Book Review Essay

BOOK REVIEW ESSAY TOO GOOD TO BE TRUE: THE POSITIVE ECONOMIC THEORY OF LAW THE ECONOMIC STRUCTURE OF TORT LAW. By William M. Landes and Richard A. Posner. Cambridge, Mass.: Havard University Press, 1987. Pp. 329. $27.50. Reviewed byJ.M. Balkin* INTRODUCTION William Landes and Richard Posner are two of the most prominent advocates of the theory that the common law promotes efficiency. In an effort to demonstrate their claim mathematically, they have collected their many articles on tort law together in a new book, The Economic Structure of Tort Law. As the authors note, this is "the first book-length study that attempts to apply [the efficiency hypothesis] to a single field of law, as well as the first book-length study of the economics of tort law" (p. vii). In fact, the book's conclusions do not diverge greatly from Judge Posner's treatment of tort law in his Economic Analysis of Law.' The difference consists mainly in the greater depth of coverage and the greater use of mathematical models to prove the efficiency of various doctrines of law. The mathematically inexperienced will not find the book easy going, and it is to the authors' credit that they always attempt to repeat in descriptive terms what they try to demonstrate mathematically. Nevertheless, anyone interested in law and economics can learn a great deal from this book, especially those persons who disagree with its conclusions. That is perhaps the highest compliment one can pay any theoretical work. However, despite the book's obvious merits, I find the argument ultimately unconvincing for a number of reasons. First, Landes' and Posner's assumptions form a cluster of beliefs that are not value free, but are intimately related to and dependent upon modem American conservative ideology. Second, the book's methodology is sufficiently controversial and sufficiently manipulable that one must doubt seriously the authors' claim to have proved that most rules of tort law are efficient. Third, a major failing of the book is its reductionism, a reduc* Assistant Professor of Law, University of Missouri-Kansas City. Harvard University, A.B. 1978, J.D. 1981. I would like to thank Ian Ayres and Dennis Corgill for their comments on previous drafts of this Article and Louis Kaplow for an enlightening discussion of economic methodology. Of course, none of them should be held responsible for what I say here. 1. R. Posner, Economic Analysis of Law (3d ed. 1986). 1447 1448 COLUMBIA LAW REVIEW [Vol. 87:1447 tionism that occurs on two levels: its attempt to view tort law as animated by a single purpose-efficiency, and its attempt to envision all human behavior as market behavior. Fourth, the book's attempt to "express the essential features of tort law in a handful of simple models" (p. viii) fails to recognize that any acceptable explanation of tort law, and indeed of the common law generally, must recognize the diversity and conflicting nature of the law's purposes and principles, which are a product of its historical evolution. Finally, I suggest that the book must fail in its project because history has made the common law heterogeneous: any theory that is able to reduce the common law of torts to a single regulatory principle is either false or so manipulable that it is of little explanatory value. I. THE PosITIvE ECONOMIC THEORY OF LAW AS AN INTERPRETIVE THEORY Landes' and Posner's basic thesis is that "the common law of torts is best explained as if the judges who created the law through decisions operating as precedents were trying to promote efficient resource allocation" (p. 1). They refer to this hypothesis as the positive economic theory of law. The word "positive" is normally used in opposition to "normative" and implies an analysis that is descriptive and value free. 2 Of course, what Landes and Posner offer us is neither of these things. Whatever one may think of the possibilities of value-free economic science in general, 3 one will not find it in this book, and it is important to understand why at the outset. A good place to begin is the concept of efficiency itself. The "efficiency" that the authors speak of is a peculiar one, with which readers of Judge Posner's earlier work will no doubt be familiar: We use ef idency throughout this book in the Kaldor-Hicks (or potential Pareto superiority) sense, in which a policy change is said to be efficient if the winners from the change could compensate the losers, that is, if the winners gain more from the change than the losers lose, whether or not there is actual compensation. (P. 16) Landes and Posner are not using efficiency in the sense of whether a policy change maximizes utility. Rather, their version of Kaldor-Hicks asks whether the change maximizes wealth: "A change is wealth maximizing if the dollar value of the gains to the winners is greater than the dollar cost of the losses to the losers. The positive economic theory of tort law holds that tort rules are efficient in the sense of wealth maximizing" (p. 16). 2. See, e.g., M. Friedman, Essays in Positive Economics 4 (1953). 3. For an introduction to the problems of defining a positive economic science, see id.; T. Hutchison, "Positive" Economics and Policy Objectives (1964); H. Katouzian, Ideology and Method in Economics (1980). 1987] POSITIVE ECONOMICS 1449 There are two important differences between wealth maximization and utility maximization, and both have to do with the fact that some people have more money than others. First, the marginal utility of income decreases as income rises: 4 this means that the utility of a dollar to a poor person is greater than the utility of a dollar to a rich person. Thus, one might increase utility simply through redistribution of wealth from rich to poor; however, this would have no effect on wealth, other than to expend it in the administrative costs of distribution. 5 Second, even if a person would gain a great deal of utility from an item, if she is too poor to afford it, she does not value it (or value it as much as a person who can afford to pay for it); value in a wealth maximization sense is defined as willingness and ability to pay (pp. 156-57). Many commentators have attacked the authors' wealth maximization criterion as immoral, inconsistent with sound economic practice, and the product of a reactionary sensibility. 6 Landes and Posner have heard all these criticisms before and remain unperturbed. After all, they are engaged in a positive economic theory of law. Their argument is that the law of torts as it presently exists maximizes wealth, whether or not that situation is morally defensible; they "are interested in explaining, rather than defending, the common law of torts" (p. 9). In one sense, Landes and Posner are surely right. The standard criticisms of normative economic analysis based upon the wealth maximization criterion-that it favors the rich over the poor, and the status quo over attempts to alter it-by themselves do not prove that the law of torts may not have had these unsavory characteristics all along. On the other hand, the nature of their project-an interpretive explanation of law-guarantees that value choices will be relevant to their work, albeit in an indirect manner. Landes and Posner are attempting what legal scholars have sought to do for generations-to offer a theory that makes the cases come out right. They are trying to "map" the received corpus of rules into a set of principles such that those principles then reproduce the rules in a logical and consistent fashion. This project is no different from what non-economic tort scholars do when they try to explain existing doctrine through theories of rights, reciprocality of risk or the like. Indeed, the project is not so far removed from what judges do when they 4. R. Posner, supra note 1, at 434; W. Nicholson, Microeconomic Theory: Basic Principles and Extensions 204 (3d ed. 1985). 5. See R. Posner, supra note 1, at 436. 6. See, e.g., Baker, The Ideology of Economic Analysis of Law, 5 Phil. & Pub. Aff. 3 (1975); Donohue & Ayres, Posner's Symphony No. 3: Thinking About the Unthinkable, 39 Stan. L. Rev. 791 (1987); Kelman, Misunderstanding Social Life: A Critique of the Core Premises of "Law and Economics," 33 J. Legal Educ. 274 (1983); Kronman, Wealth Maximization as a Normative Principle, 9 J. Legal Stud. 227 (1980); Leff, Economic Analysis of Law: Some Realism About Nominalism, 60 Va. L. Rev. 451 (1974); Symposium on Efficiency as a Legal Concern, 8 Hofstra L. Rev. 485 (1980). 1450 COLUMBIA LAW REVIEW [Vol. 87:1447 attempt to derive principles for decision from a group of seemingly conflicting cases. This "mapping" of doctrine is a special form of explanation, which necessarily combines the descriptive and the prescriptive together. Landes' and Posner's positive theory of tort law looks at first as if it were merely a scientific explanation of the forces that decide tort cases; tort doctrines are crafted so as to increase wealth. However, the book also presents wealth maximization as a unifying normative principle of tort law that could be used by judges to decide future cases, as Landes and Posner clearly think it should. It is interesting to compare Landes' and Posner's thesis with that of Ronald Dworkin, another author whose work often appears to waver between the descriptive and the prescriptive. In his latest writing, 7 Dworkin has confronted this objection head on. He argues that the goal of legal theory is to provide a creative interpretation of law as a social practice." Such an interpretation is teleological-it seeks to discover and impose an underlying purpose on the practice that both illuminates the practice and makes it the best that it can be. 9 That purpose may not be the conscious purpose of the persons who created or participate in the practice; nevertheless, the interpretation is "the most illuminating account of what [people] do [within the practice and] points out the right direction for continuing and developing the practice."' 0 Landes' and Posner's work is an interpretation of legal practice in this same sense. They argue that "the common law of torts is best explained as if. . .judges ... were trying to promote efficient resource allocation" (p. 1). Their point is that the language of the law may conceal a coherent economic logic, even if lawyers and judges themselves do not realize it, and that no other theory explains the cases in as comprehensive and enlightening a manner.11 If the positive economic theory of law is an interpretive theory like Dworkin's, it cannot be value-free. It must carry with it normative claims about what the "best" account of legal practice is. These include both judgments about what principles and assumptions constitute the best "fit" with the cases, and judgments about whether these principles and assumptions are sufficiently worthy to undergird a map7. R. Dworkin, Law's Empire (1986). 8. Id. at 45-53. For a discussion of Dworkin's theory of interpretation, see Balkin, Taking Ideology Seriously: Ronald Dworkin and the CLS Critique, 55 U.M.K.C. L. Rev. 392 (1987). 9. R. Dworkin, supra note 7, at 52. 10. Id. at 166 (discussing the interpretive theory of conventionalism). 11. See, e.g., p. 243 ("[e]conomics provides a clearer guide to understanding the structure of tort law than an approach that relies on noneconomic concepts such as causation"); p. 251 ("lawyers' references to cause may conceal economic insight"); p. 255 n.62 (noting lack of success of noneconomic attempts to give coherent accounts of the law of causation). 19871 POSITIVE ECONOMICS 1451 ping of the legal system. The latter point is likely to be overlooked. If the principles that explain a large segment of the law of torts are immoral or wicked, the "best" explanation may consist of principles and assumptions that are consistent with a smaller (or a different) class of cases but are more palatable morally. The unexplained cases would then be classified as sports or aberrations that are inconsistent with the overarching goals of the law and hence need to be eliminated from it.12 This insight also provides the key to understanding how issues of fit involve normativejudgment. No matter what principles and assumptions one uses to map the corpus of rules, they will not explain all of the cases. Thus, under any given set of principles and assumptions, some of the cases will come out "wrong." For example, in Landes' and Posner's book, the rules regarding custom, the enforceability of disclaimers of negligence, comparative negligence, contribution among joint tortfeasors and damages in wrongful death actions all fall under this category (pp. 27-28).13 However, which cases will turn out to be "wrong"-which cases do not meet the requirements of fit-depends upon the initial theoretical assumptions chosen. Those assumptions have normative implications, at least to the extent that they help determine what the "best" explanation is. For example, a very important assumption that Landes and Posner make is that people are risk neutral. In reality, most people are risk averse, a fact that follows from the decreasing marginal utility of income (pp. 55-57).14 Landes and Posner argue that the assumption is justified because (1) it simplifies their models, and (2) given perfect insurance, people will behave as if they are risk neutral (pp. 57-58 & n.8). These reasons are not particularly convincing: simplicity is gained only at the expense of a skewed analysis. In addition, the argument assumes that perfect insurance is available, which it is not. Those forms of insurance that do exist are limited to certain types of coverage and are 12. Landes and Posner recognize this fact implicitly; they defend their version of the Kaldor-Hicks criterion as "an ancient and honorable guide to social policy" (p. 17), and as a relatively noncontroversial and important social value that "has long commanded broad support" (p. 18). The import of these remarks is clear. The maximization of wealth is a "good" explanation of tort law not only in terms of fit but in terms of justification, because it is a relatively neutral value-everyone wants the size of the pie to be larger. See pp. 15-16 ("Where systematic redistribution is difficult to achieve, an interest group's best strategy is to support policies that will increase the wealth of the society as a whole, because the members of the group can be expected to share in that increase."). 13. Landes and Posner also conclude that the concept of gross negligence as a defense to contributory negligence serves no allocative purpose (pp. 91-92). 14. Given any actuarially fair gamble (for example, half the time one gets an extra dollar, half the time one loses a dollar), a risk neutral person will be indifferent as to whether to take the gamble or not. Suppose that a person values keeping the last dollar she already has more than the possibility of getting an extra dollar (so that for her the marginal utility of income is decreasing). Then she will not take the actuarially fair gamble, which is the same thing as saying that she is risk averse. 1452 COLUMBIA LAW REVIEW [Vol. 87:1447 not universally available, especially to the poorer classes. Moreover, even middle class people probably do not purchase the proper amounts of insurance necessary to make them risk neutral because of high inforunlike real insurance, mation costs.' 5 Finally, perfect insurance, 6 presents no problems of moral hazard.' Under the assumption of risk neutrality, Landes and Posner cannot explain the recent movement in the law from contributory negligence to comparative negligence and from no contribution to contribution. However, if one assumes that the marginal utility of income is decreasing (which it is) and that people are therefore risk averse (which they are), these doctrines make perfect sense, as Landes and Posner recognize (pp. 82, 211-12). The choice, then, is whether to keep the assumption of risk neutrality and have these particular doctrines be contradicted by the model, or abandon the assumption and have some different group of doctrines remain unexplained. This choice is not value-free. Some doctrines will seem more important to justify than others. If Landes and Posner could not explain the efficiency of the move to strict products liability, for example, no one would take them very seriously, for that is a major trend in twentieth-century tort law. Yet Landes and Posner have chosen to leave unexplained the movement to comparative fault and contribution, which liability, in order has been almost as universal as that to strict products 7 to preserve the assumption of risk neutrality.' Indeed, one suspects that the authors chose to assume risk neutrality for reasons quite apart from mere simplicity. The assumption of risk neutrality is central to the economic theories of the Chicago school of economics and to its conservative ideology.' 