James Roper, Michigan State University
I begin by sketching a situation I call “the reluctant gambler problem.” I maintain that my intuitions, and those of many apparently rational individuals with whom I have discussed this matter, appear to conflict with well-known normative principles of rational choice. My analysis of this puzzling situation suggests that those who take such an apparently irrational position are not necessarily forsaking their reason. By making a distinction between what I term “act probabilism” and “rule probabilism,” and adopting the latter, I argue that someone whose initial response to the reluctant gambler conundrum appears irrational can potentially justify his or her position. Standard normative theories of rational choice do not seem to accommodate such “rule probabilism.” I argue that such theories are, prima facie, not adequate to codify our intuitions about rationality.
Applying the lesson I draw from the reluctant gambler problem to an analysis of the most reasonable way for citizens of the United States to confront terrorism in a post-September 11 world, I refer to a recent poll by Professor Richard T. Curtin, Director of Surveys of Consumers at the University of Michigan, which concludes, on the basis of consumer preferences and statistical analysis, that U.S. citizens exaggerate their personal risks from terrorist attacks.I maintain that the distinction between act and rule probabilism can shed light on the exaggeration of our personal risks from terrorism. Let me be clear, I am not suggesting any sort of full analogy between the reluctant gambler issue and the threats of terrorism; rather, I believe we can use a distinction I make in my analysis of the former to help shed light on the latter.
Finally, I examine some of the political ramifications of adopting a stance regarding terrorism suggested by the distinction between act and rule probabilism. I maintain that making the distinction that Alan Dershowitz and others make between “retail” and “wholesale”—or even “apocalyptic”—versions of terrorism does not measurably change my conclusion that Americans are exaggerating their personal risks from terrorism.In this connection, I explore other, more palpable, risks we face—especially political risks associated with the loss of basic freedoms. I also suggest several circumstances that might help to explain our exaggeration of terrorist risks. The terrorism crisis is a serious issue. So too, however, is the endangerment of U.S. citizens’ fundamental civil liberties by administrative policies designed to combat terrorism. The political choices we make will determine the future of freedom in the U.S. I conclude with several recommendations, and explain why my analysis may warrant these responses.
The Reluctant Gambler Problem
Some philosophers have said a necessary condition for a person to be called rational is that he or she should “take probability as a guide to life,” and this seems intuitively plausible.In trying to clarify what it means to guide one’s life by probability, however, we encounter the following puzzle.
Imagine Ms. Jones sitting in her living room on a hot summer Sunday reading the paper. Slowly the image of a cold glass of ginger ale forms in her mind. Checking the refrigerator reveals she is out of ginger ale. Fortunately, there is a store about four miles away, and Ms. Jones decides to drive there to buy some ginger ale.
As she walks to her car, she encounters a man she recognizes as Mr. Chance. Ms. Jones knows Mr. Chance to be completely honest, but rather peculiar in other respects, as will soon be obvious. This individual reminds Ms. Jones that her computer contains a program for playing roulette. The roulette wheel that appears on her monitor can be set to have, literally, any finite number of “slots,” and the probability of the “ball” landing in any particular slot is always 1/the number of slots this computer version of a roulette wheel has been set to register.For example, if there are 10,000 slots, then the probability of the ball landing in some designated slot is 1/10,000.
Mr. Chance shows Ms. Jones very reliable evidence regarding the statistical probability that a person belonging to a narrowly specified class of drivers—which, Ms. Jones recognizes, includes her—will be killed while driving on any of a narrowly specified class of trips which, she recognizes, includes her projected trip to the store. On the basis of the evidence Mr. Chance presents to Ms. Jones and her knowledge of his intellectual integrity and the reliability of his data and analytical methods, she has absolute confidence in the reliability of his statistical probability statement—which takes into account both the fact that she has no aversion to gambling and that she attaches no positive or negative value, or utility, to the act of driving to the store.Mr. Chance’s statement also takes into account any special circumstances relevant to his statistical estimate—for example, the general condition of Ms. Jones’s car, whether she has had a good night’s sleep, is intoxicated, etc. Thus Ms. Jones finds herself presented with a statistical probability estimate that a person like her in all relevant respects, embarking on a trip such as her planned trip to the store has, for example, one chance in a million of being killed if she drives to the store for ginger ale. I assume, further, that Ms. Jones understands that Mr. Chance’s statistical analysis is accurate and that she cannot measurably alter her probability of dying should sh echoose to drive to the store herself. Mr.Chance does not need to specify these odds exactly; he need only be able to present Ms. Jones with a clear estimate. We also ignore the risk of injury, though this would not be possible in the real world.
Now Mr. Chance suggests the following wager. He and Ms. Jones, and suitable observers, are to go back into Ms. Jones’s home and set her computer roulette program for the statistical probability Mr. Chance has calculated—in our example, one in a million—she has of dying on her projected trip. If Mr. Chance has presented Ms. Jones with an estimated range, the odds would be adjusted to accommodate that fact. Next, Ms. Jones is to select a number between, in our example, one and one million and communicate it to the observers. Finally, her computer’s roulette wheel is “spun.” If her number comes up, Ms.Jones will be painlessly killed.If it does not, Mr. Chance will bring the cold ginger ale into her house, sell it to Ms. Jones at the store’s price, plus payment for the wear and tear that will not be put on her car by driving it to the store and back, and leave. Mr. Chance has assured Ms. Jones, and she knows she can believe him, that he will never return. In short, this bet is to be a one-time thing—not the first of a series of visits Ms. Jones will experience over her lifetime.
I have presented this wager to hundreds of colleagues, friends, and students over a period of many years. While a few would accept the gamble, the overwhelming majority would not. Indeed, whenever I imagine myself in Ms. Jones’s position, I find myself refusing Mr. Chance’s bet.The whole point of this wager, of course, is to pose a problem about our very concept of rational choice—“taking probability as a guide . . ..” The example is so structured that, according to Mr. Chance’s story, a rational person should be indifferent between taking the bet and driving to the store. But here is the problem: Typically, those who refuse the wager are not indifferent. Their rejection is not a matter of a coin flip; it is a serious rejection. They do not want to accept the bet. Period. But even if my “anecdotal” accounts of presenting this wager to students, colleagues, and others are not taken as credible, the key point is that I find myself wanting to reject the wager—and Iam familiar with normative theories of rational choice and probability theory, and I do not think I can change my risk of dying by driving myself.
