RUSSELL HARDIN CIVIL LIBERTIES IN THE ERA OF MASS TERRORISM (Received 30 June 2003; accepted in revised form 22 July 2003) ABSTRACT

This paper discusses the impact of the so-called war on terrorism on civil liberties. The United States government in Madison’s plan was to be distrusted and hemmed in to protect citizens against it. The terrorist attacks of 2001 have seemingly licensed the US government to violate its Madisonian principles. While the current government asks for citizen trust, its actions justify distrust. The courts, which normally are the chief defenders of civil liberties, typically acquiesce in administration policies during emergencies, and it has been during wartimes that the worst infringements of civil liberties have occurred.

KEY WORDS: civil liberties, distrust, habeas corpus, stereotyping, terrorism

The focal concern of the great civil libertarian tradition is how to design a state to protect people against each other while not interfering in their lives beyond what is necessary to maintain social order. In The Limits of State Action, Wilhelm von Humboldt suggests what is our central problem: how government can be structured both to achieve security and to restrain itself from violating individual freedom, which, for him is the unrestricted opportunity to develop one’s own capacities.1 James Madison poses the same conundrum as the central problem of the US constitution: “In framing a government which is to be administered by men over men, the great difficulty lies in this: you must first enable the government to control the governed; and in the next place oblige it to control itself.”2 Both Madison and Humboldt discuss security primarily in the normal sense of domestic security of each person against incursions from others in the same society. Theirs is essentially the ordinary issue of the state’s role in criminal law. Humboldt says the one goal that only a state can accomplish is guaranteeing security (internal and external); individuals cannot do this for themselves.3 He supposes further that those things we withdraw from the competence of the state can flourish on their own through private action. Hence, we can have economic prosperity without state management, and with greater inventiveness and variety than the state could provide.4 For him therefore the essential object of the creation of a state is to address encroachments on the rights of others.5 This is apparently a normative claim, not a historical claim about the rationale for creating actual states. Although there has long been terrorism in varied contexts, contemporary terrorism poses a new and ominous problem: how to structure government to protect us against terrorism while also protecting us against the government itself. This is a surprisingly new issue. We have long considered how to protect citizens against invading armies, and this might be the arena that is most nearly analogous to protection against terrorists. But foreign terrorists are not so transparent as invading armies, and the issue of protecting ourselves against them necessarily involves judgments on how to distinguish them from peaceful citizens and alien residents. When foreign war is the issue, we face decisions on whether to punish Charles Schenck and Elizabeth Baer for openly leafleting against the means to fight the war in 1917.6

For terrorists, who must in the nature of their purpose be secretive, such openness is unlikely, and we may be able to discover them before they act only if we resort to invasive and substantial surveillance, infiltration, screening, and so forth, all of which might be misused against innocents far more often than they are well used against potential terrorists. Here I wish to address what problems such trade-offs pose for our moral and political theories. It is often argued that we face new problems that do not readily fit our moral theories: such issues as genetic intervention, abortion, and nuclear deterrence. Our political theories are especially tested by the rise of international terrorism, in part simply because these theories were devised to deal with domestic issues and institutions.7 Because of their simplistic conception of what an action is, some moral theories cannot handle politics and cannot even handle personal morality in contexts of strategic interaction, which likely include most contexts in which morality

3 von Humboldt, The Limits of State Action, p. 39. 4 von Humboldt, The Limits of State Action, p. 35. 5 von Humboldt, The Limits of State Action, p. 38. 6 Schenck v. United States; Baer v. United States. 249 U.S. 47: 39 S. Ct. 247, 248; 63 L. Ed. 470; 1919 U.S. 7 I address a related issue of how our moral theories are to deal with stochastic processes in Russell Hardin, “Ethics and Stochastic Processes,” Social Philosophy & Policy 7 (1989), pp. 69–80.

might be at issue.8 The best political theories are about such niceties as how democracy should work, not about how we deal with threats of mass murder. There are at least two distinctly different problems of terrorism. The vast majority of terrorist actions, reasonably defined, against U.S. citizens have been virtually one-on-one actions, many of them apparently spontaneous. These actions must commonly be dealt with in the way we deal with murder and other felonies. We apprehend the terrorists after the fact and sentence them for their crimes. The majority of all deaths of U.S. citizens and their employees by terrorism in the two decades before September of 2001 were, however, from well-organized attacks. One of these was by the U.S. citizen Timothy McVeigh in Oklahoma City (OK) in 1995. The other three were by Arabic-Muslim terrorists: the bombings of PanAm flight 103 over Lockerbie in 1988 and of the U.S. embassies in Kenya and Tanzania in 1998.9 Taken together, the perhaps 1200 U.S. deaths from these and from less well organized attacks were dwarfed by the roughly 3000 deaths in the coordinated simultaneous attacks in September of 2001. It is such planned attacks that surveillance might prevent and it is the devices for prevention that raise the most grievous issues for civil liberties.

