Prohibition, Compensation, and Risk
INDEPENDENTS AND THE DOMINANT PROTECTIVE AGENCY
LET us suppose that interspersed among a large group of persons who deal with one protective agency lives some minuscule group who do not. These few independents (perhaps even only one) jointly or individually enforce their own rights against one and all, including clients of the agency. This situation might have arisen if native Americans had not been forced off their land and if some had refused to affiliate with the surrounding society of the settlers. Locke held that no one may be forced to enter civil society; some may abstain and stay in the liberty of the state of nature, even if most choose to enter (§ 95).1
How might the protective association and its members deal with this? They might try to isolate themselves from the independents in their midst by forbidding anyone permission to enter their property who hadn’t agreed to forgo exercising rights of retaliation and punishment. The geographical territory covered by the protective association then might resemble a slice of Swiss cheese, with internal as well as external boundaries.* But this would leave acute problems of relations with independents who had devices enabling them to retaliate across the boundaries, or who had helicopters to travel directly to wrongdoers without trespass upon anyone else’s land,† and so on.
Instead of (or in addition to) attempts at geographically isolating independents, one might punish them for their misenforcements of their rights of retaliation, punishment, and exaction of compensation. An independent would be allowed to proceed to enforce his rights as he sees them and as he sees the facts of his situation; afterwards the members of the protective association would check to see whether he had acted wrongly or overacted. If and only if he had done so, would they punish him or exact compensation from him.2
But the victim of the independent’s wrongful and unjust retaliation may be not only damaged but seriously injured and perhaps even killed. Must one wait to act until afterwards? Surely there would be some probability of the independent’s misenforcing his rights, which is high enough (though less than unity) to justify the protective association in stopping him until it determines whether his rights indeed were violated by its client. Wouldn’t this be a legitimate way to defend their clients? 3 Won’t people choose to do business only with agencies that offer their clients protection, by announcing they will punish anyone who punishes a client without first using some particular sort of procedure to establish his right to do this, independently of whether it turns out that he could have established this right? Is it not within a person’s rights to announce that he will not allow himself to be punished without its first being established that he has wronged someone? May he not appoint a protective association as his agent to make and carry out this announcement and to oversee any process used to try to establish his guilt? (Is anyone known so to lack the capacity to harm another, that others would exclude him from the scope of this announcement?) But suppose an independent, in the process of exacting punishment, tells the protective agency to get out of his way, on the grounds that the agency’s client deserves punishment, that he (the independent) has a right to punish him, that he is not violating anyone’s rights, and that it’s not his fault if the protective agency doesn’t know this. Must the agency then abstain from intervening? On the same grounds may the independent demand that the person himself refrain from defending himself against the infliction of punishment? And if the protective agency tries to punish an independent who punished a client, independently of whether their client did violate the independent’s rights, isn’t the independent within his rights to defend himself against the agency? To answer these questions and hence to decide how a dominant protective agency may act toward independents, we must investigate the moral status within a state of nature of procedural rights and of prohibitions upon risky activities, and also what knowledge is presumed by principles about the exercise of rights, including especially rights to enforce other rights. To these issues, difficult ones for the natural-rights tradition, we now turn.
A line (or hyper-plane) circumscribes an area in moral space around an individual. Locke holds that this line is determined by an individual’s natural rights, which limit the action of others. Non-Lockeans view other considerations as setting the position and contour of the line.4 In any case the following question arises: Are others forbidden to perform actions that transgress the boundary or encroach upon the circumscribed area, or are they permitted to perform such actions provided that they compensate the person whose boundary has been crossed? Unravelling this question will occupy us for much of this chapter. Let us say that a system forbids an action to a person if it imposes (is geared to impose) some penalty upon him for doing the act, in addition to exacting compensation from him for the act’s victims.* Something fully compensates a person for a loss if and only if it makes him no worse off than he otherwise would have been; it compensates person X for person Y’s action A if X is no worse off receiving it, Y having done A, than X would have been without receiving it if Y had not done A. (In the terminology of economists, something compensates X for Y’s act if receiving it leaves X on at least as high an indifference curve as he would have been on, without it, had Y not so acted.)† Shamelessly, I ignore general problems about the counterfactual “as well off (on as high an indifference curve) as X would have been if Y’s action hadn’t occurred.” I also ignore particular difficulties; for example, if X’s position was deteriorating (or improving) at the time, is the baseline for compensation where he was heading or where he was then? Are things changed if X’s position would have worsened anyway the next day? But one question must be discussed. Does the compensation to X for Y’s actions take into account X’s best response to these actions, or not? If X responded by rearranging his other activities and assets to limit his losses (or if he made prior provision to limit them), should this benefit Y by lessening the compensation he must pay? Alternatively, if X makes no attempt to rearrange his activities to cope with what Y has done, must Y compensate X for the full damage X suffers? Such behavior on X’s part may seem irrational; but if Y is required to compensate X for his full actual loss in such cases, then X will not be made worse off by his own noncoping, nonadaptive behavior. If so required, Y might lower the amount of compensation he must pay by paying X to respond adaptively and so to limit losses. We shall tentatively adopt another view of compensation, one which presumes reasonable precautions and adjusting activities by X. These activities would place X (given Y’s acts) on a certain indifference curve I; Y is required to raise X above his actual position by an amount equal to the difference between his position on I and his original position. Y compensates X for how much worse off Y’s action would have made a reasonably prudently acting X. (This compensation structure uses measurement of utility on an interval scale.)
A person may choose to do himself, I shall suppose, the things that would impinge across his boundaries when done without his consent by another. (Some of these things may be impossible for him to do to himself.) Also, he may give another permission to do these things to him (including things impossible for him to do to himself). Voluntary consent opens the border for crossings. Locke, of course, would hold that there are things others may not do to you by your permission; namely, those things you have no right to do to yourself.5 Locke would hold that your giving your permission cannot make it morally permissible for another to kill you, because you have no right to commit suicide. My nonpaternalistic position holds that someone may choose (or permit another) to do to himself anything, unless he has acquired an obligation to some third party not to do or allow it. This should cause no difficulty for the remainder of this chapter. Let those who disagree imagine our discussion to be limited to those actions about which (they admit) the position does hold; and we can proceed along together, having factored out that divisive and, for immediate purposes, irrelevant issue.
Two contrasting questions delimit our present concern:
1. Why is any action ever prohibited, rather than allowed, provided its victims are compensated?
2. Why not prohibit all crossings of the moral boundary that the party impinged upon did nor first consent to? Why ever permit anyone to cross another’s boundary without prior consent? 6
Our first question is too broad. For a system allowing acts A provided compensation is paid must prohibit at least the joint act of doing A and refusing to pay compensation. To narrow the issue, let us suppose there exist easy means to collect assessed compensation.7 Compensation is easily collected, once it is known who owes it. But those who cross another’s protected boundary sometimes escape without revealing their identity. Merely to require (upon detection, apprehension, and determination of guilt) compensation of the victim might be insufficient to deter someone from an action. Why wouldn’t he attempt continually to get away with it, to gain without paying compensation? True, if apprehended and judged guilty, he would be required to pay the costs of detecting, apprehending, and trying him; perhaps these possible additional costs would be sufficiently great to deter him. But they might not be. So one might be led to prohibit doing certain acts without paying compensation, and to impose penalties upon those who refuse to pay compensation or who fail to identify themselves as the crossers of certain boundaries.
