|
Direct Protection of Liberty |
Copyright Ó 2002/2003, by Alec Rawls
Liberty can be protected directly, by articulating the full ideal of protected liberty and placing it in the Constitution, or it can be protected indirectly, by placing restrictions on what can be done to enforce the laws. Our own system of constitutionally protected liberty is mostly indirect. Only the First and Second Amendments articulate any protected activities. In addition, the "takings clause" of the Fifth Amendment and the "contracts clause" of the Article I §8 provide some direct protection for economic liberty, while the Thirteenth Amendment protects against enslavement. All other protections are indirect.
Unfortunately, indirect protection of liberty is highly inefficient. First, restrictions on law enforcement necessarily protect what we don't mean to protect along with what we do. Suppose, for instance, that a part of the ideal of liberty that we want to protect is consensual activity between mentally competent adults. An indirect way to protect such victimless activities would be to protect privacy, not looking behind closed doors without probable cause on the assumption that, so long as no one is making the activity public by complaining, no one is being victimized. Constitutional restrictions on search have indeed been interpreted as protecting privacy.(1) While this protection for privacy (whether implicit or explicit) does provides substantial de facto protection against prosecution for victimless crimes, it also has the wholly negative consequence of impeding the discovery and prosecution of many crimes of victimization. Those who are murdered behind closed doors are unable to complain and activities that are consensual between some may aim at victimizing others.
All indirect protections are ill targeted in this way. The only way to limit protection to a specific class of activities is to name the class, which means protecting liberty directly. But this is only half the problem with protecting liberty indirectly rather than directly. In addition to protecting what we don't mean to protect, indirect protection also does a poor job of protecting what we do mean to protect.
Suppose again that the goal is to protect consensual adult activity. No matter what indirect protections are put in place to make such activities hard to uncover and prosecute, those searches that are not barred will still be prone to uncover consensual activity, allowing the indirectly protected activities to be prosecuted. The ineffectual hit-or-miss nature of indirect protection is exacerbated by the fact that indirect protection of liberty creates a terrible conflict between liberty and crime control. In order to protect liberty, prosecution of all crimes must be compromised, compelling society to keep indirect protections to a minimum.
Direct protection of liberty would have none of these liabilities. We would have to able to figure out how to articulate the full ideal of liberty in order to protect liberty directly, but if we can manage this feat, then in contrast to the hit and miss of indirect protection, direct protection for liberty would be thorough and systematic. All laws that violate the articulated full ideal of protected liberty would be unconstitutional. Members of Congress would swear an oath not to pass such laws, the President would swear not to sign them, and judges would swear to strike them down. With liberty-infringing laws off the books, protection of liberty would not depend on restrictions on law enforcement. No matter what kind of searches were allowed, no one would prosecuted under illiberal laws. The conflict between liberty and crime control—the signal feature of our present system of indirect protection—would be almost completely eliminated, allowing each to be optimized according to its own internal imperatives.
Direct protection of liberty would not allow for the complete elimination of indirect protections because restrictions on the courts and the police do more than protect against illiberal laws. Most obviously, they also protect the rule of law. Constitutional protection against illiberal laws only protects liberty to the extent that the laws (in particular the Constitution) are abided by. Thus it is also necessary to protect against assaults on the law itself, be they petty exercises of arbitrary power, attempts by would be tyrants to usurp our democracy, or attempts by tyrannical majorities to ignore the Constitution and pass illiberal laws. Various of our existing indirect protections of liberty focus on the need to protect against usurpation and maintain the rule of law. The need to maintain these safeguards places limits on how far the direct protection of liberty would allow indirect protections to be relaxed.
Another effect and presumable purpose of existing indirect protections is to provide some protection of privacy per se. Several indirect protections of liberty operate by carving out personal space that law enforcement has difficulty penetrating: the home, papers and affects, a person's mind. While this protection of a sphere of privacy does offer some protection from prosecution under unjust laws, it also keeps people's properly private matters from being scrutinized unnecessarily, even where there is no threat of unjust prosecution. This effect, whether intended or not, has value that needs to be accounted and given priority as far as countervailing concerns allow. We could get rid of all unjust laws and still not want to allow Big Brother type video monitoring of the inside of everyone's house, even if it would have some law enforcement benefit. This concern for privacy itself places additional limits on how far indirect protections of liberty should be relaxed when moving to the direct protection of liberty. But even with these limitations, the relaxations of restrictions on law enforcement that direct protection of liberty would enable are tremendous.
Consider a concrete example that illustrates the structure and size of the available efficiencies: the Fifth Amendment right not to testify against oneself. This right is of great value to the guilty, who would otherwise have to spin lies if they wanted to assert their innocence, and would risk getting caught in their lies. At the same time, it is totally worthless to innocents, who invariably yearn to testify to express their innocence, to account for their actual whereabouts, etcetera. This advantageousness primarily to the guilty is not a mistake. It is by design. The right evolved as a protection for those who were guilty as charged of violating a particular set of unjust laws: the English blasphemy laws of the sixteenth and seventeenth centuries.(2)
A right not to incriminate oneself is a very uncertain way to protect things that we want to protect, like religious freedom. Someone else could always testify to one's blasphemy. Recognizing this, the authors of our Constitution enumerated religious freedom for direct protection. Unfortunately, they made no attempt to articulate the full ideal of protected liberty, as they made explicit in the Ninth Amendment's assertion that not all rights had been enumerated. That left the door open for unjust laws to stand, which meant that the right not to incriminate oneself had to be retained as protection for those who violate unjust laws. Unfortunately, this right also protects thieves, rapists, murderers and terrorists.
With liberty protected directly, defendants could be required to answer questions about those activities that are certainly not part of the full ideal of protected liberty, such as any heinous crime, without posing any risks to properly protected liberties. The right not to incriminate oneself is somewhat unique in that it is the only right that developed explicitly to impede the enforcement of unjust laws, but all indirect protections have this effect and in some degree, presumably, this intent (as the Court assumed when it discovered an implied right to privacy). Thus the same efficiencies that are available in the case of the right not to incriminate oneself are available generally. If we would protect liberty directly, we could relax or eliminate many restrictions on law enforcement, allowing much more effective crime control (and hence much greater protection for our system of liberty), while enjoying the greatly enhanced protection for personal liberty that direct protection provides.
The current terrorist threat, and the increasing potential lethality of terrorism as technology progresses, greatly increase the cost of restrictions on law enforcement. So long as we continue to rely on such restrictions for our protection of liberty we are stuck with a terrible choice. If we fail to relax restrictions on law enforcement in important ways the result could easily be additional terrorist successes, on any possible scale. If instead we relax restrictions on law enforcement so as to effectively uncover and root out terrorist threats, this relaxation of restrictions will undo our indirect protection of liberty, threatening the liberty that defines our way of life. The direct protection of liberty offers a way to avoid both of these evils. Instead of having to sacrifice some liberty out of military necessity, we can loosen restrictions on law enforcement quite a bit further than we would ever contemplate under indirect protection while at the same time greatly increasing protection for liberty.
While the direct protection of liberty is especially valuable in the context of a war against terrorism, it is important to recognize that a move to direct protection in no way depends on military exigency for its justification. This is a proposed reform to civilian law, called for by the great advantages it offers to both of the civilian concerns that it affects: crime control and the protection of liberty. We should be figuring out how to move in this direction in any case. The current threat simply adds a note of urgency to the enterprise.
Note also that, while direct protection of liberty would allow us to in many ways to untie the hands of law enforcement without compromising liberty, it offers no guarantee that the war on terrorism won't call for us to go further still and in some ways compromise liberty. In the calculus of constitutionally protected interests, the weight that any particular liberty or constitutional protection places in the scale can easily be outweighed if preservation of the nation sits in the other side of the scale. The tree that bears the fruit must be protected above the fruit. This was the essence of Supreme Court's ruling in Korematsu v. United States, 323 U.S. 214 (1944), where it upheld the internment of the Japanese immigrant population during WWII. Japanese-Americans were being treated unequally as a group, but when the constitutional guarantee of equal protection was weighed against the other constitutionally recognized interests at stake (the war-fighting powers and the need to use them to preserve the nation) unequal treatment was held to be justified.
Obviously, we want to avoid such tradeoffs as far as possible and this is what the direct protection of liberty enables. Military means, like the use of secret tribunals and the holding of enemy combatants and prisoners of war (whether citizens or foreign nationals) will always be necessary in times of war, whether liberty is protected directly or indirectly. We are faced today, however, not just with these longstanding points of divergence and contention between civilian justice and military necessity, but with greatly heightened conflicts. Infiltration of the civilian population is our terrorist enemy's primary means of attack. He hides behind our protections for civilian liberty, awaiting an opportunity to commit mass murder. Routing such an enemy may require investigation of the civilian population in some extreme ways, from the systematic to the highly invasive.
So long as our primary method for protecting liberty is through the restrictions we place on law enforcement, true unleashing of law enforcement would attack the whole landscape of civilian liberty and justice, while the failure to unleash it threatens our survival. But this tradeoff only arises because our protection for liberty is indirect. If we would amend the Constitution to focus on the "what" of liberty instead of the "how," we could get much better protection of the "what" with much less restriction on the "how." Some tradeoffs we still won't be able to eliminate, but what we can eliminate, we should.
Section II offers an articulation of the full ideal of protected liberty. Luckily, Thomas Jefferson and John Stuart Mill already did most of the work, and a strong argument can even be made that the proposed ideal should be held to be protected by the Constitution already. Section III develops a scheme for protecting this full ideal of liberty comprehensively in the law. This implementation scheme is designed to first eliminate the great bulk of laws that would violate the full ideal of liberty then, where illiberal laws might slip through, insure that such laws at worst pose no greater a threat to liberty than at present. Thus the resulting scheme should be efficient in the Pareto sense of allowing vast improvements in both crime control and the protection of liberty while insuring that liberty is nowhere worse protected than by the current scheme of indirect protections. Section IV considers how far indirect protections should be retained pursuant to the protection of privacy per se. Section V considers how far they must be retained in order to maintain protection for rule of law. Section VI looks at threats to rule of law that stem, not from tyranny by usurpation, but from tyranny of the majority. And finally, to test the resulting full scheme of direct protection, the concluding section VII applies it to the hardest of hard cases: racial profiling.
Part II: Mill's principle of liberty
On the full ideal of protected liberty, the Ninth Amendment points the way. Of all the candidates for unenumerated rights "retained by the people," first in line must be the rights to "life, liberty and the pursuit of happiness" proclaimed in the Declaration of Independence ten years before the Constitution was written. Because these rights were asserted in the context of equality, protecting them in the Constitution would mean that one person's life and liberty interests should only give way where they run up against the life and liberty interests of others. That is, we get John Stuart Mill's great principle of liberty: "that the only purpose for which power can rightfully be exercised over any member of a civilized community is to prevent harm to others," where "harm to others" in Mill's usage was limited to harm to the security or autonomy (the life or liberty) of others.(3)
Adding equal rights to the pursuit of happiness only reinforces this implication. It can't encompass the happiness of telling other people how to live because it isn't possible for everyone to have a right to tell everyone else how to live. Less technically, a right to pursue happiness is a right to discover one's own values and script one's own life. To interpret it as a right to deny this liberty to others would pervert its clear meaning.
