This is from a version on my hard disk, and may differ in detail from the publishe version

Beyond the Tort/Crime Distinction


I take it that the chief purpose of Professor Seipp's paper is to establish a proposition about the history of the tort/crime distinction: that it goes back very far in English law and is based on whether the principal consequence of conviction was compensation of the victim or punishment of the offender. To me, however, the paper is interesting for two other reasons: the similarities, in function more than form, between medieval English law enforcement and eighteenth century English law enforcement, and the implications of both for the tort/crime distinction.


The Return to Private Enforcement

The medieval appeal of felony was a private action. As Blackstone put it, "on an indictment, which is at the suit of the king, the king may pardon and remit the execution; on an appeal, which is at the suit of a private subject ... the king can no more pardon it than he can remit the damages recovered in an action of battery."[

By Blackstone's time, the appeal remained in law but had almost completely vanished from practice.[2] The disappearance of the appeal of felony did not imply a shift to public prosecution of crime on the modern model. In the eighteenth century almost all crimes were privately prosecuted. The actions took the form of public prosecution-the case would be "Rex v Y" rather than "X v Y"-but a private party, usually the victim, initiated the action, located witnesses, arranged for them to appear in court, and fulfilled almost all of the functions we associate with a public prosecutor.[3]

This suggests an interesting historical question to which I have no good answer: why, after abandoning private prosecution in one form, did the English legal system reintroduce it in another?[4] It also provides an opportunity to observe how similar problems worked themselves out in two legal systems several centuries apart, different in form but similar in substance.

One such problem concerns the incentive to prosecute. A modern tort suit, if successful, results in a damage payment to the victim, giving victims an incentive to sue. But a successful private criminal suit, whether a fifteenth century appeal or an eighteenth century indictment, produces only a penalty for the defendant. Why, under such circumstances, did anyone prosecute?

The conventional answer is vengeance. In a recent piece on eighteenth century law enforcement, I offer two other explanations, both also relevant to the earlier period. One is that private prosecution was a way of producing deterrence as a private good. The other is that private prosecutors commenced cases in the hope of being paid to drop them.

The logic of private deterrence is simple. Whether someone chooses to commit a crime against me depends, in part, on what he believes will happen if he is caught. If he expects to be prosecuted, convicted, and executed, that may be a good reason not to commit the crime. It is therefore in my interest for potential criminals to believe that I will energetically prosecute crimes committed against me. One way of achieving such a reputation is by prosecuting such crimes.

That is a practical strategy if I am frequently a victim of crimes, and thus have frequent opportunities to prosecute them.[5] It is less workable for the potential victim for whom actual crime is a rare occurrence. The eighteenth century solution to that problem was an institution called an association for the prosecution of felons. A group of potential victims, usually residents of the same town, would each contribute a few pounds to a common fund to be used to pay the cost of prosecuting a felony committed against any of them. The names of the members of the association would be published in the local newspaper, in part, one presumes, for the benefit of local criminals. Thousands of such associations existed in England in the eighteenth and early nineteenth centuries.[6]

So far as I know, there were no prosecution associations in medieval England. But a society of small villages and low mobility makes possible more informal approaches to reputation and commitment. A reputation for allowing oneself to be pushed around can be a serious liability, and can be avoided by demonstrating a willingness to retaliate, either directly or through legal action. One way of interpreting the "revenge" explanation for prosecution is as the working out of a rational commitment to retaliation, designed to deter potential aggressors.[7]

A second explanation for private prosecution is that the objective was not conviction but an out of court settlement. The possibility is suggested by Professor Seipp in the medieval context, and I have made the same point for the eighteenth century.[8] In both cases there were legal rules designed to prevent private prosecutors from dropping their charges in exchange for payment,[9] but in both there is reason to believe that those rules were not always enforced.

This second explanation brings me to the one point on which I would take issue with Professor Seipp. He accepts, with some qualifications,[10] the claim that the medieval English legal system was biased towards rich defendants, since the victim of a felony by a rich felon would prefer an appeal (or indictment) of trespass to an indictment for felony, and the rich felon would thus be able to preserve his life at the cost of some or all of his wealth.