RETRIBUTIVE AND DETERRENCE THEORIES OF PUNISHMENT
A person’s option of crossing a boundary is constituted by a (1 — p) chance of gain G from the act, where p is the probability he is apprehended, combined with the probability p of paying various costs of the act. These costs are first, the compensation to the victim over and above returning whatever transferable thing may be left from the ill-gotten gains, which we shall label C. In addition, since any nonremovable benefit from carrying out the act (for example, pleasure over fond memories) also will be exactly counterbalanced so as to leave none net, we may ignore it in what follows. Other costs are the psychological, social, and emotional costs of being apprehended, placed on trial, and so on (call them D); and the financial costs (call them E) of the processes of apprehension and trial which he must pay since they were produced by his attempt to evade paying compensation. Prospects for deterrence look dim if the expected costs of a boundary crossing are less than its expected gain; that is, if p × (C +D +E) is less than (I —p) × G. (Nevertheless, a person may refrain from a boundary crossing because he has something better to do, an option available to him with even higher expected utility.) If apprehension is imperfect, though inexpensive, additional penalties may be needed to deter crimes. (Attempts to evade paying compensation then would be made prohibited acts.)
Such considerations pose difficulties for retributive theories that set, on retributive grounds, an upper limit to the penalty that may be inflicted upon a person. Let us suppose (on such theories) that R, the retribution deserved, equals r × H; where H is a measure of the seriousness of the harm of the act, and r (ranging between 0 and 1 inclusive) indicates the person’s degree of responsibility for H. (We pass over the delicate issue of whether H represents the harm intended or the harm done or some function of both of these; or whether this varies with the type of case.)* When others will know that r = 1, they will believe that R = H. A person deciding whether to perform some harmful action then faces a probability (1 — p) of gain G, and a probability p of paying out (C + D + E + R). Usually (though not always) the gain from a boundary crossing is close to the loss or harm it inflicts on the other party; R will be somewhere in the neighborhood of G. But when p is small, or R is, p × (C + D + E + R) may be less than (1 — p) × G, often leaving no deterrence.*
Retributive theory seems to allow failures of deterrence. Deterrence theorists (though they wouldn’t choose to) would be in a position to gloat at retributivists’ squirming over this, if they themselves possessed another theory. But “the penalty for a crime should be the minimal one necessary to deter commission of it” provides no guidance until we’re told how much commission of it is to be deterred. If all commission is to be deterred, so that the crime is eliminated, the penalty will be set unacceptably high. If only one instance of the crime is to be deterred, so that there is merely less of the crime than there would be with no penalty at all, the penalty will be unacceptably low and will lead to almost zero deterrence. Where in between is the goal and penalty to be set? Deterrence theorists of the utilitarian sort would suggest (something like) setting the penalty P for a crime at the least point where any penalty for the crime greater than P would lead to more additional unhappiness inflicted in punishment than would be saved to the (potential) victims of the crimes deterred by the additional increment in punishment.
This utilitarian suggestion equates the unhappiness the criminal’s punishment causes him with the unhappiness a crime causes its victim. It gives these two unhappinesses the same weight in calculating a social optimum. So the utilitarian would refuse to raise the penalty for a crime, even though the greater penalty (well below any retributive upper limit) would deter more crimes, so long as it increases the unhappiness of those penalized more, even slightly, than it diminishes the unhappiness of those it saves from being victimized by the crime, and of those it deters and saves from punishment. (Will the utilitarian at least always select, between two amounts of penalty that equally maximize the total happiness, the option that minimizes the unhappiness of the victims?) Constructing counterexamples to this bizarre view is left as an exercise for the reader. Utilitarian deterrence “theory” could avoid this consequence, it seems, only by giving lesser weight to the punished party’s unhappiness. One would suppose that considerations of desert, which deterrence theorists had thought avoidable if not incoherent, would play a role here; one would suppose this if one weren’t bewildered at how to proceed, even using such considerations, in assigning the “proper” weight to different persons’ (un)happiness. The retributive theorist, on the other hand, doesn’t have to say that a felon’s happiness is less important than his victim’s. For the retributivist does not view determining the proper punishment as a task of weighing and weighting and allocating happiness at all.*
We can connect the retributive framework with some issues about self-defense. According to the retributive theory, the punishment deserved is r × H, where H is the amount of harm (done or intended) and r is the person’s degree of responsibility for bringing about H. We shall assume that the expected value of the harm to be visited upon a victim equals H (which fails to hold only if the person’s intentions fail to fit his objective situation). A rule of proportionality then sets an upper limit on the defensive harm which may be inflicted in self-defense on the doer of H. It makes the upper magnitude of the permissible defensive harm some function ƒ of H, which varies directly with H (the greater H is, the greater is f(H)), and such that f(H)>H. (Or at least, on any view, f(H)≥H.) Notice that this rule of proportionality does not mention the degree of responsibility r; it applies whether or not the doer is responsible for the harm he will cause. In this respect it differs from a rule of proportionality which makes the upper limit of self-defense a function of r × H. The latter sort of rule yields our judgment that, all other things being equal, one may use more force in self-defense against someone whose r is greater than zero. The structure we present here can yield this as follows. One may, in defending oneself, draw against the punishment the attacker deserves (which is r × H). So the upper limit of what one may use in self-defense against a doer of harm H is f(H) + r × H. When an amount A in addition to f(H) is expended in self-defense, the punishment which later may be inflicted is reduced by that amount and becomes r × H – A. When r = o, f(H) + r × H reduces to f(H). Finally, there will be some specification of a rule of necessity which requires one not to use more in self-defense than is necessary to repel the attack. If what is necessary is more than f(H) + r × H, there will be a duty to retreat.*
DIVIDING THE BENEFITS OF EXCHANGE
Let us return to the first of our two questions: why not allow any boundary crossing provided full compensation is paid? Full compensation keeps the victim on as high an indifference curve as he would occupy if the other person hadn’t crossed. Therefore a system that allows all boundary impingements provided that full compensation in paid is equivalent to a system requiring that all prior agreements about the right to cross a border be reached at that point on the contract curve 8 most favorable to the buyer of the right. If you would be willing to pay as much as $n for the right to do something to me, and $m is the least I would accept (receiving less than $m places me on a lower indifference curve), then there is the possibility of our striking a mutually advantageous bargain if n ≥ m. Within the range between $n and $m, where should the price be set? One cannot say, lacking any acceptable theory of a just or fair price (witness the various attempts to construct arbitration models for two-person, nonconstant sum games). Certainly, no reason has even been produced to think that all exchanges should take place at that point on the contract curve one of the parties most favors, to make the benefits of the exchange redound solely to that party. Allowing boundary crossing provided only that full compensation is paid “solves” the problem of distributing the benefits of voluntary exchange in an unfair and arbitrary manner.*
Consider further how such a system allocates goods. Anyone can seize a good, thereby coming to “own” it, provided he compensates its owner. If several people want a good, the first to seize it gets it, until another takes it, paying him full compensation. (Why should this sort of middleman receive anything?) 9 What amount would compensate the original owner if several persons wanted a particular good? An owner who knew of this demand might well come to value his good by its market price, and so be placed on a lower indifference curve by receiving less. (Where markets exist, isn’t the market price the least price a seller would accept? Would markets exist here?) Complicated combinations of subjunctive conditionals and counterfactuals might perhaps succeed in disentangling an owner’s preferences from his knowledge of the desires of others and the prices they are willing to pay. But no one yet has actually provided the requisite combinations.† A system cannot avoid the charge of unfairness by letting the compensation paid for a border crossing equal that price that would have been arrived at had a prior negotiation for permission taken place. (Call this compensation “market compensation.” It will usually be more than merely full compensation.) The best method to discover this price, of course, is to let the negotiations actually take place and see what their upshot is. Any other procedure would be highly inaccurate, as well as incredibly cumbersome.