Henceforth, therefore, I will be concerned only with such terrorism and the policies of its prevention. The central issue here is the misfit of the scales of the terrorism we have experienced and of the restrictions and violations of civil liberties provoked by terrorism. The scale of annual harms from SUVs (sport utility vehicles) in the U.S. is reputedly greater than all terrorist actions together. I think we should do something about the dangers of SUVs and about possible future terrorism, but we should keep the actions in proportion to the dangers.

SAMPLING FOR TERRORISTS

Part of the fundamental problem of policing terrorism, especially before actions that tip off who the terrorists are, is a difficulty of sampling theory that comes up in many other contexts, such as (perhaps aptly) screening for cancer. Suppose that, if our anti-terrorist police agency investigates people according to some set of stereotypical or surveillance criteria, it will be extremely accurate in turning up the terrorists. That is, if the agency investigates you because you meet such criteria and you are a terrorist, 8 Russell Hardin, Morality within the Limits of Reason (Chicago: University of Chicago Press, 1988), pp. 68–70. 9 Rodger Doyle, “The American Terrorist,” Scientific American (June, 2001), p. 28. 80 RUSSELL HARDIN the agency will certainly nail you for more thorough in-depth investigation.

Suppose also that, in screening jargon, our test will have very high specificity of 95%. That is to say, only 5% of non-terrorists will be caught up in our dragnet. In screening jargon, they will be “false positives” (by idealized supposition, there will be no false negatives, which would be failures to include potential terrorists among those selected for further investigation). Let us do the simple arithmetic with its perhaps surprising implications. If only 1 in 1,000,000 U.S. residents are potential terrorists, the high 95% specificity of our criteria for initial arrest and further investigation is still radically too low for practical purposes. For every 1,000,000 people we select to put through in-depth investigation, the one who is a potential terrorist will be found out, but 5% of those selected, or 50,000, will not be terrorists. These 50,000 will initially test positive and we will have to undertake further, more in-depth investigation to determine whether any of them is a terrorist. Even if one in 100,000 residents are terrorists (an implausible number of about 2,000 total),10 our criteria will turn up 5,000 non-terrorists for every terrorist we catch. Moderately zealous agents could make life miserable for tens of thousands of innocent people. Anyone distressed by the rate at which DNA evidence has retroactively freed men who were scheduled to die must react with dread to the prospect of having police agents going after tens of thousands of suspected terrorists. Clearly, we need preliminary criteria that are better than this. Two stand out: membership in a domestic right-wing militia group and Arabic or Islamic backgrounds. Few people would object to using the first criterion because it seems to be politically relevant to possible acts of terrorism. But the second criterion raises dark ghosts in U.S. law from the era, perhaps not yet past, in which driving, walking, shopping, and virtually any other activity while African-American was taken to be evidence of likely association with any crime in the vicinity. Evidently, we are now on the way to stereotyping or profiling young male Arabs much as we have previously profiled young male African-Americans. It seems likely that the U.S. will now use stereotyping to identify those to be investigated. Anyone who was subjected to the seemingly dopey screening at airports for the first year or so after September of 2001 must have wished for some profiling. It would only have made sense to be especially attentive to young adult males, for example, and yet eighty-year-old 10 The U.S. Federal Bureau of Investigation has reputedly floated the estimate of 5000 Al Qaeda operatives in the U.S. [Noam Scheiber, “Is There Really an Army of Terrorists in Our Midst?” New York Times Magazine (16 February 2003), p. 10]. This seems implausible. That number might seem to justify substantial intrusions into civil liberties.

81 women weighing less than a hundred pounds were subjected to intensive searches in at least two of the queues I went through while hulking athletic males were passed without concern. This was the sad policy of Secretary of Transportation Norman Y. Mineta, who still remembers the abuse his family suffered from the racial profiling in the interning of JapaneseAmericans during World War II and who insisted on no racial profiling for airport security.11 “Investigating means following hunches,” says Charles Fried. “The notion of having rules about that is truly insane.”12 Unfortunately, prejudices are hunches. If it happens that most well-organized terrorist attacks are by Arabs or other Middle Easterners, recent U.S. federal case law would seemingly allow focusing investigative efforts on them if the authorities use that statistical association as a rationale for their focus, so that it is not simply prejudice against such people that they are subjected to heightened attention in this particular context (but not therefore in other, unrelated contexts). If the ethnic statistical association for mass terrorism against the U.S. is accurate, then the sampling problem could be substantially reduced, so that the initial pool of people to be investigated in an initial cut would be only the pool of Arabic and Middle Eastern males. If the number of these is 2,000,000, 95% specificity would mean we would do in-depth investigations of 100,000 people. If, as above, there are actually 2,000 potential terrorists, we now investigate 50 people for every terrorist we find out. The major attacks mentioned above have taken about 100 innocent lives for every terrorist directly involved, and the trend in the scale of attacks may well be toward even higher numbers killed.13 Very substantially hassling 50 people to find one terrorist is not likely to sound offensive to many citizens.14 We are entering a new era in which civil liberties will be stereotypical. I and most U.S. citizens who are likely to read this paper enjoy greater protection of our civil liberties than do other U.S. citizens who are easily profiled as Arabic-American or, sadly still, African-American. Suppose we as ordinary citizens or even members of the U.S. Congress want to know how good its policing agency is. The only possible source of systematic data on the quality of the agency’s surveillance is likely to

