II. EARLY ANALYSES: LAW, ECONOMICS, AND JUDICIAL DECISION MAKING

The articles that mark the beginning of “new” or “modern” law and economicsFootnote ⁵—“The Problem of Social Cost” (Coase Reference Coase1960) and “Some Thoughts on Risk Distribution and the Law of Torts” (Calabresi Reference Calabresi1961)—both gave judges a central role in particular with respect to efficiency.Footnote ⁶

In “The Problem of Social Cost,” Ronald Coase explained that judges and courts were important because they “have often recognized the economic implications of their decisions and are aware (as many economists are not) of the reciprocal nature of the problem” (Reference Coase1960, p. 19)—a condition for economic efficiency. He insisted that, “from time to time” (p. 19), and even if it is “largely unconscious and certainly not very explicit” (p. 22), judges think “of the economic consequences of alternative decisions” (p. 20) and “take these economic implications into account, along with other factors, in arriving at their decisions” (p. 19). Thus, judges “compare costs and benefits” or, put in different terms, “introduce economic efficiency considerations in their deliberations” (Bertrand Reference Bertrand2015, p. 414), and it can be said “that efficiency enlightens the judge’s path” (ibid.).Footnote ⁷ For his part, Guido Calabresi explained that the judges who in the nineteenth century had chosen to apply a principle of fault liability had understood—in a “rough and ready, non-economist’s, way” (Reference Calabresi1961, p. 517)—that it was the most efficient liability rule for the structure of the economy, “that nonfault liability would deprive our land of the benefits and promises of industrial expansion” (p. 517), and “that industry was simply not ready to bear all of its costs, and that the country would in the long run be better off if it did not” (p. 517).

However, despite the importance judges had for them, neither Coase nor Calabresi entered into the details of how judges did make their decisions, or analyzed judicial behavior. For Coase, the explanation has partly to do with his conception of economics and of law and economics. Indeed, to Coase, economists should limit themselves to the analysis of the working of the economic system, a point he repeatedly stressed.Footnote ⁸ As a consequence, legal cases were important for economists to “study … both to learn about the details of actual business practices (information largely absent in the economics literature), and to appraise the impact on them of the law” (Reference Coase1996, p. 104). He looked for judicial decisions to illustrate an economic claim—namely, the reciprocal dimension of nuisances, damages, or harmful effects: “I (and no doubt others) have used the legal cases to illustrate the economic problem” (Reference Coase1996, p. 104). But, to Coase, the working of the legal system and judicial decision making did not belong to the subject matter of economics. Hence, although judges were indeed important to take into account because of the influence their decisions have on economic activities, they should nonetheless not be taken as “objects” of analysis. That was not possible because economists do not analyze non-market behaviors—and, to Coase, saying or making the law is a non-market activity—and also because, to Coase, individuals are not rational utility maximizers. He criticized the theories of utility, “which are abstract because they represent individuals as atomised rational utility maximisers (Coase Reference Coase1978, p. 244; see also Coase Reference Coase2012; and Coase and Wang Reference Coase and Wang2013). Thus, in Coase’s article, “the specific motivations and abilities of judges remain unexplained” (Bertrand Reference Bertrand2015, p. 431).

They were not an object of analysis and their behavior remained unexplained for Calabresi, too, even though he was using economics to analyze a legal problem—that is, if his analysis was of a kind that could have allowed him to analyze judicial behavior.Footnote ⁹ But Calabresi’s goal was to understand which liability rule—strict or no-fault—was to be preferred in order to minimize the costs of accidents. Or, to put it in other words, he was trying to explain which liability rule was the most efficient. He was, at least implicitly, convinced that judges made and would make efficient decisions. But he did not explain why, according to him, judges would choose to assign liability to one party or the other. He simply denied that judges could explicitly follow a “rather complicated economic theory” (Reference Calabresi1961, p. 516) or make an explicit cost-benefit calculus, and stated that judges make “guesses” that are efficient because they are “practical … men” (Reference Calabresi1961, p. 515; emphasis added). In other words, Calabresi evacuated the question of how judges make decisions and did not try to rationalize it. More broadly, this also corresponds to his skepticism towards the assumption that individuals behave rationally: “the whole ‘rational economic man’ approach strikes me as so unreal,” he wrote (Reference Calabresi1961, p. 515). This assumption could not be used to model the behavior of workers or motorists. Why could it be applied to judges? He thus reached a conclusion similar to Coase’s, and that was not so different from what the (very) few economists who studied “legal” behaviors with economic tools wrote.