The further considerations that militate against freely allowing all acts provided compensation is paid, in addition to those concerning the fairness of the exchange price, are in many ways the most interesting. If some injuries are not compensable, they would not fall under a policy of being allowed so long as compensation is paid. (Rather, they would be allowed provided compensation was paid, but since the compensation could not be paid by anyone, in effect they would be unallowed.) Leaving that difficult issue aside, even some acts that can be compensated for may be prohibited. Among those acts that can be compensated for, some arouse fear. We fear these acts happening to us, even if we know that we shall be compensated fully for them. X, learning that Y slipped in front of someone’s house, broke his arm, and collected $2,000 after suing for compensation for injuries, might think, “How fortunate for Y to have that happen; it’s worth breaking one’s arm in order to get $2,000; that completely covers the injury.” But if someone then came up to X and said, “I may break your arm in the next month, and if I do I will give you $2,000 in compensation; though if I decide not to break it I won’t give you anything,” would X dwell upon his good fortune? Wouldn’t he instead walk around apprehensive, jumping at noises behind him, nervous in the expectation that pain might descend suddenly upon him? A system that allowed assaults to take place provided the victims were compensated afterwards would lead to apprehensive people, afraid of assault, sudden attack, and harm. Does this provide a reason to prohibit assaults? Why couldn’t someone who commits assault compensate his victim not merely for the assault and its effects, but also for all the fear the victim felt in awaiting some assault or other? But under a general system which permits assault provided compensation is paid, a victim’s fear is not caused by the particular person who assaulted him. Why then should this assaulter have to compensate him for it? And who will compensate all the other apprehensive persons, who didn’t happen to get assaulted, for their fear?
Some things we would fear, even knowing we shall be compensated fully for their happening or being done to us. To avoid such general apprehension and fear, these acts are prohibited and made punishable. (Of course, prohibiting an act does not guarantee its noncommission and so does not ensure that people will feel secure. Where acts of assault, though forbidden, were frequently and unpredictably done, people still would be afraid.) Not every kind of border crossing creates such fear. If told that my automobile may be taken during the next month, and I will be compensated fully afterwards for the taking and for any inconvenience being without the car causes me, I do not spend the month nervous, apprehensive, and fearful.
This provides one dimension of a distinction between private wrongs and wrongs having a public component. Private wrongs are those where only the injured party need be compensated; persons who know they will be compensated fully do not fear them. Public wrongs are those people are fearful of, even though they know they will be compensated fully if and when the wrongs occur. Even under the strongest compensation proposal which compensates victims for their fear, some people (the nonvictims) will not be compensated for their fear. Therefore there is a legitimate public interest in eliminating these border-crossing acts, especially because their commission raises everyone’s fear of its happening to them.
Can this result be sidestepped? For example, there would not be this increase in fear if victims were compensated immediately, and also bribed to keep silent. Others wouldn’t know the act had been done, and so it wouldn’t render them more apprehensive by leading them to think that the probability of its happening to them was higher. The difficulty is that the knowledge that one is living under a system permitting this, itself produces apprehension. How can anyone estimate the statistical chances of something’s happening to him when all reports of it are squelched? Thus even in this highly artificial case it is not merely the victim who is injured by its happening in a system that is known to allow it to happen. The widespread fear makes the actual occurrence and countenancing of these acts not merely a private matter between the injurer and the injured party. (However, since victims compensated and bribed after the fact will not complain, enforcing the prohibition on these crimes which leave satisfied victims will illustrate the problems about enforcing prohibitions on so-called crimes without victims.)*
A system which allows fear-producing acts provided their victims are compensated, we have said, itself has a cost in the uncompensated for fear of those potential victims who are not actual victims. Would this defect of the system be avoided by someone who announced he would do a certain act at will, and not only would he compensate all of his victims, if any, but he would also compensate everyone who felt fear as a result of his announcement, even though he hadn’t actually done the act to them? This would be so expensive as to be beyond the means of almost everyone. But wouldn’t it slip through our argument for prohibiting those border crossings whose allowance (with compensation) would produce a general fear for which the populace would not be compensated? Not easily, for two additional reasons. First, persons might have free-floating anxiety about attack, not because they had heard some particular announcement, but because they know the system permits these attacks after announcement, and so worry that they have not heard some. They cannot be compensated for any they have not heard of, and they will not file for compensation for the fear these caused. Yet they may be the victims of someone whose announcement they haven’t heard. No particular announcement caused such fear without a specific announcement as its object, so who should compensate for it? Thus our argument is repeated one level up; but it must be admitted that at this level the fears may be so attenuated and insubstantial as to be insufficient to justify prohibiting such announcements. Secondly, in line with our earlier discussion of fair exchange prices, one might require someone who makes such an announcement to make not merely full but market compensation. Full compensation is an amount sufficient, but barely so, to make the person afterwards say he’s glad, not sorry, it happened; and market compensation is the amount that prior negotiations to get his consent would have fixed upon. Since fear looks very different in hindsight than it does while being undergone or anticipated, in these cases it will be almost impossible to determine accurately what is the amount of market compensation, except by actually going through the negotiations.
Our argument for prohibiting certain actions, such as assaults, assumes that merely to require an attacker to compensate his victim for the effects of the attack (though not for any general anticipatory fear) would not sufficiently deter attacks so as to leave people unfearful. The argument from fear fails if that assumption is mistaken. (There would remain the argument about the division of the benefits of exchange.) We might wonder whether the punishment deserved (according to retributive theory) for violating the prohibition on doing certain acts might similarly fail to provide sufficient deterrence of the acts so as to eliminate the fear and apprehension. This is unlikely if the probability of capture is high, and the punishment itself is a feared alternative; which punishment would not be illegitimate for feared wrongful acts. Even for persons who benefit much more from an act than its victims are hurt (and so, more than the punishment inflicted upon them), this will cause no difficulty. Recall that a retributive theory holds that a person’s ill-gotten gains are to be removed or counterbalanced, if any remain after he has compensated his victims, apart from the process of punishment.
The actual phenomenon of fear of certain acts, even by those who know they will receive full compensation if the acts are done to them, shows why we prohibit them. Is our argument too utilitarian? If fear isn’t produced by a particular person, how does it justify prohibiting him from doing an action provided he pays compensation? Our argument goes against the natural assumption that only the effects and consequences of an action are relevant to deciding whether it may be prohibited. It focuses also on the effects and consequences of its not being prohibited. Once stated, it is obvious that this must be done, but it would be worthwhile to investigate how far-reaching and significant are the implications of this divergence from the natural assumption.