This argument misses an important respect in which the system was biased against rich defendants--one familiar to any observer of the modern American tort system. If seeking damages, there is much to be said for seeking them in the deepest available pockets. To the extent that private prosecutions were motivated by the hope of gain, whether payment for compounding an appeal of felony or damages awarded in a writ of trespass, victims had a stronger incentive to prosecute rich criminals than to prosecute poor ones. There may well have been situations in which a felon would be convicted and fined if rich and convicted and executed if poor. But there were surely also situations in which it was worth prosecuting a rich felon but not worth prosecuting a poor one.


Reasons for Change

The history of English law enforcement, from the middle ages to the present, shows a gradual and uneven shift from private to public enforcement; it starts with an Anglo-Saxon law based largely on private feud[11] and ends with a system of public police and public prosecutors. This raises the question of why these changes occurred. At least three different classes of explanation are worth considering.

The first interprets legal history as a steady improvement, a very slow learning by doing. That view strikes me as unpersuasive. I am aware of no evidence that human intelligence has increased significantly over the past thousand years. Englishmen in the fifteenth century, or the eighteenth, were better acquainted with the institutions of their society than we are, and so better equipped to create suitable institutions for that society.

A second possibility is that legal institutions changed in response to changes in other features of the society. Thus it might be argued that a law enforcement system without public police was workable, perhaps optimal, in rural England, but became increasingly unworkable in the anonymous, mobile, urban society of late eighteenth century and early nineteenth century London. Similarly, private prosecution motivated by private deterrence might be workable in a society where potential criminals knew the identity and reputation of their victims, but become unworkable with larger and more mobile urban populations.

This is a possible, even a plausible, explanation for much of the change that occurred. I would like, however, to suggest a less optimistic possibility--that the changes reflected a continued, and ultimately successful, attempt by the state to benefit itself at the expense of victims of crime.

Consider the shift during the Middle Ages away from private actions privately settleable towards public actions. If the victim of a felony initiates an appeal and then drops it in exchange for payment, the money goes to the victim. If the grand jury initiates an indictment of felony and the crown then agrees to pardon the convicted felon in exchange for a fine, the money goes to the crown. So one possible reason for legal rules designed to prevent compounding, both by punishing it and by permitting the grand jury to convert a dropped appeal into an indictment, was to increase the income of the crown at the expense of victims of crime. One would expect the result of such rules to be a reduction in prosecutions, a reduction in payments by felons to victims, but an increase in payments by felons to the crown.

The middle ages in general had a realistic attitude towards the economics of justice. The right to run courts and collect fines was a valuable property right, part of the bundle of rights held by a feudal lord. Both kings and barons regarded the power to enforce the law as a potential source of income.[12]

It may still have been a source of income of a sort, although not necessarily monetary income, in the eighteenth century. Prosecution was private, but the nominally public form of the action meant that the crown retained the power to pardon. While pardons were no longer purchased by cash payments to the crown, they may have been purchased at the price of implicit obligations, incurred by the friends and relatives who petitioned for a pardon and owed to those in a position to obtain one.[13]

A similar argument may also help explain the shift towards professional police in the early nineteenth century, with the objective presumably power, not money. The opposition to public police (and later to public prosecutors) seems to have been largely rooted in the belief that they would increase the power of the central government at the expense of the general population.[14] One source of support for public police was concern with the ability of the central government to put down public disturbances, as during the Swing riots.[15]

A scholarly analysis of the possibility that changes in English law enforcement were in part the result of attempts by the state to gain revenue and power would require more space than this comment provides, and more historical knowledge than its author commands. My purpose here is merely to suggest that the conjecture is worth considering.


Lessons for the Crime Tort Distinction


Many people, including several participants in this conference, have tried to answer the question "what is the essential difference between tort law and criminal law." One important lessons of historical studies such as Professor Seipp's paper is that that question has no answer.[16]


Institutions for enforcing law can vary in many dimensions. Examples include whether prosecution is controlled by the state or the victim,[17] whether punishment is designed to impose costs on the offender or to transfer wealth from him, whether enforcement is intended to prevent offenses or to price them,[18] and whether conviction is or is not accompanied by moral stigma. The legal institutions we are used to represent combinations of choices from the menu of alternatives--but not the only possible combinations.