Are all who would refuse such an offer just acting irrationally—refusing to take probability as a guide to life? This certainly seems possible. After all, it appears that both the probability and the payoff—hence, the expected value—would be the same either way we choose.The cost of the ginger ale and the expense of driving the car are the same for both choices, and, by assumption, there appears to be nothing else that would allow us to value taking the bet over rejecting it, or vice versa. If we are rational, it appears we should be indifferent between Mr. Chance’s options. Unfortunately, however, this approach would exclude from the class of rational persons many who most people believe should be included. Indeed, I would count myself among those so excluded.
We might reject the idea that taking probability as a guide to life is, on any plausible interpretation, a necessary condition for rationality, but on this view we would be hard put to say what ‘rationality’ means.The following analysis neither labels those who reject Mr. Chance’s bet
irrational nor rejects the idea that guiding one’s life by probability is necessary for rationality.
Analysis of the Reluctant Gambler Problem
When those who reject the wager are first presented with it, they usually say such things as: “If I stopped doing things like driving to the store for ginger ale, life would become so dull that it would not be worth living, whereas I can have a very full and satisfying life without doing such things as accepting Mr. Chance’s wager.” These remarks are suggestive, but they need clarification and analysis. Remember, we granted Mr. Chance’s claim about the statistical probability of being killed while driving to the store. In addition, the wager assumes that those approached have no aversion to gambling and participate in a variety of activities known to have statistically specifiable risks of dying associated with them. Therefore, someone who says there is something different about accepting Mr. Chance’s offer must, first, distinguish acts such as accepting his wager from other acts and, second, explain how we are justified in firmly refusing Mr. Chance’s bet and yet getting into the car and going after the ginger ale.
To begin answering these questions, I suggest the reluctant gambler problem reflects a pre- analytic intuition about rational action—relating especially to the way in which we use probability to guide our lives. To be more specific, I believe we can make a distinction between applying probability assessments to individual actions and applying them to rules. In effect, I am making a distinction between what I will call “act probabilism” and “rule probabilism.” This intuition is not typically reflected in standard theories of rational choice. In effect, I am using Mr. Chance’s thought experiment to challenge such standard normative theories of choice.
To help clarify the distinction between act and rule probabilism, I note a suggestive parallel.Most philosophers are familiar with the way in which the ethical theory of traditional, or “act,” utilitarianism has been distinguished from modified versions of that theory that are typically labeled “rule” utilitarianism, but a brief recounting of this distinction will elucidate our analysis of the reluctant gambler problem. Traditional, or act, utilitarianism is the theory that says one ought to do that action, among the possible alternatives, that maximizes the happiness, or “utility,” of the greatest number of people. There are variations and refinements, but this is the main idea. Actions are justified directly by being subsumed under the principle of utility. Rule utilitarianism, on the other hand, involves a two-step process for justifying actions. First, one defends a set of rules by showing that following these rules—or, rather, this set of rules—will lead to maximizing the happiness or utility of society.Justasanindividualcanbeanactutilitarian,anindividualcanbearuleutilitarian.This is controversial because some philosophers believe that the appeal to rules entails that a rule utilitarian must be part of a group or community.The point here, however, is not to argue for rule utilitarianism as an ethical theory; it is rather to draw a loose parallel between rule utilitarianism and certain aspects of the theory of rational choice regarded as a theory of individual action. In this venue, the issue is less problematic.
One further point. Though the matter is often ignored, as I have ignored it here for simplicity, it is clear that even rule utilitarians would not want to preclude ever dealing with special situations as if they were functioning as act utilitarians. Very important decisions, whether they primarily affect the decision makers as individuals or have profound implications for the larger society, might be approached, even by a rule utilitarian, as individual choices—which warrant careful and detailed calculations. Similarly, rule probabilists may, and likely will, want to consider some decisions in detail, rather than simply subsuming them under a set of rules. But, in making this point, I am not abandoning the distinction between “act” and “rule” probabilism. On the contrary, these exceptions actually strengthen my position. As soon as we contemplate what it means to do a sustained calculation of the probabilities associated with a particular action, it is clear that no human could consistently deal with every action in the manner of an act probabilist.Moreover, as I explain below, the proper response to terrorism may be to revise some of our rules. This is much different from treating cases individually, in the manner of an act probabilist. Though it may appear to some that such a program of rule revision involves treating cases individually, this is simply not true. There are still rules; they are just altered to deal with a new reality. The act probabilist I envision would not utilize rules. Every action would be individually considered.
The distinction developed above sheds light on our analysis. Is wagering with Mr. Chance something one needs to do as part of a normal life? To clarify how it might make sense for someone who takes probability as a guide to life to reject such a bet, I propose that such an individual be regarded as typically applying probability estimates to rules, rather than directly to actions. In short, such an individual would be a sort of “rule probabilist,” as opposed to an “act probabilist.” In effect, I suggest that a parallel between the application of the principle of taking probability as a guide to one’s life and rule utilitarianism, regarded as a theory of individual action, may function as a sort of “philosophical explanation” of how a rational individual can be justified in rejecting Mr. Chance’s offer.
With this view in mind, we can make sense of the typical reply that one simply does not have to deal with things like Mr. Chance’s proffered bet in order to live one’s life. Individuals who reply in this way might be regarded as rule probabilists whose rules for living rationally do not include becoming involved in wagers like that proposed by Mr. Chance. This is what they mean when they say driving to get ginger ale is part of their normal life whereas gambling with Mr. Chance is not. Such individuals “frame” the issue differently in virtue of being rule probabilists. The reluctant gambler problem is a puzzle precisely because it frames the issue in question in a way that ignores the possibility that people are rule probabilists. Instead of thinking of Mr. Chance’s proposal as more or less equivalent to driving to the store themselves, they think of Mr. Chance as proposing that they abandon the principles of rationality by which they order their lives and consider doing something that seems, prima facie, bizarre.