Thus Mill's principle of liberty ought to be held to be protected under the Ninth Amendment. At the same time, Mill's principle is solidly grounded in moral reason (read Mill). Hence it stands as a principle of natural law (of what is right in itself, regardless of statutory or constitutional law).(4) The only other question is whether it captures the full ideal of protected liberty or whether further protections (in addition to the enumerated rights of the First and Second Amendments) would also be needed before indirect protections of liberty could be relaxed.
To come to grips with the sufficiency question, note that Mill's principle can be sharpened considerably by transcribing it into slightly different terms. Mill called people's security and autonomy interests their "direct interests, " as distinguished from indirect or "other-regarding" interests in the behavior or others, which might be described as vicarious interests in what others do or think.(5) Using this terminology, Mill's principle says that a law is improper if it gives anyone's indirect interests precedence over anyone else's direct interests. (It is simply coincidence that the priority of direct over indirect interests happens to facilitate the direct protection of liberty. There is no particular relation between these different concepts of directness.)
The distinction between direct and indirect interests can look problematic in that a person's indirect interests can affect his direct interests. My distress over your lifestyle might cause me a heart attack. Distress over American lifestyles apparently sends Islamic radicals into conniption fits by the thousands, prompting them to commit suicide (which certainly affects their direct interests) in order to commit murder. But this difficulty is easily handled. When direct interests are affected through ones indirect interests the interest is to be considered indirect. The distinction is perfectly workable. Essentially, it embodies the wisdom that every child is taught by its mother: "Sticks and stones will break my bones but names will never hurt me." Psychic interests are not to weigh against material interests. This is what allows pluralism. We don't all have to agree with each other. If we just refrain from attacking each other's material interests, then we can all thrive together. Thus not only is the distinction workable, but it seems to get to the nub of the issue.
Just for emphasis, notice that the priority of direct over indirect interests is exactly what the Islamist terrorists reject. The closest they come to listing any harm to their direct interests is their protest that, when Arabs try to murder Jews, the Jews defend themselves. Some grievance. The substance of the radical Moslem complaint is just the hatred they feel towards people who think or live differently than themselves (infidels). They hate our freedom and our influence. They hate the liberation of women. They hate the progress and prosperity that a free people create. Our success and example is a slap to their tenth century world-view. In response to this injury to their opinions, they wage war. They are a photo negative of the priority of direct over indirect interests that establishes liberty and enables progress.
Rights when direct interests conflict: slavery and self-defense
At first blush, it may look as if Mill's principle of liberty only asserts a priority of direct over indirect interests. If this were so, it would certainly be insufficient for articulating the full ideal of liberty, since some conflicts between direct interests must be resolved as a matter of right, not turned over to whatever majority rule decides. Slavery, for example, involves direct interests on both sides. The slave has a direct interest in being free and the slave-owner has direct interests in keeping him as a slave, yet the right not to be enslaved is one of our most hard won constitutional rights, a right for which Abraham Lincoln was willing to supersede the Constitution of his time in order to secure the natural rights of the Declaration of Independence.(6)
As it turns out, Mill's principle is not silent on conflicts of direct interests. Consider the principle of liberty from the positive side. The principle is stated negatively, delineating where a moral society must forswear the use of force. Looked at positively, the implication is that a moral society can use force on the other side of this line. Specifically, Mill's principle says that a person's direct interests can be interfered with to prevent harm to the direct interests of others. That is, it asserts a collective right to self-defense, which in turn implies an individual right to self-defense. If we don't have a right to defend ourselves individually then we can't have a right to defend ourselves collectively, because a collective is just a collection of individuals. A collective has no independent morality.
Once Mill's principle is seen to be built on an individual right to self-defense in this way, powerful implications for conflicts of direct interests emerge, because self-defense is fundamentally about conflicts of direct interests. For instance, the fact that Mill's principle is premised on a right to self-defense means that the majority cannot say that a murderer's interest in killing you (to take your wallet, or to eliminate you as a rival) supersedes your interest in your life, such that you are not allowed to defend yourself. Similarly with the case of slavery. Because Mill's principle is premised on the right to self-defense, it cannot view the direct interests of the slave-owner and the slave as similar quantities that it cannot decide between. Rather, it must treat the slave-owner as a thief. He is the aggressor, taking liberty from the slave.
Articulating self-defense for direct protection is a pretty simple matter. On the one hand, self-defense occupies only a small corner of modern society. As mankind has moved from a state of nature into a civilized state, legal principles of ownership and contract have minimized scope for legitimate disagreement over ownership of property and other claims. At the same time, mechanisms for resolving disputes peacefully have been set in place. These developments keep legitimate conflicts from leading to violent conflicts where self-defense can arise, leaving people only to defend themselves against common crime: the thief, the mugger, the rapist, the rival. This narrowing of sphere of self-defense, while leaving the concept and the right intact, makes it easier to demarcate.
Also, the right to self-defense is one of the few liberty rights that are already protected directly. The principle of self-defense is thoroughly imbedded in both statutory and common law; it is expressly protected in most state constitutions; and it is implicit in the Second Amendment right to keep and bear arms. (A right to defend oneself implies the right to the means to defend oneself. Formal protection for the latter is the practical implementation of the former.) This history of direct protection for the right of self-defense leaves a whole body of precedent for interpreting the right to self-defense and establishing its boundaries. Thus there is not a lot that needs to be sorted out in this area to enable a move to the direct protection of liberty in general. We just need to make sure that, in moving to the direct protection of liberty, our historic common-law conceptions of self-defense (and the right to keep and bear arms that it implies) are carried over intact to the sphere of protected liberty.
Unfortunately, the principle of self-defense is today under relentless attack by the same people who hate gun rights. England has virtually outlawed self-defense (with the consequence that it now has a far higher rate of violent crime than the United States), and self-defense is similarly under attack here in the U.S.(7) Still, the concept and principles of self-defense are well defined and can readily be articulated for direct protection. As to justification, not only does self-defense have its own long history of legal and philosophical support, but Mill's principle of liberty conforms to and reinforces these existing understandings.
There is one other place where conflicts of direct interests are sometimes resolved as a matter of right: in the area of economic rights. Note that economic behavior generally involves direct interests on both sides. Each person has a direct interest in being allowed to compete economically, offering his value to the market. At the same time, the act of competing typically hurts the direct interests of other competitors, even as it benefits the direct interests of most other participants in the economy. In spite of this conflict of direct interests, we know that some economic liberty rights emerge, otherwise we would be back into slavery. But how far do these rights go? How far do people have a right to compete economically without being impeded by protectionist laws?
As far as the present project goes, it is not necessary to answer this question. Yes, economic rights are a part of the full ideal of protected liberty, but they are matters of civil law, not criminal law, hence they do not need to be resolved in order to reform criminal law. The only proviso is that, so long as economic rights are not fully articulated, relaxed restrictions on crime control must not be used to prosecute civil violations.
Ultimately it is important to understand, articulate, and afford direct protection to, the full proper scope of economic liberty. Liberty is just as important here as elsewhere, and civil law is just as important as criminal law. Full articulation and direct protection of economic liberty would enable the same gains in the area of civil law as it offers in the area of criminal law. Restrictions on law enforcement could be relaxed, enhancing enforcement of the laws, while greatly increasing protection for properly protected economic liberty. Articulation of the full ideal of properly protected economic liberty would also add confidence to the present project. If our principles of right in the area of economic rights are at odds with our principles of right for criminal law, that would seem to indicate that something is wrong. Thus as a matter of thoroughness it will ultimately be desirable to conduct this check on Mill's principle, comparing it with other claims to right, particularly with such a close cousin as economic liberty rights.
Still, this analysis is best saved for another paper, so as not to muddy the waters about what is needed to justify the direct protection of liberty. Just as Military exigency is a supporting subject rather than a necessary part of the argument for a move to direct protection, so too the extent of economic liberty rights is only a supporting subject, the conclusions of which are not a necessary part of the argument for direct protection. The first step is to establish sufficient conditions for a switch to direct protection to be efficient. At that point supporting but unnecessary conditions can be addressed without the imputation that any disputes about these supporting arguments impugn sufficiency.
So far Mill's principle is looking pretty good. It comprehensively and compellingly articulates where conflicts between direct and indirect interests can be resolved as a matter of right. It also articulates where rights emerge in cases where direct interests conflict, at least as far as matters of criminal law are concerned (the concept and limits of self-defense). It does not resolve conflicts between direct interests in the area of economic liberty, but these are matters of civil law that need not be resolved in order to reform criminal law. That leaves only conflicts between indirect interests still to address.
If the only interests at stake on either side of a conflict are indirect interests then the Millian priority of direct over indirect interests is not in play. That means no liberty rights are established and the matter can be decided by majority rule. So where does this kind of conflict arise? Where does it happen that both the person engaging in an act and the people affected by it only have indirect interests in it? This odd animal is a creature of the public sphere.
Consider the example of public nudity. Each person has only an indirect interest in whether others dress in public, but the person who wants to go undressed in public is also interested in what other people think (by definition, if his interest is in publicity per se). Hence it is indirect interests on both sides. In contrast, while people don't always have a direct interest in performing a particular activity in public, they have a direct interest in being allowed to perform virtually any activity in private, as part of their liberty to pursue happiness. According to the principle of liberty, that means people are to have a right to do whatever they want in private, so long as other people's direct interests are not being harmed. The upshot is a right to privacy, not based on the value that anyone places on privacy for its own sake, but as a way to allow some outward looking behaviors and still consist with the priority of direct over indirect interests.
Not all interests in publicity are indirect. For instance, while the principle of liberty (or the priority of direct over indirect interests that it is based on) would allow the majority to ban public nudity, it does not allow the majority to make women wear bags over their heads, as Islamic fundamentalists seek to impose, because such a requirement hinders participation in public life, including the attracting of possible mates. These are direct interests, and hence take precedence over anyone's vicarious interests. Any Muslim woman who finds the Taliban's version of Islam compelling is free to cover her face, but a society that abides by Mill's principle cannot force her to do so.
In general, people have direct interests in being able to communicate with each other, but so long as plenty of scope is provided for people to search out mutually welcome communication, limitations on public expression can be imposed. If someone is into lewd come-ons, he can wear a tee shirt that says, "I'm a dirt-bag. Come talk to me." (More practically, he can go on-line.) He doesn't have to be allowed to talk dirty to people on the street. With direct interests allowed for, his remaining interests are indirect and can be interfered with. A liberal society isn't required to enforce public decency, but neither does it have to leave itself open to all manner of public incivility, as America has largely chosen to do.
