1 The word “expected” indicates that a probability measure is multiplied by the dollar value of an outcome. Here the expected cost of a crime is the probability of detection, apprehension, and conviction times the monetary value of the criminal sanction (e.g., a fine or the opportunity cost of being incarcerated). See Robert D. Cooter & Thomas S. Ulen, Law and Google Scholar

Economics 515-32 (Chicago: Scott, Foresman & Co., 1988), for a summary of the literature on the economics of crime and punishment.Google Scholar

2 The hallmark of rationality, as taught to economics graduate students, is that consumers have transitive preferences. This means that if, at a given moment of time, some good or bundle of goods denoted A is preferred to another good or bundle of goods denoted B and B is preferred to a third good or bundle of goods denoted C, then it must be the case that A is preferred to C. By contrast, if it were the case that A is preferred to B, B is preferred to C, and C is preferred to A, we would find that distinctly odd indeed, irrational. This definition of rationality as the transitivity of preferences seems so unobjectionable that most economists never doubt it and are puzzled by those who do. Moreover, in a famous article, “Irrational Behavior and Economic Theory,” 70 J. Pol. Econ. 1 (1962), which most graduate students read as part of the first-year microeconomic theory course, Gary becker showed that even if there were consumers who behaved irrationally, in the sense of having intransitive preferences, the standard predictions of price theory (as micróeconomic theory is sometimes called) would still hold. Since that article, there have been more formal demonstrations that the conclusions of price theory and of welfare economics are not much affected by the presence of even a large number of consumers with intransitive preferences.CrossRefGoogle Scholar

As to the economic behavior of the suppliers of goods and services, the general presumption is that they are rational profit maximizers. If they are not, then either they will go out of business or be taken over by those who are.Google Scholar

3 Ronald Ehrenburg & Ronald Smith, Modem Labor Economics (4th ed. Chicago: Scott, Foresman & Co, 1991).Google Scholar

4 William Landers, “Deterrence,”Pub. Interest, Summer 1973, at 119-20.Google Scholar

5 Frederick M. Scherer, Industrial Market Structure and Economic Performance (2d ed. Boston: Houghton Mifflin, 1980).Google Scholar

6 See Griffin, J. M. & Gregory, P. R., “An Intercountry Translog Model of Energy Substitution Responses,” 66 Am. Econ. Rev. 845 (1976); V. B. Hall, “Industrial Sector Interfuel Substitution following the First Major Oil Shock,” 12 Econ. Letters 377 (1983); and Richard B. Howarth & Lee Schipper, “Manufacturing Energy Use in Eight OECD Countries: Trends through 1988” (Working Paper, Organization for Economic Cooperation and Development, 1991).Google Scholar

7 I shall very briefly describe one of these market failures (viz., public gods) below.Google Scholar

8 See Harvey Leibenstein, Beyond Economic Man 48-67 (Cambridge, Mass.: Harvard University Press, 1976).Google Scholar

9 This is the familiar difference between movements along and movements of a demand curve.Google Scholar

10 The five chapters are entitled “Pari-mutuel Betting Markets,”“Calendar Effects in the Stock Market,”“A Mean Reverting Walk down Wall Street,”“Closed-End Mutual Funds,” and “Foreign Exchange.” For example, chapter 11 reports on work by Professors Lakonishok, Haugen, Levi, and Smidt showing a peculiar pattern of returns on the stock market: “Abnormal price returns occur around the turn of the year, the turn of the month, the turn of the week, the turn of the day, and before holidays” (at 148). No rational explanations for these pattern have yet been found.Google Scholar

11 “Nonrivalrous consumption” means that if A is consuming a public good, that leaves no less of the good for B, C, and others to consume. National defense is the classic example. The high costs of excluding nonpaying beneficiaries are illustrated by a fireworks display.Google Scholar

12 “Two ingredients are necessary to produce a convincing anomaly: a theory that makes crisp predictions and facts that contradict those predictions” (at 2).Google Scholar

13 The experiments are reported in Marwell, Gerald & Ames, Ruth, “Economists Free Ride, Does Anyone Else! 25 J. Pub. Econ. 295 (1981). Thaler's chapter 2 cites numerous other experiments of this type.CrossRefGoogle Scholar

14 Andreoni, James, “Why Free Ride? Strategies and Learning in Public Goods Experiments,” 37 J. Pub. Econ. 291 (1988).CrossRefGoogle Scholar

