|

Measuring Judging

by

Summer 2015 | Volume 99 Number 1 | Download PDF Version of Article
Illustration of Modern scale on white background

Lee Epstein, William M. Landes, and Richard Posner have written a book that is monumental in its scope and yet falls frustratingly short in achieving its aspirations. Actually, it is best understood as two books in a single volume: one stunning in the information and insights it presents about the federal judiciary, while the other seeks to do the impossible in offering a unified account of judging behavior.

The former is about three-quarters of the book and is a detailed empirical analysis of the workings of the Supreme Court, the federal courts of appeals, and the federal district courts. Although this, of course, is not the first empirical analysis of the workings of these courts, it is by far the most comprehensive. An enormous amount can be learned from the analysis contained in chapters 2 through 8 of the book, which examine a broad range of topics, including the role of ideology in federal judging, the effects of group dynamics on a multi-member court’s decision making, the impacts of heightened pleading standards, and the ways that a judge is affected by the possibility of being appointed to a higher court.

The latter is an attempt by the authors to present a unified theory of judicial decision making. They offer a “market theory of judging,” in which judges are rational economic actors in a labor market motivated by the same types of pecuniary and nonpecuniary influences as other workers. Chapter 1 of the book presents this analysis and the authors offer a “judicial-utility function” in which they attempt to describe judicial decision making in an equation that they say accounts for the many factors that determine how cases are decided.

The book’s empirical examination of the federal courts, discussed in part 1 of this review, is likely to be the basis for analysis of the federal judiciary for years to come. The concluding chapter of the book provides a very valuable description of topics for future study and identifies empirical analyses that the authors did not address, usually because of gaps in data. But the book’s attempt to present a formula that describes judging, discussed in part 2 of this review, is perplexing at best. It is unclear what the formula seeks to accomplish. It cannot be used to predict decisions or even describe them. Nor does it describe how judges should decide cases, as the work is explicitly positive, not normative. At most, the formula identifies the variables that affect a judge’s behavior, but here the authors are far less original or insightful than in their empirical analysis. It also is unclear what is gained by presenting this as an equation rather than as a listing of the factors that influence judicial decision making.

The book is dense with equations and charts, yet it is clearly written and very readable. The authors are painstaking in explaining their methodology and even those without a strong foundation in empirical work can follow their explanations and analyses. The chapters are sufficiently distinct that the book can, and likely will, be used as a reference tool for those researching decision making in the federal courts. I have read it twice and picked up a great many insights on the second reading. Simply put, Epstein, Landes, and Posner are the best at this type of work, and this book stands out as the best of its type.

Measuring The Behavior Of Federal Judges

Many political scientists and law professors have attempted to empirically analyze judicial outcomes, especially the Supreme Court’s decision making, and Epstein, Landes, and Posner greatly rely on their databases. But their work is different, in part, because of their methodology. As they explain, they “make extensive use of regression analysis, rather than just simple correlations.” (p. 14) They note that “[c]laims about judicial behavior based on simple correlation are commonly made in law journals and reported in the media, but studies that do not control for potentially relevant variables, as regression analysis enables us to do, often produce unsound results.” (Id.) the justices of the last 80 years in rank order from most conservative to most liberal.

Epstein, Landes, and Posner also present a clear picture of the ideology of the lower federal courts. They show that 33 percent of the federal courts of appeals judges are “strongly conservative”; 26 percent are “moderately conservative”; 31 percent are “moderately liberal”; and 10 percent are “strongly liberal.” (p. 177) It will be interesting to see how this changes by the end of the Obama presidency. The authors provide empirical confirmation for what long has been intuitive: “all the strongly conservative judges ex ante were appointed by Republican Presidents, all the strongly liberal ones were appointed by Democratic Presidents, almost all the moderately conservative judges were appointed by Republican Presidents, and an even larger percentage of moderately liberal ones were appointed by Democratic Presidents.” (p. 180) Presidential elections thus matter enormously in determining the ideological composition of the federal judiciary.

The Effects of Panel Composition. One of the most fascinating aspects of the book is its careful analysis of the effects of panel composition on decision making in the federal courts of appeals. The authors note that “[a] growing literature finds panel composition effects in the federal courts of appeals.” (p. 82) For example, they note that “[w]hen white judges sit on a panel with a black judge, the odds of the whites’ voting in favor of liability in voting rights cases rises. And men are significantly more likely to rule in favor of a plaintiff in a case alleging employment discrimination against a woman when the other member of the three-judge panel is a woman.” (p. 82) The dynamics of a small group inevitably influence decision making in any context, but the authors have provided an empirical confirmation of this for federal court of appeals rulings. It provides very strong support for the importance of diversity in the federal courts of appeals.

