Journal of Value Inquiry 34 (2/3): 287-297 (2000)
Libertarians typically object to the state’s dealing with law and order for several general reasons: it is inefficient; it is carried out at the taxpayers’ expense; it punishes ‘victimless crimes’.[1] Exactly what the observance of liberty implies with respect to the treatment of tortfeasors and criminals is more controversial among libertarians. A general theory of libertarian restitution[2] and retribution is mainly what is attempted here. However, this general theory alone raises some practical problems that require some immediate response if the theory is to appear at all plausible. The main assumption is the hypothetical one that if liberty is to be observed as far as possible, then the imposition of a cost on (that is, the infraction of the liberty of) another person calls for libertarian rectification[3] of some kind, as far as possible. What follows attempts both to derive and defend the consequences of this assumption. I also assume that the internalizing of illiberal externalities which this conception of liberty entails, is the best long-term strategy for maximizing welfare (in the sense of want-satisfaction); so I deal with any apparent problems with this that are peculiar to libertarian rectification. I assume throughout, to avoid unnecessary complications, that people have an initial libertarian claim to their persons and property (that is, that there is no imposition on others involved in these claims) before any impositions occur. I shall first outline how I interpret the general libertarian account and then look at two important problem areas.
i. ‘Torts’ and ‘Crimes’
At present, a ‘tort’ is, roughly, a ‘wrong, or harm, that is actionable in law for restitution’. A ‘crime’ is, roughly, an ‘act punishable by law’. I propose stipulatively to define these in ways that I shall argue are libertarian (though not standardly so). My stipulative definitions of ‘tort’ and ‘crime’ are utterly different from the usual legal ones; but the conceptual links to the originals terms, and the libertarian purpose behind this translation, should become clear. A ‘tort’ becomes an ‘imposed cost, for which restitution is due’, and a ‘crime’ becomes a ‘foreseen imposed cost, for which restitution or “retributive restitution”[4] is due’. Unlike the present system, all ‘crimes’ (in my sense) entail financial restitution (if that is preferred to physical retribution by the victim or his assigns). So it will be convenient to break the discussion into what I shall call ‘non-criminal torts’ (concerning unforeseen impositions) and ‘crimes’. I hope this will not be too confusing.
First, the case of unforeseen impositions, which I call ‘non-criminal torts’, that the imposer admits to (to fail to reveal such a tort, such as after a car accident without witnesses, is itself to go beyond a mere tort and become a ‘crime’). Although such torts are unforeseen, there must always be an element of imposition: a pure accident would not be tortious (such as falling on someone as a result of a freak gust of wind blowing one over a bridge: it was completely unforeseeable). Such pure accidents are highly rare. The imposing factor in non-criminal torts is usually an element of negligence (though it could be engaging in a normally safe activity, such as flying, that just happens to damage others). And there is a continuum of increasing negligence between non-criminal torts and crimes that passes through criminal negligence.[5] If liberty is to exist, then (to the extent that it is a one-sided non-criminal tort without contributory negligence on the part of the injured party) it will normally suffice that the injured party receive restitution amounting to the value of the damage (including any indignity that might be caused by being imposed on negligently) plus any other costs involved. I write ‘value’ because it is not necessary, and might not be possible, somehow to make things physically as good as they were. The imposition is ultimately psychological (but in flouting our preferences when we are not imposing on others thereby, not in being mental states of disutility) and that is what has to be rectified. In fact, even though no accident ensues, ‘merely’ risky activities (which might, after all, be significant risks of severe damage) are tortious to the extent of the imposed risk (which is, roughly, how much one would have to be paid ex ante to bear such a risk).[6] Not to allow the enforcement of restitution for non-criminal torts is effectively to allow people mistakenly to impose costs on others. That cannot be libertarian.
