How does the public evaluate Supreme Court nominees? The literature argues people consider party, ideology, and legal qualifications. But are they also capable of considering judicial philosophy? We believe they are. Our data show that when presented with information about judicial philosophies, the public uses that information to evaluate nominees.
The day after Chief Justice John Roberts swore in Justice Neil Gorsuch, Roberts made the following remark: “It is very difficult, I think, for a member of the public to look at what goes on in the confirmation hearings these days, which is a very sharp conflict in political terms between Democrats and Republicans, and not think that the person who comes out of that process must similarly share that same sort of partisan view of public issues and public life.”Footnote 1
His fear, of course, was that the public will view the Supreme Court wholly through the lens of partisanship or ideology. And who could blame them if they did? After Senate Democrats killed the filibuster for lower-court nominations, Republicans followed suit for Supreme Court nominations, giving the appearance that elites simply put ideologically simpatico nominees on the Court and that they need only care about the ideology and perceived partisanship of those nominees.
Still, we wonder: What do the American people consider when evaluating these nominees? Despite excellent scholarship on how the public evaluates Supreme Court nominees (Kastellec, Lax, and Phillips Reference Kastellec, Lax and Phillips2010; Bartels and Johnston Reference Bartels and Johnston2011; Hoekstra and LaRowe Reference Hoekstra and LaRowe2013; Collins and Ringhand Reference Collins and Ringhand2016; Sen Reference Sen2017; Chen and Bryan Reference Chen and Bryan2018; Rogowski and Stone Reference Rogowski and Stone2018), we still know nothing about how nominees’ perceived judicial philosophies influence the public’s evaluation of them. This is surprising since, as we show below, much of the media’s discussion over Supreme Court confirmations focuses on judicial philosophy.
Employing a cutting-edge conjoint experiment, we examine whether the public responds to characterizations of Supreme Court nominees’ judicial philosophies. The data reveal that while partisanship and ideology strongly influence their public support, nominees’ judicial philosophies also matter. The public supports nominees who are described as “originalists” (i.e., those who look to the intent of the framers to reach conclusions about the constitution’s meaning) and oppose nominees who purport to apply a living constitution approach (i.e., those who look to the evolving views and values of the American public when interpreting the Constitution). The data also uncover significant polarization between Democrats and Republicans over the use of these judicial philosophies. Finally, the data show that while judicial philosophy matters for Democrats and Republicans, it seems to motivate Republicans more.
Supreme Court Confirmation Politics
Politics have nearly always affected the federal judicial confirmation process (Epstein and Segal Reference Epstein and Segal2005). Just as the human brain has two interacting hemispheres, the confirmation process has always had both a legal hemisphere and a political hemisphere. For example, George Washington’s nomination of John Rutledge sank, in large part, because Federalists distrusted Rutledge’s opposition to the Jay Treaty (Unger Reference Unger2014). For political reasons, the Whig Senate killed Andrew Jackson’s nomination of Roger Taney to be an associate justice. Herbert Hoover failed to win over enough senators to confirm John Parker to the High Court, largely because of his ties to railroad companies (Maltese Reference Maltese1995). We could add to this list, but the point obtains: historically, politics have often influenced Supreme Court nominations and confirmations (Abraham Reference Abraham1992, Reference Abraham1999).
Yet even while acknowledging the ancient connection between politics and judicial appointments, there is no escape from the belief that today’s confirmation process is more uniformly rotten than ever. There seems to be a fairly general consensus that the modern ideology-heavy confirmation process is a “broken” (Cornyn Reference Cornyn2003) “mess” (Carter Reference Carter1994). One need look no further than the treatment Brett Kavanaugh and Neil Gorsuch received from some Senate Democrats or the treatment Elena Kagan and Sonia Sotomayor received from some Senate Republicans. Only three Democratic senators voted for Gorsuch. Only five Republicans voted for Kagan. Since the debacle of the Bork nomination, most nominees have gone through hypercharged confirmation battles. The parties’ recent changes to the filibuster have laid bare the brass-knuckle politics of our modern confirmation process.
For our purposes, though, we are more concerned with how nominee characteristics influence the public’s views of those nominees. Perhaps not surprisingly, empirical scholarship strongly supports the notion that partisanship and ideology influence the public’s views of Supreme Court nominees. Many scholars find that ideological proximity to the nominee or copartisanship with the appointing president are the leading causes of public support for a nominee (Hoekstra and LaRowe Reference Hoekstra and LaRowe2013; Badas and Stauffer Reference Badas and Stauffer2018; Engst, Gschwend, and Sternberg Reference Engst and Sternberg2018). Chen and Bryan (Reference Chen and Bryan2018) argue that the most important factor predicting public support for a nominee is ideological agreement with him or her. Bartels and Johnston (Reference Bartels and Johnston2011, 108) find that the public believes nominees “should be required to state their views on controversial issues” and that the president and the Senate should consider “how the nominee might vote on controversial issues.” Armaly (Reference Armaly2018) shows that Justice Scalia’s death—and President Obama’s consequent possibility to fill that vacancy—led to an increase in support for the Court among Democrats. These studies show that partisanship and ideology strongly motivate support for or opposition to High Court nominees.