It was the case of Harold Demsetz (Reference Demsetz1964) or Simon Rottenberg (Reference Rottenberg1965), to start with. Neither of them hesitated to assume that individuals behave self-interestedly and rationally—which, to them, meant maximizing utility—even when they were involved in non-market transactions. Rottenberg, in particular, used this assumption to analyze liability. Convinced that accident law could affect individuals’ behaviors—precisely because individuals react to incentives—he claimed that accidents, and the subsequent damage, could be prevented if individuals were made to pay for the costs their behavior imposed on others (Reference Rottenberg1965). What matters for our purpose is that Rottenberg referred to the role of judges and juries in the way the rule of law is applied, stressing that “the outcome of every case depends upon the subjective judgment of judge or jury” (Reference Rottenberg1965, p. 110). It was not, however, a move towards an analysis of judges’ behavior—to the contrary. Here, Rottenberg mentioned judges and juries to emphasize how different their reasoning was compared with economists’. To him, judges and juries rely on how “a hypothetical ‘reasonable man’ would act in like circumstances” (ibid.), while economists use “a different standard, and this is the standard of the maximizing man” (ibid.). Rottenberg thus did not enter into the “subjective judgement of judges.” As with Calabresi, judicial decision making remained outside of the analysis.

It also remained outside of the analysis led by Gary Becker in his 1968 article on crime. Becker was the first to assume that individuals choose to commit a crime by making, as he put it, an “economic calculus” (Reference Becker1968, p. 209).Footnote ¹⁰ But Becker did not clearly refer to the rationality of criminals or to the fact that they would maximize a utility function. Here, “rationality” simply meant comparing the costs with the benefits of a crime.Footnote ¹¹ From this perspective, the decision to commit a crime or to behave illegally depends on the costs of punishment and on the probability of being arrested and convicted. And, quite interestingly, Becker made a remark similar to the one made by Rottenberg about the role of judges and juries. He mentioned, in a footnote only, that both the probability of conviction per offense and the punishment per offense “depend on the judge, jury, prosecutor, etc., that j [the potential criminal] happens to receive” (p. 176). Later in the article, Becker added that “judges or juries may be unwilling to convict offenders if punishments are set very high” (p. 184). For which reason? Becker did not say, not trying to analyze why certain judges or other individuals involved in the process of conviction could decide differently from others.

Thus, no more than Coase, Calabresi, or Rottenberg did Becker discuss the motives of judges or how they make decisions to convict criminals. This is all the more surprising given that, obviously, how individuals are convicted, if they are, affects not only the decision to commit a crime but also the efficiency of the legal system, and Becker was precisely aiming at understanding how criminals behave as well as determining “how many resources and how much punishment should be used to enforce different kinds of legislation” (p. 170). Isaac Ehrlich, one of Becker's students who was working on cases of illegitimate activities, adopted the same kind of approach. He did not distinguish between apprehension and conviction and did not discuss the behavior of judges and judicial decision making (see, for instance, Ehrlich Reference Ehrlich1972, Reference Ehrlich1973).

The first to discuss the behavior of one of the actors of the enforcement side of the law in terms of rational behavior was another of Becker’s students, William Landes.Footnote ¹² Yet, with regards to judges, Landes followed Becker’s and Ehrlich’s, or even Rottenberg’s, path. He did not assume that judges’ behavior could affect the functioning and the efficiency of justice. He stopped just before entering the court, focusing on pre-trial settlement and on the behavior of prosecutors. Landes’s interest in prosecutors came from his dissertation—on the impact of fair employment laws on the well-being of non-whites who had been discriminated against—in which he used expected utility to model the decision of firms to comply with or violate the anti-discrimination law (see Landes Reference Landes1967). After having finished and defended his dissertation, it was in 1966 that Landes was looking for topics that could be usefully analyzed with the same type of expected utility model. He read a newspaper article on plea bargaining that pointed out that less than 10% of criminal cases went to trial, and found that this phenomenon could be expected by using the same framework as in his dissertation.Footnote ¹³ Two versions of the paper were published in 1969 and, under the title “An Economic Analysis of Courts,” in 1971. But a preliminary version was ready earlier. The paper was presented first in 1967 at the labor workshop at the University of Chicago, and, more importantly, a second time at a “Round Table on Allocation of Resources in Law Enforcement” during the 1968 annual meeting of the American Economic Association.

Analyzing “the conditions under which a pretrial settlement or trial will take place” (Landes, in Downs et al. Reference Downs, Landes, Rottenberg and Hoffman1969, p. 505), Landes made two assumptions. First, following Becker (Reference Becker1968), he assumed that the suspect (Reference Landes1968) or the defendant (Reference Landes1971) maximizes a utility function. Much more interesting and original was the second assumption Landes made and which bore on prosecutors. Let us start by noting the specific place prosecutors occupy in a legal system: they are lawyers—at least, they possess a law degree—but are employed by an office of the government. In other words, they can be viewed as bureaucrats as much as they can be viewed as lawyers. Even if, in those years, economic analyses of bureaucratic behaviors were still in their infancy, the assumption that bureaucrats, as other public officials, were rational was in the air since the origins of public choice at the end of the 1950s.Footnote ¹⁴ This may explain why Landes did not hesitate to reason in terms of rationality and in terms of utility maximization: “The basic assumption of the model is that both the prosecutor and the defendant maximize their utility, appropriately defined, subject to a constraint on their resources” (Reference Landes1971, p. 61). More precisely, the utility of prosecutors is assumed to depend on “the expected number of convictions weighted by their respective sentences” (Reference Landes1971, p. 63). This is how prosecutors are supposed to rationally choose, according to Landes’s model, between negotiating a pre-trial settlement, or not, and then offering a certain sentence to a suspect by maximizing the expected number of convictions.