There remains a puzzle about why fear attaches to certain acts. After all, if you know that you will be compensated fully for the actual effects of an act, so that you will be no worse off (in your own view) as a result of its having been done, then what is it that you are afraid of? You are not afraid of a drop to a less preferred position or a lower indifference curve, for (by hypothesis) you know that this won’t occur. Fear will be felt even when the total anticipated package is positive, as when someone is told that his arm may be broken and that he will be paid $500 more than the amount sufficient to compensate fully. The problem is not one of determining how much will compensate for the fear, but rather why there is any fear at all, given that the total package anticipated is viewed as desirable on the whole. One might suppose that the fear exists because the person is unsure that only a broken arm will be inflicted upon him; he does not know these limits will be observed. But the same problem would arise if it was guaranteed that the person would be compensated for whatever happened, or if an arm-breaking machine was used in the task, to eliminate the question of overstepping the limits. What would a person given such guarantees fear? We would like to know what sort of harms people actually are afraid of, even when they are part of a total package that is viewed as desirable on balance. Fear is not a global emotion; it focuses upon parts of packages, independently of “on-balance” judgments about the whole. Our present argument for the prohibition of compensable border crossings rests on this nonglobal character of fear, anxiety, apprehension, and the like.10 An answer specifying the types of harms might come in terms of ordinary notions such as “physical pain,” or in terms of the notions of some psychological theory such as “unconditioned aversive stimuli.” (But one should not leap to the conclusion that when it is known that compensation will be paid, only physical injury or pain is feared and viewed with apprehension. Despite knowing that they will be compensated if it occurs, people also may fear being humiliated, shamed, disgraced, embarrassed, and so on.) Secondly, we should like to know whether such fears are due to alterable features of the social environment. If people had been raised where great numbers of certain acts were randomly and unpredictably performed, would they exhibit great apprehension and fear of the risk of these acts, or would they be able to tolerate the risks as part of the normal background? (It would be difficult to detect or measure their apprehension if it expressed itself in heightened general tension. How does one measure how jumpy people generally are?) If people growing up in such a more stressful environment could develop a tolerance for certain acts, showing few symptoms of fear and stress, we would not have a very deep explanation of why certain acts are prohibited (rather than allowed provided compensation is paid). For the fear of these acts, which our explanation rests upon, would not itself be a deep phenomenon.11
The argument from general fear justifies prohibiting those boundary-crossing acts that produce fear even when it is known that they will be compensated for. Other considerations converge to this result: a system permitting boundary crossing, provided compensation is paid, embodies the use of persons as means; knowing they are being so used, and that their plans and expectations are liable to being thwarted arbitrarily, is a cost to people; some injuries may not be compensable; and for those that are compensable, how can an agent know that the actual compensation payment won’t be beyond his means? (Will one be able to insure against this contingency?) Do these considerations, combined with those about not unfairly distributing the benefits of voluntary exchange, suffice to justify prohibiting all other boundary-crossing acts, including those that do not produce fear? Our discussion of the first question we posed near the beginning of this chapter—“Why not permit all boundary crossings provided compensation is paid?”—has led us to the second question posed there—“Why not prohibit all boundary crossings to which the victim has not consented in advance?”
The penalization of all impingements not consented to, including accidental ones and those done unintentionally, would incorporate large amounts of risk and insecurity into people’s lives. People couldn’t be sure that despite the best of intentions they wouldn’t end up being punished for accidental happenings.12 To many, it also seems unfair. Let us put aside these interesting issues and focus upon those actions the agent knows will or might well impinge across someone’s boundary. Shouldn’t those who have not gotten their victims’ prior consent (usually by purchase) be punished? The complication is that some factor may prevent obtaining this prior consent or make it impossible to do so. (Some factor other than the victim’s refusing to agree.) It might be known who the victim will be, and exactly what will happen to him, but it might be temporarily impossible to communicate with him. Or it might be known that some person or other will be the victim of an act, but it might be impossible to find out which person. In each of these cases, no agreement gaining the victim’s permission to do the act can be negotiated in advance. In some other cases it might be very costly, though not impossible, to negotiate an agreement. The known victim can be communicated with, but only by first performing a brain operation on him, or finding him in an African jungle, or getting him to cut short his six-month sojourn in a monastery where he has taken a vow of silence and abstinence from business affairs, and so on; all very costly. Or, the unknown victim can be identified in advance only through a very costly survey of the whole population of possible victims.
Any border-crossing act which permissibly may be done provided compensation is paid afterwards will be one to which prior consent is impossible or very costly to negotiate (which includes, ignoring some complications, accidental acts, unintentional acts, acts done by mistake, and so on). But not vice versa. Which ones then may be done without the victim’s prior consent provided compensation is paid afterwards? Not those producing fear in the way described earlier.* Can we narrow it down further? Which nonfeared activities which do, or might, cross a border may permissibly be done provided compensation is paid? It would be arbitrary to make a hard distinction between its being impossible and its being very, very costly to identify the victim or communicate with him. (Not merely because it is difficult to know which a given case is. If the task used the United States GNP, would it be “impossible” or extremely costly?) The rationale for drawing a line at that particular place is unclear. The reason one sometimes would wish to allow boundary crossings with compensation (when prior identification of the victim or communication with him is impossible) is presumably the great benefits of the act; it is worthwhile, ought to be done, and can pay its way. But such reasons sometimes will hold, as well, where prior identification and communication, though possible, are more costly even than the great benefits of the act. Prohibiting such unconsented to acts would entail forgoing their benefits, as in the cases where negotiation is impossible. The most efficient policy forgoes the fewest net beneficial acts; it allows anyone to perform an unfeared action without prior agreement, provided the transaction costs of reaching a prior agreement are greater, even by a bit, than the costs of the posterior compensation process. (The party acted upon is compensated for his involvement in the process of compensation, as well as for the act itself.) But efficiency considerations are insufficient to justify unpenalized boundary crossings for marginal benefits, even if the compensation is more than full so that the benefits of exchange do not redound solely to the boundary crosser. Recall the additional considerations against permitting boundary crossings with compensation mentioned earlier (p. 71). To say that such acts should be allowed if and only if their benefits are “great enough” is of little help in the absence of some social mechanism to decide this. The three considerations of fear, division of the benefits of exchange, and transaction costs delimit our area; but because we have not yet found a precise principle involving the last and the considerations mentioned earlier (p. 71), they do not yet triangulate a solution in all its detail.
We noted earlier that a risky action might present too low a probability of harm to any given person to cause him worry or fear. But he might fear a large number of such acts being performed. Each individual act’s probability of causing harm falls below the threshold necessary for apprehension, but the combined totality of the acts may present a significant probability of harm. If different persons do each of the various acts in the totality, no one person is responsible for the resultant fear. Nor can any one person easily be held to cause a distinguishable part of the fear. One action alone would not cause fear at all due to the threshold, and one action less would probably not diminish the fear. Our earlier considerations about fear provide a case for the prohibition of this totality of activities. But since parts of the totality could occur without ill consequence, it would be unnecessarily stringent to ban each and every component act.13
How is it to be decided which below-threshold subsets of such totalities are to be permitted? To tax each act would require a central or unified taxation and decision-making apparatus. The same could be said for social determination of which acts were valuable enough to permit, with the other acts forbidden in order to shrink the totality to below the threshold. For example, it might be decided that mining or running trains is sufficiently valuable to be allowed, even though each presents risks to the passerby no less than compulsory Russian roulette with one bullet and n chambers (with n set appropriately), which is prohibited because it is insufficiently valuable. There are problems in a state of nature which has no central or unified apparatus capable of making, or entitled to make, these decisions. (We discuss in Chapter 5 whether Herbert Hart’s so-called “principle of fairness” aids here.) The problems could lessen if the overall states (totality below the threshold, and so on) can be reached by the operation of some invisible-hand mechanism. But the precise mechanism to accomplish this has yet to be described; and it would also have to be shown how such a mechanism would arise in a state of nature. (Here, as elsewhere, we would have use for a theory specifying what macrostates are amenable to production by what sorts of invisible-hand mechanisms.)
