enacts the core criminal proscriptions of homicide, rape, robbery and theft. Under the deadline of the sunset provision, moralistic legislators could maintain their threat only by voting against the liberal bill. Blame for any lapse into anarchy would then clearly belong to the moralists, not to the liberals. Political incentives therefore suggest that the blackmail threat is more apparent than real.

B. Even-Handed Enforcement

When majority moral sentiments threaten individual liberty, liberals have traditionally sought the protection of the courts. In the context of substantive criminal law, however, American courts have been very reluctant to subject legislative choices to searching scrutiny. This reluctance follows from the absence of any specific textual limits on the legislature's police power to define criminal offenses. The controversy over substantive due process, whether exemplified by Lochner(77) or by Roe v. Wade,(78) cautions against any radical extension of judicial review of penal legislation based on its substantive justice.(79)

But there is room for meaningful judicial review of penal legislation that does not turn on the content of the penal code. If the courts required that the criminal law be enforced even-handedly--so that common but technically criminal activity on the part of the rich, the white, and the male was investigated and prosecuted as often as identical activity on the part of the poor, the black, and the female-one of two things would happen. First, police and prosecutors might permit such laws to fall into desuetude for all classes of society. Second, if police and prosecutors did enforce a criminal prohibition evenhandedly, legislatures would soon be moved to take a more skeptical look at defining widespread behavior as criminal. As things stand, police and prosecutors spare legislators the political pain that should attend criminalization. Two examples illustrate the point.

First, the drug laws are disproportionately enforced against black Americans. While survey data suggest that black and white rates of marijuana and cocaine use are comparable, on a per capita basis a black citizen is five times more likely than a white citizen to be punished for a drug offense.(80) This ratio can exist only so long as investigative resources and prosecutorial priorities are concentrated on minority communities.

Second, prostitution laws are typically enforced by the arrest and prosecution of the prostitutes, not their customers.(81) Occasional sting operations target johns, but these are exceptions to the general pattern.

In both situations, equal protection analysis suggests that the government should not enjoy the power to prosecute when, but for the defendant's race or gender, no charges would be brought. Existing law is unreceptive to such claims; it requires that the defendant prove that the investigating or prosecuting officials acted with invidious intent and has taken a chary view of inferring invidious intent from disparate impact.(82)

Precisely how equal protection doctrine might be mended so as to pose a meaningful barrier to discriminatory law enforcement is a subject for another paper. Suffice it to say that equal protection doctrine could be reformed to pose just such a barrier, most obviously by adopting a burden-shifting approach that would force the government to prove the absence of discrimination upon a showing of disparate impact.(83) My point here is that a constitutional requirement of even-handed enforcement would go a long way toward limiting the substantive criminal law, without involving the judiciary in any assessment of the wisdom of particular laws.

If white teenagers regularly were arrested and prosecuted for using marijuana, there would be a sober reconsideration of the marijuana laws. If middle-class men regularly were prosecuted for patronizing prostitutes, there would be a sober reconsideration of the prostitution laws. If straight people were prosecuted for engaging in oral sex, there would be a sober reexamination of the sodomy laws.

I have no quarrel with permitting a criminal law to fall into genuine desuetude, but many of our criminal laws have achieved a peculiar status indeed--selective desuetude. Voters who demand tough-on-crime policies such as mandatory minimum sentences for drug offenses know no one who will ever suffer from such policies even though they know many people who violate the applicable laws. If the courts threw out cases against members of unpopular or powerless groups until such time as cases were brought against offenders from all strata of society, the democratic process would be stimulated to reconsider the wisdom of many of our criminal laws.

Such a stimulus would not, like Lochner-style substantive-due-process adjudication, impair democracy. On the contrary, a requirement of evenhanded enforcement would force the democratic process to grapple with the merits of morals legislation, rather than rely on police and prosecutors to muffle political controversy by restricting enforcement to the relatively powerless members of society.

Conclusion: Can Liberalism Be Pragmatic?

Can you be a liberal without defending the harm principle? I think the answer is yes, at least if we go with Mencken(84) and define a liberal as a person who values liberty. I have to confess that if I were a legislator, I would vote for seatbelt and motorcycle-helmet laws. I would certainly vote to prohibit gladiatorial games. But surely these concessions leave room for rejecting familiar morals legislation. I find the Millian case for liberty very powerful; I just do not think it powerful enough to bludgeon the skeptical into accepting the categorical harm principle.

