Over the past two decades, criticism of Supreme Court confirmation hearings has intensified considerably. In particular, there is a growing sense among Court observers that nominees are now more reluctant to answer questions during their Senate testimony, and that the quality of the hearings has deteriorated as a result. Recent proceedings have been described as “a vapid and hollow charade” (Reference KaganKagan 1995: 941), an “exercise in obfuscation” (Reference Yalof and SaratYalof 2008: 141), and even a carefully choreographed “ ‘kabuki’ dance” (Reference FitzpatrickFitzpatrick 2009). In short, the message from critics is clear: Nominees now say “nothing of value” during their testimony (Reference LemieuxLemieux 2010), and the hearings are no longer what they used to be.
The origins of this recent trend toward nominee evasiveness are generally traced to Robert Bork's 1987 hearings (Reference LemieuxLemieux 2010; see also Reference CarterCarter 1994; Reference KaganKagan 1995). Observers argue that Bork's lengthy and candid answers doomed his nomination, and that subsequent nominees have avoided a similar fate by becoming more cagey and unresponsive in their testimony. In this regard, the approach taken by Ruth Bader Ginsburg during her 1993 confirmation proceedings is now generally regarded as paradigmatic. By declining to discuss any issue that might come before the Court, Ginsburg is thought to have charted a new course for post-Bork nominees. Indeed, the so-called “Ginsburg rule”—where nominees invoke their right not to answer questions about potentially unsettled legal debates—has become shorthand for the idea that contemporary nominees duck difficult questions in a way that their predecessors did not (e.g., Reference FroomkinFroomkin 2010; Reference TurleyTurley 2009).Footnote 1
But how accurate is this widely accepted version of events? Have nominees simply become less candid in recent years? Regrettably, answers to these questions have been elusive. While existing studies of the hearings have shed some light on changes in nominee testimony, these efforts have generally been limited to a small number of justices or to narrow bands of questions (see, e.g., Reference ComiskeyComiskey 2004; Reference Czarnezki, Ford and RinghandCzarnezki et al. 2006; Reference Guliuzza, Reagan and BarrettGuliuzza et al. 1994; Reference RinghandRinghand 2008; Reference Watson and StookeyWatson & Stookey 1988; Reference Williams and BaumWilliams & Baum 2006). Thus, although it is possible that the Bork proceedings were a turning point, and that things have gone rapidly downhill since that time, we simply do not yet know whether this is a fair assessment of the hearings. Perhaps the questions themselves have become more difficult to answer over the years. Or perhaps recent nominees only appear to be evasive because of the way in which they frame their responses. Clearly, given the intense interest that the hearings generate—to say nothing of the important role that they play in our political system—it is vital for those who study courts and the judiciary to know whether these proceedings have in fact changed over time—and if so, why.
To that end, this article presents the results of a content analysis of every Supreme Court confirmation hearing since 1955, the year that the hearings before the Senate Judiciary Committee became a regular part of the confirmation process. For each hearing, we code all of the exchanges between a senator and the nominee, recording things such as the type of question asked, the degree to which the answer was candid and forthcoming, and the reasons the nominees might give for not answering more fully. Using this original dataset, we then test a series of hypotheses about nominee responsiveness (or what we call “candor”) in the face of Senate questioning.Footnote 2 Our results suggest that contrary to popular belief, recent nominees are not drastically more evasive than their predecessors. Instead, we find that specific changes in the types of questions that are being asked, changes in the topics of those questions, and in the ways in which nominees answer them have all likely helped to fuel the perception that candor has declined. These findings suggest to us that the chorus of criticism directed at recent nominees is based more on an impression of increased evasiveness than on empirical evidence.
A Brief History of the Hearings
Before turning to our analysis, it is important to sketch a brief history of the confirmation hearings for Supreme Court justices and to highlight some important changes that have characterized the proceedings over the past several decades. Some of this historical information is well known, but other aspects have been largely undocumented. Indeed, even the most attentive students of the Court may be surprised to learn that confirmation hearings before the Senate Judiciary Committee did not become the norm until the 1955 nomination of John M. Harlan. Prior to that time, nominations were generally moved through the Committee with much less debate; and, in fact, until 1868 they were not even referred to the Committee. Even after 1868, when the Senate Judiciary Committee involvement became a standard part of the process, the hearings were not public, and the nominees themselves rarely appeared. There were exceptions: Louis Brandeis's nomination in 1916 had a hearing, although he did not testify; Harlan Fiske Stone was the first nominee to appear at his hearing, in 1925; and Felix Frankfurter and Robert Jackson had full, open hearings in 1938 and 1941, respectively (Reference ThorpeThorpe 1969; Reference Ringhand and CollinsRinghand & Collins 2010). But the Harlan nomination—which was, not coincidentally, held in the immediate wake of the Court's controversial decision in Brown v. Board of Education (1954)—marks the beginning of a string of hearings that has remained unbroken for more than a half century.
While the hearings have thus been a fixture of the Supreme Court nomination process since 1955, it would be a mistake to assume that they had not changed considerably during that time. Much of the focus of our study, of course, will be on the ways in which the Committee's questions and the nominees’ answers have evolved over the years. For now, however, we pause to point out three significant changes that operate at more of a structural level.
The first of these changes is the number of questions that each nominee is asked by the Senate Judiciary Committee. To be sure, most Court watchers are probably aware that the hearings have gotten longer over the years, and that the amount of scrutiny that nominees now face from the Committee is greater than in earlier periods. Moreover, this is a logical development, given that the hearings have attracted more attention over time, thus providing senators with an attractive opportunity to serve their constituents and bolster their quest for reelection.
Still, the full extent of this change is quite striking. As shown in Figure 1, the number of exchanges—essentially a question and its corresponding answer—has grown dramatically during the past 50 years.Footnote 3 Early nominees, such as John M. Harlan, William Brennan, and Charles Whittaker, were asked fewer than 100 questions, and Byron White was asked only six. Today, by contrast, the number is exponentially larger, with nominees such as John Roberts and Samuel Alito facing more than 700 questions each.
Figure 1. Level of Senatorial Scrutiny Given to Supreme Court Nominees over Time.
Moreover, as indicated by the line on Figure 1, there is a strong relationship between time and the number of questions asked of each nominee, suggesting that the increase in scrutiny given to each nominee has been both dramatic and steady. In fact, apart from the length of the hearing of two early outliers, Marshall and Haynsworth, the ten longest hearings have all occurred since the early 1980s. Also, we note that while Bork's hearing predictably had the most questions, the increase in senatorial scrutiny started to occur prior to Bork's hearings. Thus, while conventional accounts of the hearings often identify Bork as the tipping point, this particular change in the level of scrutiny clearly has its roots in the 1960s and 1970s.
The next important structural change in the confirmation hearings is that the questioning has gotten much more balanced between the two parties over time. Early hearings were often controlled by one party. For instance, Democrats asked 164 of the 174 questions (94.3 percent) during the Potter Stewart hearings.Footnote 4 However, those days are now gone. Figure 2 illustrates this trend by tracking the percentage of questions asked by a senator who is from the opposite party of the president. Thus if the hearing is relatively balanced, the percentage of questions asked by “opposition” senators should be around 50 percent. If a hearing is unbalanced, there will be a large deviation from 50 percent (either above or below). As shown, parity between the parties increased significantly around the early 1980s.
Figure 2. Changing Balance of Senatorial Scrutiny over Time.
Here again, these findings do not really come as a surprise. Given that seats on the federal bench are generally viewed as being highly valued, a seat on the Senate Judiciary Committee will allow senators to shape the judiciary. Additionally, because the hearings are now routinely televised, members of Congress can use their Judiciary Committee seat to ask questions that might appeal to their constituencies. Nevertheless, we think it is important to note the degree to which this balance has taken shape in recent years, and to see the degree to which the hearings were much more free-form prior to the 1980s.
One final historical development is relevant to our study: the advent of televised coverage of the hearings. This trend began with Sandra Day Reference O'ConnorO'Connor's 1981 proceedings, which aired on C-SPAN. Since that time, all hearings have gotten increased broadcast exposure, first from public television, which has provided gavel-to-gavel coverage since William Rehnquist's 1986 hearing to become chief justice, and later from CNN and other cable networks, which joined the fray with the Bork hearing in 1987. Today the hearings can now be seen on numerous cable and traditional broadcast channels, as well as via the Internet. We note this development because we believe that many of the changes we will see in the hearings—and many of those we have already discussed—coincide with the beginning of televised coverage.
In addition to these structural changes to the hearings themselves, several recent innovations to the pre-hearing process have dramatically increased the amount of information that senators have about nominees before testimony actually begins. For example, senators now submit written questionnaires to nominees several weeks prior to the start of the hearings. Nominees now also make “courtesy calls” to each member of the Judiciary Committee, as well as to other senators. Finally, the nominees' past decisions, speeches, and other public remarks are now much more readily available to senators through television and the Internet.
Taken together, these changes to the pre-hearing period are important to note because they may have an impact on the way that senators approach questioning once the hearings begin. For instance, it is possible that in the interest of time, senators who have gotten answers from nominees prior to the start of the hearings will avoid asking those same questions again, and will instead ask follow-up questions—some of which may be more probing and difficult than the original question. Thus it may be that in recent years, senators come to the hearings better prepared to ask more difficult questions of the nominees.
Having documented these three structural changes, we pause briefly to review the fate of the nominees who have gone through the confirmation process during the “Hearings Era.” Of the 30 prospective justices who have had hearings between John M. Harlan in 1955 and Elena Kagan in 2010, all but five have been confirmed by the Senate.Footnote 5 In Reference Fortas1968, Justice Abe Fortas had his nomination for chief justice withdrawn after his hearing. This, in turn, nullified the nomination of Homer Thornberry, who was slated to take Fortas's vacated associate justice seat. One year later, the Senate rejected Clement Haynsworth, and in 1970, another Nixon nominee, Harrold Carswell was rejected as well. The most recent nominee to have a hearing and not be confirmed was, of course, Robert Bork in 1987, whose now infamous hearing is often credited with encouraging subsequent nominees to be more reserved and evasive in their responses.
Prior Work on Confirmation Hearings
Existing work on Supreme Court nominee responsiveness falls into three broad categories. The first focuses on the increased ideological nature of the hearings over the years. Most authors suggest that it was the Bork hearings that changed these proceedings dramatically in this regard (e.g., Reference CaldeiraCaldeira 1989; Reference DavisDavis 2005; Reference Epstein, Lindstadt, Segal and WesterlandEpstein et al. 2006; Reference Martinek, Kemper and Van WinkleMartinek et al. 2002). Others argue that changes started before Bork when ideology started to play a more important role during the early 1980s (Reference Krutz, Fleisher and BondKrutz et al. 1998). Meanwhile, Reference BorkBork (1990) himself identifies the critical changing point as beginning much earlier, during the politicization of the Court during the late 1950s and early 1960s when the Warren Court started to play a bigger role in many Americans’ lives. Reference Epstein, Lindstadt, Segal and WesterlandEpstein et al. (2006) also find support for this argument. Overall, however, the consensus among these authors is that the hearings at some point became more ideological—an idea to which we return in the analysis that follows.
The second main literature on this subject comprises empirical investigations of the hearings’ transcripts and testimony. Some of these studies explore changes in Senate questioning over time. Most notably, a recent large-scale investigation by Reference Ringhand and CollinsRinghand and Collins (2010) documents changes in the issue areas of the Committee's questions since the hearings began, and finds that a senator's political party plays a role in the kinds of topics that he or she asks. Interestingly, Ringhand and Collins also discover that minority and female nominees are subjected to more substantive questioning than their white male counterparts. This report expands upon an earlier effort by Reference Williams and BaumWilliams and Baum (2006), which focused largely on questions about a nominee's past judicial decisions, and found that those types of inquiries had increased over time. Meanwhile, other empirical investigations look at nominee responses. For instance, Reference RinghandRinghand (2008) examines how the justices on the Rehnquist Natural Court answered questions about past Supreme Court cases and finds that the responses tended to be more general than specific. Reference Czarnezki, Ford and RinghandCzarnezki et al. (2006) also study the content of the Rehnquist justices’ hearings, and conclude that nominee answers during the hearings are not a reliable indicator of how the justice will vote on cases.
The last relevant literature is more prescriptive in nature. Some scholars argue that in the interest of judicial independence, Supreme Court nominees should not be required to answer questions about how they would rule in specific cases. Reference CarterCarter (1988), for instance, urges a circumscribed role for Senate questioning, arguing that senators should confine themselves to learning about a nominee's basic qualifications and her “background moral vision and the capacity for moral reflection” (Reference CarterCarter: 1199). Others envision a slightly more robust role for the Senate that includes inquiries into the nominee's judicial philosophy or general approach to constitutional interpretation (Reference EisgruberEisgruber 2007; Reference GoldbergGoldberg 2004; Reference TribeTribe 1985, Reference Tribe2008, but see Reference ComiskeyComiskey 2004).Footnote 6 A third group, spearheaded by Post and Reference Post and SiegelSiegel (2006) contends that senators should be allowed to ask nominees about how they would have voted in cases that have already been decided by the Supreme Court (see also Reference Yalof and SaratYalof 2008). As Reference RinghandRinghand (2008) laments, however, this last approach runs into trouble because nominees will often claim that cases are still “unsettled,” despite their having been ruled upon by the Court. The most famous examples of this, as Ringhand points out, came during Justice Scalia's and Chief Justice Rehnquist's hearings: both nominees refused to answer questions about Marbury v. Madison on the grounds that some of the issues from that landmark 1803 case might arise again, thus impeding their judicial independence.
Taken together, these three bodies of work support the general hypothesis that Supreme Court confirmation hearings have changed over time. Regrettably, this existing work does not provide the kind of data that one would need in order to determine what, if anything, has in fact changed over the past half century with regard to nominee testimony. The studies by Reference Czarnezki, Ford and RinghandCzarnezki et al. (2006) and Reference RinghandRinghand (2008) probably come closest in this regard, but both studies are limited to the nine justices who served under Rehnquist, and the response data the authors report are limited to narrow sets of questions, rather than all of the questions that senators ask. Thus, while these and the other analyses provide useful building blocks, they do not provide a sufficiently longitudinal or thorough set of observations about nominee candor. In Reference Williams and BaumWilliams and Baum's 2006 report, for example, the total number of observations is 240; by contrast, the analysis we conducted comprised nearly 11,000 exhanges.
Again, the value of these earlier reports—and in particular, the recent Ringhand and Collins investigation into the changes in questioning over time—should not be underestimated. However, a pressing need remains for a comprehensive examination of nominee responses to Senate questioning. To that end, in the next section we explain how we content analyzed the Supreme Court confirmation hearings, and how we developed measures for the type and content of senators’ questions, the level of candor in the nominees’ responses, and the reasons why nominees are being evasive.
Coding Methodology
A quick glance at the hearing transcripts suggests that nominees have been evasive for some time. Just two questions into his 1955 testimony, for example, John M. Harlan avoided answering a question on civil rights (Reference HarlanHarlan 1955: 137). William Brennan did the same, declining to answer a question on communism just as the hearings began (Reference BrennanBrennan 1957: 17). In fact, similar tactics can be found within the first few pages of nearly every transcript from the 1950s onward. Meanwhile, long-standing concerns from Committee members about nominee evasiveness are also not hard to find. For example, during Thurgood Reference MarshallMarshall's 1967 hearing, Senator Sam Ervin quipped, “How can the Senate perform its duty and ascertain what your constitutional or judicial philosophy is without ascertaining what you think about the Constitution?” (Reference MarshallMarshall 1967: 54). A year later, Ervin again expressed his frustration, this time during Judge Homer Thornberry's testimony, charging that the proceedings had “virtually created a new right not found in the Constitution, which might well be designated as the judicial appointee's right to refrain from self-incrimination” (Reference ThornberryThornberry 1968: 274).
Clearly then, there is anecdotal evidence to support the idea that even early nominees were not fully forthcoming during their confirmation hearings. But to what extent do these anecdotes reflect a broader pattern in the hearings? To help answer this question, we developed a detailed coding scheme that allowed us to explore as many aspects of the hearing transcripts—the questions that senators ask and the ways in which nominees respond—as possible. We then applied this method to the printed transcripts for all confirmation hearings between 1955 (Harlan) and 2010 (Kagan).Footnote 7
Coding Scheme
Our basic unit of analysis was an exchange. An exchange was defined as a question from a senator plus a response from a nominee.Footnote 8 In most situations, this meant a single question and a single response. On occasion, we included multiple back-and-forths within a single exchange, such as when the nominee and senator were clearly talking over each other (or at the same time).Footnote 9 Two coders (one of the principal investigators and one graduate student coder) then recorded the following variables for each exchange.Footnote 10 As noted earlier, the total number of exchanges included in our analysis (n = 10,883) was much larger than the datasets used in previous published reports.
(a) Question of Fact or Question of View We divided all questions into one of two main groups. Questions of Fact (QOF) are questions that seek basic factual information about a topic or issue. Questions of View (QOV) seek nominees’ opinions, thoughts, assessments, interpretations, or predictions.Footnote 11 For example, the question, “Where did you go to law school?” would be a Question of Fact, whereas “Do you think the Constitution protects the right to privacy?” would be a Question of View. While we recognized that this distinction would be obvious in most instances, we also anticipated that it would help us differentiate in some cases between questions that were more likely to generate forthcoming responses and those that were not. For example, a question about when a case was decided (QOF) is more likely to be answered without reservation than a question about how a nominee would have voted in that case (QOV). Coding both of these simply as a question about a past case, rather than distinguishing between them, would constrain our ability to track an important potential difference in the way that nominees respond.
(b) Question of Fact topics Questions of Fact fall into one of the following four main categories: (1) factual questions about a nominee's legal education; (2) factual questions about a nominee's personal biography or family; (3) factual questions about a nominee's nonlegal employment history; and (4) factual questions about past cases, as well as factual questions about the nominee's writings, speeches, previous testimony, and other issues that did not fit into the first three main categories.Footnote 12
(c) Question of View topics Questions of View seek a nominee's views on one of the following topics: (1) past Supreme Court rulings or a lower court ruling; (2) hypothetical cases; (3) approach to judging and constitutional interpretation; (4) powers of Congress and the president; (5) federalism and states’ rights; (6) judicial power and administration; (7) peace, security, law and order; (8) individual rights and liberties; (9) other topics not identified above. Questions that cover more than one issue were coded with the main topic first, followed by secondary topics, if any.
(d) Candor of nominee response To assess the level of candor/evasiveness, coders assigned each nominee response to one of five categories:
1. Fully/Very Forthcoming: Nominee answered the question that was asked without any qualification or evasion.Footnote 13
2. Qualified: Nominee indicated some reason for not answering the question fully, but then gave at least a partial response to the question.
3. Not Forthcoming: Nominee chose not to answer the question at all.
4. Interruption: Nominee was interrupted by a senator before s/he even had a chance to give a partial response.
5. Non-Answer: Nominee gave a nonsubstantive response (e.g., “Senator, you ask difficult questions”) to a substantive question. Or nominee gave a factual answer to a Question of View.Footnote 14 Or, the nominee answered the question with a question (e.g., “Is that what you're asking me?”). This should not be confused with the Not Forthcoming option (number 3, above).
We consider each of these categories to be a crucial part of capturing the candor or evasiveness of nominees. That is, rather than simply relying on two or three categories, we designed our coding to capture a more complete range of nominee responses. In particular, we anticipated that the last category, Non-Answer, could be especially important for our analysis because it might represent an important evasive tactic for nominees. Our thinking was that nominees are likely aware that senators are constrained in their questioning time. Thus, by giving a Non-Answer, the nominee can decrease the total number of difficult questions about their views—a subtle but important type of answer evasion.
(e) Reason for Qualified or Not Forthcoming response If a nominee response was coded as Qualified or Not Forthcoming, coders then identified one of six reasons or explanations:
1. Nominee expressed concerns about answering a question about a case or issue that was before the Court or could be before the Court.
2. Nominee said the issue should be handled by another branch of government.
3. Nominee expressed general concerns about conflict of interest and maintaining judicial independence.
4. Nominee claimed s/he did not have enough information, or could not remember enough about the issue, to give more than a partial response.
5. Nominee claimed s/he did not have enough information, or could not remember enough about the issue, to give any response.Footnote 15
6. Other, reason unclear, or unspecified.
Examples
Examples drawn from the transcripts themselves help illustrate how coders distinguished the level of candor or Forthcomingness of nominee responses.
(a) Fully Forthcoming response:
Sen. Kohl: Is there a general constitutional right to privacy? And where is the right to privacy, in your opinion, found in the Constitution?
Judge Sotomayor There is a right of privacy. The court has found it in various places in the Constitution, has recognized rights under those various provisions of the Constitution. It's founded in the Fourth Amendment's right and prohibition against unreasonable search and seizures (Reference SotomayorSotomayor 2009).
(b) Qualified response:
Sen. Tydings: Judge Burger, do you feel that a judge should assume the role of chief judge of a circuit court or a chief judge of a district court after he passes the age of 65?
Judge Burger: Well, I do not think I am qualified really to pick a fixed age. I think there certainly is much to be said for limitations. And the Congress created the limitation about 10 years ago at age 70. At that time I think some of the proposals were to make it 65, and the compromise was made on 70. I do not have a firm view on it. I really cannot go beyond that, Senator Tydings (Reference BurgerBurger 1969: 12).
(c) Not Forthcoming response:
Sen. McClellan: Then I take it you disagree with that philosophy of that opinion.
Judge Marshall: I am not saying whether I disagree with it or not, because I am going to be called to pass upon it. There is no question about it Senator. These cases are coming to the Supreme Court (Reference MarshallMarshall 1967: 9).
(d) Interrupted response:
Sen. Biden How about 1969?
Judge Scalia: Well, that's not 1803. All I can say is—
Sen. Biden: I am really trying to get a sense of time (Reference ScaliaScalia 1986: 104).
(e) Non-Answer:
Sen. Hatch: Do you see any inconsistency in giving parents the right to consent but denying the similar protection or privilege to the father of the child?
Judge O'Connor: Senator, my recollection of the Utah statute is that it was not one that provided for parents consent but rather for notification to the parents without a consent aspect. In fact, I think that the Supreme Court in an earlier decision had held that a statute from another state which required parental consent for minor to obtain an abortion was invalid (Reference O'ConnorO'Connor 1981: 87).
Expectations of Change over Time
As we noted earlier, recent nominees are routinely charged with being evasive or unhelpful when answering questions, leading to rhetoric in the public sphere about the health and proper functioning of the confirmation process and, more broadly, of our democratic government.Footnote 16 The problem with this criticism (or any other criticism of either the nominees or senators), however, is that it is difficult to discern whether evasiveness is primarily attributable to the nominees’ reluctance to speak out on issues for fear of costing themselves confirmation votes, or whether it is primarily attributable to the changing nature of questions asked by the senators. In other words, recent nominees may be responding in a similar fashion as earlier nominees, but the scope or qualitative nature of the questions and hearings may have changed. To complicate this perception even further, it is even possible that the nominees’ overall level of candor has remained relatively constant over time, but that they have changed the way they frame their answers in a way that has contributed to a perception that candor has decreased.
To help disentangled these interrelated parts of the nominee evasiveness conundrum, we test a series of hypotheses aimed at determining whether levels of candor have simply decreased over time or whether other factors may be contributing to the perception that nominees are now less forthcoming than they were in the past.
Nominee Candor Over Time
We begin with the central question driving this study: Has nominee candor simply decreased over time? As confirmation battles have become scrutinized more closely by the media pundits, and those with partisan and ideological interests more firmly entrenched in their camps, the conventional wisdom among scholars and Court watchers suggests that the level of evasiveness has also increased (e.g., Reference BensonBenson 2010; Reference ComiskeyComiskey 1994; Reference KaganKagan 1995). On the basis of this body of existing work, therefore, we hypothesize that (H1): The level of nominee candor will decrease over time. To test this hypothesis, we count the number of Fully Forthcoming, “qualified,” “not forthcoming,” and “non-answer” responses for each nominee and plot the percentages against an x-axis that lists the justices in chronological order.
As Figure 3 illustrates, we find only partial support for this first hypothesis. On the one hand, contrary to the conventional portrayals of the hearings in serious decline, we find that Fully Forthcoming responses—the category where nominees are not being evasive in any way—have not decreased dramatically over time. On the other hand, candor has not been entirely static over the years. Indeed, there has been a small decline over the past 20 years since the Bork hearings. Thus it appears that at least part of what critics have described in terms of a recent change is accurate.
Figure 3. General Level of Nominee Candor in Response to Senatorial Scrutiny: All Types of Questions.
However, when one focuses more closely on the Fully Forthcoming line, a few things become clear. First, the downward trend is less pronounced and dramatic than we would have expected based on existing accounts of the hearings. Second, there appear to be at least three separate, downward trends: one that runs from Goldberg to Fortas (Chief Justice), one from Haynsworth to Scalia, and one from Bork to Kagan.Footnote 17 Thus the picture that emerges is a pattern of ebbs and flows, not an isolated, steep drop-off since the late 1980s.
Relationship between Candor, the Types of Questions, and Who Asks Them
From what we have learned thus far, it appears that recent nominees have not simply adopted a more evasive posture during their hearings. To the extent that our analysis reveals other, similar downward trends during the past half century, it seems clear that there may be other factors driving changes in the degrees of candor. In other words, because of the fact that the recent drop-off is not unique, we think that whatever decline in candor we have seen in the past two decades has likely been caused by factors other than simply a reaction to the Bork hearings.
To explore this possibility, we elect to look more closely at the questions that nominees are being asked: the type of question (fact or view), its topic, and who is asking it. More precisely, we expect that three particular types of questions were more likely to trigger nominee evasiveness: (1) questions that focus on a nominee's views and beliefs (as opposed to questions that focus on facts); (2) questions that focus on controversial issues such as civil liberties; and (3) questions that are asked by a senator from an opposing party or by an ideologically distant senator. Taken together, these three expectations point to our second hypothesis (H2): Probing, controversial, and challenging questions are more likely to trigger nominee evasiveness. We examine the three central elements of H2 in turn.
(1) Questions of View Will Drive Down Candor
Here our thinking is straightforward. The literature finds that the hearings have become more ideologically charged over time (Reference Epstein, Lindstadt, Segal and WesterlandEpstein et al. 2006; Reference Krutz, Fleisher and BondKrutz et al. 1998). This kind of environment encourages senators to ask questions that help them gather substantive, ideological information about the nominees (QOVs). It also means that nominees will be reluctant to provide that kind of ideological information, since it has a higher chance of derailing their bid for the Court than would benign factual information. The result of this dynamic, we believe, is that nominees will be less likely to answer QOVs than QOFs.
As illustrated by Figure 4, our expectations about the role of questions of view were largely met. In most cases, questions directed towards a nominee's beliefs and ideas are more likely to trigger nonforthcoming responses than questions about basic factual information. There were some nominees for whom this was not true (Whittaker, White, and Goldberg, for example), but there were others for whom the predicted disparity was especially dramatic (Fortas, Marshall, and Thornberry). Overall, however, we see a clear tendency among nominees to answer QOVs more fully than they do QOFs, providing us with strong initial support for our second hypothesis.
Figure 4. Nominee Candor: Percentage of “Very Forthcoming” Responses, by Question Type.
Additionally, Figure 4 also highlights another interesting finding about nominee candor. Specifically, on the right hand side, we note that candor on QOV was relatively high for Kennedy (immediately following Bork), but then experienced a declining trend, starting with Souter and continuing through Kagan. This is noteworthy because nominee candor on QOF held relatively constant during this same time span. This also partly explains some of the apparent perception that these hearings have declined in candor over time. We also note, however, that there were periods during the 1960s and 1970s that experienced declining levels of candor, only to have them increase later.
One additional aspect of the role of QOVs needs to be discussed. Specifically, perhaps nominees appear to be answering fewer questions today simply because they are being asked more QOVs than in years past and not because they have adopted a more evasive approach in general. This prompts us to ask an obvious question: Are more QOV being asked today? Again, given the increasingly ideological nature of the proceedings discussed above, we would expect to see an upward trend in QOVs. Figure 5 confirms this result, showing a general upward trend in the percentage of questions directed at a nominee's ideas, thoughts, beliefs, and attitudes over time. The line in Figure 5 is a Lowess smoother that is designed to provide a weighted best fit between the points over time. This also implies that the proportion of factual questions has decreased.
Figure 5. Changing Type of Senatorial Scrutiny over Time: Increasing Importance of a Nominee's Views.
While we are not yet prepared to draw any firm conclusions based on this interesting observation, we do believe that it at least opens up the possibility that recent nominees may appear to be less forthcoming than their predecessors because of the increased number of QOVs rather than any change in the approach taken by nominees over time. We will return to this issue in our concluding comments.
(2) Questions on Civil Liberties Issues Will Drive Down Candor
The second expectation within H2 is closely related to the first. Here we anticipate that a particular type of QOV—namely, questions involving controversial issues in the areas of civil liberties and civil rights—will exert an especially strong downward influence on nominee candor.Footnote 18 Our thinking here follows the same general logic described above: Nominees know that cases involving hot-button issues such as abortion, affirmative action, religious freedom, and gay rights have helped place the Supreme Court—and, consequently, prospective justices as well—at the “storm center” of American politics (Reference O'BrienO'Brien 1986). As such, we expect nominees to exhibit less candor when questions involving these sorts of issues arise.Footnote 19 Thus within the family of QOV topics outlined above, we expect that questions targeting a nominee's views on civil liberties and civil rights issues are more likely to trigger evasive responses.Footnote 20
Here again, we find our expectations largely met. As Figure 6 demonstrates, the candor level of civil liberties questions is lower than for judicial philosophy questions in all but three of the 30 justices. Moreover, even when measured against the other QOV topics (grouped together in Figure 6 because none of them individually came up that frequently) civil liberties questions generated fewer forthcoming responses in nearly all of the hearings.
Figure 6. Nominee Candor about Their Views by Topic Area.
We note that during the late 1960s and early 1970s, there was a large decrease in candor on questions on civil rights and liberties. We also note that recently, there has once again been a widening discrepancy between forthcoming responses to civil liberties questions and judicial philosophy questions: candor in response to questions on civil rights and liberties has steadily declined since Kennedy (85 percent), reaching a low with Sotomayor (40 percent). This is interesting because candor in response to questions on judicial decision making has held fairly constant at around 80 percent for the same time span, dipping a bit lower during the Ginsburg hearings. Overall, therefore, it appears that while candid responses to civil liberties questions has varied over time, these types of questions tend to be associated with less forthcoming responses.
As with the first element of H2, therefore, the question then becomes whether it is possible that an increase in the prevalence of civil liberties questions has helped fuel the impression of nominee evasion. We explore this possibility simply by measuring the various percentages for the different topics of QOVs. As Figure 7 illustrates, there is no pronounced trend for any of the three most common question topics.Footnote 21
Figure 7. The Changing Type of Scrutiny over Time: The Topic of Viewpoint Questions.
There has been perhaps a slight increase in civil liberties questions over the past half-century, but it is not a marked change. Thus, unlike the previous element of H2, where there was fairly strong support for the idea that the increase in QOVs was making nominees seem less candid, there is less support for that idea here. Having said that, we note that the topics have oscillated over time, and that civil liberties questions make up a substantial portion of the full slate of QOV topics for all nominees. These two factors suggest that the perception of nominee evasiveness may be at least in some measure informed by less-than-forthcoming responses to civil liberties questions.
(3) Party and Ideological Distance Will Drive Down Candor
The third element of our second hypothesis involves party and ideological distance. Specifically, we expect that when nominees are asked questions by senators from the opposing party—i.e., the opposite party from the president who put up the nominee—their level of candor will decrease. Similarly, we expect that when nominees are asked questions by senators whose ideology is distant from their own, their candor level will decrease. Our thinking here is once again rooted in the literature establishing the ideological nature of the confirmation hearings. To the extent that the hearings represent an opportunity for senators to score political points by undermining a nominee with opposing views, the nominees will be on guard more against senators who they perceive to be “out to get them.” This, we believe, will cause nominees to be less forthcoming when asked a question by a “hostile” senator.
Figure 8 examines the percentage of “very forthcoming” responses to senators from the same and from the opposing party. The striking finding in Figure 8 is that there does not appear to be any significant divergence between the two lines, nor does there appear to be any growing divergence over time. For a short time, a small gap appeared for Breyer, Roberts, and Alito, with those nominees being more forthcoming to senators of the same party as the nominating president, but that gap completely disappeared during Sotomayor's and Kagan's hearings. A similar gap appeared earlier during the Fortas (AJ), Marshall, and Fortas (CJ) hearings, but it quickly disappeared as well. This suggests that the likelihood that nominees are to give a Fully Forthcoming response to a senator from the same party as the president compared to a senator from the opposing party has not changed much over time. In fact, nominees do not appear to treat senators much differently based on their partisan status, but appear more concerned with the institutional status differences. This is consistent with research by Reference SegalSegal (1987) who found institutional politics were at least as important as partisan politics.
Figure 8. Nominee Candor in Response to Senatorial Scrutiny by Party Type.
Figure 9 examines the percentage of Fully Forthcoming responses to senators who are either ideologically close or distant. To determine who is an ideologically close or distant senator, we use transformed Common Space ideology scores similar to Reference Epstein, Segal and WesterlandEpstein, Segal, and Westerland (2008), updated through the 111th Congress to include new senators and Sotomayor.Footnote 22 Once a Common Space ideology score was calculated for all senators and nominees that put them in the same ideological space, we calculated the (Euclidean) ideological distance between senators and respective nominees (similar to Reference Epstein, Segal and WesterlandEpstein et al. 2008). Using this distance measure, we divided the observations into two groups; those above the mean ideological distance split (distant senators), and those below the mean (close senators). The idea is that senators who are more similar ideologically to a nominee would ask questions that a nominee would find more compatible to give a Fully Forthcoming answer compared to an ideologically distant senator.
Figure 9. Nominee Candor in Response to Senatorial Scrutiny by Ideological Distance.
The results in Figure 9 are surprising for the lack of divergence. In fact, only two times do we see what could be described as a pattern of ideological divergence: Fortas and Marshall in the 1960s, and Roberts and Alito more recently. Thus, contrary to expectations, the ideologically close and ideologically distant senator lines actually mirror each other over time. This suggests that any change in candor over time does not appear to be driven by ideological differences.
Note, we are not arguing that partisanship and ideology are unimportant. In fact, nothing in our results suggests that ideology and partisanship are unimportant for confirmation hearings. We are only arguing that partisanship and ideology do not appear responsible for the small, over time change in candor that we observed in Figure 3.
Changes in Nominee Responses and Perceptions of Evasiveness
Thus far we have found support for the idea that nominees have not simply become more evasive over time, but rather that the kinds of questions that are now being asked more frequently—namely, questions about their views and beliefs—are more likely to generate evasive responses. In other words, the results of H1 and H2 suggest that observers of the hearings may perceive nominees to be less forthcoming not because the general approach of the nominees has changed, but because senators now ask tougher questions more often. If true, this version of events represents an obvious challenge to the conventional wisdom that the quality of the hearings has deteriorated markedly over the past few decades, and that it is the nominees who are to blame.
Prompted by these initial findings, we pursue this line of inquiry one step further by shifting our focus from the senators’ questions to the nominees’ responses. Specifically, we suspect that the ways in which nominees answer (or, more precisely, do not answer) questions may have changed, and that this may be having an effect on the public's notions about the hearings. That is, we believe that the impression of nominee evasiveness might be, at least in part, a result of nominees who are becoming increasingly explicit in the types of reasons they give for not answering Committee questions. Thus we hypothesize that (H3): The ways in which nominees decline to answer questions are now more direct and explicit.Footnote 23
Our third hypothesis flows directly from the extensive criticism, discussed earlier, of the so-called Ginsburg rule and its alleged effects on the quality of the hearings. Recall that the dominant view of recent hearings is that recent nominees have followed Justice Ginsburg's self-imposed “rule” of not answering questions about things that could come before the Court at some point. We think this criticism is quite revealing. Specifically, we believe it shows that the explicit way in which Ginsburg and those who have appeared after her have declined to answer questions has made a profound impression on those who watch the hearings. From this, we speculate that the use of this specific evasive tactic, the so-called Ginsburg rule, has increased over time.
To test this hypothesis, we tracked the three reasons most given by nominees for not being fully forthcoming: (1) the issue is currently before the Court or could be before the Court at some point; (2) the question raises general concerns of judicial independence; and (3) they lack sufficient knowledge or memory.Footnote 24 Again, it is the first reason that represents the explicit Ginsburg rule technique. The other two recorded reasons, while hardly benign, do not seem to have generated the type of backlash in the public sphere as the Ginsburg rule has. Thus our focus here is on whether the use of that first technique has increased in recent years.
The results in Figure 10 suggest support for our hypothesis. The “before the Court” excuse, that is, the Ginsburg rule, appears to have been on the rise since the early 1980s. This is noteworthy not only because it seems to confirm our expectations that this particular tactic has increased in recent years, but also because it actually appears to have started prior to its namesake, Justice Ginsburg. We note that the origin of this trend is more reliably traced to Justice O'Connor's hearings, which, perhaps not coincidentally, were the first proceedings to be televised.
Figure 10. Nominees’ Reasons for Being Evasive.
Two other findings strike us as significant: The first is that the judicial independence excuse is used less than are the other two excuses. It appears to have been used about 10 to 20 percent of the time by nominees Powell through O'Connor, but has been on the decline since then. Second, the most common explanation that nominees give for not answering a question is that they have either no knowledge or not enough knowledge to provide a response.
In sum, these findings support our third hypothesis that the reasons for declining to answer questions have changed. Nominees since the early 1980s have increasingly used an excuse that is more explicit and direct than are other types of excuses. Again, we hesitate to do more than simply speculate that this has fueled the impression of decreased nominee candor. But if the large body of criticism of the Ginsburg rule is any evidence, we feel confident in saying that the proliferation of this tactic has had at least some effect on those who watch the hearings.
Conclusions and Discusssion
Based on our analysis of nearly 11,000 exchanges between senators and nominees, we believe that we can now draw a series of conclusions about the Supreme Court confirmation hearings. Broadly speaking, our view is that the conventional wisdom—namely, the idea that nominees have become markedly more evasive since the late 1980s—needs to be rethought. In its place, we propose a different account of nominee performance over the years. We also propose an explanation for why there may be such a strong impression that the hearings have become less useful and informative in recent years.
We begin by acknowledging that some of what critics have observed about the recent hearings is accurate: there has indeed been a downward trend in nominee candor since the late 1980s. However, where most commentators have simply blamed this on the events surrounding the Bork hearings, we believe there is a better explanation for this latest modest decline.Footnote 25 Specifically, we think that what has actually occurred is that senators are now asking a higher percentage of questions about nominees’ views on controversial issues, and that this is what accounts for the recent downward trend. In other words, it is not that nominees since Bork have adopted a more evasive posture when it comes to the hearings, but rather that they are being asked more difficult questions more often.
At the same time, there are parts of the chorus of criticism that are simply at odds with the empirical record. Simply put, whatever nominee evasiveness has been observed in recent years is not nearly as dramatic as we are often led to believe. First and foremost, the overall level of candor has actually been fairly high, even since the 1980s, with most prospective justices answering between 60 and 70 percent of their questions in a fully forthcoming manner. Likewise, completely evasive answers make up only a small fraction of the responses for each nominee. Second, the downward trend itself is less steep than the criticism seems to suggest. Specifically, there has been a drop of roughly 20 percentage points from the high-water mark of Bork to the most recent level of Kagan. Needless to say, this is hardly a trivial change, but it also does not strike us as the kind of drop-off that should prompt Court watchers to announce the demise of the hearings. Lastly, the downward trend we have seen over the past 20 years is not unique. Other periods of decline in candor have been observed since 1955, suggesting to us that it is somewhat unfair to label the post-Bork nominees as unprecedented in their evasiveness.
Taken together, what these findings point to is that the perception of recent nominee evasiveness does not fairly reflect the reality of recent nominee performance. On its own, we think this is an important finding. However, we also believe that we can offer at least some reasons for why this gap between perception and reality might be so pronounced. One factor, of course, is television. It simply cannot be a coincidence that charges of evasiveness have picked up as viewership of the hearings has increased. Had commentators been watching early hearings as closely as those since the 1980s, we suspect they would have leveled charges against Justices Harlan, Brennan, and Goldberg the way they have against Justices Ginsburg, Roberts, and Kagan.
But beyond televised coverage, we believe that our analysis has uncovered another factor fueling the impression of nominee evasiveness. Specifically, nominees in recent years have become increasingly likely to invoke the so-called Ginsburg rule, stating that they do not want to talk about issues that could come before the Court. We think that this explicit invocation of judicial privilege—a trend that, despite its name, actually started with Justice O'Connor and not Justice Ginsburg—contributes the impression that nominees today are answering far fewer questions than they did in years past.
To summarize: our view is that Supreme Court confirmation hearings have been largely misunderstood by critics, especially in recent years. Rather than showing a simple story of post-Bork recalcitrance on the part of nominees, our findings reveal a much longer and more nuanced history of ebbs and flows in nominee candor. Moreover, we think that the recent downturn in candor is not nearly as dramatic as it is often described. We also think that while most of the blame has fallen on the shoulders of recent nominees themselves, the kinds of questions that senators ask today are actually more difficult and probing. And lastly, we find strong reasons to believe that the perception of nominee evasiveness has been fueled at least in part by the particular ways in which nominees avoid answering questions—most notably, by explicitly invoking the Ginsburg rule or something similar to it, rather than by using more subtle forms of evasion.
A number of important implications flow from these conclusions. Most notably, it appears that the confirmation hearings for Supreme Court justices have been marked by a certain degree of evasiveness since they began over a half-century ago. While we remain agnostic on the normative questions that attend this finding, we certainly recognize that anyone concerned about issues of accountability and transparency in the judicial nomination process would be troubled by this discovery. Given the tremendously important role that Supreme Court justices play in shaping our legal, social, and political landscape, it is certainly understandable that citizens would want more complete answers from the nominees. What our study can add to this debate is simply to acknowledge an awareness that if the hearings are “broken,” they have been that way from the start. Whether critics find this comforting or distressing remains to be seen.
We also recognize that there is quite a bit of additional work to be done with our data. The most pressing question, we believe, involves the extent to which party and ideology play a role in determining the candor of nominees. We addressed this in our current analysis and found little effect. However, our sense is that there is more to this part of the story than we have uncovered thus far. In particular, we suspect that when factors such as divided government and control of the Senate are factored in, we will be better able to predict how forthcoming a nominee is likely to be to Committee questioning. We have begun the task of developing this model, and hope to report results in the near future.
Until then, we are satisfied that our analysis has provided an important correction to the existing literature on Supreme Court confirmation hearings. Recent nominees may not be models of candor and openness, but they are hardly the first to avoid answering questions. Thus, while it may be true that the hearings are “vapid and hollow,” this is not by any means a recent development. Furthermore, identifying the problems that beset the hearings and exploring all of the factors that contribute to these problems may offer a first step in moving to solve them.