In the case of Mr. Chance’s proffered wager, for example, I find myself considering the general rule that I will drive in all of the “normal” ways required to function in the society in which I live—including doing my job, raising my children, nurturing my family, and so forth. This is a very general formulation, and it will no doubt include and exclude different things depending on whose rule it is and how it fits into the set of rules by which that person’s decisions are governed. Such a rule clearly incorporates probabilities because, in adopting it, I must accept the very real, though hopefully small, probability that I will be killed or injured in the course of following this rule. Moreover, at least in my case, this rule does not incorporate the belief that I can arbitrarily alter these probabilities when I am frightened—for example, by Mr. Chance’s wager. As I have argued, the distinction between act and rule probabilism is intended as a philosophical explanation of how it is possible for a human being to reject Mr. Chance’s bet without jettisoning his or her claim to rationality.This particular rule is only an example, based on my own sensibilities, but I believe many would identify with it.
This rule probabilism construal of “taking probability as a guide,” however, appears to oblige us to explain why we should accept this interpretation rather than one modeled on act utilitarianism. Mr. Chance might reply to the above argument: “What you say makes sense if people are rule probabilists, but you still have the burden of explaining why most people are not more like act probabilists. After all, act probabilism seems to be a better fit with traditional theories of rational choice.”
I present two related answers to this challenge. First, the main point of the (Mr. Chance) thought experiment is to challengetraditional ways of thinking about rational choice. To argue that the burden of showing why alternative approaches should be considered is wholly mine is, essentially, to beg the question I raise. Someone who objects to the move to rule probabilism has at least as great a burden to justify his or her position as I do—even if it can be shown that act probabilism better accords with traditional normative theories of choice.
But, second, and more importantly, I maintain that even the most rational of us are rule probabilists by default. Perhaps Mr. Data, the android in Star Trek: The Next Generation would have the computational speed necessary to apply probability estimates to every action, whatever that might mean in light of the vagaries of the theory of action.But humans, I urge, do not come even close to having the ability to apply probability directly to every decision. Even the best calculators among us simply lack the capacity to be act probabilists.Since “ought implies can,” even where the “ought” is part of a normative theory of rationality, we are rule probabilists by default—unless, of course, one jettisons the earlier assumption that we do, in some sense, take probability as a guide to our lives.This consideration helps explain why, as we said above, those who reject Mr. Chance’s wager, frame the matter in terms of their (sets of) rules rather than as a straight choice among individual actions—as an act probabilist would.
The resolution of the reluctant gambler puzzle turns on rejecting the assumption that people are act probabilists. Recall that the whole point of the reluctant gambler problem is to put those who advocate taking probability as a guide to life in a situation where it seems that they should be indifferent between accepting Mr. Chance’s wager and going for the ginger ale themselves. The problem arises because many supposedly rational individuals would firmly refuse the wager—as would the rational thought experimenter—in this case, myself. Hence these individuals are anything but indifferent between the wager and driving to the store themselves. We circumvent the problem by showing that rejection of the wager can be philosophically explained in terms of the fact that humans are rule probabilists by default.
To conclude this analysis of the reluctant gambler problem, I have outlined a strategy to enable an individual to justify rejecting Mr. Chance’s proffered wager while continuing to “take probability as a guide to life.” In the next section, I show how this strategy is useful in dealing with the new reality of terrorism directed against the United States. While there are important similarities between the case of Mr. Chance and that of Mr. Terrorist, the applicability of our strategy to terrorism in no way depends on any sort of perfect match, or analogy, between these two cases. The point of this paper is to show that dealing with the “reluctant gambler” problem suggests a theoretical approach for dealing with recent terrorist threats.
“Terrorism” and the Reluctant Gambler
I begin this section with a reference to an opinion poll conducted by one of the leading polling experts in the United States. Professor Richard Curtin, Director of Surveys of Consumers at The University of Michigan, states that his most recent polling of U.S. consumers shows that they “exaggerate their personal risks from terrorism.”The importance of this finding for this paper cannot be exaggerated. Empirical data by one of the most respected polling experts in the country supports the position that American consumers believe they are at greater risk from terrorism, in this post September 11 world, than serious statistical analysis indicates is warranted. This result provides a framework of fact for the more speculative analysis of this paper. It also clearly signals a difference between terrorism and the reluctant gambler, where we posited that Mr. Chance’s probability estimate was true.
It is as if Mr. Terrorist came to us on September 11 and following and offered us a sort of wager. He confronted us with our own mortality by placing us in a situation where the maxims of our lives—the sorts of rules we live by as rational rule probabilists—seem to be cast into question by the new reality spawned that fateful day. In this respect, the parallel with the reluctant gambler problem is clear. Mr. Chance forces us to examine our probabilistic rules pertaining to certain sorts of driving. Rational individuals know about the risks associated with driving but have come to terms with them—perhaps driving a safer car, driving more carefully, or being more selective about where and when to drive, and incorporating these decisions into their rule, or rules, about driving. Mr. Chance forces us to think about these driving risks, and perhaps that is why most people prefer to follow their old (set of) rules and drive to the store. We simply do not have to deal with Mr. Chance’s wager as part of our normal life, so we reject it. Similarly, Mr. Terrorist calls attention to the probabilistic rules by which we order our lives. Ordinarily, we would not engage in a critical reevaluation of these rules. Over time, we have arranged our lives to reduce serious risks while engaging in activities we find fulfilling and necessary for our usual functioning. It is as if Mr. Terrorist came to our door on September 11, 2001, and said, “Look, you simply cannot rely on your old ways of doing things, including your old ways of incorporating probability judgments into your decisions. The rules you have been using to organize your lives are now useless. What we have done, and threaten to do, changes everything! Pay attention to the warnings! Dramatically alter your lives! Be afraid! You cannot hide behind the fact that you are not soldiers. We want to kill you! (Remember, ‘terrorism’ is the more or less random killing of civilians for the purpose of making a political statement.)”
Mr. Chance offered us a rather strange bet that threatened to disrupt one very specific aspect of our lives simply by making us aware of the ordinary (and real) risks associated with driving to the store. Mr. Terrorist, on the other hand, challenges the basic strategies by which rational rule probabilists would order most aspects of their lives. He is not saying that we can go on with our lives knowing that the risks associated with living are essentially the same as they were; in effect, he is claiming that we face much more serious risks now that he is targeting us for terrorist attacks. In this respect, Mr. Terrorist’s offering appears to differ from that of Mr. Chance, who makes clear that our probability of dying is the same whether we take his bet or not. We can reject Mr. Chance’s bet. At worst, we may come to believe we are less rational than we thought we were—though I have challenged that perception. However, many believe we cannot reject Mr. Terrorist’s bet without placing ourselves and others in harm’s way.