Second, consider where the imposition is clearly foreseen (though it can be foreseen and yet be an unintended and unwanted by-product of some chosen activity – such as continuing to sell goods after one has discovered them to be unsafe). The foreseen imposition is a ‘crime’, in the only libertarian sense of that word.[7] The imposing party has effectively treated another’s person or goods as though they were his own. If liberty is to be observed, then as far as possible the imposer must give his victim the choice of financial restitution (now including an extra amount for the side-effects of its being foreseen: the indignity, and so on) or an equally-valued (to the imposer) claim against the imposer’s person or goods. Unless the criminal also gives the choice of such equally-valued claims against his person or goods, we are left with the view that financial restitution can always be sufficient for maximum rectification.[8] This would inadequately reflect the crucial distinction between what I have called ‘non-criminal torts’ and ‘crimes’: between mistakenly imposing and wittingly imposing. Wittingly imposing gives rise to a qualitatively different kind of imposed cost, because its foreseen and usually deliberate nature often gives rise to indignity and even fear. In fact, one imposes such a cost even if one unsuccessfully attempts a crime, analogously with those torts that impose a ‘mere’ risk.[9] Suppose that someone tries to shoot you dead but misses. Surely the very attempt is an imposed cost: it is your life and you disvalue attempts to end it without your permission (whether you discover them or not, given that costs relate to preferences and not to mental end-states). To attempt murder but fail due to a pure accident means that one creates a claim to a similar attempt against oneself (that is, with a statistically similar chance of success) by one’s victim or his assigns.[10] If exercised, that is almost certainly to cede one’s life to the victim.[11] To allow the enforcement of only, even if punitive, financial restitution for crimes is effectively to allow people to pay money to impose foreseeable costs on others; to allow the compulsory purchase of crimes. That cannot be libertarian.
So assuming the enforcement of liberty and no attempt to evade capture, which I deal with later, for you knowingly to take or damage another’s goods is to create the claim that he take or damage your goods up to the same value. Knowingly to damage, violate, or use another’s person is to create the claim that your victim have some physical damage, violation, or use done to you up to the pain, inconvenience, and indignity that you imposed. This is the libertarian and civilizing rule of lex talionis, the law of retaliation.[12] More precisely and libertarianly interpreted, lex talionis does not imply ‘an eye for an eye’ but rather ‘an equal imposition for an imposition’: the loss of a blind eye or the eye of a person soon to die of cancer, and so on, is not of equal disvalue to the loss of the typical eye (it seems that the ancient Jews may well have interpreted lex talionis this way).[13] To calculate the disvalue it is necessary to have an approximate interpersonal comparison of disutility (see 4.2), probably comparing the likely values of ‘reasonable men’.[14]
With foreseen damage to goods and persons, the imposition-rectifying claims – if exercised by the owner or his assigns – will amount to libertarian retribution or punishment (whether or not there is an attendant intention to cause suffering). This is so even though the claims might all be taken in the form of money for the imposed damage, including the indignity and fear, or money negotiated not to perform an appropriate retributive act (if that is a greater sum). Both the latter financial settlements will still amount to retribution because the option of personal retribution exists (so the criminal cannot simply choose to ‘buy crime’) and because the criminal pays a greater sum than with a non-criminal tort merely for the foreseen aspect (to the extent that this causes indignity, fear, and so forth). In the case of their goods, I guess that many people would opt for a financial settlement that added the imposed cost of any indignity suffered owing to the foreseen, and perhaps deliberate, nature of the imposition. In the case of their persons, I guess that many people would want at least some personal retribution: partly to symbolize the greater disapproval of the act and partly out of vengeance.
