Obstruction and the Supreme Court confirmation process

Tools of obstruction provide senators with the capacity to impress their preferences upon the legislative process. Following existing scholarship, I conceptualize of obstruction as a range of tactics that seek to delay or entirely block legislative business (Fong and Krehbiel Reference Fong and Krehbiel2018; Patty Reference Patty2016). Obstruction in the context of confirmation proceedings is implemented through many forms and can be employed by both majority and minority party members. Historically, most prominent among these tools was the filibuster, implemented either via explicit action or a threat (Koger Reference Koger2010). The increasing use of the filibuster to obstruct judicial nominations ultimately came to a head when Senate Democrats exercised the “nuclear option” that removed the filibuster in 2013 for lower court nominees, followed by Senate Republicans for Supreme Court nominees in 2017.Footnote ⁵ Apart from the filibuster, senators of both parties have a number of other tactics by which they can exercise delay. Members of the Judiciary Committee are empowered to place a hold on a committee nomination vote and can also call for documents or records from the nominee’s past to be released and considered by the committee.Footnote ⁶ The blue slip provides a home-state senator with an opportunity to delay or altogether veto a lower court nomination (Binder and Maltzman Reference Binder and Maltzman2002; Binder Reference Binder2018). Legislators may also pair these tactics with public statements related to obstruction or delay to signal to their constituencies, donors, or interest groups about their obstruction activity (Patty Reference Patty2016).Footnote ⁷

Majority party senators are empowered with significant control over the chamber’s agenda and can employ this power to obstruct a judicial nominee’s path through the chamber. The majority’s agenda-setting power can ensure that the votes on judicial nominees that are held advance the will of the party. The majority leader and the chair of the Judiciary Committee set the agenda for nominee hearings and votes inside the committee and on the floor. The committee chair has control over calling for subpoenas of documents or records and decides the way in which the blue slip will be employed under their tenure (Binder Reference Binder2018).

Public attitudes toward obstruction and delay

There is reason to expect that the public learns about and forms attitudes toward the obstruction of nominees to the Supreme Court. Existing scholarship shows that the public learns about and critically evaluates how members of Congress behave. Ansolabehere and Kuriwaki (Reference Ansolabehere and Kuriwaki2022) show that public evaluations of members of Congress are shaped by perceptions of their voting on important legislation. Both a member’s own depiction of their behavior (Grimmer, Westwood, and Messing Reference Grimmer, Westwood and Messing2014) and media coverage (Moskowitz Reference Moskowitz2021; Snyder Jr and Strömberg Reference Snyder and Strömberg2010) shape constituency knowledge of and attitudes toward representatives. The public has preferences over the ins-and-outs of lawmaking and congressional procedures, including the filibuster, “hardball” tactics, and bipartisanship (Curry Reference Curry2019; Doherty Reference Doherty2015; Flynn and Harbridge Reference Flynn and Harbridge2016), and public opinion responds to high-profile within-chamber events, including investigations (Kriner and Schickler Reference Kriner and Schickler2014).

More specific to the judiciary, a long line of scholarship illustrates how the American public is aware of and has well-formed attitudes toward Supreme Court nominees and major decisions. These attitudes can be explained by political preferences, support for democratic values, and extant support for the judiciary (Bartels and Johnston Reference Bartels and Johnston2013; Caldeira and Gibson Reference Caldeira and Gibson1992; Gibson and Caldeira Reference Gibson and Caldeira2009b). These attitudes can be reasonably sophisticated in nature; for example, Krewson and Owens (Reference Krewson and Owens2021) show that support for judicial philosophies shapes evaluations of judicial nominees. These attitudes are consequential; electoral support for members of Congress depends on how they handle Supreme Court nominations (Badas and Simas Reference Badas and Simas2022; Bass, Cameron, and Kastellec Reference Bass, Cameron and Kastellec2022) and reform (Davis and Hitt Reference Davis and Hitt2024) and support for individual cases and nominees can shape evaluations of Court legitimacy (Christenson and Glick Reference Christenson and Glick2015, Rogowski and Stone Reference Rogowski and Stone2021). The media and prominent politicians likely shape what Americans know and think about obstruction of Court nominees, much as they do for other aspects of the judiciary (e.g., Hitt and Searles Reference Hitt and Searles2019; Ramirez Reference Ramirez2008). In Appendix Section A, I present data from two prominent sources of political information for Americans – senators and the news media – that illustrate how public-facing discussion of the nomination and confirmation process routinely covers obstruction; this is especially true for contexts where obstruction was a major component of the process. Tactics of obstruction tend to be discussed in detail and framed in terms of their political implications. As one illustration, the PBS NewsHour broadcast from the day of Neil Gorsuch’s nomination devoted significant attention to the political implications of the filibuster, with discussion of “the strategy from the Democrats’ side” and how Democrats “don’t have the votes to defeat [Gorsuch] outright” and were considering the filibuster, which may result in Republicans “abolish[ing] the filibuster rule for Supreme Court nominees.”Footnote ⁸ These data help illustrate how the American public is exposed to and has opportunities to form attitudes toward obstruction in the context of Court nominations.