Factors other than partisanship and ideology also influence the public’s support for nominees. The public appears to base some support for nominees on their legal qualifications. Hoekstra and LaRowe (Reference Hoekstra and LaRowe2013) find the public supports nominees who attend top law schools, receive high ratings from the American Bar Association, and are praised for their years of service as sitting judges. Kaslovsky, Rogowski, and Stone (Reference Kaslovsky, Rogowski and Stone2018) find evidence that legal experience and attending elite Ivy League schools bolsters public support. Perhaps as a consequence of the public’s support for legal qualifications, senators cast their confirmation votes, in part, on the basis of nominee qualifications (Epstein and Segal Reference Epstein and Segal2005; Epstein et al. Reference Epstein, Segal, Staudt and Lindstadt2005). Some scholarship also suggests the public bases its support for nominees on demographic characteristics. Badas and Stauffer (Reference Badas and Stauffer2018) conclude that shared identity increased support for Clarence Thomas among liberal African Americans, for Elena Kagan among conservative women, and for Sonia Sotomayor among conservative Latinos—all groups not otherwise disposed to support them but for their demographic ties. Kaslovsky et al. (Reference Kaslovsky, Rogowski and Stone2018) find similar effects in an experimental setting (but see Sen Reference Sen2017).Footnote 2 In short, ideology, partisanship, legal qualifications, and demographic considerations affect how the public evaluates High Court nominees.
But what about the nominees’ judicial philosophies? Judicial philosophy, after all, should be one of the most important factors the public uses to evaluate judicial nominees. And it is something that receives attention during the confirmation process. Yet, surprisingly, we know nothing about its role in public support for nominees.
Public Evaluations and Judicial Philosophy
Before we proceed, we address an important question: Do most Americans hold views about and preferences for judicial philosophies? Probably not initially—at least without some prompting or explanation. In public opinion research, a revealed opinion or preference is an approximation of a latent attitude. And an attitude is a “psychological tendency that is expressed by evaluating a particular entity with some degree of favor or disfavor” (Eagly and Chaiken Reference Eagly and Chaiken1993, 1). True attitudes tend to be stable and well defined. It seems hard to believe that Americans, who generally have little information about the Court, hold deep-seated, stable, and well-defined preferences about judicial philosophy. Indeed, some scholars question whether they have strong attitudes about anything at all (Zaller Reference Zaller1992; Converse Reference Converse2006).
Yet that cannot be the end of the story, for we know that Americans hold particular views toward particular institutions, and nowhere is this truer than for their views of the Court. Americans tend to trust the Court more than they do the other branches of government.Footnote 3 They evaluate judicial nominees on the basis of their perceived legal qualifications. (Just ask Harriet Miers.)Footnote 4 Americans recognize the need for an independent judiciary. They generally are unwilling to seek changes in the personnel and structure of the Court simply out of disagreement with its specific decisions (Gibson and Caldeira Reference Gibson and Caldeira1992). Indeed, Gibson and Nelson (Reference Gibson and Nelson2017) argue that diffuse support for the Court is a function of the public seeing the Court as unlike an ordinary political institution.
While it is clear that citizens recognize judicial decision-making involves some extralegal considerations, they nevertheless believe justices typically heed legal norms (Gibson and Caldeira Reference Gibson and Caldeira2011). For example, the public proffered more supportive views of the Court when they believed Chief Justice Roberts upheld the Affordable Care Act out of sincere legal reasoning and were less supportive when they “learned” that he voted strategically and contrary to his own personal ideology (Christenson and Glick Reference Christenson and Glick2015).Footnote 5 Thus, in multiple ways, Americans distinguish between the Court as a legal institution and other institutions. And so they may have general thoughts about proper judging (i.e., judicial philosophy) that can be activated. We believe that when presented with information about a nominee’s judicial philosophy, the public will use that information to inform their evaluations.
The next question, then, is whether the public actually receives information about nominees’ judicial philosophies. The answer is yes. Judicial philosophy is a major component of media coverage (Guliuzza, Reagan, and Barrett Reference Guliuzza, Reagan and Barrett1992; Comiskey Reference Comiskey1999; Ringhand and Collins Reference Ringhand and Collins2010). For example, in their exhaustive study, Ringhand and Collins (Reference Ringhand and Collins2010, 601) report that “three issues have dominated the substance of [confirmation] hearings: civil rights, judicial philosophy and criminal justice.” Ringhand and Collins estimate that judicial philosophy “has consistently been in a range of approximately 10% to 20% of hearing comments.” And that count probably underestimates the attention judicial philosophy receives in confirmation hearings, as Ringhand and Collins acknowledge (618).