Full financial restitution is often not possible in extreme cases, such as murder or maiming, in the sense that the victim would almost always be better off if it had not happened. But receiving as much financial restitution as possible, with any balance as physical retribution, is clearly more libertarian than the victim’s receiving anything less. And so is the victim’s opting for full physical retribution and no financial restitution, or any ratio of the two, or simply letting the criminal go if that is what he or she prefers.[15] One cannot have full financial restitution and full physical retribution: that would be double counting the same imposed cost.[16]
Those who are tried without their agreement, by a court they have no contract to recognize, would clearly be having a cost imposed on them if the judgement went mistakenly against them. But if people contract in, then as long as the judicial procedure takes place as agreed, any outcome will ultimately be libertarian – at least in the sense that everyone has voluntarily accepted the system in the first place and thereby bound themselves.[17] An unavoidably imperfect libertarian system must take the chance of imposing on, or at least contractually convicting, some innocent people in order to deter even greater impositions on the innocent by criminals. This is certainly not using the innocent for ‘libertarian consequentialist’ reasons, any more than is engaging in many other activities with the risk of negative externalities, such as driving a car. It will be a regrettable mistake whenever a false judgement occurs. To please its customers any private system will probably offer compensation for loss, to the extent that this is possible, when errors are discovered. So it seems that all processes of prosecution are likely to be made very safe given the large restitution that would likely follow a mistake with even a contractual conviction, and given the contingency-fee lawyers who would therefore have a great incentive to expose such cases.
Although defending a separate conception of ‘crime’ – which many libertarians think is inherently statist or otherwise unlibertarian – I see no sound libertarian reason to defend any aprioristic distinction in standards of evidence for non-criminal tort and crime. The libertarian degree of evidence that would be necessary for both has to be whatever minimizes the infractions of freedom (imposed costs). However, if, as seems possible, it is contingently the case that mistaken criminal convictions impose far greater costs (or, if contractual, are at least less preference utilitarian) than mistakenly not convicting, and this gap is larger than the same one with non-criminal torts,[18] then it is contingently libertarian (or preference utilitarian, at least) to prefer higher standards of evidence for criminal convictions. The market is likely to take account of these contingencies: profits will reflect people’s choices and views, as people can contract into their individually preferred systems of adjudication. Technology might also change the optimal (imposed-cost minimizing) standards for evidence.
We shall now look at two major problem areas with this general account: first briefly, as there are fewer purely conceptual or new issues here, the effect of differing levels of wealth; then the crucial idea of risk-multiplier rectification.
ii. Can Wealth Levels Distort Rectification?
Pure financial restitution is indeed rather like ‘buying crime’ by paying afterwards, or possibly even before: ‘I am going to give you this large check and then break your nose, and there is nothing you can do about it except cash the check’. Given that someone has little money, it seems that it might sometimes be worth it for a rich man to impose on him and then pay mere financial restitution. But with libertarian retribution the criminal creates a claim against himself to treatment as severe as he imposed on his victim. With wealthy people this would typically be something that they are prepared to pay more to avoid than most people could afford. This means that the victim is put in a better position with wealthy criminals. He will often, if he wishes, be able to have some ratio of financial restitution and physical retribution without suffering financial hardship. Does this, then, discriminate against the wealthy? I do not see how. If they wish, the wealthy can decline to buy off the physical retribution and so fully suffer it, as the less wealthy might be obliged to do (or, perhaps, pay only the court-ordered ‘retributive restitution’ if the victim was bluffing).
What of the empirical charge that the rich will simply buy the best lawyers and thereby get away with crimes?[19] This is not the place to go into the details of the likely operation of free-market law and order,[20] but there seems no reason to suppose that in a free-market society lawyers will not often wish to take cases on the basis of contingency fees. And because of the possibility of making them pay much more to avoid retribution, it will generally be better to win against a rich person than a poor one. For those who worry that private institutions are inherently more open to corruption than state ones, consider private banks: they do not habitually steal or counterfeit money (despite inefficient state policing and punishment[21]), as those with even less understanding of the market might suppose – but the coercive state effectively does both. Competition in the provision of law and order, consumer scrutiny, full retribution or restitution, that a lawyer who knowingly defends a guilty person is guilty of fraud and liable to prosecution himself, that previous offenses may well be revealed in court and never wiped from the records, … all will go to ensure higher standards than are currently found, or likely to be introduced, with the normal state monopoly in this area.