Prosecution Victim/State
Transfer/ Punish
Prevent/ Price
Moral Stigma?

Auto Tort
Appeal of Felony
Indictment of Felony
Appeal of Trespass
Writ of Trespass
Speeding Ticket
Punitive Damagaes

Table 1

The typical modern tort action (Column A on Table 1), a liability suit for an automobile collision for example, is prosecuted by the victim (through his attorney), designed to transfer wealth from tortfeasor to victim, designed to price offenses, and does not impose moral stigma. The typical criminal action (Column B on Table 1), such as a murder prosecution, is publicly prosecuted, designed to impose costs on the offender, designed to prevent offenses, and imposes moral stigma. Observers familiar with such actions conclude that tort and crime are natural categories. Philosophers are likely to view the defining characteristic as presence or absence of moral stigma, economists to view it as private or public prosecution.

The clearest evidence that both are wrong, that there is (outside of the accidents of a particular legal system at a particular time) no natural category of tort or crime and thus no essential distinction, is the example of other legal systems. Consider the one described by Professor Seipp. Precisely the same action, a killing or rape, say, can be prosecuted in four different ways. An appeal of felony (Column C on Table 1) is private prosecution with punishment,[19]

an indictment of felony is public prosecution with punishment (Column D on Table 1), a writ of trespass is private prosecution with transfer (Column E on Table 1), an indictment of trespass is public prosecution with transfer (Column F on Table 1).[20] Thus a single action--with, one presumes, a single moral status--can result in legal actions corresponding to all possible combinations of the choices shown by the first two rows of the table.

One can also find examples in the current American legal system that defy any simple attempt at defining the difference between tort and crime. Consider tickets for parking or speeding (Column G on Table 1). They are publicly prosecuted, designed to transfer wealth (to the state), designed to price offenses, and impose no moral stigma. Or consider punitive damages (Column H on Table 1). They are privately enforced, transfer to the victim, are designed to prevent, and attempt to impose moral stigma.

The right question to ask is not what the essential distinction is between tort and crime; there is none. The right question is how choices in the different dimensions are related to each other, and what makes particular combinations of choices more or less appropriate for dealing with particular kinds of offenses in a particular society. A full answer to that question would require a book,[21] not a comment. But let me offer a few fragments of an answer.

Perhaps the most obvious connection is between control of prosecution and form of punishment. A legal system, such as criminal enforcement in eighteenth century England, that attempts to combine private prosecution with state punishment faces a serious problem: how to prevent the private prosecutor from converting a de jure criminal punishment into a de facto damage payment. Control of prosecution need not include the right to drop charges, but it includes the ability to prosecute badly. So in order for such a system to prevent de facto out of court settlements, it must not only make rules against them, it must spend substantial resources enforcing them. This suggests that the typical pattern of modern tort law, which combines private prosecution by the victim with damage payments to the victim and makes no attempt to prevent out of court settlements, has significant advantages.[22] The same argument explains why the writ of trespass, which resulted primarily in a damage payment to the victim, was privately prosecuted, while the indictment of trespass, resulting in payments primarily to the crown, was publicly prosecuted.

One odd feature of tort law, at least from the viewpoint of an economist, is its failure to take account, in calculating damages, of uncertainty of apprehension and conviction. Suppose I take an act which imposes a cost of a thousand dollars on you. Further suppose that my chance of being detected and successfully sued is only one in four. If I am sued and lose, I am required to make you good--to pay you a thousand dollars. That means that my expected damage payment, the probability of having to pay times the amount paid, is only two hundred and fifty dollars.[23] So it is in my interest to take the act as long as the gain to me is more than two hundred and fifty dollars. But the efficient rule would be for me to take the act only if my gain was greater than your loss--which would mean only if my gain was more than a thousand dollars.