Arguments for general opposition to obstruction

While it is likely that the American public holds attitudes toward obstruction of the Supreme Court confirmation process, existing theoretical perspectives and empirical evidence generate competing predictions over the nature of these attitudes. Components of this scholarship suggest that Americans should be generally unsupportive of the use of obstruction in this context.

A prominent perspective in the study of mass attitudes toward the courts holds that the American public thinks about judges through an apolitical lens, and suggests that the public will expect the nomination and confirmation process to unfold according to norms of procedural fairness. The American public thinks of judges as “principled” and distinct from “ordinary” politicians (Gibson and Caldeira Reference Gibson and Caldeira2011), evaluates nominees based upon legal criteria (Gibson and Caldeira Reference Gibson and Caldeira2009b; Krewson and Owens Reference Krewson and Owens2021; Krewson and Schroedel Reference Krewson and Schroedel2020), and expects judges to follow procedurally fair decision-making processes (Baird and Gangl Reference Baird and Gangl2006; Ramirez Reference Ramirez2008; Simon and Scurich Reference Simon and Scurich2011). In light of this scholarship, it is plausible that the public carries over these views to develop expectations that the Senate conducts the confirmation process by giving nominees full, apolitical procedural consideration. Indeed, Bartels and Johnston (Reference Bartels and Johnston2012) show a majority of the public prefers that the Senate vote without considering a nominee’s political positions, while Badas and Simas (Reference Badas and Simas2022) illustrate how primary voters punish copartisan Senate candidates who prioritize blocking politically distant Court nominees.

Studies of public attitudes toward Congress and congressional process provide further reason to expect that Americans will be unsupportive of obstruction and delay of Court nominees. The public has largely negative views toward Congress, driven in large part by a dissatisfaction with congressional procedure and internal congressional behavior (Hibbing and Theiss-Morse Reference Hibbing and Theiss-Morse1995). Americans view Congress as inefficient and unresponsive, dislike inter-party bickering, and express lower evaluations of Congress after exposure to such behavior (Brady and Theriault Reference Brady, Theriault, Hibbing, Theiss-Morse and Kuklinski2001; Hibbing and Theiss-Morse Reference Hibbing and Theiss-Morse1995, Reference Hibbing and Theiss-Morse2002). Further, Curry (Reference Curry2019) and Flynn and Harbridge (Reference Flynn and Harbridge2016) show that Americans dislike “hardball” procedural tactics and reward taking legislative action on bipartisan “consensus” issues. Finally, evidence from Doherty (Reference Doherty2015) suggests voters disapprove of the use of the filibuster. Taken together, these literature suggest that it is unlikely that Americans will express support for the use of obstruction in the judicial confirmation process.

Arguments for polarized attitudes toward obstruction

Contrasting perspectives suggest that individuals should polarize along partisan lines in their attitudes toward the use of obstruction for Court nominees. A growing body of research suggests that politics shapes public evaluations of the judiciary and judicial nominees. Bartels and Johnston (Reference Bartels and Johnston2013) and Christenson and Glick (Reference Christenson and Glick2015) provide evidence that satisfaction with the ideological outputs of the Court is associated with evaluations of the Court’s legitimacy; Glick (Reference Glick2023) provides similar evidence from the context of the Ginsburg vacancy. Politicians and the news media serve as a source of information that generates these partisan evaluations. Media framing and partisan cues engender politically motivated evaluations of Court decisions (Hitt and Searles Reference Hitt and Searles2019), nominees (Sen Reference Sen2017), and the institution (Rogowski and Stone Reference Rogowski and Stone2021). Furthermore, the public values and expects to learn about the political leanings of judicial nominees (Bartels and Johnston Reference Bartels and Johnston2012) and rewards senators for voting for co-ideological nominees (Badas and Simas Reference Badas and Simas2022; Bass, Cameron, and Kastellec Reference Bass, Cameron and Kastellec2022). Finally, support for reforming the judiciary, including procedures such as court packing, is also associated with an individual’s personal politics (Badas Reference Badas2019; Bartels and Johnston Reference Bartels and Johnston2020; Clark and Kastellec Reference Clark and Kastellec2015).