Reversal Aversion. One variable that the authors attempt to measure is how concern about the possibility of reversal by a higher court affects decision making in the federal district courts and federal courts of appeals. The authors acknowledge that this varies among individual judges, yet overall has a significant effect. They conclude: “District judges appointed by a Democratic President and reviewed by a court of appeals dominated by Democratic Presidents have been found to impose on average a prison sentence four months shorter than district judges also appointed by a Democratic President but who face the prospect of appellate review by a court dominated by judges appointed by Republican Presidents.” (p. 83) The authors note that this has the effect of diminishing the apparent effect of the ideology of the district court judge in sentencing.

Supreme Court decisions matter in the federal district courts. The authors focus on two recent, major Supreme Court decisions and show their powerful impact in the federal district courts. In United States v. Booker, the Supreme Court held that the Sixth Amendment required that the federal sentencing guidelines be advisory and not mandatory in sentencing criminal defendants in federal courts.1 The result was substantially more discretion for federal district court judges in sentencing. The authors note that “district judges changed their sentencing behavior almost immediately after Booker.” (p. 242) In the first year after Booker, the proportion of below-guideline sentences rose from “28.4 percent to 36.8 percent (a 33 percent increase over the pre-Booker 2005 percentage).” (p. 242)

The other Supreme Court decision that dramatically changed behavior in the district courts was Ashcroft v. Iqbal, where the Supreme Court tightened the standards for what needs to be pled in a complaint for a case to go forward in a federal court.2 The authors document a “highly significant effect” of Iqbal in leading to the dismissal of cases and that civil rights cases are “17 to 19 percent more likely to be dismissed than civil cases as a whole.” (p. 231)

These examples powerfully show that district courts carefully follow Supreme Court decisions, even very quickly after they are announced. The authors also confirm what many have noted without empirical confirmation: Booker has had a substantial effect on sentencing by federal district court judges and Iqbal has significantly increased the likelihood of a motion to dismiss being granted.

Auditioning. One of the most fascinating chapters in the book examines the extent to which judges are influenced by their desire for appointment to a higher court. The authors identify federal court of appeals judges who have had a realistic chance of being considered for appointment to the U.S. Supreme Court and measure how the prospect of appointment affects decision making. They call these judges “auditioners” and observe that:

… auditioners are significantly less likely to vote for defendants in capital punishment and street-crime cases than either non- or ex-auditioners are. There are no significant differences in the white-collar category, where the emotions of the public are less likely to be aroused by the reversal of a conviction or a sentence.” (p. 361)

The authors use regression analysis to control for other variables and document that the possibility of being considered for a Supreme Court vacancy substantially affects the decision making of federal courts of appeals judges. The authors conclude:

Our results suggest that court of appeals judges who have a good shot at the Supreme Court tend to alter their judicial behavior in order to increase

Just as important, the analysis in the book is “broader than that found in most of the existing literature” (p. 15) in two respects: the variety of courts reviewed and the volume of questions examined. Chapters 3, 4, and 5, respectively, look at the Supreme Court, the federal courts of appeals, and the federal district courts. Chapter 6 focuses on dissents in the Supreme Court and the federal courts of appeals, and what the authors call “dissent aversion,” what accounts for judges choosing to not dissent. Chapter 7 considers questioning at oral argument in the Supreme Court and offers an empirical analysis of these proceedings, including which justices ask the most and the longest questions and how questioning relates to the outcome of cases. Chapter 8 seeks to assess how a judge is affected by the possibility of being appointed to a higher court — federal district court judges to a federal court of appeals and especially federal court of appeals judges to the Supreme Court. As mentioned above, the last chapter identifies many questions for future empirical work.

Any attempt in a short review to summarize the analyses or conclusions could not possibly do justice to the breadth and depth of the book. Instead, this review provides several examples of the most important insights of the authors’ empirical analysis.

The Role of Ideology in Judging. Epstein, Landes, and Posner provide empirical support to confirm what seems intuitive: ideology matters more in “higher” courts than in “lower courts.” Ideology plays a greater role in the Supreme Court than in the federal courts of appeals or the district courts, and again a greater role in the federal courts of appeals than in the federal district courts. The authors note, “As cases rise in the judicial hierarchy, the possibility of deciding them on legalistic grounds decline.” (p. 237) In part, this is because lower courts are constrained by precedent from higher courts; it also reflects the differences in workloads among these courts. Only the Supreme Court exercises discretion in selecting the vast majority of its docket. This feature of federal court jurisdiction has two effects on the proportion of Supreme Court cases in which ideology plays a major role. First, justices have the opportunity to vote on certiorari based on their ideological preferences, unlike other courts. Second, because the Supreme Court has discretion in defining its workload, it tends to hear more high profile cases involving issues giving rise to ideological preferences as a share of its docket than other courts.