11 Albert B. Southwick, “Racial Profiling in Wartime Often a Two-Edged Sword,” Worcester Telegram and Gazette (8 September 2002), Section C, p. 3. 12 Larry C. Johnson, “The Declining Terrorist Threat,” New York Times (10 July 2001), op. ed. page. 13 Two months before September of 2001, a former U.S. State Department counterterrorism specialist argued that the threat of terrorism was declining [Larry C. Johnson, “The Declining Terrorist Threat,” New York Times (10 July 2001), op. ed. page]. 14 With the FBI estimate of 5000 Al Qaeda terrorists, only 20 Arab-Americans would have to be hassled for each potential terrorist captured.

be the agency doing the surveillance. One need not be paranoid about government and policing agencies in general to think this is a grim fact.15 Part of the objection to racial profiling is that it silences some of the likely opposition to careless investigation and even sentencing, because those who do not fit the profile are not at much risk of abuse, while those who do fit it are at great risk. Note that the sampling-theory argument here is not widely understood. The high specificity of some cancer screens – for example, for ovarian cancer – seems to most people to justify using them for every potential victim every few years. Doing so, however, would overload the system with false positives and would be very harmful to many of those false positives, whose lives would be disrupted, often severely. The samplingtheory argument is also likely to be too complex for most of the agents of any policing organization, who are perhaps all too likely to think that those caught up in an initial dragnet are, with high probability, guilty. Additionally, note the radical difference between preventing terrorist actions and finding and sentencing actual criminals, such as murderers. The pool of possible suspects for an actual murder is commonly far smaller than the pool of potential terrorists before they attack. Still, the rate of false positives in arresting and even in convicting and sentencing supposed murderers is apparently very high. The discussion above assumed only that the false positives would be put through perhaps exhaustive investigation, not that they would be sentenced to long prison terms or to capital punishment. The evidence of incompetence in prosecuting supposed criminals in the U.S. should give anyone pause about the prospects for decent policing of potential terrorists.16 15 Any reader of Chief U.S. Supreme Court Justice Rehnquist’s book on civil liberties in times of crisis must be especially worried. That was originally a book of legal history, but now it reads as a troubling account of how to deal with current events. Rehnquist holds that presidents are virtually free agents in war time, and one suspects he thinks they should be. He supports former U.S. president Lincoln’s formulation that preserving habeas corpus in his time would have meant allowing “all the laws but one to go unexecuted, and the government itself go to pieces, lest that one [habeas corpus] be violated” [William H. Rehnquist, All the Laws But One: Civil Liberties in Wartime (New York: Knopf, 1998), p. 38]. 16 Even the conservative Economist (15 June 2002), p. 25, which has been very supportive of current U.S. president Bush and his war on terrorism, characterizes the Bush policy as “if in doubt, jail him.”

CIVIL LIBERTIES IN THE ERA OF MASS TERRORISM – TRADE-OFFS

For many people, the trade-off between security and civil liberties is an easy one. We want far greater protection against terrorism and we are willing to have the civil liberties of many people violated in order to get such protection. If you think you are essentially sure not to be a suspected terrorist, you may think the costs of increased state monitoring and controlling of behavior will have little onerous effect on you. Alan Dershowitz argues forcefully for trading some civil liberty for protection against terrorism.17 He supposes even that torture can be justified by a sufficiently dreadful threat. It can be too easy to make such arguments if one focuses on the cases which, in the end, prove to have been genuine. Much of the worry of civil libertarians is that licensing the authorities to act in such cases entails a far more general license, because the authorities are unlikely to be perfect in their judgments of who is or is not a terrorist threat or even of who has or has not committed a horrendous terrorist attack. The uses of torture by the Argentine military junta were probably justified in terms that would fit the defense that Dershowitz and others use to defend its use. Against any such assessment, the dreadful fates of the Argentine “disappeareds” must upset the easy calculus of the value of torture in narrowly defined circumstances. Even the recent flurry of discoveries, after the advent of DNA testing, that men on death row were innocent of the violent crimes for which they had been convicted to die is evidence of how much we – or at least some of us – should fear actions by authorities who are licensed to take actions against citizens who are suspected of committing crimes. The big trade-offs in this relationship have long been supposed to be the potential conflict between internal order and individual freedom and that between welfare and freedom. Humboldt poses the latter issue especially sharply. He asks whether the state should extend its activities to affect the positive welfare of the nation, or restrict itself to providing security.18 His answer is a combination of quasi logical and somewhat facile causal claims. Quasi logically, he supposes that all welfare programs must have harmful consequences that impair freedom.19 Empirically or causally, he argues that such programs tend to produce uniformity, weaken