Then, Landes went on, that thanks to this behavioral assumption, one would be able to “predict charges would be dismissed when the prosecutor sees little chance of conviction regardless of his resource input into the trial, or given a conviction he expects a negligible sentence” (Reference Landes1971, p. 64). Or, in other words, one may say that prosecutors do not offer the sentence they find just or unjust. There is an optimal level of conviction that depends on the amount of resources they devote to the cases. But this does not mean that these rational decisions may be detrimental to the society. On the contrary, and this is important to note, rational decision making also contributed to the efficiency of the legal system and even to the maximization of “the community’s welfare for a given resource level” (p. 63). Indeed, a sentence is the price “the community charges for various offenses” (p. 63). Therefore, a prosecutor chooses what is best to do for the society, not for his own sake. Prosecutors are not supposed to have ideological preferences. Actually, Landes did not say much about these preferences. The only precision Landes gave was that the prosecutor “prefers longer to shorter sentences” (p. 63).

III. FROM LANDES TO POSNER, FROM PROSECUTORS TO JUDGES

When his 1971 article on courts was published, Landes was working for the National Bureau of Economic Research, which he had joined in 1968. The year 1971 was also when the bureau launched a program in law and economics in which Isaac Ehrlich, Becker himself, and Posner were involved. The latter had just been hired at the Law School of the University of Chicago and had made personal acquaintance with Gary Becker. Influenced by, and following, Becker, Posner started to view an economic analysis of law as “the application of economic theory to law” (Reference Posner1971a, p. 22).Footnote ¹⁵ He started to emphasize that economics should be viewed as a “tool” (Reference Posner1971b, p. 202).Footnote ¹⁶ Precisely, one of the first applications of economics to the law that Posner made involved judges—it was his first contribution to the topic—and also bore on the efficiency of the law. In “Killing or Wounding to Protect a Property Interest” (Reference Posner1971b), Posner analyzed from an economic perspective the legitimacy of the use of deadly weapons to defend private property. On the one hand, he put the emphasis on the optimal allocation of resources: for instance, he started with the premise that “the dominant purpose of rules of liability is to channel people’s conduct, and in such a way that the value of interfering activities is maximized” (Reference Posner1971b, p. 223). On the other hand, Posner assumed the rationality of individuals involved in the decision process and, among them, judges.

But there was no reference to a possible judicial utility function that judges would maximize. Following Becker or Coase, rather than Landes, Posner interpreted “rationality” in the terms of a cost-benefit or, to use Becker’s words, as an “economic” calculus. And, as Coase had already claimed, Posner argued that judges do take into account the costs of applying a rule when they decide a case and compare them with its benefits. At the same time, it was equally a first statement about the fact that judges seem to display a certain propensity towards efficiency. That judges “are guided by concern with economic efficiency” (Reference Posner1971b, p. 223) and “think in economic terms” (p. 224) was a premise upon which the rest of Posner’s analysis rested. To be more precise, Posner admitted that judges do follow legal principles but that they also take into account the consequences of such applications: “I expect that most judges, before deciding a case, conceive it in highly practical terms.... I mean that they consider the probable impact of alternative rulings on the practical concerns underlying the applicable legal principles” (p. 208). By “impact,” Posner meant “economic impact.” And judges’ concern with efficiency is such that they even take into account administrative costs in their calculus:

In 1972, Posner gave more precise and deeper presentations of his claim. The same year, he published many articles and conceived the first edition of his masterpiece, Economic Analysis of Law. It is therefore difficult to know which work was written first. However, for the purpose of our paper, these works nicely complement each other and the order of publication is not an issue. In “A Theory of Negligence” (Reference Posner1972a), Posner explained “Judge Learned Hand’s famous formulation of the negligence standard” (p. 32) and advanced a claim about the propensity of judges to promote efficiency: “In a negligence case … the judge (or jury) should attempt to measure three things: the magnitude of the loss if an accident occurs; the probability of the accident’s occurring; and the burden of taking precautions that would avert it” (p. 32). He then concluded that “Hand was adumbrating, perhaps unwittingly, an economic meaning of negligence” (p. 32). To Posner, Hand did not only speak of the economic content of a liability rule. He also stressed that judges should make their decisions by comparing the costs and benefits of each option. In other words, he emphasized that judges should promote efficiency. And this was also what Posner himself stressed in the very first edition of Economic Analysis of Law (Reference Posner1973a, §23.1), when he insisted that one can “assume that judges make their decision in accordance with the criterion of efficiency” (p. 325; emphasis added) and wrote that “rights and liabilities continue to be assigned, on the main, on the basis of a politically neutral comparison of costs” (p. 326).Footnote ¹⁷