15 1 refer to the literature on tit-for-tat. This is a strategy for dealing with the famous prisoner's dilemma. That dilemma arises in situations in which it would, in fact, be better for two people to cooperate than not but in which each person may be driven by a fear of the other person's defecting from an agreement to cooperate to cheat, rather than to cooperate. The game has been explored in single-play and repeated-play settings. The general conclusions have been that in a single-play prisoner's dilemma, it pays each player to cheat. In a repeated-play setting in which the number of plays is known beforehand, it also pays to cheat from the beginning. In a repeated-play setting in which the number of times the game will be played is not known when the game begins, it pays to establish a reputation as someone who keeps his agreements, until the final play is announced, and from that point on it pays both players to cheat. See, eg., Robert D. Cooter & Thomas S. Ulen, Law and Economics ch. 2 (2d ed. New York: Harper/Collins, 1994). In The Evolution of Cooperation (New York: Basic Books, 1984) Robert Axelrod reported on a computer tournament he held to see which strategies could most effectively deal with the prisoner's dilemma. The winner was an extremely simple strategy called “tit-for-tat.” Tit-for-tat cooperates on a play of the game if the other player cooperated OR the last play and defects if the other player defected on the last play of the game. To this extent tit-for-tat seems to embody Thaler's norm of cooperation.Google Scholar

Thaler has a brief summary of the tit-for-tat literature at 13-14. He does not mention a problem that arises in implementing tit-for-tat in the real world viz., that the strategy is too volatile and unforgiving. It sets off a period of cheating (or noncooperation) at the slightest defection of the other side. Nor does tit-for-tat allow for the resumption of cooperation until the other side first cooperates. These and other issues in dealing with the prisoner's dilemma are discussed in Avinash Dixit & Barry Nalebuff, Thinking Strategically 89-118 (New York: W. W. Norton & Co., 1991).Google Scholar

16 Thaler at 9, quoting Hirshleifer, , “The Expanding Domain of Economics,” 75 Am. Econ. Rev. 53 (1985).Google Scholar

17 The first experiments of the ultimatum game were reported in Werner Guth, Rolf Schmittberger, & Bernd Schwarze, “An Experimental Analysis of Ultimatum Bargaining,” 3 1. Econ. Behuuiur & Org. 367 (1982).CrossRefGoogle Scholar

18 This is the conclusion of a famous paper by Arid Rubinstein, “Perfect Equilibrium in a Bargaining Model,” 50 Econometrica 97 (1982).CrossRefGoogle Scholar

19 Thus, if the stakes were $20, the mean proposal made by Player 1 was that he would keep.63 ×$20=$12.60 and that Player 2 would have 37×$20=$7.40. Players 1 apparently recognized that they held a strategically advantageous position and used that advantage, but they did not push that advantage too far.Google Scholar

20 In a further experiment the subjects were given a week to think about the division. By comparison to the group that was not allowed this extra time, those with more time apparently made slightly less generous offers of division but only slightly less so. The mean division proposed by those who were not allowed the extra time was, recall, 63-37. The mean division proposed by those who were given the extra week was 68-32 (at 23).Google Scholar

What would be fascinating to know is whether Players 2 would reject or accept a one sided proposal by Player 1 in Player 2's favor. That is, suppose Player 1 proposed, “I take $1; you take $19.” Rationally self-interested Players 2 should, of course, uniformly accept the proposal, but I am not sure that they would. To my knowledge, this sort of experiment has not been performed.Google Scholar

21 Guth, Schmittberger, and Schwarze had concluded that game theory is “of little help in explaining ultimatum bargaining behavior” (at 25).Google Scholar

22 See Kahneman, Daniel, Knetsch, Jack & Thater, Richard, “Fairness as a Constraint on Profit Seeking: Entitlements in the Market,” 76 Am. Econ. Rev. 728 (1986), and id., “Fairness and the Assumptions of Economics,” 59 J. Business S285 (1986).Google Scholar

23 Binmore, Ken, Shaked, Avner & Sutton, John, “Testing Nonco operative Bargaining Theory: A Preliminary Study,” 75 Am. Econ. Rev. 1178 (1985).Google Scholar

24 This style of backward induction (sometimes referred to as “looking forward and reasoning backward”) is a common technique in game theory. Here it leads to a solution to the game (viz., that Player 1 should offer a division that is equal to or trivially higher than the amount that Player 2 can assure herself in the second stage of the game) that is referred to as a “subgame perfect equilibrium.”Google Scholar

