A court deciding a § 1983 action in which the defendant pleads qualified immunity faces two possible questions: (1) whether a constitutional right of the plaintiff was violated; and (2) whether that right was “clearly established” at the time the conduct occurred. If the court answers the first question “no,” the defendant prevails because the plaintiff has failed to successfully allege a constitutional violation. If the court answers the second question “no,” the defendant is entitled to qualified immunity, barring the plaintiff’s recovery. The question at the heart of the line of cases leading to Pearson is whether a court must confront the constitutional question regardless of the outcome of the “clearly established” qualified immunity prong, or whether a court may skip the substantive constitutional issue altogether when the answer to the “clearly established” prong supports granting qualified immunity.

The stakes are high because the difference between mandatory or discretionary sequencing may bear on the frequency with which courts address substantive constitutional rights questions, which in turn impacts the “rate” at which constitutional rights are “clearly established” through precedents. While the Supreme Court’s jurisprudence has spanned the spectrum from providing no guidance about sequencing, to suggesting it, to requiring it, and, after Pearson, again only suggesting it, there have been until recently no empirical studies that examined the relationship between the Supreme Court’s position on the qualified immunity sequencing issue and the behavior of lower courts resolving § 1983 claims.

Our Note examines a random sample of 741 appellate qualified immunity cases, representing 901 § 1983 claims, drawn from thirty-two years of qualified immunity jurisprudence (1976-2008). Ours is the largest data set constructed on the topic to date.

Thus far, the scholarship has generally divided the evolution of qualified immunity doctrine into three doctrinal periods. In the first period, the Supreme Court issued little guidance to lower courts on sequencing. In the second period, beginning after the Court’s decision in Siegert v. Gilley, the Court encouraged confrontation of the constitutional question before the “clearly established” question. That approach is typically thought to have lasted until the Court’s Saucier decision in 2001, which mandated that the constitutional rights question be answered first. Saucier’s requirement represented a high water mark in the Court’s efforts to create a uniform and obligatory procedure for all courts to adjudicate § 1983 actions. But Saucier became the target of criticism, sparking a debate (“the sequencing debate”) in academia and the judiciary about the virtues of mandatory sequencing. Finally, the fourth period was ushered in with the Court’s latest qualified immunity decision, Pearson v. Callahan.

The key question our empirical analysis set out to address was how mandatory sequencing affects constitutional articulation—the judicial process of confronting and resolving substantive constitutional questions—and thereby affects the judiciary’s understanding of what sets of facts and conduct represent rights violations. Although the Saucier Court insisted that sequencing was central to providing explanations about the “law’s elaboration from case to case,”² there has been little dispute and, until recently, little empirical research on the relationship between mandatory sequencing and constitutional articulation as it is borne out in the courts.

How lower courts have approached sequencing during three phases of Supreme Court guidance was the subject of only one empirical study prior to 2009. That study, authored by Professor Thomas Healy in 2005, is of limited use to understand the impact of the evolution of qualified immunity doctrine on the behavior of lower courts because it examined only a small set of appellate decisions issued after the Saucier opinion. Two contemporary quantitative studies of qualified immunity were published in 2009. In one, Paul Hughes sampled appellate dispositions during three discrete time intervals falling within the three doctrinal periods, and documented the expected finding that the shift to mandatory sequencing corresponded to a decrease in the frequency with which appellate courts skipped the substantive constitutional question. In a subsequent study, Nancy Leong sampled appellate and district court opinions from three discrete time intervals and concluded that Saucier has increased the quantity of constitutional articulation, but at the expense of constraining plaintiffs’ constitutional rights.

Echoing some of the same sentiments as Hughes and Leong, our study also provides evidence that although a mandatory sequencing regime may disadvantage plaintiffs bringing § 1983 actions, it may also have a generally rights-affirming effect, thereby benefiting potential future plaintiffs bringing similar § 1983 claims. Our analysis shows that: (1) the imposition of Saucier’s mandatory sequencing regime was associated with a decreased frequency of outcomes where a court granted qualified immunity without addressing the substantive constitutional question; (2) after Saucier, there was an increase in the frequency of outcomes where a court denied a constitutional violation, but that change was not statistically significant; (3) after Saucier, there was a statistically significant increase in the frequency of outcomes where a court found the plaintiff had successfully alleged a constitutional violation; but (4) in the pre-Saucier period, plaintiffs found by a court to have successfully alleged a constitutional violation were more likely (by eleven percent) to ultimately recover damages than their counterparts after Saucier, also a statistically significant observation.