Actions that risk crossing another’s boundary pose serious problems for a natural-rights position. (The diversity of cases further complicates the issues: it may be known which persons will undergo a risk or merely that it will happen to someone or other, the probability of the harm may be known exactly or within a specified range, and so on.) Imposing how slight a probability of a harm that violates someone’s rights also violates his rights? Instead of one cutoff probability for all harms, perhaps the cutoff probability is lower the more severe the harm. Here one might have the picture of a specified value, the same for all acts, to mark the boundary of rights violation; an action violates someone’s rights if its expected harm to him (that is, its probability of harm to him multiplied by a measure of that harm) is greater than, or equal to, the specified value. But what is the magnitude of the specified value? The harm of the least significant act (yielding only that harm for certain) that violates a person’s natural rights? This construal of the problem cannot be utilized by a tradition which holds that stealing a penny or a pin or anything from someone violates his rights. That tradition does not select a threshold measure of harm as a lower limit, in the case of harms certain to occur. It is difficult to imagine a principled way in which the natural-rights tradition can draw the line to fix which probabilities impose unacceptably great risks upon others. This means that it is difficult to see how, in these cases, the natural-rights tradition draws the boundaries it focuses upon.*
If no natural-law theory has yet specified a precise line delimiting people’s natural rights in risky situations, what is to happen in the state of nature? With regard to any particular action that imposes a risk of a boundary crossing upon others, we have the following three possibilities:
1. The action is prohibited and punishable, even if compensation is paid for any boundary crossing, or if it turns out to have crossed no boundary.
2. The action is permitted provided compensation is paid to those persons whose boundaries actually are crossed.
3. The action is permitted provided compensation is paid to all those persons who undergo a risk of a boundary crossing, whether or not it turns out that their boundary actually is crossed.
Under the third alternative, people can choose the second; they can pool their payments for undergoing risk so as to compensate fully those whose boundaries actually are crossed. The third alternative will be plausible if imposing the risk on another plausibly is viewed as itself crossing a boundary, to be compensated for, perhaps because it is apprehended and hence imposes fear on the other.* (Persons voluntarily incurring such risks in the market are “compensated” by receiving higher wages for working at risky jobs, whether or not the risk eventuates.)
Charles Fried has recently suggested that people would be willing to agree to a system that allows them to impose “normal” risks of death upon each other, preferring this to a system that forbids all such imposing of risk.14 No one is especially disadvantaged; each gains the right to perform risky activities upon others in the pursuit of his own ends, in exchange for granting the others the right to do the same to him. These risks others impose upon him are risks he himself would be willing to undergo in the pursuit of his own ends; the same is true of the risks he imposes on others. However, the world is so constructed that in pursuing their ends people often must impose risks upon others that they cannot take directly upon themselves. A trade naturally suggests itself. Putting Fried’s argument in terms of an exchange suggests another alternative: namely, explicit compensation for each risk of a boundary crossing imposed upon another (the third possibility listed above). Such a scheme would differ from Fried’s risk pool in the direction of greater fairness. However, the process of actually carrying out the payments and ascertaining the precise risk imposed upon others and the appropriate compensation would seem to involve enormous transaction costs. Some efficiencies easily can be imagined (for example, keep central records for all, with net payments made every n months), but in the absence of some neat institutional device it remains enormously cumbersome. Because great transaction costs may make the fairest alternative impracticable, one may search for other alternatives, such as Fried’s risk pool. These alternatives will involve constant minor unfairness and classes of major ones. For example, children who die from the eventuating of the risks of death imposed upon them receive no actual benefit comparable to that of the risk imposers. This situation is not significantly alleviated by the facts that every adult faced these risks as a child and that every child who reaches adulthood will be able to impose these risks on yet other children.
A system that compensates only those upon whom risks eventuate (the second possibility listed above) would be far more manageable and would involve far smaller costs of operation and transaction than one which pays all those upon whom the risk is imposed (the third possibility above). Risks of death present the hardest issues. How can the magnitude of the harm be estimated? If the harm of death cannot actually be compensated for, the next best alternative, even apart from any issue of fear, might be to compensate all those upon whom its risk is imposed. But though postmortem payment to relatives or favorite charities, upkeep of elaborate cemetery arrangements, and so forth, all have obvious flaws insofar as the deceased is concerned, an individual himself can benefit from a system of postmortem compensatory payment to the estates of victims. While alive, he can sell the right to this payment, should it have to be made, to a company that purchases many such rights. The price would be no greater than the right’s expected monetary value (the probability of such payment multiplied by the amount); how much lower the price would be would depend upon the degree of competition in the industry, the interest rate, and so on. Such a system would not compensate fully any actual victim for the measured harm; and others not actually harmed also would benefit from having sold their collection rights. But each might view it, ex ante, as a reasonably satisfactory arrangement. (Earlier we described a way of pooling payments and transforming the third possibility into the second; here we have a way of transforming the second into the third.) This system also might give an individual a financial incentive to raise his “life’s monetary value” as measured by the compensation criteria, to increase the price for which he could sell the right to compensation.15
Even when permitting an action provided compensation is paid (the second or third possibilities above) is prima facie more appropriate for a risky action than prohibiting it (the first possibility above), the issue of its being prohibited or permitted to someone still is not completely settled. For some persons will lack sufficient funds to pay the required compensation should the need arise; and they will not have purchased insurance to cover their obligations in that eventuality. May these persons be forbidden to perform the action? Forbidding an action to those not in a position to pay compensation differs from forbidding it unless compensation is paid to those actually harmed (the second possibility above), in that in the former case (but not in the latter) someone who lacks provision for paying compensation may be punished for his action even though it does not actually harm anyone or cross a boundary.
Does someone violate another’s rights by performing an action without sufficient means or liability insurance to cover its risks? May he be forbidden to do this or punished for doing it? Since an enormous number of actions do increase risk to others, a society which prohibited such uncovered actions would ill fit a picture of a free society as one embodying a presumption in favor of liberty, under which people permissibly could perform actions so long as they didn’t harm others in specified ways. Yet how can people be allowed to impose risks on others whom they are not in a position to compensate should the need arise? Why should some have to bear the costs of others’ freedom? Yet to prohibit risky acts (because they are financially uncovered or because they are too risky) limits individuals’ freedom to act, even though the actions actually might involve no cost at all to anyone else. Any given epileptic, for example, might drive throughout his lifetime without thereby harming anyone. Forbidding him to drive may not actually lessen the harm to others; and for all anyone knows, it doesn’t. (It is true that we cannot identify in advance the individual who will turn out harmless, but why should he bear the full burden of our inability?) Prohibiting someone from driving in our automobile-dependent society, in order to reduce the risk to others, seriously disadvantages that person. It costs money to remedy these disadvantages—hiring a chauffeur or using taxis.
Consider the claim that a person must be compensated for the disadvantages imposed upon him by being forbidden to perform an activity for these sorts of reasons. Those who benefit from the reduction in risks to themselves have to “make it up” to those who are restricted. So stated, the net has been cast too broadly. Must I really compensate someone when, in self-defense, I stop him from playing Russian roulette on me? If some person wishes to use a very risky but efficient (and if things go well harmless) process in manufacturing a product, must the residents near the factory compensate him for the economic loss he suffers from not being allowed to use the possibly dangerous process? Surely not.
Perhaps a few words should be said about pollution—the dumping of negative effects upon other people’s property such as their houses, clothing, and lungs, and upon unowned things which people benefit from, such as a clean and beautiful sky. I shall discuss only effects on property. It would be undesirable, and is not excluded by anything I say below, for someone to channel all of his pollution effects high above anyone’s property volume, making the sky a murky grey-green. Nothing is gained by trying to transform the second type of case into the first by saying, for example, that someone who changes the way the sky looks dumps effects on one’s eyes. What follows in this note is incomplete in that it does not treat the second type of case.