The obvious way to protect privacy under a system of indirect protection for liberty is to make it hard for law enforcement to find out what people are doing in private. This is exactly how our system of liberty operates at present. The Court in Griswold v. Connecticut and in Roe v. Wade was not wrong to find an implicit right to privacy in the enumerated indirect protections of liberty.(8) The problem is that, while the object of this protection of privacy is to shield activities that are properly no one else's business, they also shield rape, murder, terrorism, and other activities which, by attacking the direct interests of others, are very much other people's business. Such is the nature of indirect protection. It necessarily protects what we don't mean to protect along with what we do.
In contrast, direct protection of Mill's principle would only protect what ought to be protected, where it ought to be protected. Mill's principle would make it explicit that anything that does not harm other people's direct interests can never be banned in private. We could then allow law enforcement easier access to people's private domains without raising the threat of prosecution under unjust laws.
Many readers will want to skip the next two sections, which discuss the different approach to privacy taken by J. S. Mill (in the form of his different approach to public decency laws). They contain a lot of somewhat interesting mucking around with what are ultimately small differences just to make sure that everything works out. Skip
Mill's approach to public decency laws
The above analysis of privacy and publicity, treating conflicts of indirect interests as properly subject to majority rule, proceeds logically from what we can say about what matters more than what, and the limits of what we can say about matters more than what. Where the priority of direct over indirect interests is not in play, the value of democracy should be deferred to, allowing substantial scope for public decency laws. There is however, another approach to public decency laws that bears mentioning: the approach that Mill himself took.
Mill also found public decency laws to be acceptable, but on different grounds than I have offered here. He viewed public acts as involving "manners," which he then categorized as involving "harm to others."(9) Unfortunately, interpreting mere "manners" as involving direct interests has a potentially disastrous liability: it threatens to annihilate Mill's distinction between direct and indirect interests. The principle of liberty states that the mere likes and dislikes of society are not sufficient grounds for violating anyone's liberty. (10) But if the psychic interests involved in manners are held to be direct interests then "mere likes and dislikes" are being used to justify violations of liberty and the principle of liberty protects nothing.
The problem with Mill's approach is that, in matters of properly regulable civility or decency, harm to direct interests will often not be the operative concern. Consider the case of Berkeley's Naked Guy (a UC Berkeley undergraduate who insisted on conducting his public life butt naked). Even in Berkeley, which usually embraces any kind of personal liberty (so long as it leans left) the Naked Guy garnered little sympathy, undoubtedly because his explicit motivation was to intrude on others. That is, he had no direct interests at stake, only a desire to engage others against their will in his own peculiar obsession, which people easily rank below their desire not to be engaged against their will.
Contrast this response with the much greater reluctance that people (not just in Berkeley) have shown in restricting panhandling. Being accosted for money engages people against their will much more directly than a passerby's nakedness does, yet public nudity is seen as much more certainly regulable. The difference is that going naked in public is gratuitous while panhandlers have a direct interest in panhandling, leading to greater tolerance for panhandling. This shows sound moral reasoning on the part of the public. The fundamental principle of moral reason is to account all value.(11) Thus the more stake that people have in a behavior, the higher should be the hurdle for interfering with it.
Mill's approach to decency violates this principle by allowing an activity to be subject to democratic regulation only when it harms the direct interests of others, regardless of what the actor has at stake. The gratuitousness of an intrusive behavior should lower the hurdle for regulating it, making it moral to subject it to regulation even when direct interests are not harmed, but Mill does not allow for this variation. Add that people are for the most part moral—that they have moral sense and feel a drive to act on it—and the result will be pressure through the democratic system for the appointment of judges who will ratify their judgement. If we require judges to rule that direct interests are violated before they will uphold a law, the result will be democratic pressure for the debasement of the concept of direct interests. In order to ban Naked Guy, people will exert democratic pressure to appoint judges that will classify their indirect interests in not having to look at Naked Guy classified as direct interests, at which point everything has become a "direct interest."
This would be extremely dangerous in the context of a move to direct protection. If the concept of direct interests is debased then the direct protection of liberty fails to protect liberty, and with the relaxation of indirect protections of liberty, liberty would be vulnerable to annihilation. The lesson is that if we go too far in protecting liberty, trying to protect more than is properly protected given what we have grounds to assert about what must be given priority over what, the consequence could well be less protection for liberty, not more.(12) Principles placed in the Constitution must align with correct moral judgement, or democratically expressed values will tend to undermine the Constitution. Any move to direct protection would have to be careful not to admit any errors like this that could undermine the whole project.
What Mill's approach to public decency gets right
Despite its dangers, Mill's approach to public decency laws at least glances the nail head. Surely there are cases where the operative grounds for public decency laws is to stop harm to the direct interests of others. Normal conflicts of direct interests, like who has right of way, can segue into issues of manners. Consider the Earth-shaking car stereo, which imposes on others, not through their vicarious interests, but directly, by forcing a claim on people's attention that obstructs their own use of public space. This is pretty clearly a matter of direct interests, and it is not implausible that the same interpretation could be stretched to some matters of visual as well as auditory imposition. Fornicating in public can plausibly be seen as forcing other people to be voyeurs, and it is certainly plausible to hold that people have a direct interest in being able to choose not to be voyeurs. Further, this claim to direct interests could be necessary for upholding public decency laws in some circumstances. Suppose some couple has no place to fornicate except in public. If the interests of passersby in not seeing public fornication were only indirect, direct protection of Mill's principle of liberty would mean that the couple would have to be allowed to fornicate in public. Mill's approach has a role to play in this case because direct interests are in plausibly involved on the offended party's side. The mistake is to limit democratic choice to those cases where direct interests are at stake. When all that is at stake on the perpetrators side is indirect interests, the indirect interests of the majority should be allowed to take precedence.
Not that the pure Millian approach to public decency laws is without its own appeal. Not allowing any activity to be criminalized unless it harms the direct interests of others would create a raucous public sphere where almost anything goes and no one has any choice but to put up with it. Some may well see this as the American ideal. Also, the gap between Mill's criterion and the principled criterion (where conflicts of indirect interests are turned over to majority rule) diminishes once speech rights and concern for minors are factored in.
Speech rights, having an extreme history in this country (at least until the recent criminalization of political speech), would have to remain on the wide-open side in any case, at least if we want to make sure that the move to direct protection nowhere diminishes liberty. This requirement would bring the principled criterion for public decency laws closer to Mill's criterion in the area of speech. At the same time, concern for minors constitutes a direct interest that lowers the bar for public decency laws when direct interests are required to be at stake. Absent concern for minors, Mill's criterion might require tolerance for Naked Guy in public. It does not require such tolerance once concern for minors is accounted.
Thus all things considered, there might not be much to choose between the principled criterion and Mill's criterion for public decency laws. Mill's criterion would have some tendency to cause the concept of direct interests to get watered down, but this could probably be mitigated by the simple expedient of being very clear about what direct interests are, perhaps including an admonition in the Constitution that if society rebels at what incivility may result from using Mill's criterion, its obligation is to change the Constitution, not pervert it by compromising the distinction between direct and indirect interests. Altogether then, Mill's criterion would probably work fine, and in any case should not be too dangerous. Still, it seems not the best, and there is certainly no allegiance due to Mill. His principle of liberty is a great achievement but the man was not an oracle, just a source, like the Constitution, to be taken from and left behind as advancing understanding calls for.
In sum, Mill's principle treats (as far as necessary for our purposes) four categories of cases. In two of them—where indirect interests conflict with direct, and where conflicts of direct interests involve matters of self-defense—it implies liberty rights. Conflicts between indirect interests it turns over either to majority rule (if we go with principle) or to liberty (if we go with Mill's criterion for public decency laws). Conflicts between direct interests that are not matters of self-defense it turns over to majority rule (leaving aside the subject of economic rights, which involves matters of civil law that do not have to be resolved in order to reform criminal law).
Through this typography, Mill's principle establishes a comprehensive scheme for deciding what is a matter of liberty rights and what can be left to majority rule. Comprehensiveness goes a long way towards sufficiency. It means we always face a well-posed question: what category does a given situation fall into? But sufficiency also requires being able to answer that question, and here too Mill's principle excels. The qualitative nature of the distinctions it makes turns hard cases into easy ones. Can the small interests of many people outweigh the greater interests of one? Instead of a quantitative answer we only need a qualitative one: are the interests direct or indirect? The envy and hatred of the entire Islamist world could not justify any harm to a single "infidel."
Articulating a comprehensive principle of liberty is one part of the puzzle. In addition, implementation of direct protection must be comprehensive. Loopholes and gray areas must be dealt with in a systematic way, guaranteeing a) that anything that clearly falls under the ideal of protected liberty will receive full protection, and b) anything that conceivably should be protected will receive at least as much protection as it does when liberty is protected indirectly. Not that the great gains to be had from direct protection couldn't justify some minor losses to liberty. Obviously they could, but the object here is to eliminate conflict so that we don't have to tolerate losses.
One key to implementation is the principle of necessity. In addition to distinguishing the kinds of interests at stake, we also have to be careful about what it means for an interest to actually be at stake. If there is a way to keep someone's actions from harming the direct interests of others without infringing his liberty, then harm to the direct interests of others is not really at stake and should not be held to justify loss of liberty. That is, Mill's principle implies a necessity test.
Consider the case of drug use. Drug use affects other people's direct interests by leading a fraction of drug users to commit crimes against people and property. If the only way to crack down on crimes against people and property by drug users was to crack down on drug use per se, then drug laws would pass the necessity test. But crimes against people and property can also be attacked by investigating and prosecuting crimes against people and property. Indeed, the war on drugs has made property crime the lowest priority in most jurisdictions. Few police departments even collect crime scene evidence in property crime cases. They only take reports, "secretaries with guns," as one San Francisco resident recently complained after having to arrest a thief himself when the police weren't interested.(13)
Because crimes against people and property prompted by drug use can be addressed by prosecuting crimes against people and property, interdiction of these crimes cannot justify punishment of those whose drug use does not harm the direct interests of others. Effects that are "merely contingent," Mill insisted, should be taken up "in the sequel," that is, in sequence, or when they occur. (14) If they can't be, then there is a case to be made that punishing an only sometimes causally related behavior is necessary in order to prevent harm to others. But in the case of crimes that follow from drug use, cracking down on crimes against persons and property has proved very effective, with petty thieves finally getting serious jail time under "three strikes" laws.(15) Thus according to the principle of necessity, prosecuting drug use itself is not justified. At the same time, the criminalization of drugs greatly increases the level of crime by creating a violent black market for drugs. What increases crimes against persons certainly cannot be held necessary to the objective of decreasing crimes against persons.(16)
Not only is a necessity test implied by Mill's principle of liberty, it is also a general principle of constitutional law. No constitutional provision is to give way, even to another constitutional provision, unless it has to. Where accommodation is possible, it is required.(17) At present, constitutional interpretation recognizes no general protection of liberty—no significant hurdle that must be met before any activity can be outlawed, unless it is directly protected by the First or Second Amendment.(18) This would change, however, if Mill's principle were adopted into the Constitution, either by amendment or as a way of embracing the rights of the Declaration under the Ninth Amendment. Then life, liberty and the pursuit of happiness would be protected generally and all limitations on liberty would have to pass a necessity test as a matter of constitutional principle.
