Introduction

Judicial Reputation: A Comparative Theory

Nuno Garoupa and Tom Ginsburg. Chicago, 2015. 273 pages.

 

"The way to gain a good reputation is to endeavor to be what you desire to appear." Socrates1

"I don’t give a damn ‘bout my bad reputation." Joan Jett2

As we write this review, the United States is gripped by one of the most contentious and outlandish presidential elections in recent memory.3 Justice Antonin Scalia, one of the most conservative justices in the history of the modern Supreme Court, died in the midst of this political whirlwind,4 and the nomination process for his successor has been sucked into the vortex. President Barack Obama expressed frustration with Senate Republicans, who hold a majority and have promised not to allow a vote, hold hearings, or even meet the candidate: “At that point, the judiciary becomes a pure extension of politics. And that damages people’s faith in the judiciary.”5 By implication, the public’s view of courts matters.

Law professors Nuno Garoupa and Tom Ginsburg have published an ambitious book that seeks to account for the great div­ersity of judicial systems based, in part, on the public’s opinion of courts. The structural features of courts, such as whether the judges are permitted to (and do) publish dissents, Garoupa and Ginsburg explain, can have a significant impact on the public’s opinion of courts. Drawing on their own prodigious writings, the authors propose a reputation theory “to explain how judges respond to the incentives provided by different audiences and how legal systems design their judicial institutions to calibrate the locally appropriate balance between audiences” (p 7).6 Judges care about their reputations. And reputation serves as both cause and effect of the design of courts.

Judicial reputation in Garoupa and Ginsburg’s book operates on many levels. A judge has a reputation, but so too does a court (p 19). Reputation in their theory captures not only the public’s view of the judiciary (as a single court or as separate judges), but also judges’ opinions of each other (p 23). Finally, reputation has a coherent meaning that transcends state boundaries and cultures and operates across time.7 Judicial reputation is challenging to define and even more difficult to measure. But this is the task that the authors set for themselves.

We begin this Review by offering a description of the book. A responsible book reviewer should, at a minimum, give the reader a good feel for the authors’ project. Part I is not a substitute for reading this fine book, but hopefully will facilitate a deeper reading of the book. We then move in Part II to our evaluation. We find much to like in this book. But we also have questions about the ability of the theory to hang together in a unified manner and to do the work assigned to it. Part III considers what motivates judges. Part IV suggests an alternative account.

  • 1. Edward Parsons Day, Day’s Collacon: An Encyclopædia of Prose Quotations 789 (International Printing and Publication Office 1884) (attributing this quotation to Socrates). Professors Nuno Garoupa and Tom Ginsburg begin their book with two quotes which preview their thinking about reputation: Benjamin Franklin’s line that “[i]t takes many good deeds to build a good reputation, and only one bad one to lose it,” and Michael Iapoce’s less well-known observation that “[r]eputation is character minus what you’ve been caught doing” (p vii). Hopefully our quotes offer a similar insight to our modest project here.
  • 2. Joan Jett, Bad Reputation (Boardwalk Records 1981), lyrics available at http://perma.cc/L22C-B6PU.
  • 3. See Patrick Healy and Maggie Haberman, Donald Trump and Hillary Clinton Win Easily in New York Primary (NY Times, Apr 19, 2016), online at http://www.nytimes.com/2016/04/20/us/politics/new-york-primary.html (visited May 6, 2016) (Perma archive unavailable).
  • 4. See Wilson Andrews, Jeremy Bowers, and K.K. Rebecca Lai, How Scalia Compared with Other Justices (NY Times, Feb 14, 2016), online at http://www.nytimes.com/interactive/2016/02/14/us/supreme-court-justice-ideology-scalia.html (visited Apr 29, 2016) (Perma archive unavailable) (showing, based on work by Professors Lee Epstein, Andrew D. Martin, and Kevin Quinn, that Scalia was the second most conservative justice during nearly all of his tenure and holds roughly the same spot—second or third—among all justices who have served since 1937).
  • 5. Transcript and Video: President Obama’s Interview with NPR’s Nina Totenberg (NPR, Mar 18, 2016), online at http://www.npr.org/2016/03/18/470869897/transcript-and-video-president-obamas-interview-with-nprs-nina-totenberg (visited Apr 29, 2016) (Perma archive unavailable).
  • 6. Prior scholars have also emphasized the importance of understanding judicial behavior as a function of the different audiences judges are seeking to speak to, albeit not in the context of a reputational theory drawn as sharply as that of Garoupa and Ginsburg. See generally, for example, Lawrence Baum, Judges and Their Audiences: A Perspective on Judicial Behavior (Princeton 2006) (arguing that judicial behavior can be examined by exploring judicial relationships with various audiences, including fellow judges and the general public, but also including the social groups with which judges identify); Thomas J. Miceli and Metin M. Coşgel, Reputation and Judicial Decision-Making, 23 J Econ Behav & Org 31 (1994) (offering a model of judicial decisionmaking that includes a judge’s reputation as a factor).
  • 7. Finding coherent meaning in any comparative scholarship poses many problems, including very basic ones: What is a “court” and who is a “judge” for purposes of drawing comparisons?

I.  The Core Claim

Professors Garoupa and Ginsburg begin their book with an unusually valuable introduction to the substantive chapters which follow. The introduction is framed by two (in)famous examples of judicial action: one that strengthened a court’s reputation and one that damaged a court’s reputation. The book opens with Chief Justice John Roberts’s 2012 Obamacare decision (pp 1–2). Roberts not only joined the five-justice majority that upheld the constitutionality of the Affordable Care Act (Obamacare), but also authored the Supreme Court’s opinion.8 Roberts’s policy preference, by any conventional measure, should have been to overturn Obamacare (p 1). Thus, he must have chosen to vote against his preferred outcome in the immediate case in order to achieve other goals. Roberts’s actions appear to have strengthened the view of the Court as independent and of himself as a statesman chief in the model of Chief Justice John Marshall.9

Garoupa and Ginsburg draw a sharp and deliberate contrast with the sensational Italian murder trial of American exchange student Amanda Knox for the killing of her roommate (p 5). The Italian judiciary drew unwanted negative attention with its inconsistent rulings involving multiple courts, multiple trials, multiple findings of guilt, and ultimately a declaration of Knox’s innocence by the country’s highest court (p 5). The fallout from this incident and Italy’s woeful position on a World Bank ranking and other rankings of judicial quality (it fares worse than Haiti on at least one measure) has included calls for serious reforms to the Italian judicial system.10 Writing elsewhere, the authors warn that “[a] judiciary with a poor reputation . . . will find itself starved of both resources and respect” (p 15). The authors demonstrate that, even in two different countries, two dramatically different kinds of cases, and two starkly different court structures, judicial reputation matters. But how does it matter?

The authors argue that reputation is the joist in the construction of any judicial system and continues to play that central supporting role as courts operate and evolve (pp 7, 59–65, 188). A reputation-centered theory, therefore, can explain the range of court structures seen throughout the United States and the world (pp 28–44). Moreover, court structure itself influences how courts are perceived (pp 9–10). The result is a feedback loop between the judicial system and the internal and external perception of the system. In order to further this argument, the authors propose a reputation theory of courts and then seek to empirically test that theory (p 13).

A.    Theoretical Development

To build their theory, Garoupa and Ginsburg look to two well-established theories of political behavior generally and judicial behavior specifically: a political economy account and an institutional account. The first substantive chapter—“A Theory of Judicial Reputation and Audiences”—explicitly lays out these theoretical foundations (pp 14–49). The later chapters do not consistently refer back to this theoretical framework. But if reputation theory is indeed a theory (rather than a description), then the role of these two grand theories is important to understand.