8 In particular, it allows a facile move from utility maximization to wealth maximization. The claim that people are risk neutral-that the marginal utility of income is constant instead of decreasing-allows Landes and Posner to assume that a dollar in the hands of a rich person creates as much utility as a dollar in the hands of a poor person.' 9 Thus, it simultaneously avoids 15. On the general problem of consumer information and risk assessment as obstacles to optimal economic decisionmaking in the insurance market, see G. Calabresi, The Costs of Accidents: A Legal and Economic Analysis 55-64 (1970); Wortham, The Economics of Insurance Classification: The Sound of One Invisible Hand Clapping, 47 Ohio St. LJ. 835, 870-74 (1986). 16. A. Polinsky, An Introduction to Law and Economics 54-55 (1983). 17. Even though Landes and Posner are able to explain the broad outlines of strict products liability under their analysis, they do so only by postulating that assumption of risk and contributory negligence are still invoked by courts in their most pristine form. Thus, they have some difficulty demonstrating the efficiency of modem doctrines like foreseeable misuse, which apply even when the victim is the lower cost avoider of the accident (pp. 300-01). 18. See Hovenkamp, Antitrust Policy After Chicago, 84 Mich. L. Rev. 213, 235-36 (1985). 19. Landes and Posner write the utility function as U = a + Hi, where I is income and a and b are constants. The statement that b is a constant is the (false) assumption 19871 POSITIVE ECONOMICS 1453 the conclusion under their theory that any rule of tort law could be justified either on grounds of income redistribution or risk spreading: there is no efficiency gain in income redistribution because extra income is as valuable to the rich as to the poor. There is no efficiency gain from risk spreading because all persons are assumed to be risk neutral. Landes and Posner admit that the assumption of risk neutrality would be unjustified if we were analyzing an institution patently designed to reduce risk, such as insurance. But there is no compelling reason to assume that common law judges, in formulating efficient rules of accident control, would think it important to try to reduce risk as well as accident costs and accident-avoidance costs. (P. 57) This justification is circular. How do we know that risk spreading is not a2 purpose of tort law? A lot of torts books I have read suggest that it is. 0 Of course, under Landes' and Posner's model, the purpose of tort law is efficiency, and under the assumption of risk neutrality, risk spreading leads to no additional efficiency, so it can't be a purpose of tort law; but this simply begs the question of what justifies that particular model. Note as well the illicit move equating what judges thought they were doing to what the best explanation of the purposes of the common law is. Landes and Posner do not claim that judges consciously thought they were maximizing efficiency 2 '-therefore it is no argument that they did not recognize that they were spreading risk either. Even if one accepted the premise, the conclusion is false. Many judges, especially in this century, have specifically based their decisions that the marginal utility of income is constant, or that people are risk neutral (p. 58). Landes and Posner then assume that all people have identical utility functions, so that a and b are identical for everyone. They simplify the expression by setting a=0 and b= 1. The result is that U=I, that is, that utility equals income. Thus, maximizing utility maximizes income and vice versa. Judge Posner gives his own half-hearted attempt at explaining why income redistribution is unlikely to increase social utility in R. Posner, supra note 1, at 436: "It seems at least as plausible.., to assume that income and the marginal utility thereof are positively correlated-that the people who work hard to make money and succeed in making it are on average those who value money the most .. " This simultaneously explains why children of very wealthy people have champagne tastes-they're brought up that way-and why very poor people don't really enjoy extra income very much-they wouldn't know what to do with it if they had it. Judge Posner's keen insight into the nature of the human heart allows him to perceive that an extra pair of designer shoes made Imelda Marcos much happier than $100 to a welfare mother ever could. Moreover, according to Judge Posner, your average mine worker doesn't put in nearly as tough a day as your average child of affluence. This is probably due to all those extra cocktail parties the rich have to attend. 20. See, e.g., G. Calabresi, supra note 15; W.P. Keeton, D. Dobbs, R. Keeton & D. Owen, Prosser & Keeton on the Law of Torts 24-25 (5th ed. 1984) [hereinafter Prosser & Keeton]. 21. See infra note 80 and accompanying text. 1454 COLUMBIA LAW REVIEW [Vol. 87:1447 upon a desire to spread risk. 22 The point of this example is simple: the dimension of fit-the choice of what to explain and what to leave unexplained in an interpretive theory, has a normative component. Landes' and Posner's choices are tied to their assumption of risk neutrality; this is not a purely descriptive or simplifying assumption, but one which is convenient to their political agenda and to the particular conclusions they would like to reach. II. THE IDEOLOGY OF THE POSITIVE THEORY OF LAW The above remarks suggest that the positive economic theory, far from being a dispassionate value-free inquiry into the structure of tort law, serves an important ideological function, or at least is consistent with a set of ideological views. In fact, Landes' and Posner's work bears a deep relationship to the American conservative tradition and to modem conservative ideology. This section argues that the particular cluster of assumptions and arguments found in their book is intertwined with that ideology. I begin with a bit of terminology. In tort law, we can classify positions as relatively individualist or relatively communalist. A position is relatively individualist if it tends to limit the defendant's liability, or deemphasizes the defendant's responsibility for an injury to the plaintiff. A position is relatively communalist if it tends to expand the defendant's liability, or emphasizes the defendant's responsibility to others. 23 Since the late nineteenth century, American political conservatism has generally been individualist in economic matters; it has (at least in comparison with American liberalism) viewed freedom of contract as a fundamentally important value and resisted attempts to regulate it.24 Moreover, there is a clear analogy between individualist attitudes toward economic regulation and tort duties. In tort law, economic individualism implies a preference for lower standards of liability: 22. See Prosser & Keeton, supra note 20, at 536-37 (noting modem attitude among courts towards risk spreading). 23. For a discussion of these concepts and their larger significance in legal argument, see Balkin, The Crystalline Structure of Legal Thought, 39 Rutgers L. Rev. 1 (1986); Balkin, supra note 8. The terminology originally comes from Kennedy, Form and Substance in Private Law Adjudication, 89 Harv. L. Rev. 1685, 1713-22 (1976) (distinction between individualism and altruism). We can consider these terms from the standpoint of the plaintiff as well; the relatively individualist position emphasizes the plaintiff's responsibility for her own injury, while the relatively communalist position deemphasizes it. Thus, negligence is a more individualist standard than strict liability, and no duty is more individualist still. 24. C. Rossiter, Conservatism in America 136-37, 183, 189-91 (1955); Balkin, Federalism and the Conservative Ideology, 19 Urb. Lawyer 459, 474-76 (1987). Of course, with respect to other liberties like freedom of expression, American political conservatism has been relatively communaist-that is, it has been more willing than liberalism to regulate this freedom in the interests of communal ends. Id. at 461-63. 1987] POSITIVE ECONOMICS 1455 negligence (as opposed to strict liability), market solutions to problems of duty (in preference to non-disclaimable duties), and defenses such as negligence, assumption of risk, and the fellow sercustom, contributory 5 2 vant rule. The connection between American conservatism and economic individualism explains why law and economics has found such a warm reception among conservative academics. The basic insight that gives rise to the modern law and economics movement-the Coase Theorem-holds that in the absence of transaction costs all rules of liability result in efficient resource allocation, although their distributive effects may differ. 26 Because the general assumption of economic individualists is that the market works well if not perfectly, the Coase Theorem suggests that lower levels of tort duty are as good as higher levels. Therefore, higher levels of duty are superfluous, and only waste socieindividualist rules are tal resources in administrative costs. Relatively 27 considerations. efficiency by thus dictated One might object that this argument only considers efficiency in terms of resource allocation. More communalist rules might serve a redistributive or insurance function in addition to a resource allocation function, and redistribution of wealth and risk spreading might increase total social utility and hence total efficiency. However, the ideology of economic individualism provides standard responses to each of these concerns. To begin with, economic individualism denies that liability rules are an efficient method for the redistribution of income. The redistributive effects of liability rules are often not predictable to any useful degree, and to the extent that they are predictable, the long term effects often do not benefit the classes they are designed to help but instead result in higher costs to those groups.2 8 Alternatively, the distributive effects of liability rules are diffused throughout society and do not ben25. On the other hand, because of the relatively communalist attitude of conservatives towards freedom of speech (as opposed to freedom of contract), conservatives tend to favor higher standards of liability in libel suits, and liberals tend to argue for relatively individualist positions like the actual malice rule. 26. Coase, The Problem of Social Cost, 3J.L. & Econ. 1 (1960). 27. E.g., p. 76 (defense of contributory negligence justified where both parties fail to take due care because it saves administrative costs of cases where no allocative purpose would be served by awarding damages). Similarly, Landes and Posner argue that strict liability has greater claim costs than negligence (p. 65). These are defined as the costs of processing and collecting a legal claim other than determining the defendant's level of care (p. 65). However, they note that where information costs (the costs of determining whether the defendant violated due care) exceed claim costs, strict liability is preferable to negligence (pp. 70-71). 28. E.g., Posner, The Ethical and Political Basis of the Efficiency Norm in Common Law Adjudication, 8 Hofstra L. Rev. 487, 500 (1980) ("(O]nly a naive analysis of the economic consequences of refusing to enforce the leases that poor people sign with presumably wealthier landlords would conclude that the poor would be better off under such a regime."). 1456 COLUMBIA LA W REVIEW [Vol. 87:1447 efit any identifiable group. 29 A common version of this claim is that the division between plaintiffs and defendants in tort suits does not fall even roughly along class lines; a wealthy industrialist can be a plaintiff in an automobile collision case and a gas station attendant can be a defendant in a negligence suit. Thus, the individualist argument concludes that because of the inherent inefficiency of tort liability as a redistributive mechanism, redistribution is best accomplished through the tax and welfare systems. This is often combined with an institutional argument that common-law judges possess less competence and authority to engage in redistribution than majoritarian bodies. In fact, economic individualism opposes redistribution of wealth whether practiced by judges or legislatureseach situation equally abridges economic freedom.3 0 Thus, concerns with the comparative efficiency of institutions must be understood as a means of arguing against judges adding their redistributive efforts to those of legislatures. Similarly, economic individualism denies that liability rules are an efficient method of risk distribution. Plaintiffs who wish to insure themselves against risks can do so by purchasing private insurance, which is inexpensive and widely available. This argument reflects three different aspects of the ideology of economic individualism: (1) the preference for free-market solutions and the assumption that the market works perfectly in providing commodities to those persons who value them; (2) the emphasis on the plaintiff's responsibility to take care of herself as an alternative to imposing responsibility on the defendant to take care of the plaintiff; and (3) the preference for individual rather than societal decisionmaking about the necessity and extent of expenditures for personal security, It is important to understand how these individualist arguments are connected to Landes' and Posner's assumption of risk neutrality and to the philosophy of wealth maximization. If the marginal utility of income is decreasing, risk spreading and redistribution become potentially useful methods of increasing social utility. If one assumes risk neutrality, however, neither practice increases total utility. By claiming that an equivalent of risk neutrality occurs when ideal insurance is freely available in the market, and that redistributive goals can be furthered by the legislature, economic individualism sidesteps the problem of divergence between wealth and utility maximization. It can then argue that the proper task of common-law judges is to increase wealth, which is the same thing as increasing utility if one assumes a perfect market for perfect insurance and a legislature ready to redistribute with a vengeance. 29. E.g., id. at 504-05; p. 15. 30. See, e.g., Judge Posner's remarks in R. Posner, supra note 1, at 436 ("Involuntary redistribution is a coerced transfer not justified by high market-transaction costs; it is, in efficiency terms, a form of theft."). 19871 POSITIVE ECONOMICS 1457 Of course, these assumptions are wholly implausible. There is no perfect market for insurance (a regulated industry in most states precisely because of its imperfections), and American legislatures are not about to level everybody's income. Indeed, the very same persons who trumpet the virtues of wealth maximization will be the first to oppose any serious attempt at wholesale wealth redistribution.31 However, the theoretical possibility that wealth maximization could be the same thing as utility maximization gives economic individualism all the comfort it needs. I have argued that economic individualism's delegation of redistributive goals to the legislative branch is disingenuous; that it is a device really designed to avoid redistribution in common-law rules. This leads me to another aspect of American conservative ideology that is much older than its association with economic individualism: its distrust of majoritarian interference with property rights. Throughout history, American conservatism has dreaded nothing so much as popular majorities 2using their power to take money from the rich and give it 3 to the poor. It is interesting to contrast conservatism's traditional distrust of majoritarian interference with property rights with its virtual deification of the common law and the common-law process of adjudication. The conservative love of the common law is not merely due to a reverence for the old and established. Common-law adjudication was, and continues to be, a nonmajoritarian process. Legislatures can alter common-law rules, but judges create them to begin with. Moreover, the common law is the source of the basic protections of property rights and the principle of freedom of contract, two things that modern conservatives hold dear. Finally, especially to nineteenth century thinkers, the common law provided a seemingly neutral and apolitical framework of contract and property rights; the common law was the Archimedean point from which all changes could be judged in terms of both efficiency and redistributive consequences. It followed that majoritarian 31. See supra note 24. 