Liberals who cling to the harm principle need to ask themselves about the future. What will the criminal law look like in a hundred years? Surely there will be political panics against outgroups. What should liberals fight for to prevent the next round of witch hunts, prohibitions, and prudery? If we start now to work toward practical institutional limits on the criminal sanction, perhaps at least some of the disasters waiting to happen might be avoided. We might even be able to undo some of the mischief that a century of polemical brilliance in defense of the harm principle has failed to prevent.

Liberals who want to realize Mill's vision need to work for constitutional limits on the criminal sanction other than the harm principle. I close with a tautology made eloquent by Isaiah Berlin. "Everything is what it is: liberty is liberty, not equality or fairness or justice or culture, or human happiness or a quiet conscience."(85) Or, I would add, the harm principle.

NOTES

Thanks to Ian Ayres, Carlos Ball, Mike Kelly, Suzanna Sherry, Daniel Yeager, and Chris Wonnell for valuable comments on a prior draft. This paper also benefitted enormously from the comments received at a faculty workshop at the University of San Diego in October 1997. Thanks also to CJE's reviewers, whose comments were acute and very helpful.

(1) J.S. MILL, ON LIBERTY AND OTHER WRITINGS 13 (Stefan Collini ed. 1989).

(2) The principal contributions to the debate, in chronological order, are J.S. MILL, supra note 1; J.F. STEPHEN, LIBERTY, EQUALITY, FRATERNITY (first published 1873; Chicago, 1991); Report of the Committee on Homosexual Offenses and Prostitution (The Wolfenden Report) (Stein & Day ed. 1963); P. DEVLIN, THE ENFORCEMENT OF MORALS (1965); H.L.A. HART, LAW, LIBERTY, AND MORALITY (1963) (Hart responds to lectures delivered earlier, but published later, by Lord Devlin); J. FEINBERG, THE MORAL LIMITS OF THE CRIMINAL LAW (1990); R. GEORGE, MAKING MEN MORAL (1993).

(3) J. FEINBERG, HARM TO SELF 3 (1986) (emphasis added).

(4) Wonnell, Problems in the Application of Political Philosophy to Law, 86 MICH. L. REV. 123 (1987).

(5) Id. at 125.

(6) Id.

(7) On precommitment strategies, see J. ELSTER, ULYSSES AND THE SIRENS (1979); T.C. SCHELLING, CHOICE AND CONSEQUENCE (1984), chs. 3 & 4.

(8) J.S. MILL, supra note 1, at 13.

(9) See id. at 8 ("Society can and does execute its own mandates: and if it issues wrong mandates instead of right, or any mandates at all in things with which it ought not to meddle, it practices a social tyranny more formidable than many kinds of political oppression, since, though not usually upheld by such extreme penalties, it leaves fewer means of escape, penetrating much more deeply into the details of life, and enslaving the soul itself").

(10) J. FEINBERG, HARM TO OTHERS 4 (1984).

(11) J.S. MILL, supra note 1, ON LIBERTY, ch. 3.

(12) Id. at 83-93.

(13) C.L. Ten has suggested that Mill's "doctrine of individuality does not imply a maximization of the realization of individuality," MILL ON LIBERTY 79 (1980). On Ten's account, Mill distinguished coercive frustrations of individuality from the passive failure of the individual to realize herself. Even on this account, however, the individual deserves immunity from coercive frustration of harmless behavior because individuality is a good. If many laws violate the immunity without diminishing the good of individuality, people may well question the justification of the immunity.

More generally, if Ten is right that "there are significant nonutilitarian elements in [Mill's] case for liberty," id. at 5, the harm principle might be saved from underjustification only at the price of overjustification. In other words, if Mill is a crypto-rights theorist (a proposition I find colorable but on the whole less likely than not) Millian defenses of the harm principle would encounter the same obstacles to widespread internalization as have been encountered by contemporary deontological defenses of the harm principle.

(14) See J.S. MILL, Supra note 1, ON LIBERTY, ch. 2.

(15) On the dissatisfaction with utilitarianism following the Second World War, and the contemporary dissatisfaction with deontological liberalism, see I. SHAPIRO, POLITICAL CRITICISM 3-4 & 8-10 (1990).