17 Alan M. Dershowitz, Why Terrorism Works: Understanding the Threat, Responding to the Challenge (New Haven: Yale University Press, 2002). 18 von Humboldt, The Limits of State Action, p. 17. 19 von Humboldt, The Limits of State Action, Chapter 3. 84 RUSSELL HARDIN

vitality20 by displacing the active energy of individuals,21 fail to fit individual differences, and hinder individuality.22 We could load such losses on the cost side of government action and then we have to weigh the tradeoffs with lost freedom from lack of opportunity (not his language).23 His most general claim is quasi logical: such programs neglect creative powers and just concentrate on results, such as achieving egalitarian distribution.24 Humboldt goes on to argue against the use of the state to provide welfare benefits,25 and he insists that the state can address welfare only to the extent necessary for security.26 For example, the prevention of crime must be directed at its causes.27 This inherently implies some welfarism, but very limited.28 Immediately, one might ask whether the state should address the mere possibility of transgression, or only the actual instance of it? Humboldt argues that potentially the state must restrict actions harmless in themselves when the probability of harm is high enough.29 This is a variant of John Stuart Mill’s harm principle,30 with a stochastic twist. There are two very important omissions of Humboldt’s discussion. The first is any attempt to assess how government can be structured to both give it power to provide security (internal and external) and restrain it from violating individual freedom. In general, devices for internal security are probably those that put us at greatest risk. In some degree, this claim seems confounded by the current U.S. concern with homeland security, with the external threat to security as justification for internal actions. The second

20 von Humboldt, The Limits of State Action, p. 18. 21 von Humboldt, The Limits of State Action, p. 20. 22 von Humboldt, The Limits of State Action, p. 27. 23 von Humboldt, The Limits of State Action, pp. 28–29. 24 von Humboldt, The Limits of State Action, p. 31. 25 von Humboldt, The Limits of State Action, Chapter 3. 26 von Humboldt, The Limits of State Action, p. 33. 27 von Humboldt, The Limits of State Action, p. 117. 28 von Humboldt, The Limits of State Action, p. 118. 29 von Humboldt, The Limits of State Action, p. 91. Humboldt wants as little law as possible on the side of prevention because he wants to protect the virtue of free will (as in Kant) or the personal development of character (von Humboldt, The Limits of State Action, pp. 120–121). 30 “That principle is that the sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their number is selfprotection. That the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others” [John Stuart Mill, On Liberty (Indianapolis: Hackett Publishing Company, 1978), Chapter 1, p. 9].

omission is lack of definition of or derivation of the specific rights that the state is supposed to protect. His discussions of rights are far too facile – as is typical of rights theorists since he wrote as well as of those, such as John Locke, before his time. I have begun with Humboldt first in order to accentuate an ambivalence in Mill’s liberalism on these issues. Mill generally seems to think of the freedom of personal choice as a matter of one’s welfare. He argues that, say, I as an individual generally know better where my interests lie than others can know for me or than the state can know for me. The interest that society has in any individual “(except as to his conduct to others) is fractional and altogether indirect, while with respect to his own feelings and circumstances the most ordinary man or woman has means of knowledge immeasurably surpassing those that can be possessed by anyone else.”31 Here, the concern seems clearly to be that I should be free to act in my own interests because I know my interests better than others can (we can make allowances for my going to a professional to tend to some of my interests when I know that I likely have interests that I cannot fully causally understand, so long as it is I who choose to go to the professional). But Mill also says that, if “a person possesses any tolerable amount of common sense and experience, his own mode of laying out his existence is the best, not because it is the best in itself, but because it is his own mode.”32 Humboldt seems to hold the second of these views, but he does not explicitly reject the first, welfarist view of the value of individual freedom. I will treat civil liberties as grounded in welfare. There is at least one very compelling reason for this. Civil liberties are seen as protections of us; their realization should make us better off. It would be odd to defend them if they made us demonstrably worse off, as surely they could under some circumstances.

Note, however, that this argument could involve a fallacy of composition. Among people who are not and would not be terrorists, laws that suppress terrorism might affect some severely and others likely not at all. There is likely no such thing as a law that will penalize or restrict the actions only of terrorists. For example, Arab-Americans today are probably more at risk of surveillance, harassment, and mistaken arrest than are other U.S. citizens. There is also an analytical advantage in treating all rights as welfarist, an advantage that is not itself a reason for seeing them as welfarist. Tradeoffs between rights are often necessary (for example, privacy trades off with free speech whenever your privacy can block my speech or vice (31 Mill, On Liberty, p. 74. 32 Mill, On Liberty, p. 64. 86 RUSSELL HARDIN versa), but no serious rights theorist has a principled way of making such trade-offs. Courts tinker with rights at their edges, hemming them in special circumstances. Perhaps the most frequently cited instance of such trade-offs is the Supreme Court’s conclusion that the right to free speech does not entail the right to yell fire in a theater when there is no fire or to speak in a way intended to incite to riot. This analogy was used by Oliver Wendell Holmes in a remarkably awful opinion criminalizing the action of circulating leaflets against a wartime draft in 1917.33 As Dershowitz says, this may be “the only jurisprudential analogy that has assumed the status of a folk analogy,” although yelling fire in such a context is not at all analogous to the kind of speech in which Schenck was engaged.34 A very general version of the conclusion that rights trade-off against each other in certain circumstances is already enunciated in general by William Blackstone.35 Lack of a principle for trading off rights in conflict is fundamentally problematic for the trade-off between protecting us against terrorism and defending our civil liberties, because any absolutist claims for civil liberties have to be wrong in principle.