Related research from the study of Congress suggests that Americans realize the political implications of the use of obstruction and will polarize in attitudes towards its use. Smith and Park (Reference Smith and Park2013) find that partisans diverge in their attitudes over the use of the filibuster as the policy consequences of its use become clearer.Footnote ⁹ Evidence from Flynn and Harbridge (Reference Flynn and Harbridge2016) suggests that partisans will reward their party’s representatives in Congress if they employ tools of delay and obstruction to secure an ideological victory. Finally, Patty (Reference Patty2016) shows that obstruction can serve as a valuable signal to voters about a legislator’s skill as a policymaker. Taken together, this scholarship thus predicts that attitudes toward obstruction of Court nominees should split across partisan lines.

Variation by type of obstruction

There are many tools that can be employed to obstruct or delay a Supreme Court nominee’s path through the Senate. Existing scholarship offers reasons to expect that the public will support the use of these tactics at varying rates, depending on how well they fit into their expectations over how the confirmation process ought to proceed. Forms of obstruction that preclude [don’t preclude] procedural consideration of nominees should [should not] violate these expectations.

The public holds legalistic and procedurally fair expectations over how judges behave (Baird and Gangl Reference Baird and Gangl2006; Gibson and Caldeira Reference Gibson and Caldeira2011), and may extend these expectations to evaluating the different types of procedural tactics the Senate employs to evaluate nominees. Similarly, the public dislikes partisan conflict that leads to bickering or gridlock (Hibbing and Theiss-Morse Reference Hibbing and Theiss-Morse2002; Flynn and Harbridge Reference Flynn and Harbridge2016), and thus may prefer the Senate avoid these behaviors when evaluating nominees. This implies that the public will expect that nominees are given fair and full procedural treatment, are not treated like “ordinary” politicians, and are not subjected to overtly partisan tactics. This would render more extreme instances of obstruction that violate these expectations (such as refusing to hold any hearings on a nominee) unacceptable in ways that more routine and relatively apolitical procedures (such as calls for additional documents) or more consequential tools that still provide nominees with at least some procedural consideration (such as the filibuster) might not. Further, to the extent that forms of obstruction such as document requests are viewed as a way to secure greater information about a nominee’s legal qualifications, these tactics may be valued by a public that cares about a nominee’s legal credentials (Krewson and Schroedel Reference Krewson and Schroedel2020).Footnote ¹⁰ In this way, even if partisanship is a dominant frame through which Americans evaluate confirmation politics, Americans’ expectations over how they want the Senate to consider nominees may to lead them to differentially evaluate tactics of obstruction.

Data and methods

I conduct three complementary analyses to assess whether Americans hold attitudes toward the obstruction of Supreme Court nominees, test between competing expectations of how these attitudes are structured, and investigate if these attitudes vary as a function of the type of obstruction employed. In the first, I draw upon nationally representative surveys from two Supreme Court vacancies, in which I study the relationship between an individual’s partisan alignment with the nominating president and support for delay and obstruction of the nomination. I also assess how approval of the Court and knowledge of the vacancy condition this relationship. In the second and third, I employ observational survey data and a survey experiment to assess the relationship between the intensity of obstruction that is employed and partisan attitudes toward its use. Table 1 summarizes the research questions, political contexts, surveys, and analyses of the paper.

Table 1. Explaining Attitudes Toward Obstruction: Research Questions, Evidence, and Analyses

Note. The table presents information on the research questions, contexts, surveys, and analyses in this paper.

I begin by drawing upon two nationally representative surveys conducted during two vacancies on the Supreme Court. The first, with a sample of 1,951 respondents, was fielded immediately after the death of Antonin Scalia. The second, with a sample of 1,402 respondents, was fielded roughly one month after Anthony Kennedy’s announcement that he planned to retire at the end of the Court’s 2017–2018 term.Footnote ¹¹

In both surveys, respondents were asked about their attitudes toward the nomination and confirmation process when it was salient and unfolding in real-time, helping bolster the external validity of the attitudes these surveys measure. Providing further external validity to the measurement of these attitudes is that methods of obstruction were widely discussed and ultimately employed in both contexts. In the days following Scalia’s death, a major talking point in coverage of the vacancy was whether Senate Republicans would employ procedural tactics to delay a vote on a possible Obama replacement nominee and ensure the next president could make the nomination.Footnote ¹² Discussion of the confirmation politics surrounding President Trump’s nomination of Brett Kavanaugh highlighted the tactics Senate Democrats could and did employ to slow down his confirmation process, even as Kavanaugh was ultimately confirmed.Footnote ^13, Footnote ¹⁴