In applying the test of necessity, it is important to be realistic about what it means for harmless alternatives to be available. Under our system of indirect protection of liberty, where unjust laws (laws that violate Mill's principle of liberty) are allowed to remain on the books, virtually every law enforcement technique is prone to uncover evidence of, and lead to the prosecution of, activities that ought to be allowed. Since these injuries could be avoided by shifting to the direct protection of liberty all current law enforcement techniques would seem to fail the necessity test. So long as the conflict between liberty and crime control can be eliminated by protecting liberty directly the conflict isn't real and hence can't justify harms to liberty.
But is direct protection of liberty really available at this time? The scheme for protecting liberty directly is only here being proposed. It doesn't become a real option until people have had a chance to understand it. For the short term, we are stuck with the indirect protection of liberty and the conflict between liberty and crime control that comes with it. Current inefficiencies were not chosen them over a more efficient alternative, hence it makes no sense to describe them as unjust. In the long run, however, as we discover how to move to a more efficient system, we have an obligation to do so. If we can but don't, the resulting unnecessary harms to liberty start to become moral crimes. This is true no matter where the necessity test is seen to originate: either as an implication of Mill's principle of liberty, or as a general principle of constitutional interpretation. If we can figure out how to move to the direct protection of liberty, we must.
Gray areas and mala prohibitem
If Mill's principle of liberty were protected directly, joining to it a necessity test would not imply that all gray areas should be given to liberty when deciding what laws to strike down (or not pass, or not sign). That would constitute a preference for erring on the side of liberty, causing a proliferation of improper "rights" and placing them above the will of the people. What we really want is to minimize errors. Congress should try to discern and abide by the best possible judgement about the implications of Mill's principle of liberty (assuming it is Mill's principle that we place in the Constitution as our full ideal of liberty) and judges should be instructed either to do the same, or to only strike down laws that infringe on clearly protected liberties (so long as they are assertive about how much Mill's principle does clearly protect).
Either way—so long as judges are not instructed to give all gray areas to liberty—a residual of possibly (but not very) unjust laws will be allowed to stand. If these laws are prosecuted under relaxed restrictions on law enforcement, there will be an increase in the prosecution of unjust laws in these areas. Liberty would not be everywhere at least as well protected as under indirect protection. The remedy is to give all gray areas to liberty at trial. Marginally questionable laws can stand, but whenever it is at all conceivable that a prosecuted activity could possibly be protected by the full ideal of liberty, all procedural restrictions should be retained in full force, as far as possible. This would mean, for instance, that any evidence gathered through any kind of search that would not meet current fourth amendment standards would be excluded from court. Neither would defendants be required to answer incriminating questions.
This way of treating gray areas acts as a fail-safe device. First we locate the full ideal of protected liberty as best as we can. As a result, some gray areas about what ought to be protected end up being protected directly. The remaining gray areas receive the same protections they did under indirect protection, sealing against any losses of liberty.
One element of this caution would be to treat a charged act as gray unless it is clearly wrong in itself (mala in se, or mala in se) instead of wrong in the face of man made arrangements or laws (mala prohibitem). This natural law distinction does not line up exactly with the distinction between just and unjust laws. Violation of tax law, for instance, is always mala prohibitem, but there can be just taxes. Why then, require mala in se before allowing evidence from relaxed procedural restrictions at trial? There are two reasons.
First, the justice of mala prohibitem laws necessarily stands on relatively sophisticated and obscure grounds, relying on the theory of democracy and the concepts of social justice that a democracy invokes when coming up with things like tax law. Such laws really may be just, but the grounds for this determination are much more extensive and uncertain than in cases of mala in se. Not only does this introduce inevitable elements of gray, it raises questions about whether people who violate even the most just mala prohibitem know that they are doing wrong.
When people refuse to pay taxes because their ancestors were denied forty acres and a mule, they may not have right on their side, but the grounds for this determination are much more extensive and uncertain than in cases of mala in se. If someone commits murder because his ancestor didn't get forty acres and a mule, he has simply committed a crime against an innocent. Because the act is mala in se, there is no possibility of its being right and hence no possibility of injustice in prosecuting it. If he refuses to pay taxes, there is an element of gray about where right lies, at least until the entire structure of right, including economic rights and social justice, can be resolved as certainly right, to the satisfaction of the entire nation, to the point where it has been placed in the constitution as a completion of and complement to the full ideal of protected liberty.
The same point can be understood directly in terms of the Ninth Amendment. The broadest way to understand "rights retained by the people" is as a general appeal to natural law. Thus anything that is not black according to natural law (mala in se) should be considered gray in the scheme of constitutionally protected liberty. When we move to a system of direct protection of liberty, gray must not be prosecuted under relaxed restrictions on law enforcement, or else we cannot be sure that protection of liberty is at all points being increased. Ergo, no prosecution of mala prohibitem under relaxed restrictions.
Can gray area protection by itself justify relaxation of restrictions?
To take the measure of this scheme for dealing with gray areas, consider the case where we fail to come up with a satisfactory articulation of the full ideal of liberty. In effect, this would leave a great big unresolved gray area about what should be protected. Could we still employ the general scheme for protecting liberty directly, complete with all the relaxed restrictions on crime control, just by leaning really hard on the scheme for treating gray areas? To some degree, yes, because the gray area about what ought to be protected has two boundaries, and an inability to articulate the full ideal of liberty only affects one of these boundaries. Difficulty in being sure where to draw the line between what clearly should be protected and what possibly should not be protected does not imply any difficulty in specifying many things that are clearly not to be protected. Terrorism and murder are clear mala in se. So is stealing a bicycle. Wherever there is not clear mala in se, the scheme for dealing with shades of gray says that liberty is to be protected by retaining indirect protections of liberty. Stated conversely, restrictions on crime control could be relaxed in the prosecution of clear mala in se, even if we cannot come up with an articulation of the full ideal of liberty.
Leaning entirely on gray area protection like this would be grossly inferior to a scheme in which we are able to articulate the full ideal of liberty for direct protection, since the absence of direct protection would mean that there was no increased protection for liberty. The best that such a scheme could do is leave liberty no worse off while improving crime control, which in itself is plenty important. The first steps in this direction would have to be an improvement over the system of blanket indirect protection that we have now. By relaxing restrictions only in the clearest cases of mala in se, much harm could be interdicted while little threat would be posed to any activity that the full ideal of liberty would conceivably protect.
The late John Kaplan urged a step in this direction when he proposed that the exclusionary rule be relaxed in the prosecution of the most serious crimes.(19) That is, he limited the relaxation of restrictions, not only to cases of clearest mala in se, but to those cases of clearest mala in se that also involve great harm to others (no bicycle theft). He also did not advocate relaxing restrictions very far. Where he called for the exclusionary rule to be set aside, he also called for alternative mechanisms for enforcing indirect restrictions to be put in place. That is, he did not advocate any intentional relaxing of police procedures, only the use of some evidence from police mistakes. Kaplan's suggestion is too conservative, but caution is certainly appropriate here. Indirect protections cannot be relaxed very much so long as our only protection of liberty is indirect.
Gray area protection and drugs
Recreational drug use is clearly protected under Mill's principle of liberty, yet it seems that a majority of the electorate leans strongly against drug legalization. We have had fluctuation on such issues in the past. Prohibition came and went. But suppose drug legalization were to remain a no go. Would that have to be a deal breaker for the move to the direct protection of liberty, or would it be tenable to classify drug use as a gray area of protected liberty at the constitutional level, even though according to Mill's principle it isn't? The legislature would then be able to criminalize or decriminalize drugs as it wishes, but no evidence gathered through the use of relaxed restrictions could be used to prosecute drug activity amongst consenting adults. This might be a workable solution. It allows a lot of benefits, but it also presents grave risks.
The main benefit is that it would allow direct protection to proceed. Restrictions on law enforcement could be relaxed and liberty in other areas would receive enhanced protection. Drug use would receive no greater protection than at present, but neither would it receive less protection. Symbolically, gray area protection could in itself be seen as a degree of decriminalization. At least society would be recognizing as a matter of law that drug use is not clearly wrong in itself. The risk is setting a precedent taking something that is not a gray area as a matter of natural law and treating it as a gray area as a matter of political compromise.
As with the case hypothetical case where everything is treated as a gray area (when we know that, if the full ideal of liberty were specified, not everything would be gray) such tentative protection for liberty would call for extra caution in the relaxation of restrictions on law enforcement. Still, if this is the only such issue, it should be containable. The move to direct protection will have to be accomplished by constitutional amendment in any case, with the full ideal of liberty fully articulated in the constitution. If drug use is enumerated as a gray area, it need not compromise the scheme of direct protection more generally. We would continue to suffer the tremendous costs and injustices that the drug war imposes, but there is no implication of further damage.
In general, the more we compromise on protection of the full ideal of liberty the more we need to compromise on relaxing restrictions on law enforcement, but as noted above, even a stand pure gray area scheme would allow some substantial relaxations of restrictions. Not protecting the full ideal of Millian liberty is clearly an inferior solution, but some political compromise in order to secure the great benefits of direct protection may be necessary, and would still be a huge step in the right direction, possibly establishing a respect for principled liberty that would one day lead to the full embrace of Millian liberty.
Example: a bio-metric national identification system
In the short term, the possibility of a stand alone gray area scheme could be pushed to the fore just by the fact that the direct protection of liberty is not here yet. Consider one likely half-measure, already being discussed, where indirect protections may well be relaxed in order to combat terrorism.
One very useful law enforcement tool, both for the war against terrorism and for fighting common crime, would be a national identification card that referenced bio-metric information—fingerprints, retinal scans, DNA "fingerprints" and such—stored in a central data-base. If all residents of the nation and all and visitors had their bio-metric information entered into the system then terrorists, domestic or foreign, would no longer be able to use aliases to conduct any business where identification was asked for. The system could also be used to identify any fingerprints or DNA evidence left at a crime scene. No one would ever get away with stranger rape again.
It is one thing to talk about relaxed restrictions in the abstract. Looking at an example is a chance to take the measure of the actual forces that would be unleashed, driven outwards by the good they can do. In a cancer medicine, ruthless effectiveness is the holy grail. But with ruthless effectiveness in hand, it could be awfully hard to stop from using it wherever the majority was aroused. The dynamics that would be set in motion by universal bio-metric identification illustrate the point.
Note first that instituting a bio-metric national i. d. might, and almost certainly should, require relaxing restrictions on law enforcement. At present, it is unclear the degree to which the collection of bio-metric information might conflict with Fifth Amendment rights against self-incrimination, or the Fourth Amendment requirement of probable cause for search. On the Fifth Amendment objection, the Supreme Court ruled in 1966 that the collection of fingerprints does not violate constitutional protections because fingerprints are is "not testimonial or communicative in nature," and this logic has been extended by state courts to DNA collection.(20) On the Fourth Amendment question, the U. S. supremes denied certiori in Jones v. Murray (1992) where it had been ruled that "the minor intrusion of a blood test is outweighed by the strong state interest" in solving criminal cases.(21) Whether Fourth and Fifth Amendment claims will become viable if testing is extended to people who are not suspected of a crime is a question that is yet to be adjudicated.