32. R. McCloskey, American Conservatism in the Age of Enterprise 22-25 (1964); A. Paul, Conservative Crisis and the Rule of Law: Attitudes of Bench and Bar 18871895, at 229-30 (1969); see also C. Rossiter, supra note 24, at 33-34, 61, 65 (1955) (conservative distrust of majority tyranny); Balkin, supra note 24, at 466-67, 475. The conservative solution to this problem has taken many forms. It began with direct opposition to the expansion of suffrage and voting rights. See The Conservative Tradition in American Political Thought 80-83, 115-19 (J. Sigler ed. 1969) (containing selections by John Adams, Distrust of Democracy, and by James Kent, The Argument Against Universal Sufferage). After the Jacksonian era, with the adoption of universal white manhood suffrage and the dropping of property qualifications, conservatives shifted to a strategy of constitutional limitations on popular interference with property rights, enforced by the federaljudiciary. R. McCloskey, supra, at 22; A. Paul, supra, at 230-37. After 1937, conservatives have generally resisted redistributive programs in the legislative arena, trying either to cut governmental expenditures to these programs or to divert monies to national defense. E.g., B. Goldwater, The Conscience of a Conservative 70-77 (1960). COLUMBIA LAW REVIEW [Vol. 87:1447 alterations of common-law rights were likely to be redistributive in effect, if not in purpose. Given these premises, the characterization of common-law rights as constitutionally protected rights by conservative judges in the nineteenth and early twentieth centuries is not at all surprising. The use of common-law categories to define the constitutional rights of contract and property simultaneously foiled majoritarian attempts at economic regulation while preserving the ability of common-law judges to make law without majoritarian interference. In more modem times, the conservative attack against majoritarian regulation and in favor of the common law has been made on grounds of efficiency, and here once again lies an important connection between conservative ideology and Landes' and Posner's work. It is virtually an article of faith among conservative thinkers that legislative or administrative alterations of common-law rights are almost always inefficient, and are more likely than not attempts by majorities or special interest groups to redistribute income to themselves.3 3 According to conservative ideology, popular majorities and special interests will do anything to grab goodies for themselves; they will even disturb the pristine beauty and efficiency of the common law in their lust for wealth, thus shrinking the size of the pie for everyone.3 4 We see again a variant on the disingenuous institutional argument: popular democracies would be better off leaving the common law alone, and fighting instead over tax and welfare reform. 33. Compare Landes' and Posner's description of "the dominant theory of the state among economists," actually a variant of the conservative theory: "interest groups struggle for a place at the public trough; government intervenes in the economy to redistribute wealth from politically less powerful to politically more powerful groups" (p. 15). For examples of conservative arguments that common-law adjudication is superior to legislation, see R. Posner, supra note 1, at 495-507; M. Friedman, Capitalism and Freedom 177-95 (1971); F. Hayek, Law, Legislation and Liberty: Rules and Order 94-144 (1973); B. Leoni, Freedom and the Law (1961); Rubin, Why is the Common Law Efficient?, 6J. Legal Stud. 51, 61 (1977); Stigler, The Theory of Economic Regulation, 2 BellJ. Econ. 3 (1971). 34. Indeed, so firmly is this idea embedded in conservative ideology that Landes and Posner profess bewilderment at the modem movement to statutes providing for contribution among joint tortfeasors: Because contribution seems to be a less efficient rule than no contribution, the question arises why so many states have abandoned the common law approach. We do not have a complete answer to that question, as we have no idea what politically effective interest group is benefitted by contribution among joint tortfeasors. However, whatever the benefits may be, presumably they are weighted against the efficiency losses, which are costs to some group, in the political decision-making process. (P. 219) Implicit in this statement is the assumption that only rent seeking could lead to a change in the common-law rule. Landes and Posner do not consider the possibility that modern legislatures have moved to contribution because it is deemed substantively fairer. Note also the irony of this analysis when constrasted to Landes' and Posner's earlier statement that in general, rules of tort law do not aid particular classes in society (p. 15). POSITIVE ECONOMICS 1987] 1459 There is a still further connection between Landes' and Posner's positive economic theory of law and conservtive ideology, one that also shows how deeply this "positive" theory is related to normative issues. If tort law is a predominantly efficient system of rules, or if its longterm trend is efficiency, that might be a very good reason to adopt a normative economic theory as well-that tort law ought to be efficient. This is not merely confusing an "is" with an "ought." If, despite ourselves, we have managed to produce an efficient system of common-law rules, why should we muck it all up? More importantly, there is nothing that we can do about it-the natural tendency of the common law is towards efficiency, and as long as we let judges decide cases, they will (eventually) stumble upon the most efficient rules. Of course, we can change common-law rules by legislation or administrative regulation, but as stated earlier, conservative ideology views such attempts as almost always inefficient. Therefore, if we confine ourselves to judgemade law, we might as well accept the inevitable-an efficient system of rules, which is not such a bad thing after all. Landes and Posner are not unaware of the normative implications of their work: For those readers for whom economic efficiency is not merely a descriptive theory but a call for action, we have suggested several areas in which tort law might be changed to make it conform more closely to the theory; perhaps these changes would just nudge it a little faster along its natural line of development. (P. 314) Landes and Posner thus assure us that not only are their ideas morally appropriate, but that history is moving in the same direction anyway. Nothing is more comforting than knowing that no matter what anyone does, history is on your side. And nothing is more amusing than listening to conservatives make claims about the inevitability of history that might sound familiar coming from Marxists. In sum, the positive economic theory of the common law has a decidedly Panglossian air about it: this is the most efficient of all possible worlds. This attitude ties in nicely with the antidemocratic vision of conservative ideology where property rights are concerned, and with the paradigmatic conservative belief in the comparative advantages of the status quo over proposals for change. The positive theory carries with it a subtle background message: the common law has been doing quite well on its own, thank you, and we should be more appreciative of its merits. III. THE MANIPULABILITY OF THE POSITIVE THEORY Unfortunately, there is a fly in the ideological ointment. The common law that conservative ideology cherishes so much is not the common law as it exists today. Like so many love objects, it is an idealized version whose rough edges and faults disappear in the dreamy haze in which it is viewed. It resembles much more the common law of the late 1460 COLUMBIA LA W REVIEW [Vol. 87:1447 nineteenth century, before the arrival of workers' compensation and modem doctrines like unconscionability, when the law of torts was at the peak of its movement towards economic individualism. It is no accident then, that when Judge Posner, in one of his most famous articles, looked for proofs that the common law was efficient, he chose as his 5 sample fifteen hundred cases decided between 1875 and 1905.3 It is also no accident that Landes and Posner wax eloquent about the virtues of products of the nineteenth century-the defenses of custom, contributory negligence, assumption of risk, and the fellow servant ruleas proof of the modem efficiency of the common law (pp. 131-33, 137, 139-42, 296-300, 308-11). It is as if a doting parent showed us baby pictures of a convicted felon to establish that he really was a good boy after all. The development of the common law since the nineteenth century provides a never-ending source of difficulty for Landes and Posner. If markets are so efficient, why have tort rules become increasingly communalist? Why has freedom of contract been more and more circumscribed by the law of tort, and why are there more nondisclaimable duties than ever? Faced with a conflict between ideology and history, Landes and Posner have conceded some things and ignored others. They recognize that they must fit the move to strict products libability within their theory, because it is a major source of modem torts litigation.3 6 Yet little is said about the increasingly communalist developments in landlord-tenant law,3 7 or the law of owners and occupiers of land,38 while the movements to comparative fault and contribution among joint tortfeasors are dismissed as mere anomalies (p. 314). This brings me to another example of how criticisms of the normative economic theory of law are relevant to the positive economic the35. Posner, A Theory of Negligence, 1 J. Legal Stud. 29 (1972). 36. Landes and Posner show that strict products liability is an efficient doctrine by arguing that information costs to consumers are relatively high where the risk of injury from a defective product is small (pp. 280-81, 293). Landes and Posner thus demonstrate the efficiency of strict products liability without the need to assume risk aversion (p. 273), that markets are not competitive (p. 274), or that "consumers have psychological traits that cause them to misperceive risks systematically" (p. 274). Once again, the choice of assumptions is consistent with the authors' ideological views. No self respecting Chicago economist would assume that markets were not competitive; the importance of risk neutrality has already been discussed. As for the last assumption, the Coase Theorem does not operate if consumers systematically misperceive risks. Thus, if one assumed misperception of risk, one could not guarantee that unregulated contractual solutions would produce an efficient solution in many of the situations that Landes and Posner describe as involving low transactions costs. 37. See, e.g., Kline v. 1500 Massachusetts Ave. Apartment Corp., 439 F.2d 477 (D.C. Cir. 1970) (recognizing duty of landlord to protect tenants from criminal assaults on premises). 38. See, e.g., Rowland v. Christian, 69 Cal. 2d 108, 443 P.2d 561, 70 Cal. Rptr. 97 (1968) (establishing general duty of reasonable care towards trespassers, licensees, and invitees). 19871 POSITIVE ECONOMICS 1461 ory. One of the most important criticisms of Posner's normative economic analyses has been not that they are immoral but that they are manipulable; that they alternatively adopt and discard assumptions as they are necessary to prove whatever the analyst seeks to show. The normative criticism asserts that law and economics is a method for giving reactionary policies a veneer of scientific certitude. Of course, the need to make assumptions and simplify models is an occupational disease of economists generally. 3 9 Law and economics does not become a useless endeavor because all economic models are susceptible to improvement. 40 Economic science is only possible if models can be constructed. Rather, the point here is that Landes and Posner are doing something in their "positive" theory that is not too different from what Posner has been criticized for in his normative analyses: they are trying to get to a certain conclusion (the law of tort is efficient) by constructing an economic model that guarantees that the law of tort is efficient. I can best illustrate the problem with the following story. One day, an officer of the Tsar's army was looking for able-bodied peasants he could drag off to twenty years' service. As he rode through the forest, he spotted a dozen trees with concentric circles drawn on them; on each tree was a bullet hole marking an exact bull's eye. He stopped an old man walking by and asked who was responsible for this. "Oh," said the old man, "that would be Misha, the carpenter's son. But he's a little peculiar-you wouldn't want him." "I don't care what he's like," said the officer. "Anyone who can shoot like that ought to be in the Tsar's army." "Well, that's just the point," said the old man. "First Misha shoots the bullet. Then he draws the circles around the hole." Landes and Posner are like Misha the carpenter's son. They are armed with enough ambiguity and ad hoc assumptions to succeed at whatever they want to attempt. Indeed, one suspects that the fact that any doctrines exist that don't fit into their positive theory is due more to their dislike of those doctrines and their unwillingness to adopt assumptions inconsistent with conservative ideology than to any real constraints on their analysis. 39. For a good discussion of the role of models and assumptions in law and economics literature, see Lachman, Knowing and Showing Economics and Law (Book Review), 93 Yale LJ. 1587 (1984) (reviewing A. Polinsky, An Introduction to Law and Economics (1983)). 40. See M. Friedman, supra note 2, at 15 ("the relevant question to ask about the 'assumptions' of a theory is not whether they are descriptively 'realistic,' for they never are, but whether they are sufficiently good approximations for the purpose in hand."). Friedman's argument not only explains the justifiable need for abstraction in economic theory, but also unwittingly establishes the value-laden element in the building of economic models-what is "sufficiently good ... for the purpose at hand" is not a valuefree choice. See Lachman, supra note 39, at 1598 n.59 (definition of efficiency relates to goals). COLUMBIA LAW REVIEW (Vol. 87:1447 Landes and Posner recognize that this is a possible criticism of their work: We are mindful of the dangers of rationalization. Knowing what the rules of tort law are, we are in a position to construct an economic rationale for each one, working backward. It is partly to reduce this danger that we have kept our economic model extremely simple, as by assuming risk neutrality, which greatly reduces our degrees of freedom. . . . [Nevertheless, t]he large fraction of tort rules, weighting number by importance, that conform to our economic model are evidence that the model captures an important part of legal reality. (Pp. 313-14) This is precisely the basis of criticism, both of the normative analysis and the positive theory. The very fact that a few simple and unrealistic assumptions lead to the conclusion that so many diverse rules are efficient (all without the need to assume that bugaboo of conservatism, that wealth and utility maximization are divergent), argues that (1) perhaps all of these rules are really inefficient when we take more realistic assumptions into account, or (2) there are more degrees of freedom in the model than Landes and Posner are letting on-sufficient degrees of freedom, in fact, to allow the required conclusions to be reached. One can speculate indefinitely on the possibilities. However, one cannot demonstrate the failings of the authors' model without entering into the complexities of the model itself. Accordingly, I would like to give a few examples of how the positive economic theory might be criticized for its assumptions and methodology. These are by no means the only points of disagreement I have with the book, but to list them all would require the publication of a new book in itself. A. The Choice Between Negligence and Strict Liability A useful example of the importance of assumptions in economic modeling comes early in the book, when the authors discuss the comparative advantages of negligence and strict liability. According to their model, an efficient rule of liability minimizes total social loss. The authors argue that, if due care is defined as that degree of care which minimizes total social loss, a negligence standard and a strict liability standard give equal incentives to use due care. 4 1 The argument for negligence is as follows: under a negligence standard, if a defendant takes less than due care, she must pay out money damages for the amount of injury her lack of due care causes, while if she takes due care, she pays nothing. Her total costs are the sum of the costs of care she 41. That is, when the level of activity under each rule is disregarded (pp. 64, 66). In the following example, I follow Landes and Posner in assuming that optimal plaintiff care is zero (p. 63). Where optimal plaintiff care is greater than zero, one can generalize the argument to show that negligence and strict liablity with a defense of contributory negligence are equally efficient (pp 73-80). 