The political economy story begins with the familiar rational actor. Judges are rational actors who make decisions that they believe will maximize the probability of the attainment of their ends (pp 22–23). Their goals include developing their reputations, because reputation has an ultimate as well as an instrumental (or intermediate) value. That is, judges, like other people, desire that (certain) audiences view them favorably. All else constant, judges will make a decision that improves their reputation over one that worsens their reputation. Judges also understand that being held in high esteem can help them attain other ends that they want (pp 4–5). Thus, a good reputation is instrumentally valuable. The idea of a judge as a rational maximizer is familiar to students of judicial behavior.11

Garoupa and Ginsburg seek to stake out an original theory by exploiting the principal-agent model in political economics (pp 59–65).12 This model has been used regularly to evaluate the structure of courts and the work of judges.13 But Garoupa and Ginsburg recast the principal-agent framework by treating reputation as the critical “interaction of an agent and audiences” (or principals) (p 6). To understand how this works, we review the basic principal-agent model as applied to courts (pp 59–60). The public wants disputes resolved and law made and/or interpreted. To get those tasks accomplished, the public creates courts and assigns to those courts the responsibility to do this work. But, like any principal who delegates responsibility to an agent, the public worries that judges will act in their own interest, rather than in the public’s interest. The public, however, has limited capacity to directly monitor courts to ensure that judges behave properly (p 15). Moreover, the public does not want (in an ex ante sense) judges to be doing exactly what the public wants (in an ex post sense), because a crucial element of good judicial behavior tends to be judicial independence from public preferences. The design problem, therefore, is that the system has to be set up so that judges have incentives to behave in the manner that the public wants, without the public being able to evaluate the decisions of the judges, and without rewarding judges based on whether the public likes the outcomes in those decisions or not (pp 60–61). The solution judicial system designers have hit upon, Garoupa and Ginsburg explain, is to create judicial accountability via what one might call low-powered evaluations; that is judicial reputation—“the stock of assessments about an actor’s past performance” (p 15)—which serves as an accurate (or “noiseless”) indicator of judicial quality (pp 23–24).14

Judicial reputation is complicated, however, by the context in which it is created and experienced. The authors explain that reputation is both individual and collective (p 22). Individual reput­ation is held by the judge while collective reputation is held by the court. However, the authors argue that individual reputation and collective reputation are not independent because courts operate as teams (p 22). This team production feature ties the reputation of judges together. As the authors put it, the “size of the pie to be divided among individual judges” is determined by collective reputation (p 23). As a result, judges care about both individual and collective reputation.

The principal-agent model explains why the principal—the public—will rely on reputation to monitor judges. And it also accounts for why judges will be attentive to their personal reputations as well as to their courts’ reputations, because of the latter’s role in the evaluation of the individual judge’s work (pp 24–25). But the principal-agent model does not fully explain the system of incentives and disincentives that also drives Garoupa and Ginsburg’s reputation theory. In order to bring those into their theory, the authors look to the role of institutional design (p 29). The authors provide what can be characterized as a new institutional account.15

Institutional theory posits that individual behavior is a product of institutional constraints. Institutional structures impact political behavior. Institutional theories of judicial behavior adopt the rational actor assumption that judges seek to achieve their individual preferences but argue that judges, in order to attain their policy preferences, must and do consider the preferences and likely actions of other relevant actors.16 Hence, institutional theorists emphasize the influence of strategic factors, such as interactions with other judges (internal dynamics)17 or reactions of external actors, most notably those with some power over judges’ work (exogenous constraints).18

Garoupa and Ginsburg use the insights of institutional theory to explain the dynamics surrounding certain judicial actions. For example, the authors analyze the decision to dissent (pp 30–35). A judge may dissent to gain individual reputation. But dissents can be costly in terms of the effort required to write them and the possibility of creating friction with colleagues. Judges, therefore, will be more likely to dissent when they may influence subsequent development of the law (p 31). But the dissent also poses costs within the court and may even harm the court’s collective reputation as a decisionmaking body. During Marshall’s leadership, the Supreme Court wrote with a single voice in order to maximize the collective reputation of the federal judiciary at a time when courts were not held in high regard (pp 32–33). In Garoupa and Ginsburg’s account of separate opinion writing, judges use institutional design governing separate opinions in order to maximize reputation (p 33).19 They offer a similar consideration of rules governing oral argument, appeals, case selection, pay, judicial discipline, and other basic features of courts (pp 29–44).

To summarize, the theoretical structure Garoupa and Ginsburg posit is a dynamic one. On one side of the dynamic, the social planner sets up judicial institutions as a function of local contexts (needs, preferences, capabilities, etc.) so that the desire for certain types of reputation will push the agents within the system to work effectively (p 49). On the flip side though, the types of reputations that the judges already have and the degree to which they are concerned about maintaining these reputations (and even improving them) will be crucial variables to help the social planner determine which institutional structures are optimal.20

B.    Empirical Examination

The authors seek to demonstrate the role of reputation through the different mechanisms theorized using political economic and institutional accounts. They explicate their theory by applying it to a set of specific issues that most judicial systems have to face, such as the extent to which judges should be permitted to take on nonjudicial roles (pp 75–97), the uses of judicial merit commissions (pp 98–140), the effects of the increased globalization of law (pp 167–78), the proliferation of rankings of the rule of law (pp 175–78), and the interaction of courts within a country (pp 141–66). The tools used to explore the issues they flag largely fall into the category of case studies.

Legal scholarship is principally motivated by a desire to situate individual cases selected based on some shared characteristic into a coherent argument. In the current book, the cases do essentially the same work. Garoupa and Ginsburg begin their first substantive chapter with the US Supreme Court decision in 2000 that decided the presidential election between then-Governor George W. Bush and Vice President Al Gore (p 14). They note Justice Sandra Day O’Connor’s subsequent assessment that the Court should have stayed out of the process because of the harm the decision did to the Court’s reputation (p 14). Like the Obama­care decision and the Knox trial, Bush v Gore21 is only one of a number of individual cases used by the authors to support their arguments, but it is the best suited to their theory. And it shows the value of using individual cases to explain difficult concepts and persuade the audience. But the choice of case also reveals how the American judicial system—in particular federal courts and, even more specifically, the US Supreme Court—frames the reputation theory here even as the authors strive for a comparative theory.

Garoupa and Ginsburg undertake a comparative institutional analysis to separate out the effects of individual reputation from the effects of collective reputation on courts. They argue that “career” judiciaries emphasize collective reputation whereas “recognition” judiciaries emphasize individual reputation (pp 29–30). The form of judiciary influences the structure of the judicial system, which in turn reinforces the type of reputation that dominates. To understand how this works, the authors compare the United States and Japan, which have historically selected judges in very different ways, reflecting cultural and political differences between the countries and sharply different judicial practices (pp 44–48). Japan has a career judiciary reflecting cultural norms favoring collective quality over individual status. The Japanese career system is bureaucratic, selects its members at a young age when they do not have strong outside reputations of their own, and thereafter deemphasizes individuality and hence individual reputation (p 45). And Japanese court practices also favor collective reputation by relying on unsigned opinions and rotating assignments such that judges (as opposed to courts) have little meaningful external identity (p 45). The United States, by contrast, has relied on a “recognition system” of judicial selection, according to Garoupa and Ginsburg; that is, judges are chosen in recognition of individual accomplishment (pp 44–45). All of the many methods of judicial selection in the American states are based on a concept of individual merit, whether evaluated by voters at the polls, elected officials at appointment, or a merit commission on selection (pp 44–45). American judges often have significant reputations even before they take on their judicial roles and subsequently sign opinions, are visible in media coverage of cases, and stake out specific positions in their rulings and public appearances (pp 44–45). Interestingly, both countries have in recent years experienced a shift in which American judges are increasingly on a judicial career path22 and Japanese judges are increasingly identified based on external evaluation (pp 47–48).

The usual explanation for whether countries have a recognition or career judiciary is history or “tradition.” The authors challenge the legal-tradition narrative by examining judicial “pockets”—exceptions to the general system in place in a specific jurisdiction (pp 50–53, 59). The presence of outliers undermines the historical inevitability at the heart of the legal-tradition account. France is a civil-law system that has a career judiciary. But its commercial courts are staffed by elected lay people (p 55). The United States is a common-law system that has a recognition judiciary. But its administrative law and military judges represent a career judiciary operating like the system in Japan (pp 57–58). These exceptions not only undercut legal-tradition theory, they also demonstrate the explanatory power of the principal-agent model when applied to courts (pp 59–65). Career judiciaries are better suited to administrative adjudication, in which the principal concern is moral hazard (p 62). Constitutional adjudication is better served by recognition judiciaries, because judges are chosen based on proxies for the judges’ preferences (and their alignment with society’s preferences) and judges disclose more information about their preferences (initially) and about their decisions (subsequently) (pp 54, 61).

In the US federal system and nearly all American states, a single court serves as the court of last resort in the jurisdiction and has both constitutional and nonconstitutional authority (p 147). By contrast, many other countries have constitutional courts separate from their courts of last resort (p 141). The constitutional court is staffed by politically appointed judges, while law courts are staffed by career judges (pp 53–54). That method of selection is consistent with the principal-agent account discussed above. But that account does not consider the interaction between the two high courts and their competition for jurisdiction and power. Reputation theory hypothesizes that the two high courts, even though they are in the same country, will develop different procedures and norms. For example, the constitutional court will be more fragmented, with separate opinions and an emphasis on individual reputation, while the supreme court will be more uniform, operating by consensus and creating collective reputation (pp 146–49). The authors capitalize on the differences to examine the interaction between constitutional courts and supreme courts within twenty-two countries that have both (pp 154–74). They find that the greater the disagreement within a constitutional court, the higher the probability of conflict between the constitutional court and the supreme court (p 155). While the sample size is admittedly too small to draw strong conclusions, the authors strengthen their claim of a causal relationship by focusing on specific instances of conflict (pp 156–64).