Thus, Posner’s primary goal was broader than analyzing how judges make their decisions or in arguing that judges were making efficient individual decisions. Posner was rather interested in the efficiency of the entire legal system. But Posner could not simply state the efficiency of the common law. He had to explain it to give his claim a certain credibility. Now, since judges are the main source of the law and of legal change in common law systems,Footnote ¹⁸ Posner naturally involved judges in the discussion and came to consider the role of judges and to analyze their behavior. In other words, Posner’s analyses of judges’ behavior were a by-product of his primary interest in the efficiency of the legal system. Once this had been done, Posner had also to explain why judges are biased towards efficiency. He then sketched—in less than ten pages—an analysis of judicial decision making, which is not without ambiguity. For the first time, he made a further logical step and tried (unsuccessfully) to explain the reasons why judges ought to foster efficient legal changes. But it is worth stressing that he came to the behavior of judges and to judicial decision making because of the possible connection between their propensity towards efficiency and “judges’ self-interest” (Reference Posner1973a, p. 325). Here, Posner was simply asking if it was possible to explain “the promotion of efficient resource use” by assuming that judges are self-interested and behave according to the standard economics paradigm.

However, this answer turned out to be negative, since a major difference exists between decisions made on markets and decisions made by judges. Comparing the behavior of judges with that of consumers, he noted that the latter are “motivated by a desire to maximize his satisfactions, a goal affected by relative costs” (p. 325) but denied that it could be the case with judges. He spoke of “the aloof disinterest of the judge” (p. 322), and insisted that the legal system was designed to guarantee the expression of such disinterest: judges are “insulate[d] ... from any pecuniary interest in the outcome of the case” (p. 325). Or, “[t]he method by which judges are compensated and the rules of judicial ethics are designed to assure that the judge will have no financial or other interest in the outcome of a case before him” (p. 322). But he also added that, in certain circumstances, when “constraints are loose” (p. 326), judges may act “as agents for carrying out the desires of the dominant political authority” (p. 326), and he seemed to suggest that judges could try to maximize their personal utility.

Thus, there was a sort of hesitation and ambiguity between two types of behaviors, self-interested and rational—but not utility maximizers—(when constraints are loose), or disinterested (when constrained by the rules of judicial ethics). But there was no ambiguity in his analysis of “the behavior of administrative agencies” (Reference Posner1972b). Indeed, Posner did not hesitate to treat those agencies as individual entities that are rational utility maximizers. He was explicit about this assumption when he wrote, for instance, that the goal of these “agencies … is assumed to be to maximize the utility of its law-enforcement activity” (Reference Posner1972b, p. 305),Footnote ¹⁹ which was not the goal judges aim to achieve. Thus, Posner drew a frontier between “judges” and “bureaucrats.” The latter are rational and self-interested and make decisions by trying to maximize their utility or their income. Judges pursue a more neutral and less self-interested goal. They try to promote economic efficiency through their decisions by comparing costs and benefits. However, at this stage, the intrinsic motivation of judicial conduct was not clear, which accounts for the ambiguity. From these early works it was not possible to understand whether judges, while promoting efficiency, behave in a disinterested way or, more prosaically, are constrained by the institutional system that prevents them from pursuing their own egoistic goals—something that judges would otherwise be inclined to do.

The lack of conclusive evidence with respect to judicial behavior equally affected his attempt to supply strong foundations for his primary theory on common law’s efficiency. Only after having achieved his major objective with respect to legal change (discussed in the next section) was Posner able to unlock this ambiguity regarding judicial behavior. It seems that, once relieved from his main concern about common law, Posner became free to solve the puzzle behind judges’ decision making.

IV. EVOLUTIONIST ANALYSES AND NEW VIEWS ON JUDICIAL DECISION MAKING

Thus, for both Landes and Posner, prosecutors (before the trial) and judges (during the trial) contribute to the efficiency of the legal system. Indeed, both Landes and Posner had in mind and were primarily concerned with the efficiency of the entire legal system, which was a consequence of rational behavior, and this led them, more or less incidentally, to analyze how judges actually behave. But their analysis left a gap between a behavioral assumption—judges are biased towards efficiency—and a conclusion—the common law is efficient—that could be, to a certain extent, disappointing. At least, this is what Paul Rubin wrote. To him, Posner “is less persuasive in his explanation of why this is so—his argument is essentially that judges may as well decide in terms of efficiency, since they have no other criteria to use. To an economist accustomed to invisible hand explanations of efficiency in the marketplace, this justification seems weak” (Reference Rubin1977, p. 51). The additional explanation, which consisted in arguing that judges could indeed make efficient decisions by comparing the costs and benefits because they are independent from political pressures (see also Landes and Posner Reference Landes and Posner1975), was equally insufficient. A more persuasive theory was thus required. But, from the preceding quotation, one understands very well that Rubin’s objective, or concern, was to demonstrate why the common law is efficient, rather than to clarify or substantiate Posner’s argument about judges. This is what he, and others after him, started to struggle with in the second half of the 1970s with evolutionary analyses of the law. Those analyses focused on the efficiency of a legal system and concluded that legal systems could be efficient even if judges were not behaving efficiently. Thus, it was not necessary to explain how judges behave to explain the efficiency of the system. Despite the apparent remoteness with our purpose—the development of economic models of judicial behavior—these works are important because they eventually set the premises for a more precise economic analysis of judicial decision making.