There can be at most one absolutist principle that never yields to any other principle when they conflict, and civil liberties cannot possibly be that one because they are a collection of principles that can conflict. We need a defense of civil liberties that allows trade-offs between them and between any of them and security. If all of these are grounded in welfare, such trade-offs are, in principle, fairly simple, although in fact welfare is an extremely difficult concept, especially in comparisons across people. One might suppose that security is itself a right, and in domestic contexts of criminal law, legal protections of contracts, and so forth, it is a positive right in the law. Thomas Hobbes, Blackstone, and many other early theorists in the development of the modern vision of civil liberties counted protection against intrusions by others into one’s life or in taking one’s property as rights that should be part of the law or the state’s prerogative. It makes no sense to say, in an analogous way, that security against foreign terrorists is a right. I have a right under the laws of our government

33 Schenck v. United States; Baer v. United States. 34 Dershowitz, Why Terrorism Works, p. 142. 35 William Blackstone speaks of absolute rights but then relativizes them with the claim that they can be limited by law. For example, he says, “The laws of England have never abridged [the right of personal liberty] without sufficient cause,” His category is therefore vacuous [William Blackstone, Commentaries on the Laws of England, 4 volumes (Chicago: University of Chicago Press [1765–1769] 2002), Book 1, Chapter 1, p. 29, emphasis added].

to protection against you, my fellow citizen. Terrorists from abroad should be deterred to some extent, but this is a welfarist claim, not a claim of rights under law. A sort of grain of truth – solely pragmatic – in the absolutist position on civil liberties is that we should probably give greater force to civil liberties than we might suppose just because it is easy to weaken them and very hard to strengthen them; and the moments when we think to weaken them are almost always going to be moments of crisis when the longer-run interest in civil liberties gets over-shadowed by the urgency of the moment. John Ashcroft was a civil libertarian on internet issues while he was senator from Missouri but, after he became attorney general under current president Bush and was in charge of enforcement of restrictions, and of choosing against whom to enforce them, he became anti-civil libertarian on those issues.36 That seems likely to be the way these trade-offs will go. But this defense of elevating civil liberties to a special status is grounded in empirical rather than theoretical concerns. It does not mean that civil liberties are absolute or that they take priority over other concerns but only that their defense is relatively fragile so that their maintenance may require heightened scrutiny and even vigilance. As Humboldt notes for domestic matters of crime, the security of citizens requires that all pursuit of redress be transferred to the state.37 The alternative would be a murderous system of vengeance. Protection against foreign terrorists is not strictly analogous. We can fill out Humboldt’s theory by noting that only the state can possibly protect us against terrorism, and therefore such protection is required of the state. We might therefore add attempted control of terrorism to control of crime and defense against foreign invasion as now the third matter with respect to which the individual is virtually helpless and must call on the state. Insofar as the terrorism or the potential terrorists are acting within the control of our state, this makes good sense as a simple analogy with legal action against crime and with national defense. Because of the international character of terrorism, however, it may be hard to show that its policing should be delegated exclusively to the state. Suppose we eventually do have international law on these matters or some form of reciprocal law between particular states. With such a regime in

36 As senator, Ashcroft wrote: “The government’s police-state policy on encryption is creating hindrances and hurdles that will eventually injure our ability to compete internationally” [John Ashcroft, “Keep Big Brother’s Hands Off the Internet,” USIA Electronic Journal 2 (October, 1997), final paragraph. See David Corn, “The Fundamental John Ashcroft,” Mother Jones (March/April, 2002)]. 37 von Humboldt, The Limits of State Action, p. 106. 88 RUSSELL HARDIN

place we might be able to say that action taken by individuals or groups who are not licensed by states would be too harmful to allow such action, which might be criminalized in international law. But if there is no positive law or even any strong system for intervening against terrorists wherever they might be, we cannot on Humboldt’s argument conclude that action by essentially free agents would be blocked. Because such action would be hopelessly inadequate, we would presumably want states to take action. If the terrorism comes from foreign bases, however, the role of the state in fighting it is not analogous to its role in crime and national defense. The state then is essentially the analog of a far more powerful individual taking action on its own as an unconstrained free agent. The response of the Bush administration to international terrorism has taken on the character of the action of an unconstrained free agent, a lawless cowboy, and that response has therefore had deleterious effects on international efforts to build a system of extra-state control of some problems. Advocates of international law could hope that the anarchic quality of interstate relations had been reduced in recent decades. That hope has been demoralized by the rancor over how to deal with terrorism and by the go-it-alone actions of the U.S.