25 Binmore et d., 75 Am. Econ. Reu. at 1180.Google Scholar

26 See text at note 14 supra,Google Scholar

27 A famous pair of experiments on the Coase Theorem by Hoffman, Elizabeth & Spitzer, Matthew, “The Chase Theorem: Some Experimental Results,” 25 ]. Law & Econ. 73 (1982). and id., “Entitlements, Rights and Fairness: An Experimental Examination of Subjects' Concepts of Distributive Justice,” 14 I. Legal Stud. 259 (1985), showed the importance of instructions in affecting the outcome. In the first experiment students played a variation on the ultimatum game in which the roles of Player 1 (who, as before, initially proposed a division) and Player 2 (who, again as before, either accepted or rejected the offered division) were determined by the flip of a coin. Player 1 was asked to choose between an outcome in which she received $12 and Player 2 received nothing and one in which, if both players consented to an offered division, the two could divide $14 according to a proposal made by Player 1. The theory predicts that Player 1 will prefer the second option on the ground that she can do at least as well in that option as she can in the first option (by proposing a split of $12 for her and $2 for Player 2) and may well do better. AU the players proposed and accepted an even split of the $14. Hoffman and Spitzer's second experiment tried to explain this anomalous result. They gathered a new group of subjects and again asked Player 1 to choose between $12 and a division of $14 and Player 2 to either accept or reject the proposed division. However, there were two new conditions imposed. First, the role of Player 1 was determined either by a coin flip (as before) or by the subjects playing a simple game whose winner became Player 1. Second, the Player 1 selected by either the coin flip or the game was told either that she had “earned” the role of Player 1 or that she had been “designated” that role. These two new conditions were then varied so that four combinations resulted e.g., (1) Player 1 was deter. mined by a coin flip and was told that she had earned that role; (2) Player 1 was the winner of a simple game between the subjects and was told that she had been designated as Player 1; and so on. As in the first experiment, almost all the subjects chose the second option, that of dividing $14. But here the division was not equal. It turned out that it was not significant whether the role of the initiator (Player 1) was chosen by a coin flip or by winning a simple game. But there was a significant difference in the proposed division depending on whether Player 1 believed that she had “earned” that role or merely been “designated” for that role. In instances where Player 1 had been told that she had “earned” the role, she proposed a much more one-sided division of the $14.CrossRefGoogle Scholar

28 Note also that in both stages the initial offers were above the equilibrium amount. Werner Guth & Reiner Tietz, “Ultimatum Bargaining for a Shrinking Cake: An Experimental Analysis” (unpublished, 1. W. Goethe Universitar, 1987), and Jack Ochs 6 Alvin Roth, “An Experimental Study of Sequential Bargaining” (unpublished, University of Pittsburgh, 1988). These and other variations on ultimatum bargaining games are discussed in 1. Keith Murnighan, Bargaining Games: A New Approach to Strategic Thinking in Negotiations 103-21 (New York: Morrow, 1992).Google Scholar

29 Capen, E. C., Clapp, R. V. & Campbell, W. M., “Competitive Bidding in High-Risk Situations,” 23 J. Petroleum Tech. 641 (1971).CrossRefGoogle Scholar

30 James C. Cox & R. M. Isaac, “In Search of the Winner's Curse,” 22 Econ. Irquiry 579 (1984).CrossRefGoogle Scholar

31 An interesting question raised by the winner's curse is this: “If you are to be engaged in an auction and you know of the winner's curse, how should you bid?” As we shall see, there is no simple answer to this question.Google Scholar

32 Max H. Bazerman & William F. Samuelson, “I Won the Auction but Don't Want the Prize,” 27 J. Conflict Resolution 618 (1983), and Samuelson, William F. & Bazerman, Max H., “The Winner's Curse in Bilateral Negotiations,” 3 Res. Experimental Econ. 105 (1985).Google Scholar

33 John P. Dessauer, Book Publishing: The Basic Innoduction (3d ed. New York: Continuum, 1993).Google Scholar

34 Cassing, James & Douglas, Richard W., “Implications of the Auction Mechanism in Baseball's Free Agent Draft,” 47 Southern Econ. J. 110 (1980).CrossRefGoogle Scholar

35 Richard Roll, “The Hubris Hypothesis of Corporate Takeovers,” 59 J. Business 197 (1986).Google Scholar

36 See Capen et d., 23 J. Petrokum Tech. Google Scholar

37 1 shall develop the implications of this anomaly for the Coase Theorem in the next section.Google Scholar