When he wrote “Why Is the Common Law Efficient?” (Reference Rubin1977), Rubin had already written articles on law and economics (Rubin Reference Rubin and Rottenberg1973; Kau and Rubin Reference Kau and Rubin1979). He was more particularly interested in crime and deterrence and had demonstrated that deterrence depended more on the probability of conviction than on the length of the sentence. The argument echoed and gave more empirical legitimacy to Becker’s claims about punishment.Footnote ²⁰ At the same time, this claim gave less credence to Landes’s analysis in which prosecutors have a preference for longer rather than shorter sentences and, therefore, for which the length of the sentence could be of importance. That was clearly not the case for Rubin. But this does not account for why, as with Becker or Ehrlich, James Kau and Rubin had not taken into account the role of how judges behave and the impact their behavior would have on the individuals’ participation in illegal activities. Indeed, judges should have been central in all these models on deterrence since the “probability of conviction” depends on the decision made by a judge. Yet, judges did not receive a central role in these models. The focus was put on the system and how it works rather than on individual agents.

The analysis that Rubin was developing on the efficiency of the common law extends this reasoning: the common law system does not really need judges to be efficient. Yet, how could it be possible to explain the tendency of the legal system towards efficiency without making room for the actors who were particularly important, according to Posner? Rubin solved the problem by using the biological analyses that were becoming controversially fashionable in economics after the publication of Michael Ghiselin’s bioeconomics (Reference Ghiselin1974) and even more after Edward Wilson’s sociobiology (Reference Wilson1975). Rubin had read a review of Sociobiology published in Scientific American that incited him to read the book, which then led him to biology.Footnote ²¹ The particularly favorable review, written by John Bonner, was published towards the end of 1975, the same year Kau and Rubin’s article was published. Not long after, Rubin started to write his article that gave birth to a set of works about the efficiency of the common law that “shifted from a Posnerian view that efficiency is a result of the wisdom of the judge” (Rubin Reference Rubin1982, p. 205).Footnote ²² Thus, these works shifted away from a view in which efficiency is the consequence of the voluntary action of individuals. This is exactly what a biological or, one should rather say, an evolutionary perspective says: a society does not evolve or change because of the intentional and voluntary acts of the individuals. Evolution is a process that is led by no one in particular, very much like a market process. And this was exactly what was put forward in the literature to which Rubin’s article gave birth.

The common argument was thus that the common law—as a whole, globally—is efficient because its functioning rests on an evolutionary process, based on a sort of natural selection that allows only the most efficient rules to “survive” legal change (Goodman Reference Goodman1978; Priest Reference Priest1977, Reference Priest1980; Rubin Reference Rubin1977, p. 739, and Reference Rubin1980; Terrebonne Reference Terrebonne1981). The thesis defended in these works—which became known as “selective litigation”—was that the result of the common law process depends on the behavior of litigants.Footnote ²³ This then meant, as a corollary, that judges no longer had the central role Posner had given them. The evolutionary forces that are supposed to push the system towards efficient outcomes should work independently from judges favoring efficiency. This sort of “invisible hand” applied to legal change ought to hold even if judges are “ignorant” of the consequences of their decisions (Priest Reference Priest1977, p. 72; Cooter and Kornhauser Reference Cooter and Kornhauser1980, p. 140) or “decide cases randomly” (Goodman Reference Goodman1978, p. 394). Thus, as an unsurprising consequence, these analyses do not provide any insight regarding an economic analysis of judicial decision making. John Goodman, for his part, claimed that “no particular assumption about the motivation of judges (judges may be initially neutral with regard to the issue of economic efficiency)” (Reference Goodman1978, p. 394) ought to be formalized. And, as Robert Cooter and Lewis Kornhauser made clear, by commenting on assumptionsFootnote ²⁴ that were considerably reducing “the insight and learning ability of the judiciary” (Reference Cooter and Kornhauser1980, p. 143), “[w]e make these assumptions because we wish to characterize a process of blind evolution, not because we believe that they are true or that legal evolution is blind” (Reference Cooter and Kornhauser1980, p. 143).

The reason why we linger on this debate, seemingly deviating from our initial focus, is that we believe that it was crucial in the evolution of Posner’s thinking. In fact, in 1979, Landes and Posner published a paper entitled “Adjudication as a Private Good,” in which they took a position in favor of the evolutionary theory. At the end of a very accurate analysis of the reasons motivating the existence of public judicial systems, the authors claimed that they “cannot conclude that private provision, with all its problems, is less efficient than public” (Reference Landes and Posner1979, p. 240), while at the same time they argued that judges “do not automatically generate efficient rules” (p. 284). Landes and Posner also supplied various examples (customary law in primitive societies and commercial arbitrations) of evolutionary processes in which “rules emerge by a competitive or evolutionary process without need for a formal organ to promulgate them in a deliberate or self-conscious fashion” (p. 244). They argued that “where parties can feasibly stipulate the forum, public or private, for adjudicating disputes arising between them, competition is feasible and we would expect efficient rules of substantive law to emerge” (p. 257). Although limiting the range of the evolutionary theory to this area in which private and public supplies of justice are competitive, Landes and Posner stressed that “the tendency of the common law toward efficiency is accelerated even if the judges are indifferent to the loss of business that contracting around entails” (p. 262).