Most judicial systems, and perhaps all, constrain judges from engaging in at least some nonjudicial activities (p 76). This constraint, the authors argue, can be understood from the perspective of reputational theory. Judiciaries, the argument goes, build reputations over time—specifically, reputations for matters such as impartiality, honesty, and the ability to sort out complicated facts and get close to the truth (p 81). Actors outside the judiciary, then, have an incentive to see whether they can take advantage of this built-up reputation for their nonjudicial purposes, by hiring the judges for jobs that require a credible signal of matters such as honesty and credibility (p 83). These jobs can include things such as running private arbitration, being on a board of directors, serving on an investigative commission, and even taking high political office (p 79). The risk that the foregoing poses, though, is that judges will undermine the carefully and painstakingly built reputation of the judicial system by pursuing private gain that takes advantage of what is supposed to be a public good (pp 82–83). And that, the authors argue, is why judicial systems around the globe tend to constrain nonjudicial activities (pp 76–78, 96–97).

C.    Prescriptions

Garoupa and Ginsburg embrace the responsibility to make meaningful policy recommendations based on what they have learned about how courts function. Their commitment to normative relevance is one of the many strengths of their book. Their approach to understanding courts informs their prescriptions. They contend that we can improve our courts by simultaneously incentivizing strong individual performance and encouraging effective collaboration (pp 188–97).

Garoupa and Ginsburg mention several possible reforms that would induce greater individual effort, including variable pay, transparency, and a market for judges (pp 191–93). “In an ideal world,” they say, “we would compensate judges for their marginal contribution to judicial reputation” (p 190). But variable compensation has several theoretical as well as proven difficulties. Measuring an individual judge’s performance is the most obvious challenge, but probably not the most difficult. As we have seen in other settings, pay for performance can undermine contributions to team effort and erode professionalism (pp 190–91). And, despite some apparent success in a limited, historical setting in England, pay-for-performance efforts in other countries have been heavily criticized (pp 190–91). While acknowledging these issues, Garoupa and Ginsburg make a plea for more study of how compensation structures can encourage healthy competition (p 192). They also suggest that jurisdictions should consider allowing direct competition among judges. Judges could, for example, be allowed to advertise their relative virtues to forum-shopping parties (p 192). Courts could also create special panels or assignments for which judges could compete internally (pp 192–93). We already see this practice in the United States, where, for example, federal judges position themselves for selection as a transferee judge in multidistrict litigation23 or as a member of a special assignment court or judicial conference committee.24

The authors also weigh possible reforms for inducing greater collective effort, including random de novo appeal (pp 193–94), a transnational market for judges (pp 194–95), enhanced judicial disciplinary systems (“[c]leaning [h]ouse”) (pp 195–96), and active management of the media (pp 196–97). Random audits are routinely done across industries and activities. The authors argue that even though they can find no record of its use in courts, random de novo review of a sample of cases could serve much the same purpose in the judicial system as it serves elsewhere (p 194). They also consider the suggestion of Professors Jens Dammann and Henry Hansmann that, in essence, justice-rich countries should be able to rent their judges to justice-poor countries (p 194).25 This would create a global market for courts, in which adjudication is the product and reputation is the measure of value. The benefits could be similar to those gained from open borders for other types of products.

In sum, the authors contend that reputation is a positive aspect of judicial institutions because reputation allows judges to provide information to audiences and allows outsiders to monitor judges (pp 187–88). Therefore, greater transparency, strong judicial disciplinary systems, and competition among judges would be beneficial as they would allow more information about the quality of judges and courts to be shared and create greater confidence in courts themselves.

II.  Evaluating and Testing the Core Claim

The authors’ core argument, as we see it, has three crucial steps. First, judicial systems need to have a high reputation in order to function effectively because judges will, for many of their most important decisions, lack police or military power to ensure enforcement of their dictates (p 3). Or, even if they have access to military or police power, they will risk losing it to the extent they are prosecuting the people with authority over those branches of the state.26 Second, the type of system—whether it emphasizes individual judicial reputations, collective ones, or some hybrid—that will generate the kind of reputation that the state in question needs will differ depending on context (pp 16–17). And, finally, social planners (who represent the populace), understanding these considerations, will make choices about how to structure their judicial systems as a function of what type of reputation is needed to maximize social welfare at any given point in time (p 17).

Professors Garoupa and Ginsburg’s claim is a departure from how most scholars think about the structural features of judiciaries, which is primarily as a function of historical happenstance at the starting point eons ago and inertia for the many intervening decades since then (p 18). Change, to the extent it ever occurs, is glacial. Historical origins, inertia, and path dependence, though, are but bit players in the story that Garoupa and Ginsburg tell; for them, the structures of judiciaries can and should be primarily understood through a functional lens—specifically, the lens of maximizing the type of reputation the judicial system in question most needs (pp 8–9).

Garoupa and Ginsburg deserve enormous credit for advancing a new paradigm—that judicial institutions around the world are structured and modified largely as a function of their needs for particular types of reputation—and calling into question the legal origins or path-dependence school of thought. And even if one does not buy their theoretical frame one bit, their book is a delightful read in terms of its thick description of the various characteristics of judicial systems around the world. The discussions in the chapters on judges engaging in nonjudicial activities (pp 75–97), the varying uses of judicial councils (pp 98–140), the reasons for the rise and proliferation of constitutional courts (pp 141–66), the increasing relevance of cross-country rankings of judicial quality (pp 167–86), and the varying uses of modern media to brand the court system (pp 196–97) are all gems in their own right. Further, the book is full of rich country-specific descriptions from Japan, Spain, Italy, Canada, Pakistan, and more that will be new to many readers (most were new to us).

Being academics, we of course have our quibbles, but they are more accurately put as requests to the authors that they consider explicating and extending their work in certain directions that might help further their project.

A.    A Fuller Theory, Please

Where the book falls short for us is in failing to articulate a theoretical frame strong enough to generate a wide range of testable hypotheses. To be fair, Garoupa and Ginsburg are clear in their introduction that they are not setting out a testable theory; instead, theirs is more of a description (p 13). But to give them a pass on that would be too easy, especially because they have an intriguing theoretical claim and it should be possible to articulate it in a parsimonious-enough fashion such that it can be subject to testing. As we see it, the theory could be sharpened on two definitional fronts: (1) the meaning of reputation within the Garoupa and Ginsburg framework; and (2) how it varies (in terms of what type is needed) by context.

1.   Theorizing reputation.

Garoupa and Ginsburg state at the outset that reputation is “the stock of judgments about an actor’s past behavior” (p 4). But judgments about what? And whose judgments? A court that builds the right kind of reputation, Garoupa and Ginsburg explain, is more effective—it has greater credibility and its dictates are followed (pp 2, 16). And one can imagine a self-fulfilling dynamic in which courts that have high reputations can do their work with less and less need for monitoring and policing expense and take more risks with their decisions in trying to make rules that improve social welfare, which all in turn enables even better and cheaper decisionmaking (and ultimately puts the system on a high-growth path). But we cannot test the Garoupa and Ginsburg claim unless they provide a measurable definition of reputation (which in turn would require defining the relevant audience whose judgments are to be measured).27 And this, we think, should not be an impossible task—particularly given the theoretical claims being made about the importance of reputation in determining the effectiveness of judicial systems. We note two reasons why below.

First, take the theoretical claim about the importance of reputation. There is a well-established literature from scholars like Professors Lisa Bernstein,28 Robert Ellickson,29 Avner Greif,30 Barak Richman,31 and others on how, in certain contexts, formal legal sanctions and top-down monitoring can be unnecessary because of the effective operation of nonlegal sanctions. These stories are, in effect, stories about how reputation-based systems can work effectively. And, as we see it, Garoupa and Ginsburg are in effect making the argument that reputation can play a similar role within judicial systems (albeit on a far larger scale than any of the aforementioned authors tried). The key to all of those stories of the effective operation of reputation in holding a system together, though, was the high quality of information that was available to all the relevant players.32 And, logically we think, if there is high-quality information available to all, then reputation should be measurable. Or, put differently, for a reputation-based system to work effectively, participants have to be able to calculate reputation. And that in turn should mean that it is easily measurable, at least for local participants, and, therefore, perhaps even for external observers. In small communities of traders, for example, one indicator of an individual’s loss of reputation is often the refusal of others to trade with that person or to engage in large (valuable) trades with her.33 Or it might be something more idiosyncratic, such as the denial of entry into the local temple.34 For small groups in which reputation matters crucially for the internal sanction mechanisms to work, it has to be possible for members of the group to easily identify those who have lost reputation (and conversely those who have gained it). Likewise, if reputation is crucial to the operation of courts, then judges (and interested local audiences such as lawyers and politicians) must be able to evaluate changes in a judge’s reputation.