19871 POSITIVE ECONOMICS 1463 does take plus the injury costs to the plaintiffs caused by her lack of due care. Landes and Posner demonstrate that under a negligence standard, a rational defendant will always invest in safety up to the point where the marginal cost of care equals the marginal decrease in expected losses to plaintiffs. This is also when total social loss is minimized; therefore defendants have incentives to take due care (that is, to minimize total social loss) under a negligence standard (pp. 60, 63-64). This argument assumes, however, that expected accident losses caused by defendant's lack of due care equal expected damage awards-that is, that everybody she injures sues and collects full compensation for the harm done to them. But there is no reason to think that this happens in the real world. Many people do not sue when they are injured, either because (1) they do not realize that they have been injured by someone (rather than by an accident for which no one is responsible), (2) they do not realize that there is legal redress available, (3) they do not know how to go about getting legal advice, (4) they believe (correctly or incorrectly) that the cost of getting legal advice and/or representation will be prohibitive, or (5) they estimate they have little chance of winning, or so slight a chance that the aggravation and expense of a lawsuit is not worth it to them. Even if the injured party sues, she may not get full compensation, either because (1) she settles for less than the real amount of the injury, (2) her lawyer is inept and she recovers less than her damages or loses the case outright, (3) the defendant's lawyer is very good and reduces the damages or wins the case outright, (4) the evidence of negligence is lost or destroyed, or is unavailable, or unconvincing, or (5) the jury simply makes a mistake and awards too little or nothing at all. If we add these more realistic assumptions to Landes' and Posner's model, expected damage awards will be only a percentage of expected accident losses to plaintiffs. In that case, if the defendant equates her marginal cost of care to marginal expected damage costs, she will spend less on care than she should, because expected damage costs are less than the total social loss suffered by plaintiffs. Hence negligence will lead to an inefficient investment in safety. Moreover, under this set of assumptions, strict liability might be more efficient than negligence. Under Landes' and Posner's model of strict liability, the defendant faces the cost of care plus the cost of money damages paid out for accidents caused by lack of due care, plus the costs of money damages paid out for accidents defendant caused when due care was exercised (under negligence this last amount is zero). Once again, the defendant will spend up to the point where the marginal cost of care equals the marginal decrease in expected damage awards. Because accidents that cannot be avoided through the exercise of due care are precisely those accidents where the marginal cost of care exceeds the marginal savings in expected accident losses, the result in strict liability will be the same as for negligence. The defendant 1464 COLUMBIA LAW REVIEW (Vol. 87:1447 will spend up to the point where she is exercising due care. She will pay the costs of nonnegligent accidents because doing so is cheaper than attempting to avoid them through extra units of care (p. 63). However, if we recognize that expected damages are only a fraction of actual plaintiff loss, strict liability will not lead to an efficient investment in safety. On the other hand, strict liability might be more efficient than negligence, because the expected damage awards in strict liability cases are closer to the actual loss that plaintiffs suffer than is true in negligence cases. This is because more people are likely to sue for damages if they know that they will not have to prove negligence, and defendants are less likely to avoid paying full compensation under a strict liability regime, either because they will lose more cases or will 42 settle more quickly for larger amounts. Thus, under both a negligence standard and a strict liability standard, only a portion of the people who were injured by lack of due care by the defendant will sue and collect damages. If all of these plaintiffs sued, the defendant would have proper incentives to take due care under either standard. However, under strict liability more of this group will sue (along with the group of plaintiffs whose accidents could not have been avoided by the exercise of due care). Hence the defendant will spend more money to reduce her liability from those accidents that could be avoided through the exercise of due care, and this result 43 is more efficient. The point of this example is that questions of comparative efficiency may be subject to a number of factors, each of which, by itself, may be enough to tip the scales in favor of one rule over another. If these factors were relatively small, we could justifiably neglect them, and we would have a more apt analogy to the type of theorizing done in the physical sciences. But the factors I have added concerning the operation of the legal system must strike any experienced lawyer as palpable and significant. The costs of litigation, and the obstacles and barriers that both plaintiffs and defendants face in the determination of compensation are very real, often more real than the rules of liability themselves. And to hope, as Landes and Posner seem to do, that these various factors will all cancel each other out seems to me to be mere whistling in the dark. 42. This is implicit in Landes' and Posner's argument that strict liability has higher claim costs than negligence (p. 65). It costs more to administer a strict liability standard because more people will sue. 43. Of course, I have not even begun to show how much more complicated this model can get. For example, I should really take into account, under both negligence and strict liability, the cost that defendants will have to pay in legal expenses. This amount is likely to be different for negligence and strict liability standards-more people will sue under strict liability, while negligence presents more varied issues of proof. When the differing cost of legal expenses are added to expected damages and the cost of exercising care, we can see that the question of which rule is actually more efficient is becoming hopelessly complicated. 19871 POSITIVE ECONOMICS 1465 B. The Assumption of Full Compensation One of the most troubling assumptions that Landes and Posner make is the assumption that tort suits give full compensation. 4 4 Under this assumption, it follows that the plaintiff is indifferent between being injured by the defendant or receiving monetary compensation. On its face, this assumption seems wildly improbable. Its ideological function is to justify relatively individualist doctrines like contributory negligence, assumption of risk, and the fellow servant rule. These doctrines are all based on the need to give plaintiffs incentives to take care of themselves. The assumption of full compensation is necessary in order to make plausible the claim that plaintiffs would have insufficient incentives to protect themselves if such compensation-denying doctrines did not exist. Thus, according to Landes and Posner, plaintiffs apparently need the doctrine of contributory negligence as a deterrent; for otherwise they would happily crash into cars driven by negligent defendants; indeed, if it cost them some effort to swerve, they would have an economic incentive to collide with the defendant (p. 76). Perhaps recognizing the unreality of their assumption, Landes and Posner later concede that "tort compensation is not always full compensation" and argue that without a defense of contributory negligence, victims simply will have insufficient incentives to take care (p. 80). However, if the plaintiff knows that if she does not exercise due care she will be in a worse position (because she will not be fully compensated by the tort system), she would still have an incentive to exercise due care. If my choice is between not being hit by the car or receiving one half of my medical expenses and lost wages three years down the road, I will still choose not to get hit in the first place. Consistent with their practice throughout the book, Landes and Posner drop the assumption of full compensation when it proves inconvenient. Thus they explain the award of punitive damages in intentional tort actions on the grounds that full compensation is unlikely, so that defendants need extra sanctions to ensure that they engage in socially desirable levels of intentional injury (p. 160). C. The Presence of Transaction Costs A third example of the potential manipulability of the positive the44. E.g., pp. 45-46 ("Under a reasonableness standard an injunction will be issued [in a nuisance case] if damages are for some reason inadequate, a general principle of remedies law that we shall ignore."); p. 63 ("Under strict liability (with no defense of contributory negligence), the potential victim... has no incentive to take care, because he is fully compensated for his injury .... "); p. 76 (where care is sequential, "[i]f the injurer happened to take less than due care ... the victim, observing this, would have no incentive to take due care unless there was a defense of contributory negligence. For without such a defense he would expect compensation if he was injured; hence any expenditures on care would reduce his income without producing offsetting benefits to him."). 1466 COLUMBIA LAW REVIEW (Vol. 87:1447 ory concerns the existence or nonexistence of high transaction costs. This issue is fundamental to law and economics analyses because of the nature of the Coase Theorem: where transaction costs are low, all rules of liability will lead to an efficient allocation of resources. It follows that in situations of low transaction costs, more individualist rules are just as efficient as more communalist rules, and save on administrative costs in addition. In other words, if transaction costs are low, the Coase Theorem tells us that tort law is superfluous; allocative efficiency can be achieved through the marketplace alone, and society will avoid the additional administrative costs of litigation. Hence, the positive theory of tort law predicts that the law will impose no duty (or disclaimable du45 ties) in situations of low transaction costs. This analysis, however, seems contradicted by several doctrines in the law of torts. One is the defense of custom, which is today no longer considered a defense to negligence except in cases of medical malpractice: We are led to predict that compliance with custom will not be a defense in accident cases where transaction costs are high but will be where those costs are low. The legal pattern approximates this pattern, but only very roughly. Custom is rejected as a defense in cases where transaction costs are high but was, traditionally at least, a defense in two of the three most important areas where accidents arise out of contractual relationships: industrial-accident cases (accidents to workers on the job) and professional (especially medical) malpractice cases. It was not a defense in the third area, products cases, but.., there are good reasons for regarding that as an area of high transaction costs. (P. 133) Here we see at work the inevitable temptation to draw the bull's eye after the shot has been fired. We know that the law rejects custom in products liability cases; therefore there must be high transaction costs in those cases. We also know that the law permits the defense of custom in medical malpractice cases; therefore transaction costs must be low here. However, the choice of medical malpractice as an area in which there are low transaction costs demonstrates how little plausibility the economic model has. There are few other situations where the difference in access to information is greater, and few where the patient pays more deference to the superior authority of the defendant as a 45. Landes and Posner argue that: If transaction costs are low, an optimal allocation of resources to safety as to other activities will be achieved by negotiation regardless of the liability rule in force. In these circumstances whatever is customary is, at least prima facie, optimal.... If a higher standard of safety were optimal, it would pay the manufacturers to provide it, because the incremental price of the [safety improvement] would more than offset the incremental cost of additional safety. (As a first approximation, consumers will be willing to pay a higher price equal to the expected reduction in accident damages from a safer [product].) (P. 132) 1987] POSITIVE ECONOMICS 1467 result of the difference in knowledge. Patients do not haggle with doctors about the cost of each phase of their treatment; generally they sit there and take what the doctor gives them. Even when the doctor atconsent, the information tempts to create an atmosphere of informed 47 costs 46 to the plaintiff still remain enormous. The greatest embarrassment to the prediction of a custom defense in low transaction cost situations is that the very case which announces the rule that custom is not a defense, The T.J. Hooper,4 8 is a case involving low transaction costs. Landes and Posner try to avoid this conclusion in two ways. First, contrary to their habit in other cases, they do not take the appellate court's discussion of the facts at face value, but instead read the lower court's opinion to state that there was a prevailing custom to carry radios on board tugboats (p. 134). However, even assuming thatJudge Hand's decision on the particular facts of the case was right for the wrong reasons, The T.J. Hooper still stands for an inefficient proposition of law. Thus Landes' and Posner's second response is that The T.J. Hooper doesn't really mean what it says; in fact the rule of the case is normally only applied in cases 46. Landes and Posner use the term "information costs" in two senses: (1) the cost of determining whether a defendant exercised due care (p. 65), and (2) the cost to the parties to a transaction of determining whether the terms of the transaction are in their best interest (pp. 139, 280-82, 284-87). The latter is the more traditional use of the term, and I use it in this sense. Information costs are thus a type of transactions cost which would prevent the Coase Theorem from operating. 47. The examples of malpractice cases that Landes and Posner give in their book bear witness to the unreality of their assumption. In Quintal v. Laurel Grove Hosp., 62 Cal. 2d 154, 397 P.2d 161, 41 Cal. Rptr. 577 (1964), the plaintiff was operated on by an opthamologist and suffered cardiac arrest while under general anesthesia; the charge of malpractice was based upon the failure of the hospital to provide a surgeon capable of doing open heart massage at every operation where general anesthesia was used. In Lucy Webb Hayes Nat'l Training School v. Perotti, 419 F.2d 704 (D.C. Cir. 1969), the plaintiff's decedent was a mentally ill person who broke free from his escort and threw himself out of a closed window. Landes and Posner argue that "the... deference given customary standards of care in both cases are... appropriate from an economic standpoint. The existence of a preexisting voluntary relationship between doctors and patients suggests that transaction costs were not prohibitive and hence, via the Coase theorem, that customary standards were efficient" (p. 107). We may well ask, however, why there were not very high information costs in Quintal: the patient was unlikely to bargain for an extra surgeon in the operating room unless he knew that (1) cardiac arrest was a common occurrence under general anesthesia; (2) the opthamologist could not perform open heart massage if cardiac arrest occurred; and (3) an additional surgeon would not be provided as a matter of course. Of course, all the patient had to do to get this information was ask, but how is the patient supposed to know that these questions should be asked in the first place? Similarly, in Lucy Webb, it is ridiculous to assume that insane patients will have the wherewithal to bargain rationally about measures designed to protect them from themselves. A better argument is that their saner relatives will do the bargaining for them, but even these persons are unlikely to possess enough information to ask the right questions. 48. 60 F.2d 737 (2d Cir. 1932). In The T.J. Hooper, the plaintiff argued that the defendant's failure to keep a working radio aboard its tugboat led to the loss of plaintiff's cargo during a storm. 1468 COLUMBIA LAW REVIEW (Vol. 87:1447 of high transaction costs (pp. 135-36). 