(16) J. RAZ, THE MORALITY OF FREEDOM 373 ("No one can control all aspects of his life.... All that has to be accepted is that to be autonomous a person must not only be given a choice but he must be given an adequate range of choices."); id. at 382 ("As with the conditions of autonomy, so this awareness [the awareness of one's options one must have to be autonomous] is a matter of degree").

(17) Id. at 308-13 & 391.

(18) Id. at 418-19.

(19) Raz, Liberty and Trust, in NATURAL LAW, LIBERALISM AND MORALITY 113 (R. George, ed. 1996).

(20) See R. GEORGE, MAKING MEN MORAL, ch. 6 (1993).

(21) See, e.g., id. at 188 ("The danger of interfering with morally acceptable choices is a consideration that counts against antipornography legislation in the practical reasoning of prudent legislators. But it may not be a conclusive reason").

(22) Wolfe, Being Worthy of Trust: A Response to Joseph Raz, in NATURAL LAW, LIBERALISM AND MORALITY, supra note 19, at 131.

(23) See Dworkin, Liberalism, in LIBERALISM AND ITS CRITICS 60 (M. Sandel ed. 1984).

(24) See R. NOZICK, ANARCHY, STATE AND UTOPIA 33 (1974)

(25) J. RAWLS, POLITICAL LIBERALISM, lecture IV (1993).

(26) Id. at 144.

(27) Id. at 160-64.

(28) See, e.g., Holmes, John Rawls and the Limits of Tolerance, NEW REPUBLIC, Oct. 11, 1993, at 46 ("Rawls cannot deny that different people in the same culture can hold conflicting concepts of justice, for he allows that `there are many variant liberalisms,' that `the public political culture may be of two minds at a very deep level' and even that his own conception of justice is `deeply controversial'"); Sandel, Book Review, 107 HARV. L. REV. 1765, 1783 (1994) ("Political liberalism must assume not only that the exercise of human reason under conditions of freedom will produce disagreements about the good life, but also that the exercise of human reason under conditions of freedom will not produce disagreements about justice").

(29) Compare J. RAWLS, POLITICAL LIBERALISM, supra note 25, with Sandel, Book Review, supra note 28.

(30) Compare Dworkin, Liberalism, supra note 23, with J. RAZ, THE MORALITY OF FREEDOM, chs. 5 & 6 (1986).

(31) Compare B. ACKERMAN, SOCIAL JUSTICE AND THE LIBERAL STATE (1980) with R. BORK, SLOUCHING TOWARD GOMORRAH: MODERN LIBERALISM AND AMERICAN DECLINE (1996).

(32) H.L.A. HART, supra note 2, at 32.

(33) J. RAZ, supra note 16, at 422.

(34) J. FEINBERG, HARMLESS WRONGDOING 325-31 (1990).

(35) J. FEINBERG, HARM TO SELF 134-42 (1986).

(36) This point is not novel. See, e.g., H. SIDGWICK, THE METHODS OF ETHICS 477-78 (7th ed., 1907, reprint 1981); Nagel, The Enforcement of Morals, HUMANIST, May/June 1968, 20, 21-23. Its institutional implications, however, are underappreciated.

(37) J. FEINBERG, supra note 10, at 215.

(38) J. FEINBERG, OFFENSE TO OTHERS, ch. 7 (1985).

(39) Id. ch. 8, at 93-94.

(40) See J.S. MILL, supra note 1, at 84; H.L.A. HART, supra note 2, at 46-47.

(41) J. FEINBERG, supra note 38, at 63-64.

(42) See, e.g., Kennedy, Distributive and Paternalist Motives in Contract and Tort Law, with Special Reference to Compulsory Terms and Unequal Bargaining Power, 41 MD. L. REV. 563, 624-49 (1982); M. KELMAN, A GUIDE TO CRITICAL LEGAL STUDIES 126-41 (1987).

(43) See Nagel, supra note 36, at 22-23.

(44) R. GEORGE, supra note 20, at 186-87.

(45) I think George neglects the possibility that citizens have an obligation founded on fairness to obey only certain categories of laws (including some they would vote to repeal) and that morals legislation might reasonably be viewed as outside the category of laws deserving content-independent moral respect.

(46) See J. FEINBERG, supra note 35, at 139-41.

(47) Epstein, The Harm Principle -- and How it Grew, 45 U. TOR. L. REV. 569 (1995).