From the point of view of the direct protection of liberty, the problem with universal collection of bio-metric information is that it enables the harms that the Fourth Amendment is (at least in part) trying to protect against: the enforcement of unjust laws. Universal bio-metric indentification would give law enforcement a powerful tool for reaching into what is properly people's private business and enforcing all laws, not just those that legitimately involve the interests of society at large. The proper course would be to protect liberty directly (thereby striking down unjust laws) before allowing universal collection of bio-metric information, but direct protection is not yet on the radar screen, while in the wake of the attack on the Trade Towers, the tremendous law enforcement potential of bio-metric identification can no longer be overlooked. Civil libertarians all across the political spectrum are uniting to resist it (as in some degree they certainly should, when liberty is not protected directly) but law enforcement and the electorate as a whole are powerfully behind it.(22)
This brings the inefficiency of indirect protection into stark relief. Must we give up such a useful and even necessary instrument because we can't trust ourselves not to do wrong with it? Are we really that incontinent? Even if we cannot figure out where to draw the line on what government should not suppress, surely we can at least specify some things that government must suppress, and limit relaxed restrictions to prosecution of those activities.
Suppose we take the first step in this direction and allow universal bio-metric identification to be used only to track down, interdict and prosecute terrorists. In terms of the scheme for protecting liberty directly, everything but terrorism would be considered a gray area, where indirect protections of liberty are to remain in full force. All pairing of DNA or fingerprint samples with names would be excluded from court in the prosecution of anything but terrorist activity, unless the defendant became a suspect on independent grounds that would justify collection of fingerprint or DNA evidence under Fourth Amendment standards, or had been previously been convicted of a crime (implying probable cause to have been searched for these particulars). This first step would be nothing but positive. Society would gain an effective tool for combating terrorism without ensnaring anyone but the terrorists, and those common criminals who can be caught under existing Fourth Amendment restrictions.
But civil libertarians know it won't stop there. If a national DNA database identifies a serial rapist, we would be crazy not to use it to prosecute the rapist, even if he has not in any other way come to the attention of the police. Similarly for the murderer, the mugger, the embezzler. Once we start down this road we won't stop because it doesn't make sense to stop. It is a fundamentally correct direction. That means that before we start down it—before we start relaxing our indirect protections of liberty—we had better get to work on direct protection. Otherwise we could easily lose our liberty.
In peacetime, this would be a caution. When military exigency is forcing us down this road, it is an alarm. If we go beyond Kaplan's scheme and relax, not just the enforcement of restrictions on law enforcement but the restrictions themselves, we are playing with fire. The huge net gains from further incremental relaxation of restrictions will carry the relaxation of restrictions forward, sending it roaring towards town, towards our liberty. Not that we shouldn't play with fire. The human race would still be in darkness if we hadn't played with fire, but progress comes in learning to control the forces we unleash. In the present case, the means of control is the direct protection of liberty. Then we can bring the fire on into town, fully contained, even quietly muffled, like an internal combustion engine, doing the work of civilization for us without intruding on properly protected liberty.
A national identification system and economic rights
Actual containment of the forces unleashed by relaxing restrictions of law enforcement involves getting many details right. Consider, for instance, the treatment of cases that involve unresolved questions of economic rights and associated mala prohibitem. If a national i.d. card incorporating bio-metrics is adopted, it will very quickly become the standard identification for financial transactions, large and small. The government could easily collect and compile all this data on who is buying what and use it for such purposes as enforcing tax law, identifying who is living beyond their declared means. By requiring that vendors identify buyers even for cash sales, the government could clamp down on the underground economy, identifying, for instance, who was buying building materials on a commercial scale without having a contractor's license. But remember the general principle for when we can relax restrictions without losing our liberty. Restrictions can only be relaxed where we can determine that the laws being enforced are in full compliance with all relevant principles of right.
With criminal law, that means relaxed restrictions must not be allowed to contribute the to the prosecution of possibly unjust laws. Where there is gray, restrictions are to remain in place. With tax law, it means we must first resolve all the principles of a just tax system and guarantee them in the Constitution before relaxed restrictions can be used to enforce tax law. With the underground economy, it means that the full scope of economic liberty must be resolved and protected in the Constitution. If all that is protected directly is liberty as it concerns criminal law, these other areas of prosecution must remain off limits for the use of information gathered under relaxed restrictions.
That doesn't mean the information can't be gathered. For tracking down terrorists, it might well be crucial to be able to track people's transactions, both for identifying suspicious activity and for tracing the past and present movements, associations and activities of suspects. We just have to be very firm on the mala prohibitem exclusion of evidence from trial. Until such time as we are able to more fully explicate the implications of right, so that we can identify what mala prohibitem laws embody certain right, mala prohibitem must not be prosecuted under relaxed restrictions. Ultimately, our objective should be to identify what certain right is possible within mala prohibitem. It is important, for instance, that all pay their fair share of taxes. We just have to be certain that a fair share is in every case what we are asking for before we can fully unleash the power to enforce it. Such, at least, is the implied principle behind the mechanisms of indirect protection that our Constitution currently employs. Defendants are not required to answer questions because they might be guilty as charged of violating unjust laws. Until we get rid of the unjust laws, we cannot unleash enforcement without injuring properly protected liberty.
The prospect of a bio-metric national identification system also raises a question about Fourth Amendment protections that needs to be settled before we can figure out how these protections could be relaxed under a system of direct protection. The requirement of probable cause is simply an instrument for restricting search. It is not a principle of right in itself. But what about the requirement that search be reasonable? Reasonableness is in itself a basic element of right that we would not want to give up. Does a compulsory bio-metric identification system violate reasonableness?
Reasonableness can be understood in contrast to the broader concept of rationality, which is defined in terms of suitability to the pursuit of a specified end. An action or mechanism is a rational if it efficiently does what it is trying to do. Reasonableness is much more specific concept. Where rationality pertains only to the suitability of means to ends, reasonableness critiques ends as well as means, and it does not just look at consequences for a particular end, but at consequences for all ends, weighed as full information about the value of different ends demands. Reasonableness accounts all of what is at stake and all that can be said about what matters more than what.
Focussing only on the goal of crime control, it can be rational to admit almost any kind of search, but reasonableness is a whole other question. In the absence of the direct protection of liberty many kinds of search aren't reasonable because their tendency to promote the enforcement of unjust laws would do more damage to public and private good (in the form of enforcement of unjust laws) than can be offset by their legitimate crime control benefits. But because reasonableness accounts every relevant thing, it can change as circumstances change. Thus a search that is unreasonable when liberty is not protected directly can become perfectly reasonable when liberty is protected directly, and this is just the claim that is here made for the direct protection of liberty.
Once direct protection against unjust laws is in place, the harms from search are greatly diminished, allowing reasonable search to greatly expand. The principle of reasonableness remains the same, but judgements of reasonableness are conditional.
Part IV: Protection of privacy per se
Suppose that we are satisfied with Mill's principle as an expression of the full ideal of liberty, and we have been careful to implement it in a legally comprehensive way. These steps would secure protection against unjust laws. If protection against unjust laws were the only concern, direct protection of Mill's liberty would mean that all of our existing indirect protections of liberty could simply be dispensed with. They would be completely superfluous, since the protection they offer against unjust laws would be greatly exceeded at every point by direct protection. But protection against unjust laws is not the only concern. First, there is the threat of tyranny to worry about. Those in power may not obey the Constitution and the laws. Thus it isn't enough to get injustice out of the laws. To protect against injustice, it is also necessary to protect the rule of law. Second, some restrictions on law enforcement—especially the restrictions on search and the right not to incriminate oneself—afford some protection for privacy per se. Consider the latter first.
Earlier a right to privacy was derived as a means for protecting, not privacy, but liberty. When the only conflict is between indirect interests, Mill's principle (or the priority behind Mill's principle) allows the majority to rule, yet a person always has a direct interest in doing what he wants in private. Thus according to Mill's principle, the indirect interests of the majority cannot invade a person's private space. This protection of privacy is not based on any recognition of the value of privacy per se. The logic of this protection for privacy would be perfectly well served if law enforcement were allowed to invade people's privacy in almost every way, even keeping tabs via Big Brother type video monitors, to insure that no mala in se was occurring behind closed doors, so long as consensual activity were not punished. To object to snooping that does not threaten prosecution under unjust laws it is necessary to go further and bring in the value of privacy per se. Privacy must have value, not just as a means, but as an end in itself, and of course it does. It matters to us to be able to choose whom to reveal ourselves to.
How should we view this value within the Millian framework? Is it a direct interest, or as an indirect interest? Formally, it looks like an indirect interest, since it is an interest in what other people think, yet intuitively, most people would probably insist that privacy per se is a direct interest. If others snoop into your house, if they invade your private moments, then they are invading your liberty and security. Further, if privacy is only an indirect interest, it would seem unable to justify any limit on Big Brother type surveillance. Virtually any type of surveillance is of some value for prosecuting mala in se. If all that stands against this direct interest is an indirect interest in privacy, then according to the priority that underlies Mill's principle—the priority of direct over indirect interests—the law enforcement interest should take precedence. True, Mill's principle makes no such positive statements about how government power should be used, only negative statements about how collective force should not be used. Still, the underlying priority clearly militates in this direction. If liberty is to be protected directly by Mill's principle, then giving privacy the protection we intuitively want to give to it seems to require interpreting privacy as a direct interest.
In this way, the analysis of privacy per se is similar to the earlier analysis of privacy as a means. Both tread the boundary between direct and indirect interests. Both offer a temptation: whether to stretch the concept of direct interests, so as to justify public decency laws (in the earlier case), or to increase protection for privacy (in the present case). If this stretching is inappropriate, it runs the risk of undermining the distinction between direct and indirect interests, placing the entire protection of liberty in peril. Earlier, the principled course was to resist temptation and hold to a narrow concept of direct interests. Here I think principle goes the other way. The crux of the issue is the same in both cases, but one crucial element is turned around: the principle of self-defense.
Invasion of privacy and self-defense
The person whose interest is in publicity is choosing to go out into the world and affect what other people think about her. A person's interest in privacy, in contrast, is her desire not to have worries about what other people will think forced upon her by an invading world. These are almost opposite concerns. One is aggressive. One is defensive. And as we saw earlier, Mill's principle implicitly embodies the principle of self-defense. A right to self-defense is a right to use force. As we move to rule of law, any instance where an individual would have a right to self-defense is an instance where the collectivity can legitimately substitute its force. Essentially, the principle of self-defense offers an opportunity to look through the binoculars of Mill's principle from the other side. Where self-defense is justified, direct interests are implicit.
Consider freedom of speech. Certainly people have very broad direct interests in speech. But people can also have direct interests in being protected from speech. When a person who won't take "no" for an answer accosts passersby, their interests are impinged upon, not through their interests in what other people think, or through their vicarious interests in other people's lives, but through the accoster's interference with the conduct of their own lives. This is a direct interest. At some point, speech that imposes on others regardless of their desire not to be engaged harms direct interests and may be a proper subject for regulation, even when the speaker has direct interests of his own.