News and polling organizations also tend to focus on nominees’ judicial philosophies. In particular, they often discuss originalism and living constitutionalism. (We describe these philosophies in the next section.) In September 1987, for example, the ABC/Washington Post and Gallup/Newsweek polls asked respondents whether they preferred a Supreme Court that “[sticks] closely to founding fathers’ intentions” or one that “makes decisions based on a modern interpretation of what the Constitution means.” From 2005 to 2018, at least 18 polls administered by seven major polling organizations asked similar questions regarding which of the two philosophies the public preferred.Footnote 6 The consistency with which this topic has appeared in surveys by major polling organizations is another indication of its relevance to the public. As the news media and others highlight judicial philosophy, they may prime the public to evaluate judicial nominees using those same criteria (e.g., Baumgartner, Morris, and Walth Reference Baumgartner, Morris and Walth2012).
Now, returning to the question we raised at the beginning of this section, we think it unlikely (although possible) that most of the public has deep-seated beliefs about judicial philosophies. Nevertheless, they do think the Court is unique, and they will, we believe, consider a nominee’s judicial philosophy when provided information about it. Whether they learn about these philosophies from the news, from copartisans, or from elite opinion drivers is outside the purview of this study. That is, we are agnostic as to whether people have ex ante beliefs about specific judicial philosophies or whether people react to information about those philosophies when provided to them. Our goal is simply to determine whether people evaluate nominees on the basis of their perceived judicial philosophies when provided information about them. Given their views that the Court is a unique institution, we believe they will. We expect members of the public—when provided with information about a nominee’s judicial philosophy—will evaluate a judicial nominee on the basis of judicial philosophy (as well as partisanship, ideology, and legal qualifications).
Three Judicial Philosophies
We focus on how the public evaluates nominees and their views on three general judicial philosophies: originalism, living constitutionalism, and stare decisis.Footnote 7 We focus our analysis on these three prominent judicial theories because they are the most systematically well known and discussed (see, e.g., Scalia Reference Scalia1997).
Originalism
We told some respondents that a nominee would “look to the intent of the Framers to reach conclusions about the meaning of the Constitution” (what could be considered original intent) and others that nominees “consider what the words in the Constitution meant at the time those words were crafted” (what could be considered original meaning). These theories argue that judges should interpret the law by looking at what those who crafted it intended. Justices rely on original intent because, supporters argue, it is the only neutral way of deciding cases and preventing justices from imposing their own preferences. Supporters also suggest it enhances stability in the law. Our times may change, but the past is done. And so we can look to the framers as a fixed point around which our constitutional decisions revolve. For example, in District of Columbia v. Heller, 554 U.S. 570 (2008), Justice Scalia determined the meaning of “arms” in the Second Amendment thusly: “The 18th-century meaning [of ‘arms’] is no different from the meaning today. The 1773 edition of Samuel Johnson’s dictionary defined ‘arms’ as ‘[w]eapons of offence, or armour of defence.’”Footnote 8 Opponents point out that determining framer intent is impossible. There were so many framers with different views that selecting one framer channels the outcome toward the direction sought. And so it may not be a neutral Sherpa guiding judges through confusing legal terrain.
Living Constitutionalism
We told some of our respondents that a nominee would “look to the evolving views and values of the American public when interpreting the Constitution.” Supporters of this judicial philosophy argue that it keeps the Court and the Constitution up with the times and therefore enhances their legitimacy. So, whereas the death sentence for burglary may not have been viewed as cruel and unusual at the time of the framing, many people today might see it as so. In Trop v. Dulles (1958), for example, the Court interpreted the Eighth Amendment’s “cruel and unusual” language not in terms understood at the time of its creation, but in a modern context, when it found that revoking citizenship as a punishment for a crime was unconstitutional. Opponents point out that the approach, while framed in terms of updating the Constitution in order to protect its legitimacy, actually disrespects the Constitution. The Constitution separates powers among the branches of government but favors the elected branches. It created the legislature, not the courts, to make policy. If the Constitution itself needs to be updated, people can do so through the amendment process. Otherwise, they say, judicial policy making is simply an illegitimate power grab by unelected judges.
Stare Decisis
We also told some of our respondents that nominees “interpret the Constitution based on precedents established in previous cases.” Chief Justice Roberts famously explained this approach during his Senate confirmation hearing, stating: “Judges are like umpires. Umpires don’t make the rules; they apply them” (New York Times 2005). This philosophy argues that judges should behave incrementally and follow the collective wisdom of previous judges. Supporters of this view argue that it binds judges to previous decisions and thereby enhances the rule of law through stability. Still, the approach has its detractors. First and foremost are empirical scholars who argue that judges, at least appellate court judges, do not actually adhere to this approach (Segal and Spaeth Reference Segal and Spaeth1996, Reference Segal and Spaeth2002). Others suggest that even if judges did adhere to a norm of precedent, the Supreme Court hears so many cases of first impression that existing case law will not be particularly useful. And, of course, there is the fact that the Supreme Court has control over its own agenda and can therefore select cases without controlling precedent. At any rate, following precedent is an ancient judicial philosophy to which many claim fidelity, and so we included it in our experiment. And while stare decisis is not quite a judicial philosophy in the same way as originalism or living constitutionalism, it is the most widely discussed aspect of judging, which led us to include it in the survey.