On the other hand, how will the impecunious pay restitution? I guess, as do many other libertarians, that it will usually be a case of the victim’s insurance company paying the money to the victim and then attempting to recover the purchased debt from the tortfeasor or criminal. Such a debt could always be sold for immediate restitution, even if there were no insurance. If a tortfeasor or criminal then has extreme debts and cannot easily work productively enough to pay them off, then the free market is bound to be inventive; and competition will increase the imposer’s productivity while minimizing his cost. At the extreme, for instance, there is the possibility of selling his bodily organs or putting him to gladiatorial combat on pay-television. This might sound rather awful but then he will probably have done something rather awful to owe this much. Moreover, he is only really not being allowed to get away with imposing the cost of his actions on others. Those who seek damages in this way need not defend, intend, or use any theory of punishment. They can accurately assert that they are merely seeking libertarian rectification for what they are owed; although, as mentioned earlier, from the viewpoint of the criminal’s experience he will in effect be being punished in libertarian proportion to his crime.
Generally, the libertarian system would be far more humane than the state system. In particular, the libertarian system would almost entirely avoid the current state system of long prison sentences for relatively petty offenses. Only a tiny minority, consisting of those who are a continuing threat to others in ways that cannot fully be compensated for, would bring long-term incarceration upon themselves (and maybe only then if electronic tagging is not adequate): mainly seriously violent criminals. Even for such criminals the incarceration itself, unless involved with some extra aspect of retribution, may be as pleasant as they can afford to pay for (for only the criminals themselves must bear the cost of what they have made necessary to protect liberty).[22]
iii. Risk-Multiplier Rectification
With most crimes it might even be wondered whether libertarian rectification could offer enough deterrence to maximize liberty and welfare. Surely it would often pay to commit a crime of, say, theft if one usually gets away with it and only faces little more than paying back what one stole. I cannot see by what clear libertarian principle one can simply add on all the costs of security devices, policing, and so on. For if this policy were taken seriously, then there would be no upper limit to passing on such costs for even the most trivial of offenses. There must come a point where the criminal himself is being imposed on by going beyond any claim against him he might have created by his own criminal behavior.[23] At this point the role of criminal and victim becomes reversed. But where is this point if it goes beyond any apparent damage done by the act itself? There is, it seems to me, a fully libertarian way of determining, at least in principle, the correct degree of restitution and retribution in such cases. This is where Mane Hajdin’s (1987) gambling theory of crime comes in; though, contra Hajdin’s interpretation, I apply it individualistically and also to retribution.
When one imposes a cost on another the full cost to him need not, despite superficial appearances, be merely the obvious damage done by the act itself (including indignity, anxiety, and so on), as Hajdin thinks (1987, 81). The full cost must also take account of the likelihood that the criminal might have got away with the crime. This is another kind of imposed risk. If a criminal imposes ten ducats worth of damage (including indignity, anxiety, and so on) with a one in ten chance of capture, then that is to impose a cost in excess of ten ducats in the sense that it would probably not pay the average victim to agree beforehand to such a risk unless he were sure of at least one hundred ducats if the thief were caught (real attitudes to risk would need to be empirically determined). So the full imposition, in monetary terms, is probably in excess of one hundred ducats. The full imposition is the sum one would have to be paid ex-ante to accept such a risk. For want of a better expression, I call this ‘risk-multiplier rectification’. The precise amount will vary with the circumstances of the victim. Some ‘reasonable man’ interpretation of the imposition will have to be made to rule out inflated claims. Such proportionality is clearly related to liberty, in the sense of rectifying imposed costs, in a way that Rothbard’s ‘double the extent of theft’ is not.[24] Such doubling is an unlibertarian limit as far as I can see: it will occasionally be too much and very often too little for full libertarian rectification.
Several problems with risk-multiplier rectification readily spring to mind. I cannot begin to tackle anything like an exhaustive list. I shall deal with a few that strike me as the most pressing.
One problem concerns the likelihood of capturing the criminal. If determinism is true, then in a sense the criminal is either bound to be caught or bound to escape, and that cannot do the job required here.[25] If a purely subjective estimate of probability were allowed then that would be too indeterminate and subject to biased interpretations. It seems necessary to have some kind of statistical approach based on the number of such crimes where the criminal is caught. The criminal himself will typically be operating with some such statistical likelihood in mind. We cannot have the certainty of capture, but at least we can have the well-advertised certainty that crime will be a ‘bad bet’ for criminals (which is apparently not the case with many offenders facing only state punishments and capture rates).