Harm to direct interests can only occur when people are imposed upon against their will. When imposition against one's will reaches a point where self-defense is justified, interests must be considered direct. But self-defense from public speech is an iffy concept. It isn't like a physical attack or an attack on one's property where the invasion of self is manifest. Further, no one can claim to own any public space, while all can claim rights to fair use, making the weights on the scale very hard to pin down. Laws against aggressive panhandling have been upheld, and other restrictions on highly invasive speech are certainly tenable, but this is a risky road to go down.
The situation changes dramatically when the context changes from the public sphere to the private sphere. Here, when one is impinged upon, there is ownership, and rights to self-defense become clear. One element of the concept of self-defense is the principle of retreat. It is not a necessary element. A right to defend oneself means a right to stand and defend oneself. A person should not have any legal obligation to run away. Still, retreat is an important supporting concept. If a person does retreat, and is advanced upon, his claims to self-defense are clearly strengthened. When a person's privacy is invaded, his claim to self-defense on this grounds is as strong as can be. He has retreated from the public sphere, and has been pursued. If non-violent impositions against a person's will in public can possibly qualify as harming direct interests, then invasions of privacy in private have a much stronger claim to harm to direct interests because they are much more clearly matters of self-defense.
A similar case is that of libel. The value at stake in a libel case is reputation, which does affect one's direct interests, but only indirectly, through what other people think. As such, it looks at first glance that libel should be classed as harm to indirect interests, which would not justify any use of force to stop it. After all, no one has a right to have others think of him as he wishes. Looking more closely though, people do have a right to defend themselves from attack by force or its substitute, fraud. Libel is fraud, hence it is properly looked at as a direct interest. Invasion of privacy is force, hence it too is properly looked at as a direct interest.
Least invasive law enforcement first
If privacy per se is considered a direct interest then, when law enforcement invades privacy in search of mala in se, there are direct interests on both sides. The priority of direct over indirect interests neither calls for invasion of privacy, nor forbids it. What is implied is a preference. If any law enforcement tools are to be eschewed out of deference to privacy per se, it should be those that do much to invade privacy while doing little to interdict crime. At the one end of the spectrum there are a variety of what might be called non-invasive types of systematic or random search that peer very little into people's lives while being highly valuable to law enforcement, such as collecting DNA samples from everyone which can then be tested against evidence left at crime scenes. Some uses of this information could invade privacy if they were allowed. As we learn more about what people's genes say about them a person's genome becomes very much a matter for privacy, but allowing law enforcement to use genes solely for the purpose of identifying suspects does not nose in on anything that is properly private.
The same principle could be followed more broadly. Even the Big Brother video screen could be non-invasive if a system were in place to insure that the surveillance would only be examined where a crime was independently evidenced or alleged, or examined by non-invasive techniques, such as computer searches for the faces of missing persons about whom foul play is suspected. The problem is that, if inappropriate access to this kind of surveillance were to occur, it would be highly invasive. Even if government leaders were to show complete integrity in trying to keep potentially invasive surveillance from being abused, no guarantees are possible, and the potential injury is large. The possibility of having one's private embarrassments made public would be like a reverse lottery ticket, where you get to lose as much as you can imagine. The reality of these costs (people buy lottery tickets all the time) means that no kind of potentially invasive systematic or random search can ever be allowed without powerful law enforcement rationale, no matter how non-invasive it would be under ideal circumstances.
So long as we start with those kinds of search that have the greatest law enforcement bang per risk-to-privacy buck, the kinds of surveillance that are most threatening to privacy will surely not be justified. With all the law enforcement mechanisms that would already be enabled, the marginal enforcement value of watching people we have no reason to suspect should be negligible.(23) More invasive techniques are also likely to be more expensive, so monetary considerations should reinforce this conclusion.
Of course the war on terrorism could change all of these calculations. If unknown people are working on weapons of mass destruction in our midst, self-defense may require turning over every stone, but this issue really doesn't need to be considered here. If an extraordinary measure is needed for military purposes, it can be called for by a president invoking his war powers. No provision need be made ahead of time when framing a system of direct protection. Neither are exigencies of war particular to direct protection. They will occur no matter how liberty is protected. What we can say is that, if Orwellian mechanisms ever are called for as an exigency of war, direct protection will minimize the harms, keeping the extraordinary surveillance from being used to prosecute unjust laws.
Privacy per se and self-incrimination
The history of the Fifth Amendment right not to incriminate oneself makes clear that its primary purpose was as a protection for those who were guilty as charged of violating unjust laws, in particular, the English blasphemy laws. Under the direct protection of liberty this purpose would still be in effect so long as there were any uncertainty whether a prosecuted behavior might properly fall under the full ideal of protected liberty. According to the implementation scheme, all indirect protections, such as the right not to incriminate oneself, would be retained in all cases of gray. Outside of any gray area, when dealing with certain mala in se, the right not to incriminate oneself would become entirely superfluous. No possibility of prosecution under unjust laws exists and so the right not to incriminate oneself does nothing to stop this injustice.
At the same time, the right not to incriminate oneself is extremely detrimental to crime control. Without the Fifth Amendment, guilty suspects and defendants who want to try to get away with crime would have to spin lies, and risk getting caught in their lies, to the great advantage of justice. Being able to hide information is extremely useful to the guilty and of relatively little value to the innocent. Thus unless there are other purposes at stake besides protection against unjust laws, the right not to incriminate oneself should be eliminated when moving to the direct protection of liberty, retained only as a gray area protection, when the prosecuted activity is not clear mala in se.
There are other purposes at stake besides protection against unjust laws. One of the original purposes of the right not to incriminate oneself was to counter the oath ex officio, employed by Star Chamber and the Court of High Commission. This was an oath to truthfully answer all questions asked on any subject, whether or not there was any grounds to suspect the violations that were inquired after. If the fruits of such questioning was only used to prosecute clear mala in se, such questioning could be looked at as a kind of universal surveillance, like bio-metric database searches or Orwellian video surveillance, except where those kinds of search are only potentially invasive, the oath ex officio is inherently invasive, publicly exposing the surveillance in toto and demanding that the suspect or defendant actively participate in this exposure. Thus it necessarily and powerfully conflicts with concern for privacy per se.
In relaxing the right not to incriminate oneself, it is necessary to account where this might allow invasion of privacy to occur, then limit relaxation to where the value of privacy is outweighed by more important direct interests in law enforcement. At trial, this can be (and is) accomplished by judicial oversight, as judges require that questions be directly relevant to the charged violation. With this oversight in place (and liberty protected directly) the right not to incriminate oneself could be completely dispensed with at trial when the charges are clear mala in se.
That is big, but it only covers one of the areas where the right not to incriminate oneself comes into play. Another area of Fifth Amendment protection is a right to remain silent during police questioning, which the Court has expanded into a right not to even be questioned, if a person chooses to exercise it. Relaxing these restrictions would mean some degree of mandatory submission to police questioning.
The same general assessment applies here as to a right not to incriminate oneself at trial. A right not to answer questions from the police is far more valuable to the guilty than the innocent. Once direct protection has eliminated prosecution under illiberal laws, a right not to answer questions is about the last restriction it makes sense to impose, in terms of protection of the innocent per cost to crime control. Yet police questioning raises its own special problems. For instance, on the matter of privacy, so long as police questioning is to the point in a reasonable investigation of mala in se, concern for privacy per se is outweighed. But what is going to keep police questioning on point?
Judicial oversight of police questioning
To avoid crossing over into the oath ex officio, mandatory submission to questioning would have to be accompanied by some oversight mechanism to keep questioning on point. At present, restrictions on police procedures are enforced by excluding improperly gleaned evidence at trial. Presuming that police are motivated to convict criminals, this creates a disincentive to break the rules. The same structure could be used to enforce rules for mandatory submission to questioning. What would change is the rules that are being enforced. Now the issue is voluntariness of answers. Under the direct protection of liberty, answering would not be voluntary. The issue would be the grounds for and relevancy of questioning. Note that any questioning that leads to evidence in the matter under investigation or any related matter would presumptively not be excluded. The questioning would have proved relevant. The purpose of the restriction is not to hamper police investigations in any way, but to keep the police from randomly invading people's privacy.
Exclusionary rules could also be used to protect the innocent from questioning techniques that are prone to elicit false confessions. Now the line is drawn at any kind urging that questions be answered. Under direct protection, pressure to answer questions would be a given and the issue would be whether a suspect was pressured, not to speak the truth, but to give a particular answer, whether or not it is the truth. It is important not to define the pejorative concept of an unreliable "forced confession" too broadly. When a guilty person is pressured to answer questions truthfully, the truth is a confession. So long as the pressure is to answer truthfully, there is nothing unreliable about the confession and no reason to exclude it. Pressure to answer truthfully gives justice to both the guilty and the innocent.
The problem is that some of the most useful questioning techniques are explicitly not fixed on the truth. For instance, it can be very useful, in questioning a guilty party who is not forthcoming, to ask him to speculate on what might have motivated whoever did commit the crime in question. Once they start talking, guilty parties sometimes turn out to have a lot they want to unload. On the other hand, innocent people who, trying to be cooperative, will sometimes indulge in the asked for speculation and in doing so manage to make themselves sound like criminals. There are many documented cases where this kind of questioning has led to prosecution and even conviction of the innocent.(24) Such techniques are inherently unreliable, similar to the repeated questioning of small children, lie detector tests, and questioning under hypnosis or sodium pentathol. Testimony gathered through these techniques should be admitted under advisement of unreliability, if at all.
Excluding testimony for unreliability is a very different thing than excluding testimony because rights were violated. At present, self-incriminating testimony that is not judged to be completely voluntarily is excluded by right. As a consequence, any physical or other evidence uncovered as a fruit of that testimony is also excluded at trial. Under direct protection there would no longer be a right not to incriminate oneself, hence exclusion of testimony on grounds of unreliability would not taint the fruits of that testimony. Physical and other evidence uncovered through questionable questioning would be admitted on its own merits. The doctrine of "fruit of the poisoned tree" would no longer apply in cases of certain mala in se. This allowance would do much to catch the guilty while posing an insignificant threat to the innocent. Such is the general tendency of relevant information, which is why, when liberty is protected directly (removing all reason to protect those who are guilty as charged) relaxed restrictions are called for.
What kind of pressure should be brought to bear on suspects who are uncooperative in answering questions? Since cooperativeness would be required, it would be appropriate, with proper judicial oversight, to compel cooperativeness by such mechanisms as judges use now to compel cooperativeness in court. Those who demonstrate contempt of court can be held indefinitely until they choose to be forthcoming, though the activity under investigation would have to be clear mala in se, otherwise the right not to incriminate oneself would remain in force. Similarly, if the questions asked are not directly relevant to the investigated activity, a person would retain the right not to answer, and could not be properly be held in contempt for refusing to answer. This would extend to the answering of hypothetical questions. Only the truth is relevant. Asking people to imagine themselves in the person of the perpetrator may be asking the guilty to tell the truth, but it is asking the innocent to engage in a fiction, which is a lie. Accordingly, it cannot be compelled. Regarding matters of fact, however, the threat of contempt procedures is appropriate.