Second, as a positive matter, the most common kind of reputation Garoupa and Ginsburg seem to be talking about—the one that gives courts legitimacy in the eyes of the public—is of the kind that comes from being willing to stop the state from harming those who lack power.35 This is a measurable characteristic. If the populace has good information about whether their judiciary is likely to stop those in the executive branch or the military from overreaching their authority, and is confident in that belief, that should be measurable via surveys—after all, the claim is that populations do have this kind of information and that their having this kind of information matters for the effective working of the system. Further, if one wanted to get more fine-grained, one could try to measure public perceptions of the quality of decisionmaking, which would include matters such as speed of disputation, as well—and this also seems to be important to the Garoupa and Ginsburg analysis (pp 23, 187–88). The quality of decisionmaking of courts or judges could be, and is already, measured in a variety of ways (citations to decisions by other courts, surveys of litigants, and so on).36 The bottom line is that a sharper definition of reputation is necessary, even if multiple kinds of reputation need to be specified. That in turn will enable testing, something that Garoupa and Ginsburg presumably want for their ambitious theory.

2.   Describing the boxes.

Assuming now that Garoupa and Ginsburg were to clarify the different kinds of reputations that they have in mind and how to measure them, the next specification question in their model is which types of contexts require which types of reputation. This second specification is needed because the core of the Garoupa and Ginsburg claim is that judicial structures are (1) initially chosen and (2) subsequently modified over time as a function of the system’s need for different types of reputation (pp 50–53). The response that Garoupa and Ginsburg might give is that the answer depends on context, and that different social, economic, and political contexts require the relevant judiciaries to pursue different types of reputations (pp 17–18, 189). But surely Garoupa and Ginsburg can give us some boxes or broad categorizations.

For example, do countries or regions with highly educated and secular populations, and with income levels above a certain amount, tend to find judiciaries more legitimate if they provide their judgments in individualized and nuanced ways? Would, for example, a 5–4 decision of the type that was handed down in Bush have undermined confidence in the judiciary to a sufficient extent to cause a structural breakdown if it had happened in a different context than the United States in 2000? Similarly, perhaps there is a set of conditions under which Garoupa and Ginsburg would predict that a population would give more legitimacy points to a judiciary that provides collectivized decisions, without any nuance or dissension, than it would to a judiciary that showed dissension among the individual judges. They seem to hint as much, to our reading, but do not sketch out the categories or boxes of circumstances.

Once those different boxes or sets of stylized circumstances are set forth, Garoupa and Ginsburg could then explain, drawing from their theory, why those particular contexts call for particular types of reputation-generating structures. For example, is there a prediction regarding the kind of reputation a judicial system needs in a newly formed nation that has not inherited, from its prior incarnation, well-established institutions? And does this need vary as a function of the degree of overall wealth or wealth disparities or racial and ethnic diversity in a society?

Another set of factors that might impact an audience’s interpretation of a court’s decision has to do with the structural features of the court. Garoupa and Ginsburg, as we read them, would say that an individualistic system like the US Supreme Court will produce a different level of reputation than a collective system like the European Court of Justice (pp 28–44, 180–81). At first cut, this seems intuitively right. There are surely going to be situations—for example, when the court is telling the public that the government in question is illegitimate and needs to leave office because the evidence shows that it rigged elections—in which the public would like the members of the court (regardless of political preferences) to speak with one voice. And in a context in which these kinds of decisions need to be made frequently by the judiciary (perhaps a country with a weak rule-of-law culture and a high level of governmental corruption), one might jump to the conclusion that the judiciary should be constrained to speak with one voice. However, if one scratches the surface, it is not clear that that is the case.

Take a court whose judges often disagree and speak separately, with starkly differing views. If and when these judges are in unanimity about something, the public will notice—and particularly so if it is a matter of great public importance (like the need to throw out an illegitimate government). But is the public going to notice, or take any special meaning, from the fact that the judges are speaking in unison if the rules require that they do? We suspect that the message—and the reputational impact—of a court’s members speaking in unison will be close to zero if the court is mandated to speak with a single voice. Indeed, the differential reputational value of speaking in unison versus individually probably occurs only in systems in which the judges have the option to do either. In that context, the choice that the judges make sends a message. If the judges have no choice, then there is no message from the choice to speak with one voice.

And even if a court were speaking in unison in a context in which the judges have the option to speak individually, it is by no means clear what the message would be. Dissenting requires effort cost. Overworked courts lack time and enthusiasm for extra work. Thus, busy judges will write fewer dissents than nonbusy judges regardless of the level of underlying disagreement in their respective courts.37 In the context of a judiciary that is overburdened with work, speaking in unison might just be an indicator of the fact that the judges cannot handle their work as a group and have just agreed to agree even when they disagree so that the work can get handled more effectively. In the high-workload context, though, the occasions when the judges speak individually could plausibly be interpreted by their audiences as more meaningful.

The point of the foregoing is that unless one has a theory of why and how the relevant audience interprets judicial or court behavior, one cannot build a meaningful theory of judicial reputation—at least not one that is amenable to empirical testing. We want more meat on the bones of the core Garoupa and Ginsburg theory in order to run a horse race between it and the historical origins or inertia theory in terms of which one can better predict what we see in judicial systems around the world. More than that, it would help to understand how and why the Garoupa and Ginsburg theory is different from existing institutional theories of judicial behavior and structures.38 Institutional analysis of the kind articulated by Professors Lee Epstein and Jack Knight in their classic book, The Choices Justices Make, would likely explain Chief Justice Roberts’s decision to vote to uphold the Affordable Care Act in very much the same manner that Garoupa and Ginsburg do in their opening two pages: as a strategic decision born out of a concern about preserving the court’s position vis-à-vis the other branches of government and the public (pp 1–2).39 The difference between the Epstein and Knight variety of institutional theory and the Garoupa and Ginsburg version, though, as we understand it, is the emphasis that Garoupa and Ginsburg put on reputation. But the two sets of theories are indistinguishable unless Garoupa and Ginsburg provide a sharper and more precise definition of reputation that shows how their theory explains and predicts in a different fashion from the more traditional institutional analysis.

Below, we try to set out some ways in which the theory might be better specified and tested. First though, we think there is another gap in the theoretical frame that needs to be filled: the question of judicial motivation within the model.

B.    Where Are the Judges (and What Motivates Them)?

Missing from the Garoupa and Ginsburg framework is a theory of judicial behavior and, in particular, a story about what motivates judges to pursue reputation at an individual level. (And this is where we need to know more about what reputation is.) This is a significant gap in their story because most judicial systems around the world, almost by definition, need to be given a significant degree of independence and discretion (effectively, protection from oversight). The flip side of independence, however, is that judges have considerable room to misbehave. But assuming—we think plausibly—that they do not misbehave, at least not to the same extent that most people probably would if given that level of job security, there is a puzzle: Why do judges work as much and as well as they do, given that they have considerable leeway to shirk and misbehave?

When we began reading Garoupa and Ginsburg, we thought that maybe they were going to use the desire for reputation as the key to explaining why judges behave well, despite the relative lack of constraints on them. The reason this is important, given their claims about the overall importance of reputation as a driving force for judicial system design, is that their system depends on the individual actors within it acting in the pursuit of reputation (either individualistic or collective), without significant carrots or sticks to make them do that. So, for their overall claims to be plausible, they need to tell a persuasive story about why it is not only in the interest of the system designers to structure judicial institutions to generate reputation, but why it is plausible to think that the soldiers within these institutions (the judges) will act in ways that help achieve the system designers’ goals.

Our pointing out this gap in the analysis is meant as a friendly amendment. As a starting point, it is not plausible to us that judges might be more motivated by reputation than most of the rest of us. That is, such a claim is not plausible unless a good case could be made that social planners had constructed judicial-selection systems to select individuals who would not only be intrinsically motivated, but also motivated by the pursuit of reputation. And this cannot be any reputation, but the kind of reputation for high-legitimacy or high-quality decisionmaking that Garoupa and Ginsburg are positing. If this is plausible though, that would put the onus on Garoupa and Ginsburg to delve deeper into judicial-selection systems and tell a plausible story about how these systems operate in a manner consistent with their reputation story.