MADISONIAN PHILOSOPHY OF GOVERNMENT

The U.S. with its Madisonian philosophy of government initially came close to Humboldt’s vision of how government should be used. Madison distrusted government. This stance of liberal distrust is a defining element of U.S. constitutional principles.38 In the twentieth century, the U.S. government expanded its role to provide welfare benefits of many kinds, although in this it lagged behind Germany and many other democracies. For Humboldt that was already a violation of the purpose of government. For civil libertarians, it opened new arenas for violation of individual prerogatives, although most often the net result may have been expansion rather than contraction of individual liberties. We may now therefore be in the first long-run era of a general contraction of civil liberties since Madison. Let us think ourselves back to the era in which Montesquieu and Madison wrote.39 Montesquieu argued that the best form of government would be republican and must therefore be in a small community, in 38 Russell Hardin, “Liberal Distrust,” European Review 10 (2002), pp. 73–89. 39 See further, Russell Hardin, Liberalism, Constitutionalism, and Democracy (Oxford: Oxford University Press, 1999), Chapter 3.

which he supposed that the confluence of interests would enable consensual government (That was an astonishing assumption for one who knew the history of Greek and Italian city states). Madison actually lived in such a community, if one counts only the land-owning males, who were the only citizens who did count politically. He thought Montesquieu was fundamentally wrong. He supposed that the worst threat to liberty in his world was from the governments of the thirteen states. He supposed that a national government would give the best protection against the tyrannies of local majorities.40 He therefore wanted a national government that would override the state governments, even perhaps abolish them. He wanted it to be an otherwise relatively weak government. For Montesquieu the point of separation of powers between the legislative, executive, and judicial branches is to reduce the odds of the arbitrariness that might follow from having any one of the branches fully in control.41 The argument is essentially the claim that requiring multiple branches to act to deprive one of liberty adds to one’s security through regression towards the mean. It is far less likely that two or three branches will coordinate in an arbitrary suppression of someone’s liberties than that a unified single branch will be arbitrary. It is too easy to suppose that Montesquieu’s and Madison’s42 concern with tyranny – the arbitrary use of power – is a problem of old-fashioned monarchy. But if minor executives are not controlled by courts and the legislature, it is potentially an even more grievous problem of big government. The number of officials involved in ferreting out terrorists in the U.S. today is likely in the thousands, and all of them might act tyrannously in their own small venues, all with the good purpose of getting their jobs done thoroughly. A tyrannous monarch can hardly avoid exposure; thousands of officials operating in closed venues can hardly be exposed. In some degree, therefore, Madison’s argument that liberty is safer in a larger society is violated. Bush’s response also violates Montesquieu’s separation of powers. Bush gives executive orders that have the force of law (he legislates); he acts under those orders (he administers); and he judges. The constraint imposed by the separation of powers on abuse of power is likely to be severely weakened in a time of great crisis, as during all

40 Alexander Hamilton, John Jay, and James Madison, The Federalist, No. 10 (The Gideon Edition), eds. George W. Carey and James McClellan (Indianapolis: Liberty Fund, 2001), pp. 42–49. 41 Charles de Secondat, baron de Montesquieu, The Spirit of the Laws (Cambridge: Cambridge University Press, 1989), Book 11, Chapter 6, p. 157. 42 Alexander Hamilton, John Jay, and James Madison, The Federalist, No. 47 (The Gideon Edition), eds. George W. Carey and James McClellan (Indianapolis: Liberty Fund, 2001), pp. 249–255. 90 RUSSELL HARDIN

of the major wars fought by the U.S. During the U.S. Civil War, president Abraham Lincoln suspended the right of habeas corpus. During World War I, Chief U.S. Supreme Court Justice Holmes issued his awful opinion in Schenck v. United States. In World War II, tens of thousands of JapaneseAmerican citizens were interned in the western deserts. The decision in which Holmes invoked his bad analogy with shouting fire in a theater was provoked by what he thought was a grim crisis for U.S. security – sufficiently grim that it justified jailing someone for distributing a leaflet that largely cobbled together passages from the U.S. Constitution and other liberal documents to support criticism of conscription for a “great” war. If the judiciary can bend itself to so great a violation of civil liberties, we cannot be optimistic that civil liberties will win the political support they require if they are to prevail over the momentary passions of majorities faced with threats of terrorism. The U.S. courts have long shown themselves to be inclined to let Congress and the executive branch override civil liberties in times of great crisis. The immediate U.S. response to exporting international terrorism to the U.S. has been to go into deep crisis mode. The majoritarian response to crisis seems to be to countenance the use of state power to handle the crisis. This makes eminently good sense when, on Humboldt’s principle, it is only the state that can handle the crisis (Quite possibly not even the national state can handle terrorism, but presumably nothing short of the national state can do so). But giving the state anti-terrorist powers runs counter to Madisonian principles of distrusting the state and attempting to build in safeguards against abuse of power and against Montesquieu’s principle of separation of powers. Normally we face the problem that giving the state power to do anything gives it the power to do many other things, even things contrary to the purpose of the grant of power. For example, giving it the power to protect civil liberties probably gives it power that might be used to violate civil liberties. In the face of terrorism, the power we must give the state is virtually directly the power to violate civil liberties, at least selectively. Virtually every inventive move to control terrorists inside the U.S. will be a move to enable easier violations of civil liberties of ordinary citizens who are not ever going to be terrorists. The object of criminal law, especially in utilitarian theory, is to deter and prevent crime (through interdiction, surveillance, and so forth), not to pre-arrest those who are thought likely to commit a crime. The object of national defense against foreign powers is similarly to deter them or prevent them from attacking and, failing that, to defend against them. The apparent object of much, maybe even most, anti-terrorist policy is to criminalize actions leading to terrorist action and  to apprehend the potential terrorists before they act. Some terrorist acts are perhaps deterred for reasons of fear of retaliation, although that is an irrelevant fear for suicidal terrorists, such as those of September of 2001.