Since it would exclude too much to forbid all polluting activities, how might a society (socialist or capitalist) decide which polluting activities to forbid and which to permit? Presumably, it should permit those polluting activities whose benefits are greater than their costs, including within their costs their polluting effects. The most feasible theoretical test of this net benefit is whether the activity could pay its way, whether those who benefit from it would be willing to pay enough to cover the costs of compensating those ill affected by it. (Those who favor any worthy activity that fails this test can make charitable donations to it.) For example, certain modes of airplane service impose noise pollution on homes surrounding airports. In one way or another (through lower resale value, lower rent obtainable for apartments, and so on), the economic value of these homes is diminished. Only if the benefits to air passengers are greater than these costs to airport neighbors should the noisier mode of transportation service go on. A society must have some way to determine whether the benefits do outweigh the costs. Secondly, it must decide how the costs are to be allocated. It can let them fall where they happen to fall: in our example, on the local homeowners. Or it can try to spread the cost throughout the society. Or it can place it on those who benefit from the activity: in our example, airports, airlines, and ultimately the air passenger. The last, if feasible, seems fairest. If a polluting activity is to be allowed to continue on the ground that its benefits outweigh its costs (including its polluting costs), then those who benefit actually should compensate those upon whom the pollution costs are initially thrown. The compensation might encompass paying for the costs of devices to lessen the initial pollution effects. In our example, airlines or airports might pay for soundproofing a house and then pay compensation for how much less the economic value of that house is than the value of the original unsoundproofed house in the neighborhood as it was without the additional noise.
When each of the victims of pollution suffers great costs, the usual system of tort liability (with minor modifications) suffices to yield this result. Enforcing other people’s property rights will, in these cases, suffice to keep pollution in its proper place. But the situation is changed if individual polluters have widespread and individually minuscule effects. If someone imposes the equivalent of a twenty-cent cost on each person in the United States, it will not pay for any one person to sue him, despite the great total of the cost imposed. If many persons similarly impose tiny costs on each individual, the total costs to an individual then may be significant. But since no single source significantly affects one individual, it still will not pay any individual to sue any individual polluter. It is ironic that pollution is commonly held to indicate defects in the privateness of a system of private property, whereas the problem of pollution is that high transaction costs make it difficult to enforce the private property rights of the victims of pollution. One solution might be to allow group suits against polluters. Any lawyer or law firm may act for the general public and sue, being required to distribute a proportion of the amount collected to each member of the included public who claims it from them. (Since different people are differently affected by the same polluting acts, the lawyers might be required to distribute different amounts to those in different specified groups.) The lawyers’ income would come from those who do not write in to claim their due, and from earnings of the money of those who do not claim promptly. Seeing some receiving great income in this way, others would go into business as “public’s agents,” charging a yearly fee to collect and turn over to their clients all the pollution payments to which they were entitled. Since such a scheme gives great advantage to a lawyer who acts fast, it insures that many would be alert to protect the interests of those polluted. Alternative schemes might be devised to allow several to sue simultaneously for distinct sets of persons in the public. It is true that these schemes place great weight on the court system, but they should be as manageable as the operation of any government bureaucracy in determining and distributing costs.*
To arrive at an acceptable principle of compensation, we must delimit the class of actions covered by the claim. Some types of action are generally done, play an important role in people’s lives, and are not forbidden to a person without seriously disadvantaging him. One principle might run: when an action of this type is forbidden to someone because it might cause harm to others and is especially dangerous when he does it, then those who forbid in order to gain increased security for themselves must compensate the person forbidden for the disadvantage they place him under. This principle is meant to cover forbidding the epileptic to drive while excluding the cases of involuntary Russian roulette and the special manufacturing process. The idea is to focus on important activities done by almost all, though some do them more dangerously than others. Almost everyone drives a car, whereas playing Russian roulette or using an especially dangerous manufacturing process is not a normal part of almost everyone’s life.
Unfortunately this approach to the principle places a very great burden on the scheme used to classify actions. The fact that there is one description of a person’s action that distinguishes it from the acts of others does not classify it as unusual and so outside the sphere of application of the principle. Yet it would be too strong to say, on the other hand, that any action falling under some description which almost every other person also instantiates is thereby shown to be usual and to fall within the compass of the principle. For unusual activities also fall under some descriptions that cover actions people normally do. Playing Russian roulette is a more dangerous way of “having fun,” which others are allowed to do; and using the special manufacturing process is a more dangerous way of “earning a living.” Almost any two actions can be construed as the same or different, depending upon whether they fall into the same or different subclasses in the background classification of actions. This possibility of diverse descriptions of actions prevents easy application of the principle as stated.
If these questions could be clarified satisfactorily, we might wish to extend the principle to cover some unusual actions. If using the dangerous process is the only way that person can earn a living (and if playing Russian roulette on another with a gun of 100,000 chambers is the only way that person can have any enjoyment at all—I grant these are both extravagant suppositions), then perhaps this person should be compensated for the prohibition. By having the only way he can earn a living forbidden to him, he is disadvantaged as compared to the normal situation, whereas someone is not disadvantaged relative to the normal situation by having his most profitable alternative forbidden to him. A disadvantage as compared to the normal situation differs from being made worse off than one otherwise would be. One might use a theory of disadvantage, if one had it, in order to formulate a “Principle of Compensation”: those who are disadvantaged by being forbidden to do actions that only might harm others must be compensated for these disadvantages foisted upon them in order to provide security for the others. If people’s increased security from a contemplated prohibition would benefit them less than those prohibited would be disadvantaged, then potential prohibitors will be unable or unwilling to make sufficiently great compensatory payments; so the prohibition, as is proper in this case, will not be imposed.
The principle of compensation covers the cases falling under our earlier statement which involved messy problems about classifying actions. It does not avoid completely similar questions concerning the circumstances under which someone is especially disadvantaged. But as they arise here, the questions are easier to handle. For example, is the manufacturer who is prevented from pursuing his best alternative (though having other profitable alternatives) especially disadvantaged if everyone else may pursue their best alternatives, which happen not to be dangerous? Clearly not.
The principle of compensation requires that people be compensated for having certain risky activities prohibited to them. It might be objected that either you have the right to forbid these people’s risky activities or you don’t. If you do, you needn’t compensate the people for doing to them what you have a right to do; and if you don’t, then rather than formulating a policy of compensating people for your unrightful forbidding, you ought simply to stop it. In neither case does the appropriate course seem to be to forbid and then compensate. But the dilemma, “either you have a right to forbid it so you needn’t compensate, or you don’t have a right to forbid it so you should stop,” is too short. It may be that you do have a right to forbid an action but only provided you compensate those to whom it is forbidden.
How can this be? Is this situation one of those discussed earlier, in which a border crossing is permitted provided that compensation is paid? If so, there would be some boundary line that delimits forbidding people to do certain risky acts, which it would be permissible to cross if the party trespassed upon were compensated. Even if so, since in the cases under discussion we can identify in advance the particular persons being forbidden, why are we not required instead to negotiate a contract with them whereby they agree not to do the risky act in question? Why wouldn’t we have to offer them an incentive, or hire them, or bribe them to refrain from doing the act? In our earlier discussion of border crossing we noted the absence of any compelling theory of just price or compelling reason why all of the benefits of voluntary exchange should go to one of the parties. Which of the admissible points on the contract curve was to be selected, we said, was a question appropriately left to the parties involved. This consideration favored prior negotiation over posterior payment of full compensation. In the present subclass of cases, however, it does seem appropriate uniformly to select one extremity of the contract curve. Unlike exchanges in which both parties benefit and it is unclear how these benefits are to be divided, in negotiations over one party’s abstaining from an action that will or might endanger another person, all the first party need receive is full compensation. (The payment the first party could negotiate for abstaining, were he allowed to perform the action, is not part of his loss due to the prohibition for which he must be compensated.)