4 9 Landes' and Posner's discussion of The T.J. Hooper leads them to a still more general problem: why should courts ever imply tort duties other than those the parties have agreed to where there are low transactions costs? The authors answer that standard terms economize on transaction costs, and since most parties will bargain for due care anyway, the courts should simply read the economically efficient rule into all contracts (p. 136). However, this solution simply leads to a further puzzle: why don't courts permit explicit disclaimers of negligence liability for personal injury when transaction costs are low? At this point, Landes and Posner concede that information costs may be high: "Once transaction costs are seen to include information costs as well as the free-rider and holdout costs that analysts of transaction costs usually focus on, it becomes apparent that negotiations regarding safety may be prohibitively costly even if the parties have a contractual relationship" (p. 139). Yet they argue that "[t]his analysis seems not to hold in the medical setting" (p. 139); ironically, this is the very setting in which the inequality of access to information between the parties is most pronounced, and in which the information costs to the plaintiff are greatest. Landes and Posner thus find themselves in a theoretical bind: The Coase Theorem appears to demonstrate why custom should be used in medical malpractice cases, but under that theory, disclaimers of liability should be enforceable because transaction costs are low. On the other hand, the law does not permit disclaimers of liability, which makes perfect sense if transaction costs are high, but in that case, medical malpractice cases should not be governed by custom. The answer to this dilemma is that the assumption of a perfectly operating market for medical services is incorrect. Custom is the standard used in medical malpractice cases because otherwise the costs of determining what is due care would be too high, not because it is presumed that market forces have made medical care efficient. Thus the medical profession is held to the administratively simple standard of custom, but there are still high information costs to patients, so that disclaimers of liability should not be enforceable. 49. The authors attempt to prove this by examining the facts of every tenth case that cites Judge Hand's opinion. Of course, this raises serious methodological problems. It is unclear if inspecting every tenth case proves anything at all about the rule of The T.J. Hooper. It would perhaps have been more convincing if one read not one tenth, but almost all of the cases, and demonstrated that the case is regularly distinguished in situations of low transactions costs. If one assumes, as Landes and Posner apparently do, that the general trend of the law is towards efficient rules, and if an exception for low transaction costs situations is efficient, one would think that it would eventually appear in the cases, first by distinguishing the old precedent repeatedly, then by bolder doctrinal moves. However, because the rule of The T.J. Hooper is much newer than the older rule permitting custom as a defense, the more natural inference is that it is a more efficient rule than its predecessor. 1987] POSITIVE ECONOMICS 1469 Given the crucial importance of transaction costs to their economic analysis, Landes and Posner routinely and conveniently discover and disregard the existence of transaction costs when (under their other assumptions) rules of law can only be explained by their existence or absence. This practice is possible because, as I shall now argue, the concept of transaction costs is much more manipulable than generally supposed. D. The Manipulation of Transaction Costs A good example of manipulative technique involves the authors' attempt to justify the rule of private necessity of Vincent v. Lake Erie Transportation Co.5 0 In Vincent, the defendant was moored at the plaintiff's dock when a storm arose. In order to save the ship, the defendant's servants kept it moored to the dock even though the ship pounded against the dock and severely damaged it. The defendant was held privileged to use the dock in the emergency but was required to compensate the plaintiff for the damages incurred.5 1 The Coase Theorem would seem to indicate that such a rule is unnecessary if the parties can agree beforehand-if the ship is worth more than the dock, then presumably the shipowner will bribe the dockowner to let him destroy the dock. However, Landes and Posner argue that the result in cases like Vincent is justified on the grounds that "[t]he emergency precluded a transaction in which the defendant would have negotiated with the plaintiff for compensation for using the pier."' 52 Consistent with this analysis, the authors state confidently: "[w]e know of no case in which a defense of private necessity has been recognized where transaction costs would have been low" (p. 181). However, the authors need only look to the Vincent case itself. The plaintiff dockowner and the defendant shipowner had an ongoing business relationship. The defendant had moored in the plaintiff's dock before the storm; it did not do so as a trespasser, but because there was a contract of dockage between the parties. Landes and Posner claim that "[n]o provision of the contract between the parties obligated the plaintiff to provide a berth for the defendant's ship during the storm" (p. 178). This interpretation is not at all obvious-one might interpret the contract to state that the plaintiff agreed to keep the ship docked for a reasonable time in return for an hourly or daily fee, in which case the plaintiff could only refuse to honor the contract on the grounds of impossibility or frustration of purpose. However, even accepting Landes' and Posner's assumption 50. 109 Minn. 456, 124 N.W. 221 (1910). 51. Id. at 222. 52. P. 179 (discussing Ploofv. Putnam, 81 Vt. 471, 71 A. 188 (1908)). Landes and Posner argue that the same difficulty would arise whether the shipowner sought to protect property (as in Vincent) or personal safety (as in Ploof) (p. 179). 1470 COLUMBIA LAW REVIEW [Vol. 87:1447 that there was no prior agreement on this point, the natural question to ask is why the defendant did not bargain for the right beforehand. Before the storm there is no emergency, no problem of bilateral monopoly. 53 It is only when we look at the transaction during the height of the storm that the transaction costs the authors rely on appear-to be high. This last example raises a general methodological problem. Throughout the book, Landes and Posner envision the presence of transaction costs alternatively from a narrow time frame and from a broad time frame. 54 Thus, in Vincent, we can look at the possibilities of a transaction before the storm, when the defendant first enters the dock, or in the context of a long course of dealing between the parties (a broad time frame); on the other hand, we can freeze the moment when the storm is at its height (a narrow time frame) and consider whether transaction costs are high at that point. The inherent manipulability of this strategy consists in the fact that transaction costs are almost always lower when viewed from a broad time frame. 55 In this way, Landes and Posner can conveniently choose a broad or narrow time frame whenever necessary in order to make their theory's predictions match received case law. For example, consider the issue of implied consent. Suppose that a surgeon discovers in the course of an operation that another operation is necessary, and performs it without the patient's consent. Should this be considered a battery?5 6 Landes and Posner argue that in emergency situations this is a case of high transaction costs where consent should be implied: "Characterizing a situation as 'one of unforeseen emergency, critical in its nature' is a way of saying that the costs of an 53. The concept of bilateral monopoly is used to explain how two parties in a face to face transaction might not reach an optimal economic solution, because of hard bargaining, strategic behavior, or general intransigence. The Coase Theorem, of course, predicts that in a situation of low transactions costs, an optimal bargain will be struck. However, where there is only one buyer and one seller (because of an emergency situation, for example), bargaining strategies may preclude the solution that the economic analyst deems efficient. Hence, this is called a situation of"bilateral monopoly," defined to be a type of transaction cost or market imperfection that prevents the Coase Theorem from operating. The irony of calling face to face bargaining a situation of high transaction costs has not been lost on commentators. See, e.g., J. Murphy &J. Coleman, The Philosophy of Law: An Introduction toJurisprudence 258-62 (1984) (defining bilateral monopoly); cf. Lachman, supra note 39, at 1587, 1592-98 (1984) (decision to label strategic behavior in situations of bilateral monopoly inefficient involves normative assessment). 54. See Kelman, Interpretive Construction in the Substantive Criminal Law, 33 Stan. L. Rev. 591 (1981) (demonstrating manipulability of broad and narrow views of act, intent, causation and time in the criminal law). 55. Similarly, the existence of bilateral monopoly also depends upon the time frame in which the situation is viewed. At the height of the storm, there is only one buyer and seller. Before the storm, however, there may have been many places for boats to dock; that is why there is a market for dockage services. 56. See, e.g., Mohr v. Williams, 95 Minn. 261, 104 N.W. 12 (1905). 1987] POSITIVE ECONOMICS 1471 explicit transaction with the patient are prohibitively high. In such a case the law properly allows the market to be bypassed."' 57 This is a judgment from a narrow time frame-the surgeon has cut the patient open; the patient is unconscious, or if conscious, is in no position to haggle. Transaction costs could not be higher. Yet, viewed from a broad time frame, it is a situation of low transaction costs, because the patient can always fill out a general consent form before the operation takes place. Indeed, proof that transaction costs are low is the fact that to fill out consent forms of this type behospitals often do ask patients 58 cause of the fear of lawsuits. IV. THE PROBLEM OF REDUCTIONISM The positive economic theory of law requires that most if not all doctrines can be explained by the single concept of economic efficiency. Landes and Posner argue that the building blocks of legal docand reasonableness-are trine-legal concepts like intent, causation 59 reducible to considerations of efficiency. This reductive strategy has both advantages and disadvantages. Its advantages are that it attempts to bring a deeper logic and intellectual unity to many different legal rules. Its disadvantages are that such an explanation may be unduly impoverished-it deliberately disregards the many conflicting concerns that may animate a body of law. Different visions of human nature and different moral concerns that give rise to a multiplicity of doctrines all must be shoehorned into the model and reexplained as facets of a unitary principle. Thus, the reductionist when in fact strategy views conflict and diversity as mere appearance 60 they may be fundamental features of legal thought. The poverty of the reductive strategy is nowhere more evident than in the authors' treatment of intentional torts. According to their model, intentional torts differ from unintentional torts in that the de57. P. 172 (quoting Prosser & Keeton, supra note 20, at 119). 58. Even where this is done, market imperfections still exist, not because of the time frame, but because of high information costs. It is important to distinguish the bilateral monopoly claim about transaction costs from the claim about inequality of access to information. In the former case, we can easily manipulate the result by varying the time frame. The latter case is not immune from time frame manipulation, but it occurs only with considerably greater difficulty. It is true that given a sufficiently large amount of time, a patient might learn enough about medicine to make an informed judgment about her operation, but she still has to know what questions to ask in the first place. 59. E.g., p. 153 (intentional action occurs when costs of avoidance by injurer are low in comparison to social benefits of the activity); p. 229 ("the idea of causation can largely be dispensed with in an economic analysis of torts"). The reduction of"reasonableness" to economic terms-the equation of reasonable care to the Learned Hand Test to wealth maximization-is by now familiar (pp. 85-87). 60. I have argued that tort law is best understood as presenting a series of conflicting moral principles which oppose each other in each and every area of doctrine. Balkin, supra note 23; Balkin, supra note 8. 1472 COLUMBIA LAW REVIEW [Vol. 87:1447 fendant normally has to expend resources in order to injure the plaintiff (p. 153).61 Consequently, the total social loss from an intentional tort includes the cost to the defendant to commit the tort, the gain G to the defendant from committing the tort, the damage D that the plaintiff suffers as a result, and the cost to the plaintiff to prevent the tort from occurring (p. 153). If the gain to the defendant is less than the damage to the plaintiff, (G < D), total social loss is minimized when the defendant does not attempt to commit the intentional tort and the plaintiff takes no precautions against its commission (pp. 153-59). Hence, under Landes' and Posner's model, when G < D, the defendant should be held liable for her intentional tort; this will deter persons from committing such torts to a societally useful degree. However, this model raises an interesting question: what happens if G > D, that is, if the defendant gains more than the plaintiff loses when the defendant commits an intentional tort? Assume (to use one of the authors' examples), that B comes across a deserted cabin owned by A, that B is starving and therefore breaks into the cabin and steals some food (pp. 155-56). Here, Landes' and Posner's theory distinguishes two cases. Although liability should be the same in both, the reasons for liability are different in each. In the first case, B has enough money that B could buy the food from A if A were present, while in the second, B is too poor to meet A's asking price. In the first case, the intentional tort of conversion maximizes wealth, because the food is worth more to B than to A. Only transaction costs prevent the bargain from being struck (p. 156). This suggests that there is nothing inefficient about such transactions per se. However, Landes and Posner suggest that there should be liability nevertheless, because without liability plaintiffs would attempt to expend resources to prevent conversions, and "[t]he law does not want to encourage potential victims to spend resources on preventing this kind of taking" (p. 156). With a guarantee that they will be fully compensated in a later tort suit, however, plaintiffs will not expend resources on preventing these kinds of conversions and society's wealth will be 62 maximized. Surely something is wrong here. Under this model, there is nothing improper about conversion per se if you could have afforded to purchase the stolen item in the first place; in fact, such a conversion actually maximizes society's wealth. The only reason to compensate the plaintiff at all is the fear that otherwise the plaintiff might try to 61. The more general case is where the "costs of avoidance to the injurer are low relative to the social benefits of the activity" (p. 153). 62. Because, as we know to be the case, suits for conversion do not offer full compensation-either because of the costs of bringing suit, the inadequacy of compensation, and the more than occasional inability of our system ofjustice to locate the valuemaximizing defendant-plaintiffs will still purchase alarm systems, guard dogs, and the like, in an apparently misguided attempt to protect their property. 1987] POSITIVE ECONOMICS 1473 prevent the conversion from taking place. This reduction of questions of intentional tort liability to issues of efficiency misses the intuition that deliberate theft is wrongful because it is deliberate and because it is theft. The right to compensation does not arise from any notion of moral desert stemming from injury to the plaintiff, but rather from the fear of retribution from the plaintiff. If we did not think plaintiffs would try to get their property back, there would be no reason to redress the wrong, and indeed, there would be no wrong to redress. The authors' logic becomes even more bizarre when we consider the second case-where B is too poor to meet A's asking price for the food. Here the authors' assumption that efficiency means wealth maximization and not utility maximization reappears with a vengeance. In this case, argue Landes and Posner, the transaction is not wealth maximizing, because B does not have enough money. In this case G > D in a utilitarian sense, but ... not in the narrower sense in which we compare gains and losses in this book. Our concept of efficiency excludes utility not accompanied by willingness to pay. In terms of our model, this is a case in which G is actually less than D. (Pp. 156-57) The curious result of this logic is that the richer you are, the more likely it is that stealing something maximizes society's wealth; the poorer you are, the more likely it is that stealing a loaf of bread to feed your starving family is inefficient. Only persons with years of special63 ized training could be able to convince themselves of this conclusion. Finally, Landes and Posner consider the case of intentional torts like rape, where the value in the intentional act is precisely the fact that it is coerced. Under this theory, the enjoyment the defendant derives from the rape would be decreased if the defendant had to gain consent through the market. Landes and Posner argue that, on the contrary, "there is no increase in value in the rape case because it is not the kind 63. Landes' and Posner's insistence upon translating every doctrine of law into market terms occasionally gives their work an air of unreality. Thus, the authors justify the award of nominal and punitive damages in Alcorn v. Mitchell, 63 Il1. 