So, if one takes the different US states that have considerable variation in terms of how they select local judges (elections, merit selection, and a range of structures in between) (pp 100–01), Garoupa and Ginsburg should be able to tell a story for (1) the differences in the types of reputations different states might need as a function of context, and (2) why the variation in judicial-selection systems within the United States helps produce those different types of reputation. And then, even better, they should be able to show that when the context within a state changes (for example, when a state goes from having a relatively racially homogeneous population with an agricultural base for its economy to a relatively racially diverse population that has technology as its economic core), the judicial-selection systems also change.40

As noted earlier, though, Garoupa and Ginsburg’s theoretical frame will need to be better specified for it to be used in a horse race against the standard inertia explanation for why state judicial systems are the way they are.

C.    A Possible Testing Ground

Garoupa and Ginsburg present the US federal judiciary as the paradigm of an individualized judiciary and compare it to the more bureaucratized judicial systems in much of the rest of the world (p 28). We have no complaint with that perspective; one of the best aspects of the book for those of us who focus mostly on the US judicial system is the exposure to an analysis of a number of foreign legal systems by two leading comparatists. What we want to suggest, though, is that Garoupa and Ginsburg’s model might be usefully tested, and perhaps pushed further, if it were to be applied to a part of the US judiciary that goes largely unmentioned in their book—the state judicial systems.

In the frame of Garoupa and Ginsburg’s book, the US federal judiciary sits at the highly individualized end of a continuum that has career judiciaries, such as those in France and Japan, at the other end (p 28). But the spectrum can perhaps be extended in the other direction as well. Specifically, if compared to the US state judicial system, in which judges in many states are elected, the US federal judiciary looks to be more of a careerist or bureaucratic system (pp 100–01). After all, a judge who has to individually stand for election every few years has to make sure to develop a very different type of reputation with the voting public (arguably, a thicker and more locally oriented reputation) than a federal judge who for the most part will be largely unknown by the local public. The question, then, is whether Garoupa and Ginsburg’s functional theory can help explain why different states in the United States have their particular elected, quasi-elected, or nonelected judicial systems. If it can, Garoupa and Ginsburg will have exploded a standard assumption regarding the reason for why so many states in the United States have elected or quasi-elected judiciaries, which is that they are a product of a populist move in the early days of the nation and that once the structure was set it was extremely difficult to change.41

The foregoing assumption is one that many in the large literature on elected versus appointed judiciaries are particularly fond of for perhaps functional reasons. For those on the policy side, who are writing diatribes about why elected and quasi-elected judiciaries are obviously bad, the claim that these were the product of some bizarre historical anomaly is convenient in that it fits the claim that no one in their right mind would ever want an elected judiciary.42 Academic empiricists studying the age-old question whether elected or appointed judiciaries are best for society are, for their part, also fond of this assumption because it means they can assume that the state system in question was put in place independent of any functional reason (therefore enabling them to justify not having to do empirical corrections for the dreaded endogeneity problem).43 For our purposes, what the foregoing means is that if Garoupa and Ginsburg (or their successors) can show that the functional explanation better explains why state systems are the way they are, that will call into question a whole host of empirical studies of the state judiciaries in the United States. And that would be quite exciting. And as an aside, we suspect that a deeper understanding of the state systems might be arrived at if examined through Garoupa and Ginsburg’s functional lens. That lens says that if one looks closely enough at the various judicial systems that are supposed to be structured the way they are because of historical happenstance combined with inertia (pp 7–8), one can detect equally, if not more, plausible functional explanations.

III.  Motivating the Judges (and Doing It Better)

Given the story that Professors Garoupa and Ginsburg tell about how social planners design judicial institutions to pursue reputation, they need to also tell a story of why it is plausible to think that the key actors in their play, the judges, will cooperate in this pursuit of reputation. And this story has to come in two parts. In the individualistic systems, judges need an incentive to pursue their own personal reputations, whereas in the collective systems, judges need incentives to pursue the enhancement of the system’s reputation and put their self-enhancement aside. The question then is whether we think it plausible that judges in either system will act in these manners.

The rub here is that there does not seem to be, in either system, a clearly identifiable set of rewards that go to judges from their bosses as a reward for pursuing the relevant kind of reputation that is needed. And this is because judging, the world over, is almost always designed as a kind of priesthood. To become a judge is, for the most part, a decision to abdicate being incentivized by the things that motivate the rest of us at our jobs—more income and more opportunities for leisure.44 In theory, the judges in both systems could be motivated by the prospect of promotions to higher-status jobs.45 But, as best we can tell, there is no judicial system that Garoupa and Ginsburg have identified that makes promotions clearly and identifiably a function of the work that the judges do on the reputational front.46

For the judges in Garoupa and Ginsburg’s model to be pursuing reputation, either the individualistic or collective version, the incentives need to be intrinsic (or inculcated). And that then leads to the question whether we think it plausible that either the individualistic or collective system selects or socializes judges to pursue the relevant kinds of reputation. We briefly take each system in turn.

A.    Plausibility of the Garoupa and Ginsburg Claim

The individualistic system, of which the US federal judiciary is the prime example that Garoupa and Ginsburg use, typically selects judges from the practicing bar at more advanced ages, after these individuals have already been lawyers for a number of years (pp 44–45). If the goal was to have a set of judges who would embark on the pursuit of individual reputation of the type that Garoupa and Ginsburg posit, it would have to be that the selection system was designed to identify and select the types of individuals who were already predisposed to pursue judicial superstardom (regardless of whether it was plausible for them to achieve it). So, the question then is whether the US federal system, Garoupa and Ginsburg’s exemplar, is designed to select individuals of this type. We would like to think so—to think that the politicians doing the selection are looking to select those who will pursue reputation—and certainly there have been some exemplars such as Justice Benjamin Cardozo, Judge Richard Posner, Judge Henry Friendly, Judge Frank Easterbrook, Justice Oliver Wendell Holmes and a handful of others. And Garoupa and Ginsburg suggest that this is indeed the system we have; they see the structural feature of choosing lawyers who are midway through their legal careers as indicative of a system that seeks out the star lawyers to then make judges (pp 9, 14, 29). But as two researchers who have spent a great deal of time interviewing and studying data on US federal judges, we suspect that our politicians are not at all looking to select judges of the Cardozo, Posner, and Friendly type. They are the aberrations, who somehow got through the selection process almost by mistake. Politicians, if anything, do not seem to like them very much precisely because they have pursued judicial stardom and not the interests of the politicians (put differently, they turned out to be too damn independent and too damn smart).47 Most judges on the federal bench, we suspect, care very little about building their individual reputations in a Cardozo-Friendly-Posner fashion; they are happy to do a good job, but they are not pursuing individual stardom.

Moving to the collective-reputation systems, which we readily admit we know a lot less about than the US federal system, we suspect that the view of judges as pursuing their reputation in some other-regarding or altruistic fashion is even more implausible. In these systems, Garoupa and Ginsburg explain, selection to the judiciary tends to be at a young age, and then judges become cogs in a big bureaucratic system (p 29). Nothing that we know about the selection systems used around the world tells us that the individuals selected tend to be particularly other-regarding. Nor do we know of any plausible accounts of how these systems succeed in socializing the judges at a young age to be altruistic in terms of pursuing the system’s collective needs and abnegating their own needs for self-advancement. That is, there are no plausible accounts unless the systems are designed to identify and reward the generation of collective reputation by individual judges—but we have seen no claim to that effect.

Garoupa and Ginsburg pose their inquiry as a positive one; they are investigating why judiciaries are structured the way they are and not whether they should be structured in one way or the other (pp 7, 14–19). But we read an undertone of approval from Garoupa and Ginsburg for judicial systems that are designed to pursue reputation. Elsewhere in this Review, we have tried to push Garoupa and Ginsburg on their positive claims. But even if the positive claims turn out to be theoretically flawed (maybe judicial systems are not designed to ensure the optimal amount of reputation), Garoupa and Ginsburg, in their final chapter, do raise the question of whether social planners in the future might do well to consider reputation enhancement when they design new structures (pp 187–97). Certainly, the two of us accept the argument that a judiciary that has a high reputation will be more effective and that social planners should keep reputation generation in mind when designing judicial institutions and selecting judges. But we are skeptical that they do that very much currently, anywhere.