38 Thaler at 63. The scholarly article on which this chapter in The Winner's Curse is based is reproduced in full in Quasirational Economics. Google Scholar

39 See William Samuelson & Richard Zeckhauser, “Status Quo Bias in Decision Making,” 1 J. Risk B Uncertainty 7 (1988).CrossRefGoogle Scholar

40 Daniel Kahneman & Amos Tversky, “Choices, Values, and Frames,” 39 Am. Psychologist 341 (1984).CrossRefGoogle Scholar

41 This is, of course, like the prediction that approximately half the subjects in the experiment just recently described would engage in a transaction. Here there is no presumption that half the drivers in New Jersey and in Pennsylvania will have the cheaper policy. Rather, it is that the distribution of the two types of policies will be proportionately the same (approximately) in the two states, even though the residents start from different default points.Google Scholar

42 John Hershey, Eric Johnson, Jacqueline Meszaros, & Matthew Robinson, “What Is the Right to Sue Worth?” (unpublished, Wharton School, University of Pennsylvania, 1990).Google Scholar

43 Thaler at 70. The aversion to loss was first proposed by Kahneman and Tversky. They asked a group of subjects to answer the following question: “You have purchased tickets to a play for $10, and after arriving at the play you discover that the tickets are lost. Would you pay an additional $10 to buy replacement tickets or go home?” They then asked a second group of subjects: “You intend to go to a play for which the tickets cost $10. Upon arriving at the theater you discover that you have lost $10 from your wallet. Would you still purchase tickets to the play or would you go home?” According to the rational-choice theory of decision making, there should be no difference in the responses of the two groups. Losing $10 in cash is identical to losing tickets whose value is $10. But Kahneman and Tversky found that while a majority of those in the first group said they would go home rather than attend the performance, 88% of those in the second group said they would attend the play. The only explanation that Kahneman and Tversky could give was that people in the first group found the loss of the tickets to have been imposed a cost on them of more than $10. See Amos Tversky & Daniel Kahneman, “The Framing of Decisions and the Psychology of Choice,” 211 Science 453 (1981). There is an excellent discussion of loss aversion and its economic implications in Robert Frank, Microeconomics and Behavior 229-31 (New York: McGraw-Hill, 1991).Google Scholar

44 Presumably an implication of the loss-aversion hypothesis is that change occurs only when the benefits of a change are significantly larger than the costs. Of course, once a change is affected, the new position becomes the status quo.Google Scholar

45 Oliver Wendell Holmes, “The Path of the Law,” 10 Haw. L. Reu. 457, 477 (1897).Google Scholar

46 It certainly does not mean that someone who prefers pizza to a hamburger at lunchtime but a hamburger to pizza at dinnertime has unstable preferences. Nor is someone irrational who switches her preferences to B and away from A on learning that A is carcinogenic but back to A when she learns that the information was erroneous.Google Scholar

47 See Amos Tversky & Daniel Kahneman, “Loss Aversion and Riskless Choice: A Reference-dependent Model,” 106 Q. J. Econ. 1039 (1991).CrossRefGoogle Scholar

48 See, e.g., Maurice Allais & 0. Hagen, eds., Expected Utility Hypotheses and the Allais Paradox: Contemporary Discussions of Decisions & Uncertainty with Allais' Rejoinder (Boston: D. Reidel Publishing Co., 1979), and Amartya Sen, “Rational Behavior,” in John Eatwell, Murray Milgate, & Peter Newman, eds., The New Palgrave: Utility and Probability (New York: W. W. Norton & Co., 1990) (“Eatwell et d, New Palgrave”).Google Scholar

49 Lichtenstein, Sarah & Slovic, Paul, “Reversals of Preference between Bids and Choices in Gambling Decisions,” 89 J. Experimental Psych. 46 (1971), and “Response-induced Reversals of Preference in Gambling: An Extended Replication in Las Vegas,” 101 J. E+'mend Psych. 16 (1973). See also Willem Wagenaar, Paradoxes of Gambling Behnvior (Hills-dale, N.J.: Lawrence Erlbaum & Associates, 1988).CrossRefGoogle Scholar

50 The figures are dramatic. Thaler reports that in a recent replication of the experiment that used the values given in text, 71% of the subjects preferred H but 67% priced L above H.Google Scholar