The obvious solution to this problem is to scale up the successful plaintiff's award to make his expected return, allowing for the probability of a successful suit, equal to the damage done. In the example that would mean an actual award of four thousand dollars, giving an expected award of one thousand. But our civil law contains no such rule.[24]

One possible reason may be that tort law is privately enforced by the victim, and it is the victim who collects the damages. A rule which routinely awarded victims a substantial multiple of their actual damages, thus making them much better off as a result of the tort, would be an invitation to fraud. A "victim" could arrange an actual or fictitious tort in the presence of witnesses loyal to him, then prosecute and collect much more than the damage done. We avoid that problem with a legal rule--damage payment equal to actual damage done--under which even a successful plaintiff is no better off than if the offense had not occurred. To the extent that this rule restricts the ability of the tort system to deal with offenses that have only a low probability of producing a conviction, it provides an argument for prosecuting such offenses in some other way.

A related problem arises for offenses such as robbery, where the prosecution's key witness is often the victim. A legal system which awards damages to the victim gives him an incentive to commit perjury if doing so is necessary to secure a conviction.[25] The existence of that incentive is a good reason for the jury to discount his testimony--which makes it harder to convict even guilty defendants. A similar problem arose in the eighteenth century, as a result of rewards for conviction designed to encourage private prosecution. Juries were sceptical of testimony by witnesses who might well end up sharing in the reward. That seems to have been one of the factors responsible for the eventual scaling back of the reward system.[26]

It is tempting to look for the solution to both of these problems in public prosecution. But there is no obvious reason to assume that public prosecutors are less willing than private prosecutors to convict the innocent, if it is in their interest to do so.[27] An alternative possibilty is to deter such offenses with inefficent punishments--punishments designed to impose a cost on the offender without benefitting anyone else. But that solution brings back the problem of preventing out of court settlements--settlements that replace an inefficient punishment with a (possibly secret and illegal) payment to the prosecutor.

As these examples suggest, making sense out of the complicated interactions among the different dimensions of law enforcement is a hard problem. Tort law and criminal law as we know them provide two examples of particular sets of legal rules applied to particular offenses. Articles such as Professor Seipp's provide additional examples from other times and places. From such examples, combined with relevant theoretical considerations,[28] it may be possible to produce a coherent analysis of both the positive question of why societies use particular combinations of legal rules to convict and punish particular sorts of offenders and the normative question of what combinations of legal rules societies ought to apply to what offenses.

[1]4 William Blackstone, Commentaries *316.

[2] "As this method of prosecution is still in force, I cannot omit to mention it: but as it is very little in use ... ." 4 William Blackstone, Commentaries *312.

[3]For a more detailed description of these institutions, see David D. Friedman, Making Sense of English Law Enforcement in the 18th Century, University of Chicago Roundtable (forthcoming June 1995).

[4]This assumes that the public action, the indictment, was publicly prosecuted in the middle ages. An alternative possibility is that both the nominally public action (indictment) and the nominally private (appeal) were, in practice, privately prosecuted in the middle ages just as the indictment was in the 18th century.

[5]A familiar modern example is the case of a department store with an announced policy of prosecuting shoplifters.

[6]In addition to Friedman, supra note 3, see David Philips, Associations for Prosecution of Felons and P.F. R. King, Prosecution Associations and their Impact in Eighteenth-Century Essex in Policing and Prosecution in Britain: 1750-1859, (Douglas Hay and Francis Snyder, Eds., 1989).

[7]Commitment strategies based on emotions, such as vengefulness, are discussed in David D. Friedman, Price Theory: An Intermediate Text 288-290 ( 1990) and Robert Frank, Passions Within Reason: The Strategic Role of the Emotions (1988).

[8]Friedman, supra note 3.

[9] "The compounding of penal actions, originally allowed, was made illegal by 18 Eliz.c. 5, made perpetual by 27 Eliz. c. 10." 2 Leon Radzinowitz, A History of English Criminal Law and its Administration from 1750 138 n.2 (1957). Also see 4 William Blackstone, Commentaries, chapter 10 parts 10 and 14.