I measure the public’s attitudes toward obstruction in the context of the Scalia vacancy with a question that captures respondents’ attitudes toward Obama or the winner of the 2016 presidential election nominating the next Supreme Court Justice.Footnote ¹⁵ This question closely parallels the divide amongst Republican and Democratic politicians at the time of the survey as to whether the 2016 election should be treated as a referendum on the vacancy, or whether the Senate had an obligation to comprehensively consider a nominee for the vacant seat. In the context of the Kennedy vacancy, I use a question that asks respondents whether the Senate should consider and vote on Trump’s nominee before or after the 2018 midterm elections.Footnote ¹⁶ From these questions, I create the outcome variable Support for Delay, which takes a value of 1 if respondents expressed support for the next president appointing or next Congress considering and voting on the nominee.Footnote ¹⁷

The different contexts and question wordings of these surveys provide my analysis with important variation for understanding the nature of public attitudes toward obstruction. Examining attitudes toward these confirmation processes in tandem allows me to avoid drawing conclusions from a single vacancy for which particular contextual considerations may not apply in other circumstances. As these vacancies occurred under different presidents of different partisan identities, under both unified and divided government, and at the beginning and end of a presidential term, any patterns in attitudes toward obstruction that emerge from both contexts will allow me to be more confident that these patterns are not an artifact of any single nomination. This variation can also provide leverage for understanding any differences that do emerge in attitudes from each context.Footnote ¹⁸ Finally, as the specific survey questions vary in how they tap into the underlying concept of delay (one asks about the appropriate nominating president, the other asks directly about the Senate’s behavior), this allows me to be more comprehensive in speaking to the general attitudes Americans have toward delay and obstruction rather than being limited to a single real-world measure.

I employ OLS regression models to investigate the extent to which an individual’s political leanings drive their attitudes toward the use of obstruction in these contexts.Footnote ¹⁹ To test whether individuals of different political orientations differ in their attitudes toward obstruction, I create the binary explanatory variable Party Alignment, which takes on a value of 1 if the respondent is of the same party as the president making the nomination (e.g., Democrats in the context of the Scalia vacancy receive a 1 for this variable) and a 0 otherwise.Footnote ²⁰ In my main analysis, I focus on the differences between the president’s copartisans and outpartisans and omit independents from my analysis; results are substantively similar when treating independents as outpartisans. If the theoretical expectations about partisan polarization in attitudes toward obstruction are correct, the coefficient on Party Alignment should be negative. However, if Americans approach the judicial confirmation process through an apolitical lens, the coefficient should be small in magnitude and statistically indistinguishable from zero. In addition, my models also include respondent-level demographic variables plausibly associated with attitudes toward obstruction.

In my full model, I also include two respondent-level characteristics that both of my surveys ask about and that research on public attitudes toward the judiciary suggests might condition the main relationship of interest: knowledge of the vacancy and approval of the Court. Familiarity with and support for the Court is associated with a proclivity to evaluate the Court through a legal, rather than political, lens (e.g., Gibson and Caldeira Reference Gibson and Caldeira2009b), which suggests that Court approval may mitigate differences between the president’s copartisans and outpartisans in evaluations of the use of delay. Knowledge of the vacancy indicates a greater level of attentiveness to politics and the Court, which suggests the individual may be better aware of the political stakes of judicial nominations and the nomination at hand and increase any differences between a president’s copartisans and outpartisans in evaluations of the use of delay (e.g., Park and Smith Reference Park and Smith2016). I capture these respondent-level characteristics with the variables SC Approval and Vacancy Knowledge, binary indicators of an individual’s Court approval, and knowledge of the vacancy.Footnote ²¹

The political nature of attitudes toward obstruction

Before presenting the regression results, I note that mean levels of support are consistent with the argument that partisanship shapes attitudes toward obstruction. Across both nominations, the mean level of support for obstruction for all respondents is 50.5 percent. Mean support is 24.2 percent for copartisans, 56.9 percent for independents, and 76.5 for outpartisans.Footnote ²²