553 (1872), where the defendant spat in the plaintiff's face, on the grounds that "the court was implicitly encouraging 'transactions' of this type to take place in the market" (p. 161). In Vosburg v. Putney, 80 Wis. 523, 50 N.W. 403 (1891), a schoolboy was held liable for battery for giving another child a playful kick on the shin. Landes and Posner argue that "[e]ven if the defendant's gain had exceeded the plaintiff's loss, the economic argument for liability would still have been compelling because the injury occurred in a setting of low transaction costs" (pp. 168-69). But we may well ask: where is the market for kicking and spitting on people? Moreover, even accepting this market analysis, in both Alcorn and Vosburg, one might easily argue that transaction costs are quite high because of the problem of bilateral monopoly: in Alcorn, the person who made the defendant mad enough to spit was the plaintiff, in Vosburg, the defendant wanted to tease this particular classmate and not another. In such cases, we might argue, the rule of punitive damages is inefficient-by forcing the parties to engage in "market transactions" in a situation of high transaction costs, there will be insufficient amounts of kicking and spitting. 1474 COLUMBIA LA W REVIEW (Vol. 87:1447 of coercive act that improves the operation of the market, as the theft [of food] from [a deserted] cabin does" (p. 157). Landes and Posner confess that "[f]or reasons unclear to us, this rather recherch6 example has become a focus of criticism of the adequacy of the positive economic theory" (p. 157 n.9). It may be useful to stop for a moment and consider why this objection keeps popping up. The assumption that Landes and Posner make in rebutting the rape example is that one must "tie the idea of value to the voluntary processes of the market" (p. 157). But this begs a central question-can value exist outside of a real or shadow market? For Landes and Posner, the answer is obviously no: a thing has value only to the extent that it can be bought and sold in a real or hypothetical market transaction. Yet this reduction of all value to market value is precisely what Landes' and Posner's detractors find so upsetting. Thus, the rape example returns again and again because it is symptomatic of a more general dissatisfaction with the explanatory power of the reductive model: The theory of value underlying the economic analysis is not only incomplete, but seems completely extraneous to the real issues involved in a rape. These are issues of bodily security and personal dignity, the sort of things that people generally do not think should be the subject of market transactions. Thus, critics keep focusing on the example of rape because of the not unreasonable belief that one's right to be free from rape does not depend on whether rapes improve the operation of the market. At first glance this criticism might seem irrelevant to a positive economic theory-after all, Landes and Posner are arguing that the law is consistent with their conclusion that the rapists should be liable in tort, because, under their assumptions, rape is not wealth maximizing. Whether wealth maximization is a good idea, a theory based upon wealth maximization produces results consistent with the rules of positive law. However, if the reasons why rape is tortious under the positive economic theory seem wildly at odds with the reasons we believe rape should be tortious, that fact alone should lead us to doubt the explanatory capacity of the theory. Suppose that I offered a positive theory to explain tort law that suggested that rules of law maximized hours of sleep. This might be part of a broader theory in which I would demonstrate that all human activity is designed to maximize hours of sleep: I might point to evidence that people eat in order to feel sleepy, that they work in order to become tired at the end of the day so that they will go to sleep, that persons deprived of sleep for long periods become insane, and so forth. Suppose I then argued, under the assumptions of my model, that the laws against rape were sleep maximizing because, (1) knowing that rapes are illegal, men and women sleep more peacefully at night, (2) women who have been raped tend to suffer from insomnia, and (3) outlawing rapes tends to encourage consensual sexual relations, which as 1987] POSITIVE ECONOMICS 1475 we all know tend to take place in bedrooms, make both subjects very tired, and thus lead to more sleep. A reasonable person would probably respond that this explanation is ridiculous-rape is prohibited by our society because it offends our most deeply held beliefs about personal dignity and autonomy, not because it maximizes the hours we sleep. This skeptical conclusion would not change even were I able to rebut counterexample after counterexample attempting to refute my sleep maximization hypothesis. No matter how well I was able to provide a coherent positive theory of sleep maximization, the mere fact that my teleological explanation of human behavior seems so far removed from common sense would tend to 64 make one doubt its validity. Of course, there is a fundamental difference between a wealth maximization theory and a sleep maximization theory. An economic model of human behavior seems germane to many areas of accident prevention, especially if it is based upon utility maximization as opposed to its bastard cousin, wealth maximization. However, in many other aspects of our lives an economic model based on wealth fails us, and a broader notion of human practical reason must be invoked. This is a basic shortcoming of Landes' and Posner's reductionist model. A theory of economic behavior is merely a subset of a more general theory of human values and rational choice (or irrational choice). Landes' and Posner's attempt to reduce all human behavior to market behavior and all human values to the maximization of wealth creates the air of unreality that one so often finds in their analyses. Landes' and Posner's work involves a "privileging"-an assertion that a certain form of behavior-rational market behavior-is the standard case, the normal, the routine, the foundational. Other forms of behavior are peripheral and exceptional-they all are subsidiary forms that can be explained in terms of market behavior. Like all privilegings, however, this one can be deconstructed: We may ask whether market behavior is not itself simply a special case of human behavior-whether it too, is only one of a number of different forms of human choice, which in turn depend upon many different forms of human valuation and motivation. 65 Human values and goals may take wealth maximiza64. This shows once again the role that ideology plays in interpretive explanations. Even if there is a high degree of fit between the theory and experience, the "weirdness" of the explanatory theory makes us reject it. However, "weirdness" is not an inherent property of the explanation, but a function of who is explaining what to whom. Imagine a group of"sleepologists," who had spent their entire academic careers studying human sleeping habits, and whose lives were totally absorbed in this facet of human life. To some of these scientists, the sleep maximization hypothesis might not seem bizarre at all. However, our likely response would be that those sleepologists have a narrow and incomplete picture of reality which tempts them to believe that all human problems can be reduced to their particular field of study. 65. The simplest version of this deconstruction notes the dependence of market behavior on other forms of human practical reason. For example, in a market, people 1476 COLUMBIA LAW REVIEW [Vol. 87:1447 tion into account, but they may not be exclusively or even primarily concerned with it. Human action and human decision may rest only in part on the type of reasoning acceptable to Landes' and Posner's reductive vision. Ironically, then, the greatest problem with wealth maximization as a theory of human practical reason may be that it is insufficiently rich. V. THE CAUSAL EXPLANATION OF THE PosITIvE ECONOMIC THEORY OF LAW Landes and Posner argue that their positive theory of law must be correct because no other existing positive theory can explain all or even most of the cases. 66 But this begs a fundamental question: why should we assume that any positive theory of the common law can explain all of the cases? After all, the common law is the product of many minds working over many generations. It would be remarkable indeed if there 67 were a single explanatory principle manifested in all of its doctrines. The best explanation of the common law may be that it has no single guiding principle, but rather contains a combination of conflicting principles. This is partly due to the fact that common-law doctrines were created at many different times in many different places; it is partly due 6 8 It to fundamental oppositions inherent in our moral consciousness. is not a sufficient reason to accept a positive economic theory of law that no one has offered a better theory that explains all the cases. Rather, that is a reason to be suspicious of any theory that purports to explain them all. Hence, Landes' and Posner's strongest argument-that the positive economic theory of law offers the most coherent and consistent explanation of the law of torts-may also be its weakest argument. choose rationally to maximize the satisfaction of desire, but market behavior presumes that there are other methods of choice to determine tastes and preferences. One cannot know what wealth is in a society until one relates wealth to a system of desires and values. Thus the rationality of market bahavior ultimately depends upon nonrational decisions (or other forms of nonmarket choice) used to create these desires, values, and preferences. As David Hume expressed it, "reason is ... the slave of the passions." D. Hume, An Enquiry Concerning Human Understanding § XII, pt. I (Selby-Bigge ed. 1894). The above deconstruction has argued that the value to be maximized-wealthdepends upon other values. Similarly, we can argue that rationality of choice in the economic sense is only a special case of a more general method of human decisionmaking. For a useful compendium of social science research demonstrating the many different strategies that people use to make decisions (some quite inconsistent with Landes' and Posner's model of rational market behavior) see R. Nisbett & L. Ross, Human Inference: Strategies and Shortcomings of Social Judgement (1980). 66. P. 21 ("there is no well-developed theory of what motivates or explains tort law besides efficiency"), pp. 252-55 (noting failure of legal writers to devise noneconomic theory of causation). 67. See Balkin, supra note 8, at 415-16. 68. Balkin, supra note 23; Balkin, supra note 8, at 408-22. 19871 POSITIVE ECONOMICS 1477 Everything we know about the development of the common law suggests that it is a patchwork of principles and counterprinciples woven together through history. Thus, the authors present us with a puzzlethey give us no convincing explanation of how the law could have ended up the way it did-with most rules promoting economically efficient results. 6 9 This is not necessarily a fatal defect in a theoryLandes and Posner clearly believe that their book demonstrates an undeniable tendency towards wealth maximization in the law of torts, and if one accepted this, one might be quite content to accept the thesis now and wait for someone else to provide the causal explanation later. The problems with this approach are twofold. First, the positive theory seems both too manipulable and too ideologically convenient to establish a clear tendency towards efficient rules except to those persons already convinced of the theory. Second, Landes' and Posner's theory has the same failing as the sleep maximization hypothesis-even if one demonstrates a considerable degree of fit, it will still remain implausible unless a reasonable causal connection can be demonstrated. There are several possible causal explanations of the positive theory. One is that judges consciously attempted to make the law efficient. A second is that judges, laboring under some form of ideological delusion, attempted to promote whatever values they thought were best but actually promoted efficiency. A third possibility is that regardless of what judges wanted or thought they wanted, the process of litigation itself produced efficient results. Before attempting to consider which one of these explanations is best, we must consider Landes' and Posner's thesis more closely. Landes and Posner do not appear to be arguing that the actual outcome of rules in society is efficient. They contend only that most rules "creat[e] incentives for parties to behave efficiently rather than that 0 they actually behave so" (p. 314).7 Moreover, the authors only appear to claim that the rules of tort law promote efficiency each by itself certeris paribus, and not in the con71 text of all the other rules, both judicial and legislative, in our society. 69. More precisely, because the authors' theory is interpretive, the problem is how the rules of tort law could have developed so that an outside observer would find that the best explanation of the rules is that they promoted efficiency. 70. This raises a serious methodological difficulty for Landes and Posner. How can we know if incentives towards efficiency are created unless we know that the actual outcomes are efficient? Landes' and Posner's strategic retreat on this point may gain plausibility only at the cost of making their "positive" theory effectively nonfalisfiable. See M. Friedman, supra note 2, at 13. 71. Compare Landes' and Posner's attempt to rebut (as their theory requires) the conservative claim that strict products liability law is inefficient: actual case outcomes may be quite different [than the model predicts] and support the conservative criticisms. This is a general limitation of our analysis. We can (or we think we can) assess the efficiency of the substantive rules of tort law and illustrate those rules with many cases that make economic sense. But we 1478 COLUMBIA LAW REVIEW [Vol. 87:1447 This is an important point, for a rule may produce efficient outcomes if all other factors in society are efficiency promoting, yet produce highly 72 inefficient results if any of these other factors are altered. For example, as conservative law-and-economics types are fond of telling us, we live in a world riddled with inefficient governmental regulation. Moreover, as they are not so fond of telling us, we live in a world full of spillover effects in situations where transaction costs are high. 73 In such circumstances, a demonstration that a particular rule of tort law, judged only by itself, promotes overall efficiency, may be quite misleading; the rule may actually have the opposite effect. Landes' and Posner's assumption, then, appears to be only that individual rules of tort law promote efficiency certerisparibus, and not necessarily when the problems of second best are taken into account.7 4 In short, Landes' and Posner's theory seems to have less to do with actual efficiency than with the appearance of efficiency. With this in mind, we can return to the question of a causal mechanism. One possible theory, which has been much discussed and criticized in recent years, is a litigant-centered evolutionary theory. Several versions of this theory have been proposed, most prominently by Paul Rubin, 7 5 George Priest, 76 and John Goodman. 77 The common theme in all of these theories is that litigants relitigate inefficient rules more often than efficient ones, resulting in a long-term trend towards effi78 cient rules. Such a theory cannot form the basis of Landes' and Posner's positive economic theory of law. First, all of these theories are designed to make no attempt in this book to evaluate the overall efficiency of anyfield of tort law, which depends on the actual administration of the law by judges, juries, lawyers, insurance claims adjusters, and others. (P. 27 (emphasis added)). 