B.    Making It Happen

Garoupa and Ginsburg suggest some intriguing and creative possibilities that, to a limited extent, are already being utilized, but perhaps could be utilized better to enhance judicial reputation. These include improvements to brand management, better use of social media by the courts and individual judges, mechanisms to clean house when the entire judicial system of a country is corrupt, and so on (pp 195–97). The mechanism that interested us in particular is one that Garoupa and Ginsburg are rather skeptical of: the relevance of postjudgeship employment and how those employment opportunities might help incentivize judges, while they are employed as judges, to pursue the right kinds of reputational enhancements (pp 75–97).48 As a general matter, at least in the US judicial system, the phenomenon of judges retiring and going on to work in the private market is viewed with a degree of contempt (such judges are often referred to as “rent-a-judges”).49 But maybe judges should be encouraged to pursue such postjudgeship employment. Maybe it would produce good incentives to generate the kind of reputation Garoupa and Ginsburg suggest that judiciaries should be pursuing (and we are skeptical that they are). Below, we mention three sets of jobs judges might take (we draw these examples in part from Garoupa and Ginsburg), although the list is by no means exhaustive.

First, there are the jobs as independent investigators or on independent review commissions, doing things like investigating scandals and writing reports on appropriate reforms that the government should undertake (pp 79, 87, 89). These are quasi-judicial high-status jobs in which the rewards go only to those judges who have built a reputation of holding those in power to account (after all, these commissions call for “independent” review usually). These are jobs that might be particularly attractive in systems in which judges have to retire at a specified age and in which they still have some years of productive work ahead of them. The government in question, which is facing some sort of scandal, might not want the investigations to occur, but needs the public to be persuaded that a fair and impartial investigation has occurred. Here, the hiring of a judge who has built a reputation for legitimacy and fearlessness in the face of government pressure might be the best solution. And if the jobs are attractive enough, and the possibility of getting them is real, then judges will strive to generate the kind of reputation that will help them get these jobs after they take retirement. Indeed, as Garoupa and Ginsburg explain, there is sometimes a demand for judges to move from their judicial roles into political roles when the political system is seen as having become so flawed that what it needs is a leader with the kind of reputation that a career in the judiciary generates (p 75).

Second, there are the jobs that judges might perform doing so-called private judging—that is, sitting on arbitral tribunals (pp 76, 79). Arbitration, because it is private and has less access to the state’s monopoly on force, has to depend even more on the legitimacy of the decisions its judges make than does the regular, formal state system. The judges who get offered arbitrator jobs, therefore, are necessarily those who built the right kind of reputation; the measure of their reputation is the willingness of private parties to abide by their decisions. There is a fly in the ointment here, which Garoupa and Ginsburg point out, in that the presence of high-quality arbitral options may undermine the incentive for the local state-sponsored system to develop a high reputation (p 83). But this dynamic could easily cut the other way, as well; the competition from the arbitral system might make state systems work better, and particularly so if the arbitral system picks its judges from amongst those who perform the best in the state system.

Third, and less utilized than options one and two, but perhaps even more intriguing, is the possibility for judges who have built reputations in one setting—either individual or collective—to be hired by other judicial systems that want to, in effect, buy a judicial reputation (pp 194–95). For example, one might imagine, quite plausibly, a country or region that has decided that it wants to build the kind of judicial system that will attract foreign investors. One can also imagine, equally plausibly, that one of the key things that foreign investors are worried about is expropriation of their assets by some future populist government that runs on a platform of “take from the evil foreigners.”50 In this scenario, the country seeking to attract foreign investment might consider offering positions in its judiciary to judges who have built strong judicial reputations—and particularly reputations for standing up to state misbehavior—elsewhere. The Dubai International Financial Centre is one such example, in that it uses a common-law-modeled court system with judges hired from a variety of common-law jurisdictions, explicitly to attract foreign investment, while being located within the United Arab Emirates’ civil-law system.51

To be clear, Garoupa and Ginsburg do not go as far as we do in seeing the positive value for judicial-reputation creation via the incentives provided by the prospect of future employment in alternative settings. They express concern about adverse selection (that the wrong kind of person will be attracted to the judiciary) (p 82) and negative effects on reputation (perhaps if judges are seen as working harder in order to enhance their prospects for future employment, as opposed to working hard in order to pursue justice or another “pure” goal, that will cause the public to think worse of judges) (p 83). We are not so concerned. Adverse selection problems can be ameliorated by having better selection systems. And even if the wrong kinds of candidates get through (the candidates who care more about self than others), that is not such a big problem if these judges can be given the right kinds of incentives to pursue the right kind of reputation. Garoupa and Ginsburg, we think, could be more eager to embrace the possibilities for using the prospect of lucrative or prestigious postjudicial employment as a means of enhancing the incentives for judges to pursue reputation.

IV.  Alternative Functional Theories

One reason why we are a bit skeptical of the functional claims that Professors Garoupa and Ginsburg make is that we do not think that judicial systems around the globe are as yet designed to enhance reputation to the extent Garoupa and Ginsburg suggest. Plus, we think it plausible that there are alternative functional explanations for why judicial institutions are the way they are, and Garoupa and Ginsburg have not persuaded us that these alternative explanations are not also playing a role (perhaps alongside reputation). Below, we sketch out one such explanation.

Garoupa and Ginsburg tell us that the reason different judiciaries have different structural features is that these different structures are aimed at optimizing the proper amount of reputation that a particular judiciary needs to serve its society best (p 2). And Garoupa and Ginsburg’s foil here is the path-dependence or inertia explanation that dominates the literature (pp 7–8, 59). We applaud the moves both to provide a functionalist explanation and to question the standard path-dependence story for why we have the institutions we do. But what if there were other functionalist explanations—ones that did not rest on reputational considerations?

One such explanation is the holdout problem. Systems that have strong norms or requirements as to unanimous decisionmaking are by definition more concerned with holdouts, because a requirement of unanimous decisionmaking necessarily gives considerable power to anyone willing to threaten to hold out.52 By contrast, there is relatively little danger of holdouts in a highly individualized judicial system. In such systems, the judges make up their own minds and their decisions get aggregated at the end; there are no holdout problems, just winners and losers.

Once we see the decision to use one structure or the other to design a judicial system in terms of the risk of holdouts, then the next step is to ask whether different contexts call for the social planners to be more or less concerned about holdouts. Imagine, for example, a society that is highly polarized along race, religion, and income grounds. Holdouts might be highly likely in such a society in cases involving certain subject areas (for example, human or civil rights). And so one might expect that courts tackling these issues would be individualized. However, one might also imagine certain types of topics (for example, issues having to do with the banking sector’s stability) for which neither the public nor the judiciary has strong preferences. For these types of topics, the risk of a holdout might be minimal and the enhancement of the quality of the decision by forcing the members of the court to discuss and compromise might be considerable.

We are by no means certain that the foregoing holdout problem really is one that drives the difference in structures that one sees around the world. The point, though, is that it is a plausible alternate functional explanation, and it has nothing to do with reputation creation.

Conclusion: The Future (Likely) Relevance of Garoupa and Ginsburg’s Reputational Theory

Our final observation is that regardless of whether Professors Garoupa and Ginsburg’s view accurately represents the reality of why current judicial systems are structured the way they are today (we suspect that there is much more path dependence in the equation than Garoupa and Ginsburg are ready to concede), we think that the reputational model is going to become more and more important as time goes on. And that will happen not because social planners realize that judicial systems should be designed to pursue certain kinds of reputations, but because of the rapid growth in recent years in the external monitoring, evaluation, and relative ranking of judicial systems. As Garoupa and Ginsburg describe in their penultimate chapter, the past two decades have seen phenomenal growth in the number of relative rankings of judicial institutions around the world (p 176). Among the most prominent of these are those produced by the World Bank and Transparency International.53 But there are also rankings produced by academics and for-profit organizations.54 There are weaknesses in these rankings, as numerous researchers (including one of the authors of this book) have pointed out.55 But they are being constantly critiqued and improved and, most importantly, are taken seriously by the markets and therefore the countries that are being rated. Credit rating agencies use them and governments—particularly emerging-market governments—frequently feel compelled to report them to foreign investors.56 A recent bond issue by Mozambique (its first) is particularly illustrative. The country, in its bond prospectus, reported its rankings on both the World Bank’s Worldwide Governance Indicators and Transparency International’s 2014 Corruption Perceptions Index, and also discussed the rankings of neighboring countries in Africa, such as South Africa.57 Put simply, if the quality of a country’s judiciary is important to its cost of borrowing on the international markets, the country will make sure to invest in improving that system and its ranking—and that in turn means setting up the right incentives for judges to pursue the right kind of reputation. Equally important for many poor countries is the fact that development agencies often look to these indicators in determining where to allocate funding.58

But that is not all. As more and more information becomes available about individual judges and courts and their performances, it becomes easier to rate them on factors such as the speed with which they produce decisions, the rates at which those decisions are reversed by higher courts, the rates at which they are cited by other courts, the credentials of those who are selected to be judges, the satisfaction levels of those who have had cases before them, and so on. Again, there are criticisms that can fairly be leveled at all of these ratings.59 But the bottom line is that we, as a global society, are moving in the direction of being able to increasingly evaluate not only the overall quality of a judicial system, but also individual courts and judges. In turn, to the extent judges care about their evaluations (and we suspect they care, or can be made to care), the result will surely be an enhanced pursuit of the kinds of reputation contained in these measures. And that fact, Garoupa and Ginsburg posit, will affect the structuring of judicial systems.