[10] "One interpretation of the choice between crime and tort is that it created one law for the rich and another for the poor. ... but the simple equation of tort for the rich and crime for the poor fails to account for the real differences between victims seeking vengeance and victims seeking compensation." Seipp draft pp. 28-9
[11]For the workings of a similar system in Iceland, see David Friedman, Private Creation and Enforcement of Law -- A Historical Case, Journal of Legal Studies 399 (1979).

[12]For examples of revenue from law enforcement under the Angevins, see W. L. Warren, King John, p. 177.

[13]See Friedman, supra note 3, and Property, Authority and the Criminal Law by Douglas Hay in Douglas Hay, Peter Linebaugh, John G. Rule, E. P. Thompson, and Cal Winslow, Albion's Fatal Tree: Crime and Society in Eighteenth-Century England 48-49 (1975). To the extent that out of court settlements between defendant and prosecutor were practical even if illegal, some of the gains from prosecution in the 18th century would have gone to the prosecutor, usually the victim.

[14] "In the mid eighteenth century a `gendarmerie' was an unthinkable incursion on English liberties," Douglas Hay and Francis Snyder, Using the Criminal Law, 1750-1850, in Hay and Snyder, supra note 5, 4.

[15]Hay and Snyder, supra note 5, 6-7, 10.

[16]This is not Professor Seipp's conclusion. But his interest is in the historical development of concepts, not the logic of different ways of enforcing law.

[17]There are other alternatives; in eighteenth century England the private prosecutor was not necessarily the victim.

[18]This is an oversimplification, but adequate for the purposes of this comment. For a more detailed analysis of optimal deterrence see David D. Friedman, Reflections on Optimal Punishment or Should the Rich Pay Higher Fines?, Research in Law and Economics 185 (1981), Should the Characteristics of Victims and Criminals Count? Payne v Tennessee and Two Views of Efficient Punishment, 34 Boston College Law Review 731 (1993).

[19]Out of court settlements between private prosecutors and their victims seem to have been common during some parts of the period and illegal during others; there may have been times when they were both.

[20]In the case of a writ of trespass, the transfer was both to the victim and to the crown; in the case of an indictment of trespass, it was usually to the crown alone, although occasionally to the victim as well. Both appeal and indictment of felony resulted in both a punishment (execution of the convicted felon) and a transfer of the convicted felon's property to the crown, and his land, after a year and a day in the hands of the crown, to his feudal superior.

[21]In progress. Drafts of early sections are available from the author.

[22]This argument provides a possible explanation for one of the odd features of the law of punitive damages--the fact that they are paid to the victim, even though ordinary damages are supposed to be sufficient to make good his loss. Under a system in which the excess damages went to the state instead of the victim, it would be in the joint interest of plaintiff and defendant to settle, thus eliminating the payment to the state. Preventing such agreements would require substantial changes in civil procedure designed to reduce the plaintiff's control over the case.

[23]I am ignoring here both risk aversion and the costs to me of my unsuccessful defense; including such issues would complicated the example without adding anything useful to the argument.

[24]It can be argued that punitive damages represent such a scaling rule, and has been, in William M. Landes and Richard A. Posner, The Economic Structure of Tort Law 160-162 (1987). I find the argument unconvincing, for reasons discussed in David Friedman, An Economic Explanation of Punitive Damages in 40 Alabama Law Review 1125 (1989). In any case, punitive damages have only recently come to play a significant role in tort law, so the puzzle would still apply to most of the history of the Anglo-American legal system.

[25]In modern tort law, we associate this problem with stories of victims who abandon their wheelchairs immediately after collecting large damage judgements for permanently crippling injuries.

[26]Friedman, supra note 3, pp. (fill in when the article is published)

[27]Consider, for example, current problems with civil forfeiture.

[28]As an economist, the theoretical structure of most interest to me is economic. But the same project could be attempted with other approaches. It may be an interesting philosophical question, for example, whether there is any natural connection between offenses prosecuted by the state and offenses subject to moral stigma.

Back to the list of articles.

Back to my home page.