The descriptions we provided to our respondents are similar to the nominees’ actual statements. For example, on the night President Trump nominated him to the Court, Justice Kavanaugh stated: “My judicial philosophy is straightforward: A judge must be independent and must interpret the law, not make the law. A judge must interpret statutes as written, and a judge must interpret the Constitution as written, informed by history and tradition and precedent” (New York Times Reference Matthijsse, De Leeuw and Hox2018). Justice Gorsuch made similar remarks, stating: “in our legal order it is for Congress and not the courts to write new laws. It is the role of judges to apply, not alter, the work of the people’s representatives. A judge who likes every outcome he reaches is very likely a bad judge” (New York Times Reference Matthijsse and Hox2017). In short, these descriptions generally fit what people tend to see and hear about actual nominees.
On the whole, we have no expectation regarding which judicial philosophy the public will prefer generally. On this point we are admittedly atheoretical. We do expect, however, that individual preferences over judicial philosophy will be a function of one’s partisanship or ideology. Conservatives are likely to prefer judicial philosophies that emphasize a preference for the status quo. Originalism accomplishes that. Liberals are likely to desire judicial philosophies that allow for a loose construction of the law and more progressive interpretations. The living constitution approach accomplishes that.
It is possible, of course, that judicial philosophy acts also as an additional ideological or partisan signal. A declaration of adherence to originalism or living constitutionalism might signal to the public a seriousness regarding the nominee’s commitment to conservative or liberal judicial decision-making. Given the inconsistent voting behavior of some justices (e.g., Kennedy, O’Connor, and Souter), the public may seek additional signals that a nominee is truly committed to deciding cases in a particular fashion. Sending the signals of originalism or living constitutionalism might help assuage uncertain Court watchers that a nominee is the “real deal.”
The Conjoint Experiment
To determine whether, and to what extent, the public evaluates nominees on the basis of their perceived judicial philosophies, we executed a conjoint experiment. Because a conjoint experiment differs from traditional and more well-known experiments, it requires a bit of explanation. A traditional experiment isolates a causal effect by “treating” a randomly assigned subset of individuals and then comparing outcomes between the treated and untreated groups. One significant weakness of this traditional approach, however, is that testing for the independent and interactive effects of additional variables is time-consuming, requires a large number of participants, and is therefore costly.
A conjoint experiment circumvents these limitations. In a conjoint experiment, the researcher can present respondents with multiple items to evaluate and thus increase sample size in a cost-effective, yet appropriate, manner. The conjoint allows researchers to test the causal effects of a number of item attributes simultaneously. An attribute is a characteristic of the item being evaluated. In our context, it is a characteristic of a nominee. Each attribute has a predetermined number of levels. For example, the levels of the attribute sex are “male” and “female.” The levels of the attribute religion of the nominee could be “Christian,” “Jewish,” or “not religious.” Any number of attributes or characteristics may be specified.Footnote 9 (We provide an example of the experiment shortly.)
Another benefit is that the conjoint experiment allows us to limit the impact of social desirability bias. In conjoints, respondents encounter profiles that consist of multiple attribute levels. That is, researchers need not vary only one single characteristic to measure a causal effect. Like list experiments (see, e.g., Blair and Imai Reference Blair and Imai2012), conjoints provide individual respondents with multiple rationales for selecting a candidate, some of which may be based on a socially undesirable characteristic. So, respondents can reveal their true preferences without tipping their hands, and we can uncover them. Traditional experiments may bias results by emphasizing a specific attribute and deemphasizing others. The complete randomization of attribute levels for each profile evaluation in the conjoint makes it unlikely that respondents will focus on one particular attribute in their evaluations.
Although conjoint experiments first became popular in market research, they are becoming popular in political science. For example, Hainmueller, Hopkins, and Yamamoto (Reference Hainmueller, Hopkins and Yamamoto2013) showcase the value of conjoint experiments for questions in which political scientists are interested. Their study presented survey respondents with eight side-by-side descriptions of two would-be presidential candidates and asked the respondents to evaluate the candidates. The experiment allowed the authors to test the effects of various descriptors on the probability a respondent voted for one candidate over another. (They also performed a similar experiment in which respondents chose to admit one of two potential immigrants into the country.) Similarly, Ono and Burden (Reference Ono and Burden2017) investigate whether voters use a candidate’s sex as a voting heuristic. They presented survey respondents with side-by-side profiles of candidates for elected office. The profiles included personal, party, issue, and polling information. The experiment allowed them to test the simultaneous impact of these traits and, most importantly, uncover the causal effect of sex when presented alongside other common information. Sen (Reference Sen2017) used a conjoint experiment to analyze the impact of copartisanship on evaluations of Supreme Court nominees.