51 Paul Slovie & Sarah Lichtenstein, “The Relative Importance of Probabilities and Payoffs in Risk-Taking,” 78 j. Experimental Psych. (Monograph Supp., pt. 2) 1 (1968). See also id., “Preference Reversals: A Broader Perspective,” 73 Am. Econ. Rev. 596 (1983).Google Scholar

52 An alternative view is that the experiments reveal that the expected-utility model is a normative model of decision making under uncertainty, not a descriptive model. This normative view of expected-utility theory is, I think, the theme of much of decision theory. For an excellent introduction to how one ought to make decisions under uncertainty, see Robyn Dawes, Rational Choice in an Uncertain World (San Diego: Harcourt Brace Jovanovich, 1988).Google Scholar

53 The existence of the preference-reversal phenomenon is robust. David Grether and Charles Plott set out in 1979 to demonstrate through a series of experiments that the reversal was specious. But they merely confirmed, indeed strengthened, the evidence for the existence of the phenomenon. Grether, & Plott, , “Economic Theory of Choice and the Preference Reversal Phenomenon,” 75 Am. Econ. Rev. 623 (1979).Google Scholar

54 See note 3 supra. Google Scholar

55 A formal presentation of the independence axiom (and of the other axioms of expected utility theory) may be found in Mark Machina, “Expected Utility Hypothesis,”in John Eatwell et d., eds., The New Palgrave 79, 87.Google Scholar

56 Tversky, Amos, Slovic, Paul & Kahneman, Daniel, “The Causes of Preference Reversal,” 80 Am. Econ. Rev. 204 (1990).Google Scholar

57 Paul Slovic, Dale Griffin, & Amos Tversky, “Compatibility Effects in Judgment and Choice,” in Robin Hogarth, ed., Insights m Decision Making: Theory and Applications (Chicago: University of Chicago Press, 1990).Google Scholar

58 Economists have calculated that the purchase of the lower-priced, less efficient appliances implies a discount rate of between 45% and 130% at low energy costs and between 120% and 300% at higher energy costs. Either set of discount rates is absurdly high.Google Scholar

59 Suppose that someone is asked to choose among three options: $100 today, $200 at the end of three years, and $500 at the end of five years. The accepted method of making this comparison is to discount the future rewards to present value. That is, in essence, the same as asking, How much money must I invest today at the prevailing rate of interest in order to have $200 at the end of three years or $500 at the end of five years? Once having reduced these future values to present values, one then chooses whichever option has the greatest present value (subject, naturally, to one's tastes about risk). The thrust of Thaler's report is that, although this method of making comparisons among present and future values is widely accepted among specialists, it is not an accurate description of how people actually behave when asked to make these comparisons. They apparently use lower discount rates for higher future values and higher discount rates for more distant rewards.Google Scholar

60 Thaler at 98. Professor Robert H. Strotz first discussed this myopia, as he felicitously termed it, in a famous article published in 1955: “Myopia and Inconsistency in Dynamic Utility Maximization,” 23 Rev. Econ. Stud. 165 (1955).CrossRefGoogle Scholar

61 See, e.g., Jon Elster, Ulysses and the Sirens (Cambridge: Cambridge University Press, 1979); Shelling, Thomas, “Self-Command in Practice, in Policy, and in a Theory of Rational Choice,” 74 Am. Econ. Rev. 1 (1984); and George Loewenstein & Jon Elster, eds., Choice over Time (New York: Russell Sage Foundation, 1992).Google Scholar

62 Hersh Shefrin & Richard H. Thaler, “The Behavioral Life-Cycle Hypothesis,” 26 Econ. inquiry 609 (1988). The notion of mental accounting and how that affects savings-consumption decisions is further explored in chapter 9 of The Winner's Curse. Google Scholar

63 3 J. Law & Econ. 1 (1960). See Ulen, “Flogging a Dead Pig: Professor Posin on the Coase Theorem,” 38 Wayne L. Reo. 91, 91-92 (1991), for a discussion of the impact of the Coase Theorem on legal scholarship.Google Scholar

64 See, e.g., the discussion in Cooter & Ulen, Law and Economics ch. 4 (2d ed. cited in note 15).Google Scholar

65 Bob Cooter first raised this possibility and called it the Normative Hobbes Theorem. See Cooter, , “The Cost of Coase,” 11 J. Legal Stud. 1 (1982), and the discussion in Cooter & Ulen, taw and Economics ch. 4 (2d ed. 1994).CrossRefGoogle Scholar