Appendix Table C.3 presents results from four linear regression models that assess the relationship between an individual’s political leanings and attitudes toward delay in the confirmation process.Footnote ²³ Models (1) and (4) include respondents from both surveys, while model (2) is a subset to the Scalia vacancy and model (3) the Kennedy vacancy. In the baseline model (1), the negative coefficient on Party Alignment ( $ \mathrm{p}<0.01 $ ) reveals that political alignment with the president is a significant determinant of attitudes toward delayed consideration of the president’s judicial nominees. The president’s copartisans are predicted to be roughly fifty-three percentage points less likely to support delay than the president’s outpartisans. Furthermore, the negative coefficient on SC Approval ( $ \mathrm{p}<0.01 $ ) indicates that supporters of the Court are less supportive of delay than those unsupportive of the Court. Finally, knowledge of the vacancy is negatively associated with support for delay ( $ \mathrm{p}<0.14 $ ). These results persist when examining the Scalia and Kennedy vacancies on their own, as models (2) and (3) denote.Footnote ²⁴

To test whether Court approval and knowledge of the vacancy condition the partisan differences in evaluations of obstruction, I conduct a triple interaction of the SC Approval and Vacancy Knowledge variables with Party Alignment and present the results from this regression in column (4) of Table C.3. Results are similar when running two models and conducting single interactions in each (see Table C.6). As the regression coefficients on triple interactions are difficult to interpret, I illustrate the relationship between these variables by plotting a set of marginal effects in Figure 1. The y-axis presents the estimated effect of moving from a 0 to 1 on SC Approval (left panel) and Vacancy Knowledge (right panel) on support for delay. Average marginal effects are plotted in black for the president’s outpartisans (circles) and copartisans (triangles). To fully depict the nature of the triple interaction, I also plot marginal effects broken down by level of vacancy knowledge (left panel) and Court approval (right panel) in gray.

Figure 1. Partisanship, Court Approval, Vacancy Knowledge, and Support for Delay.

Note. Figure shows the estimated marginal effect of SC Approval (left panel) and Vacancy Knowledge (right panel) on support for delay for outpartisans (circles) and copartisans (triangles) of the president. Average estimated marginal effects are plotted in black. Estimated marginal effects by vacancy knowledge (left panel) and Court approval (right panel) are plotted in gray. Values are calculated using the margins package in R.

The results show that Court approval and vacancy knowledge play an important role in shaping support for obstruction. The left panel illustrates that specific support mitigates support for obstruction among outpartisans of the president. Outpartisans who are supportive of the Court are on average 17 percentage points ( $ \mathrm{p}<0.01 $ ) less likely to support obstruction than outpartisans who are not (6 percentage points less for those without vacancy knowledge, and 22 percentage points less for those with vacancy knowledge).Footnote ²⁵ Intriguingly, and possibly due to a floor effect, the effect of Court approval in decreasing support for obstruction appears restricted to the president’s outpartisans; copartisans exhibit no distinguishable differences in support for obstruction as a function of Court support (an estimated difference of just 2 percentage points [ $ \mathrm{p}<0.49 $ ]).

Second, the right panel shows how knowledge of the vacancy is associated with decreased support for delay among the president’s copartisans. Copartisans aware of the vacancy are estimated to be roughly 13 percentage points ( $ \mathrm{p}<0.01 $ ) less likely to support delay than those unaware of the vacancy.Footnote ²⁶ However, vacancy knowledge has little to no impact in shaping outpartisan support for obstruction; outpartisans knowledgeable of the vacancy are predicted to be 2 percentage points ( $ \mathrm{p}<0.41 $ ) more likely to support obstruction.

Intensity of obstruction shapes attitudes toward its use

The above surveys provide a number of benefits for studying the relationship between politics and support for obstruction, including high-quality representative samples, comparable questions about obstruction asked in realistic contexts, and measures of two important respondent-level characteristics associated with attitudes toward the judiciary (Court approval and vacancy knowledge). Respondents in both, however, evaluate a substantively similar type of obstruction: entirely delaying consideration of the nominee until the next Congress takes session. This leaves unanswered an important question: whether the type of obstruction employed matters in shaping public evaluations of its use. There are many types of obstruction, ranging from the relatively innocuous (e.g., document requests) to the intense (e.g., outright refusal to consider a nominee). Theoretically, there is reason to believe the public’s attitudes toward the use of tools of obstruction vary depending on the type under consideration.

I marshal evidence from two sources to answer this question. First, I draw upon 22 questions from nationally representative surveys conducted during seven historical Supreme Court vacancies that ask respondents about a wide range of different types of obstruction. Second, I employ an original survey experiment in which I manipulate the intensity of obstruction respondents are exposed to in a hypothetical nomination context. This allows me to combine the external validity of a wide range of historical surveys with the internal validity of an experiment. When taken together, these data sources provide valuable evidence to speak to how the type of obstruction under consideration shapes how the public evaluates its use.