(48) See, e.g., American Booksellers Association v. Hudnut, 771 F.2d 323, 329 (7th Cir. 1985), affirmed, 475 U.S. 1001 (1986).

(49) See, e.g., Dixon, Why We Should Ban Handguns in the United States, 12 ST. LOUIS U. PUB. L. REV. 243, 257 (1993).

(50) See, e.g., WHITE HOUSE CONFERENCE FOR A DRUG FREE AMERICA, FINAL REPORT 5 (June 1988).

(51) See Love v. Bell, 465 P.2d 118, 122 (Co. 1970):

It is also clear that the helmet requirement, as well as the requirement for adequate protection for they eyes, bears a reasonable relationship to the health, safety, and welfare of the rest of the motoring public. As other courts have noted, it is not unreasonable to contemplate that rocks picked up and thrown by the tires of automobiles or some overhead obstruction might strike the motorcyclist on the head causing him to lose control of his vehicle. Such loss of control presents a direct threat to the safety of those who use our public highways.

The hypothesis that such an accident might occur because of failure to wear an approved protective helmet has been upheld as reasonable by the highest courts of appeal in Louisiana, Massachusetts, North Carolina, North Dakota, Oregon, Rhode Island, Texas, Washington, and Wisconsin. Only the Supreme Court of Illinois has held to the contrary. (Citations omitted.)

(52) See, e.g., Vaughn, Kappeler, & del Carmen, A Legislative and Constitutional Examination of "Abuse and Lose" Juvenile Driving Statutes, 19 AM. J. CRIM. L. 411, 412 (1992) ("Scientific studies dealing with controlled substances and marijuana intoxication affecting teenagers operating motor vehicles, under actual road conditions, indicate that these substances adverse-ly affect driving performance") (footnote omitted); Moore, Drugs: Getting a Fix on the Problem and the Solution, 8 YALE L. & POL'Y REV. 8, 9 (1990) ("Inattentiveness or recklessness caused by drug induced intoxication results in serious accidents not only on the road, but also at work and at home").

(53) See, e.g., Copetas, Handguns Without Child Safety Devices--Defective in Design, 16 J. L. & COM. 171,172 (1996) ("Each year, 365 American children under the age of 15 die as a result of accidental shootings with handguns") (footnote omitted).

(54) See, e.g., Gowen & Spyerer, Compulsive Gambling and the Criminal Offender: A Treatment and Supervision Approach, 59 FED. PROB. 36, 37 (Sep. 1995) ("Compulsive gambling can often lead to illegal activity").

(55) U.S. DEPT. OF HEALTH & HUMAN SERVICES, THE HEALTH CONSE-QUENCES OF INVOLUNTARY SMOKING: A REPORT OF THE SURGEON GENERAL (1986).

(56) See, e.g., W. G. MANNING, E.B. KEELER, J.P. NEWHOUSE, E.M. SLOSS, & J. WASSERMAN, THE COSTS OF POOR HEALTH HABITS (1991).

(57) See, e.g., Finnegan, New York City's Watershed Agreement: A Lesson in Sharing Responsibility, 14 PACE ENVTL. L. REV. 577, 610 (1997); Jones, The Coastal Barrier Resources Act: A Common Cents Approach to Coastal Protection, 21 ENVTL. L. 1015, 1023 (1991).

(58) See, e.g., O'Kane & Challer, Injuries from Errant Golf Ball: Liabili-ty Theories and Defenses, 37 FED. INS. & CORP. COUNSEL Q. 247 (1987).

(59) See Kates, Schaffer, Lattimer, Murray, & Cassem, Guns and Public Health: Epidemic of Violence of Pandemic of Propaganda?, 62 TENN. L. REV. 513, 570 (1995) ("twenty times as many children under age five drown in bathtubs and home swimming pools as are killed in handgun accidents") (footnote omitted).

(60) According to the Kaldor-Hicks criterion, "a change from the present social state should be undertaken if the gainers from the change could compensate the losers in such a way that everyone would be better oft, in that way satisfying the Pareto criterion," E. STOKEY & R. ZECKHAUSER, A PRIMER FOR POLICY ANALYSIS 279 (1978).

(61) American Booksellers Association v. Hudnut, 771 F.2d 323, 329 (7th Cir. 1985), summarily affirmed, 475 U.S. 1001 (1986).