66 Of course, many economists have long recognized that there is no uniquely efficient distribution of even a fixed amount of two goods among two consumers. It all depends on the starting point or initial endowment. See Thomas S. Ulen, “An Economic Appreciation of the Bill of Rights: The Limits and Potential of Law and Economics in Discussing Constitutional Issues,” 1992 U. III. L. Rev. 189, 191-98.Google Scholar

67 There has not been as much controversy about where the starting point or default rule should be. For example, in the literature on insider trading briefly discussed in the text, both sides seem to agree on starting from the rule that insider trading is illegal, but then differ on whether anyone should be allowed to opt out of this prohibition. On a broader view of these sorts of issues, one can well imagine a trade-off between the choice of starting point and the issue of mandatory versus opt-out. See Symposium, “Contractual Freedom in Corporate Law,” 89 Colum. L. Rev. 1395 (1989), and Frank Easterbrook & Daniel Fischel, The Economic Structure of Corporation Law (Cambridge, Mass.: Harvard University Press, 1991) (“Easterbrook & Fischel, Economic Structure”).Google Scholar

68 For a summary of the literature that leans toward broader allowances for insider trading, see Easterbrook & Fischel, Economic Structure 253-75.Google Scholar

69 For an excellent introduction, see Peter Asch, Consumer Safety Protection: Putting a Price on Life and Limb (Oxford: Oxford University Press, 1988). There is an excellent summary of cognitive imperfections and their effect on economic decision making at 70-100. The bibliography, at 151-166, is also highly recommended.Google Scholar

70 For studies that are persuasively critical of current risk regulation along these lines, see W. Kip Vicusi, Smoking: Making the Risky Decision (Oxford: Oxford University Press, 1992) (“Vicusi, Smoking”); id., Fad Tradeoffs: Public and Private Responsibilities for Risk (Oxford: Oxford University Press, 1992); and Stephen Breyer, Breaking the Vicious Circle (Cambridge: Cambridge University Press, 1993).Google Scholar

71 W. Kip Viscusi, “The Value of Risks to Life and Health,” 31 J. Econ. Lit. 1912, 1912–13 (1993).Google Scholar

72 See, eg., Viscusi, Smoking 37. One of the reforms for which Viscusi has argued is that the government's creation of a uniform vocabulary for talking about risk. This is, I believe, a marvelous suggestion that promises substantial social benefits. See Viscusi, Product Risk Labeling (Washington, D.C.: AEI Press, 1993).Google Scholar

73 The literature on cognition is growing very rapidly. In addition to the Thaler books and the articles and books cited in their comprehensive bibliographies, I can also recommend some others. On the implications of the recent cognitive psychological literature for studies of jury behavior, see Reid Hastie, ed., Inside the Juror: The Psychology of Juror Decision Making (Cambridge: Cambridge University Press, 1993). In the political science area, see George A. Quattrone & Amos Tversky, “Contrasting Rational and Psychological Analyses of Political Choice,” 82 Am. Pol. Sci. Rev. 719 (1988). For a general discussion of the cognition literature on the theory of explaining individual and social behavior, see the works of Jon Elster, e.g., The Cement of Society: A Study of Social Order (Cambridge: Cambridge University Press, 1989), Nuts and Bob for the Social Sciences (Cambridge: Cambridge University Press, 1990), and Solumonic Judgments: Studies in the Limitations of Rationality (Cambridge: Cambridge University Press, 1989). See also Robert Frank, Passions within Reason (New York: W. W. Norton, 1988). The recognition that there are human errors in cognition and the exercise of judgment has begun to have an impact on the business administration literature, specifically in the area of negotiation. See Max H. Bazerman & Margaret A. Neale, Negotiating Rationally (New York: Free Press, 1992) and Cognition and Rationality in Negotiation (New York: Free Press, 1991).Google Scholar

74 Friedman, Essays in Positive Economics (Chicago: University of Chicago Press, 1953). The example that Friedman gives is of an expert billiards player, who does not know the rules of physics that govern the paths that will be taken by moving objects that have collided, but nonetheless plays a superb game of billiards as if he knew those rules. Just as the expert billiards player may not know what the rules of physics are, so, Friedman argued, the economic decision maker whose behavior is investigated in economic theory does not have to know that her preferences are transitive and reasonably stable and semi-strictly quasi-convex.Google Scholar