Historical evidence on intensity and attitudes toward obstruction

First, I turn to a set of survey questions that ask about tactics of obstruction that were either proposed or used during seven Supreme Court vacancies.Footnote ²⁷ To gather these questions, I drew upon the Roper iPOLL database, a repository of over 700,000 questions from national public opinion surveys conducted since the 1930s. For the purposes of this project, I searched the database for questions that asked individuals to evaluate the use of any tools of obstruction or delay as applied to Court nominees. This provides me with 22 questions that measure Americans’ attitudes toward obstruction and delay in the Court confirmation process.Footnote ²⁸

Importantly, these questions vary in the type of obstruction under consideration. Existing scholarship establishes that the American public views judges as distinct from other types of politicians, holds expectations of procedural fairness when evaluating the judiciary, and dislikes partisan bickering and gridlock in Congress (Baird and Gangl Reference Baird and Gangl2006; Flynn and Harbridge Reference Flynn and Harbridge2016; Gibson and Caldeira Reference Gibson and Caldeira2011). When applied to how Congress treats Court nominees, these views suggest that the public has expectations over giving nominees full procedural consideration and that the public may evaluate methods of obstruction differently based upon the degree to which the methods fit these expectations.

Guided by this extant scholarship, I classify tactics of obstruction asked about in these survey questions into three categories. The first, which I call weak intensity, captures relatively innocuous attempts to slow down the process that do not preclude nominees getting full consideration. In my data, the questions I have that fit into this category are calls for additional documents, but similar tactics (such as calls for additional hearings) would fit into this category. These tactics are within the established procedures of the Senate and commonly used to handle Court nominees. These tactics are unlikely to violate expectations of procedural fairness, lead to gridlock or severe partisan disagreement, or lead the public to think the Senate is treating judges like ordinary politicians. The second, which I call moderate intensity, captures methods of delay that can lead to a nominee receiving some but not full consideration. In my data, the questions I have that fit into this category include the use of the filibuster and the capacity of the majority party to prevent a final floor vote on a nominee. While still well-established Senate procedures to handle Court nominees, these tools may more plausibly lead to gridlock, suggest that the Senate is treating judges like ordinary politicians, and lead to violations of procedural fairness expectations. The third, which I call severe intensity, captures tactics that are associated with entirely blocking consideration of a nominee. In my data, these are questions about foregoing any consideration of a nominee or not holding hearings on a nominee. More than any other attempt at obstruction, these tactics are not well-established Senate tools for handling nominees and are the most likely to lead to gridlock, convey to the public that the Senate views the nominee as a political actor, and deny the nominee procedural consideration.Footnote ²⁹ A list of the questions I draw upon and how I classify the method of obstruction they ask about is available in Tables D.1 and D.2.Footnote ³⁰

I leverage the variation in the type of obstruction respondents evaluate to assess my theoretical expectations as to whether Americans are more supportive of the use of weaker forms of obstruction than more severe forms. Due to data availability limitations, I can only access the individual-level data for 14 of the 22 survey questions I have that ask about obstruction. Thus, in order to draw upon all 22 questions, I begin by looking at aggregate support for obstruction by partisan groups. For each survey question, I determine the percentage of copartisan, independent, and outpartisan respondents who support the use of obstruction (of all respondents from each group who reported an opinion on obstruction).Footnote ³¹ I then average these aggregate percentages to get an average level of percentage support for obstruction for each of the three intensity levels and partisan groups. I present the results in Figure 2. The y-axis denotes the average percentage of respondents who support obstruction, and the x-axis denotes the intensity of obstruction under consideration. I present percentages for the president’s copartisans (circles), independents (triangles), and outpartisans (squares).

Figure 2. Intensity, Partisanship, and Support for Obstruction (Aggregate Analysis).

Note. The plot shows average levels of percentage support for the use of obstruction across weak, moderate, and severe instances of obstruction from 22 nationally representative survey questions. Plotted numbers and p-values reflect the difference in percentage support between weak and severe instances of obstruction for each partisan group.