The authors confirm the importance of ideology in Supreme Court decision making. They state: “[U]tilizing an expanded and corrected dataset, our analysis establishes, we hope with rigor and precision, the existence of a substantial ideology effect in Supreme Court decisions.” (p. 149) They use several different measures to assess the ideology of the individual justices and conclude that “Rehnquist and Thomas rank as the most conservative justices and Sutherland, Alito, and McReynolds as the most conservative in economic cases (economic regulation, labor, and tax). At the other end of the spectrum, Marshall, Douglas, Brennan, and Murphy rank as the most liberal justices overall.” (p. 106) These conclusions are not surprising (except perhaps as to the extent of Justice Alito’s conservativism), but it is still interesting to see the authors use several different methodologies to list the justices of the last 80 years in rank order from most conservative to most liberal.

Epstein, Landes, and Posner also present a clear picture of the ideology of the lower federal courts. They show that 33 percent of the federal courts of appeals judges are “strongly conservative”; 26 percent are “moderately conservative”; 31 percent are “moderately liberal”; and 10 percent are “strongly liberal.” (p. 177) It will be interesting to see how this changes by the end of the Obama presidency. The authors provide empirical confirmation for what long has been intuitive: “all the strongly conservative judges ex ante were appointed by Republican Presidents, all the strongly liberal ones were appointed by Democratic Presidents, almost all the moderately conservative judges were appointed by Republican Presidents, and an even larger percentage of moderately liberal ones were appointed by Democratic Presidents.” (p. 180) Presidential elections thus matter enormously in determining the ideological composition of the federal judiciary.

The Effects of Panel Composition. One of the most fascinating aspects of the book is its careful analysis of the effects of panel composition on decision making in the federal courts of appeals. The authors note that “[a] growing literature finds panel composition effects in the federal courts of appeals.” (p. 82) For example, they note that “[w]hen white judges sit on a panel with a black judge, the odds of the whites’ voting in favor of liability in voting rights cases rises. And men are significantly more likely to rule in favor of a plaintiff in a case alleging employment discrimination against a woman when the other member of the three-judge panel is a woman.” (p. 82) The dynamics of a small group inevitably influence decision making in any context, but the authors have provided an empirical confirmation of this for federal court of appeals rulings. It provides very strong support for the importance of diversity in the federal courts of appeals.

Reversal Aversion. One variable that the authors attempt to measure is how concern about the possibility of reversal by a higher court affects decision making in the federal district courts and federal courts of appeals. The authors acknowledge that this varies among individual judges, yet overall has a significant effect. They conclude: “District judges appointed by a Democratic President and reviewed by a court of appeals dominated by Democratic Presidents have been found to impose on average a prison sentence four months shorter than district judges also appointed by a Democratic President but who face the prospect of appellate review by a court dominated by judges appointed by Republican Presidents.” (p. 83) The authors note that this has the effect of diminishing the apparent effect of the ideology of the district court judge in sentencing.

Supreme Court decisions matter in the federal district courts. The authors focus on two recent, major Supreme Court decisions and show their powerful impact in the federal district courts. In United States v. Booker, the Supreme Court held that the Sixth Amendment required that the federal sentencing guidelines be advisory and not mandatory in sentencing criminal defendants in federal courts.1 The result was substantially more discretion for federal district court judges in sentencing. The authors note that “district judges changed their sentencing behavior almost immediately after Booker.” (p. 242) In the first year after Booker, the proportion of below-guideline sentences rose from “28.4 percent to 36.8 percent (a 33 percent increase over the pre-Booker 2005 percentage).” (p. 242)

The other Supreme Court decision that dramatically changed behavior in the district courts was Ashcroft v. Iqbal, where the Supreme Court tightened the standards for what needs to be pled in a complaint for a case to go forward in a federal court.2 The authors document a “highly significant effect” of Iqbal in leading to the dismissal of cases and that civil rights cases are “17 to 19 percent more likely to be dismissed than civil cases as a whole.” (p. 231)

These examples powerfully show that district courts carefully follow Supreme Court decisions, even very quickly after they are announced. The authors also confirm what many have noted without empirical confirmation: Booker has had a substantial effect on sentencing by federal district court judges and Iqbal has significantly increased the likelihood of a motion to dismiss being granted.