Should other victims automatically receive a share of the risk-multiplier restitution? No: once full financial restitution (if that is preferred to any physical retribution) is paid to the individual victim, then there will be no ‘extra money’ for distribution to other victims of crime.[26] It is true that there are imposed externalities of crime that affect the general public; most obviously, the uninsured victims where the criminals escape rectification. As there is no proof that a convicted criminal also committed unsolved crimes, there is no libertarian reason that he should compensate those victims (even if he did not already owe, as he does owe, a libertarian debt for all the risk-multiplier to his actual proven victims or their assigns). However, such victims can minimize the imposition by insuring themselves. Uncompensatable crimes aside, all they will then suffer is roughly the price of the insurance and security measures. In this case the insurance company is then owed the risk-multiplier whenever the criminal is caught, for it will have bought the debt.[27] As it will use this money to pay its claims, this will spread the risk-multiplier money around to all insured victims in a similar way to that which Hajdin wants, though on a completely libertarian basis. But I do not see how it can be libertarian or preference utilitarian (in practice, rather than as a mere logical possibility) to disallow people the free choice of not paying insurance (and thereby 1. saving money, or 2. losing from crime, or 3. gaining all the risk-multiplier for themselves).[28]
Are crimes not also against society? I admit that insurance and security measures are effectively an imposed cost by all criminals on all who pay for them. I do not see that this means that particular crimes are really against ‘society’.[29] Considered in himself, the single criminal cannot constitute a significant enough threat to make his own share (of any such restitution due to everyone) a collectable amount. Nor should we overlook the probability that the criminal suffers very similar costs and concerns, regarding crimes against himself, to those which most people do. Therefore the vanishingly small amount we are considering here is, in any case, more or less cancelled out. Given that the captured criminal must pay the actual victims or their assigns the amount – or suffer the retribution – that fully internalizes illiberal externalities between them, I see no serious reason to doubt that this will be optimally libertarian and preference utilitarian. While crime exists, though, it will be impossible to have perfect libertarian rectification such that all victims are no worse off.[30]
What if, after any victim’s choice of a restitution-retribution ratio, a dangerous criminal is still at large? If he is so likely to re-offend that this risk itself imposes more of a cost on the general public than his incarceration would impose on him, then we do not need to wait for the next victim: a class action for the imposed risk to the public might imprison him on libertarian grounds. It is somewhat like banning home-made nuclear devices, except that the person makes himself the danger. Until this risk disappears he has to be electronically tagged or, at the extreme, incarcerated or sent to some escape-proof wilderness. Despite appearances, this is not a punishment (if it were then alternative ‘suffering’ or restitution might replace it) but, rather, the denial of the benefits of free association due to being too great a danger to others. I guess that professional expertise will be required to calculate just how long incarceration or transportation needs to be.
Granted the risk-multiplier, might violent criminals be liable to suffer many times the physical imposition they are convicted of if that is what their victims choose? I do not see why not. After all, violent criminals will on average get away with similar crimes just that many times without being caught. Given the likely involvement of insurance companies, though, it seems that receiving money is going to be generally preferred to intentionally inflicting suffering.[31] Then is there a danger that great restitution or retribution could cause criminals to be more dangerous? Might a violent criminal with little to lose choose to ‘destroy the evidence’ – you? Or faced with such severe punishments, perhaps the jury – if it is a (private) jury system – will be reluctant to convict. I should say first that it would not be unlibertarian if, in order to stop worse crimes from being committed, people were to choose private judicial systems that limited their own libertarian claims to restitution or retribution. As admitted already, I am far from asserting that in practice illiberal externalities can be perfectly internalized even using libertarian principles with respect to ‘crime’. On the other hand, it might well become apparent from market competition that full risk-multiplier rectification is an effective incentive for the criminal to avoid such crimes, or to turn himself in so that he can avoid the risk-multiplier (plus any interest that would be owed on the overall debt), or at least to avoid doing more imposing than a crime ‘requires’. So the chosen judicial systems, whatever they are, are likely to be both libertarian and preference utilitarian as far as possible.