72. This is the problem of second best, see Lipsey & Lancaster, The General Theory of Second Best, 24 Rev. Econ. Stud. 11 (1957), which is conspicuous by its absence in this book as in most of the other writings of Landes and Posner I have encountered. 73. See Rizzo, The Mirage of Efficiency, 8 Hofstra L. Rev. 641 (1980). 74. "[O]ur theory... is a theory about judicial behavior rather than global optima "..(p. 98). Thus, they admit that "Courts ordinarily do not consider technological possibilities not urged by one of the parties, and ... this may mean that judicial decisions do not always reach optimal results" (p. 98). The positive theory requires only that "[t]he approach Ijudges] employ is efficient within the framework of the facts presented to and recited by the court in its opinion" (p. 98). 75. Rubin, Why is the Common Law Efficient?, 6J. Legal Stud. 51 (1977). 76. Priest, The Common Law Process and the Selection of Efficient Rules, 6 J. Legal Stud. 65 (1977). 77. Goodman, An Economic Theory of the Evolution of the Common Law, 7 J. Legal Stud. 393 (1978); see also Terrebonne, A Strictly Evolutionary Model of Common Law, 10J. Legal Stud. 397 (1981); Note, The Inefficient Common Law, 92 Yale L.J. 862 (1983). The latter note advocates a theory with precisely the opposite conclusion-that the common law is getting less efficient over time. 78. For an excellent discussion of the economic evolutionary theories, and of evolutionary theories of the law in general, see Elliott, The Evolutionary Tradition in Jurisprudence, 85 Colum. L. Rev. 38 (1985). 1987] POSITIVE ECONOMICS 1479 explain an evolution towards actual efficiency, rather than towards the mere appearance of efficiency certeris paribus that is involved in Landes' and Posner's theory. Second, these theories at best demonstrate a modest claim-that there is some long-term tendency towards effibecome efficiency. They do not even establish that most rules7 will 9 cient, much less that the vast majority presently are. Landes' and Posner's positive theory is judge-based; that is, rules promote efficiency (when viewed in isolation) because certain forces operate upon judges to adopt such rules. However, Landes and Posner do not claim that judges consciously endeavor to promote efficiency goals. That claim seems contradicted by the language ofjudicial opinions themselves, and by our knowledge of the judicial process. Many judges, if not most, whether conservative or liberal, work on the assumption that they are following precedent, doing justice, aiding the poor or aiding their former law partners-in short, everything but consciously promoting efficiency. Landes and Posner concede this point, for they argue that People can apply the principles of economics intuitively-and thus "do" economics without knowing that they are doing it. We think that economic principles are encoded in the ethical vocabulary that is a staple of judicial language, and that the language ofjustice and equity that dominates judicial opinions is to a large extent the translation of economic principles into ethical language. (P. 23)80 Landes' and Posner's explanation, therefore, must be ideological: regardless of what judges believe they are doing, the dominant political ideology of England and America has manuevered judges into thinking that what promotes efficiency is just, consistent with previous prece79. E.g., Elliott, id. at 64-71; Clark, The Interdisciplinary Study of Legal Evolution, 90 Yale LJ. 1238, 1266-72 (1980); see also Cooter & Kornhauser, Can Litigation Im- prove the Law Without the Help ofJudges, 9 J. Legal Stud. 139, 144-45 (1980) (decisions by litigants without the help of judges are insufficient to bring about tendency towards efficient rules). Professor Priest has modified his views since his original article and is now less convinced that a clear tendency to efficiency exists. Priest, Selective Characteristics of Litigation, 9 J. Legal Stud. 399 (1980); Priest & Klein, The Selection of Disputes for Litigation, 13 J. Legal Stud. 1 (1984). Landes and Posner themselves have concluded that this literature has overstated the tendency of a common law system to produce efficient rules, although areas can be identified where such a tendency can indeed be predicted on economic grounds. Viewed as a contribution to the emergent literature on the positive economic theory of law, our finding that the public courts do not automatically generate efficient rules is disappointing, since it leaves unexplained the mechanism by which such rules emerge as they seem to have done in a number of the areas of Anglo-American judge-made law. Landes & Posner, Adjudication as a Private Good, 8J. Legal Stud. 235, 284 (1979). 80. See also p. 14 (source of norms on which conduct is deemed to be wrongful may be economic, in which case desire to punish wrongful behavior has an economic basis). 1480 COLUMBIA LAW REVIEW [Vol. 87:1447 dent, and otherwise serves all the various interests that judges over the centuries have sought to promote. Landes and Posner apparently are arguing that the dominant political ideology of common law countries creates a separation between distributive and nondistributive rulemaking that maps closely the distinction between legislative and commonlaw rulemaking: [B]ecause the courts, especially in such fields as tort law, are poorly equipped to develop and implement doctrines that will effectively redistribute wealth to particular interest groups (rather than randomly), the forces that control the political system will find it in their self-interest to give the courts the function of maximizing the size of the economic pie and to other institutions the function of rearranging the slices of the pie in accordance with the balance of political power. (P. 313) Tort law, according to Landes and Posner, is a public good (like national defense, the police force and the court system itself), created by a political system to benefit the entire community (p. 15). A wealth maximizing system of tort law would be supported by all major groups in society, because an interest group's best strategy is to support policies that will increase the wealth of society as a whole, because the members of the group can be expected to share in that increase. Hence it is consistent that the AFL-CIO should support both the minimum wage and a strong national defense. The former is a redistributive policy that favors its members; the latter is a nonredistributive policy that also benefits its members, although no more than other members of society. Tort law is the same kind of policy. (Pp. 15-16) Although Landes and Posner are attempting to show how the dominant political ideology might create a wealth-maximizing system of tort law, their argument has distinctly ideological overtones itself. Here we see most of the standard conservative ideological moves outlined at the beginning of this Article-the belief that the common law is a nonredistributive benchmark against which all regulation is to be measured, that legislative action is at the mercy of rent-seeking interest groups, that no politically active group in society has an economic interest in a particular rule of tort law, and that wealth maximization is an uncontroversial social policy with no inherent bias in favor of the rich. The very example Landes and Posner give creates a problem for their argument. The AFL-CIO might well oppose increased spending on defense if the need to control budget deficits created a choice between a new missile system and a public works project. Landes and Posner neglect the fact that even public goods have private consequences, as any congressman from a district with a defense plant will tell you. Similarly, tort law has redistributive consequences, even if these are often diffuse. Nevertheless, the distributive consequences of tort rules are predictable enough that the American Medical Associa- 19871 POSITIVE ECONOMICS 1481 tion, the nuclear power industry, and consumers' and manufacturers' lobbies have attempted to get state and federal legislatures to monkey with the rules of tort law, rather than doing the "logical" thing and seeking relief through the tax and welfare systems.81 Furthermore, Landes' and Posner's argument is embarrassingly ahistorical-neglecting the very different understandings of the respective roles of courts and legislatures that have existed between 1100 and the modem era. Even in our own era the division of functions is unrealistic, as modem attempts to modify rules of tort law attest. How are we to explain the Codification Movement, the Uniform Commercial Code and other uniform acts, the Lousiana Civil Code, workers' compensation statutes, the Federal Employer's Liability Act, and the PriceAnderson Act? There are a number of theoretical puzzles here as well. Ifjudges do not adopt rules that actually increase social wealth, but only rules that do so certerisparibus, why would legislatures not act to correct their mistakes? Indeed, given that legislative activity is a major reason why common-law rules that appear to maximize wealth might not do so in fact, legislative modification of judicial rules should be relatively commonplace. Moreover, if legislatures have incentives to create efficient tort rules, why don't they just do it themselves? Why has this task been delegated to an antimajoritarian body with no particular expertise in economics? One cannot respond that legislatures are likely to muck it up, since we have already postulated that it is in everyone's interest to have wealth-maximizing rules of tort law. On the other hand, if it is not in everyone's interest to have wealth-maxmizing rules, wealthmaximization is not an uncontroversial social policy, and why would legislatures willingly delegate the task of creating such rules to common-law judges? Even putting these difficulties to one side, Landes and Posner still have not identified a causal mechanism. What ideological force creates the unconscious delegation of functions that their theory requires? In other words, how can we be sure that judges will do what they are supposed to, especially if they are not aware of the peculiar function assigned to them by Landes and Posner? The authors recognize that, even if wealth-maximizing tort law is a public good, we have not explained the incentive of judges to cooperate in 81. Here Landes and Posner are making yet another version of the classic conservative argument against using tort law for redistributive purposes. Landes and Posner argue that it "is all the more plausible [that tort law is a public good] because it deals with activities (mainly accidents) that do not lend themselves well to redistribution in favor of politically influential interest groups" (p. 15). Have the authors never heard of products liability, landlord tenant law, medical malpractice, and employer's liability? Of course, because some groups, like the poor, have little political clout, it may be true that there is no politically influential lobbying group that speaks for them. However, this does not prove that rules of tort law do not have redistributive consequences that affect them. 1482 COLUMBIA LAW REVIEW [Vol. 87:"1447 the production of this good. To regard judges as simply the agents of legislators who have decided to provide an efficient law of torts as they have decided to provide for the national defense ignores the fact that the judicial office is hedged about with various safeguards designed to make judges independent of legislative preferences. But of course judges are not completely independent; and persons are not likely to be elected or appointed as judges if they do not share the basic values of the dominant political groupings in society. We shall not attempt in this book to develop a theory of judicial incentives and relate it to the positive economic theory of tort law, but we feel on fairly safe ground in assuming that judges are good enough agents of society's dominant groups that if an efficient system of tort law is demanded judges will supply it-although less consistently than if they were perfect agents. (P. 19) Landes and Posner offer the problem but no hint of a solution, other than to reiterate that judges must be under the thumb of the dominant political forces in society. I am unsure whether this statement is meant to be comforting or disturbing. This last point suggests a possible (and quite cynical) explanation that would solve all of Landes' and Posner's theoretical difficulties. Suppose, as critics of law and economics have been telling us for years, wealth maximization is "biased in favor of the wealthy, oblivious to questions of distributive justice, and in general disregards all human valuations or motivations that are not responsive to considerations of price, or cost, in a sense approximately measurable by methods available to economic science." 8 2 In that case, if the wealthy are in control of the executive and legislative branches (a rare occurrence in history, I admit), they can appoint judges who reflect the same biased ideology. Then they can claim that wealth maximizing rules are neutral, apolitical standards of conduct which incidentally maximize the size of the pie for everyone, and that to alter the common-law rules by legislative action would be to politicize what is neutral, apolitical, and in the best interests of all concerned. Under this view, Landes and Posner are quite correct-the law does maximize wealth, with that expression understood as a code for "aiding the interests of the privileged classes." Under this interpretation, Landes and Posner have been unwitting 3 Marxists all along.8 I mention this interpretation of Landes and Posner not to suggest that they really are closet Marxists, but to show how impoverished and incomplete their explanation of social events is. Their theory of how 82. Michaelman, A Comment on Some Uses and Abuses of Law and Economics, 46 U. Chi. L. Rev. 307, 311 (1979) (citing Baker, The Ideology of Economic Analysis of Law, 5 Phil. & Pub. Aff. 3 (1975)). 83. Cf. Horwitz, Law and Economics: Science or Politics?, 8 Hofstra L. Rev. 905, 905 (1980) ("Like vulgar Marxism, Law-and-Economics treats law as 'superstructure,' merely reflecting what is 'real' in the 'base' of economic rationality."). 1987] POSITIVE ECONOMICS 1483 tort law could become wealth maximizing is riddled with inconsistencies, unexplained factors and ad hoc assumptions. Only by recasting it as a vulgar Marxist theory does it gain any measure of plausibility at all, and certainly not the sort of plausibility that the authors would have hoped for. In sum, we must understand Landes' and Posner's theory less as a serious attempt to explain how the common law became efficient than as a method of making the world conform to their conservative ideology. The common law should be wealth maximizing, and all interest groups should recognize that it is in their best interest not to alter it; that is how it might come about that the common law is wealth maximizing. Wishes have become horses, so that the authors may gallop to their conclusions. VI. THEORIES OF LEGAL CHANGE AND THE POSITIVE THEORY OF LAW One should not gather from the above remarks that economics can tell us little about the development of common-law rules. Changes in population, technology, and economic life do affect rules of law, just as rules of law affect them in return. To these factors we must add others equally important: religion, culture, and moral belief. A theory that purports to explain the law must take into account both the ongoing evolution of legal doctrine and the historical factors that motivate legal change. The problem with Landes' and Posner's work in this regard is not that they view economics as a force for legal change, but that they have not taken the mechanisms of legal change seriously enough. Indeed, their theory of law seems inconsistent with what we already know about the way law evolves over time. Landes and Posner have hoped to find an explanation of legal rules by demonstrating that all legal rules have one good thing in common-efficiency. From that conclusion, they hope to argue that some mechanism produces that good thing in all or almost all of the rules. However, a plausible theory of legal evolution and legal change is unlikely to produce this result. A. The Misplaced Searchfor Equilibrium Landes and Posner are misled by their understanding of economic science, which for them is based upon the belief that human forces acting independently both seek and achieve efficient outcomes. If this is true for corn and soybeans, why not for legal rules? Within their world view markets are the paradigm of all human endeavor, and markets do not fail us. Their unstated line of reasoning goes something like this: markets are efficient, all human activity is market activity, legal rules are the result of human activity, therefore, legal rules are efficient. However, as