In sum, our data suggests that Saucier was not necessarily a seismic moment at which the mandatory period began among appellate courts. Rather, there is quantitative and qualitative evidence that appellate courts considered themselves bound to sequence even before the Saucier decision. In other words, Saucier is not a clean proxy for the point at which appellate courts suddenly begin perfect compliance with sequencing requirements, and this is important for the empirical study of qualified immunity because it suggests a new timeline along which to study lower court behavior in relation to Supreme Court decisions.

Thus, it is our contention that Pearson should be cast as the start of a new period in the sequencing debate, rather than as a reversion to the status quo ante Saucier. After Pearson, lower courts should understand without ambiguity that they have discretion in the handling of the key questions in § 1983 qualified immunity actions. How courts will employ that discretion is not a matter that may be bounded or predicted on the basis of past behavior. The post-Pearson period will represent, at last, a period in which courts may unambiguously understand themselves to have discretion in the disposition of § 1983 qualified immunity actions, and, therefore, will be an ideal period to continue an empirical analysis of how discretionary sequencing influences the articulation and refinement of constitutional rights in the § 1983 context.

Acknowledgments:

Copyright © 2010 Stanford Law Review.

Greg Sobolski and Matt Steinberg are 2009 graduates of Stanford Law School.

This Legal Workshop Editorial is based on the following Law Review Article: Greg Sobolski & Matt Steinberg, Note, An Empirical Analysis of § 1983 Qualified Immunity Actions and Implications of Pearson v. Callahan, 62 STAN L. REV. 523 (2010).

The romantic image of the generalist, however, is not without its costs. The phenomenon of generalism deprives the judiciary of potential expertise, which could be extremely useful in cases involving complex doctrines and specialized knowledge. It also undermines efficiency, a goal that is difficult to ignore in an era of crowded dockets and overworked jurists. Indeed, many state courts have increasingly turned to specialization or a subject-matter rotation system for these reasons, yet the federal judiciary remains unflinching.

But does it really? Despite the frequent rhetoric against specialization, an empirical look at opinion assignments in the federal courts of appeals from 1995 to 2005 reveals “opinion specialization” to be an unmistakable part of everyday judicial practice. In short, the generalist judge is largely a myth. But while some may deplore this subversion of a long cherished judicial value, the development may indeed be a beneficial one. As it turns out, opinion specialization may actually achieve many of the benefits of specialized courts without incurring their costs.

I.
Opinion Assignment

One way of studying judicial attitudes toward specialization is to observe if judges become specialists when given the chance. Random case assignment eliminates most such opportunities, but the process of opinion assignment provides a rare instance in which federal circuit judges can specialize in certain subjects.

To construct the dataset used in this study, I combined the Federal Judicial Center’s well-known Federal Courts database and a database extract generously provided by Thomson West. The resulting dataset included all opinions written between 1995 and 2005 in the United States Courts of Appeals for all circuits except the Federal Circuit. To detect instances of specialization, I modeled the number of expected opinions that a judge should write in each subject area given that judge’s caseload and the circuit’s overall docket patterns. The expected frequencies were then compared against the actual frequencies using Pearson (standardized) residuals.

The figures below graphically summarize the most likely instances of specialization in the First, Seventh, and D.C. Circuits. Each horizontal line represents a subject matter, or in the case of the D.C. Circuit, an agency being reviewed. Each dot represents a judge-subject pairing. For purposes of the study, residuals with absolute values above three were defined to indicate instances of specialization, with positive values showing preference and negative values showing aversion.

Figure 1: Subject matter specialization, 1st Circuit, 1995-2005

Figure 2: Subject matter specialization, 7th Circuit, 1995-2005

Figure 3: Agency specialization, D.C. Circuit, 1995-2005

II.
Many Judges Specialize

As the graphs show, specialization appears to be alive and well in the federal appellate judiciary. Opinion assignments are not randomly distributed, and frequently the rate at which certain judges write in a subject area is highly disproportionate to that of their colleagues.

One important question is whether these results might occur purely as a matter of chance. After all, with so many judge-subject pairings, some statistical outliers are inevitable. A number of reasons, however, suggest that some non-random phenomenon is at work. For one, statistical simulations indicate that, under random opinion assignment conditions, residuals greater than 3.0 are exceedingly rare. For example, for the Seventh Circuit under random assignment, we statistically expect to see less than two residuals greater than 3.0. Instead, Figure 2 shows twenty-four such instances.

In addition, many of the specific instances of specialization make intuitive sense based on the judges’ backgrounds. For example, Judge Michael Boudin of the First Circuit, a former deputy assistant U.S. attorney general in the Antitrust Division of the Department of Justice, writes a disproportionate number of antitrust opinions. Judge Frank Easterbrook of the Seventh Circuit, known for his academic work in antitrust and corporate law, appears to specialize in antitrust and securities regulation. On the D.C. Circuit, Judge Harry Edwards, who was a labor law scholar and arbitrator before entering the judiciary, specializes in labor cases. Judge Douglas Ginsburg, who specializes in Federal Communications Commission (FCC) cases, is a longtime author of a casebook on telecommunications law, and Judge Stephen Williams, who specializes in Federal Energy Regulatory Commission (FERC) cases, is formerly an oil and gas law professor.

The explanation for these specialization patterns is likely an amalgam of factors, including individual preferences (both conscious and unconscious), internal court dynamics, and caseload pressures. Experts may prefer cases in their fields of choice not only because they are more interesting, but also because they can write opinions more efficiently and with less concern about errors. Similarly, non-experts may be willing to defer given that specialized subjects may appear less interesting, more time-consuming, and rife with potential pitfalls.

III.
A Loophole or a Keeper?

Should we be concerned about opinion specialization? The structure of the federal courts exhibits a long-standing preference for generalist judges, and opinion specialization clearly runs counter to that spirit. However, one perhaps should not be too quick to close the apparent loophole through randomized opinion assignments or other measures. Circuit judges, after all, are experienced and intelligent actors, and their chosen practices thus deserve careful consideration.

The core of the debate is whether the benefits of specialization outweigh the costs. As mentioned previously, specialization increases judicial expertise, improving efficiency and accuracy. On the flip side, specialized courts past and present have often highlighted the considerable drawbacks of specialization, including the danger of special interest capture, a tendency toward complex or arcane doctrines, and a loss of judicial prestige.

In this debate over specialization, however, opinion specialization offers an intriguing compromise. It of course captures many of the benefits of specialization. Whenever a “specialist” writes an opinion, the parties and the circuit’s jurisprudence benefit from the specialist’s expertise. Besides being more efficient, the specialist is more likely to produce opinions that are more consistent with the existing legal framework, and he or she may be better positioned to solve problems more creatively. At the same time, the opinion may enjoy greater legitimacy, since parties may give greater deference to a judge who “understands” the stakes and complexities in a field or industry. Even when a specialist is not writing the opinion, other judges can benefit from the specialist’s perspective (if on the same panel), or, at minimum, from a more coherent and well-developed body of precedent.

At the same time, because opinion specialization operates informally and flexibly, it minimizes many of the known drawbacks of more formal specialization schemes. With no formal and exclusive concentration of cases, interest groups have far less incentive to become enmeshed in the appointments process. Judges are less likely to develop tunnel vision, because they continue to handle diversified dockets and are required to both write and vote in areas outside their expertise. Finally, because judges can experiment with greater or less specialization depending on individual preference, opinion specialization mutes concerns about repetitive caseloads and a loss of prestige.

That said, opinion specialization does have potential dangers. If nonexpert judges become too deferential to their expert colleagues, the result could be anathema to the right of appeal, which is partly to protect litigants from the potentially arbitrary decisions of a single judge. An even more serious problem is the potential bias that may arise because specialties are self-selected. For example, some of the study results suggest a possible correlation between specializing in criminal law and being a former prosecutor. If judges without a criminal law background avoid writing criminal law opinions, and former defense attorneys seldom become judges because of electoral politics, then in essence only former prosecutors will direct the future of criminal law.

Some of these problems can be (or are already) addressed by the nature of the panel system or by judicial norms and procedures. However, an understanding of the actual ramifications of opinion specialization needs to be fleshed out in further examinations of this subject.

IV.
A Mechanism for Reform

On a final note, it bears mentioning that beyond merely a quirky practice among a subset of judges, opinion specialization offers a new avenue of reform for those who have long argued for specialized courts. For proponents of specialization, the most important attribute of opinion specialization is that it is modest. It does not require a radical restructuring of the federal courts or an act of Congress. Instead, it can develop informally and incrementally through everyday judicial practice, a critical advantage whenever actors are wedded to the status quo. Faced with enormous caseloads and increasingly complex cases in specialized areas, judges will opt for opinion specialization simply because it is a convenient and useful way for the judiciary to help itself.

Whether solution or affliction, opinion specialization reveals an unexplored tension in the federal judiciary. Circuit judges appear to be more conflicted on the issue of specialization than the frequent posturing might initially suggest. Exposing this fault line will hopefully encourage judges and commentators to reexamine their attitudes toward specialization. After all, archetypes like the generalist judge are powerful mental images that constrain the imagination. Dispelling the myth may therefore liberate jurists and reformers alike from their traditional boxes.

Acknowledgments:

Copyright © 2009 Stanford Law Review.

Edward K. Cheng is Associate Professor of Law at Brooklyn Law School.

Many thanks to Aran McNerney for research assistance, and the Project on Scientific Knowledge and Public Policy and the Brooklyn Dean’s Summer Research Fund for generous support.

To contact Edward K. Cheng, please email him at edward.cheng@brooklaw.edu.

This Legal Workshop Editorial is based on the following Article:   Edward K. Cheng, The Myth of the Generalist Judge, 61 STAN. L. REV. 519 (2008).

The Supreme Court, it turns out, can help answer this question.  But not in the way you might think.

I.
By the Numbers: Knowing Where to Look

For decades, the FCC has maintained a set of ownership regulations that limits the number of media outlets one entity can own.  Newspapers, such as Murdoch’s New York Post, come under the purview of the FCC’s “cross-ownership rule,” restricting common ownership of newspapers and broadcast stations in a market.

Most of federal law on the matter is predicated on an assumption that consolidation will reduce so-called “viewpoint diversity.”  Put another way, viewpoints may converge with common ownership.  Yet economic or communications theory doesn’t squarely provide a conclusion to that premise.  Over the past decade, recognizing the theoretical ambiguity, the courts and the FCC have increasingly required empirical evidence in support of this convergence hypothesis.

The trouble is that the evidence so far has been, well, flimsy.

The concept of viewpoint diversity, as the courts have recognized, is elusive.  And when, in 2002, the FCC commissioned a handful of empirical studies on the connection between ownership and viewpoint diversity, it didn’t find much.  Indeed, this elusiveness led Commissioner Jonathan Adelstein to conclude that the FCC’s work was “like submitting a high-school term paper for a Ph.D. thesis.”³

But the lack of conclusive evidence may be the result of either poor measurement of viewpoints or that ownership and viewpoints aren’t directly related.  Little consensus exists as to which story is right.

Fortunately, rapid advances in statistics are making rigorous assessment of the convergence hypothesis possible.  While “viewpoint” is an elusive concept, it does have observable consequences—in the same way that elusive concepts of “ability” or “intelligence” have observable implications.  The virtue of standardized tests, such as the SAT, is that each test answer can be viewed as a noisy indicator of a student’s underlying intelligence.  Similarly, as political scientists have recognized, we can summarize legislators’ views based on their voting records on common bills.  The crucial step is collecting information about answers (or votes) to common questions, just as SAT test-takers answer the same exam.

Where might we look for answers to common questions about viewpoint diversity, when newspaper editors don’t sit for a test, such as an SAT?  Here’s where the Supreme Court comes in.  Supreme Court justices vote on the merits of roughly one hundred cases each term.  And newspapers regularly editorialize on these decisions.  Connecting newspaper editorials to the opinions of the justices solves the difficult problem of quantifying editorial viewpoints, which the FCC has recognized as a crucial component of viewpoint diversity.

With a large research team at Harvard and Stanford, we collected every editorial position on a Supreme Court decision by the top twenty-five newspapers from 1994-2004 (roughly 1600 editorial positions) and coded these as agreeing with the majority or minority on the court.  Supreme Court cases are ideal for this study as they represent a staggering array of discrete issues.

With some refined statistical adjustments, this evidence allows us to scale newspapers in terms of their comparability on a single dimension.  One can think of it as running from “liberal” to “conservative.”  The scale tells us how each newspaper would have voted as a tenth justice and allows us to assess how viewpoints change with mergers and acquisitions of newspapers.  Essentially, the results reveal what a reasonable reader would infer after reading the editorial pages of twenty-five newspapers and the opinions in some 500 Supreme Court cases over a period of ten years.  It is in this sense that the Supreme Court is helping us learn about newspapers.

Figure 1: Estimates for Viewpoints for Select Newspapers and Supreme Court Justices. The left panel presents the median viewpoint estimate for each justice of the natural Rehnquist court. On the same scale, the right panel presents the viewpoints of The New York Times, New York Post, Washington Times, and Washington Post over time, based each paper’s editorials.  The solid lines represent our median estimate of editorial viewpoints, and the colored bands visualize the uncertainty of those estimates. The New York Post is estimated starting in 1997 because electronic versions of the paper were unavailable earlier.

Figure 1 presents some sample results for The New York Times, New York Post, Washington Times, and Washington Post.  The results quantify editorial viewpoints (and uncertainty as represented in the bands) meaningfully: The overall probability that the Washington Post is to the right of The New York Times is nearly 1.  The New York Post’s phantom jurisprudence most resembles that of Justice Scalia.  More importantly, our analysis allows us to examine the dynamic evolution of newspapers. The New York Times, for example, has been consistently trending to the left of Justice Stevens.

II.
The Merger Question

So what happens with a newspaper merger?  One important test is the merger of the editorial boards of the Atlanta Journal and the Atlanta Constitution in 2001 to form the Atlanta Journal Constitution.  This merger appears to corroborate the convergence hypothesis: the Journal Constitution’s viewpoint lands squarely between the two prior papers. But they arrive at that middle position in an unusual way.

In 1995, both the Journal and the Constitution supported the five-justice majority in United States v. Lopez, which struck down a federal statute prohibiting guns in school zones.  But shortly thereafter, the papers diverge considerably.  In 1999, for example, the Constitution argued the court “ruled wisely and well” when it found that a school could be liable for discriminatory acts committed by students, while the Constitution charged that the decision “opened yet another floodgate to lawsuits.”  The viewpoints of the editorial board members differed so sharply between the two papers that the merged Journal Constitution faced difficulty forging a consensus position on cases.  Around 2006 the paper became one of the first major U.S. newspapers to disband the practice of unsigned editorials.  The individual columns reflected diverging liberal and conservative viewpoints in line with those followed prior to the merger.  Paradoxically, then, the merger may have unified Atlanta’s readership, with the net effect of exposing more readers to more viewpoints.

Figure 2: Editorial Viewpoints for the Atlanta Journal, Atlanta Constitution, and the combined Atlanta Journal Constitution. This figure illustrates the divergence in viewpoints between the two editorial boards prior to merging. After the merger, the viewpoint of the combined board falls between those of the two former papers. As in Figure 1, the solid lines represent median viewpoints, and the color shading captures the uncertainty in estimates.

Of course, the Atlanta experience may be unique.  Examining all acquisitions occurring between the newspapers in our data, effects were varied and depended on the circumstances of the ownership change: for chain acquisitions (e.g., Hearst’s acquisition of the San Francisco Chronicle), editorial viewpoints remained stable; but after the The New York Times acquired the Boston Globe, the papers switched positions.

III.
So What?

Our analysis suggests three lessons.  First, consolidation does not inexorably cause convergence or divergence in viewpoints.

Second, our analysis points to the promises and perils of empirical assessment in law and regulation.  Using tools developed across applied statistics allows thorny questions of public policy and regulation to be examined with data.  If, for example, consolidation systematically diversified viewpoints, there would be little use in maintaining various ownership regulations.

On the other hand, such inquiry isn’t easy.  Courts and agencies shouldn’t expect too much.  Our approach, for example, does not assess viewpoints expressed in news reporting, nor can we realistically examine the effects of vast changes of federal regulation.  Judges and policymakers don’t necessarily have the luxury of making decisions after the data have been systematically gathered and analyzed.  This difficulty of evaluation suggests a type of precautionary principle: incremental, as opposed to wholesale, modification of federal regulation facilitates policy evaluation.

Last, our study sheds light on and informs what factors the FCC should consider in applying its waiver policy to the likes of Rupert Murdoch.  Whether media consolidation stifles viewpoints may ultimately turn on the minutiae of the acquisition: e.g., the terms of organizational restructuring, guarantees of editorial independence, and employment conditions.

That’s the trouble when you face the data.  It might show you that the devil’s in the details.

Acknowledgments:

Copyright © 2009 Stanford Law Review.

Daniel E. Ho is Assistant Professor of Law & Robert E. Paradise Faculty Fellow for Excellence in Teaching and Research at Stanford Law School.

Kevin M. Quinn is Associate Professor in the Department of Government and Institute for Quantitative Social Science at Harvard University.

This Legal Workshop Editorial is based on the following Article:   Daniel E. Ho & Kevin M. Quinn, Does Media Consolidation Stifle Viewpoints?: How the Supreme Court Can Provide an Answer, 61 STAN. L. REV. 781 (2009).

This Editorial previously appeared in the Stanford Lawyer: Daniel E. Ho & Kevin M. Quinn, Does Media Consolidation Stifle Viewpoints, STAN. LAWYER, Fall 2008, at 38.