Preventive arrest bears little relation to past notions of prevention of crime, which is done by deterrence or by changing incentives (as through the welfare policies that Humboldt commends). Ashcroft says that, to deal with terrorism, “the Department of Justice has added a new paradigm to that of prosecution – a paradigm of prevention.”43 Unfortunately, what we cannot readily do is restrict the state’s use of this anti-civil-libertarian power to be sufficiently selective in the use of its new powers, to restrict it to attacks on likely terrorists. The sampling theory argument above must provoke doubts that the government will be very civil libertarian as it attacks terrorism at home (and perhaps far less so still when it attacks terrorism in its foreign home bases). Far more urgently than has ever been true in any other context, we now face Madison’s problem of obliging the government to control itself in the ways that worried him. In large part, the state must now perhaps control itself by simply not using its powers, rather than by having institutional safeguards designed to block its potential transgressions. And it may even have to do that in the face of majoritarian sentiments against restricting it. Humboldt comments that state welfare programs face the prospect that the state’s solicitude can prevail forever with deleterious effects.44 The state’s inventiveness in the surveillance of possible terrorists is even more likely to entail a permanent change in the nature and quality of government. We are unlikely soon to repeal any license we grant now and we are not likely ever to reverse new inventions for invading civil liberties. Perhaps that is a reasonable or necessary cost of dealing with terrorism, but it seems short-sighted to reach that conclusion in a time of great crisis and without much debate. We are reaching that conclusion not through reason but through action, action that is likely to be irreversible.

MORAL THEORIES AND TERRORISM

International terrorism poses seemingly new issues that undercut traditional understandings of the trade-off in principle between government power to act and constraints on government action, of the vision of the separation of powers into institutions that can counter each other, and of

43 Adam Liptak, “Under Ashcroft, Judicial Power Flows Back to Washington,” New York Times (16 February 2003), Section 4, p. 5. 44 von Humboldt, The Limits of State Action, p. 34.

the level at which action should be taken (individual, state, or suprastate). It also heightens our attention to the old debate over moralities that are primarily about the goodness of outcomes or states of affairs and those that are primarily about rights, rules for behavior, or personal character. The first concern with terrorism is likely to be the trade-off between controlling terrorism and enabling government to intrude massively into citizens’ lives. Most moral theorists are averse to making such trade-offs. One might suppose it is largely for that reason that most moral theories are not easily expanded into political theories. If we are to address terrorism normatively, we must do so in chief part at an institutional, political level. A “principled way” to deal with trade-offs between two conflicting rights, as in Blackstone, or between civil liberties and protection against terrorism, as in our new era, is not likely to be found. Acknowledgement that these pairs do trade off and that our preference is that the trade-off be allowed to tip either way depending on the conditions may be the most we can expect. This position, however, may be congenial only to welfarists, who are accustomed to think that everything potentially trades off with everything else. Other moral theories often are posed as hostile to any such trade-offs, as though, for example, there were inherently no issue of conflicts among a set of supposedly absolute principles. Without substantial revision of their ground assumptions, such theories are irrelevant for the judgment of devices for dealing with terrorism in its new international, world-wide, high tech form in which the costs of delivering pain can be dwarfed by the scale of the pain delivered. As was remarked of an Irish Republican Army bombing in London’s financial district, a thousand pounds of semtek did a billion pounds (sterling) of damage.