Such a mechanism would raise the question of how uncooperativeness is to be determined. Uncooperativeness may be overt, as when someone simply refuses to speak, or gives snide responses. But what of the suspect who claims not to know anything or says he can't remember? How can it be determined whether he is telling the truth or is being recalcitrant? Other evidence may sometimes point to knowledge, as when a thief says he can't remember the name of an accomplice who got away, or says he had no accomplice when other evidence indicates he did. Also, lie detector tests could be applied. Falsely claiming to not know or remember are just lies like any other and sometimes lie detectors can show these things up, giving a judge grounds to hold a suspect, or if the matter is urgent, to order sodium pentathol, or other questioning techniques that society should not want police to engage in without judicial oversight.
There are many cases where such measures are absolutely called for according to a weighing of the different concerns at stake: the criminal conspirator who won't reveal his accomplices, the hijacker who misses the plane and refuses to say a word about his terrorist network. No concern for unjust laws is in play (even when liberty is not protected), and concerns for privacy per se are completely dominated by weightier concerns. Protection for the innocent is a concern only on the other side of the balance. When information about mass murder plans is being withheld, even torture is not a hard question.(25) What holds us back from acting on rational priorities now is simply that we are hamstrung by indirect protections which by their nature apply to what we don't meant to protect along with what we do. Once we separate these categories we can limit protection to what we do mean to protect.
Other areas of Fifth Amendment protection might be due for relaxation as well. Attorney client privilege is an extension of the right not to incriminate oneself. Not only are suspects and defendants allowed to keep even relevant facts hidden, they can also enlist help in doing it. Another derivative of the Fifth Amendment is the asymmetry of discovery in criminal trial, where only the state is required to turn over its evidence. As with all incarnations of the right not to incriminate oneself, these restrictions on law enforcement are far more valuable to the guilty than to the innocent. Still, that doesn't mean we should get rid of them.
There is substantial weight on both sides. If privilege is out the window then to get the benefit of effective council, a suspect or defendant has to come clean. In cases of mala in se, that is exactly what we want. For those who want to keep trying to get away with things that are not possibly protected by natural right, there is no interest on the part of justice for them to have any kind of assistance in this endeavor. Still, there could be significant costs to giving up the privilege. People who are guilty as charged of things that might possibly be protected under the full ideal of liberty very much need good lawyering, but if they don't know what will be held to be possibly protected they won't know whether privilege will be allowed and hence would have reason to fear revealing anything to a lawyer.
A solution might be to draw a Kaplanesque line, only eliminating attorney client privilege in cases, not only of clearest mala in se, but also of extreme magnitude, or extreme magnitude combined with continuing threat. This would set the line for elimination of privilege well away from any legitimate questions of about what is and is not clear mala in se. Thus no defendant who had legitimate questions about whether his case would be treated as clear mala in se would have any reason to fear seeking the aid of an attorney. There would no longer, in these cases, be any reason retain attorney-client privilege and so it should be dispensed with.
According to this standard, it is perfectly appropriate for Attorney General John Ashcroft to refuse to let terrorist suspects have private conversations with their lawyers. No legitimate interest is served by allowing attorney client privilege in these cases, any more than by allowing terrorism suspects not to testify at trial. Neither those who are prosecuted under unjust laws, nor those who are innocent as charged, reap any benefits from this privilege, it does much to inhibit prosecution of the guilty. In the case of terrorism, attorney-client privilege can also give terrorist suspects an avenue to transmit strategic information back to terrorist collaborators, making it appropriate to suspend this privilege for terrorist suspects even when a system of direct protection of liberty is not in place.
Under direct protection, and when a Kaplanesque standard is not met, attorney-client privilege should almost certainly be retained. In general, as restrictions on law enforcement are relaxed, crime control will become more effective until at some point, further incremental relaxations will achieve little reduction in crime. If we start with those relaxations that yield the greatest crime control bang per risk to the innocent, and per concern for privacy per se, and per concern for rule of law, and per dollar (ultimately per whatever combination of benefits is delivered, converted into dollars or some other chosen numeraire), then almost any restriction that is of substantial value in protecting the innocent will probably be worth retaining.
In particular, limiting relaxing attorney-client privilege to cases, not just of clear mala in se, but also of a severe magnitude of mala in se, would probably do only minor damage to crime control. On the plus side, those who are guilty as charged of acts that are not certain mala in se, but who fear that their acts may be ruled certain mala in se, would not be deterred from confiding in an attorney, allowing them full access to the greater protections and leniency that we should extend in cases that do not involve certain mala in se. Altogether, a Kaplanesque line on attorney client privilege seems correct.
Part V: Maintaining rule of law
The framers of the Constitution were very much concerned about the threat of tyranny. Many of our indirect protections of liberty—habeas corpus, warrants issued only for probable cause, the prohibition on ex-post facto laws, the right not to incriminate oneself—are partly intended as obstacles to arbitrary power. When liberty is protected directly, these indirect protections of liberty will still have some value as bulwarks against tyranny. This is a direct interest, as interests in crime control generally are, meaning that neither crime control nor concern for tyranny automatically trumps the other. It is a matter of how to serve each without too much damage to the other. The strategy, then, is to identify for possible retention those restrictions on law enforcement that offer the biggest bang in terms of protection against tyranny per forgone-crime-control buck.
As a benchmark, notice that, in general, threats to the rule of law should be greatly mitigated by a shift to the direct protection of liberty. Just as indirect protections of liberty have the additional benefit and purpose of protecting the rule of law, so too does direct protection of liberty have this additional benefit and purpose, and just as direct protection protects liberty better than indirect protection, so too does it present more of an obstacle to tyranny.
Consider the protection from tyranny afforded by the requirement of probable cause for the issuance of warrants for search. What obstacle does this requirement pose to incumbents who would use the law to attack legitimate political opposition? It presents no obstacle at all when there is probable cause to suspect that unjust laws are being broken. Do your political opponents do drugs? Just wage a war on drugs, as President Nixon initiated at the federal level in order to wage war against the hippie counterculture that hated him. When probable cause to suspect violation of unjust laws is missing, incumbents can still use unjust laws to attack their opponents if a compliant judiciary is willing to be lax on probable cause. No wide conspiracy to usurp power would be necessary to attack political opponents.
Contrast this with a system of direct protection of liberty, where restrictions on probable cause will be relaxed or eliminated, but where unjust laws will be systematically purged. In this environment, attacking opponents with unjust laws would require both a complicit legislature (to pass unjust laws) and a complicit judiciary (to let them stand), or a compliant military (to make opposition by legislative and judicial branches irrelevant). In short, a wide conspiracy would be necessary, wider than is necessary when liberty is protected indirectly.
Width of necessary conspiracy is the crucial criterion for how easy it is to usurp power. If the separated powers conspire to usurp, then the only recourse is for The People to exercise the first purpose of their Second Amendment rights. Prior to a citizen uprising to defend the republic, the main obstacle to tyranny is how wide a conspiracy it requires.
Arbitrary prosecution under unjust laws is the tool that it is most necessary to keep out of the hands of would be usurpers. If political opposition can be relatively safe just by not being genuinely criminal then the republic is relatively safe. Since direct protection provides much stronger protection against unjust laws than indirect protection, it is inherently much stronger against usurpation. On the other hand, usurpers might find it marginally easier under the direct protection of liberty to use mala prohibitem to attack political opponents.
An example here would be IRS abuse. Relaxed restrictions on law enforcement could make it easier to find out who was cheating on their taxes, making it easier to target political opponents with tax law. But the IRS is already in a position to audit political opponents if it is willing to break the law. All that the easing of indirect protections would do is make it easier for investigators to look behind a tax return, making audits more punishing. Under the scheme of direct protection, information uncovered under relaxed restrictions on law enforcement is not to be used in prosecuting mala prohibitem (including tax law). Thus the more punishing abuse that is possible under direct protection requires the complicity of the courts (in this case, the tax courts). Still, this could constitute a minor weakness compared to the system of indirect protection of liberty.
Where direct protection would definitely make it easier to attack political opponents is in the prosecution of clear mala in se. In general, prosecution of mala in se becomes much easier to prosecute when liberty is protected directly and indirect protections are relaxed. If we are going to stop making law enforcement jump through hoops, the increased discretion will necessarily afford executive officers more leeway make sure political opponents aren't committing clear mala in se. To the extent that this might happen, it is a question whether it would be a good thing or a bad thing. Legitimate political opposition does not commit mala in se. Should the Black Panthers—a criminal street gang that branched out into political demagoguery—have been a target for prosecution? Absolutely, simply as a matter of unbiased law enforcement. The fact that there were a few dupes in with the murderers shouldn't have kept the police from going after the murderers (as for the most part it did). Legitimate opposition groups will also have their bad elements, but with all real criminals being much more efficiently prosecuted than at present, it wouldn't take politically motivated investigations to catch them. Thus the residual advantage that could be gained from politically motivated investigations would seem to be vanishingly small.
Altogether then, so long as the separated powers do not conspire to usurp the republic, direct protection of liberty is not more vulnerable to tyranny than indirect protection, even if restrictions on law enforcement are eliminated entirely in the prosecution of clear mala in se. The enhanced protection against unjust laws that direct protection affords does far more to restrain tyranny than the enhanced power to enforce just laws does to enable it. Still, it is also necessary to consider the case where separated powers do conspire to usurp the republic. It is over-simple to say that in this case the usurpers can do whatever they want and so it does not matter what government is already empowered to do. If instruments that would be especially useful to tyrants and totalitarians are already in place, restrained to proper purposes only by law, they could become the instruments by which tyrannical power is achieved.
Restricting what could aid the rise of a tyrant
This is the ultimate argument against Big Brother video surveillance of private places. With liberty protected directly, legal restraints on examination of the surveillance could render it non-threatening, even to privacy per se, so long as rule of law is maintained. But if the nation ever does find itself in the breach, fighting against a would-be tyrant, universal surveillance would give a tremendous advantage to the would-be usurper, allowing him to pinpoint his enemies, or more mundanely, to find out which of his political opponents are having affairs, etcetera, that can be used to attack them politically. Given that the marginal law enforcement value of such measures should be negligible (with all the other law enforcement mechanisms that would already be enabled), the threat from tyranny dominates and Big Brother should not be allowed in private places, no matter how access to this surveillance was limited by law.
Similar arguments can be made against the keeping of certain kinds of information. Once a tyrant gains power she will try to find out who the loyal democrats and republicans are so she can suppress them. Obviously that information should not be kept in government records, already available where a would-be usurper can use it to more easily achieve tyrannical power. Such information has great potential to be misused and zero legitimate law enforcement value. Thus no information should be kept except what is relevant to the prosecution of mala in se.