Auditioning. One of the most fascinating chapters in the book examines the extent to which judges are influenced by their desire for appointment to a higher court. The authors identify federal court of appeals judges who have had a realistic chance of being considered for appointment to the U.S. Supreme Court and measure how the prospect of appointment affects decision making. They call these judges “auditioners” and observe that:

… auditioners are significantly less likely to vote for defendants in capital punishment and street-crime cases than either non- or ex-auditioners are. There are no significant differences in the white-collar category, where the emotions of the public are less likely to be aroused by the reversal of a conviction or a sentence.” (p. 361)

The authors use regression analysis to control for other variables and document that the possibility of being considered for a Supreme Court vacancy substantially affects the decision making of federal courts of appeals judges. The authors conclude:

Our results suggest that court of appeals judges who have a good shot at the Supreme Court tend to alter their judicial behavior in order to increase their chances, though this is just an average tendency — we do not suggest that all court of appeals judges in what we are calling the promotion pool audition for the Supreme Court.” (p. 363)

These, of course, are only some of the many insights that the authors’ empirical analysis provides. Although the empirical analysis of the federal courts is stunning in its scope and depth, there are limits to the book’s methodology that should be noted. First, the authors only analyze published decisions of the federal courts. This limitation does not harm the Supreme Court analysis because all of its decisions are published. But for the federal courts of appeals, in fiscal year 2010, 84 percent of the decisions were unpublished. (p. 155) In other words, the authors’ analysis of the federal courts of appeals is based on just 16 percent of their rulings. The authors assume that the unpublished rulings are most likely routine, easily decided, and less important. Yet this assumption is questionable for many unpublished decisions. Often, for whatever reason, appellate courts write lengthy unpublished decisions that may include concurrences and dissents. In fact, because all decisions are placed on easily accessible data bases, there is no need to focus just on those that are designated for publication by the judges. Technology has ended any reason for distinguishing between published and unpublished decisions. Understanding the work of the federal courts of appeals necessitates looking at all of their work, not just 16 percent of it. It would be fascinating to examine the differences, if any, between “published” and “unpublished” decisions.

Second, though the book is titled The Behavior of Federal Judges, it omits an enormous part of the workload of the federal judiciary from its consideration. As the authors acknowledge, they have not evaluated non-Article III courts, that is courts where judges don’t have life tenure, including bankruptcy judges (who serve 14-year terms), magistrate judges (who serve eight-year terms), or judges on tribunals such as the Tax Court (who serve 15-year terms). (p. 397) This is a huge percentage of the workload of the federal courts. For example, in 2011, Justice Breyer explained: “The volume of bankruptcy cases is staggering, involving almost 1.6 million filings last year, compared to a federal district court docket of around 260,000 civil cases and 78,000 criminal cases.”3 The book more aptly should have been titled, The Behavior of Article III Federal Judges.

Third, the empirical analysis curiously treats disparate cases alike and does not account for important differences among them. The role of ideology in the Supreme Court is understated if its entire docket is treated identically, but it plays a more prominent role if the analysis focuses on cases involving socially and legally divisive issues, such as abortion, campaign finance, the death penalty, gay and lesbian rights, race, rights of detainees, and separation of church and state. The same, of course, is likely to also be true of the lower courts even though such divisive issues are a much smaller part of the docket.

I do not want to overstate these limitations in the methodology, but they are important to keep in mind in reading the book’s analysis and conclusions.

A Labor-market Theory Of Judging

The authors are not content with presenting detailed empirical analysis of decision making in the federal courts. They seek to offer a unified theory of judging and present an equation — which they call a “judicial utility function” — that they contend accounts for the many variables in judicial decision making. (p. 48) I am far more skeptical of this aspect of the authors’ enterprise.

The authors begin by contrasting two models of judging. One they term “legalism,” though they note that the more common label is “formalism.” (p. 2) They say that “in its simplest form, judges are said merely to apply law that is given to them to the facts; their task is mechanical, at best a form of engineering …, and involves no exercise of discretion.” (Id.) They contrast this with “realism,” which they say is a conception of judging in which “legal pretensions are mere rhetoric, designed to conceal the political character of their rulings.” (Id.) The authors attribute this view to proponents of “critical legal studies” and feminist legal theory and critical race theory scholars.

The problem, of course, is that each of these two accounts of judicial behavior is far too simplistic. Judges at all levels have discretion. In constitutional law, for example, deciding what is a “legitimate” or an “important” or a “compelling” government interest to justify government action is inherently a value choice of the judges. Deciding whether diversity is a compelling interest justifying affirmative action or whether limiting marriage to a man and a woman serves a legitimate interest — to pick two high-profile illustrations – is a function of the judge’s values. In criminal procedure, judges on a daily basis need to decide what is an “unreasonable” search or arrest. A highly formalistic account of judging was rejected long ago by the Legal Realists, and few today believe it accurately depicts judicial decision making.