If I buy a good or service from you, I benefit from your activity; I am better off due to it, better off than if your activity wasn’t done or you didn’t exist at all. (Ignore the complication that someone once might sell a bona fide good to another person he generally harms.) Whereas if I pay you for not harming me, I gain nothing from you that I wouldn’t possess if either you didn’t exist at all or existed without having anything to do with me. (This comparison wouldn’t do if I deserved to be harmed by you.) Roughly, productive activities are those that make purchasers better off than if the seller had nothing at all to do with them. More precisely, this provides a necessary condition for an unproductive activity, but not a sufficient condition. If your next-door neighbor plans to erect a certain structure on his land, which he has a right to do, you might be better off if he didn’t exist at all. (No one else would choose to erect that monstrosity.) Yet purchasing his abstention from proceeding with his plans will be a productive exchange.16 Suppose, however, that the neighbor has no desire to erect the structure on the land; he formulates his plan and informs you of it solely in order to sell you his abstention from it. Such an exchange would not be a productive one; it merely gives you relief from something that would not threaten if not for the possibility of an exchange to get relief from it. The point generalizes to the case where the neighbor’s desire does not focus only upon you. He may formulate the plan and peddle his abstention around to several neighbors. Whoever purchases it will be “served” unproductively. That such exchanges are not productive ones, and do not benefit each party, is shown by the fact that if they were impossible or forceably prohibited so that everyone knew they couldn’t be done, one of the parties to the potential exchange would be no worse off. A strange kind of productive exchange it would be whose forbidding leaves one party no worse off! (The party who does not give up anything for the abstention, or need not because the neighbor has no other motive to proceed with the action, is left better off.) Though people value a blackmailer’s silence, and pay for it, his being silent is not a productive activity. His victims would be as well off if the blackmailer did not exist at all, and so wasn’t threatening them.* And they would be no worse off if the exchange were known to be absolutely impossible. On the view we take here, a seller of such silence could legitimately charge only for what he forgoes by silence. What he forgoes does not include the payment he could have received to abstain from revealing his information, though it does include the payments others would make to him to reveal the information. So someone writing a book, whose research comes across information about another person which would help sales if included in the book, may charge another who desires that this information be kept secret (including the person who is the subject of the information) for refraining from including the information in the book. He may charge an amount of money equal to his expected difference in royalties between the book containing this information and the book without it; he may not charge the best price he could get from the purchaser of his silence.* Protective services are productive and benefit their recipient whereas the “protection racket” is not productive. Being sold the racketeers’ mere abstention from harming you makes your situation no better than if they had nothing to do with you at all.
Our earlier discussion of dividing the benefits of voluntary exchange, thus, should be narrowed so as to apply only to those exchanges where both parties do benefit in the sense of being the recipients of productive activities. Where one of the parties does not so benefit and is unproductively “served,” it is fair that he merely barely compensates the other, if any compensation is due the other party at all. What of those cases where only the first condition of unproductive exchange is satisfied, not the second: X is no better off as a result of the exchange than if Y didn’t exist at all, but Y does have some motive other than selling abstention. If from Y’s abstention from an activity X gains only a lessened probability of having his own border crossed (a crossing whose intentional performance is prohibited), then Y need be compensated only for the disadvantages imposed upon him by the prohibition of only those activities whose risk is serious enough to justify prohibition in this manner.
We have rejected the view that the prohibition of risky activities is illegitimate, that through prior agreements and open negotiations people must be induced to agree voluntarily to refrain from the activities. But we should not construe our case merely as compensation for crossing a border that protects another’s risky action, with the requirement of prior negotiation obviated by the special nature of the case (it doesn’t involve any productive exchange). For this does not explain why all are not returned to the indifference curve they would occupy were it not for the prohibition; only those disadvantaged by a prohibition are to be compensated, and they are to be compensated only for their disadvantages. If a prohibition of risky acts had two separate effects on someone, the first making him worse off though not disadvantaged as compared to others and the second disadvantaging him, the principle of compensation would require compensation to be paid only for the second. Unlike an ordinary border crossing, the compensation in these cases need not raise the person to the position he was in before he was interfered with. In order to view the compensation under the principle of compensation as ordinary compensation for a border crossing, one might try to redefine or relocate the border so that it is crossed only when someone is disadvantaged. But it is more perspicuous not to distort our view of this compensation situation by assimilating it to another one.
That it is not to be assimilated to the border-crossing sort of compensation situation does not, of course, foreclose deriving the principle of compensation from deeper principles. For our purposes in this essay we need not do this; nor need we state the principle exactly. We need only claim the correctness of some principles, such as the principle of compensation, requiring those imposing a prohibition on risky activities to compensate those disadvantaged through having these risky activities prohibited to them. I am not completely comfortable presenting and later using a principle whose details have not been worked out fully, even though the undeveloped aspects of the principle do not appear to be relevant to the issues upon which we shall wield it. With some justice, I think, I could claim that it is all right as a beginning to leave a principle in a somewhat fuzzy state; the primary question is whether something like it will do. This claim, however, would meet a frosty reception from those many proponents of another principle scrutinized in the next chapter, if they knew how much harder I shall be on their principle than I am here on mine. Fortunately, they don’t know that yet.
* The possibility of surrounding an individual presents a difficulty for a libertarian theory that contemplates private ownership of all roads and streets, with no public ways of access. A person might trap another by purchasing the land around him, leaving no way to leave without trespass. It won’t do to say that an individual shouldn’t go to or be in a place without having acquired from adjacent owners the right to pass through and exit. Even if we leave aside questions about the desirability of a system that allows someone who has neglected to purchase exit rights to be trapped in a single place, though he has done no punishable wrong, by a malicious and wealthy enemy (perhaps the president of the corporation that owns all the local regular thoroughfares), there remains the question of “exit to where?” Whatever provisions he has made, anyone can be surrounded by enemies who cast their nets widely enough. The adequacy of libertarian theory cannot depend upon technological devices being available, such as helicopters able to lift straight up above the height of private airspace in order to transport him away without trespass. We handle this issue by the proviso on transfers and exchanges in Chapter 7.
† Lacking other avenues of redress, one may trespass on another’s land to get what one is due from him or to give him what he deserves, provided that he refuses to pay or to make himself easily available for punishment. B does not violate A’s property rights in his wallet by touching it, or by opening its seal if A refuses to do so, in the course of extracting money A owes him yet refuses to pay or transfer over; A must pay what he owes; if A refuses to place it in B’s possession, as a means to maintaining his rights, B may do things he otherwise would not be entitled to do. Thus the quality of Portia’s reasoning is as strained in holding that Shylock is entitled to take exactly one pound of flesh but not to shed a drop of Antonio’s blood as is the quality of her mercy as she cooperates in requiring that to save his life Shylock must convert to Christianity and dispose of his property in a way hateful to him.
* This sufficient condition for prohibiting or forbidding an action is not a necessary one. An action may be forbidden without there being any provision for its victims to be fully or at all compensated. Our purposes here do not require a general account of forbidding and prohibiting.
† When is a person to be indifferent between the two situations—the time at which compensation is paid (which would encourage boundary crossing, since time heals wounds), or the time of the original act?
* We also pass over whether the retribution includes a component representing the wrongness of the act it responds to. Those retributive theories that hold the punishment somehow should match the crime face a dilemma: either punishment fails to match the wrongness of the crime and so doesn’t retribute fully, or it matches the wrongness of the crime and so is unjustified.
* Recall that C + D + E + R measures the agent’s loss as compared to his initial position, not as compared to his position after gaining from the other party by inflicting damage upon him. We ignore here the question of whether the cost imposed shouldn’t be C + D + 2E + R, with the second E deserved for attempting to impose a cost of fruitless search upon the apparatus of detection and apprehension; or rather whether the R in C + D + E + R shouldn’t also contain this second E as a component.