(62) Id. (footnote omitted).

(63) See, e.g., California v. Greenwood, 486 U.S. 35 (1988) (no "search" to rummage through suspect's garbage); United States v. Miller, 425 U.S. 435 (1976) (no "search" to compel bank to disclose suspect's bank records to investigators).

(64) See, e.g., Veronia School Dist. 47J v. Acton, 515 U.S. 646 (1995) (random urinalysis of public school sports participants to test for drug use "reasonable"); Griffin v. Wisconsin, 483 U.S. 868 (1987) (search of probationer's home, without warrant or probable cause, "reasonable").

(65) See U.S. Const. Art. I, sec. 7 ("All Bills for raising Revenue shall originate in the House of Representatives).

(66) See id. Art. I, sec. 3, cl. 6 ("The Senate shall have the sole Power to try all Impeachments.... and no Person shall be convicted without the Concurrence of two thirds of the Members present").

(67) See id. Art. II, sec. 2, cl. 2 (the President "shall have power, by and with the Advice and Consent of the Senate, to make Treaties, provided two thirds of the Senators present concur").

(68) T. AQUINAS, 28 SUMMA THEOLOGIAE question 96, article 2, reply, 123-24 (Thomas Gilby ed. 1966) (Law is laid down for a great number of people, of which the majority have no high standard of morality. Therefore it does not forbid all the vices, from which upright men can keep away, but only those grave ones which the average man can avoid, and chiefly those which do harm to others and have to be stopped if human society is to be maintained, such as murder and theft and so forth").

(69) J.F. STEPHEN, LIBERTY, EQUALITY, FRATERNITY 154-55 (1873; Chicago 1991) ("In practice this is subject to highly important qualifications, of which I will only say here that those who have due regard to the incurable weaknesses of human nature will be very careful how they inflict penalties upon mere vice, or even upon those who make a trade of promoting it, unless special circumstances call for their infliction").

(70) P. DEVLIN, THE ENFORCEMENT OF MORALS 11 (1959) ("there must be toleration of the maximum individual freedom that is consistent with the integrity of society").

(71) R. GEORGE, supra note 20, at 42:

Taking a cue from Aquinas, we can identify other prudential (and, as such, morally significant) considerations which might militate in favor of a policy of tolerating certain moral evils: for example, (1) the need to avoid placing dangerous powers in the hands of governments that are likely to abuse them; (2) the danger that criminalization of certain vices may have the effect of placing monopolies in the hands of organized criminals who will market and spread the vices more efficiently; (3) the risk of producing secondary crimes against innocent parties; (4) the risk of diverting police and judicial resources away from the prevention and prosecution of more serious crimes; (5) the concern that the power to enforce moral obligations will be exploited by puritanical, prudish, or disciplinarian elements in society to repress morally legitimate activities and ways of life whose genuine value these elements fail to appreciate; (6) the danger of establishing too much authority and creating a situation in which people relate primarily to a central authority whom they must constantly work to avoid offending, thus discouraging them from building genuine relationships with each other to the point of true friendships and valuable communities.

(72) Sandel rejects the harm principle because he believes that public life shapes individual identity in a way that warrants collective coercion to promote virtue. See M. SANDEL, DEMOCRACY'S DISCONTENT 86-90 (1996) (arguing that legal protection for pornography or hate speech "presupposes a controversial theory of personhood and speech" and that "protecting speech by insisting that local com-munities bracket moral judgments carries costs for self-government"). A sunset provision for criminal laws would force the polity to engage in the type of debate about virtue that Sandel envisions. See id. at 106-108 (urging defenders of gay rights to base their case on the moral goodness of gay life, not on a political right to lead such a life). The supermajority requirement departs from strict democratic theory, but Sandel voices no disapproval of the Senate, judicial review, filibusters, committee systems, and other antidemocratic but well-established features of our system. He seems far more concerned with how the debate is conducted than with how the votes are counted.

(73) See Bloch, Congressional Self-Discipline: The Constitutionality of Supermajority Rules, 14 CONST. COMMENT. 1 (1997); Leach, House Rule XXI and an Argument Against a Constitutional Requirement for Majority Rule in Congress, 44 U.C.L.A. L. REV. 1253 (1997); Leiber, On Supermajorities and the Constitution, 83 GEO. L.J. 2347 (1995); Ackerman et al, An Open Letter to Congressman Gingrich, 104 YALE L.J. 1539 (1995); McGinnis & Rappaport, The Constitutionality of Legislative Supermajority Requirements: A Defense, 105 YALE L.J. 483 (1995).