There does appear to be a parallel between the payoffs that Messrs. Chance and Terror offer us. Just as Mr. Chance offers only the same ginger ale we wanted in the first place, Mr. Terrorist is offering us, in a sense, exactly what we have but—and here the similarity breaks down—with a ghastly difference: If we take his “bet,” we will experience fear, arguably paralyzing fear.
Shouldn’t we be afraid? The World Trade Center and the Pentagon were attacked, anthrax laced the mail, and who knows what is next? Our own national government issues alert after alert, urging us to be vigilant, but offering no specific directions about how to do that. Mr. Terrorist wants us to be worried and afraid. Yet, contrary to what Mr. Terrorist wants us to believe, our probability of dying in a terrorist attack is so small that it is essentially the same as it was before the attacks.Some may face greater threats from terrorists, depending on where they live, the requirements of their jobs, and so on. Yet even these individuals’ risks of being killed by terrorists are negligible if viewed in the context of the risks they faced before September 11. Since September 11, 2001, fewer than thousand people have died in the worst terrorist attacks in this nation’s history. No one in their right mind fails to think this an atrocity—especially given that these people were murdered to make a political point. On the other hand, this year, twenty thousand people will die of the flu, forty-five thousand will die and hundreds of thousands will be injured in auto accidents, countless numbers will die of various “natural” diseases, and so on.An important difference between the reluctant gambler problem and Mr. Terrorist’s imposed wager is that the first is structured to be a one-time thing, and it is so specific and isolated that we can ignore it without having to dislocate our lives, as I explained above. Mr. Terrorist, on the other hand, presents us with the prospect of continuing “wagers” and that makes it difficult to ignore him and continue to live our lives. However difficult it may be to ignore Mr. Terrorist, he does not appreciably increase our risks. What Mr. Terrorist fears is that we will note our risks are largely unchanged, refuse to be terrified, and only change those maxims of living that will create greater difficulty for him.
In his book, Fear Less, Gavin De Becker argues that the generalized “alerts” from the government are probably a bad idea. De Becker is a risk expert with special interest in terrorism and he offers a number of specific suggestions of steps people in different occupations and walks of life can take in the present situation.In effect, De Becker is suggesting how we might remain rule probabilists but revise our rules with the terrorists’ “new reality” in mind—though he does not approach the problem in terms of rule versus act probabilism. It is not part of this paper to rehearse suggestions like those of DeBecker, but merely to call attention to this literature and stress its significance. Although I argue that the risks associated with a post September 11 world are negligible when placed in the context of the variety of risks we face, I do not deny that we can, and probably should, make changes in the rules by which we bring probability to bear on our decisions—changes which will decrease somewhat the risks terrorists pose for us. (Remember, I have argued these risks are already incredibly small.) But then there are ways to reduce most risks we face, for example, the risks associated with driving, disease, and smoking—some of which are not small. What I maintain we should not do is attempt the impossible task of becoming an act probabilist in the face of terrorist risks that are not appreciably greater than we faced prior to September 11—or, for that matter, in the face of other risks we face—though we might revise our rules to reduce these risks.
If I am right about the real risks we face from terrorist attacks, why should there be any issue at all?Why would rational individuals allow themselves to be drawn into the terrorists’ web of fear? Because of the random nature of their attacks, terrorists place us on the front line by the appalling proposition with which they confront us: That we are all “enemy soldiers” they want to kill—men, women, children, everyone is a target. These are the sorts of arbitrary attacks that Israelis have had to live with for over fifty years and that are currently intensifying. On September 11, all Americans were given notice that we too were potential targets, even in our own country, and we are not comfortable with this message.
This is precisely the importance of the lesson we learn from the reluctant gambler: Though we should probably revise many of our rules to make ourselves more difficult targets for terrorists, we must not permit ourselves to be turned toward act probabilism. I explained earlier that such a position is impossible for human beings to adopt. The lesson we learn from the reluctant gambler is that humans, to the extent that they take probability as a guide to their lives, must be rule probabilists. Unlike Mr. Data, we simply haven’t the capacity to calculate the risks of each individual action— whatever that might mean. To the extent that Mr. Terrorist can provoke us into such an attempt, he has terrorized us into a sort ofdecision paralysis. Should this happen, we will have been drawn away from our usual set of maxims, which incorporate the (albeit rough) guidance of probability; and we will find ourselves attempting to calculate the risks attending all of our actions—or so many of them that our lives are disrupted. If this happens to a sizable portion of the United States populace, the terrorists will have succeeded in terrorizing this nation.
Explanation and Justification
Early in this paper I alluded to an empirical study by Professor Richard Curtin, whose most recent empirical study of U.S. consumer behavior shows that American consumers “exaggerate their personal risks from terrorism.” These results suggest there is no justification for the outright panic many experience when they consider the terrorist attacks that have been made on the United States and future terrorist attacks that may occur.These reactions of panic and fear might be explained, however.Like a good deal of recent work in the social sciences, my speculations presuppose something like the account of rationality that is commonplace to economists, political scientists, psychologists, technically trained philosophers, and others, though I make no formal reference to it here.Indeed, this effort is best regarded as a philosophical explanation of the sort discussed by Robert Nozick.Idescribecircumstancesthatcould,inthecontextofastandardtheoryofrational choice, lead to the reactions in question.
After the horrors of September 11, George W. Bush urged Americans to be vigilant but to “live their lives.” Many people seem not to be reassured by Bush’s rhetoric. They are still frightened.Indeed, the major TV news networks spent—and continue to spend—a great deal of time reminding us how vulnerable we are to terrorist attack and how terrible such attacks are apt to be. Often they are echoing remarks made by the President, the Vice President, or other high- ranking officials. Ana Marie Cox, in her nationally syndicated newspaper column, maintained that “worse-case scenario claims have become [a] cottage industry in [the] media since Sept. 11 attacks, overwhelming calmer voices.”