But that said, for lower courts there are cases that are clearly resolved by prior precedents and cases where virtually any judge would come to the same conclusion. There are “easy” cases where the facts and the law are clear and where there is no doubt as to the application of the law to the facts.

Neither the “legalist” nor the “realist” approaches to judging account for both the discretion and the constraints that are a part of day-to-day judicial decision making. Epstein, Landes, and Posner explicitly recognize this. But it is curious that they begin the book by painting these two caricature approaches to judging.

It is into this gap that the authors offer their labor-market theory of judging and their judicial-utility function. The authors establish a model in which federal judges are rational economic actors behaving in a labor market the same as any other workers. Epstein, Landes, and Posner explain this:

We explain that a judge conceived of as a participant in a labor market can be understood as being motivated and constrained, as other workers are, by costs and benefits, both pecuniary and nonpecuniary, but mainly the latter: nonpecuniary costs such as effort, criticism, and workplace tensions, nonpecuniary benefits such as leisure, esteem, influence, self-expression, celebrity (that is, being a public figure), and opportunities for appointment to a higher court, and constrained also by professional and institutional rules and expectations and by a ‘production function’ — the tools that the worker uses in his job and how he uses them. (p. 5)

The judicial-utility function seeks to express all of these variables in an equation. (p. 48) The authors say that “the judge seeks to maximize his utility subject to a time constraint.” (Id.) Epstein, Landes, and Posner thus seek to offer a unified theory of judging in the federal courts.

While the variables they identify certainly influence how a judge decides cases, little to nothing is gained by placing all of these variables in an equation, or as they call it, a “judicial-utility function.” The equation creates the appearance of a precision that obviously it cannot provide. These variables and the weight they place in decision making are different for each judge, and likely for any given judge at different times. In other words, the ideology of the judge matters enormously for some cases and little for others, and more for some judges and less for others.

Thus, the authors’ judicial-utility formula does not offer any basis for describing how a judge has decided a case or how he or she will decide a matter. Nor, of course, does it provide any basis for discussing how judges should decide cases. So what is gained by presenting this as an equation rather than simply a list of the many factors that influence judging, to a greater or lesser extent depending on the case, the judge, and the circumstances? I understand the great value of the book in attempting to measure the importance of many of these factors in the decision-making process. That is the great strength of the book. But it is the attempt to provide an overall account of judicial behavior – the authors’ labor-market theory – that I find puzzling and likely an impossible quest.

Conclusion

A few years ago, I had a conversation with a prominent professor at a prestigious law school. She said that she does not like constitutional law being taught in the first year of law schools because it causes students to think that the outcomes of cases depend on the identity of the judges. I was shocked to hear her say this. I expressed my view that results, of course, often depend on who is on the bench. When I am arguing a case in a federal court of appeals, I want to learn the identity of my panel at the first possible moment. When lawyers tell me that they have argued a case, my first question is “who was your panel?” This is not to say that every case depends on the identity of the judges; obviously in many instances any judge would come to the same conclusion. But as Epstein, Landes, and Posner explain, judges at all levels of the federal courts have substantial discretion, and the identity of the judge matters greatly in how that discretion will be exercised. Epstein, Landes, and Posner have written a brilliant book that ultimately shows the importance of the individual judge or justice to the outcome of cases. The Behavior of Federal Judges shows that who is on the bench matters enormously. It is a work that will shape discussions and research about the federal courts for many years to come.

1 543 U.S. 220 (2005).

2 556 U.S. 662 (2009).

Stern v. Marshall, 131 S. Ct. 2594, 2620 (2011).

Erwin Chemerinsky is the founding Dean and Distinguished Professor of Law, and Raymond Pryke Professor of First Amendment Law, at the University of California, Irvine School of Law, with a joint appointment in Political Science. He is the author of eight books, including The Case Against the Supreme Court, published in 2014.

About Erwin Chemerinsky

Erwin Chemerinsky is Dean of the University of California Berkeley School of Law and the Jesse H. Choper Distinguished Professor of Law. He previously served as the founding Dean and Distinguished Professor of Law, and Raymond Pryke Professor of First Amendment Law, at University of California, Irvine School of Law, and, before that, was a professor at Duke University and at the University of Southern California Law School. He is the author of eleven books, including leading casebooks and treatises about constitutional law, criminal procedure, and federal jurisdiction, and a fellow of the American Academy of Arts and Sciences.