The findings illustrate that outpartisans of the president and independents, although not copartisans, express decreasing levels of support for the use of obstruction as its intensity increases.Footnote ³² For outpartisans, a movement from weak to severe forms of obstruction leads to a 30.0 percentage point ( $ \mathrm{p}<0.01 $ ) decrease in support for obstruction, whereas the same change for independents is 19.0 percentage points ( $ \mathrm{p}<0.01 $ ). Substantively, this reflects a significant change in attitudes, especially in light of the powerful predictive role that partisanship plays in contemporary political attitudes. The president’s copartisans exhibit no statistically distinguishable differences in support for obstruction by intensity; this may be the result of a floor effect, as a sizable majority of copartisans are unsupportive of obstruction in all forms. Furthermore, these surveys corroborate the above analysis in showing the considerable partisan differences in evaluations of obstruction in the nomination and confirmation process.Footnote ³³

As noted above, I am able to access the individual-level data for 14 of the 22 survey questions I employ in this analysis. For these questions, I can assess individual-level support for obstruction in a regression framework. Such an analysis allows me to draw upon a greater number of observations (individual respondents rather than aggregate surveys) and account for other possible factors that may shape evaluations of obstruction (e.g., respondent-level characteristics or the nominating president). However, this individual-level analysis does come with limitations, as the results will be more susceptible to individual survey questions and contexts.Footnote ³⁴ Happily, despite the smaller number of questions, I still have full coverage of questions about each intensity of obstruction in the individual-level data.Footnote ³⁵ For each survey, I subset the data to include only those respondents with an opinion on obstruction and classify respondents into supporters or opponents of obstruction. I determine whether respondents are copartisans of the nominating president, outpartisans, or independents (coding leaners as partisans except in a small number of surveys where a measure of leaners is not available). I create indicators for which intensity of obstruction the respondent evaluated. I also control for the nominating president and two respondent characteristics that are available across all 14 survey questions (gender and race). Then, I conduct an OLS regression of support for obstruction as a function of respondent partisan alignment with the president, the intensity of obstruction, and the interaction of the two. This allows me to assess how these two variables shape support for obstruction.

For interpretability and comparability to the above aggregate analysis, I present the results from my full model in Table 2 as a set of predicted probabilities.Footnote ³⁶ In short, the results largely parallel the findings from the aggregate analysis above in both substantive and statistical significance. As the intensity of obstruction increases, outpartisans and independents are predicted to be increasingly less supportive of obstruction. Outpartisans exhibit the greatest difference, with 25.8 percentage point ( $ \mathrm{p}<0.01 $ ) predicted less support for severe as opposed to weak forms of obstruction. There are statistically significant but substantively small predicted differences between how the president’s copartisans evaluate weak as opposed to moderate and severe obstruction; as in the aggregate analysis, copartisans are predicted to be largely unsupportive of obstruction in all forms.Footnote ³⁷

Table 2. Intensity and Predicted Support for Obstruction (Individual-Level Analysis).

Note: Predicted probabilities come from OLS regressions of support for obstruction as a function of intensity, partisan alignment with the president, and their interaction. p-values are calculated using the margins package in R.

Finally, while most surveys only ask about one intensity of obstruction, one of my surveys asks about both moderate and severe forms of obstruction. In a within-respondent analysis in Section D.5, I show that respondents are significantly more supportive of moderate as opposed to severe obstruction. This result provides further support to my conclusion that, as obstruction increases in intensity, Americans are less supportive of its use.

Experimental evidence on intensity and attitudes toward obstruction

The wide range of contexts and types of obstruction I draw upon in my observational surveys help me tell a generalizable story about how the intensity of obstruction is associated with support. However, as with any observational analysis, a limitation of these data is the possibility for other factors apart from the obstruction (e.g., attitudes toward the president or characteristics of the nominee) to shape evaluations of its use. Accounting for partisanship, the nominating president, respondent characteristics, and support for the nominee – as I do in various analyses above and in the Appendix – can help assuage concerns about these factors shaping my conclusions.Footnote ³⁸ An additional solution to these challenges is a survey experiment. Randomly assigning respondents a nominating president, nominee, and method of obstruction alleviates concerns about any other factors shaping evaluations of obstruction. Of course, there are also limitations that come with survey experiments. A single-shot experiment is necessarily of a particular point in time and is divorced from the real world. However, by combining the external validity of observational surveys and the internal validity of a survey experiment, I can tell a more complete story of the relationship between intensity of obstruction and support for its use.