To perform our conjoint experiment, we surveyed 1,000 adult respondents in April and May 2018 (before the Kennedy vacancy), using YouGov.Footnote 10 We weighted survey responses by matching respondents’ 2016 presidential vote choice, gender, age, race, and education with the US general population (according to the American Community Survey). We provided each of our 1,000 respondents five pairs of possible (but fictional) nominees with randomly assigned attributes.Footnote 11 The attributes included the nominees’ judicial philosophy, age, race, sex, professional background, religion, law school, ideology, and American Bar Association rating. We also informed respondents of whether the nominee ever held elective office and the party of the president who nominated him or her. For each pair of nominees presented, respondents indicated which nominee they most preferred to see on the Court. They also rated each nominee on a five-point scale. (Our empirical results are substantively similar regardless of whether we analyze respondents’ discrete choice or their relative rankings of nominees.)Footnote 12
We offered respondents a pair of nominees to evaluate—rather than a single nominee at a time—for two reasons. First, people tend to make decisions with a reference point in mind (Kahneman and Tversky Reference Kahneman and Tversky1979). They tend to make decisions relative to other possibilities. Asking them to rate nominees while looking at other possible nominees allows us to take advantage of that natural human condition. Second, it is common for news outlets to compare nominees with the justices whom they replace. For example, the New York Times posted 155 articles discussing Brett Kavanaugh within 30 days of his nomination. Of those stories, 49.7% mentioned the justice he replaced (Justice Kennedy). Similarly, the Wall Street Journal did so 42.5% of the time. Thus, readers of these popular news outlets were, at the very least, implicitly encouraged to evaluate the nominee with another person in mind. Our experiment incorporates this reality.
Of course, forcing people to choose between two alternatives is not new to scholarship. Forcing the choice often reveals useful information about the respondent. Indeed, since 1992, the American National Elections Study has asked respondents to choose qualities they believe children should have. The qualities usually are a forced choice between independence and respect for elders, obedience versus self-reliance, curiosity versus good manners, and being considerate or being well behaved. The answers people give are thought to reveal their latent authoritarianism. At any rate, Hainmueller et al. (Reference Hainmueller, Hopkins and Yamamoto2013) argue that choice-based conjoint analysis is more common than rating-based analysis.
Our primary interest is whether respondents evaluate nominees on the basis of their judicial philosophies. To that end, we asked respondents to choose among nominees who espoused various judicial philosophies, including the three prominent philosophies we described above: originalism, living constitutionalism, and stare decisis. We also accounted for other features that might lead members of the public to support or oppose Supreme Court nominees. We included the party of the nominating president (Democratic or Republican) and the ideology of the nominee (liberal, moderate, or conservative). In order to increase the realism of the experiment and align it with the current political environment, Republican presidents nominated only moderates or conservatives. Democratic presidents nominated only moderates or liberals.
We employed various measures for the nominees’ legal qualifications. We varied the nominees’ legal education to account for whether they received their law degrees from an elite law school (Yale, Stanford, Harvard, Chicago, Columbia, or Cornell) or a nonelite law school (Vanderbilt, Iowa, Notre Dame, Arizona State, George Washington, North Carolina, Texas, Wisconsin, or Connecticut). Using weighted randomization, we randomly selected the university appearing in a profile, with higher-ranked universities appearing more often than lower-ranked universities.Footnote 13 We included the American Bar Association rating as not qualified, qualified, or well qualified. Weighted randomization favored more qualified nominees.
We also include nominee demographic characteristics. We varied the nominees’ age from 40 to 65 years old. We also varied the nominees’ sex (male or female). We varied the nominees’ race to take on one of the following options: White, Black, Asian, Native American, or Hispanic. Considering the prevalence of nominees and current judges who are white, race was assigned a 37.5% chance of being White, 25% chance of being Black, 12.5% chance of being Asian, 12.5% change of being Native American, and a 12.5% of being Hispanic. We next varied the nominees’ religion, with nominees assigned equal probabilities of being Catholic, Lutheran, Methodist, nondenominational Christian, Jewish, or not religious.
We further accounted for whether the nominee ever held elected office. Nominees were more likely not to have held elected office in the past (66% probability). We also controlled for the nominees’ current position. Nominees were equally likely to be an attorney in private practice, an attorney for the United States, a state supreme court justice, a federal trial court judge, a federal appellate court judge, or a law professor.