Terrorism has literally become the bread and butter of television’s twenty-four hour news channels—the main source of information for most Americans. Ana Marie Cox’s article, cited above, points out that many of the so-called experts on terrorism who regularly appear on these news channels have questionable credentials. What they do share, though, is a penchant for hysteria. One commentator said, for example, that the probability that al Quaeda would “decimate” this country was “50-50.”It is not difficult to understand why television networks and commentators would want to frighten us. Scared viewers are more likely to stay glued to their televisions, fearful they will miss key information needed to help survive the next onslaught of the terrorists. These television stations use their coverage of terrorism and its future possibilities in much the same way as terrorists use their attacks: To gain national attention. The 1970 era film Network tells the story of an industry in which the news divisions were in the process of being turned into “shows” to gain market share.AlthoughNetwork is a work of fiction, its references to shows about terrorist groups seem almost quaint compared to what is actually showing on the 24-hour news channels. Today, it is almost as if a sort of “Clippy”—the little paperclip “helper” from Microsoft who graces the screens of my computers—appears on all of our media and constantly reminds us of the new reality of terrorism.
In addition, the Bush administration has been under fire because of its ties to corporate scandals associated with Enron, World Com, and others and because of allegations that more was known about the terrorist attacks of September 11 than has been divulged. As soon as the government was challenged on its intelligence failures, Vice President Cheney went on television and said that we would almost certainly be attacked and probably with weapons of mass destruction.One does not have to believe the Bush administration is using terrorist threats to deflect criticism to understand that these officials are clearly in “potential conflict of interest” in this matter.That is, it is in their interest to deflect criticism from areas that are potentially embarrassing, like failures of the economy and intelligence failures. Both the Whitehouse and the twenty-four hour news channels have reasons to make sure Americans are focused on the threat of terrorism—and the television stations are in a position to orient their programming to make sure that is their focus. There are other voices, but these are seldom heard.It is not surprising, then, that Americans are deeply worried about terrorist threats—worried beyond anything that is rational. Though not a formal scientific explanation of why Mr. Terrorist has been so successful in frightening us toward adopting an untenable “act probabilism,” the discussion in this section does suggest how such a result is possible.
This paper focuses on the individual actor, but ultimately the citizens of a nation, presented with a serious threat, demand that their leaders protect them. Democracies have always managed to make such adjustments in time of war.Today, we are told, we are fighting a “war on terror,” which may last many years. In light of the preceding section, it should not be surprising that many Americans are so obsessed with terrorist threats that they have acquiesced to their leaders making major changes in the country’s political infrastructure ostensibly to reduce the likelihood of terrorist attacks. At both the federal and state levels, legislation has been passed giving law enforcement agencies broad powers that clearly supersede anything envisioned in the U.S. Constitution. No price seems too high if people come to believe that paying it provides even slightly greater protection from terrorists, and they are regularly encouraged in such beliefs.
Decision theorists call the problem of moving from the individual preferences of the members of a group to a collective—or group—preference “an amalgamation problem.”To solve such a problem rationally, it is essential that we know the real “costs” of various options. The preceding sections show why most people are not in a good position to assess rationally the risks associated with terrorism. The steady diet of fear that pours from their television screens, and from many other media, whether by accident or design, seem to foster a frantic, act probabilist response—a response I have argued is literally incoherent.
In principle, though, we are in a much better position to appraise the risks of abandoning our basic legal protections as embodied in the Constitution. We know, for example, that the Constitution is a collection of legal rules. In short, it is a rule based—not an act based—document. We also know that the Constitution embodies the fundamental rights and duties that are the bases of our freedom. Furthermore, we know the rights embodied in the principles of the Constitution are difficult to achieve but very easy to lose and, thus, that we should not part with them except in the most extreme circumstances. Even then, we should make provisions that allow us to return to our Constitutional base after the crisis is over. Finally, we know that, even in the face of nuclear destruction, during the Cold War, we managed to preserve basic Constitutional guarantees. The Constitution warrants such protection because it embodies the rights that make the United States, imperfect though it may be, a free and democratic society.
I have argued that the individual risks associated with the terrorism we confronted on September 11 are not significantly greater than the risks Americans faced prior to that date.I have suggested that we may believe they are greater because our media and our leaders are telling us they are, but a rational assessment of the probabilities shows otherwise.
On the other hand, the risks of losing our basic rights and freedoms because we trade them away for what our leaders say is greater security are very real. Indeed, many rights and liberties have already been lost or severely diminished.It is consistent with this paper that Americans should not continue to give up basic rights for the sake of supposed increases in security against terrorism. The price is just too high—especially in light of both Professor Curtin’s empirical results that consumers exaggerate their risks from terrorist attack and my more theoretical argument that the risks associated with terrorism are essentially the same as we faced prior to September 11, 2001. If we want to consider reducing the risks associated with the new reality of terrorism, that may be worth doing. But we should approach this project against the backdrop of a realistic appraisal of the risks from terrorist attack—and we should be very careful to understand the costs of such risk reduction. Alan Dershowitz makes a number of proposals for diminishing terrorist risks in his excellent new book Why Terrorism Works.What is particularly valuable about this book is that Dershowitz spells out both how a non-democratic government would seek to eliminate terrorism and how a democracy should proceed, balancing the preservation of Constitutional protections against security needs. Finally, any such risk reduction must take place within the framework of rule probabilism. This is consistent with Dershowitz’s insistence that a democracy should not jettison constitutional protections even if these may need to be reviewed in light of terrorist threats.
This paper began with an analysis of the reluctant gambler problem—a problem in which the notion of guiding one’s life by probability seemed to conflict with the preferences of apparently rational people (or rational thought experimenters). To resolve that problem, I made a distinction between rule and act probabilism, on the analogy of rule and act utilitarianism. Arguing that humans are rule probabilists by default, I suggested that, if those who refuse Mr. Chance’s wager—our reluctant gamblers—are considered to be rule probabilists, their decision to reject Mr. Chance’s bet is consistent with the idea that they rationally guide their lives by probability.
Next, I stressed that the parallel between the reluctant gambler and our response to terrorist attacks was only partial. But the distinction between act and rule probabilism used to analyze the reluctant gambler problem suggested a way of confronting the new terrorist risks we faced after September 11—especially in light of the fact that the risks associated with such attacks today are essentially the same as they were before September 11.