At bottom, Garoupa and Ginsburg have made an invaluable contribution to the literature on judicial institutions not only in terms of providing an alternative paradigm for understanding why we have the institutions we do have, but also in terms of helping us predict the future shape that these institutions are likely to take.

  • 8. See generally National Federation of Independent Business v Sebelius, 132 S Ct 2566 (2012).
  • 9. See Adam Liptak, John Roberts Criticized Supreme Court Confirmation Process, before There Was a Vacancy (NY Times, Mar 21, 2016), online at http://www.nytimes.com/2016/03/22/us/politics/john-roberts-criticized-supreme-court-confirmation-process-before-there-was-a-vacancy.html (visited Apr 29, 2016) (Perma archive unavailable) (quoting Professor Akhil Amar’s assertion that by standing against the Senate’s opposition to considering a nominee, Roberts would have a “John Marshall moment,” his “third extraordinary moment of . . . showing that he is no partisan,” his first two moments being his votes to uphold Obamacare against two separate challenges).
  • 10. See, for example, Italy’s Judicial System: Justice Denied? (The Economist, July 19, 2014), archived at http://perma.cc/SCN2-8K2K (reporting on European Central Bank President Mario Draghi’s call for reforms to the Italian civil justice system); Gianluca Esposito, Sergi Lanau, and Sebastiaan Pompe, Judicial System Reform in Italy—a Key to Growth *11–14 (IMF Working Paper No 14/32, Feb 2014), archived at http://perma.cc/7SJX-ZBT5 (proposing various reforms to accelerate the pace of cases in the Italian judicial system).
  • 11. See, for example, Lee Epstein, William M. Landes, and Richard A. Posner, The Behavior of Federal Judges: A Theoretical and Empirical Study of Rational Choice 29 (Harvard 2013) (describing the economic theory of judges as rational actors and explaining how the authors’ work expands on this approach).
  • 12. For an explanation of the role of the principal-agent model in the theory of political actors and institutions, see Gary J. Miller, The Impact of Economics on Contemporary Political Science, 35 J Econ Lit 1173, 1189–93 (1997); Terry M. Moe, The New Economics of Organization, 28 Am J Polit Sci 739, 756–58, 765–72 (1984).
  • 13. For an early example, see generally Donald R. Songer, Jeffrey A. Segal, and Charles M. Cameron, The Hierarchy of Justice: Testing a Principal-Agent Model of Sup­reme Court–Circuit Court Interactions, 38 Am J Polit Sci 673 (1994).
  • 14. A high-powered incentive system would be one in which the rewards for judges are tied much more closely to their performance. For example, judges might get specific rewards based on whether the ruling elite liked their decisions in individual cases.
  • 15. See James G. March and Johan P. Olsen, The New Institutionalism: Organizational Factors in Political Life, 78 Am Polit Sci Rev 734, 741–42 (1984) (describing individual behavior as “driven by rules within a political structure” and explaining the role of “institutional rules”).
  • 16. For a collection of institutional accounts of courts, see generally Cornell W. Clayton and Howard Gillman, eds, Supreme Court Decision-Making: New Institutionalist Approaches (Chicago 1999).
  • 17. See, for example, Tracey E. George, The Dynamics and Determinants of the Decision to Grant En Banc Review, 74 Wash L Rev 213, 272–73 (1999) (concluding that an appellate court’s decision to hear a case en banc is primarily influenced by factors internal to the court system).
  • 18. See, for example, William N. Eskridge Jr, Reneging on History? Playing the Court/Congress/President Civil Rights Game, 79 Cal L Rev 613, 664–66 (1991) (discussing the dynamic interaction between the Supreme Court, the president, and Congress).
  • 19. For a similar conception of separate opinions, see, for example, Virginia A. Hettinger, Stefanie A. Lindquist, and Wendy L. Martinek, Judging on a Collegial Court: Influences on Federal Appellate Decision Making 109–17 (Virginia 2006).
  • 20. See, for example, pp 17, 49, 74 (noting both that the structures of institutions impact the texture of reputation that is generated within them and that the type of reputation that a system already has—and the preferences of the judges within it—impacts choices about modifications that need to be made to the system).
  • 21. 531 US 98 (2000) (reversing the Florida Supreme Court’s order allowing a manual recount in the 2000 presidential election).
  • 22. One indication of this is the increasing tendency in the US federal courts to make promotions, and particularly so vis-à-vis the high court, from within the system. See, for example, Epstein, Landes, and Posner, Behavior of Federal Judges at 337–41 (cited in note 11).
  • 23. See Margaret S. Williams and Tracey E. George, Who Will Manage Complex Civil Litigation? The Decision to Transfer and Consolidate Multidistrict Litigation, 10 J Empirical Legal Stud 424, 440, 449–57 (2013).
  • 24. See generally Tracey E. George and Margaret S. Williams, Venue Shopping: The Judges of the U.S. Judicial Panel on Multidistrict Litigation, 97 Judicature 196 (2014); Dawn M. Chutkow, The Chief Justice as Executive: Judicial Conference Committee Appointments, 2 J L & Cts 301 (2014).
  • 25. See generally Jens Dammann and Henry Hansmann, Globalizing Commercial Litigation, 94 Cornell L Rev 1 (2008).
  • 26. The ongoing prosecutions of key members of the ruling political party in Brazil, driven largely by the efforts of a single judge, are a vivid illustration of this tension. See Bruce Douglas and Sam Cowie, Brazil: Judge Halts Lula’s Appointment to Cabinet amid Corruption Scandal (The Guardian, Mar 17, 2016), archived at http://perma.cc/76UF-7V72.
  • 27. Garoupa and Ginsburg are explicit about not wanting to specify a core reputation function (p 16).
  • 28. See generally Lisa Bernstein, Private Commercial Law in the Cotton Industry: Creating Cooperation through Rules, Norms, and Institutions, 99 Mich L Rev 1724 (2001).
  • 29. See generally Robert C. Ellickson, Unpacking the Household: Informal Property Rights around the Hearth, 116 Yale L J 226 (2006).
  • 30. See generally Avner Greif, Reputation and Coalitions in Medieval Trade: Evidence on the Maghribi Traders, 49 J Econ Hist 857 (1989).
  • 31. See generally Barak D. Richman, How Community Institutions Create Economic Advantage: Jewish Diamond Merchants in New York, 31 J L & Soc Inquiry 383 (2006).
  • 32. The literature in question is too large to justify citing all of its key pieces. For an overview of the relevant dynamics, though, see, for example, Barak D. Richman, Firms, Courts, and Reputation Mechanisms: Towards a Positive Theory of Private Ordering, 104 Colum L Rev 2328, 2333–37 (2004).
  • 33. See Bernstein, 99 Mich L Rev at 1764 (cited in note 29) (noting that “breach of contract as to one transactor is transformed into breach of contract as to numerous market transactors for the purposes of a transactor’s commercial reputation”).
  • 34. See, for example, Richman, 31 J L & Soc Inquiry at 407 n 59 (cited in note 32).
  • 35. Although the example above is that of the judiciary standing up to the state in a manner that is welfare enhancing, there can also be instances in which the judiciary tries to block state action in a fashion that is welfare reducing. The actions of the US Supreme Court in the so-called Lochner (pre–New Deal) era are seen by many as such a case. For a description, see generally Noah Feldman, Scorpions: The Battles and Triumphs of FDR’s Great Supreme Court Justices (Twelve 2010). More generally, there is a large literature on what kind of judicial behavior vis-à-vis the government in power gives the judiciary legitimacy in the eyes of the public (and therefore power of its own). See, for example, Georg Vanberg, Constitutional Courts in Comparative Perspective: A Theoretical Assessment, 18 Ann Rev Polit Sci 167, 179–81 (2015) (discussing the role of strategic judicial behavior in interactions with other branches of government); Rafael La Porta, et al, Judicial Checks and Balances, 112 J Polit Econ 445, 457–58, 468 (2004) (analyzing the link between judicial independence, constitutional review, and economic freedom).
  • 36. There is a growing, albeit still small, literature on measuring judges and justice. See Jeffrey M. Chemerinsky and Jonathan L. Williams, Note, Foreword: Measuring Judges and Justice, 58 Duke L J 1173, 1175–79 (2009) (describing empirical research on judges and judicial decisionmaking).