As table 1 shows, respondents might choose between, on the one hand, a Native American male government attorney who believes in originalism and a Black female circuit court judge who believes in precedent, on the other. Another respondent might choose between a White female state supreme court justice who believes in a living constitution and a Hispanic male law professor who employs originalism. By varying these attributes across a large sample of respondents, the conjoint experiment allows us to determine which attributes respondents most prefer.
Results
To measure the causal effects of attribute levels in our conjoint experiment, we calculate average marginal component effects (AMCEs; Hainmueller et al. Reference Hainmueller, Hopkins and Yamamoto2013).Footnote 14 AMCEs are straightforward to interpret. They represent the average causal effect of the attribute on the probability the respondent chooses a particular nominee. So, for example, an AMCE of 0.15 for the female level of the sex attribute would suggest that a respondent has a 0.15 greater likelihood of selecting a female nominee than a male nominee (the baseline). In other words, AMCEs reflect the change in probability of voting for a nominee caused by the indicated level of an attribute compared to its baseline. We weight respondents such that the results reflect the average effect of the attribute on a sample that is representative of the US population.
Results from All Respondents
When provided with information about judicial philosophy, does that information independently influence how the general public evaluates Supreme Court nominees? The results say yes. Figure 1 shows the AMCEs for the entire population in our sample, and it reveals that when provided information about judicial philosophies, the general public evaluates nominees on the basis of those philosophies.Footnote 15 More specifically, figure 1 shows the change in the probability of voting for a nominee (X-axis) with the associated attribute level on the Y-axis. Horizontal lines represent 95% confidence intervals. The data show that the general public appears to prefer originalism over the living constitution approach. Respondents are 5 percentage points more likely to support an originalist than a living constitutionalist nominee. (It is worth pointing out that judicial philosophy has an effect on nominee support that is comparable to the effects of nominee ideology or qualifications.) This suggests to us that in the realm of “selling” judicial philosophies, originalism seems to have won out, at least for the time being.
AMCEs of profile attributes of nominees presented to all survey participants. The estimate represents the change in the probability of voting for a nominee with the associated attribute level. Horizontal lines represent 95% confidence intervals. Positive values mean respondents preferred the first attribute to the second. Negative values mean they preferred the second attribute to the first.
Some may find it surprising that the general public has preferences over judicial philosophies. After all, as we pointed out above, scholarship suggests the public generally has a low level of political knowledge (Converse Reference Converse and Apter1964; Delli Carpini and Keeter Reference Delli Carpini and Keeter1996; Somin Reference Somin2006). But our findings suggest that respondents were able to pick up nuances in the judicial philosophies we presented them. Indeed, looking deeper into the respondents’ backgrounds reveals that judicial philosophies influence respondents’ views regardless of their education levels. Those with only a high school education were just as attentive to the judicial philosophy frame as were the college educated and those with postgraduate degrees.
It could be, of course, that people do not actually prefer originalism to a living constitution approach in the application of the law. Perhaps respondents simply “like the sound” of originalism better. A person who knows nothing about judicial philosophy might prefer one that it is rooted in the framing of the Constitution more than one rooted in flexible and changing contemporary standards.Footnote 16 We have not tested how respondents react to the application of judicial philosophies in particular cases and therefore cannot opine. Nevertheless, even if the results are driven simply by a preference for the framing of originalism, that still strikes us as important for understanding how people support or oppose High Court nominees. It means there is a built-in preference for originalism. And it offers some explanation why some very conservative nominees may have made it through the Senate and why the moderately conservative (and somewhat originalist) Court retains substantial public support.
Results for some of our other variables reveal that the general public prefers moderates to both liberal and conservative nominees. More specifically, the public is approximately 4 percentage points more likely to support a moderate nominee compared to a conservative or liberal alternative. Not surprisingly, the majority of Americans prefer moderation. Having controlled for ideology, the party of the appointing president had no influence. Thankfully, respondents are much more likely to support qualified nominees than unqualified nominees.
Respondents by Party
Whereas figure 1 shows how all Americans as a group evaluate Supreme Court nominees, figure 2 breaks down the results by respondent party ID.Footnote 17 It examines how Democrats, Independents, and Republicans evaluate originalism, living constitutionalism, and stare decisis. The squares in figure 2 represent Democratic respondents, the circles represent Independent respondents, and the triangles represent Republican respondents. Horizontal lines represent the 95% confidence intervals. What should be clear is that Democrats, Independents, and Republicans all differentiate among judicial philosophies, and they do so even while controlling for ideological and partisan factors.
AMCEs of profile attributes of nominees presented to Democratic, Independent, and Republican survey participants. The estimate represents the change in the probability of voting for a nominee with the associated judicial philosophy. Horizontal lines represent 95% confidence intervals. Positive values mean respondents preferred the first attribute to the second. Negative values mean they preferred the second attribute to the first.
Democrats, however, split with Republicans and Independents over judicial philosophy. The AMCE on originalism is negative and significant for Democrats (
), meaning they prefer nominees who espouse a living constitution approach over nominees who espouse originalism. For Republicans (
) and Independents (
), the AMCE on originalism is positive and significant, showing that they prefer originalism to a living constitution.Footnote 18 Indeed, comparing the AMCEs on originalism versus partisanship and ideology reveals that Republicans consider philosophy to be as important as partisanship and ideology. Among Democrats, the effects are nearly but not quite as strong.
Figure 2 also confirms existing understanding of the public’s evaluation of nominees. Ideology and partisanship mattered to our respondents. Republican respondents were more likely to prefer Republican nominees over Democratic nominees. They were more likely to prefer moderates and conservatives to liberals. Democrats were slightly more likely to prefer liberals to moderates (
) and always more likely to prefer liberals to conservatives. Looking at the conditional results (shown in the appendix) revealed that Democratic respondents preferred female nominees over male nominees. Republican respondents preferred religious nominees over irreligious nominees. Neither Republicans nor Democrats preferred any nominees on the basis of race.
One might wonder whether support for these philosophies is just cue-taking from party leaders. That is, perhaps conservatives and Republicans are used to hearing about the virtues of originalism from their tribe, while liberals and Democrats are used to hearing about the living constitution from their tribe. This is entirely possible. Still, three things lead us to believe the story is more complicated than that. First, we receive the results we do even while controlling for ideology and partisanship. So, the effects of philosophy are shorn of some of that possible linkage. Second, the careful reader will notice we never used the words “originalism” or “living constitution” with our survey respondents. We explained each philosophy without using buzzwords that might instantly link the philosophy to partisanship. This suggests to us that respondents were able to take our descriptions and apply them to their own views. Third, Independents (i.e., those without partisan leanings) also had definite views of judicial philosophies, which leads us to believe it is not simply cue-taking by partisans, for even the least partisan respondents displayed telling effects.
The Relative Importance of Judicial Philosophy
So far, we have discovered that the American public has preferences over judicial philosophy—at least when told about those philosophies. We also discovered that originalism’s aggregate support came from Republicans and Independents. Both Republicans and Independents prefer originalism while Democrats prefer a living constitution approach. We now dig deeper to find out just how important judicial philosophy was to our respondents.
We estimate a logistic regression model to calculate the conditions under which respondents selected hypothetical nominees. Our dependent variable, select nominee, equals 1 when the respondent selected the profile for a particular nominee and 0 when the respondent did not select that nominee profile. We interact respondent party identification, party of the appointing president, and judicial philosophy in a three-way interaction. We also include American Bar Association ratings and ideology as controls in the model. We exclude from the analysis respondents who identified as either “not sure” or “other” when asked about their own partisan identification.
Because of the well-known difficulty of interpreting coefficient estimates in a three-way interaction, we display predicted probabilities. We separately calculate the probability that Democratic and Republican respondents select a nominee appointed by a Democratic or Republican president. We calculate these probabilities for each of the three judicial philosophies mentioned. (We held American Bar Association rating at well qualified and ideology at moderate when making these calculations.)
Democratic Respondents
Figure 3 displays the predictions among Democratic respondents. It shows, first, that partisanship strongly influences how Democratic respondents evaluate judicial nominees. Consider the comparison between the first and fourth plots, which focus on a living constitution approach, applied by Democratic nominees (plot 1) and Republican nominees (plot 4). The Democratic respondent has a .78 probability of supporting a Democratic nominee who espouses a living constitutionalism philosophy but a .47 probability of supporting a Republican nominee who espouses it. This 31-point difference is attributable to partisan differences. Next, consider the comparison between the third and sixth plots, which focus on originalism. A Democratic respondent has a .66 probability of selecting a Democratic nominee who espouses originalism but a .44 probability of selecting a Republican nominee who espouses originalism. This 22-point difference is also attributable to partisan differences. Put simply, partisanship is a strong predictor of support among Democratic respondents. When holding judicial philosophy constant, they are always more likely to support the Democratic nominee over the Republican nominee.
Predicted probabilities of a Democratic respondent supporting Democratic and Republican appointed nominees who espouse a living constitution, stare decisis (precedent), or originalist judicial philosophy. Bars represent 95% confidence intervals. Results obtained via logistic regression. For predictions, characteristics were held as follows: ABA rating—well qualified; ideology—moderate.
But the data also show that Democratic respondents evaluate nominees based, in part, on judicial philosophy. Consider plots one and three in figure 3. When we hold the nominee’s partisanship constant (at Democratic) but vary his or her judicial philosophy, we observe the effects of judicial philosophy. A Democratic respondent has a .78 probability of supporting a Democratic nominee who espouses a living constitution philosophy but a .66 probability of supporting a Democratic nominee who espouses originalism. This 12-point difference is attributable to judicial philosophy—and the Democratic respondents’ dislike of originalism. On the other hand, when evaluating Republican nominees (plots 4 and 6), Democrats do not distinguish among judicial philosophies. They “punish” Republican nominees equally. This suggests that a Republican president facing a Democratic senate would be unlikely to pick up public support by nominating a living constitutionalist judge.