The perception that the risks of terrorism have somehow mushroomed in light of the events of September 11 may be at least partially explained by the television coverage of the 24-hour news networks and the potential interest of the current administration in diverting public attention away from their own questionable practices and issues. While the risks of terrorism are not appreciably higher than they were prior to September 11, it might be possible to reduce these risks. Unfortunately, our national government seems to have determined that the best way to diminish such risks is to dismantle our Constitution—a document which is a testament to the value of having clear rules to protect citizens’ rights and freedoms. The important question Americans now face is how much they want to pay—especially in terms of diminished rights and liberties—to reduce a risk that is essentially the same as it was several years ago. If we continue down this path, our chances of remaining a uniquely free and open society do not look promising.
Finally, in an appendix, I differentiate ‘terrorism’ from what I call ‘terror war.’ I contest the claim that the possibility of the U.S. facing terrorist attacks using weapons of mass destruction offers a serious challenge to my analysis in this paper.
Appendix: “Terror War”
A strike on the U.S. with one or more weapons of mass destruction (WMDs) would not be in the realm of what I term ‘terrorism.’ The order of magnitude of such a strike would be qualitatively different from terrorism as we have known it. In the wake of September 11, 2001, the U.S. declared a “war on terror;” many, however, including myself, believe that terrorism as we know it, including the events of September 11, is really more analogous to a criminal assault than a war. But the use of WMDs by a country or a group of individuals clearly is an act of war. Because of the destructive power of such weapons, I suggest we label such an attack “terror war,” and this is not the topic of this paper.
Some will argue that, in restricting my topic in this way, I am really admitting that we should be terrified. “After all,” such an individual might intone, “What you are now calling ‘terror war’ is what is really scaring us.” I disagree. While I only sketch some features of my answer to this challenge, these suggestions reflect my general strategy. The destructive power of many WMDs is so great that the end result of such an attack on the United States simply overwhelms even major increases in the probability of such a result.
For example, if a nuclear attack on New York city would kill ten million and the probability of such an attack were, say, one in ten billion prior to September 11, what decision theorists call “the expected (negative) value” of such an attack will be 0.001.In short, the expected (negative) value of a one in ten billion chance of a nuclear bomb exploding in New York city and killing ten million people could be interpreted as the (actual) loss of one one-thousandth of a human life. If al Quaeda were able to increase the odds massively to one in one billion, the expected (negative) value becomes 0.01 or the (actual) loss of one one-hundredth of a human life. The difference is clearly insignificant. This happens because the (negative) end result of such a WMD’s use is so much more important (10 million dead) than the probability of such an end result (one in ten billion or one in one billion) when evaluating expected (negative) value—unless the probability increases astronomically or we assume we are starting with zero probability. While some may feel that al Quaeda is capable of such astronomical increase in the probability of a WMD being deployed against the United States or that New York city has been at zero risk of a nuclear strike since the Cold War ended, the arguments I outline below mitigate strongly against such prospects.
During the Cold War, the former Soviet Union and the United States were said to have a “balance of terror” because each could destroy the other. The strategy behind this “balance” was referred to as MAD, or mutually assured destruction. The message inherent in the MAD strategy was, and is, that an attack with WMDs rises to a level commensurate with using whatever weapons are required to defend your country—even if these weapons may kill civilians. Hence, the term “terror war.” Note, incidentally, that the MAD strategy, and the variations of it I sketch out below, is a policy—a rule that does not admit of obvious exceptions. That is what makes such a policy effective: the fact that your opponents know what to expect if they attack you, and you know that they know, and so on.
It can be argued that it is irrelevant who deploys a WMD against the U.S.; the risk associated with such a deployment is the same whoever does it. For forty years, Americans faced the prospect of nuclear annihilation in an exchange of many, many nuclear warheads with the former Soviet Union. After the Cold War ended, just over ten years ago with the breakup of the Soviet Union, most Americans seem to have forgotten that the majority of the (former Soviet) weapons were still there. While these weapons are not targeted as they were during the Cold War, they can be retargeted very quickly. During the last ten years, those weapons have continued to exist, and to pose a threat for us—albeit a significantly reduced threat. On the other hand, with the end of the Cold War, other nations began to develop WMDs, and there are now a number of other countries with such weapons. Many of them are potentially hostile to the United States. In a sense, the nuclear Damocles Sword that hung over our heads never went away completely. It was just shrunk a bit and shoved into a closet.
The conclusion is this: We have continued to face the prospect of being struck with WMDs since the late 1950s, and the probability of such a strike has always been very obscure. Al Quaeda terrorists, or other terrorists, who are intent on striking us with a WMD are just one more such threat; we have known that such threats from extremist groups existed for at least a decade. It can be argued, and this is particularly significant, that succeeding with a plan to strike the United States with a WMD will likely be significantly more difficult in light of September 11, since other states are now potentially less likely to allow a terrorist access to nuclear materials or technology. This is the thrust of our “rule” that countries are “either with us or against us” and that we will pursue something like a MAD strategy against anyone who aids terrorists in obtaining or deploying WMDs. But make no mistake, we have been continuously at risk for such a WMD attack since the late 1950s and it is very difficult to assign a probability to the risk of the U.S. being struck by a WMD.
Most experts do not think such an attack on the United States is imminent; but a WMD attack is possible.To assess this likelihood we need to consider what the U.S. response to such an unprovoked attack would likely be. If we were so attacked by a country, it is pretty clear what would happen, and no sane leader is likely to risk the nuclear annihilation of his or her nation by supporting such an assault.If a WMD is deployed against us and it is unclear who did it,how many Americans really believe our leaders would not be forced to respond, perhaps with our own WMDs, against someone? The citizens of the United States would demand it, and the world knows it.Perhaps this is why Yasser Arafat has so publicly and vehemently denounced al Quaeda. Arafat declared that (bin Laden) “never helped us, he was working in another completely different area and against our interests.”This means that any terrorist would have to think long and hard before launching that kind of attack because once the U.S. targeted a particular nation, or nations, for retaliation, it is very likely that information about who was behind the attack would be forthcoming. Let me be clear: I am not morally sanctioning such action; rather, I am saying that this is what I believe would happen were, say, ten million New Yorkers to die in a nuclear explosion. In any event, the considerations of this paper pertain to terrorism, not to terror war. Though it may be difficult to draw a precise line between the two, there are clear cases on both sides of the line. I plan a future project in which I will deal specifically with terror war and the strategies and morality pertinent to it.