We should also consider the opposite problem: will there be optimal libertarian and preference-utilitarian deterrence by restitution and retribution where there is no attempt to escape? That is, where there is no risk-multiplier can lex talionis be enough? Might someone regularly commit some crime but always turn himself in afterwards in order to minimize what he then owes? As stated earlier, anyone who is a serious threat to others in a way for which they cannot be fully compensated would be liable to long-term incarceration, or at least electronic tagging. Where the threat is a mere nuisance or against (not too expensive) physical property, there is no real problem. With full libertarian restitution people need not worry about such things because, ex hypothesi, they will not lose out. Having said this, it is hard to see why anyone would have an incentive to commit such criminal acts regularly. Full restitution and retribution would mean that the eventual expense to the criminal would almost certainly be far greater than his initial gain: crime is typically a negative-sum game whereby the criminal destroys more welfare than he personally gains.
Finally, what of the occasional possibility that a criminal might make considerable money from his crimes by, for instance, selling his story to a newspaper? Might crime not thereby become a paying proposition that will be encouraged rather than deterred? If the crime is a very serious one, then full libertarian restitution might consume all the money made in this way (no amount of money will usually be enough to compensate for murder). The criminal might even libertarianly be forced to sell his story for this very purpose. With lesser crimes, if people have been fully compensated then it is not clear what libertarian grievance people would have or why this should be a problem. However, to make public intimate details of an identifiable person’s life that one has come by illegitimately (for instance, the details of some crime someone has suffered because one was oneself the criminal responsible) would itself constitute an imposition; and one would be liable to further restitution or retribution (subject to any risk-multiplier, of course). Even an unwitting publisher would be liably imposing by publishing such details. So I see no serious objection to the libertarian approach here either.
I am keenly aware that if this theory of libertarian rectification is even approximately right, then there are myriad questions that have not been addressed. If it is not even approximately right, then I look forward to its refutation and hope thereby that clearly perceiving its errors may still help to discover the libertarian approach.
Notes
[1] Chapter twelve of Rothbard 1973 remains one of the best introductions to this line of argument.
[2] I mean ‘restitution’ in the broad sense that includes compensation, not necessarily putting things back as they were (which might not be possible).
[3] I use ‘rectification’ in only this hypothetical sense (of ‘libertarian correction of an infraction of liberty’) and not any moral sense.
[4] Benson 1996, 77.
[5] As Roger Pilon notes (1978, 355 n13).
[6] However, detecting the risk and finding the risk-victim might often be too difficult to determine in practice.
[7] The view on crime given here differs to varying extents from the accounts that have chiefly inspired it, and perhaps from all of them in that no moral views are used to draw the conclusion. See, for instance, Epstein 1977, Ferrara 1982, Hospers 1977, and Pilon 1978.
[8] Let me put it another way: on what libertarian basis could he complain, if we take a similar action against someone to that which he initially imposed on us?
[9] Contra Barnett (1980, 150), Hajdin (1987, 85), Kleinberg (1980, 277 n9, and 278), and Miller (1978, 359–60).
[10] In the case of torts, by contrast, it is hard to see how one could create the claim that someone could accidentally impose on one, as Pilon observes (1978, 354).
[11] However, where the attempt stood no reasonable chance of success things are quite different. Attempting to kill someone by sticking pins in his effigy cannot be any serious imposition, except for the highly superstitious. I would not pay a penny to stop such ‘attempts’ on my life. In such cases we have really no more than the daydream of the murder of another. Though we might allow proportional ‘effigy retribution’ (sticking pins into the daydream murderer’s effigy or requiring money not to do so) for those who insist on taking such things seriously.
[12] Lex talionis is not introduced to avoid the moral hazard of ‘buying crime’. There just seems no libertarian reason that a victim of a crime should not have an equivalent claim against the perpetrator.