We believe that once having granted a solution to the issue of common law’s efficiency and making it independent from judges’ conduct, Posner was in some sense relieved from the greater burden of linking legal change to judicial behavior. At this point, the debate became mature enough so that he could place judges at the center of his attention. Accordingly, they were no longer ancillary to other problems in Posner’s narrative. They started to be treated as autonomous “economic actors” rather than an “economic instrument” at the service of the common law.

V. WHEN RATIONALITY-AS-UTILITY-MAXIMIZATION WAS PUT AT THE CENTER OF THE ANALYSIS

If the 1970s had brought the scientific debate finally to unveil the reasons behind the efficiency of judge-made law, the following decade represents the period in which the definitive premises, for Posner (Reference Posner1993b), were made. Posner and others pushed the debate into considering judges explicitly as self-interested, rational utility-maximizer individuals for the first time (Higgins and Rubin Reference Rubin1980; Landes and Posner Reference Landes and Posner1980; Cooter Reference Cooter1983). These works are somewhat in the “middle” of the development of Posner’s view on judges, and are important steps towards the 1993 paper. These papers now place judges at the center of attention by explicitly considering incentives from an economic perspective, while in Posner’s previous articles (such as in Reference Posner1973a), the idea was just sketched. However, contrasting what Posner will do in the 1993 article with which our story ends, these papers still linked judicial behavior to legal change, although not any longer in a subordinate matter. Something was still missing in order to make Posner’s provocative claim as explicit and autonomous as in his 1993 article. Accordingly, this is the reason why we believe that these papers, although chronologically antecedent, were not equally referred to by the following literature.

Richard Higgins and Rubin were the first, in their article “Judicial Discretion,” to explicitly make and use the assumption that “judges maximize a utility function” (Reference Higgins and Rubin1980, p. 130). In a work that they conceived as a sort of reply to the previous “evolutionist” papers,Footnote ²⁵ Higgins and Rubin introduced the idea of judicial maximization by comparing long and short considerations with respect to legal process. Specifically, they claimed that if previous works “have proposed theories … which do not rely on the behavior of judges … as these theories deal with long-run equilibria, there is still room for judicial discretion in the short run” (Reference Higgins and Rubin1980, p. 130). Higgins and Rubin thus still conceive their model of judicial behavior as a derivation, a by-product of the broader debate on legal change. However, differently from before, they supply a formal model in which judges are motivated by egoistic goals: precisely, they are wishing to exploit the discretion granted to them by the legal system and maximize their chances of promotions. In their model, judges’ utility function depends on “judicial discretion” and “wealth.” Judges, it is assumed, “like to impose their values on society, which is accomplished by precedent-setting” (Higgins and Rubin Reference Higgins and Rubin1980, pp. 130–131) but do not want to be overruled and are averse to reversal, because it reduces their wealth by reducing their possibilities to be promoted. However, the empirical analysis conducted does not yield conclusive evidence that judges arbitrate between discretion and wealth. This led Higgins and Rubin to state that their “results are mainly negative” (p. 137), and Rubin, referring later to this article, to add that “[e]fforts to model or explain the behavior of such judges are notoriously unsuccessful” (Reference Rubin and Paul1983, p. 134).

Later in that year (1980), in the next issue of the Journal of Legal Studies, Landes and Posner published a paper entitled “Legal Change, Judicial Behavior and the Diversity Jurisdiction.” As in Higgins and Rubin (Reference Higgins and Rubin1980), and as they had done in their previous works, Landes and Posner focused on judicial behavior as a means to reach a broader goal: to analyze the legal process and to demonstrate the efficiency of the common law system. More precisely, the authors tried to take a further logical step in this debate: assuming that common law is efficient, the authors proposed a sort of exercise in comparative statics that “changes in economic conditions will lead to changes in common law” (Reference Landes and Posner1980, p. 367), and wished to find empirical evidence of such a claim. However, after these premises, Landes and Posner justify their focus on judicial behavior by arguing that “in a common law system … the judge … is an important agent of legal change” (Reference Landes and Posner1980, p. 368). Thus, by contrast with their previous claims, judges were an important but not unique determinant of the evolutionary process.

As mentioned above, we believe that the narrow scope assigned to judges in the evolution of common law allowed the opening of new perspectives in the investigation of their decision-making process. In fact, the authors claimed that the answers found in the previous literature—among which also Posner’s works can be included, as we have seen—on judicial incentives and behavior, although not incorrect, had not pushed the debate forward. These works, to use Landes and Posner’s words, had not sufficient “explanatory power” (Reference Landes and Posner1980, p. 368). It was to complement, or even to supplement, such inconclusive results that Landes and Posner explained that they wished to contribute “to the development of a theory of judicial behavior” (Reference Landes and Posner1980, p. 368).