(74) See, e.g., McGinnis & Rappaport, supra note 73, at 503-07.

(75) See, e.g., Dept. of Revenue of Montana v. Kurth Ranch, 114 S. Ct. 1937 (1994) (imposition of punitive tax on prohibited sale of contraband is punishment for double jeopardy purposes); United States v. Ward, 448 U.S. 242 (1980) (administrative agency's imposition of civil fine not a criminal prosecution calling for Bill of Rights procedural safeguards); In re Winship 397 U.S., 358 (1970) (juvenile court delinquency proceeding is criminal prosecution calling for proof of guilt beyond reasonable doubt).

(76) Dripps, The Exclusivity of the Criminal Law: Toward a "Regulatory Model" of, or "Pathological Perspective" on, the Civil/ Criminal Distinction, 7 J. CONT. LEGAL, ISSUES 199 (1996).

(77) Lochner v. New York, 198 U.S. 45 (1905).

(78) 410 U.S. 113 (1973).

(79) There is of course a vast literature on unenumerated rights. For a leading critique of Roe, see Ely, The Wages of Crying Wolf: A Comment on Roe v. Wade, 82 YALE L.J. 920 (1973).

(80) See NATIONAL CRIMINAL JUSTICE COMMISSION, THE REAL WAR ON CRIME 115 (Steven Donziger ed. 1996); M. TONRY, MALIGN NEGLECT 109-111 (1995).

(81) See, e.g., MacKinnon, Prostitution and Civil Rights, 1 MICH. J. GENDER & L. 13, 17 (1993).

(82) For cases involving allegations of racial discrimination in enforcement of the drug laws, see, e.g., United States v. Armstrong, 116 S. Ct. 1480,1487 (1996) ("The requirements for a selective-prosecution claim draw on `ordinary equal protection standards.' To establish a discriminatory effect in a race case, the claimant must show that similarly situated individuals of a different race were not prosecuted") (citations omitted); United States v. Avery, 1997 West Law 839275 (6th Cir. 1997) (defendant challenging search and seizure on equal protection grounds must show discriminatory intent). Armstrong is Kafkaesque; to establish racial animus, the defendant must show that white suspects were not charged, yet the Court rejected the defendant's motion for discovery because Armstrong had not come forward with evidence to suggest discrimination. Only some sort of burden-shifting analysis, modeled on Batson v. Kentucky, 476 U.S. 79 (1986), could enable criminal defendants to prove discrimination when it exists. Yet the Armstrong Court adopted a presumption that prosecutorial discretion is not discriminatory. See 116 U.S. at 1486. For judicial rationalizations of the enforcement of prostitution laws against women, see MacKinnon, supra note 81. For an unusual case of a court throwing out a prostitution prosecution on account of sex discrimination, see State v. McCollum, 464 N.W.2d 44 (1991).

(83) Batson v. Kentucky, supra note 82, supplies some authoritative support for such an approach. For a defense of the burden-shifting approach, see Schifferle, After Whren v. United States: Applying the Equal Protection Clause to Racially Discriminatory Enforcement of the Law, 2 MICH. L. & POLICY REVIEW 159, 171-80 (1997).

(84) See Mencken, Mr. Justice Holmes, in A MENCKEN CHRESTOMATHY 258, 259 (1982) ("I find it hard to reconcile such notions [as are found in some Holmes opinions] with any plausible concept of Liberalism. They may be good law, but it is impossible to see how they can conceivably promote liberty").

(85) Berlin, Two Concepts of Liberty, in LIBERALISM AND ITS CRITICS, supra note 23, at 18.

Donald A. Dripps is The James Levee Professor of Law and Criminal Procedure at the University of Minnesota Law School.
COPYRIGHT 1998 Institute for Criminal Justice Ethics
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1998 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:DRIPPS, DONALD A.
Publication:Criminal Justice Ethics
Date:Jun 22, 1998
Words:12468
Previous Article:Why Drug Legalization Should Be Opposed.
Next Article:A Dialogue of the Deaf? New International Attitudes and the Death Penalty in America.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters