HOW LAW CLERKS INFLUENCE:
INFORMATION AT THE
U.S. SUPREME COURT
by
CHRISTOPHER DAVID KROMPHARDT
JOSEPH L. SMITH, COMMITTEE CHAIR STEPHEN BORRELLI RICHARD C. FORDING EMILY H. RITTER CHRISTOPHER ZORN
A DISSERTATION
Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Political Science in the Graduate School of The University of Alabama
TUSCALOOSA, ALABAMA
2015
Copyright Christopher David Kromphardt 2015 ALL RIGHTS RESERVED
ABSTRACT
The role of law clerks at the United States Supreme Court has long been a source of curiosity among observers and scholars alike. Of particular interest are what characteristics determine which applicants are selected to clerk and what influence clerks have on the justices’ decision making. Using a two-part framework that identifies information asymmetries, where clerks possess relevant information that they can transmit to their justices, and conditions under which this information leads a justice to learn about policies’ impact, I uncover evidence of a causal mechanism by which clerks wield systematic influence over the justices’ decision making. I show that information clerks convey that is derived from their ideological preferences and from their experiences and socialization influences their justices’ votes on the merits. In the concluding chapter, I argue that these findings shed light on the broader class of principal-agent relationships of which the justice-clerk dynamic is an example and discuss how law clerks pose an opportunity for scholars to learn how principals acquire and use information from their advisors.
ii
DEDICATION
To Bud Fischer and Bonnie Irwin, who made graduate school sound like a good idea.
They were right.
iii
ACKNOWLEDGEMENTS
Personal and Professional
My family and my friends, especially Erik Uribe and Chelsie Young, are the non- political science audience that keeps me grounded, although it is no fault of theirs when my thoughts are in a dataset somewhere.
Emily Ritter, Joe Smith, and Greg Vonnahme were the project’s earliest supporters. I don’t know what they saw when I first started going around talking about law clerks, but their encouragement was vital.
Todd Peppers, Art Ward, and Chris Zorn provided the inspiration for all my good questions, and, as my work progressed, they also became valued friends and sounding- boards. A special shout-out to the University of Alabama’s Amelia Gorgas Library for stocking Todd and Art’s books, which I have had checked out since 2010.
Amanda Bryan deserves special mention. Her early enthusiasm for law clerks helped convince me that it was a cool topic. Other friends and colleagues who have inspired me to think about clerks in new ways are Erin and Max Hill, Wes Hutto, Michael
Flynn, Steve Miller, Mark Owens, Chanley Rainey, Doug Rice, Laura Sojka, James
Todd, and Joe Walsh.
Paul Collins and David Klein were key players in the project’s development:
Paul, whose data provided the alternative explanations to the story I was telling, and
David, whose exceptional editorial advice should earn him the title of unofficial sixth
iv member of my committee. Steve Borrelli, Heather Elliot, Rich Fording, Doug Gibler, and
Hong Min Park were willing to help whenever I had a nagging issue.
The project has benefited enormously from the feedback from conference and symposium audience members. The chance to present my work at Marquette University
Law School has been a high point in my academic career, and I thank Todd and Chad
Oldfather for inviting me. I was fortunate enough to sit in the back of the room with
Larry Baum, Ryan Black, and Christy Boyd, whose good humor and insightful comments
I only wish could accompany every academic panel I have attended. Others whose comments have made my conference experiences rewarding include Sara Benesh, Robert
Howard, Michael Salamone, Amy Steigerwalt, Steve Wasby, Jennifer Williams, and
Patrick Wohlfarth.
Finally, I would like to acknowledge my officemates over the years, for enduring with good humor and minimal exasperation the thinking out loud that precedes most of my work: Wes, Chanley, James, Doug Pigue and Anderson Starling.
Intellectual Debts
I could not have conducted the research presented in this dissertation without other scholars generously sharing their data. Not having to worry about the basic building blocks of my research design freed me up to focus on making the new elements that I was adding as good as possible. First, Todd Peppers provided the full list of who all have served as clerks that I needed to operationalize my concepts. Knowing their names, when they served, and for whom, allowed me to collect additional background information on
v these people, including their previous clerkship(s), sex, and later careers. These new data were necessary to describe clerk preferences and validate a new measure of clerk preferences. Second, almost from the beginning I have used Paul Collins’s data. His model has provided the consistent baseline to which I will repeatedly compared the effects of my variables. Finally, the Judicial Common Space and the Supreme Court
Database are reliable resources whose quality and ease-of-use are easy to take for granted.
This research is also indebted to work describing how clerks fit into the decision- making processes of the Supreme Court. When your goal is to try to explain what is going on inside a black box, it helps to have as much of an idea as you can about what that box looks like and how it works. The most fundamental guidance came from Todd
Peppers’ and Art Ward and David Weiden’s books. These and other sources provide the background and anecdotes in Chapter Three that lay the microfoundation for the theory of information transmission, and this theory will be the source of my hypotheses about how influence occurred. On a similar note, reading articles in formal theory has helped teach me to appreciate the value of rigor when making assumptions and deriving hypotheses. As well, adapting some of the lessons I have learned from the formal theory literature on regulatory design and function would be a natural and exciting next step to the work I have done. I talk about some possible avenues of research in Chapter Five.
Finally, work on clerk selection by Peppers and Chris Zorn and by Larry Baum and
Corey Ditslear, while strongly supporting the finding that justices usually try to hire
vi clerks who share their preferences, to me raise questions about when they choose not to hire identical agents—can this phenomena be described and what consequences for decision making stem from having clerks whose preferences diverged from the justice for whom they worked? These are the questions that stimulate and energize the inquiry presented in this dissertation.
vii CONTENTS
ABSTRACT...... ii
DEDICATION...... iii
ACKNOWLEDGEMENTS...... iv
LIST OF TABLES...... ix
LIST OF FIGURES...... x
CHAPTER ONE: INTRODUCTION...... 1
CHAPTER TWO: FIELDING AN EXCELLENT TEAM: LAW CLERK SELECTION AND CHAMBERS STRUCTURE AT THE U.S. SUPREME COURT...... 10
CHAPTER THREE: U.S. SUPREME COURT LAW CLERKS AS INFORMATION SOURCES AND JUSTICE DECISION MAKING...... 34
CHAPTER FOUR: FEMALE LAW CLERKS AND SUBSTANTIVE REPRESENTATION AT THE U.S. SUPREME COURT...... 69
CHAPTER FIVE: LAW CLERKS AND INFORMATION AT THE SUPREME COURT...... 94
REFERENCES...... 100
viii LIST OF TABLES
2.1 The number of foxhole chambers assembled by a moderate justice...... 30
3.1 Logistic Regression of a justice’s decision to cast a liberal vote, 1985-2001...... 59
3.2 Logistic Regression of a justice’s decision to cast a liberal vote, 1985-2001...... 63
3A.1 List of Supreme Court Law Clerks with Common Space Scores...... 68
4.1 Descriptive statistics for dependent and independent variables, 1985-2001...... 80
4.2 Logistic Regression of a justice’s decision to cast a liberal vote, 1985-2001...... 86
ix LIST OF FIGURES
2.1-2.2 Justice and bloc ideological scores from 1969-2007...... 26-27
3.1 Effects of increasing clerk and justice liberalism on the probability a justice casts a liberal vote, 1985-2001...... 60
3.2 Effect of increasing the average ideological distance between justice and clerks on the probability a justice casts a liberal vote in a conservative foxhole chambers, 1985- 2001...... 65
4.1-4.2 Hiring practices by sex by justice, 1969-2007...... 83-84
4.3 Effect of increasing the number of female clerks on the probability a justice casts a liberal vote in cases involving a gendered dimension, 1985-2001...... 88
4A.1 Number of female law clerks hired by all justices, 1969-2007...... 93
x Chapter One: Introduction
How Law Clerks Influence
Supreme Court law clerks are often referred to metaphorically—as courtiers, sorcerers’ apprentices, courthouse mice, or legal Rasputins. Observers know they help their bosses, the justices of the United States Supreme Court, do their work. Their curiosity leads these observers to then wonder whether clerks wield too much influence over the justices’ decisions making. We can thank William Rehnquist—former law clerk, associate justice, and chief justice—for popularizing the idea that the answer to Who
Writes the Decisions of the Supreme Court? may include the justices’ law clerks.
These metaphors make for tantalizing images, but how accurately do they depict reality? Do law clerks influence? Scholars have employed social-scientific tools to measure the extent of the influence clerks actually wield at decision-making stages including agenda sending, vote casting, and opinion writing. Their research engages a unique challenge—trying to identify the influence of an unusually secretive group on an opaque process like decision making. How can we capture the impact on a process we cannot observe (decision making) of a group of people (the law clerks) who generally do not like to talk about what they do?
Pondering how clerks influence raises more than methodological questions. As I will discuss in the chapters of this dissertation, understanding how clerks influence their justices provides additional insight into how the justices approach their jobs, how they acquire and use information in their decision making, and how the substance of the
Court’s representation of women’s rights is formed. And as I will make more concrete in the concluding chapter, the effort we put into understanding the specifics of law clerks
1 will also tell us more about the broader category of principal-agent relationships. In sum, studying the role of clerks will tell us a lot more than just how accurate the metaphors are.
What We Learn about Law Clerks
To situate this study in the literature, I take up what I see as the contributions of this dissertation to the study of law clerks. Specifically, I see three main contributions: a description of why justices seek out clerks with different preferences than their own; a causal theory of how information acquired from clerks influences votes on the merits; and tests of this general theory with information arising from two dimensions of clerk policy preferences, ideology and feminism.
Identifying Judicial Goals by the Clerks a Justice Hires
Work by Baum, Ditslear, Peppers, and Zorn finds a strong but imperfect positive relationship between a justice’s policy preferences and those of his1 clerks. This is of course consistent with principal-agent theory, but I argue that we should not ignore the occurrence of preference-mismatches. There are at least three ways that we can interpret these findings. The first is measurement error. At least some measurement error, for justice and clerk ideologies alike, should be expected, as ideology is a notoriously tricky concept to measure (Fischman and Law 2009). Another explanation is that when a justice hires an ideological mismatch that it is the result of coincidence or a mistake. However, the existence of selection committees, feeder judges, and a large number of qualified applicants across the ideological spectrum makes this explanation less plausible.
Certainly there are enough qualified clerks from lower-court judges of known ideology
1 For purposes of clarity, the feminine denotation (i.e. she, her) is used in reference to clerks and the masculine denotation (i.e. he, his) is used for judges and justices. This convention was adopted randomly.
2 that if justices were only interested in hiring ideological matches, they would be capable of doing so. The third explanation is that justices sometimes hire ideologically diverse advisors on purpose.
Substantial anecdotal evidence, which I present in Chapter Three, indicates that this third possibility—that justices purposefully hire clerks whose ideological preferences differ from their own—took place for at least ten justices serving from 1969 to 2007.
Chapter Two demonstrates empirically just how common this practice is. In Chapter
Three, I discuss two reasons why hiring clerks whose preferences are not identical to his own can help a justice realize his goals. A justice may hire a crosswinds chambers, where clerks hold a variety of viewpoints; a crosswinds chambers would serve well the justice who wants to consider the policy implications of a broad set of outcomes by drawing on information from clerks with a heterogeneous array of preferences. Chapter Two shows that crosswinds are the most common type of chambers. Alternatively, a justice may hire a foxhole chambers. Foxholes are distinctive for the similarity of the clerks’ preferences, to each other if not necessarily to the justice’s. A justice who wants to marshal as much information reinforcing the beneficial impact a particular policy will have would be well served by assembling a foxhole of clerks who desire that outcome.
This categorization of chamber structure by how similar clerks are to each other is the first contribution of this dissertation to the literature on law clerks. After relying on the anecdotes of ten recent justices to derive what goals the justices seek, I show that the chambers structure can have predictable effects on votes on the merits. When a greater number of clerks share similar preferences, they can reinforce a desired outcome and increase the chance that their justice votes to support that outcome. As I show in Chapter
3 Three, this effect is observable for conservative clerks, and in Chapter Four, for female clerks.
Accounting for How Information Influences Votes
A gap in the literature on clerk influence has been what causal mechanism leads to influence. Scholars are certainly aware that this is missing from their accounts
(Peppers and Zorn 2008, 75). Therefore, a goal of mine is to develop and test a causal mechanism. Specifically, I want to show how information provided by clerks influences the justices’ votes on the merits. This mechanism underlies the question in the dissertation’s title: how law clerks influence.
The theory I develop accounts for what information a justice acquires from his clerks and when that information should shape the justice’s behavior. The theory is outlined in greatest detail in Chapter Three. It has three benefits. First, the theory is plausible: we know that at least ten justices have sought their clerks’ input, opening their doors, so to speak, to the clerks’ information. It also has the benefit of being falsifiable: knowing the outcome clerks are likely to support and the conditions under which they advocate for that outcome, does the justice’s behavior change? If his behavior does not, then the theory is undermined. If his behavior does change as expected, however, then the theory, while not proven, is more likely to be true. Finally, the theory is generalizable. As
Sally Kenney shows in her analysis of referendaires (2000a), and as I will discuss in
Chapter Five, I believe that studying features of the case of Supreme Court law clerks can tell us more about other hierarchical relationships. In particular, I will argue that studying clerks will shed new light on the acquisition and utilization of information that takes place between superiors and subordinates in a hierarchy.
4 Testing the Theory with Two Dimensions of Clerk Preferences
This informational theory lends itself to many possible empirical tests, two of which I present in Chapters Three and Four. These chapters share the same structure and they model the research approach I outline later in this introductory chapter: specify what information clerks possess to transmit and identify those conditions that affect whether that information determines which policy a justice supports with his vote. The evidence reported in these chapters reinforces the causal mechanism.
Chapter Three focuses on information of an ideological nature. When a justice hires clerks whose ideological preferences diverge from his own, discussing a case with his clerks can lead the justice to learn about policy impacts. To put it another way, information can lead a justice to update his prior beliefs about the likely impact of his policy choices. Say that a justice is exposed to information provided by his conservative clerks that reinforces the positive impact of a proposed conservative policy. When his clerks’ message is consistent, the information they provide may make a justice more likely to adopt a policy they support. In other words, the learning is more likely to be sufficient to affect his vote when all of his clerks agree on the outcome; when they disagree, the justice will receive mixed and competing signals about which policy to adopt. This conditional expectation that influence occurs when clerks’ preferences are uniform but not when they are diverse will be partially borne out by the results: uniformly conservative clerks pull their justice to the right, whereas liberal clerks appear not to be able to do so in their preferred direction.
A similar dynamic occurs when female clerks convey information. Chapter Four reports that an increase in the number of female clerks a justice employs positively
5 affects his support for feminist policies. However, this effect has a threshold: the positive effect only becomes statistically distinguishable from zero when a justice has at least two female clerks. An analysis of what I label feminist justices adds robustness to this account by showing that feminism is unlikely to be the driving force behind both support for feminist policies and the decision to hire multiple female clerks. Chapter Four supports the theory by showing that information that can be attributed to clerks affects the justice’s support for the clerks’ desired outcome. As in Chapter Three, though, this information needs to be reinforced: whether clerks are conservative or female, there is strength in numbers.
Research Approach
The empirical approach I use to measure clerk influence consists of two major steps. First, I seek to develop an accurate picture of the information that clerks possess.
This is a critical step because it makes sense to argue that clerks wield influence only if I can show that something a clerk possesses—such as information that reinforces the clerk’s desired policy outcome—leads the justice to act in a manner he would not have absent some intervention by his clerks. Intervention is crucial because it specifies how clerks influence—the causal mechanism. Here, intervention assumes the form of the clerks conveying information to their justice. Once I have established what information clerks possess to convey, the second step is to identify the conditions under which this information will affect a justice’s behavior. What will lead to clerks’ preferences being discernible in the policies a justice endorses? By first measuring the content of the clerks’ preferences and then testing when that information affects a justice’s behavior, I show how clerks influence through their role as advisors.
6 I derive hypotheses and interpret the results this approach yields within the context of a causal theory about how information can lead to learning. Justices use information to learn, in order to reduce their uncertainty about the impact their policies will have on affected parties and on society. To preview my causal argument, the information clerks convey leads their justice to learn about the ramifications of policies from which he can choose. Learning that leads to a change in behavior—say, a justice coming to support a different litigant than whom he would have if his clerks had said nothing—constitutes clerk influence.
A major condition affecting this learning will be the strength of the informational signal his clerks transmit to him. In other words, how likely is the information to cause the justice to learn something that makes him more likely to support the clerk’s preferred outcome? There are two ways in which a signal is considered a particularly strong tool for instigating a justice’s learning: when the signal is reinforced by multiple clerks advocating for the same policy and when it is not weakened by other clerks calling for an opposing outcome. If statistical analysis uncovers behavior that expresses the clerks’ preferences when they have sent a strong signal, then we can attribute the justice’s behavior to the influence of his clerks.
While this is a plausible and parsimonious theoretical account of how clerks influence, it has not yet been tested with empirical data. To evaluate the theory’s validity,
I conduct tests using two different forms of information that are attributable to clerks: information based on the clerks’ ideologies (Chapter Three) and on their sexes (Chapter
Four). I evaluate the statistical effects of measures capturing this information in analyses that control for well-known alternative information sources, including litigant resources
7 and amicus curiae briefs filed by external parties and by the solicitor general. The results provide persuasive support for the theoretical argument that when clerks send strong information signals they can shape their justices’ behavior.
Chapter Roadmap
The dissertation consists of five chapters. This first chapter outlines the questions raised, situates the project in the literature, previews the research design, and outlines the subsequent chapters.
Chapter Two describes the role of clerk ideology in the justices’ hiring patterns from 1969-2007. The analysis uncovers surprising results in the context of the extant literature on clerk selection. I show that the justices are often willing to hire clerks whose ideological preferences differ from the justices’ own and from those of their fellow clerks. I argue that these results are evidence that studies treating the hiring of individual clerks as independent from the hiring of the other clerks that term overlooks an important, team-building aspect of the justice’s behavior. Conceptualizing a justice’s clerks as composing a team rather than a set of individuals will also highlight the conditions that enable or hinder clerk influence.
Chapter Three investigates how clerks use ideological information to influence their justices’ votes on the merits. I present and validate an innovative measure of clerk ideological preferences that offers several advantages over survey-based measures. This measure captures a clerk’s preferred policy and serves as a proxy for the information she conveys to her justice. Using a research design that accounts for how similar clerks are to each other, I show that uniformly conservative clerks working for a moderate justice successfully pull their justice’s votes on the merits to the right. This is strong but
8 incomplete support for the theory. There is no observable vote pull for clerks with diverse preferences, as predicted by the theory, or for uniformly liberal clerks.
Chapter Four considers a different form of information: that which a clerk has acquired through her socialization and experiences as a female. In an analysis similar to that presented in Chapter Two on ideology, I show that over time many justices have come to assemble teams of clerks with a greater eye toward sex representation. Statistical analysis then shows that having two or more female clerks causes a justice to exhibit greater support for feminist policies. This finding supports the theoretical expectation that female clerks convey information reinforcing their preferred policies and that this information affects the justices’ behavior.
Chapter Five presents an argument for how the research advances knowledge about principal-agent models in general. I expand the scope of the discussion to include hierarchical relationships in general, as well as what scholarship on law clerks tell us about how principals acquire and use information provided by their advisors.
9 Chapter Two: Fielding an Excellent Team: Law Clerk Selection and Chambers Structure at the U.S. Supreme Court1
Introduction
All judges rely on information to help them make decisions. Information helps them to understand the facts and the law and to decide which strategy to pursue. While unobservable psychological processes are what lead a judge to choose a particular strategy, scholars have sought to learn about the inputs to these processes. These inputs include information from numerous sources, including that contained in litigant and amicus curiae briefs Collins 2008, debated during oral argument (Johnson, Spriggs, and
Wahlbeck 2006), articulated by the solicitor general (Black and Owens 2012; Bailey,
Kamoie, and Maltzman 2005), and conveyed by law clerks in in-chambers discussions
(Kromphardt n.d). By studying these sources, scholars hope to learn how judges make decisions and, perhaps, explain their behavior.
There are unique challenges to studying the role of clerks as information sources.
First, these conversations take place in secret and are usually unrecorded, except in memo form. It is also difficult to know what information clerks possess that their judge does not. Finally, there is the challenge of disentangling the mechanism by which this information can influence the judge's decision making.
The relationship between the justices of the United States Supreme Court and their law clerks provides leverage over some of these challenges. Scholars have availed themselves of full lists of every law clerk to serve, and because these clerks often go on
1 This chapter has been accepted for publication in the Marquette University Law Review. 10 to be public figures, it is possible to collect their biographical information. When her2 biography differs from that of her justice, a clerk has the advantage of an information asymmetry and can convey information a justice does not already have. Additionally, because we know the identity of all the clerks a justice has hired, we can analyze how the information environment of a justice’s chambers affects his decision making. Research has shown that when a team of clerks holds unified preferences, some teams have pulled the justice’s vote in their direction successfully (see Chapter 3)
While we can neither sit in on meetings between a justice and his clerks, nor probe his brain as he considers his strategies, we can analyze how he assembles the team of clerks on which he relies. The makeup of these teams reveals clues about what information he seeks to aid his decision making. Some justices desire information from disparate and competing sources, pursuing the logic that the fruits of many minds often produce the best answer. Other justices seek information of a particular ideological nature; this information helps justify voting in their preferred ideological direction and may provide ammunition for persuading other justices and defusing attacks. Studying the team a justice assembles provides scholars with a rare glimpse into how he does his work.
This is not the first study on clerk selection, but to my knowledge it is the first to treat selection as the assembly of a team rather than the hiring of individuals. My subject of interest is the team a justice assembles. Specifically, I will analyze patterns in the ideological characteristics of the justices’ teams from 1969-2007. I discuss two theoretical perspectives on clerk hiring—one in which clerks are agents to the justice’s principal, and one in which clerks are tapped as sources of information—and derive
2 For purposes of clarity, the feminine denotation (i.e. she, her) is used in reference to clerks and the masculine denotation (i.e. he, his) is used for judges and justices. This convention was adopted randomly. 11 implications from each perspective that will facilitate interpretation of data on justice and clerk ideological preferences. These patterns reveal a great deal about the teams of clerks the justices assemble to accomplish their goals. In general, the analysis uncovers variance across the justices and over the justices’ tenures. In particular, the results undermine the notions that a justice’s ideology completely determines the information he seeks and that clerk ideologies always match that of their justice.
This study should be of interest beyond the narrow question of how clerks influence their justices. As I mention above, the teams a justice assembles provide clues about how he does his work. Information about their clerks should join the justices’ comments and released papers as important sources for learning about the day-to-day job of being a Supreme Court justice. The study treats justices as performing an additional role. Scholars are used to looking at justices as role-players, such as members of a collegial group (Maltzman, Spriggs, and Wahlbeck 2000), yet are unaccustomed to treating them in the role of personnel managers. Finally, the study also serves to illuminate a case of how elites engage in personnel management.
Selection of Law Clerks
Clerk selection at the Supreme Court has been the subject of much scholarly scrutiny. In particular, scholars have described the young lawyers who fill the pool from which justices select their clerks, as well as the individual characteristics or criteria on which the justices base hiring decisions. This literature contains clues about the people whom the justices hire and the information those clerks have to provide.
Insight into the motivations of students applying for federal clerkships is limited.3
3 Very few Supreme Court clerks only have federal district court or state court experience. Only 7% of Rehnquist clerks with clerkship experience had only served for a federal district court 12
Wasby (2006) looks at a federal appellate chambers to see why clerks chose that judge
While the results of Wasby’s survey of a single chambers are of limited generalizability,4 the study does provide valuable insight into the application process. In particular, the results are suggestive of treating lower-court clerks as belonging in two pools: those who are interested in applying to Supreme Court clerkships and those who are not. For this latter group, to which Goodwin’s clerks generally belong, factors like the judge’s personality, the Ninth Circuit—both its prestige and the issues and cases its judges deal with—and the Northwest geography played the biggest roles in why respondents applied.
Several scholars have used surveys to investigate what types of students apply for clerkships. Rhinehart (1994) and Avery and his coauthors (2001, 2007) survey students at top schools to learn about their attitudes and experiences with regard to clerking.5
Rhinehart identifies gender gaps in why students do not seek clerkships; in particular, she finds that women are less likely in general to be interested in clerking. While Rhinehart’s sample size is too small to determine whether this difference is statistically significant, and she is unable to draw any firm conclusions about the experiences of racial minorities, this finding implies that otherwise qualified students are not interested in applying for clerkships and that the overall pool is not representative of all otherwise highly qualified
judge (i.e did not also serve with a federal appellate judge) and 0.3% with only a state supreme court justice (Ward and Weiden 2006, 77). For work on other clerks, see Peppers, Giles, and Tainer-Parkins (2008) (district court clerks) and Swanson and Wasby (2008) (state supreme court clerks). 4 This is even more so for explaining the experience of students who go on to be Supreme Court clerks, given that Goodwin fed just two clerks. Indeed, Wasby concludes that “[t]here is one reason (for applying to clerk with Goodwin) that definitely did not apply—as a step toward a clerkship in the U.S. Supreme Court,” and even then, “the judge was not one of those seen as providing that connection” (2006, 418, emphasis in original). 5 Rhinehart only surveys students at eleven top schools who serve on the law review, while Avery and his coauthors use a sampling frame that they argue is more representative of who receives clerkships, including fewer schools but all of these schools’ students. No one to my knowledge has conducted a survey that reaches a broader sample of applicants. 13 young lawyers. If true, this finding has implications for the perspectives the justices hear from their clerks.
The Avery team conducts surveys in order to gain student- and judge-level perspectives on the application process, and to determine how this controversial process affects students’ attitudes toward the judiciary. The authors describe how the clerk hiring market faces a timing problem, where more earnest participants will inevitably circumvent deadlines looking to get a jump on the competition. Survey responses show that students and judges alike are unsatisfied with the process because many participants are left having made suboptimal matches.6 The hectic process understandably also has the effect of driving students away from applying at all: 58% of respondents who did not apply for federal clerkships said they opted out because of either the nature or timing (often fall semester of their 2L year) of the market’s activity (Avery et al. 2001, 829). As one student commented, “[The market] certainly seems like a hellish experience and that definitely contributed to my decision not to apply” (Avery et al. 2001, 894).7 Meanwhile, the authors report that 42% of judges in their survey responded to a closed-ended
6 Numerous remedies to this longstanding problem have been advanced, from creating a common deadline (Becker, Breyer, and Calabresi 1994), to implementing a system modeled off the process used to match medical graduates with residencies (Wald 1995, Oberdorfer and Levy 1992, Haruvy, Roth, and Unver 2006, Avery et al. 2001, 2007), to more market-based solutions (Kozinski 1991, Adams 2000). See also Priest (2005) and Aldisert, Kirkpatrick, and Stevens (2006). 7 While Avery and his coauthors (2001) largely paint a picture of a process whose flaws run deep, their results also show that students are strategic in how they manage the market. Because many students feel obligated to accept their first offer, or are faced with an “exploding” offer with an expiration date, 42% in one survey and 55% in another reported that they narrowed the pool to whom they apply ex ante to include only their most-preferred judges (829). Students who strongly desire a particular experience, be it a certain form of mentorship or to work for a feeder judge, will narrow their pools to include only those judges that fit the bill. Furthermore, applicants are strategic when scheduling interviews, to avoid receiving an exploding offer from a less desired judge (Katz and Stafford 2010, 481). Ample evidence supports the conclusion that students are purposeful in selecting to whom they apply.
14 question that a recommendation from professors with whom they were familiar was either the most or second-most important factor (Avery et al. 2001, 899). Comments from student- and judge-respondents alike strongly suggest that faculty recommendations play a substantial role in the hiring process (Avery et al. 2001, 900). These factors shape the body of information the justices can glean from conversations with their clerks.
Multiple judges have written odes to what they consider the ideal clerk. Chief
Judge of the D.C. Circuit Patricia M. Wald establishes what prominent feeder judges such as herself view as the stakes: “The judge-clerk relationship is the most intense and mutually dependent one I know of outside of marriage, parenthood, or a love affair…[A]n excellent versus a mediocre team of clerks makes a huge difference in the judge’s daily life and in her work product (Wald 1995, 37-38). As a consequence,
“many judges are not looking just for qualified clerks; they yearn for
neophytes who can write like Learned Hand, hold their own in a
discussion with great scholars, possess a preternatural maturity in
judgment and instinct, are ferrets in research, will consistently outperform
their peers in other chambers and who all the while will maintain a
respectful, stoic, and cheerful demeanor” (39).
On top of making the job easier and more pleasant, Wald points to the prestige that comes with being considered a feeder judge. Chief Judge of the Ninth Circuit Alex
Kozinski, who like Wald is a judge feeder and a former court of appeals- and Supreme
Court-clerk, agrees that “judges have a very substantial stake in selecting clerks who are not merely competent, but brilliant; not merely articulate, but lightning fast and prolific; not merely thoughtful, but persuasive and tactful” (Kozinski 1991, 1708); to identify
15 such clerks, he attributes “the type of drive and determination—as well as imperviousness to pain—(that) many judges look for in a clerk” to being successful in the classroom, in publishing a note, doing research, or showing proficiency as a successful moot court advocate (1710).
The foregoing studies reveal a great deal about the pool from which justices hire their clerks. It shows that clerks in their application and acceptance and justices in their offering decisions act purposefully; the pool is not filled at random, so to speak. These processes contribute to two bifurcations: feeder and non-feeder judges,8 and feeder and non-feeder clerks.9 Ideology plays a role in both feeder and non-feeder matches; applicants and judges alike, through conscious and unconscious selection, show a tendency to make ideological matches.10 And while the mechanism driving the disparity is uncertain, female and minority law students are underrepresented in all appellate clerkships.
Characterizing the pool of individuals with appellate clerkship experience is essential because, in recent years, having this experience has become a norm and virtual requirement to be selected for a Supreme Court clerkship. The genesis of this practice is
8 Feeder judges compete for the top law students, and among the reasons these students choose to clerk with feeder judges is their prestige and their track record in placing clerks at the Supreme Court (Avery et al. 2001, 875-877). 9 Nelson and his coauthors (2009) discuss how an increasing number of law students appear to be following an ideological track that continues after their clerking days are over. “A law student who self-identifies as a conservative by, for example, joining the Federalist Society can take the next step forward by clerking for a conservative federal circuit judge, moving on to a conservative Supreme Court clerkship, next serving in a conservative Justice Department, and finally becoming a litigator in a conservative practice group. After two decades in such a career, a smart lawyer will be fully prepared to be appointed to the bench as a reliably conservative judge or Justice” (1797-1798). 10 See Ditslear and Baum (2001). A published conversation between Chief Judge Kozinski and a former clerk of his discusses the phenomenon of judges hiring non-matches (Kozinski and Bernstein 1998). The dynamic between a conservative jurist and liberal clerk will provide fodder for dramatization in a play called The Originalist, about Antonin Scalia (Mark 2014). 16 attributed to a preference by Chief Justice Warren Burger that soon caught on with the rest of the justices (Peppers 2006, 31). Over the course of the Burger Court the percentage of clerks with prior clerkship experience increased from 68% to 95% (Peppers
2006, 31) and 85% of these clerks with experience had acquired it with a federal appellate judge (Ward and Weiden 2006, 77). This trend grew even more pronounced in the Rehnquist Court, with 98% of all clerks having experience and 92% of that subset gaining it with federal appellate judges (Peppers 2006, 31; Ward and Weiden 2006, 77).
The norm has become near law in the Roberts Court: through the 2013 Term, only two clerks did not first gain experience in one of the twelve regional circuits of the federal courts of appeals, although both had served clerkships in the federal judiciary—one in the
Federal Circuit and one in the D.C. District Court.
The use of feeder judges structures ideological hiring of clerks by the justices.11
Baum and Ditslear (2010) find that, “[c]ompared with the random distributions, many more judges actually provided no clerks to the justices, and some judges provided more clerks to the justices than any judge would have sent under a random distribution” (31); in other words, hiring is disproportionately concentrated among a handful of feeder judges. They further observe that “it appears that liberal and conservative justices were drawing clerks largely from different sets of feeders, sets structured by the ideological positions of the feeder judges” (38).12
11 Other factors that appear to affect clerk hiring are the overrepresentation of elite schools (Peppers 2006, 32-31; Ward and Weiden 2006, 69-76) and, as suggested by the underrepresentation of female and minority clerks discussed earlier, sex and race (Brown 1996; Peppers 2006, 20-24; Ward and Weiden 2006, 87-98). 12 In an earlier paper where the authors introduce the idea that lower-court judge ideology signals information to the justices, they find that ideological polarization, as measured by the tendency of conservative (liberal) justices to hire clerks from judges appointed by Republican (Democrat) presidents, has increased in modern times: “In choosing clerks, the justices now give much more 17
The evidence that many recent clerk hires were ideological matches is strong.
Baum and Ditslear (2010) report robust and statistically significant correlations between multiple measures of the ideologies of justices and judges for the period 1995-2004 (38
Table 4). However, their results reveal considerable variance across correlations, making it difficult to assert with confidence precisely how strong the relationship is. This difficulty is corroborated by the often substantial gaps between clerk and justice ideologies that I report later.
There are three possible explanations for this phenomenon. The first is measurement error. At least some measurement error should be expected, as ideology is a notoriously tricky concept to measure (Fischman and Law 2009). Another explanation is that when a justice hires an ideological non-match that it is the result of a mistake. The existence of selection committees, feeder judges, and a large number of qualified applicants across the ideological spectrum makes this explanation less plausible.
Certainly there are enough qualified clerks with judges of known ideology that if justices were only interested in hiring ideological matches, they would be capable of doing so.13
The third explanation is that justices sometimes hire ideological non-matches on purpose.14 Substantial anecdotal evidence suggests that this third possibility—that justices may purposefully hire clerks whose ideological preferences are different from
weight to an ideological signal than they did in the 1970s and the 1980s” (Ditslear and Baum 2001, 882-883.”
13For a discussion of the benefits of having clerks who disagree with you, see Kozinski and Bernstein (1998). 14 The placement records of feeder judges Douglas Ginsburg and J. Harvie Wilkinson may support this third explanation (although the possibility remains that liberal justices hire from Ginsburg for non-ideological reasons, such as reputation for hiring hard workers): Ginsburg is very conservative and Wilkinson less so, but both have placed clerks across the ideological range of justices (Baum and Ditslear 2010, 39 Table 5). 18 their own—does indeed take place (Chapter 3, 35-38), although systematic work has made only fleeting reference to it (Kenney 2000b, 194; Black and Boyd 2012, 23 f.n. 3) and has never explicitly tested how it affects decision making.15
In the next section, I link these observations about the role of clerk ideology in hiring to two perspectives on clerk selection. These perspectives consider the roles clerks play and how the team of clerks a justice hires suggests which role he prioritizes.
Two Perspectives on Chambers Structure
This study is the first to describe empirically how Supreme Court justices take clerks’ ideological positions into account when assembling teams.
There are two perspectives that deal with the concept that I use to characterize these teams, which I call chambers structure, or the particular arrangement of clerks’ preferences with respect to each other and their justice. I discuss these perspectives in this section, including implications that I will rely on in the following section to identify and make sense of patterns in a set of data.
Structural decisions are closely related to clerk selection (Ditslear and Baum
2001; Peppers 2006; Ward and Weiden 2006; Peppers and Zorn 2008; Baum and Ditslear
2010). Selection is both interesting and important, but the literature tends to treat the individual hiring decisions as if they are independent of one another, a choice a justice makes a few times each term. However, a justice does not hire his clerks in isolation of each other; rather, he hires a team of clerks in order to create a particular fit. The quality of this fit is determined by how his clerks help him pursue his goals.
15 An important exception is a study by Peppers and Zorn, who test the effect of chambers structure—how many Democrats a justice employs, for example—on the direction of a justice’s vote. The authors identify several mechanisms through which clerk ideology can influence a justice’s vote, but cannot draw any conclusions about the mechanism driving their findings (Peppers and Zorn 2008, 57-58). 19
These selection studies have shown that is not controversial to assert that ideology has an important function in the decision to hire a clerk. Typically, the interpretation of the results has been that the justices at least attempt to make ideological matches.
However, anecdotal evidence suggests that for at least ten recent justices, hiring decisions were purposefully also a function of the ideological positions of the other clerks who were hired (Chapter 3, 35-38). These justices profess to have the whole chambers structure in mind, not just individual clerks’ characteristics. These anecdotes point to hiring decisions also being structural decisions.
The first perspective on structural decisions treats the justice-clerk relationship as an example of a principal-agent dynamic. The principal-agent perspective is very common in clerks studies (Kenney 2000a; Ditslear and Baum 2001; Wahlbeck, Spriggs, and Sigelman 2002; Peppers 2006; Peppers and Zorn 2008; Rosenthal and Yoon 2011;
Black and Boyd 2012}. These studies assume that the justice-clerk relationship is a classic example of a principal-agent relationship. This theory predicts that the principal can reduce the costs of monitoring and the risk of shirking by hiring agents he believes share his goals (Ditslear and Baum 2001, 870). According to this perspective, we should expect that clerk viewpoints do not diverge greatly from the justice’s, because the justice seeks clerks whose preferences closely match his own.
We can also learn what kind of information a justice receives from a chambers structure in which clerks’ ideal points are unified and lie close to the justice’s. In this environment, it is unlikely that the justice will receive information that supports a position dramatically different from his own. Because of this, a justice who creates this sort of structure probably wishes to reinforce the views he already holds.
20
This discussion suggests the following implications. If we observe little divergence between clerk and justice ideological positions:
• The justice’s hiring behavior is consistent with principal-agent theory; and
• The information asymmetry held by clerks is likely small, and any learning that
takes place will serve primarily to reinforce the justice’s current ideological
position.
The second perspective treats clerks as advisors. At least ten justices in the past 45 years have discussed using their clerks as advisors, tapping them for information about the case at hand (Kromphardt n.d.) This perspective reasons that clerks are well-trained young lawyers who are incentivized to be good confidants to their justice. As a result, clerks both possess useful information and are encouraged to share that information forthrightly.
Clerks have had different life experiences than their justice, such as those acquired perceiving the world through an ideological frame or as member of a minority group.
Information based on those distinct experiences creates an information asymmetry between clerk and justice and thus the opportunity for learning by the justice. The justice learns in two ways. The first is that by tapping his clerks as sounding-boards he increases his overall supply of information, which he can draw upon when reasoning about voting strategies. The second way a justice can learn is by filtering ideological arguments through his clerks, gleaning information about a strategy’s utility by comparing a clerk’s attitude toward it with her known biases (Calvert 1985).
Depending on the information he seeks the advisory role can create reasons for a justice to make structural decisions with a greater eye toward viewpoint diversity. In the
21 next section, I discuss how structure affects information transmission. Meanwhile, the present discussion suggests the following implications. If we observe divergence between clerk and justice ideological positions:
• The justice faces a greater need to implement monitoring in order to prevent
shirking.
• The information asymmetry held by clerks may be large, and the justice can learn
a great deal by discussing cases with his clerks. Learning from his clerks may
cause an observable change in the justice’s behavior; and
• Factors that impede the transmission of information from clerks to their justice
will affect the justice’s ability to learn.
While the implications suggested by the agent and advisor perspectives are not the same, it is imperative to note that the agent and advisor roles are not mutually exclusive; in all likelihood, every justice will seek the benefits of both roles and, to some degree, hire clerks intending to use them as both agents and advisors. One reason to conduct the empirical inquiry I discuss below is to evaluate the applicability of each perspective, both in general and in the context of individual justices’ structuring practices. This is an important exercise because it reveals the environment each justice creates in which to interact with his clerks. By characterizing how justices go about the delicate task of turning to their clerks for aid with their work, we learn about how each individual approaches the job of being a Supreme Court justice.
Structure Types and Information Transmission
Chambers structure—the particular arrangement of clerks’ preferences with respect to each other and their justice—is important because it affects the impact clerks
22 have on the justice’s work. From the principal-agent perspective, a structure in which clerks’ preferences diverge from the justice’s means that, unless the justice engages in monitoring, which can be costly, clerks can shirk from their assigned tasks and pursue their own preferences.
Structure is also one of the factors that can impede the transmission of information alluded to in the third implication of the advisor perspective. This is because structure determines the type of information and affects how it is transmitted from clerks to justice. For example, clerks with broadly similar ideological preferences can convey information consistent with those preferences without interference; in other words, they can transmit a clear signal to their justice. The process is different when clerk preferences are dissimilar and multiple signals are sent. These signals can conflict with each other ideologically, and the resulting mixed signals make it difficult to predict what the justice learns. The content of these signals comes from any information asymmetry between the justice and his clerks. Such information can come from several places, whether through experience perceiving the world through an ideological frame or as member of a minority group, or through research, or through amicus curiae briefs.
This transmission of signals can be made clearer with a theoretical model in which a justice’s chambers consists of two blocs: one composed of all clerks who are to the justice’s ideological left, and one with all those clerks to his right. A model of chambers structure such as this can be used to specify the conditions under which signals can influence justice behavior when they trigger justice learning about the case at hand.
When clerks are in the same bloc, a single signal is sent, and the justice can easily learn
23 about the bloc’s position. However, when clerks are in both blocs and they send multiple, ideologically conflicting signals, the learning process becomes more complicated.
This bloc model is a useful tool for categorizing each justice’s chambers based upon its information environment. Some justices, including Lewis Powell, desired a sort of information free-for-all in which multiple perspectives competed. This sort of environment would lead to the accumulation of lots of information. A Justice would also be able to learn by using his clerks as ideological filters in the manner described earlier. I refer to this sort of chambers as a crosswinds, and it occurs when clerks are in both blocs. Other justices, for example Clarence Thomas, seem to seek a narrower flow of information that is consistent with a uniform set of preferences. The learning that results from acquiring this information reinforces a particular ideological leaning by strengthening the evidentiary base. These chambers I call foxholes, and they occur when clerks are in only one bloc. For example, a conservative foxhole is one in which no clerks lie to the justice’s ideological left.
We suspect that these chambers structures exist because the justices and their clerks say they do. By analyzing empirical data we can describe the extent to which these chambers actually occur.
The agent and advisor perspectives outlined above will help to make sense of these patterns. While the bullet-pointed generalizations do not qualify as causal hypotheses, they will help to guide my interpretation of the data and to evaluate the validity of the agent and advisor perspectives on clerk selection.
Data and Analysis
24
Figures 2.1 and 2.2 plot each justice’s Judicial Common Space (Epstein et al.
2007) (JCS) scores for the period 1969-2007. These scores are derived from the justices’ votes on the merits and are calculated for each term. This dynamic feature allows for an approximate visualization of a justice’s trending ideology over time.
I also plot clerks’ ideological positions with a proxy measure derived from their lower court clerkship experience.16 To do this I position each clerk in the correct bloc based on whether her JCS score is higher or lower than the justice’s and plot the average
JCS score for all clerks in each bloc. This measurement strategy allows bloc and justice ideologies to be situated ordinally and gives a sense of how they relate to each other.
The solid black line running through the middle of each pane connects a justice’s
JCS scores across terms.17 The points in red above this solid line depict the average ideology score of all clerks in the right bloc for that term. The points in blue below the line depict the same for clerks in the left bloc. The physical size of these points varies based on the proportion of clerks in the bloc: larger points are reflective of more clerks.
Blocs with no clerks are omitted from the figure.18
16 I assign each clerk the JCS score of the court of appeals judge for whom she previously clerked. This measure performs well under convergent and construct validation and is theoretically valid because applicants and judges select on ideology (Chapter 3, 49-50) 17 Scores are lagged by one year and capture a justice’s ideological position during the period when he hired his clerks. Because this strategy eliminates freshman terms, I plot a justice’s JCS score from his previous position when applicable. For example, John Roberts’ ideology score for 2005, his first term on the Court, is his JCS score from his tenure on the D.C. Circuit (Giles, Hettinger, and Peppers 2001). 18 I omit Hugo Black and William Douglas from the figures, who never met the coding criteria of having had at least two clerks with court of appeals experience, and John Harlan II, who met it only once. 25
26
27
The figures enable identification of the different types of chambers discussed in the previous section. A crosswinds chambers is identifiable when there is a red point directly above and a blue point directly below a justice’s point (in other words, when a straight line can be drawn through a red dot, a justice's black dot, and a blue dot for a single term). As the figure shows, in 2007, Samuel Alito had a crosswinds. A foxhole chambers is identifiable when there is only a red or blue dot but not both, as Alito had in
2006.
To my knowledge, no study has attempted to describe the structural decisions the justices make. I will draw some general conclusions from these data that constitute the first findings about the important structural decisions the justices make.
One of the most striking things about the figures is the presence of gaps between the ideologies of justices and blocs. In other words, the distance between justice and bloc ideologies is often quite great. While the nature of the data—JCS scores, while continuous, are ordinal—do not permit completely precise descriptions of this distance, it is clear that at times the blocs and the justice are not ideological matches.19 This goes against the prediction of principal-agent theory that the principal will hire agents that hold preferences similar to his own. These gaps are surprising because divergence between a justice and his clerks creates an opportunity for clerks to affect the output of a justice's chambers. Yet we can clearly see in the data that there are often substantial gaps between justice and bloc ideologies.20
19 I do not measure any actual distance between the points. Rather, I use the words “distance” and “gap” here to represent what are essentially different ranks. See Ho and Quinn 2010. 20 It is also worth noting that because the bloc scores are averages they understate the distance some individual clerks lie from a justice. 28
It is important to note that some of these justice-bloc distances are a function of ideological extremeness. According to the data, it would be impossible for a justice whose JCS score approaches the poles to hire ideological matches, much less clerks that have a more extreme ideological rank. This is due to the sample of ideology scores for the pool of courts of appeals judges. The most liberal and conservative JCS scores of courts of appeals judges who fed at least one clerk in the period 1969-2007 are -0.65 and
0.56, respectively. In the same time period, Justices Brennan, Douglas, Marshall, and
Stevens had more liberal JCS scores than -0.65 and Rehnquist, Scalia, and Thomas had conservative scores surpassing 0.56.
For less extreme justices who do not face this problem, why might they not be concerned with hiring ideological matches? One explanation is that the justices who do so seek to obtain information from their clerks that aids in learning about which strategy to pursue. These data cannot tell us for certain why justices make these hires, but future work should try to explain this curious pattern.
Also noteworthy is how bloc scores spike, or vary, over time. For example, over the period 1982-1985, O'Connor’s own score remained fairly constant while the average scores of her Left Bloc clerks were -0.014, -0.53, 0.217, and -0.41. Other justices, including Powell, Souter, Stevens, and White, at times show more attenuated spikes in bloc scores while maintaining constant, moderate scores of their own. According to the theory, these spikes would lead to the justice's receiving different types of information
29 from term to term. Any effect of such spikes may be amplified by some of these justices’ status as frequent swing-vote casters.21
This behavior can be contrasted with that of another swing-vote justice, Anthony
Kennedy. After some initial fluctuation, Kennedy shows remarkable consistency in his
Right Bloc scores. From 1996-2005, his Right Bloc scores fell within the narrow range
0.33-0.46.22 Kennedy assembled more of these chambers than any of his colleagues from
1985-2001. He also cast the swing vote in 28.3% of 5-4 decisions (Enns and Wohlfarth
2013, 1096). Furthermore, in nearly all of these terms Kennedy had assembled a foxhole chambers, which meant there was not a liberal signal being sent to balance what may have been an extremely conservative signal.
21 Enns and Wohlfarth (2013) report that, in 5-4 decisions, O’Connor cast the swing vote 23.7% of the time. Powell cast such votes 23.3% of the time, Souter 4.7%, Stevens 10.9%, and White 23.8%. (1095 Figure 1). 22 I find that this streak has consequences: the probability a moderate justice with clerks who are all more conservative than he is casts a liberal vote decreases as the average score of those clerks moves further away from the justice’s score (Chapter 3, 58-62) 30
Table 2.1 lists the number of times individual justices elected to assemble a foxhole chambers when his JCS score fell within a moderate range of scores. I limit the table to a moderate range – -.35 < j < .35 – due to the problem of ascribing purpose to structural decisions made by the sample of ideologically extreme justices. Some justices on this list are surprising – two of Scalia’s earlier terms qualified him as moderate, while
Rehnquist had seven such terms toward the end of his career. (See Figures 2.1 and 2.2 for justice trends.) I am less concerned with the validity of whether these justices really were
“moderate” than with distinguishing those justices and terms for which we might properly say their structural decisions were purposeful. This subset of ideological values is well within the sample of feeder judge scores and suits this purpose.
Not surprisingly, the vast majority of chambers in general are crosswinds. When making structural decisions, the justices seem to desire more information in the form of more signals from their clerks. If we wish to consider when structural decisions were arguably the most purposeful, it is helpful to look at those justices who made these decisions several times. It may take several tries before a justice settles on what sort of structure he prefers. Of the justices in this sample who faced the decision at least five times, Stewart assembled foxholes the highest percentage of times at 57%, followed by
Powell, Stevens, and White.23 Interestingly, all but two of the foxholes assembled by these four justices were composed of clerks who were more liberal than they were. The justices seem more comfortable making the decision to hire all more liberal clerks than all more conservative ones. That chambers where clerks’ ideologies are unified have been demonstrated to pull their justice's votes on the merits in their direction (Chapter 3) may
23The justice who assembled the lowest percentage of foxholes (and therefore the highest percentage of crosswinds) was Souter at a mere 11%, followed by Blackmun, Burger, and O’Connor. 31 not be a coincidence: the justices may be aware (or learn) of such a risk, and they may avoid it.
Discussion
What does the forgoing analysis tell us about what constitutes an excellent team for most Supreme Court justices? For more moderate justices, the answer is clear: a team of clerks that sends more signals is to be preferred. The learning that takes place after receiving these signals is difficult to predict, and there are exceptions to this trend, but a justice is unlikely to assemble repeatedly a team of all-conservative clerks.
The results also raise questions about the assumption that the justice-clerk relationship can be completely explained using principal-agent theory. The data reveal divergence between justice and clerk ideologies, which, coupled with the extent to which the justices defer to their clerks, suggests at least one of two things is at work. The justices who hire divergent clerks may implement monitoring to prevent shirking on delegated tasks. These justices may also find the risk of defiance tolerable in the face of the information benefits these clerks provide. To the extent that the latter is true, scholars need to draw from a broader theoretical toolbox to explain the role of law clerks at the
Supreme Court. The advisor theory I present in this paper can be a useful complement to principal-agent theory.
While this study is hopefully a stimulus for future work delving into the determinants and effects of the various roles clerks play, there are important limitations to the data and research design. Because the data are ordinal and of a proxy nature, they are not appropriate for making precise distinctions about the distance between a justice and his clerks. Ideology, while a vitally important concept, is notoriously difficult to measure.
32
Future work to refine these measures and the research design needs to take seriously the worthwhile challenge of capturing how clerk and justice ideologies relate to each other.
Also, the research design I employ is purely descriptive: I have not accounted for other factors besides ideology that explain the decision to hire a particular clerk. This decision is undoubtedly multivariate; race, gender, alma mater, feeder judge, and other personal factors undoubtedly play a role. Going forward, research designs should seek to eliminate any systematic factor associated with the error term.
Their limitations and tentative nature aside, the results I report here pose a number of exciting puzzles. Are conservative clerks really more effective at influencing the justices, and if so, what are the consequences and how have the justices reacted? When is a justice willing to accept fewer signals, and therefore less information that could prove useful, from his clerks? Is information obtained from these signals privileged compared to other forms, such as that obtained through oral arguments and amicus curiae briefs?
Why do some moderate justices, like Anthony Kennedy, who arguably have less of an ideological dimension to their voting behavior nonetheless appear to follow an ideological dimension in their clerk hiring?
When we see a justice as a manager and his clerks as a team (or teams), the range of behavior to be explained broadens. Better data collection and research designs guided by clear theoretical predictions will help improve our understanding of the effect law clerks have on the work of the Supreme Court.
33
Chapter Three: U.S. Supreme Court Law Clerks as Information Sources and Justice Decision Making1
Introduction
The nine individuals whose names grace the Supreme Court’s opinions have long been subjects of interest to scholars. Much less attention has been directed toward those who work most closely with the justices: their law clerks. While clerks seem to generate controversy “cyclically” over time (c.f. Kenney (2000b, 200); also see Peppers (2012)), they have inspired far less academic work. We know a great deal about the clerks’ routine tasks thanks to books by Peppers (2006), Ward and Weiden (2006), and Peppers and
Ward (2012), but we know considerably less about their preferences and backgrounds, or how these factors affect their interactions with the justices.
How a justice interacts with his2 clerks will have consequences for the work that leaves his chambers bearing his name. Clerks are believed to have a greater potential for influence when they are delegated responsibilities, such as at the cert-granting and opinion-writing stages. Justices delegate at these stages to free up their own time, and if they do not delegate carefully, clerks will shirk and influence the final product. However, clerks have another, more subtle opportunity to affect their justice’s work: through their role as advisors.
Clerks are tapped as advisors because they are sources of information, and are used in much the same way a justice relies on amicus curiae briefs (Collins 2008), the solicitor general (Bailey, Kamoie, and Maltzman 2005; Black and Owens 2012) and attorneys
1 This chapter has been accepted for publication in the Journal of Law & Courts. 2 For purposes of clarity, the feminine denotation (i.e. she, her) is used in reference to clerks and the masculine denotation (i.e. he, his) is used for judges and justices. This convention was adopted randomly. 34
(Johnson, Wahlbeck, and Spriggs 2006; McAtee and McGuire 2007; Corley 2008; Szmer and Ginn 2014). Because the justices discuss pending cases with their clerks, they open their decision making up to clerk influence. As the fruit born of this discussion, influence may be observable at stages of decision making believed to be more immune to the phenomenon, such as the justices’ votes on the merits.
This paper seeks to understand this process by identifying conditions under which the information clerks transmit can be detected at the merits-voting stage. I argue that structural features of a justice’s chambers affect the ability of information transmitted by the clerks to influence decision making. Using a novel measure of clerk preferences that I also validate, I find that information stemming from the clerks’ ideological views leads to change in voting behavior, but only in chambers where clerks’ views are similar. Further,
I find that clerks’ ability to influence is asymmetric: conservative clerks are effective while liberals appear not to be.
Books by Peppers, Ward, and Weiden expanded our understanding of the role of law clerks at the United States Supreme Court. Leveraging theirs and others’ work, I argue that it is fruitful to think of clerks as performing multiple roles. By analyzing how clerks function as advisors, I contribute to our understanding of how the justices seek out and use information to aid their decision making. The conditional and asymmetric nature of clerks’ ability to shape behavior that I find also has implications for other hierarchical relationships in which a supervisor relies on subordinates. Specifically, I find that it is important to understand how structure moderates the flow of information.
Clerks as Agents and Advisors
Supreme Court justices no longer, to paraphrase Justice Louis Brandeis, “do their
35 own work” (c.f. Rosenthal and Yoon (2010, 1307)). Their law clerks assist them in several aspects of their work, including processing cert petitions, conducting research, and writing opinions. While one reason for having their clerks perform these tasks is for the justices to reduce their own workloads, another, supported by the words of the justices themselves and their former clerks, is to tap their law clerks as useful sources of information. These distinct goals underlie two clerk roles that I label as time-savers and advisors.
The time-saver, or agent, role is motivated by the justice’s desire to ease his workload. Over the history of the clerkship institution, clerks have assumed more and greater duties. As the work of the Court has grown more complex and time-consuming, the justices have relied on their clerks to assume more of what have been called the delegable tasks (Posner 1995). The modern clerk’s delegated duties include, generally, reading and making recommendations on writs of certiorari, preparing bench memos that summarize litigants’ briefs and additional research, and writing drafts of opinions
(Peppers 2006).
This delegation of tasks fits into the time-saver role if the justice only delegates in order to cut back on his own workload; it is unlikely that direct benefits are accrued to anyone but the justice when clerks complete these tasks, because the justice becomes freer to pursue other activities. The justice who delegates faces considerable risk, however. Principal-agent theory explains that without proper monitoring, clerks are freer to shirk and pursue their own interests.3
To date, scholarly studies have found little systematic evidence of untoward,
3 Scholars have long recognized that the justice-law clerk dynamic is an example of a principal- agent relationship (Kenney 2000a; Ditslear and Baum 2001; Wahlbeck, Spriggs and Sigelman 2002; Peppers 2006; Peppers and Zorn 2008; Rosenthal and Yoon 2010; Black and Boyd 2012). 36 intentional influence arising from the clerks’ role as time-savers. Influence can be defined as action by a justice that he would not have taken absent some intervention from his law clerks. Scholars have searched for influence at multiple stages, including cert granting (Brenner and Palmer 1990; Brenner 1997; Palmer 2001; Stras 2007; Black and
Boyd 2012; Black, Boyd, and Bryan n.d.), merits voting (Brenner 1992; Palmer and
Brenner 1995; Peppers and Zorn 2008), and opinion writing (Wahlbeck, Spriggs and
Sigelman 2002; Rosenthal and Yoon 2010); research generally shows that, with the possible exception of opinion writing, the justices apparently design effective institutions to constrain clerk shirking.
An area that has received considerably less scholarly attention, despite its potential as a source of influence, is the clerks’ role as advisors. Whereas the justice’s goal with the time-saver role is to make his job easier, the advisor role is motivated by his desire to gather information.4 The justice can increase his understanding of a legal issue and its implications by obtaining his clerks’ personal input.5 For instance, when the justice is deciding how to vote on the merits, clerks may be able to provide information based on their personal experiences or research that reduces the justice’s uncertainty about the consequences of available strategies. Reducing uncertainty could help the justice pursue two goals: to help him identify the optimal outcome, without regard to ideology, or to serve to reinforce a desired ideological position by producing evidence in its favor. As I
4 Lazarus (1999) chronicles an exchange in OT 1952 that occurred between Justice Robert Jackson and his clerk, a 27-year-old William Rehnquist, that is a clear example of a clerk transcending service as a mere time-saver to function in the advisor role. During the opinion- writing process for Brown v. Allen (344 U.S. 443), a case concerning the jurisdiction of federal courts in state habeas cases, Rehnquist repeatedly lobbied Jackson by memo to adopt his (Rehnquist's) views about states rights in the concurring opinion Jackson was writing (143-146). 5 This is a manifest difference between the two roles. As time-savers, clerks are rarely expected to provide their personal input; they are supposed to mirror the justice's mind. As advisors, however, clerks providing personal input may help the justice acquire useful new information. 37 will argue in greater detail below, the way in which a justice structures his chambers contains clues about which of these two goals he pursues in his decision making.
The advisor role reduces uncertainty by facilitating information gathering in two ways. The first is that it increases the justice’s overall stock of information. Just as a justice is a product of his environment, so too his law clerks are of theirs. By virtue of having had different experiences – for example, as a member of a different gender or racial group, or of a different generation, or from viewing the world through a different ideological lens – clerks will have obtained information from such experiences that the justice may not have, and this difference may be salient to issues that come before the
Court. Key to this form of information gathering is that the clerks possess an information asymmetry from which the justice can learn. By discussing strategies with his clerks and seeking their input, the justice gains information with a fairly low-cost approach.
A second way the advisor role is a source of information is that it enables the justice to filter his clerks’ advice through their known biases. A bias that is likely to be helpful for this purpose is the clerk’s ideology.6 Calvert (1985) shows the “value of biased information”: when a decision maker consults advisors with known biases, he can interpret their advice in the context of what he expected. When the advice goes against his expectation, the decision maker may learn about a bad outcome and avoid making a mistake. This would occur, for example, when a justice interprets his conservative clerk’s begrudging endorsement of a liberal case disposition as a signal that the liberal outcome has merit.
Influence arising from the time-saver and advisor roles are distinct phenomena, but
6 Indirect (Ditslear and Baum 2001; Chapter 2) and direct (Peppers and Zorn 2008) empirical regularities suggest that ideology structures clerk hiring decisions, implying that justices are aware of their clerks' ideologies. 38 the roles themselves are not mutually exclusive. In all likelihood, most justices will use clerks in both roles, perhaps even simultaneously. However, treating the roles separately when theorizing how influence can occur raises empirical implications that are unique to the advisor role. Influence from the time-saver role is undesirable, because it is a symptom of shirking—the output bears the clerk’s fingerprints when the justice does not want it to. Influence from the advisor role, however, occurs as the justice seeks to identify or to provide reinforcement for his optimal voting strategy. As I argue below, this learning is conditional on the information clerks have to offer and the structure of clerks’ preferences. Before discussing how to model chambers structure and clerk preferences, I present evidence suggesting that clerks do actually perform the advisor role.
Evidence of the Advisor Role
While most 27-year-olds are typically not founts of jurisprudential wisdom, historical and anecdotal evidence shows that the justices nevertheless find testing their clerks’ minds to be worthwhile. This is not surprising, given the caliber of the clerks’ minds and that institutional pressures reinforce clerk allegiance. The prototypical modern
Supreme Court clerk has received training from a top law school and tutelage from a federal circuit judge. For her service as a clerk, she can expect a flurry of job offerings, and, at many firms, a hiring bonus well into six figures. Training and self-interest, in other words, render clerks uncannily dependable resources for the justices.
There is evidence that at least ten justices who served within the past 30 years relied on their clerks as advisors.7 Because of the importance of recognizing that the
7 Because of the secretive nature of the decision making process and a general unwillingness by justices and clerks to discuss intra-chambers matters, it is important to remember that an absence of evidence for some justices is not evidence of absence. In other words, just because we do not have accounts indicating a justice used his clerks as advisors does not mean that he did not in fact 39 time-saver and advisors roles are distinct phenomena, I briefly recount this evidence. A significant account comes from the chambers of Byron White. Clerk Kevin Worthen recounts that “[f]or Justice White, the judicial decision-making process was a two-step process: first make sure the problem has been fully considered, and, second, decide. The primary role of the law clerks was to assist with the consideration, not the decision”
(Worthen 2012, 305). Clerk David Kendall offers a more colorful account of this dynamic, remembering that White “wasn’t invested in an argument; if you could hit him back with a chair, intellectually speaking, he could be convinced” (Worthen 2012, 305).8
There is less vivid but still persuasive evidence from intra-chamber memos and clerk surveys intimating that William Rehnquist and Harry Blackmun’s clerks would try to influence their justice’s thinking and that these particular justices encouraged such behavior (Ward and Weiden 2006, 155, 192, 165-166). Similarly, Alex Kozinski recalls lobbying Warren Burger during his stint as Burger’s clerk, and “[s]ometimes he would switch, and sometimes he wouldn’t” (Kozinski and Bernstein 1998, 58). Comments by a former clerk in a recent profile of Anthony Kennedy illuminate how he runs his chambers: “He wants to talk [a case] over (with his clerks). He wants the roundtable”
(Calabresi and Drehle (2012, 36); see also Peppers (2006, 203)). Lewis Powell and John
Paul Stevens reputedly preferred diversity among their clerks. Federal judge and former
Powell clerk J. Harvie Wilkinson III wrote that “[J]ustice (Powell) preferred that his law clerks be of different persuasions and stir ‘crosswinds,’ as he once put it, through his chambers” (Wilkinson 2012, 346). Similarly, Peppers (2006) describes Stevens as do so. Ten can be considered a conservative estimate. For a possible eleventh, see the description in Lazarus (1999) of Justice O'Connor's chambers (391-394) 8 White clerk Stephen McAllister presents this type of exchange in a more cynical light, describing the Justice as having “humored the law clerks by discussing pending cases as some length with us" (Peppers 2006, 164). 40 engaging his clerks “in what might be best described as a Socratic dialogue, a joint inquiry toward the ‘right’ outcome” (196). A former Stevens clerk has written, “I can imagine few bosses so interested in the views of their employees, so prepared to engage in free-flowing debate, and so enthusiastic to be proven wrong” (Pearlstein (2000, 13), c.f. Peppers (2006, 197)). Finally, a study of clerk “fingerprints” in opinions from two chambers indicates that language crafted by Thurgood Marshall’s law clerks in opinion drafts was recognizable in final opinions (Wahlbeck, Spriggs and Sigelman 2002), from which we might infer that Marshall’s chambers, at least during the 1985 term, were fertile grounds for clerk influence attempts.
Drawing inferences based on hiring patterns that clerks were used as advisors enables me to add two more justices to the list. Evidence suggests that ideology plays a role in hiring decisions. That she was handpicked by the justice from among a large pool of qualified applicants should further solidify a clerk’s status as trustworthy. Previous research has demonstrated a positive correlation between the ideologies of lower-court feeder judges (Ditslear and Baum 2001; Baum and Ditslear 2010; Kromphardt 2014) or the clerks themselves (Peppers and Zorn 2008) and the justices who hire them. The most plausible explanation for this correlation is that the justices often seek to hire ideological compatriots. This makes sense from a time-saver perspective: justices will worry less about shirking if they believe the clerks to whom they delegate share their preferences.
We know that for at least one justice, hiring ideological compatriots also makes sense from an advisor perspective. Clarence Thomas only hires clerks who share his ideological vision (Peppers 2006, 200). Thomas has compared clerk hiring to “selecting mates in a foxhole” (Ditslear and Baum (2001, 883); Liptak (2010)), explaining his
41 reasoning that “[he] won’t hire law clerks who have profound disagreements with [him].
It’s like trying to train a pig. It wastes your time, and it aggravates the pig” (Savage
2011). This is not simply rhetoric: prior to the 2013 term, Thomas had never hired a clerk from a judge appointed by a Democratic president, a streak of nearly 100 clerks.9
We may, with some qualification, add Antonin Scalia to this same category, and bring the list to ten recent justices. When he was a federal circuit court judge, in order to avoid, as Liptak writes, the “ideological orthodoxy (that) can dampen the robust discussions in chambers that clarify issues and shape rulings,” Scalia would hire a token liberal clerk (Liptak 2010).10 In recent years, however, this particular practice seems to have all but ended.11
Reasons for Hiring Clerks as Advisors
It is apparent that while many justices use their clerks as advisors, they may not share the same reasons for doing so. More specifically, the evidence suggests that there are two goals underlying a justice’s decision to hire clerks in order to engage with them as information sources. These two purposes are to reinforce a desired ideological position and to determine the optimal strategy regardless of ideology.
Justices such as Thomas prize ideological harmony among their clerks. A justice who seeks clerks with broadly similar views may do so for purposes of minimizing clerk shirking, but there is another reason assembling such a team of clerks is desirable.
9 The streak ended with Thomas's hire of Rebekah Perry Ricketts, who clerked for Jose A. Cabranes, a Clinton-appointee on the Second Circuit. 10 An alternative reason Scalia may have done this would be to consider the strengths and weaknesses of “liberal" arguments in advance of conferring with his colleagues so as to fortify his own position. 11 The last Scalia clerk with a negative (liberal) Judicial Common Space score was Gil Seinfeld (OT 2002), who clerked for Guido Calabresi. 42
Thomas has been noted to engage his uniformly conservative clerks12 in conferences to discuss upcoming cases that “were the most uninhibited and wide-ranging on the Court”
(Thomas (2002, 464-465), c.f. Peppers (2006, 200); see also Benson (2007, 41, f.n. 119)).
A fair interpretation of this practice is that Thomas uses his clerks as advisors for the optimal application of his conservative views. To generalize, a justice who hires a team of clerks who share similar preferences may seek to generate a critical mass of arguments that reinforce a preferred ideological outcome. This information serves to support strongly held views and may also provide persuasive material for lobbying other justices.
Such chambers I categorize, inspired by Thomas’s memorable turn of phrase, as Foxhole
Chambers.
Other justices express a desire for ideological diversity among their clerks. The anecdotes above suggest that for Stevens, the purpose for doing so was to stimulate “a joint inquiry toward the ‘right’ outcome” (Peppers 2006, 196). While it is unclear what qualifies as an objectively “right” outcome, the method of assembling and engaging with a diverse team of clerks would enable a justice to evaluate multiple strategies across a variety of criteria. The justice could gather points of view from across the ideological spectrum. Moreover, he would have an in-house method for gauging the validity of these perspectives by filtering arguments through clerks he expects to disagree with them.
Another understanding of the “right” outcome is one that avoids making mistakes, which was Scalia’s purported reason for preferring diversity and hiring a token liberal clerk.
Whichever benefit of diversity a justice seeks, it is clear that there are good reasons to use such a method. Judge Wilkinson’s remembrance of Powell provides the label for this
12 Thomas's attitudes toward his clerks seem to be unique, including his practice of inviting his new clerks to his home for a viewing of The Fountainhead (1949) (Savage 2011) and his idiosyncratic willingness to look beyond the major feeder schools when hiring clerks. 43 category of chambers, a Crosswinds Chambers.
Scholars have written before about the advisory role clerks play (Kenney 2000a,b;
Ward and Weiden 2006; Black and Boyd 2012), but the categorization of foxholes and crosswinds and the outlining of reasons for assembling such chambers are novel. More importantly, no existing study has developed a theory explaining the advisory role’s purpose as providing information for use in decision making and tested the implications of this mechanism using empirical data.13
In the next section, I establish what sorts of information clerks realistically will provide if they operate as advisors, such as information about policy views based on their own experiences (Tate 1981; Tate and Handberg 1991), and develop a model that accounts for the information clerks convey and the structural conditions under which this information leads a justice to change his vote. This discussion motivates a research design to observe informational influence on the justices’ votes on the merits.
Ideological Blocs and the Pull Model
To claim evidence of influence, I need to show that factors attributable to clerks resulted in a change in the justice’s behavior. To depict the conditions under which information a justice acquires from his clerks can result in a change in behavior, I develop the Pull Model. The Pull Model emphasizes the structure of a justice’s chambers as consisting of blocs of law clerks, as determined by the actors’ ideal points.
By conceptualizing chambers structure in this way – as composed of two ideological clerk blocs – I can model a mechanism of information transmission and
13 Peppers and Zorn (2008) is the closest analogue to the analysis I conduct below, but the authors do not explicitly articulate an informational mechanism. That is, while their findings are consistent with the phenomenon of justices’ votes being influenced by information gleaned from their clerks, as the authors admit, their “findings offer no support for any particular causal model of that influence” (75). 44 influence. In this section, I discuss salient characteristics of these blocs and the conditions under which influence, defined as action taken by a justice that he would not have absent clerk intervention, occurs. I develop the hypothesis that structural arrangements condition when bloc characteristics will explain influence, in which a bloc pulls the justice’s vote in its direction. The hypotheses I derive can also be tested empirically, using the measure of clerk preferences I introduce in the next section. Features of these blocs will be operationalized to measure the information a bloc transmits to the justice as a signal.
Treating chambers as composed of blocs is a flexible and general approach for modeling information transmission. To begin with, every justice’s chambers consists of blocs. Blocs are characterized and distinguishable by the unique information clerks possess to convey to their justice. A bloc can have an informational advantage, or asymmetry, in which clerks possess information their justice does not. This asymmetry is important because a justice can learn from it. Recall that a justice can learn in two ways: by increasing his overall stock of information and by using his clerks as ideological filters. This new learning may lead him to change his behavior, and because clerks conveyed the information that led to a change in behavior, this is considered clerk influence.
One source of an informational asymmetry is clerks’ unique life experiences. It makes sense to characterize blocs only in terms of experiences that are likely to be relevant to decision making in a judicial context (Posner 2009; Tate 1981; Tate and
Handberg 1991), such as those arising from a clerk’s ideological views, gender, or race, to name a few. In the data I test, the content of the asymmetry is ideological information that has been accumulated as a result of perception through an ideological frame. The
45 theory, however, applies to any instance in which clerks possess an asymmetry from which their justice learns something relevant to the case he is deciding.
Before I discuss how structure conditions how clerks send information, I discuss the content of that information. A bloc, say one consisting of all liberal clerks, sends an informational signal that reinforces a specific ideological outcome; here, a liberal outcome. This signal consists of information that is relevant to the case at hand but that a justice does not already have, due to an asymmetry. Recall that there is an information asymmetry because clerks possess information as a result of their personal experiences, such as experiencing the world and politics through an ideological frame.
In addition to their content, the transmission of signals must also be conditioned on how likely that information will successfully lead the justice to change his vote. More specifically, how does the structure of clerks’ preferences affect the likelihood a justice votes in a particular direction because of the information his clerks convey? Below, I test two ways of conceptualizing structure: whether clerks are liberal or conservative and whether the clerks compose a foxhole or crosswinds. While both tests produce evidence consistent with the hypothesis that the structure of clerk preferences influences the direction of justice voting behavior, the second test has the advantage of the more specific causal mechanism.
The condition that clerks have generally harmonious preferences may be critical for predicting whether a bloc of clerks can pull their justice in their direction. When unified, clerks send a single, consistent signal advocating a specific ideological outcome. Because there is only one signal and it reinforces a particular position, the justice can learn about the clerks’ position through the signal’s content and be convinced to move his vote in the
46 clerks’ direction.
The learning process is different when clerk preferences are dissimilar and multiple blocs send competing signals.14 The justice would receive signals supporting opposing outcomes. In this context it is difficult to know what precisely the justice learns. Signals can conflict with each other ideologically, and without knowing anything of the relative quality of the signals, we cannot judge their convincingness. Without knowing which bloc makes a better argument, I am unable to form any a priori predictions about how the justice’s vote will be pulled in such chambers.
Conceptualizing Signal Content and Chambers Structure
It is possible to operationalize variables relating to signal content and structure, the independent and modifying variables, respectively, with varying degrees of precision.
While greater precision gives us a clearer picture of the causal mechanism, it requires making assumptions about measuring clerk ideological preferences. Scholars of judicial politics are well aware that measuring ideology is, paradoxically, one of its most important yet contentious tasks. Before turning to the options for operationalizing content and structure, I discuss the theoretical concerns relating to each variable.
The theoretical expectation is that the ability of clerks’ signals to pull their justice’s vote in their preferred direction is conditioned on chambers structure. This expectation would be born out by evidence that a justice’s voting behavior is distinguishable across different structural arrangements, what I have called vote pull.
14 While I hesitate to draw firm conclusions from Lazarus's Closed Chambers (see Kenney 2000b), his account, in a chapter entitled “The Triumph of Politics," of the intra-chambers debates among O'Connor's and Kennedy's clerks prior to the abortion case Webster v. Reproductive Health Services (492 U.S. 490 (1989)), is a vivid depiction of blocs clashing. The specificity of this account, backed by support from my measurement scheme predicting values that match well with Lazarus's description of specific clerks' ideologies in Kennedy's and O'Connor's chambers, leads me to accept his version as sufficiently accurate. 47
The theory is general enough to allow us to consider these structural arrangements in at least two different ways. The first requires minimal theoretical assumptions: a vote in a chambers of all liberal clerks will be distinguishable from a vote in a chambers of all conservative clerks, ceteris paribus. This is the basic test that Peppers and Zorn conduct and attribute as evidence of clerk influence. Among this approach’s virtues are the ease of interpretation and the relative lack of contestable assumptions about ideology.
However, as they readily admit, their “findings offer no support for any particular causal model of that influence” (2008, 75). That is, while theirs is evidence consistent with the theory that liberal (conservative) clerks can craft a signal that pulls the justice in their direction, it is an opaque mechanism. I have no wish to criticize Peppers and Zorn; theirs is a novel study of a hard-to-observe phenomenon like clerk influence on decision making processes. Rather, by standing on their shoulders I hope to get closer toward establishing a mechanism for how this influence takes place.
I do this in two ways. First, I use greater specificity in conceptualizing the independent and conditioning variables. I measure clerk ideology at a higher level of precision—continuously ordinal rather than binary—in order to capture relative ideological positions. I use this new measure to specify two characteristics related to the informational signal—Distance and Direction, which I operationalize in the next section—and one characteristic related to chambers structure, Proportion, which enables me to distinguish foxhole from crosswinds chambers relative to the justice’s ideological position. More specifically, the measure allows me to compare clerk and justice ideological positions ordinally in order to determine the effects of ideological extremeness in clerks; for instance, do we observe greater pull in chambers containing
48 ideologically extreme clerks? The theory I outlined above suggests we might: these clerks have larger information asymmetries compared to clerks who are more proximate to their justice (see Chapter 2, 22-23). The second extension I make is to control for a broader array of competing factors affecting the dependent variable.
Measurement and Data
The theoretical model discussed above specifies conditions under which information can influence a justice’s voting behavior. In order to observe influence, or
Vote Pull—which I operationalize as a change in the probability of a liberal vote across the range of values for the independent variable—I run a series of statistical models.
The first model is essentially a replication of Peppers and Zorn’s statistical model of clerk influence on votes on the merits. To conduct the replication, I construct an independent variable Clerk Liberalism to measure what is essentially the same latent variable as Peppers and Zorn’s main variable Clerk Partisanship. The dependent variable is whether a justice casts a liberal vote. I use logistic regression, and I control for competing explanations for the dependent variable by including the set of informational influences Collins uses in Chapter Four of his book (2008). I discuss the measurement of these variables below.
The second model is designed to be a more specific test of the informational theory I outline above, in which chambers structure modifies the effect the clerks’ informational signal can have on their justice’s voting behavior. I use the same data and dependent and control variables; what differs is the measurement of the independent and modifying variables. These variables—Direction, Distance, and Proportion—capture the informational content of the signal as well as the structure of the justice’s chambers.
49
Measuring Clerk Preferences and Bloc Direction, Distance, and Proportion
Construction of these variables requires information about clerk preferences that no available measure captures. Specifically, it requires an at least ordinal-level measure of clerk preferences that can be ranked with justice preferences. Proportion and Direction require determining whether a clerk is more liberal or more conservative than her justice and Distance requires determining how far away each clerk’s ideal point lies from the justice’s.15
The measure I employ assigns a clerk the Judicial Common Space (JCS) score of the federal circuit court judge for whom she previously clerked (Epstein et al. 2007). JCS scores are available for all federal judges and range continuously from -1 (liberal) to 1
(conservative). The measure has several advantages. First, it satisfies the requirements for constructing Direction, Distance, and Proportion. JCS scores are continuous, ordinal, and directly comparable across the different levels of the federal judiciary (Ho and Quinn
2010), which allows us to determine each bloc’s Distance and its Direction and
Proportion, respectively. The measure’s continuousness also enables more precise description than binary (Peppers 2006; Peppers and Zorn 2008) or discretely ordinal
(Ward and Weiden 2006) measures that collapse clerks into broad categories and therefore capture gradation in ideology with less specificity. For instance, a binary measure classifying a clerk as either “Democrat” or “Republican” cannot depict ideological extreme- or moderateness. Such a feature is desirable for fully describing the nature of the information asymmetries.
15 While continuous, JCS scores are derived from ordinal Martin-Quinn scores (Martin and Quinn 2002; Ho and Quinn 2010). Because of this, we cannot be sure that the distance between values is constant. As a result, the variable Distance should be interpreted as a ranking, with the caveat that the distance that results from the JCS measurement strategy may not be constant across all values. 50
A second advantage is that the measurement scheme assigns nearly every clerk a score. The scheme leverages the modern regularity, which has been recognized as a norm beginning with Burger’s tenure as Chief Justice (Peppers 2006, 31), of Supreme Court clerks having clerked for a circuit court judge. Over the course of the Burger Court the percentage of clerks with prior clerkship experience increased from 68% to 95% (Peppers
2006, 31), and 85% of these clerks with experience had acquired it with a federal appellate judge (Ward and Weiden 2006, 77). This trend grew even more pronounced in the Rehnquist Court, with 98% of all clerks having experience and 92% of that subset gaining it with federal appellate judges (Peppers (2006, 31); Ward and Weiden (2006,
77)).16 Because it assigns almost every clerk a score, the measure is not susceptible to issues arising from survey response rates, such as small sample (Ward and Weiden 2006,
103-105) or selection biases.17
Finally, the measure is not subject to respondent misrepresentation or misunderstanding. While there is no specific reason to suspect that clerks misrepresent their preferences in their responses, Ward and Wasby (2010) recommend that interview and survey data “should be supplemented with data developed through additional methodological approaches” (16). The measure takes that recommendation seriously and is constructed based on the empirically verifiable fact of each clerk’s previous clerkship(s).
The literature supports the assumption that a Supreme Court clerk’s ideology matches her circuit court judge’s ideology, by demonstrating that applicants are aware of
16 Through 2013, only 2 Roberts Court clerks did not have experience clerking in one of the twelve federal geographic courts of appeals, but both had served federal clerkships, one in the Federal Circuit and one in the D.C. District Court. 17 The authors of one study (Peppers and Zorn 2008) find no systematic bias between respondents and non-respondents of their survey. 51 and take into account judge ideology, and vice versa. Kozinski and Bernstein (1998) is a discussion between a conservative feeder judge, Alex Kozinski, and his liberal clerk. The conversation openly discusses how ideology plays a role in two selection processes: for students, to whom to apply, and for judges, whom to hire. Ideology may not always trump other factors, like judge prestige or the availability of qualified applicants (57-58), but it is a factor.18 19 Wasby (2006) presents evidence that even students who apply to moderate judges and who are not necessarily positioning themselves for feedership to the
Supreme Court, based their decision to apply on ideology (419). Wasby explains that when applying to Judge Goodwin, “prospective clerks may unconsciously have gravitated toward someone who was politically like them" (429).
In the same conversation with his clerk, after paraphrasing a liberal colleague,
Stephen Reinhardt, who he says hires only “committed liberals, who are going to spend their careers promoting liberal causes,” Kozinski described his own attitudes this way:
“I think I owe an extra measure of consideration to conservative and
libertarian law students. First of all, I feel an obligation to train conservative
and libertarian lawyers. There are a lot of liberal judges out there, not as
18 Avery et al. (2001) show that students are strategic in how they manage the market. Because many students feel obligated to accept their first offer, or are faced with an exploding offer with an expiration date, 42% in one survey and 55% in another reported that they narrowed the pool to whom they apply ex ante to include only their most-preferred judges (829). Students who strongly desire a particular experience, be it a certain form of mentorship or to work for a feeder judge, will narrow their pools to include only those judges that fit the bill. Furthermore, applicants are strategic when scheduling interviews, to avoid receiving an exploding offer from a less desired judge (Katz and Stafford 2010, 481). 19 Nelson et al. (2009) discusses how an increasing number of law students appear to be following an ideological track that continues even after their clerking days are over. “A law student who self-identifies as a conservative by, for example, joining the Federalist Society can take the next step forward by clerking for a conservative federal circuit judge, moving on to a conservative Supreme Court clerkship, next serving in a conservative Justice Department, and finally becoming a litigator in a conservative practice group. After two decades in such a career, a smart lawyer will be fully prepared to be appointed to the bench as a reliably conservative judge or Justice" (1797-1798). 52
many conservatives and libertarians. Second, there are a lot of cases, and
having a clerk who basically agrees with me makes for an easier year” (58).
This passage indicates that two prominent feeder judges, Reinhardt and Kozinski, are aware of their clerks’ ideologies when they are hired, and that they base decisions on that information. These feeder judges have the luxury of choosing among the best applicants, and do so with an eye toward burnishing and maintaining their own reputations.20 Even if there are occasional mismatches at the hiring stage, a feeder judge who depends on his reputation to attract future clerks and to maintain positive relations with the justices is likely to only recommend for feeding clerks with reliable ideologies, i.e. those that match his own, to maintain credibility.
I conduct two analyses to support the validity of assigning Supreme Court clerks the JCS scores of their lower court judges. First, I collected data on 42 former Supreme
Court clerks who became federal judges or members of Congress – clerks, in other words, for whom there are actual Common Space scores (Poole 1998).21 I correlate the assigned, feeder-judge JCS scores with the actual measure of ideology of the Common
Space score, anticipating a positive relationship if both variables measure the same underlying dimension. The moderately positive correlation of .392 supports this expectation and can be considered convergent validation (Adcock and Collier 2001, 540-
542).
The replication of Figure 4 in Peppers and Zorn (2008, 74) that I present below pulls double duty as the second validation analysis. The replication shows that the
20 See generally Katz and Stafford (2010) for extensive discussion on the importance of reputation in the market for clerks. “[J]udges and communities of jurists who consistently share clerks probably do so because the receiver either respects the judgment of his or her colleagues or otherwise shares a social connection with the senders" (482). 21 See the Appendix for the full list of these former clerks. 53 assigned measure behaves very similarly to the direct measure of clerk partisanship
Peppers and Zorn use to predict the direction of votes on the merits. This result can be considered construct validation (Adcock and Collier 2001, 542-543), and is further evidence of the validity of treating assigned JCS scores as proxy measures of clerk ideology.
In sum, among the virtues of the proposed measure are that it enables the construction of bloc characteristic variables Direction, Distance, and Proportion; is continuous, available for almost every clerk, and not survey-based; and theoretically and empirically valid.
Controlling for Other Information Sources
The primary hypothesis is the signal a bloc sends will pull a justice’s vote in its direction when the chambers structure is unified. I test this hypothesis with two sets of statistical models. The first analysis measures the signal’s content with Clerk Liberalism, which is the proportion of liberal clerks a justice has hired.22 The second analysis is composed of a set of models that measures the signal’s content with covariates for each bloc’s Proportion and Distance in each of the three chambers types: crosswinds, liberal foxhole, and conservative foxhole. I split the data along chambers type because the predictions are conditional on chambers type and interpreting a three-way interaction with at least two continuous variables—Distance and Justice Ideology—poses a daunting challenge for interpretation.
For each model, I include controls for alternative informational influences on a justice’s vote from Chapter Four of Collins (2008). I repurpose Collins’ model into one
22 I code a clerk as having a liberal ideology if her proxy JCS score is ≤ 0 and a conservative otherwise and construct a chambers-level variable that describes the proportion of liberal clerks. 54 testing for the general occurrence of vote pulling by using Peppers and Zorn’s method of pooling data across justices and terms. In other words, I am not looking for influence in one term or for only one justice; rather, I am interested in exploring the general effect of unified and disparate chambers structures on decision making.
Collins conducts a rigorous analysis of amicus curiae influence that is bolstered by a theory that, while it does not explicitly account for law clerk influence, is consistent with the logic of law clerk influence I outline. This is because amicus briefs contain information not provided by the litigants’ briefs, in the same way that clerks provide information the justice does not already possess. For instance, Spriggs and Wahlbeck
(1997) and Collins (2008) find that 67% and 70% of their samples of amicus briefs, respectively, contain new evidence, arguments, and policy ramifications not found in the parties’ briefs.
The model and data are thoroughly discussed in Collins’ book, so I will summarize them here. The unit of analysis in both of my models is the justice-vote and the dichotomous dependent variable Vote Pull is whether a justice casts a liberal vote (1 for liberal, 0 for conservative), in all orally-argued cases, including split votes, from 1985 to
2001 (Segal et al. 1995; Spaeth 2003).23
The control variables Liberal and Conservative Amicus are the number of ideological amicus curiae briefs filed in either direction (Kearney and Merrill 2000;
Collins 2008). Lower Court Direction is the direction of the lower court decision under review (1 for liberal, 0 for conservative). Solicitor Liberal and Conservative are dummy variables coded 1 if the Solicitor General submitted a brief advocating that ideological
23 These cases come from the United States Supreme Court Data Base Justice-Centered Vote dataset. I include decisions that were orally argued, a judgment was issued, or were per curiam. 55 position (Deen, Ignagni and Meernik 2003). Liberal Resources and Conservative
Resources capture the identities of the litigants (Collins 2008). Like Collins, in the both sets of models I also interact Justice Ideology with Liberal and Conservative Amicus.
In the first analysis, I measure Justice Ideology with Segal-Cover scores (Segal and
Cover 1989) for the purpose of matching Peppers and Zorn’s approach. In the second analysis, I measure Justice Ideology using JCS scores, lagged by one term to reduce circularity.24 JCS scores are transformed from the dynamic Martin-Quinn scores to a -1
(liberal) to l (conservative) scale (Giles, Hettinger and Peppers 2001; Martin and Quinn
2002; Epstein et al. 2007). While these scores carry an incumbent risk of circularity, they have several advantages. The JCS already plays an important role in the clerk ideology and bloc measurement schemes in the second analysis. JCS scores and, say, Segal-Cover scores are not easily comparable, which is necessary for construction of Proportion,
Distance, and Direction. Furthermore, because Segal-Cover scores are static using them can over- or underestimate the sway clerks actually have. By controlling for a justice’s recent ideological voting using lagged JCS scores I do not run the risk of attributing observed vote change to being pulled by clerks when a likely cause is a justice ideology score that has drifted out of date. Consider Justices Souter and Kennedy in OT 2000.
Souter’s Segal-Cover score is .325 while Kennedy’s is a slightly more liberal .365. Their lagged JCS scores, respectively, are -.414 and .179, suggesting that Souter is now much more liberal than Kennedy than at their respective confirmations. In 2000, Souter’s clerks’ scores were uniformly more conservative than his JCS score, suggesting that he
24 Thanks to a commentator for pointing out that this coding eliminates from the model any justice's freshman term that falls within my time frame. Because my theory neither predicts nor precludes any variation in justice behavior based on length of tenure, I opt to address the circularity issue here and leave analysis of a potential freshman effect to a later day. 56 could be pulled in a conservative direction. The Distance of the bloc, and therefore the magnitude of the observed pull, would be much smaller if Souter were measured as a moderate-conservative, as he would given his Segal-Cover score, than as a strong liberal, as he would given his lagged JCS score.
Finally, in the second analysis, in addition to splitting the date along chambers type—crosswinds, liberal foxhole, and conservative foxhole—I also split the data on
Justice Ideology, and test only a continuous, moderate range of values. There are two reasons for doing this. First, as a justice’s JCS score approaches either pole, it will be difficult for him to hire clerks that are more ideologically extreme than he is. As a result, it makes less sense to assume that a justice’s structuring decisions were purposeful, because clerks of particular ideology are scarce. This is evident in the data in chambers for Justices Marshall, Brennan, and Rehnquist, whose JCS scores were higher than all or nearly all active circuit court judges, making it impossible for these justices to stock their chambers with clerks more extreme than themselves. The second reason is that the theoretical model describes behavior likely to be exhibited by moderate justices. These justices are less attached to strong ideological views and may be more likely to seek out additional information, such as that which their clerks can provide.25
Results
I begin the analyses with a replication of Peppers and Zorn’s model of a justice’s votes on the merits, as presented in their Figure 4 (2008, 74). This replication serves two purposes. First, it functions as a construct validation exercise (Adcock and Collier 2001,
25 Enns and Wohlfarth (2013) find that the case-specific median justice -- who is usually, but not always, a moderate justice -- exhibits different behavior than the other justices in deciding how to vote, focusing less on other justices' preferences and more on public opinion, the Solicitor General, and separation-of-powers concerns. A natural question raised by this analysis is if the case-specific median justice is also more likely to be pulled by his clerks. 57
542-543). Because, as I discuss below, the results with the JCS measure of clerk preferences hew closely to those with Peppers and Zorn’s survey measure, we can infer that the JCS score is measuring the same latent variable: clerk ideology. Second, the replication leads directly to an analysis that is designed to uncover the mechanism of influence. These analyses are closely related. The first analysis makes minimal assumptions about how precisely clerk preferences have been measured, and after uncovering influence with this conservative test I conduct a more nuanced analysis that is designed to uncover how structure modifies the transmission of information which leads to influence on merits-votes.
Replication of Peppers and Zorn (2008)
To conduct the replication, I construct an independent variable Clerk Liberalism to measure the same latent variable as Peppers and Zorn’s main variable. They measure chambers liberalism by aggregating survey responses and estimations of clerk partisanship. Alternatively, I code a clerk as having a liberal ideology if her proxy JCS score is ≤ 0 and a conservative otherwise and construct a chambers-level variable that describes the proportion of liberal clerks.
58
Table 3.1 presents the results of a logistic regression of a justice’s decision to cast a liberal vote on Clerk Liberalism, and includes control variables from Collins’s model of informational influences (2008, Chapter 4). I note that Justice Ideology is measured with
Segal-Cover scores in order to match Peppers and Zorn’s analysis. The results show that each variable, save Collins’ interactions, is statistically significant and correctly signed.
Figure 3.1 plots the estimates obtained from the model in Table 3.1. To compare the relative effect of Clerk Liberalism, I also plot the effect of Justice Ideology on an identical y-scale.
59
60
The results (for 1985-2001) strongly resemble Peppers and Zorn’s (for 1953-
2004). They find that “a change across the full range of Clerk Partisanship from an all-
Republican cadre of clerks to one composed entirely of Democrats increases the predicted probability of a liberal vote by 0.13 (from 0.39 to 0.52)” (74-75). My statistical model predicts a similar but slightly stronger effect: moving from a chambers of all conservative clerks to all liberals increases the probability of a liberal vote by 0.18 (from
0.37 to 0.55). The difference in effect sizes—0.05—is quite small, and could be due to the different time frames or to subtle differences in the measures.
As do Peppers and Zorn, I interpret the results of this statistical model as evidence that clerk ideology influences the justices’ votes on the merits. The more liberals a justice employs, the more likely he is to cast a liberal vote even after controlling for a host of other factors. Both sets of results—Peppers and Zorn’s and mine—also point toward structure being an important factor in clerk influence: the level of variance among clerks’ ideologies affects their ability to pull a justice’s vote. This is apparent even with a conservative interpretation strategy that does not try to ascertain the effect size: the predicted probability when Clerk Liberalism equals 0 is statistically distinguishable from when it equals 1.
There are limitations to what we can interpret using this research design. As
Peppers and Zorn admit, while their findings are consistent with the phenomenon of justices’ votes being influenced by information gleaned from their clerks, their “findings offer no support for any particular causal model of that influence” (75). Similarly, while the results in Table 3.1 and Figure 3.1 are consistent with the informational account of clerk influence that I discussed above, the mechanism remains opaque. I now turn to an
61 analysis of the different types of chambers a justice assembles—foxholes and crosswinds—that tests a causal mechanism in which chambers structure modifies the effect of the information that clerks transmit to their justices.
Testing an Informational Mechanism of Clerk Influence
I expect that the flows of information from clerks to justice differ according to chambers structure.26 Foxhole chambers are characterized by a high level of agreement among the clerks. Relative to their justice, the clerks in these chambers are all either more liberal or conservative. This homogeneity means that clerks can craft an informational signal with which they all agree, because even a slight shift in the clerks’ direction will be preferable to the justice’s ideal point. The signal reaches the justice unopposed and should be likely to pull the justice. Alternatively, crosswinds chambers are characterized by a greater variety of viewpoints. Clerks to both sides of the justice will craft signals designed to pull the justice in their direction. Knowing nothing about which bloc of clerks crafts the higher quality signal in any given case, we cannot predict which is more likely to pull; sometimes the liberals will win out, sometimes the conservatives. As a result, I predict overall pull in crosswinds chambers will come out as a wash, resulting in no statistically distinguishable effect.27
As described above, the variable Distance measure the content of a bloc’s signal. In foxhole chambers, it should have a significant effect: as the signal gets stronger, it
26 A crosswinds chambers is denoted as one in which the Proportion of the left and right blocs are greater than 0. A foxhole chambers is denoted as one in which the Proportion of a bloc is 1, i.e. a liberal foxhole occurs when the Proportion_Left Bloc = 1. Clerks with identical ideology scores as their justice and clerks for whom a score is unavailable are in neither bloc. One justice-term is excluded from the analyses because a clerk had an identical score as her justice. 27 Spitzer and Talley (2013) model the role of revealed information in moderating ideological decision making. While their model focuses on three-judge panels, the dynamic is roughly similar to a justice's chambers in which the merits of a variety of perspectives are debated (648-649; 653). 62 becomes more likely to pull. Because the dependent variable is whether a justice casts a liberal vote, in liberal foxholes the effect should be positive and in conservative foxholes it should be negative. In crosswinds chambers, neither Distance nor Proportion, which drops out of the foxholes analyses, should have a significant effect, as a result of clerks’ competition.
Table 3.2 reports the results of three models, each of which tests data from 1985-
2001 for justices falling within a moderate range of ideology and in the three different
63 types of chambers: crosswinds and liberal and conservative foxhole.28 The performance of the control variables provides confidence in the overall model. Consistent with social- psychological models of decision making, Justice Ideology is significant and correctly signed in all three models. Among statistically significant effects, the rest of Collins’ variables generally perform as expected.
The results support my expectations. In crosswinds chambers, neither bloc’s
Proportion or Distance has a statistically significant effect on vote direction. This is as predicted, and we can interpret from this finding that having clerks with diverse viewpoints results in no predictable pull on the justice’s vote.
In liberal foxhole chambers, right bloc Proportion and Distance drop out of the model because, by stipulation, each has a value of 0. Left bloc Proportion also drops out because in this sample there is no variation; it is always equal to 1. Left bloc Distance is signed correctly but is not statistically significant. We therefore cannot conclude from these results that liberals have a discernible influence on their justice’s merits-votes.
It is a noticeably different story in conservative foxhole chambers. Left bloc
Proportion and Distance are both equal to 0 and drop out. Right bloc Proportion also drops out because it is always equal to 1. Right bloc Distance is correctly signed and statistically significant. This is evidence that the farther the average Distance between the justice’s ideology and his clerks, the more his vote is pulled rightward.
28 Recall that I split the data along chambers type because the predictions are conditional on chambers type and interpreting a three-way interaction with at least two continuous variables— Distance and Justice Ideology—makes interpretation challenging. 64
I plot the conservative foxhole model’s estimates in Figure 3.2 to interpret the magnitude of Distance’s pulling effect. Figure 3.2 shows that the effect is considerable.
Over the range of Distance a justice’s vote is pulled rightward by a probability of .09 to
.13 (from .57 to .44 for a liberal-leaning justice and .28 to .19 for a conservative-leaning justice). In other words, a justice with clerks who are at the maximum distance is 23-32% less likely to cast a liberal vote.
The predicted probabilities reveal a statistically and substantively significant negative relationship between the values of the bloc’s characteristics and the probability that a justice who has structured a conservative foxhole casts a liberal vote. This is strong but asymmetric support for the theory, because there is no significant effect for liberal clerks. Interestingly, this asymmetry seems to support Lazarus’s (1999) description in
65
Closed Chambers of the conservative “Cabal” being more successful than the “Dreaded
Libs” at manipulating the justices.
Discussion
My findings show that structure matters in the justices’ interactions with their clerks. Whether the justices desire the sort of influence I have found remains an open question, but there is reason to expect they do. They have great discretion about whom they hire, which should lead scholars to view a justice’s structural decisions as a purposeful assembling of a team designed to pursue his individual goals. It is noteworthy that ideology does not appear to be the sole factor in the decision to assemble a unified chambers: nine different justices in the analysis assembled at least one liberal or conservative foxhole chambers, with one justice—O’Connor—assembling one of each.
Further research should seek to uncover all the factors that guide a justice’s structural decision; in the meanwhile, the insufficiency of an obvious answer like ideology should allay concerns about sample bias or whether clerk ideology is endogenous.
Because my approach to studying clerk influence is somewhat unprecedented, my findings may be met with skepticism. I rely on a novel measure of clerk preferences to study a role played by the clerks that has been underappreciated in the literature. Skeptics should take note that the measure meets construct and convergent empirical validation, and is supported by theory and the literature. Finally, the theory of decision making I present is not new; only clerks’ role in it is. The literature has shown that the justices seek out information from attorneys, the solicitor general, and amicus curiae briefs, among other sources. I contribute to this conversation that clerks represent an additional source.
To the extent that a veil of secrecy hangs over the actions of the Supreme Court, we
66 are limited in how much we can know about what goes on inside the justices’ chambers.
To some, it is deeply concerning that young, unconfirmed lawyers play a part in deciding cases. However, when “young clerks keep judges abreast of the thinking taking place in the outside world,” there should be little doubt that when “a clerk persuades a judge to a view, so that the judge makes the view her own, this is proper functioning of the judge/clerk relationship rather than arrogation of judicial power by clerks” (Velvel 1995,
5). Critics who are concerned about clerks’ usurpation of Article III may be premature: an advisory form of clerk influence actually may be welcomed by the justices. The justices’ chambers are the workplace of a corps of law clerks that grows increasingly diverse – perhaps by ideology, but certainly in terms of gender, race, and educational background – and the effects clerks have can be observed if we carefully discern implications of institutional structures. Only once we know what influence, if any, clerks have, should we draw normative conclusions.
Appendix
List of Supreme Court Law Clerks with Common Space Scores
The following is a list of former Supreme Court law clerks with federal courts of appeals clerkship experience and who have held a position for which a Common Space score was calculated. Scores for Justice Elena Kagan and Senator Ted Cruz were not yet available.
67
68
Chapter Four: Female Law Clerks and Substantive Representation at the U.S. Supreme Court
Introduction
An important normative puzzle in American government is the extent to which the federal judiciary should be representative of the people. On the one hand, the courts should be insulated from popular pressures, so that “the independence of the judges may be an essential safeguard against the effects of occasional ill humors in the society”
(Federalist #78). In other words, there are benefits when courts are empowered to act in a countermajoritarian manner. On the other hand, the courts, especially the Supreme Court, are composed of a handful of people who nonetheless wield impressive policy-making capacity. Who is to say when, if at all, this older, wealthier, and descriptively unrepresentative few should be permitted to substitute their judgment for that of the people and its elected representatives?
This dilemma has led one pair of scholars to endorse courts that, in reference to the Goldlilocks metaphor, are “just right.” Courts should “have just enough judicial independence to check and balance the executive and legislative branches of government under most circumstances, but…lack the sky-high levels of judicial independence that enable them to make long strings of unpopular, countermajoritarian decisions without experiencing any negative consequences of those decisions,” namely, consequences such as attempts to curtail their power and autonomy (Gibson and Nelson 2015, 173). This struggle to balance accountability and representativeness with independence has taken place over the course of American history. In a notable trend, for instance, it led many states to incorporate direct elections into the selection of their judges.
69 A fundamental question involved in this debate is whether courts are ever actually representative of the people. It may be that courts never reflect the will of the people.
This is an empirical, not a normative, question, and in answering it scholars consider two types of representation, substantive and descriptive.1
The degree to which the Supreme Court takes into account the substance of public opinion and reflects it in its decision making—its substantive representation—continues to elude scholars. While most recent scholarship points to a relationship between public opinion and the Court’s outputs (Mishler and Sheehan 1993; Norpoth and Segal 1994;
Flemming, Bohte and Wood 1997; McGuire and Stimson 2004; Durr, Martin, and
Wolbrecht 2000; Epstein and Martin 2010; Giles, Blackstone and Vining 2008), debate persists over the causal mechanism(s) driving the relationship (Casillas, Enns, and
Wohlfarth 2011; Hall 2014; Bryan and Kromphardt n.d.) and how the two factors reinforce each other (Ura 2014).
There is no question that the Supreme Court fails as a representative of the
American people in a second way. Because the Supreme Court most certainly does not, so to speak, “look like” like the American people, it cannot be said to be descriptively representative. To sample a few characteristics, the Court is currently composed of five millionaires (Cline 2011), six Catholics and three Jews, past or present denizens of four of the five boroughs of New York City, and eight graduates of either Harvard and Yale
Law School, but only three women, one African-American, and one Latina.
1 According to Farhang and Wawro (2004, 301), “A political body or institution is descriptively representative by literally resembling or reflecting the constituent elements of the community that it governs. In contrast, substantive representation is concerned with what the representative actually does on behalf of the interests of the group he or she is associated with.”
70 A better case for its descriptive representation emerges when we factor in those who work most closely with the justices: their law clerks. Once charged with being
“lacking in diversity” (Mauro 1998), the corps of clerks has come to reflect the image of the American people to a greater degree, at least in terms of sex.2 Moreover, recent theoretical and empirical work suggests that enhanced descriptive representation among the clerks could affect the Court’s substantive representation—for instance, more female clerks could lead to greater support for pro-women policies in the justices’ votes. This paper probes the link between law clerk characteristics and judicial decision making by analyzing the role female law clerks play in influencing their justices’ votes on the merits. While the normative debate over what make a court “just right” rages on, the empirical question of the Court’s descriptive and substantive representation will benefit from a close analysis of everyone with a role to play in the justices’ decision making.
The Influence of Females on Judicial Decision Making
This section serves first as a preview to my argument and then as a review of the relevant literature. Law clerk characteristics, such as ideology and sex, are relevant to the decision making of the justices for whom they work when those characteristics mean the clerks posses an information advantage or asymmetry. Asymmetries can arise from having had unique experiences, such as perceiving the world through a distinct frame or being exposed to differential treatment by other people. For instance, a 27-year-old female law clerk is likely to have had experiences that a 65-year-old male justice has not.
To the extent that these experiences—for example, being a victim of discrimination— would inform how someone reacts to similar situations, a person’s background is relevant
2See Figures 4.1 and 4.2. Beginning in the mid- to late-Nineties, and roughly coinciding with Mauro’s article, the data reveal patterns of greater concentrations of female clerks in several chambers, e.g. Breyer, Ginsburg, O’Connor, Souter, Stevens, and Thomas.
71 to her decision making. As I will discuss in greater detail below, the impact likely goes further, affecting the decision making of others. Because several justices regularly engage their clerks as advisors (see Chapter 3, 35-38), asymmetric information can and likely will be transmitted from clerk to justice. When this happens the justice learns more about the clerk’s position and may decide to endorse it. This learning, because it is a direct result of clerk intervention, should be considered influence.
What do judges seek to learn and how do they go about doing so? Frequently, the law points toward several possible outcomes, and a justice may be uncertain what impact different votes will have. Under such circumstances it may not be immediately obvious to a justice whether a disposition or rule is optimal. In order to avoid making what might later turn out to be a mistake—for instance, creating a precedent that leads to unwanted consequences, according to his goals—a justice consumes information that reduces uncertainty about the strategies available to him. Information, especially from sources that the justice deems credible, can help him identify his optimal outcome and avoid making mistakes.
The literature documents the influence numerous sources of information have on judicial decision making. Among these sources are amicus curiae briefs (Collins 2008), the solicitor general (Bailey, Kamoie, and Maltzman 2005; Black and Owens 2012), law clerks (Kromphardt n.d., 2014), and attorneys (Johnson, Wahlbeck, and Spriggs 2006;
McAtee and McGuire 2007; Corley 2008; Szmer and Ginn 2014).
An information source will be valued when it is perceived to be especially credible (Johnson, Wahlbeck, and Spriggs 2006). For instance, Szmer, Sarver, and
Kaheny (2010) find that Supreme Court justices are more likely to side with female
72 attorneys (when they are facing opposing male attorneys) in cases that involve gendered issues, such as sex discrimination and reproductive rights, which the authors suggest shows that females enjoy greater credibility when arguing such cases.
Females may be more credible than males when litigating gendered issues, but do they make decisions differently? Some sort of difference that is attributable to females seems to be an important condition for asserting that female law clerks possess unique information. Evidence suggests what one set of scholars calls “individual effects”
(“whether and in what ways male and female judges decide cases distinctly”) do exist
(Boyd, Epstein, and Martin 2010, 389). Weinberg and Nielsen summarize findings reporting differential voting behavior between men and women in cases involving sex discrimination (Davis, Haire, and Songer 1993; Peresie 2005; Boyd, Epstein, and Martin
2010) and sexual harassment (Peresie 2005). Gruhl, Spohn, and Welch find an interaction between judge- and litigant-gender: “‘[w]omen are about twice as likely to sentence females to prison as men are’” (1981, 320, c.f. Weinberg and Nielsen 2011, 335).3
However, some scholars (see Kenney 2013) challenge that essentialist explanations of decision making, if they explain behavior at all, are unlikely to be biologically based.
That is, sex is unlikely to be a direct causal factor. An alternate explanation for observed differences in behavior could be found in widely shared experiences or socialization among women, especially in subsets of the population likely to be considered qualified for a Supreme Court clerkship. Evidence suggests that a subset of the population of women that would contain all qualified applicants—those with a college education—
3 While it is important to note that I do not argue that law clerks actually decide cases for their justices, these studies suggest that female law clerks may react to the same set of facts differently than their male colleagues and superiors. This reaction, were it to form the basis of an informational asymmetry, could affect their justice’s decision making should he learn from this asymmetry.
73 holds distinctive views on an important gendered issue. From 1980-2004, Gallup reports that between 40-50% of college-educated women expressed support for legal abortion under all circumstances, the least restrictive choice offered, as opposed to 29-38% of college-educated men (Saad 2010). This gender gap suggests that college-educated women’s attitudes toward the issue of abortion, and possibly other issues with a gendered dimension, are distinctive.4
An account by a former clerk to Byron White lends further validity to the proposition that law clerks in general and female law clerks in particular convey unique information and that such information would be perceived as credible. According to
Kathryn Bradley, “[White] never had more than one woman in chambers, but he always had one, and I think it was important to him to have someone there to give some broader perspective…I can't think of particular instances where my gender made a difference, but
I know that it did affect the way I looked at cases and the things that I brought to him and just the way that I interacted with him when we were discussing cases. So I do believe it was valuable” (PBS Newshour 1998).
The argument that female law clerks possess and convey unique information, at least in the context of some issues, will be developed in greater detail in the next section.
For now, it is important to recognize two assumptions supported by the literature:
1) Women are likely to have had different experiences than men; and
2) Women, on average, react to certain types of facts, such as those involving sexual
harassment, abortion, and other gendered issues, differently than men.
4 It is unimportant for the purposes of this paper to identify what drives this gender gap. What is important is recognizing that the subset from which all female law clerks come from—female college graduates—holds distinctive attitudes on gendered issues. To the extent that these attitudes shape how clerks perceive life experiences, they may lead to the formation of an information asymmetry.
74 These assumptions underlie the theory of information transmission and learning that I develop in the next section.
Distinct from individual effects, be it a judge’s or law clerk’s, are panel effects.
Panel effects occur when interaction with a person with a particular characteristic influences the behavior of others. In a judicial context, scholars have documented panel effects arising from a judge whose behavior is affected by the presence of colleagues with characteristics including partisanship (Sunstein et al. 2006; Cross and Tiller 1998; Kim
2008; Kastellec 2011a,b), sex (Boyd, Epstein, and Martin 2010), and race (Kastellec
2013). While this literature focuses on the impact of judges on their colleagues, other research suggests that exposure to non-judges possessing salient characteristics can affect judicial behavior. For instance, a phenomenon bearing a resemblance to panel effects occurs when decision making is influenced by those with whom one has personal relationships. Glynn and Sen (2014) show that having daughters affects the support courts of appeals judges exhibit for women’s rights in cases involving gendered issues, such as sex discrimination and reproductive rights.5,6 Together, studies on panel and daughter effects show that there is considerable potential for the characteristics of those with whom a judge interacts to affect his decision making. This may generalize to law clerks.
This literature supports an assumption that will be developed in detail in the next
5 Washington (2008) reports a similar effect for Members of Congress (see also Iacus, King, and Porro 2012, 14). 6 Importantly for generalization purposes, Glynn and Sen’s results point to a causal mechanism involving information transmission rather than preference change or protectionism (15-16), from which we might infer that other forms of interaction in which information is exchanged, such as that between a justice and his clerks, could lead to similar influence.
75 section: that the identities of those with whom a judge interacts can affect his7 decision making. Such effects could be driven by information conveyed by (Boyd, Epstein, and
Martin 2010; Glynn and Sen 2014) or the mere presence of (Sommers 2006, c.f.
Kastellec 2013 5-6) a person possessing the relevant characteristic. How presence affects behavior is an important question because it does not involve interaction with numerous persons bearing the relevant characteristic to affect behavior—just one would suffice.
Two Causal Mechanisms: Presence and Reinforcement
The goal of this paper is to analyze how female law clerks affect judicial decision making—if they affect it at all. I hypothesize two possible mechanisms, which I label as presence and reinforcement. In this section, I introduce these mechanisms and derive hypotheses for testing them empirically.
How might even just one female law clerk influence her justice’s behavior? We can find clues in how the Court adapted in a related context. The Court was long an all- male institution, but right around Sandra Day O’Connor’s confirmation in 1981 various practices changed. For example, up until shortly before O’Connor’s confirmation, the justices were referred to as “Mr. Justice.” While the justices voted to change the honorific to the gender-neutral “Justice” shortly before O’Connor was confirmed to the Court, they may have done so to avoid the awkwardness of having to do so after they had been joined by a “Sister” justice (Biskupic 2005, 101). Also, after O’Connor was confirmed, in briefs, the prefix used to refer to women shifted from “Miss” or “Mrs.” to the general “Ms.”
(Kramer 1990). While these changes may seem minor, they reveal the Court transitioning from long-held practices in order to accommodate the presence of a woman in its midst.
7 For purposes of clarity, the feminine denotation (i.e. she, her) is used in reference to clerks and the masculine denotation (i.e. he, his) is used for judges and justices. This convention was adopted randomly.
76 Experimental evidence suggests that a presence effect can be more substantial.
Kastellec (2013) describes an experiment that illustrates a presence effect in simulated juries, where the presence of a black juror causes nonblack jurors to adjust their behavior empathetically before any deliberation takes place (Sommers 2006). This suggests that a presence effect might take place passively: little to no meaningful conversation may be necessary for, say, a female clerk to influence her justice’s vote.
Alternatively, we can imagine a more active presence effect in which the female clerk interacts with her justice. Recall from the previous section that many modern justices use their clerks as advisors (see Chapter 3, 35-38). In this situation—talking with his clerk about case outcomes—the justice takes something away from the interaction, and it is whatever he takes away that causes a change. The key commodity here is information, as it will be for the reinforcement mechanism I discuss below.8
The reinforcement mechanism can be summarized as “more is better.” How might a larger number of female clerks affect behavior in a way distinct from the presence effect described above? As with the active form of the presence effect, the key commodity is information. The difference is that more females would mean that information unique to females has a greater shot at impacting a justice’s behavior. There is not more information, so to speak, but more favorable conditions for information to be influential. To state differently, an increase in the number of female clerks changes the chambers’ structure—specifically, the sex makeup—in such a way as to increase the probability that the unique information female clerks possess can have an effect on the
8 The distinction between passive and active presence effects is most likely empirically unimportant, as it leads to no observable implications of which the data permit testing. I include this discussion to suggest that information may not be the only force leading to influence.
77 justice’s decision making. This effect would take place when a justice is exposed to such information, and comes to be convinced to vote in support of women’s rights.9
Female clerks’ unique information arises from experiences and treatment from going through life as a female, and is most likely to be salient in issues involving a gendered dimension because of their credibility (Szmer, Sarver, and Kaheny 2010). A gendered issue is something like sexual harassment or discrimination, or abortion because policy debates over the issue involve reproductive rights. However, cases that turn on facts involving, say, antitrust or freedom of speech are unlikely to feature a gendered element.
The argument that chambers characteristics modify the transmission of information is intuitive. Increasing the number of clerks who share characteristics should increase the likelihood their information has an impact for three reasons. The larger a bloc of clerks who share characteristics is, the more likely a clerk from that bloc will be assigned to work on a given case—writing memos, conducting research, and generally serving as point-person with the justice—and a clerk working in this capacity is in the best position to influence a justice. Second, clerks who agree with each other can work together to reinforce a shared preference. Evidence presented in Chapter 3 shows that clerks who are uniformly conservative pull some justices’ votes to the right (58-62). The
9 For male justices, this may, but is not assumed to, be their first exposure to such information. The information is assumed to not be unique to female justices; however, a female justice can still be influenced if this information reinforces a feminist outcome about which she had uncertainty over its outcome. Imagine a scenario akin to a simple one involving voting fluidity. A justice is initially inclined to reverse a decision upholding stricter abortion regulations. However, because of uncertainty about policy outcomes of such a decision, she casts a preliminary vote to affirm. After being availed upon by her clerks, who provide information based on their own experiences or those of others in procuring an abortion, her uncertainty over the outcome decreases, reinforcing her initial confidence in voting to reverse; her final vote is to reverse. Maltzman and Wahlbeck report that during the Burger Court, at least one justice changed his vote in 36.6% of cases and individual justices changed their vote 7.5% of the time (1996, 587).
78 results indicate they are able to do so because they send a consistent signal advocating for a particular outcome. Similarly, more female clerks should be able to send a stronger signal reinforcing an outcome supporting women’s rights, and the stronger the signal, the more likely it is to cause the justice to adopt its position. Finally, increasing the number of female clerks necessarily reduces the chance that a female will be treated as a token representative. If he has more than one female clerk, then the hires were probably made in good faith.
These observations suggest a complex dynamic between the sex of a justice’s clerks, the structure or makeup of his chambers, and the outcome of his decision making process. One dynamic indicates that the presence of even a single female may cause a justice’s behavior to change, although it is uncertain as to in what direction. A second mechanism, one involving the number of females, argues that increases in the number of females are more likely to produce a positive outcome for women’s rights. The statistical analyses I present below are an attempt to begin “untangling the causal effect of (law clerk) sex on judging” (see Boyd, Epstein, and Martin 2010).
Data and Method
I analyze justice-votes in all decisions, including split-votes, from 1985-2001
(Segal et al. 1995; Spaeth 2003). I pool all votes, cluster standard errors at the justice level, and control for whether a case involves issues with a gendered dimension. Issues considered to involve a significant gendered dimension are abortion and sex discrimination both pertaining to and not pertaining to employment, as coded by the
Supreme Court Database. This division of issues is supported by the literature (see Glynn and Sen 2014) and the theory of information acquisition and utilization I outline in the
79 previous section. These are cases in which a female is likely to possess asymmetric information from males, due to a greater likelihood of personal experience or that of someone close to her, and in which she would be considered a credible source.
Table 4.1 presents descriptive statistics for the dependent and independent variables. The dependent variable Liberal Vote is coded 1 if the justice casts a liberal vote, which, for example, in gendered issues involves siding with the plaintiff in the sex discrimination cases and being opposed to greater restrictions on abortion, and 0 otherwise. I use logistic regression to determine whether the primary independent variables—the interaction between gendered issues and the presence of a female law clerk and between issues and the number of a justice’s clerks who are female—exert an effect on the decision to support the feminist position after controlling for competing explanations.
The first independent variable of interest is One Female x Gendered Issues, which is coded 1 when one of a justice’s clerks is female and an issue is gendered and 0
80 otherwise. To test the reinforcement mechanism, I include in the model Two+ Females x
Gendered Issues.10 I operationalize the concept as dummy variables because the theory predicts that the effect of the number of female clerks may be nonlinear. I have no a priori expectation whether the effect of one female will be positive or negative, but I expect the effect of two or more females to be positive.
Figures 4.1 and 4.2 plot the number of female clerks for every justice to serve between 1969-2007 and to employ at least one female clerk over that period.11 12 13 To interpret these scatterplots, I turn to two significant dates: 1982, the year after Sandra Day
O’Connor became the first female justice, and the late Nineties, when the justices came under harsh scrutiny for the diversity of their chambers (Mauro 1998, Ruger 2012). We might expect to see these dates serving as transition points for the justices’ practices in hiring female clerks. After O’Connor joined the Court, Justices Blackmun, Brennan,
Burger, and Powell evince a greater willingness (relative to their prior behavior) to hire females.14 The data for the mid- to late-Nineties reveal patterns of greater concentrations of females in the chambers of Breyer, Ginsburg, O’Connor, Souter, Stevens, and Thomas.
Kennedy and Rehnquist’s patterns remain consistent, while Scalia hired fewer female clerks. While these figures simply describe patterns, we can conclude from them that the tendency of several justices to assemble greater concentrations of female clerks
10 The reference category No Female Clerks and its interaction with Gendered Issues are omitted from the statistical models. 11 I conduct this same sort of descriptive analysis in Chapter 2, where I plot the chambers structure by clerk ideology for each justice from 1969-2007 (21-22). 12 Hugo Black and John Harlan II are not shown because neither had a female clerk during the range of years. 13 See Figure 4.4 in the Appendix for the overall number of female law clerks at the Supreme Court. 14 Writing in 1984, Cook reports much lower concentrations (1984, 593). The increases that took place during and after her period of analysis lend credence to the claim that O’Connor’s confirmation may have stimulated greater hiring in her colleagues.
81 coincided with and was perhaps even a response to media scrutiny.15 This variation in the number of female clerks working in the justices’ chambers is an interesting finding, particularly in the context of testing of the reinforcement mechanism.
15 An alternative explanation for these hiring patterns could be the presence of more highly qualified females in the pool, especially working for feeder judges (Baum and Ditslear 2010).
82
83
84 In order to account for competing explanations of the direction of a justice’s vote on the merits in gendered cases, I control for the variables in Collins’ (2008) model of informational influences on a justice’s vote on the merits. Briefly, Liberal and
Conservative Amicus are the number of ideological amicus curiae briefs filed in either direction (Kearney and Merrill 2000; Collins 2008). Lower Court Direction is the direction of the lower court decision under review (1 for liberal, 0 for conservative).
Solicitor Liberal and Conservative are dummy variables coded 1 if the Solicitor General submitted a brief advocating that ideological position (Deen, Ignagni and Meernik 2003).
Liberal Resources and Conservative Resources capture the identities of the litigants
(Collins 2008). Justice Ideology is coded with a justice’s Judicial Common Space score
(Epstein et al. 2007), lagged one term to reduce circularity.16 As does Collins, I interact
Justice Ideology with Liberal and Conservative Amicus. Finally, I include a dummy variable for whether a justice is female.17
Results
I use logistic regression to determine whether the data support the hypotheses that the presence of at least one female clerk and an increase in the number of female clerks influence the direction of their justice’s votes on the merits in cases with a gendered dimension.
16 Because Segal-Cover scores are static (Segal and Cover 1989), using them can over- or underestimate the sway clerks actually have. By controlling for a justice’s recent ideological voting using lagged JCS scores I do not run the risk of attributing observed vote change to being pulled by clerks. See further discussion of this point in Chapter 3 (52-53). Note that the statistical results hold when Segal-Cover scores are substituted for lagged JCS scores in the models. 17 The theory does not preclude that female justices can be influenced by their female law clerks. However, having only two female justices in the data makes it difficult to draw inferences about systematic differences based on justice sex.
85
Table 4.2 presents the results. The control variables perform basically the same way as in Collins’ original model. Justice Ideology is significant and correctly signed, which is consistent with the literature on the importance of the justices’ attitudes (see
Segal and Spaeth 2003). These findings are reassuring for the performance of the basic model.
86 How do my variables of interest perform? Recall that, because the expected relationship between increasing the number of female clerks and liberal voting may be nonlinear, I use dummy variables to describe chambers structure by clerk sex. Turning to the constitutive elements of the interactions, One Female and Two+ Females are signed negative, Gendered Issue is positive, and only One Female is statistically significant.
This means that the effect of having only one female clerk—the presence effect—is a backlash for non-gendered issues. Because all but 301 of the votes in the data are non- gendered, this is a surprising finding, but probably an unimportant one.
Turning now to the interactions, I find that One Female x Gendered is negative but insignificant and Two+ Females x Gendered is positive and significant. This indicates that there is no distinguishable difference between having no female clerks and having one (although when I plot the predicted probabilities, we will see that the probability of casting a liberal vote decreases slightly). However, there is a distinguishable positive effect for having two or more female clerks. To understand the apparently complex relationship between having zero, one, and two or more female clerks, I plot the predicted probabilities derived from the model’s estimates in Figure 4.3.
87
88 The left pane of Figure 4.3 depicts a nuanced dynamic between the number of female clerks he has and a justice’s voting support for feminist policies. When a justice has no female clerks, the left pane shows, his vote shows support for feminism, with a probability of .57. When he adds one female, the probability of support actually decreases, to .53. The downward movement reverses itself as the justice adds more female clerks, however, and the probability of feminist support shoots to .77. This is an increase of support by 35% of the original level of support, when there are no female clerks present.
What the data show us is that the concentration of female clerks is important. It is not enough to hire one female clerk and see a feminist effect; in fact, hiring just one actually decreases the probability of casting a liberal vote. Many justices may be willing to hire a female clerk, perhaps as a token, and this reduces their likelihood of voting liberally in cases involving gendered issues. However, the more female clerks a justice hires, the more likely he is to vote liberally in the same cases.
As a robustness check on this result, I conduct an additional analysis into whether the relationship between a tendency to hire female law clerks and to support feminist positions is spurious. It could be that a lurking variable—one we might call feminism or gender consciousness—influences both of these decisions. If this were to be true, we could not have faith in the findings I have reported so far.
It is often quite challenging to account for potential spuriousness. Without an instrumental variable, we might not be able to rule out whether the errors are correlated with each other. Fortunately, the scatterplots in Figures 4.1 and 4.2 undermine the notion that another covariate, justice ideology, explains both the independent and dependent
89 variables. The variety of chambers reporting two or more female clerks depicted in
Figures 4.1 and 4.2 should allay concerns about an obvious explanation like ideology— just look at O’Connor’s and Thomas’s hiring practices. To account for another proposed endogenous variable of feminism, I analyze only the data of a subset of “feminist” justices to see if having two or more female clerks still has a positive influence on support for feminist policies. The research design allows us to consider the effect of multiple female clerks on support for feminist policies for justices with a baseline of feminism.
Obviously, measuring something like a judge’s feminism is going to be a tough task. Beverly Blair Cook’s work is noteworthy for two such efforts, one asking survey questions and policy positions of judges (Cook 1981) and one, as Kenney (2013, 33) describes it, “plott[ing] each of the justices based upon their votes on women’s rights cases and the views they expressed in their opinions, other writings, and speeches” (Cook
1978). My own approach is based on the assumption that a justice’s repeated willingness to hire more than a token representation of female laws clerks evinces feminism. The justices have broad discretion over whom to hire, so those who repeatedly choose female clerks most likely desire sex representation and believe female and male are equally capable. The rule I use to operationalize feminism is whether a justice hired at least two female clerks in the same year for at least two years.18 This measure is rough, but it enables me to test the important question of whether the effect of two or more female clerks on support for feminist policies is a function of the justice being feminist.
18 These data are depicted in Figures 4.1 and 4.2. Justices who qualify as feminists by this rule and who casted votes between 1985-2001 are Blackmun, Breyer, Ginsburg, Kennedy, Marshall, O’Connor, Souter, Stevens, and Thomas. While some names on this list may be surprising, it is important to note that ideology and gender consciousness are not always correlated.
90 The results in the second column of Table 4.2 indicate that it is not. Considering only those justices who repeatedly hire more than one female law clerk, the effect of having two or more female law clerks on support for feminist policies remains positive and statistically significant. To put it differently, this analysis shows that it is not a justice’s own feminism that drives the results: variation across terms in the number of female clerks hired by the same feminist justice affects behavior.
The right pane of Figure 4.3 plots the predicted probabilities derived from the model’s estimates. The model predicts support for feminist policies occurring with a probability of .56 when a feminist justice has no female clerks. This level of support is nearly identical to that in the sample of all justices. However, unlike the sample of all justices, the effect among feminists of having one female clerk is positive, although insignificant. From predicting feminist support with probability .58 when a feminist justice employs one female, the model estimates support at .76 when a feminist justice has two or more female clerks. The primary difference between these two models is the dogleg in the estimates for the sample of all justices. The negative effect of one female clerk, while not significant, becomes positive when we only consider feminist justices. In sum, the results point to a positive reinforcement effect for feminist justices.
Discussion
Judicial decision making is a primary subject of interest to scholars of judicial politics. Uncovering systematic factors that influence voting behavior that are supported by theory provides a glimpse into the black box that is the decision making process. Due to practices of extreme secrecy, scholars face challenges in identifying any systematic influence arising from law clerks. However, the public availability of data about law
91 clerk characteristics, such as sex, allows us to use relevant theory to form predictions about how this characteristic plays a role.
The foregoing analysis may surprise some readers, as it indicates that not only do law clerks wield influence, they do so in support of feminist outcomes. While surprising, it should also be reassuring for observers who worry about the Court’s protection of minority rights. It would seem that an increasing diversification of the corps of law clerks benefit not just for the Court’s descriptive representation but also its substantive representation.
With that said, those who would be pleased (or concerned) by the effect of female law clerks on inducing feminist support should be cautious. To my knowledge, these results are the first of their kind. While the results are consistent with those on panel and daughter effects on the courts of appeals and on the influence of clerk ideology on
Supreme Court votes on the merits, this is a new area of study and thus one that will benefit from further scholarship. Future avenues of research might explore whether there exist any systematic explanations, aside from justice ideology and feminism, which are tested here, for the number of female clerks a justice hires, and if so whether such explanations are endogenous with decision making. Work might also be done looking for gendered dimensions in cases that are less obvious than sex discrimination and abortion.
Gendered issues may be partially subsumed by other issues; for example, United States v.
Morrison was a case decided along federalism lines, but concerned the Violence Against
Women Act. Finally, other law clerk characteristics that are immutable, such as race, could be tested using the informational theory I articulate.
Appendix
92
93 Chapter Five: Law Clerks and Information at the Supreme Court
What We Learned about How Principals Use and Acquire Information
The dynamic between a justice and his clerks is an example of a hierarchical relationship. By considering clerks as an illustrative case of advisors providing information to their principal, I discuss what the findings may tell us about the broader class of principal-agent relationships. Additionally, I discuss fruitful avenues of research, derived from the literature on the development of expertise in a bureaucracy, for shining new light into the black box of judicial decision making.
Clerks as Agents and Advisors
Scholars have advanced our knowledge of law clerks by treating theirs as a classic example of a principal-agent relationship. For example, scholars have demonstrated the utility of focusing on monitoring and selecting agents that share the principal’s preferences. But we can go farther. By considering how the case of law clerks illustrates a distinct type of principal-agent relationship, I believe the law clerk literature can grow in explanatory power. We can both illuminate and draw insights from a broader category of political phenomena than judicial decision making. Once we recognize that clerks function as advisors who acquire information and convey it for use by a principal, we are no longer just explaining a single facet of decision making; rather, we are now also explaining a representative case of hierarchical relationships.
One type of principal-agent relationship takes place when advisors are tasked with acquiring information for use by their boss. Scholars focus on these two stages: how information is acquired and how it is used. Because it is costly for the principal to gather
94 his1 own information, he looks to his advisors for expertise that he can draw upon when adopting policy. This dynamic parallels when a justice assigns his clerks to become acclimated with precedent and the litigant and amicus briefs so that he can consult them when deciding what policy to support in a case.
Importantly, information gathering can benefit from the advisor having different preferences than the principal. For example, Calvert shows that advisors with known biases can shape what the principal learns. When an advisor you know is biased against a policy surprises you by endorsing it, this indicates the likely utility of that policy.
Another benefit is that advisors with divergent preferences can be incentivized to acquire better information. Stephenson (2011) and Gailmard and Patty (2013) show that the preferences of advisors are endogenous to the information they acquire. When advisors are less worried about placating their boss because they do not share his preferences, they are less risk-adverse when seeking expertise. The net result is that higher quality information is acquired and available for use in policy decisions.
Admittedly, the dissertation does not tell us much about how information is acquired. To conduct the research, I assumed clerks already have information appropriate to the typical person with their ideology or sex. However, we could relax this assumption by focusing on what information a clerk is incentivized to acquire. As with the work by
Stephenson and Gailmard and Patty of other advisors and principals, what information is acquired may be endogenous to the interaction between clerks’ and justice’s policy preferences. This research would benefit from more and better knowledge of clerks’ preferences. One extension of the work begun in this dissertation that I plan to conduct is
1 For purposes of clarity, the feminine denotation (i.e. she, her) is used in reference to clerks and the masculine denotation (i.e. he, his) is used for judges and justices. This convention was adopted randomly.
95 the development of a measure of clerk preferences that draws on two sources of information: their lower-court JCS score and their response to a survey inquiring about their partisanship. Because the same latent variable underlies both sources of information—ideology—Todd Peppers and I are working on a paper that bridges the two sources into a single score. Through research into clerk policy preferences we would gain a better understanding of how clerks serve as a conduit for external information sources as they process the information for their justices’ consumption.
Acquisition is an area ripe for analysis, and scholarship can build on my and others’ work on clerk preferences. The research in this dissertation is more immediately instructive about a second aspect of advisor influence: how information is used. The results in Chapters Three and Four indicate that a justice needs to hear from multiple clerks espousing the same outcome in order to be more likely to support that policy himself. In other words, information is more likely to be used when it is reinforced by multiple sources.
This makes sense from a credibility perspective. In order to observe influence, in which a justice acts differently than he would absent clerk influence, it is necessary for clerks to call for an outcome a justice is not already guaranteed to make. This creates tension: a clerk is most likely to call for different outcomes when her preferences diverge from the justice’s, but this difference also makes the clerk less trustworthy. Homophily tells us a justice is going to be more likely to trust a clerk he knows shares his preferences, but he is also less likely to be influenced in any observable way by the clerk who is just like him. One solution to this credibility problem is for a principal to grant greater weight to a policy recommendation when multiple voices call for it. The results in
96 Chapters Three and Four lend support to this interpretation in the context of justices and clerks.
In general, a principal has to weigh the potential value of information with the credibility of the source. Besides having others call for the same outcome as you, what are other markers of credibility? This question, which is endemic to all advisor dynamics, is one that I believe the case of law clerks is uniquely situated to answer.
We know a lot more about the average law clerk than we do the bureaucrats or aides whose behavior advisor models currently describe. Clerks already have a long paper trail by the time they arrive at the Court, and they typically go on to high-profile positions in government, academia, or the private sector. By collecting background information on clerks, we can infer their policy preferences, satisfying the first step of my approach to identifying influence. Once we have an idea about what recommendations a clerk makes, we can identify how those preferences shape the choices a justice makes.
This information will also enable scholars to test hypotheses about when clerks are considered more credible and if this affects whether a justice follows their recommendations. Chapter Four shows that female clerks are influential in gendered cases, which is probably because they are considered credible sources on these issues. If they can be measured, it makes sense that ideological proximity, trust, race, expertise and other characteristics would be markers of a clerk’s credibility. By studying the interaction between clerk characteristics and case facts, we can identify what contextual features determine credibility and how this influences the justices’ use of such information.
Research shows that the justices are more likely to respond favorably to information from credible attorneys (Johnson, Wahlbeck, and Spriggs 2006; Szmer, Sarver, and Kaheny
97 2010). We should try to understand this dynamic between credibility and how justices use information with another source: their law clerks.
98 References
Adams, Edward S. 2000. “A Market-Based Solution to the Judicial Clerkship Selection Process.” Maryland Law Review 59: 129-182.
Adcock, Robert and David Collier. 2001. “Measurement Validity: A Shared Standard for Qualitative and Quantitative Research.” American Political Science Review 95 (3): 529-546.
Aldisert, Ruggero J., Ryan C. Kirkpatrick, and James R. Stevens. 2006. “Rat Race: Insider Advice on Landing Judicial Clerkships.” Pennsylvania State University Law Review 110:835-852.
Avery, Christopher, Christine Jolls, Richard A. Posner and Alvin E. Roth. 2001. “The Market for Federal Judicial Law Clerks.” University of Chicago Law Review 68 (3):793-901.
Avery, Christopher, Christine Jolls, Richard A. Posner and Alvin E. Roth. 2007. “The New Market for Federal Judicial Law Clerks.” University of Chicago Law Review 74: 447-486.
Bailey, Michael A., Brett Kamoie, and Forest Maltzman. 2005. “Signals from the Tenth Justice: The Political Role of the Solicitor General in Supreme Court Decision Making.” American Journal of Political Science 49(1): 72-85.
Baum, Lawrence and Corey Ditslear. 2010. “Supreme Court Clerkships and ‘Feeder Judges.’” Justice System Journal 31 (1):26-48.
Becker, Edward R., Stephen G. Breyer, and Guido Calabresi. 1994. “The Federal Judicial Law Clerk Hiring Problem and the Modest March 1 Solution,” Yale Law Journal 104: 207-225.
Benson, Christopher R. 2007. “A Renewed Call for Diversity Among Supreme Court Clerks: How a Diverse Body of Clerks Can Aid the High Court as an Institution." Harvard BlackLetter Law Journal 23:23-54.
Biskupic, Joan. 2005. Sandra Day O’Connor: How the First Woman on the Supreme Court Became Its Most Influential Justice. New York: HarperCollins.
Black, Ryan C. and Christina L. Boyd. 2012. “The Role of Law Clerks in the U.S. Supreme Court's Agenda-Setting Process." American Politics Research 40 (1): 147-173.
Black, Ryan C., Christina L. Boyd and Amanda C. Bryan. Forthcoming. “Revisiting the Influence of Law Clerks on the U.S. Supreme Court’s Agenda-Setting Process.” Marquette University Law Review.
99 Black, Ryan C. and Ryan J. Owens. 2012. The Solicitor General and the United States Supreme Court: Executive Branch Influence and Judicial Decisions. Cambridge University Press.
Boyd, Christina L., Lee Epstein, and Andrew Martin. 2010. “Untangling the Causal Effects of Sex on Judging.” American Journal of Political Science 54(2): 389- 411.
Brambor, Thomas, William R. Clark, and Matt Golder. 2006. “Understanding Interaction Models: Improving Empirical Analyses.” Political Analysis 14: 63-82.
Brenner, Saul. 1992. “The Memos of Supreme Court Law Clerk William Rehnquist: Conservative Tracts, or Mirrors of His Justice's Mind?" Judicature 76 (2):77-81.
Brenner, Saul. 1997. “Error-Correction on the U.S. Supreme Court: A View from the Clerks’ Memos." The Social Science Journal 34 (1):1-9.
Brenner, Saul and Jan Palmer. 1990. “The Law Clerks' Recommendations and Chief Justice Vinson's Vote on Certiorari." American Politics Research 18 (1): 68-80.
Brown, Mark R. 1996. “Gender Discrimination in the Supreme Court's Clerkship Selection Process.” Oregon Law Review 75: 359-388.
Bryan, Amanda C. and Christopher D. Kromphardt. n.d. “Public Opinion, Institutional Legitimacy, and Counter-Attitudinal Voting on the U.S. Supreme Court.” Working paper: copy on file with the author.
Calabresi, Massimo and David Von Drehle. 2012. “What Will Justice Kennedy Do?" Time June 18, 2012.
Caldeira, Gregory A. and John R. Wright. 1988. “Organized Interests and Agenda Setting in the U.S. Supreme Court." American Political Science Review 82 (4): 1109-1127.
Calvert, Randall L. 1985. “The Value of Biased Information: A Rational Choice Model of Political Advice." Journal of Politics 47:530-555.
Casillas, Christopher J., Peter K. Enns, and Patrick C. Wohlfarth. 2011. “How Public Opinion Constrains the U.S. Supreme Court.” American Journal of Political Science 55(1): 74-88.
Cline, Seth. 2011. “Ruth Bader Ginsburg, Steven [sic] Breyer Wealthiest Judges on the Supreme Court.” September 6, The Center for Responsive Politics. Available at https://www.opensecrets.org/news/2011/09/ruth-bader-ginsburg-steven-breyer/, accessed on December 5, 2014.
Collins, Paul M., Jr. 2008. Friends of the Supreme Court: Interest Groups and Judicial Decision Making. Oxford University Press.
100 Cook, Beverly Blair. 1978. “The Burger Court and Women’s Rights, 1971-1977.” In Women in the Courts, eds. Winifred Hepperle and Laura Crites. Williamsburg, VA: National Center for State Courts.
Cook, Beverly Blair. 1981. “Will Women Judges Make a Difference in Women’s Legal Rights? A Prediction from Attitudes and Simulated Behavior.” In Women, Power, and Political Systems, ed. Margherita Rendel. New York: St. Martin’s Press.
Cook, Beverly Blair. 1984. “Women Judges: A Preface to their History.” Golden Gate University Law Review 14(3): 573-610.
Corley, Pamela C. 2008. “The Supreme Court and Opinion Content: The Influence of Parties’ Briefs.” Political Research Quarterly 61(3): 468-478.
Cross, Frank B. and Emerson H. Tiller. 1998. “Judicial Partisanship and Obedience to Legal Doctrine: Whistleblowing on the Federal Courts of Appeals.” Yale Law Journal 107(7): 2155-2176.
Davis, Sue, Susan Haire and Donald Songer. 1993. “Voting Behavior and Gender on the U.S. Courts of Appeals.” Judicature 77: 129-133.
Deen, Rebecca E., Joseph Ignagni and James Meernik. 2003. “The Solicitor General as Amicus, 1953-2000: How Influential?" Judicature 87 (2):60-71.
Durr, Robert H., Andrew D. Martin, and Christina Wolbrecht. 2000. “Ideological Divergence and Public Support for the Supreme Court.” American Journal of Political Science 44(4):768-776.
Ditslear, Corey and Lawrence Baum. 2001. “Selection of Law Clerks and Polarization in the U.S. Supreme Court." Journal of Politics 63 (3): 869-885.
Enns, Peter K. and Patrick C. Wohlfarth. 2013. “The Swing Justice." Journal of Politics 75 (4): 1089-1107.
Epstein, Lee and Andrew D. Martin. 2010. “Does Public Opinion Influence the Supreme Court? Possibly Yes (But We're Not Sure Why).” University of Pennsylvania Journal of Constitutional Law 13:263-281.
Epstein, Lee, Andrew D. Martin, Jeffrey A. Segal and Chad Westerland. 2007. “The Judicial Common Space." Journal of Law, Economics, and Organization 23 (2): 303-325.
Farhang, Sean, and Gregory Wawro. 2004. “Institutional Dynamics on the U.S. Court of Appeals: Minority Representation under Panel Decision Making.” Journal of Law, Economics, and Organization 20: 299–330.
101 Fischman, Joshua B. and David S. Law. 2009. “What is Judicial Ideology, and How Should We Measure It?” Washington University Journal of Law and Policy 29: 133-214.
Flemming, Roy B., John Bohte, and B. Dan Wood. 1997. “One Voice Among Many: The Supreme Court's Influence on Attentiveness to Issues in the United States 1947- 1992.” American Journal of Political Science 41(4):1224-1250.
Gailmard, Sean and John W. Patty. 2013. Learning While Governing: Expertise and Accountability in the Executive Branch. University of Chicago Press.
Gibson, James L. and Michael J. Nelson. 2015. “Can the U.S. Supreme Court Have Too Much Legitimacy?” In Making Law and Courts Research Relevant: The Normative Implications of Empirical Research on Law and Courts. Edited by Brandon L. Bartels and Chris W. Bonneau. New York, NY: Routledge, 169-179.
Giles, Michael W., Bethany Blackstone, and Richard L. Vining. 2008. “The Supreme Court in American Democracy: Unraveling the Linkages Between Public Opinion and Judicial Decision Making.” Journal of Politics 70:293-303.
Giles, Michael, Virginia Hettinger and Todd C. Peppers. 2001. “Picking Federal Judges: A Note on Policy and Partisan Selection Agendas." Political Research Quarterly 54 (3): 623-641.
Glynn, Adam and Maya Sen. 2014. “Identifying Judicial Empathy: Does Having Daughters Cause Judges to Rule for Women’s Issues?” American Journal of Political Science 1-18.
Gruhl, John, Cassa Spohn, and Susan Welch. 1981. “Women as Policymakers: The Case of Trial Judges.” American Journal of Political Science 25.
Hall, Matthew E.K. 2014. “The Semiconstrained Court: Public Opinion, the Separation of Powers, and the U.S. Supreme Court’s Fear of Nonimplementation.” American Journal of Political Science 58(2): 352-366.
Haruvy, Ernan Alvin E. Roth, and M. Utku Unver. 2006. “The dynamics of law clerk matching: An experimental and computational investigation of proposals for reform of the market,” Journal of Economic Dynamics and Control 30: 457-486.
Ho, Daniel E. and Kevin M. Quinn. 2010. “How Not to Lie with Judicial Votes: Misconceptions, Measurement, and Models.” California Law Review 98 (3): 813- 876.
Iacus, Stefano M., Gary King, and Giuseppe Porro. 2012. “Causal Inference Without Balance Checking: Coarsened Exact Matching.” Political Analysis 20: 1-24.
102 Johnson, Timothy R., Paul J. Wahlbeck, and James F. Spriggs. 2006. “The Influence of Oral Arguments on the U.S. Supreme Court.” American Political Science Review 100 (1): 99-113.
Kastellec, Jonathan P. 2011a. “Hierarchical and Collegial Politics on the U.S. Courts of Appeals.” Journal of Politics 73(2): 345–61.
Kastellec, Jonathan P. 2011b. “Panel Composition and Voting on the U.S. Courts of Appeals Over Time.” Political Research Quarterly 62(2): 377–91.
Kastellec, Jonathan P. 2013. “Racial Diversity and Judicial Influence on Appellate Courts.” American Journal of Political Science: 1-17.
Katz, Daniel M. and Derek K. Stafford. 2010. “Hustle and Flow: A Social Network Analysis of the American Federal Judiciary." Ohio State Law Journal 71 (3): 457-509.
Kearney, Joseph and Thomas Merrill. 2000. “The Influence of Amicus Curiae Briefs on the Supreme Court." University of Pennsylvania Law Review 148 (3): 743-853.
Kenney, Sally J. 2000a. “Beyond Principals and Agents: Seeing Courts as Organizations by Comparing Referendaires at the European Court of Justice and Law Clerks at the U.S. Supreme Court." Comparative Political Studies 33 (5): 593-625.
Kenney, Sally J. 2000b. “Puppeteers or Agents? What Lazarus's Closed Chambers Adds to Our Understanding of Law Clerks at the U.S. Supreme Court." Law and Social Inquiry 185-226.
Kenney, Sally J. 2013. Gender & Justice: Why Women in the Judiciary Really Matter. Routledge: New York.
Kim, Pauline.T. 2008. “Deliberation and Strategy on the United States Courts of Appeals: An Empirical Exploration of Panel Effects.” University of Pennsylvania Law Review 157:1319.
Kozinski, Alex. 1991. “Confessions of a Bad Apple.” Yale Law Journal 100(6):1707 -1730.
Kozinski, Alex and Fred Bernstein. 1998. “Clerkship Politics." Green Bag 2: 57-64.
Kramer, Staci D. 1990. “Enter O’Connor, Exit Mr. Justice.” New York Times Nov. 16. Available at http://www.nytimes.com/1990/11/16/news/enter-o-connor -exit- mr-justice.html, accessed on November 30, 2014.
Kromphardt, Christopher D. 2014. “Fielding an Excellent Team: Law Clerk Selection and Chambers Structure at the U.S. Supreme Court.” Marquette University Law Review.
103 Kromphardt, Christopher D. n.d. “U.S. Supreme Court Law Clerks as Information Sources and Justice Decision Making.” Working paper: copy on file with the author.
Lazarus, Edward. 1999. Closed Chambers: The Rise, Fall, and Future of the Modern Supreme Court. Penguin: New York.
Liptak, Adam. 2010. “A Sign of the Courts Polarization: Choice of Clerks." New York Times September 6, 2010.
Maltzman, Forrest, James F. Spriggs II, and Paul J. Wahlbeck. 2000. Crafting Law on the Supreme Court: The Collegial Game. Cambridge University Press.
Maltzman, Forrest and Paul J. Wahlbeck. 1996. “Strategic Policy Considerations and Voting Fluidity on the Burger Court.” American Political Science Review 90(3): 581-592.
Mark, Peter. 2014. “Arena Stage Finds Material Nearby: Scalia.” Washington Post, Feb. 27, 2014, at C1.
Martin, Andrew D. and Kevin M. Quinn. 2002. “Dynamic Ideal Point Estimation via Markov Chain Monte Carlo for the U.S. Supreme Court, 1953-1999." Political Analysis 10 (2):134-153.
Mauro, Tony. 1998. “Corps of Clerks Lacking in Diversity.” USA Today, March 13.
McAtee, Andrea and Kevin T. McGuire. 2007. “Lawyers, Justices, and Issue Salience: When and How Do Legal Arguments Affect the U.S. Supreme Court.” Law & Society Review 41(2): 259-278.
McGuire, Kevin T. and James A. Stimson. 2004. “The Least Dangerous Branch Revisited: New Evidence on Supreme Court Responsiveness to Public Preferences." Journal of Politics 66(4):1018-1035.
Mishler, William and Reginald S. Sheehan. 1993. “The Supreme Court as a Counter majoritarian Institution? The Impact of Public Opinion on Supreme Court Decisions.” American Political Science Review 87:90-96.
Nelson, William E., Harvey Rishikof, I. Scott Messinger and Michael Jo. 2009. “The Liberal Tradition of the Supreme Court Clerkship: Its Rise, Fall, and Reincarnation?" Vanderbilt Law Review 62 (6): 1749-1814.
Norpoth, Helmut and Jeffrey A. Segal. 1994. “Popular Influence on Supreme Court Decisions.” American Political Science Review 88:711.
Oberdorfer, Louis F. and Michael N. Levy. 1992. “On Clerkship Selection: A Reply to the Bad Apple.” Yale Law Journal 101: 1097-1108.
104 Palmer, Barbara. 2001. “The "Bermuda Triangle?": The Cert Pool and Its Influence over the Supreme Court's Agenda." Constitutional Commentary 18: 105-120.
Palmer, Jan and Saul Brenner. 1995. “The Law Clerks' Recommendations and the Conference Vote On-the-Merits on the U.S. Supreme Court." Justice System Journal 18 (2):185-197.
PBS Newshour. 1998. “Color-Blind Court?” July 23. Transcript available at http://www.pbs.org/newshour/bb/social_issues-july-dec98-clerks_7-23/, accessed on December 5, 2014.
Pearlstein, Deborah N. 2000. “The Power to Persuade: A Year in the Life of a Supreme Court Clerk." NCJW Journal
Peppers, Todd C. 2006. Courtiers of the Marble Palace: The Rise and Influence of the Supreme Court Law Clerk. Stanford Law and Politics.
Peppers, Todd C. 2012. “Of Leakers and Legal Briefers: The Modern Supreme Court Law Clerk." Charleston Law Review 7:95-109.
Peppers, Todd C., Michael W. Giles and Bridget Tainer-Parkin. 2008. “Inside Judicial Chambers: How Federal District Judges Select and Use Their Law Clerks.” Albany Law Review 71: 623-646.
Peppers, Todd C. and Artemus Ward. 2012. In Chambers: Stories of Supreme Court Law Clerks and Their Justices. University of Virginia.
Peppers, Todd C. and Christopher Zorn. 2008. “Law Clerk Influence on Supreme Court Decision Making: An Empirical Assessment." DePaul Law Review 58:51-77.
Peresie, Jennifer L. 2005. “Female Judges Matter: Gender and Collegial Decisionmaking in the Federal Appellate Courts.” Yale Law Journal 114.
Perry, H.W. 1991. Deciding to Decide: Agenda Setting in the United States Supreme Court. Harvard University Press. Poole, Keith T. 1998. “Estimating a Basic Space From a Set of Issue Scales.” American Journal of Political Science 42:954–993.
Posner, Richard A. 1995. “The Rise of the Law Clerk." The Long Term View 3 (1):23-31.
Posner, Richard A. 2009. How Judges Think. Harvard University Press.
Priest, George L. 2005. “Reexamining the Market for Judicial Clerks and Other Assortative Matching Markets.” Yale Journal on Regulation 22: 123-204.
Rehnquist, William H. 1957. “Who Writes Decisions of the Supreme Court?” U.S. News & World Report.
105 Rhinehart, Lynn K. 1994. “Is There Gender Bias in the Judicial Law Clerk Selection Process?” Georgetown Law Journal 83: 575-604.
Rosenthal, Jeffrey S. and Albert H. Yoon. 2011. “Judicial Ghostwriting: Authorship on the Supreme Court." Cornell Law Review 96:1307-1344.
Ruger, Todd. 2012. “Despite efforts for new black judicial clerks, percentages going down.” National Law Journal, April 23.
Saad, Lydia. 2010. “Education Trumps Gender in Predicting Support for Abortion.” Gallup, April 28. Available at http://www.gallup.com/poll/127559/education- trumps-gender-predicting-support-abortion.aspx, accessed on January 21, 2015.
Savage, David G. 2011. “Clarence Thomas is His Own Man." Los Angeles Times July 3, 2011.
Segal, Jeffrey and Albert Cover. 1989. “Ideological Values and the Votes of U.S. Supreme Court Justices." American Political Science Review 83: 557-565.
Segal, Jeffrey, Lee Epstein, Charles Cameron and Harold Spaeth. 1995. “Ideological Values and the Votes of U.S. Supreme Court Justices Revisited." Journal of Politics 57: 812-823.
Sommers, Samuel R. 2006. “On Racial Diversity and Group Decision Making: Identifying Multiple Effects of Racial Composition on Jury Deliberations.” Journal of Personality and Social Psychology 90(4): 597- 612.
Spaeth, Harold. 2003. The Original United States Supreme Court Database, 1953-2004 Terms. East Lansing, MI: Department of Political Science, Michigan State University. http://web.as.uky.edu/polisci/ulmerproject/sct-data.htm (accessed December 14, 2011).
Spitzer, Matthew and Eric Talley. 2013. “Left, Right, and Center: Strategic Information Acquisition and Diversity in Judicial Panels.” Journal of Law, Economics, and Organization 29(3): 638-680.
Spriggs, James F. and Paul J. Wahlbeck. 1997. “Amicus Curiae and the Role of Information at the Supreme Court." Political Research Quarterly 50 (2): 365-386.
Stephenson, Matthew C. 2011. “Information Acquisition and Institutional Design.” Harvard Law Review 124: 1422-1483.
Stras, David R. 2007. “The Supreme Court’s Gatekeepers: The Role of Law Clerks in the Certiorari Process." Texas Law Review 85: 947-997.
Sunstein, Cass R., David Schkade, Lisa M. Ellman, and Andres Sawicki. 2006. Are Judges Political? An Empirical Analysis of the Federal Judiciary. Washington,
106 DC: Brookings Institution Press.
Swanson, Rick A. and Stephen L. Wasby. 2008. “Good Stewards: Law Clerk Influence in State High Courts.” Justice System Journal.
Szmer, John and Martha H. Ginn. 2014. “Examining the Effects of Information, Attorney Capability, and Amicus Participation on U.S. Supreme Court Decision Making.” American Politics Review 42(3): 441-471.
Szmer, John J., Tammy A. Sarver, and Erin B. Kaheny. 2010. “Have We Come a Long Way, Baby? The Influence of Attorney Gender on Supreme Court Decision Making.” Politics & Gender 6: 1-36.
Tate, C. Neal. 1981. “Personal Attribute Models of the Voting Behavior of U.S. Supreme Court Justices: Liberalism in Civil Liberties and Economics Decisions, 1946-1978." American Political Science Review 75 (2):355-367.
Tate, C. Neal and Roger Handberg. 1991. “Time Binding and Theory Building in Personal Attribute Models of Supreme Court Voting Behavior, 1916-88." American Journal of Political Science 35 (2):460-480.
Thomas, Andrew P. 2002. Clarence Thomas: A Biography. Encounter Books.
Ura, Joseph D. 2014. “Backlash and Legitimation: Macro Political Responses to Supreme Court Decisions.” American Journal of Political Science 58(1): 110- 126.
Velvel, Lawrence R. 1995. “Introduction." The Long Term View 3:2-8.
Wahlbeck, Paul J., James F. Spriggs and Lee Sigelman. 2002. “Ghostwriters on the Court?: A Stylistic Analysis of U.S. Supreme Court Opinion Drafts." American Politics Research 30 (2):166-192.
Wald, Patricia M. 1995. “Selecting Law Clerks.” The Long Term View 45-46.
Ward, Artemus and David L. Weiden. 2006. Sorcerer's Apprentices: 100 Years of Law Clerks at the United States Supreme Court. New York University Press.
Ward, Artemus and Stephen L. Wasby. 2010. “‘Get a Life!’” On Interviewing Law Clerks." Justice System Journal 31 (2):1-19.
Wasby, Stephen L. 2006. “Why Clerk? What Did I Get Out of It?" Journal of Legal Education 56 (3):411-429.
Washington, Ebonya. 2008. “Female Socialization: How Daughters Affect Their Legislator Fathers' Voting on Women's Issues." American Economic Review 98(1): 311-33.
107 Weinberg, Jill D. and Laura Beth Nielsen. 2011. “Examining Empathy: Discrimination, Experience, and Judicial Decision Making.” University of Southern California Law Review 85.
Wilkinson, J. Harvie, III. 2012. “Justice Lewis F. Powell, Jr.: A Personal View by a Former Clerk.” In In Chambers: Stories of Supreme Court Law Clerks, eds. Todd C. Peppers and Artemus Ward. University of Virginia.
Worthen, Kevin J. 2012. “Shirt-Tales: Clerking for Byron White.” In In Chambers: Stories of Supreme Court Law Clerks, eds. Todd C. Peppers and Artemus Ward. University of Virginia.
108