Based on these premises, Landes and Posner built a model of judicial decision making in which they assumed that judges are rational—now in the sense of maximizing their utility —which led them to explicitly make a parallel with how “ordinary” individuals behave. Anticipating the title of Posner’s 1993 article, Landes and Posner wrote: “judges, like other people, are rational maximizers of their satisfaction” (Reference Landes and Posner1980, p. 369). This claim was then followed by an accurate specification (also in a formalized form) of such a model. Judges are here depicted as sensitive to economic incentives related to potential career upgrades. As a consequence, Landes and Posner assumed that judges would respond to labor-related incentives, in terms of salary and tenure: the better the working conditions in which judges operate, the better their performance (in terms of quality of their precedents). This hypothesis was then corroborated by an empirical analysis conducted on precedents produced by judges serving in federal or state courts.

Therefore, the first explicit formulation of a model, which would gain glory only thirteen years later, can be traced back to 1980.Footnote ²⁶ However, it is worth emphasizing that the difference with Posner (Reference Posner1993b) lies in the different reasons that motivated both papers. The motivation of the Landes and Posner (Reference Landes and Posner1980) paper still derives from the willingness to enrich the debate on legal change, once the problem of the efficiency of common law was solved (Landes and Posner Reference Landes and Posner1979). The perspective is abandoned by Posner in his 1993 article. The focus was now shifted to the behavior of judges in itself. Indeed, and by contrast with earlier works, Posner is almost entirely devoted to proposing “a new, positive economic theory of judicial behavior” (Reference Posner1993b, p. 1). The reference to efficiency, as mentioned above, arrives only towards the end of the article. Yet, the process was not complete. Some further steps were necessary to emancipate completely the economic analysis of judicial conduct from the study of legal systems.

One last paper belongs to this phase of the debate: Robert Cooter’s “The Objectives of Private and Public Judges” (Reference Cooter1983). As in Higgins and Rubin (Reference Higgins and Rubin1980) and Landes and Posner (Reference Landes and Posner1980), Cooter proposed in his article an economic theory of judicial behavior, which was still in some ways not completely autonomous from the necessity of linking judges’ decisions to the concept of efficiency. Even if he did not explicitly cite it in the paper, Cooter continued the discussion started by Landes and Posner (Reference Landes and Posner1979) about the boundaries between the territory that ought to be assigned to private (that is, arbitrators) and public judges. Although he realized that economic analysis is more suitable for private arbitrators, since they are more exposed to incentives, he believed that “public judges will behave much like private judges” (1983, p. 129). Also, Cooter admitted that public judges were much more constrained by the institutional “insulation” discussed above. But he nonetheless hypothesized that public judges maximize their own prestige, in the same way as private judges maximize their income. With this statement Cooter solved the original ambiguity that was at the beginning of Posner’s thought (see section II). Judges do not change their conduct according to the institutional setting: they are disinterested when constraints bind and self-interested otherwise. Public judges, just like their private colleagues—or anybody else, we claim in line with Landes and Posner (Reference Landes and Posner1980) and Posner (Reference Posner1993b)—are rational and self-interested actors wishing to maximize their own personal utility. What changes is the scope of their maximization. When constrained by the legal system, judges will not be directly maximizing their income but rather other non-pecuniary determinants of their utility such as leisure, prestige, or power—which echoed, again, Landes and Posner (Reference Landes and Posner1980, p. 369).

VI. THE FINAL STEP: JUDICIAL DECISION MAKING—A QUESTION IN ITS OWN RIGHT

Thus, in substance, a Copernican revolution had already been initiated by the aforementioned papers. It now remains to explain why it was only in 1993 that the idea of judges as economic actors truly became established in the scholarly debate. As we have shown, all the “ingredients” had already been supplied by Posner and a few others. Why did it take a decade for this idea to gain wide (although still not unanimous) consensus? In this last section we aim to explain what we believe was still missing in the first half of the 1980s for this theory to successfully emerge on its own.

From the perspective of our story, 1992 is certainly an important year. On April 7, Coase delivered the Henry C. Simons Memorial Lecture at the University of Chicago Law School. The article, published one year later (Reference Coase1993) was entitled “Law and Economics at Chicago,” a title that summarizes quite well the object of the lecture and of the paper. In this article, Coase praised Posner for having “played the major role … [i]n the development of the economic analysis of law or, as I prefer to put it, of the legal system” (p. 251). This laudatory comment came at the end of a paragraph in which Coase claimed that it was he who had been the first to “point out that the judges in their opinions often seemed to show a better understanding of the economic problem than did many economists even though their views were not always expressed in a very explicit fashion” (p. 251). Posner, added Coase, “picked up what I had said about the judges and ran with it” (p. 251). Coase was right and wrong. Indeed, as seen above, judges played an important role in his analysis, but to present Posner as a follower who picked up what Coase had said seems an exaggeration: Posner’s work on judges is of a different nature from Coase’s. As William Domnarski wrote, “Posner might have been surprised by Coase’s lecture” (Reference Domnarski2016, p. 68). Later that year, in autumn, Becker was awarded the Nobel Prize in economics. He was one among many other economists from Chicago who had received the prestigious prize before him: Friedrich Hayek (1974), Milton Friedman (1976), Theodore Schultz (1979), George Stigler (1986), Merton Miller (1990), Coase (1991), and Robert W. Fogel (1993), to whom one could add Buchanan in 1986. It is hard to tell whether or not the prize affected the image Chicago had in the public and in the discipline, but it was certainly important for the type of economic analysis of law Posner had been promoting since the early 1970s.