[13] Some might suggest that even calculating such a balance makes this somewhat like a trade, which therefore undermines the proper notion of punishment. But there is no similar balance in trade, where both sides usually gain different amounts. Because lex talionis is obviously a form of balance it is unlike a trade. The crime is never made libertarian after all, but some balance with retribution is possible. In any case, even under an illiberally harsh punishment system the criminal can regard the risk and type of punishment as a ‘price’ in some sense, and respond economically to different such ‘prices’. For those set on some crime, even the certainty of such punishment can still be a ‘price’ they feel worth ‘paying’.
[14] On a related point, what of the degree of resistance that can libertarianly be offered during the course of a crime? This must not be out of proportion to the apparent imposition. To shoot a shoplifter is clearly to go so far beyond his imposition as to become yourself the criminal. We must give the victim some extra leeway though, for the criminal might otherwise escape too easily and one cannot know how far a criminal might be prepared to go with his impositions. It is his choice that he has put others in that situation. It is therefore libertarian to assume, for instance, that an armed robber might shoot to kill and so to shoot him dead before he tries this (or even to assume the worst with any physical attacker). Not to allow the victim this leeway is to give the criminal an illiberal advantage in all situations, which can only promote crime and thereby be unlibertarian.
[15] Pilon has this correct (1978, 356–7) though based on morals rather than a theory of liberty, and without a theory of proportionate libertarian punishment.
[16] I write of ‘financial restitution’ and ‘physical retribution’ rather than simply ‘restitution’ and ‘retribution’ because perhaps libertarian retribution is not fully separable conceptually from restitution but is a part or form of it. For it is surely a form of restitution (though usually imperfect in that the status quo ante was preferable) that the victim comes to own an equally-valued claim to the criminal’s person or goods. For the victim who wants this to be given anything less than this claim, is for him to be less adequately compensated for no libertarian reason.
[17] This might sound mistaken. Of course the innocent person will suffer a great personal cost if he, for instance, receives the death penalty. This would be like flying with an airline when the plane crashes. In both cases someone is dead as a result of a fault in the system he contracted into. The airline does not impose a cost, though he suffers a cost, by mistakenly killing him (unless it was negligent). Neither does a contractual judicial system impose a cost, though he suffers a cost, by mistakenly killing him (unless it was negligent). This might seem more like an ‘imposed’ cost only because in one sense they ‘deliberately’ kill him, but that description is incompatible with the definition of liberty and the analysis of contracts.
[18] If so, this is because of increasing marginal disutility.
[19] For example, Kleinberg 1980, 275 and 278–80.
[20] See Benson 1990.
[21] For the much misunderstood theory and history of more honest and efficient ‘free banking’ (that is, with private note issue and no state regulation), see the relevant writings of Kevin Dowd, George Selgin, and Larry White.
[22] Private prison factories are one way to pay for this (eating being contingent upon working).
[23] What if a shop carried a sign outside saying, ‘Those who enter contract to be shot if caught shoplifting’? Entry to the shop would probably not create a contract. A contract has to be at least implicitly agreed to by all concerned parties, and it is not reasonable to suppose that any shoplifter would agree. Neither would it create a contract if someone were allowed to enter a shop with a t-shirt saying, ‘Shops which allow me in contract to allow my shoplifting.’ However, there is an important asymmetry in that it is the would-be shoplifter who is taking the action of entering the shop, and thereby at least prima facie accepting the shop’s contract.
[24] Rothbard 1977, 263.
[25] So-called chaos theory implies that the universe is infinitely complex (‘infinite complexity’ theory is a more accurate name for it) but determinate. It is of no metaphysical significance that the epistemological problem is thereby greater than hitherto thought.
[26] As Hajdin suggests (1987, 81).
[27] This could bring down the price of crime insurance considerably; it might even approach being free of charge if the insurance company gets enough whenever a criminal is convicted.
[28] What of the moral hazard posed by the possibility of frauds claiming enormous ‘restitution’? (Pilon 1978, 351 n4) Here it must be realized that a false accuser will be liable to pay an even greater amount (roughly, the sum fraudulently claimed multiplied by, for instance, ten if there is a ten per cent chance of detection). This makes such fraud a very bad gamble for most people.