I conducted a nationally representative survey experiment of 1,250 American adults on the CloudResearch Connect platform in December 2023.Footnote ³⁹ CloudResearch has been shown to provide high-quality responses in social science research studies (Douglas, Ewell, and Brauer Reference Douglas, Ewell and Brauer2023). Respondents were quota targeted to match U.S. Census demographics on gender, race, ethnicity, and age.Footnote ⁴⁰ I was guided by examples of how politicians talk about obstruction and the above theoretical discussion in designing my experiment. First, respondents saw a prompt where they were asked to imagine that either President Biden or Trump (randomly assigned) was re-elected in 2024 and that the Senate was controlled by the opposite party of the president. Thus, the president faced divided government, an effective precondition to moderate and severe forms of obstruction at the time of the survey. Respondents were told to imagine a Supreme Court vacancy arose and then saw a hypothetical nominee profile with varying characteristics that were randomly assigned; these variations are not a part of my analysis but serve to ensure specific nominee traits are not associated with evaluations of obstruction. Finally, respondents received a prompt detailing the Senate majority’s statement on how they planned to handle the nomination. Respondents were randomized into one of four conditions: a control (no obstruction; proceeding in a timely manner with hearings and a vote), weak obstruction (a request for additional documents), moderate obstruction (holding hearings on the nominee but not a final floor vote), and severe obstruction (not holding hearings or considering the nominee). The control allows me to account for baseline support or opposition to the nominee in a non-obstruction context. For more information and full wordings, please see Section E.2. Finally, respondents were asked to evaluate the Senate’s treatment of the nominee on a five-point scale from strongly oppose to strongly support. I code respondents who reported strongly or somewhat supporting as supporters of the Senate’s behavior and the others as non-supporters.Footnote ⁴¹

I code respondents as copartisans of the nominating president, outpartisans, or independents.Footnote ⁴² For each partisan group, I assess the treatment effect of weak, moderate, and severe forms of obstruction on support for the Senate’s behavior as compared with the control condition of no obstruction. I present the results in Figure 3; plotted points and 95 percent confidence intervals are differences-in-means between the obstruction condition and the control for each partisan group.Footnote ⁴³ I also report average levels of support for the Senate’s behavior across the three obstruction treatment conditions on the y-axis for each partisan group.

Figure 3. Experimental Results: Intensity and Support for Obstruction.

Note. The plot presents treatment effects and 95 percent confidence intervals of weak, moderate, and severe obstruction treatments as compared to the control condition. Separate treatment effects are plotted for respondents who saw an outpartisan nominating president, a copartisan nominating president, and independents. Mean support for the Senate’s behavior under the three obstruction treatment conditions is presented under each partisan group label.

The results reveal no substantively nor statistically distinguishable differences between the control and weak obstruction treatment. This is evidence that the public sees weak forms of obstruction as part of the typical advice and consent process, and that these tactics do not violate procedural fairness expectations nor tap into the public’s dissatisfaction with partisan disagreement in Congress. For all partisan groups, I find a significant decrease in support for the Senate’s behavior under moderate and severe forms of obstruction as compared to the control (and weak treatment). For example, outpartisans of the president are 33.4 percentage points [35.0] less supportive of severe obstruction than the control ( $ \mathrm{p}<0.01 $ ) [weak treatment ( $ \mathrm{p}<0.01 $ )]. For outpartisans and independents, these differences benchmark fairly well to the observational results presented above. Interestingly, copartisans exhibit greater support for the weak obstruction treatment than moderate and severe, unlike in the observational analysis. One possible explanation for these different findings is that, in the weak treatment, the Senate explicitly stated that they planned to proceed with hearings and a vote for the nominee. While in the real world, these weak forms of obstruction generally do not derail the nomination, it is possible that in the real world copartisans of the president were worried about such a possibility.

Contrasting with the observational analysis above (and the within-respondent analysis in Section D.5), the experiment does not reveal distinguishable differences between how respondents evaluate moderate and severe forms of obstruction. There are at least two plausible explanations for these different findings. First, in order to create a realistic treatment with a similar setup across all treatment conditions, respondents in the moderate treatment saw a Senate statement about planning to hold hearings but not a final vote on the nominee. This pre-commitment to not holding a final vote on the nominee might have been viewed by respondents as a more intense form of obstruction than analogous real-world contexts where the Senate may have left open the possibility of holding a vote.Footnote ⁴⁴ Second, my experiment was conducted in 2023, meaning respondents had lived through a number of tumultuous real-world nomination contexts in which severe forms of obstruction were discussed. It is plausible that attitudes toward obstruction were different in 2023 than in the earlier years from which my observational questions above come from. Future experimental research should test the impact of other contextual setups and methods of obstruction on support for its use.

Putting the observational and experimental results together, I conclude that the specific means of obstruction the Senate employs when considering Supreme Court nominees plays an important role in shaping attitudes towards its use. Thus, in addition to the critical role that politics plays in shaping attitudes toward the use of obstruction, Americans also hold attitudes over the specific methods of obstruction the Senate employs when considering Supreme Court nominees.