Republican Respondents
Figure 4 displays the predictions among Republican respondents. Not surprisingly, partisanship also matters to Republicans. Consider the comparison between the first and fourth plots, which focus on a living constitution approach. A Republican respondent has a .36 probability of selecting a Democratic nominee when that nominee espouses a living constitution. But that same Republican respondent has a .68 probability of selecting a Republican nominee who believes in a living constitution. This 32-point difference reflects partisanship. We retrieve similar results when we focus on originalism. The Republican respondent has a .54 probability of supporting a Democratic nominee who espouses originalism but a .84 probability of supporting a Republican nominee who espouses it. This 30-point difference reveals that among Republicans, partisanship is a strong predictor of support.
Predicted probabilities of a Republican respondent supporting Democratic and Republican appointed nominees who espouse a living constitution, stare decisis (precedent), or originalist judicial philosophy. Bars represent 95% confidence intervals. Results obtained vis-à-vis logistic regression. For predictions, characteristics were held as follows: ABA rating—well qualified; ideology—moderate.
Judicial philosophy, however, also has a substantive effect on Republicans—and its effect is considerably greater than on Democrats. Consider plots one and three in figure 4. When we hold the nominee’s partisanship to be Democratic but vary his or her judicial philosophy, we observe the effects of judicial philosophy. A Republican respondent has a .36 probability of supporting a Democratic nominee who espouses a living constitution philosophy but a .54 probability of supporting a Democratic nominee who espouses originalism. This 18-point increase results from the nominee’s judicial philosophy. Looking at Republican nominees (plots 4 and 6) also reveals a strong effect. A Republican respondent has a .68 probability of supporting a Republican nominee who espouses living constitutionalism but a .84 probability of supporting a Republican nominee who espouses originalism. This 16-point difference again speaks to the effect of judicial philosophy on Republican respondents. Unlike Democratic respondents, Republican respondents strongly respond to variation in judicial philosophy whether the nominee is a Democrat or a Republican. This suggests to us that a Democratic president facing a Republican Senate might gain votes (and public support for a nominee) by nominating an originalist.
Conclusion
Supreme Court nominations in recent years have become increasingly polarized. They emphasize ideology and partisanship more than they did in the past. Chief Justice Roberts’s comments in the introduction underscore the point. We have shown, though, that the public can also evaluate nominees through the lens of judicial philosophy. When provided summaries of nominees’ judicial philosophies, the public can differentiate between originalism and living constitutionalism (as well as stare decisis). And the effects on their support are notable. While the general public prefers originalism to living constitutionalism, the data revealed significant polarization between Democrats and Republicans in their views. And even though respondents of both parties used judicial philosophy, it appears to have a larger effect on Republicans than on Democrats.
To be sure, these results are not without their limitations. As we stated above, we provided our respondents brief descriptions of the three philosophies and then gauged their responses. It could be that after learning in depth about the philosophies, they become more or less enamored of them. Perhaps a couple of historical examples might change their outlook. While this is possible, we hasten to point out that the descriptions we offered are in line with what they tend to hear from the nominees and from the media, providing external validity to the study.
Similarly, we wish to be clear that we believe judicial philosophy, while plainly something respondents considered in our study, is not the primary factor respondents use to evaluate nominees. Partisanship and ideology remain king. And, while these philosophies obviously have some partisan connections, the models reveal effects from philosophy that were statistically independent of partisanship and ideology. What is more, even if judicial philosophy is some kind of residual partisan or ideological signaling, our data suggest it would be a multiplier effect still worth studying.
Our results can inform current debates regarding the nomination and confirmation process. While senators appear to prefer a scorched-earth policy in their efforts to either confirm or oppose nominees, we find that the public, at least in the abstract and when so prompted, can evaluate nominees using an arguably reasonable set of criteria. Although partisanship and ideology matter to the public, so do qualifications and judicial philosophy. The more responsive the nomination and confirmation process is to the public’s (and not factional) preferences, the more likely we will see a fairer and even-tempered confirmation process.
Our claim is not that the public holds deeply rooted beliefs about judicial philosophy (although some members of the public surely do). Rather, we argue that when made aware of a nominee’s philosophy, the public uses that information to evaluate nominees. This information is often available to the public—as we discovered in Justice Kavanaugh’s nomination announcement remarks and those of Justice Gorsuch. There may be times when nominees want to highlight their philosophies and other times when they want to hide them. Divided government may lead to such behavior. These are empirical questions worth pursuing. For the time being, though, our data show that when given brief (and realistic) descriptions of judicial philosophy, respondents use the information to evaluate nominees.