  • 37. For an analysis of the relationship between workload and dissent rates, see generally Lee Epstein, William M. Landes, and Richard A. Posner, Why (and When) Judges Dissent: A Theoretical and Empirical Analysis, 3 J Legal Analysis 101 (2011).
  • 38. Garoupa and Ginsburg describe their work as fitting within the institutional-ana­lysis rubric (p 7). Their modification of the standard model, though, is emphasizing the paramount importance of reputation.
  • 39. See Lee Epstein and Jack Knight, The Choices Justices Make 138–39 (CQ 1998). For an explication of the Epstein and Knight model of judicial behavior, which views judges as acting strategically in response to a variety of audiences who have power over them while also seeking to advance personal policy preferences, see Frank B. Cross, Book Review, The Justices of Strategy, 48 Duke L J 511, 511–31 (1998).
  • 40. Garoupa and Ginsburg hint at being able to tell such an evolutionary story in their discussion of how the norms of the US Supreme Court have fluctuated over time from being individualistic (seriatim opinions initially) to collective (the norm against individual opinions under Chief Justice Marshall) and then back to more individual opinions (under Chief Justice Harlan Fiske Stone) (pp 32–33).
  • 41. Historical facts, of course, suggest a story that is more complicated that the simple populism or history explanation. See Caleb Nelson, A Re-evaluation of Scholarly Explanations for the Rise of the Elective Judiciary in Antebellum America, 37 Am J Legal Hist 190, 203–10 (1993).
  • 42. See Thomas R. Phillips, The Merits of Merit Selection, 32 Harv J L & Pub Pol 67, 71, 88–96 (2009) (noting that “[t]he type of system a state has depends largely on the date it adopted that system” and critiquing election systems in favor of merit selection systems).
  • 43. See, for example, Stephen J. Choi, G. Mitu Gulati, and Eric A. Posner, Professionals or Politicians: The Uncertain Empirical Case for an Elected Rather than Appointed Judiciary, 26 J L Econ & Org 290, 303 (2010) (noting in the description of an empirical study of appointed versus elected judges that “the selection mechanism long predates our data pool and thus mitigates concerns about endogeneity—that states adopted new mechanisms in response to changes in judicial quality”). See also generally Alexander Tabarrok and Eric Helland, Court Politics: The Political Economy of Tort Awards, 42 J L & Econ 157 (1999).
  • 44. See Richard A. Posner, What Do Judges Maximize? (The Same Thing Everybody Else Does), 3 S Ct Econ Rev 1, 2 (1993) (exploring “how to explain judicial behavior in economic terms, when almost the whole thrust of the rules governing compensation and other terms and conditions of judicial employment is to divorce judicial action from incentives—to take away the carrots and sticks . . . that determine human action in an economic model”).
  • 45. Some of these basic ideas on the economics of judging are discussed in Russell Smyth, Do Judges Behave as Homo Economicus and, If So, Can We Measure Their Performance? An Antipodean Perspective on a Tournament of Judges, 32 Fla St U L Rev 1299 (2005). For examples of papers conjecturing that reputational concerns motivate judges and/or those with power over their careers, see generally Gilat Levy, Careerist Judges and the Appeals Process, 36 RAND J Econ 275 (2005); Eli Salzberger and Paul Fenn, Judicial Independence: Some Evidence from the English Court of Appeal, 42 J L & Econ 831 (1999); William M. Landes and Richard A. Posner, Legal Precedent: A Theoretical and Empirical Analysis, 19 J L & Econ 249 (1976).
  • 46. What we mean here is a reputation for high-quality work—the kind of reputation that the principal in the principal-agent relationship would want. We do not mean a reputation for being willing to advance the policy agenda of some politician. The latter, one could argue, is indeed the kind of reputation that probably does result in advancement in the US federal court system. For an argument that courts should, as a normative matter, be designed to incentivize judges to pursue the former and not the latter, see generally Stephen Choi and Mitu Gulati, A Tournament of Judges?, 92 Cal L Rev 299 (2004).
  • 47. The research on this question is thin, but what little there is does not seem to suggest that politicians selecting judges are looking for stars (either in terms of their willingness to stand up to the government, their independence from political influence, or the quality of their decisions). See, for example, Stephen J. Choi, Mitu Gulati, and Eric A. Posner, The Role of Competence in Promotions from the Lower Federal Courts, 44 J Legal Stud S107, S129 (2015) (“[P]residents do not seem interested in promoting the district judges with the highest competence.”); Stephen J. Choi and G. Mitu Gulati, Mr. Justice Posner? Unpacking the Statistics, 61 NYU Ann Surv Am L 19, 42–43 (2005) (“We suspect that Posner himself does not think that he should be on the Supreme Court. . . . [It] is a highly political body.”).
  • 48. See pp 39, 173 (mentioning how the Spanish “superjudge” Baltasar Garzón, who led the indictment of General Augusto Pinochet, had developed such a high reputation outside his home jurisdiction that when he was indicted locally for pursuing Franco-era crimes, he received numerous other jobs—and twenty-one honorary doctorates in other countries).
  • 49. See, for example, Anne S. Kim, Note, Rent-a-Judges and the Cost of Selling Justice, 44 Duke L J 166, 168, 175–76 (1994) (arguing that rent-a-judges are “undesirable as a matter of public policy” and noting the large number of retired judges who become rent-a-judges).
  • 50. The classic work on this topic is Douglass C. North and Barry R. Weingast, Constitutions and Commitment: The Evolution of Institutions Governing Public Choice in Seventeenth-Century England, 49 J Econ Hist 803 (1989). See also generally Gary W. Cox, Marketing Sovereign Promises: The English Model, 75 J Econ Hist 190 (2015).
  • 51. See Alejandro Carballo, The Law of the Dubai International Financial Centre: Common Law Oasis or Mirage within the UAE?, 21 Arab L Q 91, 95–100 (2007); Jurisdiction: Opting In to the Dubai International Financial Centre (Linklaters, Dec 6, 2011), archived at http://perma.cc/J6EJ-3S9W.
  • 52. See Richard A. Epstein, Covenants and Constitutions, 73 Cornell L Rev 906, 922 (1988) (“The holdout problem occurs in its most extreme form when unanimous consent is necessary.”).
  • 53. See, for example, Anja Rohwer, Measuring Corruption: A Comparison between the Transparency International’s Corruption Perceptions Index and the World Bank’s Worldwide Governance Indicators, 7 CESifo DICE Report 42, 44–49 (Autumn 2009).
  • 54. For discussions of rankings of judicial systems and individual judges in the US context, see generally, for example, Thomas J. Miles, Do Attorney Surveys Measure Judicial Performance or Respondent Ideology? Evidence from Online Evaluations, 44 J Legal Stud S231 (2015); Theodore Eisenberg, U.S. Chamber of Commerce Liability Survey: Inaccurate, Unfair, and Bad for Business, 6 J Empirical Legal Stud 969 (2009). See also generally Stephen J. Choi, Mitu Gulati, and Eric A. Posner, Judicial Evaluations and Information Forcing: Ranking State High Courts and Their Judges, 58 Duke L J 1313 (2009).
  • 55. See, for example, Kevin E. Davis, Legal Indicators: The Power of Quantitative Measures of Law, 10 Ann Rev L & Soc Sci 37, 41–46 (2014). See also generally Tom Ginsburg, Pitfalls of Measuring the Rule of Law, 3 Hague J Rule L 269 (2011).
  • 56. For an example of the use of these measures by rating agencies, see Richard Cantor, Moody’s Sovereign Rating Methodology *18–19 (Moody’s Investors Service, May 31, 2012), archived at http://perma.cc/WS2Q-GRFA (listing the World Bank indicators as important determinants of a country’s credit rating).
  • 57. See Prospectus Dated 15 April 2016: The Republic of Mozambique *12–13, 27–28 (Apr 15, 2016), archived at http://perma.cc/5PLD-AVE9.
  • 58. See M.A. Thomas, What Do the Worldwide Governance Indicators Measure?, 22 Eur J Development Rsrch 31, 31, 34 (2010).
  • 59. See, for example, Joshua B. Fischman, Reuniting ‘Is’ and ‘Ought’ in Empirical Legal Scholarship, 162 U Pa L Rev 117, 130–54 (2013) (criticizing empirical studies that use citation rates, reversal rates, and interjudge disparities to measure judicial quality).