* We should note the interesting possibility that contemporary governments might make penalties (in addition to compensation) monetary, and use them to finance various government activities. Perhaps some resources left to spend would be yielded by the retributive penalties in addition to compensation, and by the extra penalties needed to deter because of less than certain apprehension. Since the victims of the crimes of those people apprehended are fully compensated, it is not clear that the remaining funds (especially those yielded by application of the retributive theory) must go toward compensating the victims of uncaught criminals. Presumably a protective association would use such funds to reduce the price of its services.
* An interesting discussion of these diverse issues is contained in George P. Fletcher, “Proportionality and the Psychotic Aggressor,” Israel Law Review, Vol. 8, No. 3, July 1973, pp. 367—390. Despite Fletcher’s claim that there is no way to say both that one may use deadly force in self-defense against a psychotic aggressor (whose r = o) and that we are subject to some rule of proportionality, I believe our structure presented in the text yields both these results and satisfies the diverse conditions one wants to impose.
* One may be tempted to delimit partially the area where full compensation is permissible by distinguishing between using something as a resource in a productive process and damaging something as a side effect in a process. Paying only full compensation would be viewed as permissible in the latter case, and market prices as desirable in the former, because of the issue of dividing the benefits of economic exchange. This approach won’t do, for dumping grounds for effects are also priceable and marketable resources.
† A similar problem arises with economists’ usual explanation of exchange. Earlier views had held that there must be equality in something or other between goods that persons are willing mutually to exchange. For otherwise, it was thought, one party would be the loser. In reply economists point out that mutually advantageous exchange requires only opposed preferences. If one person prefers having the other’s good to having his own, and similarly the other person prefers having the first’s good to having his own, then an exchange may benefit both. Neither will lose, even though there is nothing in which their goods are equal. One might object that opposed preferences aren’t necessary (even apart from questions about whether exchanges might not take place between parties indifferent between two commodities, or might not advantageously take place between two persons with identical preferences and identical initial mixed holdings of two goods when each person prefers either unmixed holding to any mixed one and each is indifferent between the two unmixed holdings). For example, in three-way baseball trades one team may trade away a player for another they prefer having less than the one they trade away, in order to trade this other player to yet another team for a third player they prefer having more than the first. It might be replied that since the first team knows that the second player can be traded for the third, they do prefer having the second (who is easily transformable into the third player, via exchange) to having the first player. Thus, the reply continues, the team’s first exchange is not for a less preferred object, nor does this exchange move the team to a lower indifference curve. The general principle would be that anyone who knows that one good is transformable into another (via exchange or in any other way) preferentially ranks the first at least as high as the second. (Omitting costs of transformation does not affect the point at issue.) But this principle, apparently necessary to explain simple three-way exchanges, conflicts with the earlier explanation of exchange in terms of opposed preferences. For this principle has the consequence that a person does not prefer having another’s good to having his own. For his own can be transformed into the other (via the exchange to be explained), and so he preferentially ranks it at least as high as the other.
The various routes out of this difficulty that suggest themselves and that survive cursory examination (remember that two different parties each can offer a commodity to someone for his) all seem to involve complicated and involuted bundles of subjunctives and counterfactuals.
* Note that not every act that produces lower utility for others generally may be forbidden; it must cross the boundary of others’ rights for the question of its prohibition even to arise. Note also that no such considerations of fear apply to a system of allowing any acts that have the prior consent of the person whose boundary is crossed. Anyone who worries that under such a system he foolishly might consent to something can ensure that he won’t, via voluntary means (contracts, and so on); secondly, others cannot reasonably be restricted to counteract a person’s fear of himself!
* An act risking a possible consequence might not produce fear, even though it would if known for certain to have that consequence, if the lessened probability dissipates the fear.
* One might plausibly argue that beginning with probabilities that may vary continuously and asking that some line be drawn misconstrues the problem and almost guarantees that any position of the line (other than 0 or 1) will appear arbitrary. An alternative procedure would begin with considerations “perpendicular” to those about probabilities, theoretically developing them into an answer to the questions about risky actions. Two types of theories could be developed. A theory could specify where a line is to be drawn without this position’s seeming arbitrary, because though the line comes at a place which is not special along the probability dimension, it is distinguished along the different dimensions considered by the theory. Or, a theory could provide criteria for deciding about the risky actions that do not involve drawing a line along the probability (or expected value or some similar) dimension, whereby all the actions falling on one side of the line are treated in one way and all those on the other side in another. The considerations of the theory do not place the actions in the same order effected by the probability dimension, nor does the theory partition actions into equivalence classes coextensive with some interval partition of the unit line. The considerations the theory adduces merely treat the question differently, and so have the consequence that some act is forbidden while another with a higher expected value of harm is permitted. Unfortunately, no satisfactory specific alternative theory of either type has yet been produced.
* Instead of compensating them, can the agent supply tranquilizers to all those upon whom the risk is imposed, so that they won’t feel very afraid? Should they have to tranquilize themselves, so that it’s not the agent’s concern at all if they neglect to do so and feel feat? For an illuminating initial tangling of such issues see Ronald Coase, “The Problem of Social Costs,” Journal of Law and Economics, 1960, pp. 1–44.
* The proposal I make here can, I think, be defended against the considerations adduced in Frank Michelman’s sophisticated presentation of a contrasting view in his “Pollution as a Tort,” an essay review of Guido Calabresi’s The Costs of Accidents, in Yale Law Journal, 80 (1917), pt. V, 666–683.
I do not mean to put forth the above scheme as the solution to controlling pollution. Rather, I wish merely to suggest and make plausible the view that some institutional arrangement might be devised to solve the problem at a fell swoop, and to commend the task to those clever at such things. (J. H. Dales proposes, in Pollution, Property, and Prices, to sell transferable rights to pollute in specified amounts. This elegant proposal unfortunately involves central decision as to the desirable total amount of pollution.)
Popular discussions often run pollution problems together with that of conserving natural resources. Again, the clearest examples of misdirected activity have occurred where there are no clear private property rights: on public lands denuded by timber companies and in oil fields under separately held pieces of land. To the extent that future people (or we later) will be willing to pay for the satisfaction of their desires, including trips through unspoiled forests and wilderness land, it will be in the economic interests of some to conserve the necessary resources. See the discussion in Rothbard, Power and Market (Menlo Park, Calif.: Institute for Humane Studies, 1970), pp. 47–52, and in the references he cites.
* But if he didn’t exist, mightn’t another have stumbled on the unique piece of information and asked a higher price for silence? If this would have occurred, isn’t the victim better off because his actual blackmailer exists? To state the point exactly in order to exclude such complications is not worth the effort it would require.
* A writer, or other person, who delights in revealing secrets, may charge differently. This consideration does not help the racketeer discussed below, even if he is sadistic and enjoys his work. The activity he threatens is excluded by moral constraints and is prohibited independently of whether it, or abstaining from it, is charged for. The example of the writer is taken from footnote 34 of my essay, “Coercion,” in Philosophy, Science, and Method: Essays in Honor of Ernest Nagel, ed. S. Morgenbesser, P. Suppes, and M. White (New York: St. Martin’s Press 1969), pp. 440–472. Contrast our view of blackmail with the following, which sees it as on a par with any other economic transaction: “Blackmail would not be illegal in the free society. For blackmail is the receipt of money in exchange for the service of not publicizing certain information about the other person. No violence or threat of violence to person or property is involved.” Murray N. Rothbard, Man, Economy, and State, vol. 1, p. 443, n. 49.