[29] As held by, for instance, Dagger (1980 and 1993), Kleinberg (1980, 277), and Miller (1978).
[30] I initially dismissed the option of passing on all and any policing and security expenses no matter how high they might be. Even though we now have the risk multiplier as a libertarian principle, it might still be thought that we have the problem of how much of the general policing and security costs can be passed on in addition. However, apart from any unforeseen additional expenses which the criminal causes (by violently resisting arrest, say), I suppose that this would be double counting. The ex-ante price that someone would reasonably be prepared to accept to suffer the risk of some crime will include any policing and security expenses that he actually does pay.
[31] What if someone steals only one ducat with a million-to-one chance of capture, but he is then caught? Can it really be libertarian and preference utilitarian to make him pay one million ducats restitution? In a world where such thefts regularly occurred, only such a ratio could be adequate to internalize illiberal externalities. This might not fit our current intuitions as to what is libertarian and preference utilitarian just because such low capture rates are a fantasy. The victim is always at liberty to exercise mercy if he wishes and waive his claim. Other people are also at liberty financially to assist any criminal for whom they feel sympathy. However, they will run a real risk of encouraging crime if they are too soft-hearted.
References
Barnet, Randy E. 1977. ‘Restitution: A New Paradigm of Criminal Justice.’ Ethics 87, no. 4:279–301.
—— 1980. ‘The Justice of Restitution.’ In Moral Issues, ed. J. Narveson. Toronto and New York: Oxford University Press, 1980.
Barnett, Randy E., and John Hagel III, eds. 1977. Assessing the Criminal: Restitution, Retribution, and the Legal Process, Cambridge, Mass.: Ballinger Publishing Co.
Barnett, Randy E., and John Hagel III. 1977. ‘Assessing the Criminal: Restitution, Retribution, and the Legal Process.’ In Barnett and Hagel, 1977.
Benson, Bruce Lowell. 1990. The Enterprise of Law. San Francisco: Pacific Research Institute for Public Policy.
—— 1996. ‘Restitution in Theory and Practice.’ Journal of Libertarian Studies 12, no. 1:75–97.
Dagger, Richard. 1980. ‘Restitution, Punishment, and Debts to Society’. In Victims, Offenders, and Alternative Sanctions, ed. Joe Hudson and Burt Galaway. Lexington, Mass.: Lexington Books, 1980.
—— 1993. ‘Playing Fair with Punishment.’ Ethics 103:473–488.
Epstein, Richard A. 1977. ‘Crime and Tort: Old Wine in Old Bottles.’ In Barnett and Hagel 1977.
Ferrara, Peter, J. 1982. ‘Retribution and Restitution: A Synthesis.’ Journal of Libertarian Studies 6, no. 2:105–136.
Friedman, David D. 1996. ‘Anarchy and Efficient Law.’ In Narveson and Sanders 1996.
Fuller, Lon L. [1964] 1969. The Morality of Law. New Haven: Yale University Press.
Gewirth, Alan. 1984. ‘A Reply to Danto.’ In Paul and Miller 1984.
Hajdin, Mane. 1987. ‘Criminals as Gamblers: A Modified Theory of Pure Restitution.’ Dialogue 26:77–86.
Kleinberg, Stanley S. 1980. ‘Criminal Justice and Private Enterprise.’ Ethics 90:270–282.
Miller, Franklin G. 1978. ‘Restitution and Punishment: A Reply to Barnett.’ Ethics 88:358–360.
Narveson, Jan and Jack Sanders, eds. 1996. For and Against the State: New Philosophical Readings. Lanham, Md.: Rowman & Littlefield.
Pilon, Roger. 1978. ‘Criminal Remedies: Restitution, Punishment, or Both?’ Ethics 88:348–357.
Rothbard, Murray N. 1977. ‘Punishment and Proportionality.’ In Barnett and Hagel 1977.
—— [1973] 1978. For a New Liberty: The Libertarian Manifesto. Rev. ed. New York: Macmillan Co.