Whether coincidence or causality, it is when Posner wrote three articles, all published in 1993: the one on judges (Reference Posner1993b), an article about Becker (Reference Posner1993a), and one about Coase (Reference Posner1993c). What is interesting is not only that Posner praised Becker on one hand and criticized Coase on the other, but also the reason for which Posner praised Becker: for having played a major, indeed decisive, role in the “law and economics movement” (Posner Reference Posner1993a). This is important because, until that date, Posner had always presented Coase as one of the founders of law and economics (see, for instance, Reference Posner1975). He was changing his mind. Becker—and, for that matter, Jeremy Bentham—was the founder of the movement.

The main reason for which Becker was particularly important for the law and economics movement was his methodology. This echoed the criticisms precisely leveled by Posner against Coase because of his flawed methodology. Indeed, Becker was the one who had claimed that economics should be viewed as an approach, as a method that could be used to analyze any kind of behavior or social phenomena—economic theory “applies to both market and nonmarket decisions” (Becker Reference Becker1971, p. viii), or “the economic approach is clearly not restricted to material goods and wants, nor even to the market sector” (Becker Reference Becker1976, p. 6). Now, the fact that behaviors on explicit as well as on implicit markets could be analyzed by adopting the same economic approach implied that the same assumptions could be used in both cases. In other words, this meant that the assumption that individuals are rational could be transposed to non-market activities and behaviors. This is exactly what Posner stressed about Becker:

Thus, Posner replaced Coase by Becker as the founder of the law and economics movement precisely because the latter had adopted a behavioral assumption that the former had refused to adopt. And, as if to make his point clearly, Posner published his article on judges the same year.

However, we believe that some “hard evidence” of the utility-maximization process on the side of judges was still missing. Since the very beginning, Posner and the other scholars we have discussed had uniquely focused on US federal judges. This narrow perspective had some downsides. In this jurisdiction, the regulatory environment had (more or less) succeeded to isolate judges from the canonical incentives that might influence judges’ behavior. The criticism raised by a part of legal scholarship (Epstein Reference Epstein1990) reflected this situation: at best, this approach constituted a rather useless academic exercise. In this sense, even the evidence supplied by Landes and Posner (Reference Landes and Posner1980) was not conclusive. First of all, in that paper the authors were not directly testing how judges maximize their utility but rather how judicial performance—in terms of precedents production—responds to incentives in the form of better work conditions (better salary or tenure length). At the same time, their analysis did not constitute an ideal empirical design to isolate such an effect, since it was not possible to control for every possible other factor that could have a role in precedents production. In other words, studying judges belonging to different jurisdictions (with different working conditions) but deciding on different issues could not supply conclusive evidence.

The “smoking gun” was found in 1988. In that year an exogenous event set the stage for a natural experiment to be exploited by scholars: the Federal Sentencing Guidelines. In 1984 US Congress established the US Sentencing Commission, an independent body with the task of writing “guidelines” for federal judges’ decisions in criminal cases. The main objective of this reform was to make criminal sentencing more predictable and thus limit the discretion of judges. The guidelines were enacted on November 1987. As a response to this threat to judicial independence, over 200 district court judges ruled on the constitutional legitimacy of the guidelines between January and June 1988. This set the ideal stage for a natural experiment: a sufficiently vast number of judges belonging to the same jurisdiction but characterized by different personal situations, ruling on the exact same issue in a short timespan.

Mark Cohen was the first to exploit this opportunity with “Explaining Judicial Behavior,” an article published in 1991. In this sense, it is ironic to highlight how the first successful attempt to find empirical evidence of judges’ utility maximization derived from a context in which judges were reacting to a reform intended to limit their discretion. The author starts with a statement from the 1986 edition of Posner’s Economic Analysis of Law and embraces the idea that judges are maximizing their utility and that “at the margin, judges are likely to act in ways to promote their own self-interest” (Reference Cohen1991, p. 184). However, Cohen does not develop a general theory but rather focuses on precise judge-specific factors that might influence judicial decisions: reputation among peers, career opportunities, or workload magnitude. The added value of his work relies on the fact that he was then able to test his hypotheses in an ideal situation: since all judges were ruling on the same issues, differences in their decisions might be more easily ascribed to the incentives to which they were subject. For example, he predicted that a judge with greater promotion potential would be more likely to rule a decision that could favor her chances of professional upgrade. His estimates allowed him to conclude that “some further evidence has been provided that the utility maximizing framework is useful in analyzing judicial behavior” (Reference Cohen1991, p. 198).

At this point all the pieces were in the right place. Judicial conduct was no longer considered as instrumental to the promotion of efficiency but instead raised to an autonomous topic of research. At the same time, the criticisms raised especially by legal scholars had been refuted by empirical evidence. The times were ready for Posner, who meanwhile was appointed chief justice of the Seventh Circuit US Court of Appeal, to launch his well-known paper.