Tversky and Kahneman’s distinction between normative and descriptive decision theory is suggestive. Amos Tversky and Daniel Kahneman, “Rational Choice and the Framing of Decisions,” Decision Making, ed. David E. Bell, Howard Raiffa, and Amos Tversky (Cambridge: Cambridge UP, 1988).
See endnote 25 below.
See Alan M. Dershowitz, Why Terrorism Works (New Haven: Yale UP, 2002).
The idea that one should “guide one’s life by probability” is intended as a sort of common sense, or pre-analytic, stand-in for a detailed theory of rational choice, incorporating both probability and utility. It might be urged that I simply state such a theory at the outset—perhaps incorporating into such a theory a solution to, or dissolution of, the “reluctant gambler problem”—more or less by fiat. I present the issue of “taking probability as a guide” in this more common sense way, however, precisely because I intend to raise questions I consider analytically prior to any precisely stated theory of rational choice. I could raise the same issues as a challenge to some standard theory of choice. Unfortunately, this approach would make the paper essentially unreadable by those not familiar with such work, and that seems inappropriate given the very general and intuitive issues the paper seeks to raise.
While the “classical” interpretation of probability is associated with certain problems—especially in regard to the principle of indifference—it is appropriate in this context. See Merrilee H. Salmon, et al., Introduction to the Philosophy of Science (Englewood Cliffs, N.J.: Prentice Hall, 1992) 74-77.
There are several technical points we should note in setting up this problem. First, it is only necessary that Ms. Jones find Mr. Chance’s statistical probability estimate acceptable. The precise statistical methodology he uses is irrelevant, as long as Mr. Chance is using generally accepted statistical methods. Second, because statistical probability technically does not apply to the “single case,” Ms. Jones must transform Mr. Chance’s statistical probability estimate into something like rational subjective probability. (See Salmon 77-84). Third, as I said in endnote 5 above, this problem can be set up using various standard theories of rational choice. The problem is constructed so that the “utility,” as determined by standard theories of rational choice, appears to be the same whichever decision Ms. Jones makes, but there are other issues that emerge below in our analysis.
It may strike us as strange that anyone would give someone the right to (even painlessly) kill them when they are not, for example, desperately ill—or even that such killing should be legal. All we need here, however, is that the reluctant gambler problem is a viable philosophical thought experiment, not that it fit our specific laws and customs.
I set the problem up very carefully for those whose opinions I canvass. They include people who understand at least the elements of the theory of rational choice and the basic interpretations of probability. In short, I am assuming well-informed, and presumably rational, observers, though most of them are not specialists in probability and choice theory. That is intentional since the problem is designed to challenge a standard interpretation of rational decision theory.
For an elementary introduction to the theory of rational choice, see Duncan Luce and Howard Raifa, Games and Decisions (New York: John Wiley and Sons, 1957).
See Luce and Raifa, op. cit.
Note: I am not drawing any sort of exact analogy between the act/rule utilitarianism distinction and that between act and rule probabilism. I merely offer the former as a suggestive motivation of the latter.
 Manuel Velasquez, Business Ethics: Concepts and Cases, 3rd ed. (Englewood Cliffs, N.J.: Prenctice Hall, 1992) 60-70. Note: I presuppose a number of refinements in utilitarian theory not included in most accounts, including Velasquez. The most important is the idea that it is not individual rules that the rule utilitarian justifies but, rather, a set of rules. I assume that a set of rules cannot be altered piecemeal without loss of utility, since the rules that make up such a group will interact with one another in various ways.
 In personal correspondence, some years ago, John Rawls made this point in connection with a challenge I made to his paper, “Two Concepts of Rules,” reprinted in John Rawls, Collected Papers, ed. Samuel Freeman (Cambridge: Harvard UP, 1999) 20-46.
Such theories are typically constructed, at least initially, as theories of individual action. The “amalgamation” of the decisions of individual actors into a sort of group action is treated as a separate problem. Again, Luce and Raifa, op. cit. is a good introduction.
Since utilitarianism is an ethical theory, I leave open the question whether a human can be an act utilitarian—though I doubt that this is possible.
Robert Nozick, Philosophical Explanations (Cambridge: Harvard UP, 1981) 8-11. Here I utilize Robert Nozick’s notion of a “philosophical explanation” regarded as an account of how something is possible. The importance of this notion where I refer to “explanation” is to highlight that I am not making straightforward empirical claims; rather, I am indicating how certain things are possible in light of various assumptions.
See endnote 5 above.
 See endnote 17 above.
I remind the reader that my analysis of rule probabalism (or, indeed, rule utilitarianism) assumes we are dealing with sets of rules that are interconnected with each other. A full explanation of this for either theory is beyond the bounds of this paper.
See, for example, Star Trek: Insurrection, dir. Jonathon Frakes, Paramount Pictures, 1998.
As noted above, I do not deny that ordinary people make individual calculations in some especially important cases. But I contend that rational people would be forced to adopt, even if implicitly, rules to cover most of their dealings because they cannot, in principle, calculate every individual action. That is what it means to say they are “rule probabilists by default.”
Refer to the preceding note and to endnote 2 above. A detailed treatment of both choice theory and utilitarianism would reveal more complexities, but exploring them would take us well beyond the limited, more or less “broad stroke,” goals of this paper.
See endnote 17 above.
Personal conversation. The results of these polls will be published in Richard T. Curtin, “What Recession? What Recovery? The Arrival of the 21st Century Consumer,” Department of Economics, The Economic Outlook for 2003 (Ann Arbor: U of Michigan P, forthcoming January 2003.) Note that Professor Curtin is in no way responsible for any logical or interpretative errors in my use of his empirical results.
For a more detailed account of what I take terrorism to be, refer to the Appendix.
See endnote 25 above. The Appendix is also relevant here.
The reference for U.S. automobile deaths is Ruth Gastel, ed., Auto Safety and Crash Worthiness (N.p.: 2002 Insurance Institute III Insurance Issues Update, August, 2002). For information on flu deaths, see “Seniors at Greater Risk as Flu Season Approaches” (Business Wire, 21 Oct. 2002).