These rationales for maintaining restrictions on law enforcement have clear limitations. Any mechanism that could be quickly deployed by an out of control executive is not something that can be limited out of concern for tyranny, because the restriction does not succeed in restricting the would-be tyrant. Thus for instance, it wouldn't make sense to place restrictions on interrogation techniques to limit usurpers because usurpers will immediately deploy whatever interrogation techniques they want. Further, when hard to deploy mechanisms have high law enforcement value—like European style video surveillance of public places—concern for tyranny will typically not outweigh concern for crime control, since concern for tyranny is highly speculative while crime control is a constant need. The wider limitation that does stand up to scrutiny is on the compiling of information from such surveillance. Dossiers should not be kept on anyone not suspected of mala in se.
Again, exigencies of war could change these calculations, and again, it isn't really necessary to address that here, except to note that the fewer the indirect protections left standing in the wake of military exigency, the greater the need for direct protection.
An example of an indirect protection of liberty that provides important protection against tyranny while imposing relatively little burden on law enforcement is habeas corpus. The power to hold people for more than a short period without charges can easily be used against political opponents, allowing descent into despotism, yet its only value in the enforcement of just laws is for holding suspects when there is insufficient evidence to bring charges. This value can be discounted because holding mere suspects is at odds with the principles that our system of criminal law ultimately acts on. We ask our juries to acquit those who are almost certainly guilty, at huge cost to crime control, rather than run more than the smallest risk of punishing the innocent. If we were going to make a concession to crime control here it should be to relax the standard of guilt, not to jettison due process by allowing mere suspects to be held without charge. Thus holding people without charges is simply a tool of tyranny and should remain banned, even when liberty is protected directly.
There is, however, some room for flexibility in the length of time that suspects can be held before charges must be brought. When law enforcement authorities are pursuing leads and have evidence that a suspect is guilty, but have not yet built a case, should they be forced to either free a suspect or to bring charges pre-maturely (possibly allowing a guilty person going free) after only two days, even when the person, if guilty, would likely flee or commit further mayhem? Of course this question can be asked within our current system of indirect protection of liberty. The question here is whether a shift to direct protection changes the calculations involved, and it seems it does.
Direct protection distinguishes between just and unjust laws and between mala prohibitem and mala in se. These distinctions can be used to treat different cases differently. In cases of clear mala in se, the habeas corpus requirement could be that the body of the defendant be brought before the court, not to enter a plea to charges, but to hear if there is reasonable suspicion of involvement in mala in se that warrants holding the suspect for up to another couple of days as the case is investigated, at which point the hearing could be repeated. If the process drags on, with several continuations, a Kaplanesque requirement should also be imposed: that the suspected crime must be grave and the risks of freeing a possibly guilty suspect must be great. This would follow the pattern of direct protection generally. To the extent that we can restrict our focus to what we clearly do not mean to protect, we can relax our mechanisms of protection.
The weights that habeas corpus puts on the scales change dramatically in times of war. Where civilian law is focused on punishment of those who have done wrong, military exigency must focus on the interdiction of harm. Habeas corpus is only relevant to the civilian standard of punishment for harms already inflicted. A specific act has been committed for which charges can be brought. It is fundamentally at odds with wartime imperatives, which call for the killing and holding of known and suspected enemy soldiers and spies without knowing what specific harms they have committed or would commit.
Relaxing the habeas corpus requirement presents some increased risk of usurpation, but this can be easily outweighed by the need to win a war. This calculation is the same whether liberty is protected directly or indirectly. Either way the Supreme Court must adjudicate between competing Constitutional concerns: the executive war powers on the one hand and the requirements of civilian law on the other. If people are being held without charge for political reasons having no legitimate connection to a war effort, or are being held on untenable military grounds, the Court can step in and limit the power to suspend habeas corpus. Switching to the direct protection of liberty would leave this process untouched.
Another example of a law enforcement mechanism that is of great value to tyrants but is of little value to law enforcement is gun registration. The difference between this example and the habeas corpus example is that neither the direct protection of liberty, nor the presence of wartime imperatives, have much chance of changing this calculation, meaning that gun registration should probably be banned in all circumstances, though provision must always be left for some unforeseen military exigency. (Maybe space aliens will give us a choice between gun registration and annihilation of the planet.)
The first purpose of gun rights is to provide a final check against tyrannical power. If the government oversteps its constitutional authority in a way not recoverable by peaceful means, We The People still might be able to throw off the illegitimate government the way we once threw off King George II. For this reason, a first concern of any tyrant will be to disarm the people, and this job will be much easier if the tyrant has access to a registry of who owns what guns. Thus gun registration carries tremendous potential for tyrannical abuse. On the other side of the scale, legitimate law enforcement rationales for gun registration seem to be close to non-existent, as can be learned from the countries that have instituted it. Canada, for instance, has invested about a billion dollars in a national gun registration program that apparently has yet to contribute to the solution of a single crime.(26)
One would think that a gun registry ought to have some law enforcement value. The futility of the Canadian system is probably due to the fact that it was never intended as a crime fighting tool, but was instituted, as gun registration always is, as a prelude to the disarmament of the law abiding people. The same would be the case in the United States. The only people lobbying for gun registration are the people who want to ban guns. At the same time, gun registration programs are very expensive. Accordingly, their bang per buck as a law enforcement tool is bound to be very low compared to such proven winners as Project Exile, which applies mandatory sentencing laws to felons caught with guns. The most cost effective crime fighting tool of all is gun rights—simply allowing the law abiding people to be prepared to defend themselves—which costs very little public money yet has a tremendous deterrent effect on crime.(27) Even with no concern for tyranny weighing against it, registration is one of the least efficient law enforcement mechanisms. Thus the concern for tyranny is not outweighed and registration should not be allowed.
Should this be called the maintenance of a restriction on law enforcement, or the addition of restriction? For present purposes it doesn't really matter. The plan here is to amend the Constitution as necessary and appropriate to institute direct protection. Since a ban on gun registration is called for, even under direct protection, we would just have to make sure to name such a ban, whether it is already implied or not. Still, it is an interesting question whether a ban on gun registration is already implied. Consider an analogous case.
A requirement that people register their books would certainly violate the First Amendment. Should an administration bent on tyranny arise, it could use such a registry to identify enemies of tyranny. Thus the practice of book registration could deter people from owning books, or from owning politically incorrect books, infringing freedom of association and raising an obstacle to authors that infringes their freedom of speech. Similarly, book registration could be looked at as a violation of either the Fourth Amendment prohibition on unreasonable searches, or the Fifth Amendment right not to incriminate oneself. Since it is better if the government does not have this information, search for it, or the demand that testimony about it be surrendered, cannot even be rational, never mind reasonable.
First or Fourth or Fifth Amendment, the same reasoning would seem to apply to gun registration. People might be deterred from gun ownership by the likelihood that, were an administration bent on tyranny ever to arise, gun owners would become targets. This deterrence of gun ownership would directly violate the Second Amendment, and being contrary to the national interest, could not even be rational, never mind reasonable, under the Fourth and Fifth Amendments.
Gun registration could possibly be dressed up as well-regulation of the militia, even though the only people advocating this interpretation are those who want the militia (the entire adult population) to be very badly regulated (disarmed).(28) Under this guise, an anti-gun Court could easily excuse gun registration, even though it is logically similar to book registration. Thus to maintain the Second Amendment's primary purpose: as a last resort against usurpation of the rule of law, a ban on gun registration should be made explicit in the Constitution, with perhaps a proviso allowing registration when it might be needed to assist in dealing with extraordinary threats.
Just because gun registration seems to have close to zero law enforcement value in normal times doesn't mean it couldn't be necessary for fighting terrorism. If the war on terrorism turns into a closely fought battle we might need every law enforcement tool possible. Terrorist operatives, legally in this country, might well purchase guns, and gun purchase records could provide valuable data points about who, what, when and where every time a gun or a suspect is found, just as hotel receipts and other purchase records can be useful.
If push comes to shove, we should be willing to concede some risk of tyranny in order to protect against outside enemies. It becomes a question of who we fear more, ourselves, or those who are trying to kill us. This, of course, is exactly the issue of gun rights, translated from the individual to the collective. Obviously we must trust ourselves more than our attackers. That is why individuals with no criminal record should be allowed to carry guns and it is why the government should be allowed to keep track of guns, if there is ever a truly powerful law enforcement reason, either temporary (in which case registration should be temporary) or permanent. There would have to first be in place unambiguous protection for the right to keep and bear arms as an individual right that is not ever to be infringed, not by federal or state or local government or by treaty, so that gun registration can never be a prelude to gun banning and confiscation, as it has been in every instance in world history to date. (29)
There should also be a general recognition that, of all the forms of monitoring that might be justified as useful for interdicting terrorism, gun registration is about the last that should ever be undertaken, since there are such powerful reasons not to monitor gun ownership, and the likely benefits are so trivial. Suppose a bunch of terrorists take up where the Virginia sniper left off and start using sniper rifles to commit random murder. What good would it do, when such people are finally caught, to be able to find out whether or not used legally registered weapons? To throw up any roadblock at all to such shootings it would be necessary to ban rifles, but such a ban is easily circumvented by smuggling, either by the terrorists themselves, or by existing black markets.
In sum, monitoring of gun registration is no more or less useful as a crime-fighting tool than monitoring hotel stays and credit card purchases. It facilitates the tracking of past movements that might reveal accomplices or other strategic information. Yet it carries a huge downside that other kinds of monitoring do not, as it is especially valuable to would be domestic tyrants. This downside becomes even larger once we consider that it is not just individual tyrants that we have to protect against, but tyranny of the majority as well.
Part VI: Tyranny of the majority
Tyranny by usurpation is not the only threat to rule of law. Tyranny of the majority poses an equally grave threat. If the majority of voters do not themselves appreciate or feel bound by the principles of right established in the Constitution, and they elect representatives who reflect their own views, then direct protection of liberty may not be very secure.
This might seem like a quixotic worry. Why would The People and/or their representatives vote for a constitutional amendment installing the direct protection of liberty if they didn't appreciate and feel bound to adhere to the principles of right it embodies? Actually, this could happen quite easily, because direct protection offers dramatic gains on multiple fronts. In particular, it offers the prospect of absolutely crushing mala in se. People who have no inclination to tolerate things that Mill's principle clearly protects, such as drug use, might nevertheless favor a move to direct protection just because they think that crushing mala in se is a lot more important than punishing drug use. Thus we could very well end up in a situation similar to that of gun rights, where a large fraction of the population fails to embrace directly protected liberties and actively works to pass laws that infringe them. Together with relaxed restrictions on crime control, this could be a very dangerous situation for liberty. Can it be handled?
At the abstract level, the problem of tyranny of the majority is no different under direct protection than under indirect protection. Where tyrants have the power (such as a tyrannical majority in a democracy) they will do whatever they want to do and none can stop them. They will get their presidents. They will get their supreme court judges. They will get their constitution. A tyrannical majority that is intolerant of things that Mill's principle protects but that is more interested in smashing mala in se could simply eliminate or overrun indirect protections without bothering to protect liberty directly. This has already happened with forfeiture rules for seizing of "ill gotten gains," rules that allow police to confiscate property without due process. It isn't that direct p