On September 2001, 19 men took the lives of about 3,000 people. Unfortunately, the grounding of many moral arguments makes them seem ill-suited to address these issues. For example, Kantian arguments arguably cannot be decisive about such things, because those arguments are inferred from what would be required of us in a kingdom of ends. If there are weird terrorist visions out there, then we are very far from the kingdom of ends, and we need principles for our world. Virtue theory in our time is often posed as a strictly individual matter, without Aristotle’s sometime functional concern with how the individual’s virtues serve the larger society through, for example, political leadership. Intuitionism, which is perhaps the most common position of ordinary people, is of no value in the analysis of issues about which there is no good reason to suppose we have natural, untutored intuitions.45 More generally, theories

45 Hardin, Morality Within the Limits of Reason, pp. 178–191.

that are anti-consequentialist can hardly make sense of the issues of our time. Personal morality is not irrelevant to our current concern with terrorism. In addition to protecting individuals from intrusions by the state, civil libertarians must be concerned with intrusions by the larger society. The reach of legally protected civil liberties probably expanded from the time of Holmes’s decision in Schenck v. United States until recently. Similarly, the protection of individuals against oppressive social intrusions has likely also grown over that period as tolerance for difference has risen. The legal and social movements have played into each other, with the courts becoming more liberal on, for example, issues of race in response to slowly changing mores and the society becoming less racist in response to court and political decisions to enhance racial equality under the law. With a level of stupidity that caught even his detractors by surprise, Jerry Falwell instantly attributed the attacks on September of 2001 to his god’s anger at the U.S. for its social tolerance and even for its civil liberties, which he supposed were contrary to the views of his bigoted god.46 We will not soon know whether the progress in civil liberties of the previous decades has been halted by the September 11th attacks, but the fact that those attacks were carried out in the name of a religion that is strange to most U.S. citizens might augur ill for further strides in tolerance as well as for civil liberties in the near term.

CONCLUDING REMARKS

It may be too early for good academic writing on terrorism and civil liberties, because we are in a heightened state of inventiveness about how to deal with terrorism and we are unlikely to anticipate all of the issues about how the inventions will work. We know that the technology of surveillance has become radically more sophisticated in recent decades and far more easily put into use, and we can probably expect its use to be expanded. But perhaps there will also be institutional design to inhibit abuses of surveillance and oppression of various individuals and groups. This suggests a game of escalation, but the escalation is likely to be highly skewed in favor of inventiveness in the means of surveillance and control of citizens rather than in the ways of controlling government use of these means. If there will be great inventiveness in devising ways to control the government in its new “paradigm of prevention,” then we cannot yet know what are the trade-offs between what we want the government to do and

46 New York Times, 14 September 2001, p. A18. 94

what its capacities are. Again, it seems likely that the rate of invention of new techniques for surveillance will far outrun the rate for new techniques of controlling the use of surveillance. The traditional concern with trade-offs at the institutional level was about making institutions powerful enough to do their jobs while making them too weak to go beyond their assigned tasks. That concern seems irrelevant for the kinds of agencies that must contest with terrorism. The U.S. National Security Agency (NSA) has massive powers of surveillance. Limiting those powers in order to protect ordinary citizens from intrusions by the state necessarily limits them in the task of ferreting out terrorists. If there are to be controls on such agencies as the NSA, those controls will have to come from outside the agencies, as in the traditional defense of the separation of powers, and not through internal devices that weaken the agencies. In the U.S. system, the most important external control in principle is surely the courts. One might think that the office of the attorney general would also be an external control, as it was used in the past to control state and local governments in the era of civil rights. But its role as the nation’s chief prosecutor makes it an implausible defender against the federal government of those who might, rightly or wrongly, be charged with crimes of terrorism; and indeed many of the attorneys general of recent times have been personally implausible defenders of individual liberties at any level. Some of them have corrupted the office by using it for partisan political advantage. Yet it is difficult to imagine an office, such as the Scandinavian ombudsman, that could have the power to stand against the NSA, FBI, CIA, and the attorney general, especially if the courts are deferential to the administration, as they generally are, especially in times of crisis. Finally, note that the current reaction of the U.S. government to terrorism reinforces the appeal of Madison’s liberal distrust. Many scholars have recently argued that U.S. citizens trust their government too little, and that trust in government is in decline.47 They wish to “restore” trust in government – although it is not at all clear that current levels of deference to and confidence in government are much lower than they were throughout most of U.S. history. Now, however, there is likely to be increasing ground to doubt the quality of government judgment in combating terrorism without grievously undercutting civil liberties.48 The 47 See contributions to Susan J. Pharr and Robert D. Putnam (eds.), Disaffected Democracies: What’s Troubling the Trilateral Democracies (Princeton: Princeton University Press, 2000). 48 The trade-off and incursions against civil liberties were already worrisome before September of 2001. Laura Donahue says: “Special courts, secret evidence, classified  Madisonian era and since have often been times of finding institutions to free the economy from state restrictions of various kinds, with the result that the economies of various nations were transformed and produced the wealthy world in which we live. It would be depressingly ironic if that technological wealth should now be put to use to reduce civil liberties just because that technological wealth enables terrorism at a new, mass scale.

New York University 726 Broadway, Room 712 New York, NY 10003-9580 USA E-mail: Russell.hardin@nyu.edu deportation proceedings, and special rules of evidence speak not of an open, liberal, democratic society, but of one cloaked in secrecy. Proposals to erode posse comitatus, alter assassination policies, use criminal informants, introduce military courts, indefinite detention, and identity cards, and widen powers of information gathering proliferate” [Laura K. Donohue, “In the Name of National Security: U.S. Counter-terrorist Measures, 1960–2000,” Terrorism and Political Violence 13 (2001), p. 48].