Gavin De Becker, Fear Less (Boston: Little, Brown, and Co., 2002). Note: The fact that we can reduce the risk is compatible with my claim that the risks from terrorism are negligible. We all face many risks every day. The great majority of them can be reduced. The question is usually whether we want to take the trouble—and pay the cost—of such risk reduction.
Again, I refer the reader to Professor Curtin’s empirical results regarding Americans’ perception of terrorist risks. See endnote 25 above.
See the Appendix on “terrorism” versus “terror war.”
See endnote 17 regarding “how possible” explanations.
Again, endnote 17 is especially relevant. Ultimately, empirical investigation will be necessary to confirm our suggestions, but such empirical research, if it is to be useful, must reflect the theoretical bases of our rationality. This section outlines how the theory of rational behavior may—when inappropriately utilized—help us understand the irrational reactions of many Americans to terrorist threats.
See endnote 17 above.
This point reflects a common stance taken both by many ordinary citizens and journalists. For support, see endnote 36 below.
Ana Marie Cox, “Stuck on Terror,” Lansing State Journal 28 July 2002: 11A.
Network,dir. Sidney Lumet, Turner Entertainment, 1976.
See Microsoft Windows 98, Second Edition (Computer Operating System).
Tom Stuckey, “Cheney Says There is No Doubt Terrorists Wish to Strike Again,” Associated Press 24 May 2002: BC cycle.
Valasquez 377-379. The news networks have an even greater potential conflict of interest than the administration, as I indicate.
Crane Brinton, The Shaping of Modern Thought (Englewood Cliffs, N.J.: Prentice Hall, 1950).
Robin Toner and Janet Elder, “A Nation Challenged,” New York Times 12 December 2001, Late Ed.: 1.
John C. Harsanyi, Rational Behavior and Bargaining Equilibrium in Games and Social Situations (Cambridge: Cambridge UP, 1977). Harsanyi is perhaps the most comprehensive reference on this subject.
Stephen Holmes and Cass R. Sunstein, The Cost of Rights (New York: W.W. Norton: 1999).
It is not that these risks cannot, like most risks, be reduced; it is, rather, that they are not appreciably greater, as they stand, than they were prior to September 11.
The references for this are vast. See endnote 4: Dershowitz speaks to this issue seriatum.See also Lynda Guydon Taylor, “Groups Worry Liberties are Compromised in Name of Anti Terrorism,” Pittsburgh Post-Gazette, 27 Oct. 2002, One Star ed.: W-3.
See endnote 4.
See endnote 4.
According to the 2000 United States Census, the population of New York City is just over 8 million; and the population of the greater New York city metropolitan area is 21.2 million. My choice of “ten million” in my example is somewhat arbitrary, but it would not alter my point if I worked with another number—say, five, eight, fifteen, or twenty million.
My remark here is intended to refer to countries likely to be directly supporting terrorist strikes against the U.S. Neither Russia nor China is very likely to provide such direct support, so I exclude them from consideration. If they were included, it would certainly complicate the U.S. response.
Many I have spoken with believe some of the nations that rushed to our side after September 11 feared we would loose a nuclear fusillade against someone in retaliation. But, then, we were not struck with a WMD.
Greg Myer, “Arafat Decries al Quaeda Tactics,” Lansing State Journal 16 Dec 2002: 1A.
Brinton, Crane. The Shaping of Modern Thought. Englewood Cliffs, N.J.: Prentice Hall, 1950. Cox, Ana Marie. “Stuck on Fear.” Lansing State Journal 28 July 2002: 11A.
Crichton, Michael. Eaters of the Dead. New York: Ballantine Books, 1976.
Curtin, Richard T. “What Recession? What Recovery? The Arrival of the 21st Century Century Consumer.” Department of Economics. The Economic Outlook for 2003. Ann Arbor: U of Michigan P, forthcoming 2003.
De Becker, Gavin. Fear Less. Boston: Little, Brown, and Co., 2002.
Dershowitz, Alan M. Why Terrorism Works. New Haven: Yale UP, 2002.
Frakes, Jonathon, dir. Star Trek: Insurrection. Paramount Pictures, 1998.
Gastel, Ruth. Ed. Auto Safety and Crash Worthiness. N.p: 2002 Insurance Institute III Insurance Issues Update. August 2002.
Harsanyi, John C. Rational Behavior and Bargaining Equilibrium in Games and Social Situations. Cambridge: Cambridge UP, 1977.
Holmes, Stephen and Cass R. Sunstein. The Cost ofRights. New York: W.W. Norton and Co. 1999. Luce, Duncan and Howard Raifa. Games and Decisions. New York: John Wiley and Sons, 1957. Lumet, Sidney, dir. Network. Turner Entertainment, 1976.
Microsoft Windows 98, Second Edition (Computer Operating System).
Myer, Greg. “Arafat Decries al-Qaeda Tactics.” Lansing State Journal. 16 Dec 2002: 1A. Nozick, Robert. Philosophical Explanations. Cambridge: Harvard UP, 1981.
44 Rawls, John. Collected Papers. Ed. Samuel Freeman. Cambridge: Harvard UP, 1999.
Salmon, Merrilee H., John Earman, Clark Glymour, James G. Lennox, Peter Machamer, J. E. McGuire, John D. Norton, Wesley C. Salmon, and Kenneth F. Shaffner. Introduction to the Philosophy of Science. Englewood Cliffs, N.J.: Prentice Hall, 1992.
“Seniors at Greater Risk as Flu Season Approaches.” Business Wire. 21 Oct. 2002.
Stuckey, Tom. “Cheney Says There is No Doubt Terrorists Wish to Strike Again.” Associated Press 24 May 2002: BC cycle.
Taylor, Linda Guydon. “Groups Worry Liberties are Compromised in Name of Anti Terrorism.” Pittsburgh Post-Gazette. 27 Oct 2002. One Star ed.: W-3.
Toner, Robin and Janet Elder. “A Nation Challenged.” New York Times 12 Dec. 2001, Late Ed.: 1.
Tversky, Amos and Daniel Kahneman. “Rational Choice and the Framing of Decisions.” Decision Making. Eds. David E. Bell, Howard Raiffa, and Amos Tversky. Cambridge: Cambridge UP, 1988.
Velasquez, Manuel. Business Ethics: Concepts and Cases. 3rd ed. Englewood Cliffs, N.J.: Prentice Hall. 1992.