George Priest has argued, the irony of this economi- 1484 COLUMBIA LA W REVIEW [Vol. 87:1447 cally influenced vision is that it is actually inconsistent with the economic theory of behavior itself. If Landes and Posner were correct: [I]t would demonstrate that in a significant area of human behavior, an economic equilibrium can result without any of the prerequisites of equilibrium: without individual maximizing decisions, without even the existence of a market or of a shadow market. Such a finding would challenge every empirical result consistent with the theory of individual behavior in markets because it would provide evidence that individual behavior is irrelevant 8to4 equilibrium. The market itself may be an epiphenomenon. A system of law does not operate in the same fashion as the corn and soybean markets. This does not mean, however, that it is totally impervious to economic factors. Rather, it means that legal rules will respond to the economic forces in place at the time the rules are created, although that response will not be of the same nature as market equilibrium. For example, Landes and Posner argue that as insurance becomes cheaper and more widely available, there is less of a need for tort law to provide insurance. Hence, there should be a move towards negligence and away from strict liability (p. 66) because "insurance through the tort system is normally more costly than market insurance" (p. 121).85 This is the mindset of persons looking for equilibrium-who believe that law is affected by economics in the manner of supply and demand curves. However, what has actually happened in the twentieth century is precisely the opposite. As insurance has become more and more widely available, people have begun to view tort law as having an insurance function, and tort law has moved from negligence to strict liability. The economic forces that have produced the growth of modem insurance markets have also contributed to the zeitgeist that allows us to see accidents in an actuarial sense, and to view tort law as a means of risk spreading. This is not to suggest that the causal connection between economic changes and human thought runs in only one direction- economic forces and ideas influence each other. Rather, the point is that the growth of insurance markets, the movement to strict liablity, and the modem actuarial vision of accidents are all manifestations of the same shift in consciousness. In contrast, Landes' and Posner's equilibrium centered view of economic influence argues that lack of developed insurance markets creates a demand for a good which strict liablity in tort supplies. This misunderstands the way in which economic forces and ideas interact with each other. There may indeed be a large scale eco84. Priest, The New Scientism in Legal Scholarship: A Comment on Clark and Posner, 90 Yale L.J. 1284, 1292 (1981). 85. See also pp. 210-12 (because early demand for maritime insurance was great, but market was underdeveloped, admiralty doctrines provided insurance component). 1987] POSITIVE ECONOMICS 1485 nomic equilibrium that leads to simultaneous changes in productive forces, legal rules and social consciousness. However, if there is, Landes and Posner are searching for it from too limited a vantage point. B. The Diachronic View of Legal Evolution It seems reasonable to accept Landes' and Posner's most basic premise that the law responds to economic forces through an evolutionary process of adaptation. Is so, the law may well have some tendency towards efficient rules, or at least towards rules that are not too inefficient.8 6 However, if there is a general tendency in the law towards efficiency, we should expect as a proof of thisfact that there will be significant numbers of doctrines that are inefficient when judged by today's standards. This seemingly paradoxical statement will make more sense if we consider the concept of an evolutionary theory in general. By "evolutionary theory," I do not mean only biological evolution, but any theory that explains change over time, whether within a species of animal life, a language, or a system of law. A theory of evolution is above all an historical theory. It attempts to describe how change occurs based upon forces that work over time. For that reason, it is always a mistake to view the results of an evolutionary process synchronically-that is, at a given slice of time. Rather, evolutionary theories must be understood diachronically-over a span of time. Landes' and Posner's book is essentially synchronic-they attempt to show that all or almost all of the doctrines of tort law as they exist at a particular moment in the 1980s are efficient. But this involves a serious misunderstanding of how evolutionary theories work. We can understand this point better if we compare legal evolution to the most familiar type of evolutionary theory-natural selection. If we viewed natural selection as Landes and Posner view legal evolution-synchronically-we would be led to a number of incorrect conclusions. To begin with, it is a fallacy to assume that if an organism has a particular feature, that feature presently assists in its survival in the environment. For example, humans have five toes on each foot, with a little toe that is barely moveable, and an appendix. These may have present evolutionary advantages, but then again, they may not. Rather, all that the principle of natural selection allows us to deduce is that a present feature may have had some relative evolutionary advantage at one point, or that it was genetically tied to another feature that had a 87 relative evolutionary advantage. 86. This concession is made with the caveat that "efficiency" is understood in more general terms not necessarily identical with wealth maximization. 87. A feature of an organism may presently serve no useful function but still remain because it has not yet become so great a hindrance that it has been bred out. It may exist in its present form because it is a vestigial organ; the feature may simply be an altered version of an originally beneficial organ that has "atrophied" over the course of 1486 COLUMBIA LAW REVIEW [Vol. 87:1447 The presence of historical remnants like vestigial organs tends to make Darwinian evolutionary theory lack confirmation when one views an organism synchronically, but actually provides strong proofs of the principle of natural selection when organisms are viewed diachronically.8 8 The same reasoning applies to an evolutionary economic the- ory of the law. If there is a general tendency for the law to become efficient over time, it does not follow that every rule or even most rules are presently efficient. It may mean only that rules that exist were (more or less) efficient at some point in time given the "environment"-the social, political and economic factors of the time, as well as the other rules in force-in which they came into being. Rules may develop and solidify at a particular time because they serve useful social functions, or because they are concomitants to other rules that when taken together as a whole, are more efficient than the alternatives presented. This suggests that new areas of law, which are steadily developing, are more likely to respond to present concerns of efficiency than well settled rules, which may have many of the quaint, puzzling, but essentially harmless features associated with vestigial organs. Of course, Landes and Posner do try to view rules historicallythey do try to show how laws change to promote efficiency 8 9-but they always assume that if there has been no change, that is because the present rule must be presently efficient. This is the synchronic fallacy, and it betrays the conservatism of the authors, whose ideology assures them (to paraphrase the old saying) that if the law didn't fix it, it wasn't broken. Thus, Landes and Posner expend great energy trying to demonstrate that rules like the defense of custom, the fellow servant rule, and the oldest rules of all-those in intentional tort, are all efficient rules. This attempt is surely misguided. To pursue the biological metatime because having it is now not particularly necessary to the organism. In the alternative, the feature may have developed a new function different from its original function and have altered itself in various ways; the confluence of different purposes the function has historically served now makes it difficult to explain how it could have arisen to serve its present function. This phenomenon is sometimes referred to by the misleading expression "preadaptation." S. Gould, Ever Since Darwin: Reflections in Natural History 107-10 (1977); S.Gould, Hen's Teeth and Horses Toes: Further Reflections in Natural History 170 (1983) [hereinafter S. Gould, Hen's Teeth and Horses Toes]. 88. No evidence for evolution pleased Darwin more than the presence in nearly all organisms of rudimentary or vestigial structures, "parts in this strange condition, bearing the stamp of unutility," as he put it .... The general point extends both beyond rudimentary structures and beyond biology to any historical science. Oddities in current terms are signs of history ....The panda's "thumb" demonstrates evolution because it is clumsy and built from an odd part .... The true thumb had been so shaped in its ancestral role as the running and clawing digit of a carnivore that it could not be modified into an opposable grasper for bamboo in a vegetarian descendant. S. Gould, The Panda's Thumb: More Reflections in Natural History 29 (1980). 89. See, e.g., p. 23 ("we regard changes in the law as important tests of the positive economic theory"). 1987] POSITIVE ECONOMICS 1487 phor further, the fellow servant rule is like a mastodon preserved in a glacier-it was rendered obsolete by workers' compensation, and, given the general trend of twentieth century tort law, there can be no question that if workers' compensation were abolished today few courts would follow the fellow servant rule in industrial accident cases. 90 The authors' attempt to explain the rules of intentional tort in terms of modem day economic conditions is similarly inappropriate. In fact, Landes' and Posner's assumptions are much less those of an evolutionary biologist than those of a creationist-one who believes that existing organisms are perfect from their inception. The evolutionary thinker knows that the best proofs of her hypothesis are the imperfections of the subjects of her study: "[r]emnants of the past that don't make sense in present terms-the useless, the odd, the peculiar, the incongruous91 are the signs of history." C. Conclusion: Bricolage and the Common Law Finally, I suggest that any evolutionary theory of intellectual production must take into account the concept of bricolage. Bricolage is a term first coined by the French anthropologist Claude Levi-Strauss to describe the manner in which primitive peoples created mythologies to explain the world to themselves. A bricoleur, in French, is a kind of handyman who does various odd jobs. When a bricoleur is called upon to fix a leaky faucet, or replace a loose shingle on a roof, the bricoleur may not have tools that are perfectly suited to the task: The bricoleur simply makes do with what is at hand. Levi-Strauss argues that this is exactly what human beings do in forming conceptual structures-they take what they are familiar with and adapt it to their present needs, later adapting that result to needs that arise in the future, and so forth. In constructing conceptual devices, says Levi-Strauss, we never write on a blank slate. We simply make do with what we have and analogize from it. Hence, our conceptual structures are the result of history and not of necessity; they are not the only possible ways of looking at things, but the result of what we happened to do with what we had at a 2 given time. 9 Defamation is a good example of bricolage in the law. The action 90. Indeed, even before the fellow servant rule was eliminated by statute in most states, common-law judges were increasingly prone to create exceptions and distinctions to mitigate its harshness. Prosser & Keeton, supra note 20, at 572; see id. at 575-76 (in remaining cases falling outside of workers' compensation statute, courts avoid application of fellow servant rule when at all possible, and some have decided to abolish it). One can only believe if the issue were litigated more often, the process of its demise would be hastened. 91. S. Gould, supra note 88, at 28. 92. C. Levi-Strauss, The Savage Mind 16-36 (1966). Biological evolutionary the- ory has its own form of bricolage. Living organisms adapt to their environment, but they do not always choose the means of adaptation that would seem most efficient if one were constructing an organism from scratch. Rather, organisms adapt their present features 1488 COLUMBIA LAW REVIEW [Vol. 87:1447 for slander developed in the ecclesiastical courts; it was imported into the common-law courts in the sixteenth century. Litigants could escape ecclesiastical jurisdiction if they could show that the slander caused temporal harm, thus paving the way for the requirement of special damages, which is still part of the law of slander today. 93 The law of libel, a product of the Star Chamber in the seventeenth century, was brought into being by the development of the printing press and the need to stamp out dissemination of seditious material. 94 Thus, libel became known as a tort involving printed material, to distinguish it from slander, which involved oral accusation, and because of its different origins, no showing of special damages was required. Later courts, attempting to make sense of the distinctions, added a complicated set of rules of pleading and proof. The doctrine of special damages, which originally served the function of establishing the jurisdiction of temporal courts, became a means of weeding out claims where no real damage had been done, or (as plaintiffs' counsel sometimes put it), a way of keeping cases from juries. 9 5 The property of bricolage requires that human conceptual structures-which include doctrines of law-are always makeshift to a large degree. They have no claim on being the best way of dealing with problems-they only represent a way that worked at a given time, given 96 what had come before. History, then, is the key to any true evolutionary theory of the common law. In an evolutionary theory, legal rules will bear the stamp of history upon them as much as the stamp of efficiency. This suggests that the proper way to look at rules is the way a paleontologist would look at a sedimentary formation: each level of rock represents a different time span, and the fossils found in each layer would be those common to that era. I have always felt that the law of torts is much like a to new needs, with the result that there are as wide a range of solutions to the problems of survival as there are species. See S. Gould, Hen's Teeth and Horses Toes 156-57. 93. Prosser & Keeton, supra note 20, at 772, 788-94. 94. Id. at 772. 95. Beginning with the 1960s, the federal courts in this country applied a constitutional gloss to the common-law rules, thus adding still another layer to the doctrines of defamation. Perhaps a complete reworking of defamation law from the ground up might have struck a better balance between first amendment interests and the protection of personal reputation, but constitutional adjudication does not operate in this fashion. Instead, the courts took the common-law framework and superimposed constitutional limitations. 96. For an expression of ideas similar to Levi-Strauss' by a legal historian, see S. Milsom, Historical Foundations of the Common Law 6-7 (2d ed. 1981). For Milsom, the law does not develop by a process of "evolution," see id. at 3-4, by which he means a teleological movement towards some overarching goal, id. at 6 ("The largest changes in the law have never been deliberate."). Rather, the law changes through the individual decisions of lawyers and clients to adapt the preexisting materials and concepts of the law to their private ends. As Milsom notes, "[t]he life of the common law has been in the abuse of its elementary ideas." Id. This is the very essence of bricolage. 1987] POSITIVE ECONOMICS 1489 sedimentary formation: the laws of intentional tort are the oldest, dating back to the middle ages, while the law of professional malpractice and common carrier liability are relics of the eighteenth and early nineteenth centuries. Much of negligence law is the product of the nineteenth and early twentieth centuries, and we have recently seen the development of new areas of the law, such as products liablity, that clearly bear the stamp of modem times. In contrast, Landes' and Posner's work offers us an economic theory of law without bricolage, without the imperfections and maladaptations that evolutionary mechanisms inevitably display. The products of evolution are too true to be good; Landes' and Posner's theory is too good to be true.