INTRODUCTION
US Supreme Court confirmation hearings before the Senate Judiciary Committee are incredibly important. With high levels of public interest and gavel-to-gavel television coverage, the hearings provide a unique opportunity for the world to observe and develop opinions about senators, nominees, and the usually opaque Supreme Court. For nominees, the hearings are their primary chance to defend their judicial record or develop credibility—all of which can help gain senator confirmation votes and develop public support for their service on the Court. For senators, the Judiciary Committee is a coveted assignment, and Supreme Court confirmation hearings offer them an excellent opportunity to increase constituent support, raise their national profile, and even launch a White House bid (Collins and Ringhand Reference Collins and Ringhand2016; Schoenherr, Lane, and Armaly Reference Schoenherr, Lane and Armaly2020).
But not all hearing participants have equal opportunity to speak unimpeded, as Supreme Court nominee Ketanji Brown Jackson found out during her 2022 hearing. In the course of just minutes of questioning from Senator Graham (R-SC), Jackson was interrupted three times.Footnote 1 Some senators also have this experience. For example, in her 2018 questioning of Brett Kavanaugh, Senator Hirono (D-HI) was interrupted multiple times by the nominee.
In this high-profile setting, interruptions matter, especially when disproportionally directed to the speech of women and person of color senators and nominees. Interruptions silence the original speaker and elevate the power of the interrupter, which may aggravate the challenges already facing women and people of color in high-profile settings. Women and person of color senators are at a numerical disadvantage in Congress and face challenges in winning elections and governing once elected (e.g., Holman, Merolla, and Zechmeister Reference Holman, Merolla and Zechmeister2011; McDermott Reference McDermott1998; Palmer and Simon Reference Palmer and Simon2006; Shah Reference Shah2015). Interruption inequities can be an additional blow. For the nominees, being interrupted may stall their efforts to overcome the competence bias women and people of color face in interviews across professions (Boyd, Collins, and Ringhand Reference Boyd, Collins and Ringhand2018; Christensen, Szmer, and Stritch Reference Christensen, Szmer and Stritch2012; Haynie Reference Haynie2002; Lawless Reference Lawless2004; Nelson Reference Nelson2015). More broadly, since legitimacy and public support rest on representative inclusion in governing institutions (Badas and Stauffer Reference Badas and Stauffer2018; Barnes Reference Barnes2016; Clayton, O’Brien, and Piscopo Reference Clayton, O’Brien and Piscopo2018; Harris and Sen Reference Harris and Sen2019; Means, Eslich, and Prado Reference Means, Eslich, Prado, Sterett and Walker2019; Scherer Reference Scherer2023; Stauffer Reference Stauffer2021; Widner Reference Widner2023), public displays of bias against women and people of color can result in lower rates of ambition for members of those groups (Fox and Lawless Reference Fox and Lawless2014; Williams Reference Williams2008) and aggravate a sense that our government does not represent them (Campbell and Wolbrecht Reference Campbell and Wolbrecht2006).
We examine interruptions coming from societal majority group members like men and white people and the likelihood that they will occur more frequently for women and person of color participants in Supreme Court confirmation hearings, particularly when those participants are of the opposite party of the would-be interrupter. We utilize almost 40,000 senator and nominee statements from 1939 to 2022 to analyze confirmation hearing interruptions. Our findings indicate that cross-party women and people of color face substantially higher rates of interruptions than others.
These results are an important addition to the existing literature in three ways. First, we are the first to show that biased interruptions come from both senators and nominees. Even in a situation in which male and white individuals are presumptively in a position of less power than their senatorial interrogators—testifying for a coveted Supreme Court seat for which they need senatorial votes—they still interrupt female senators and senators of color at higher rates than others. Second, while prior political science literature studying interruptions has, to our knowledge, exclusively focused on gendered patterns, our findings confirm that nominees and senators of color are also subject to heightened interruption rates. Third, we show that bias in interruptions occurs even in high-profile settings that garner substantial media attention and interest from the American public.
INTERRUPTIONS, RACE, GENDER, AND PARTISANSHIP
Whether in Congress, the courtroom, or elsewhere, turn-taking—where a single person serves as the speaker while others act as listeners—is the norm in group speaking situations (Beattie Reference Beattie1982). Allowing speech without interruption signals respectful discourse (Bresnahan and Cai Reference Bresnahan and Cai1996; Goldberg Reference Goldberg1990) and is cognitively necessary since we struggle to “talk and listen simultaneously” (Beattie Reference Beattie1982, 93). Interruptions break conversational turn-taking norms. When another person interrupts before the prior speaker’s turn is finished, the interruption stops the speaker from continuing and disorganizes the future speech (Smith-Lovin and Brody Reference Smith-Lovin and Brody1989). Interruptions thus can be a conversational “power play” (Mendelberg, Karpowitz, and Oliphant Reference Mendelberg, Karpowitz and Oliphant2014) that control or alter speech and grant dominance to the interrupter (Karpowitz, Mendelberg, and Mattioli Reference Karpowitz, Mendelberg and Mattioli2015; Leman and Ikoko Reference Leman and Ikoko2010).
In political settings, interruptions enable control over the discussion and achievement of outcome goals. In congressional hearings, for example, interruptions loudly signal conflict among participants, with policy conflict likely to emerge thereafter (Miller and Sutherland Reference Miller and Sutherland2023). During Supreme Court oral arguments, justices’ interruptions prevent their colleagues from coordinating or signaling their preferences (Black, Johnson, and Wedeking Reference Black, Johnson and Wedeking2012; Johnson, Black, and Wedeking Reference Johnson, Black and Wedeking2009) and reduce the likelihood of justices voting together (Jacobi and Rozema Reference Jacobi and Rozema2018).
Race and Gender Bias
Interruptions can happen among any speaker pairings. However, when majority group members (like men and white people) are the would-be interrupters, interruptions may be particularly likely to be targeted at women and person of color speakers. The combination of bias, status, negative stereotypes, and incentives to grow power and derogate subordinates can result in powerful group members inferring (consciously or not) entitlement to seize the conversational floor from women and person of color speakers (Anderson and Leaper Reference Anderson and Leaper1998; Christensen, Szmer, and Stritch Reference Christensen, Szmer and Stritch2012; Moyer et al. Reference Moyer, Szmer, Haire and Christensen2021; Smith-Lovin and Brody Reference Smith-Lovin and Brody1989). Inequities in who is interrupted may further entrench the power of majority classes and signal weakness among those with less power (Mendelberg, Karpowitz, and Oliphant Reference Mendelberg, Karpowitz and Oliphant2014).
Women and people of color face more interruptions than male and white speakers in many settings (e.g., Leman and Ikoko Reference Leman and Ikoko2010). This pattern holds in political and legal contexts, although political science research to date has largely focused on gendered interruptions and ignored race-related effects. Women—including congresspeople (Miller and Sutherland Reference Miller and Sutherland2023) and witnesses (Mattei Reference Mattei1998)—are interrupted more than men during congressional hearings (but see Kathlene Reference Kathlene1994). In courts, female judges, justices, and lawyers are interrupted more than men (Bogoch Reference Bogoch1999; Cortina et al. Reference Cortina, Lonsway, Magley, Freeman, Collinsworth, Hunter and Fitzgerald2002; Feldman and Gill Reference Feldman and Gill2019; Jacobi and Schweers Reference Jacobi and Schweers2017; Patton and Smith Reference Patton and Smith2017).
The Conditioning Effect of Partisanship
The activation of inequalities in interruptions is likely dependent on the relationship status and worldview perspectives of the speaker and would-be interrupter. Individuals who do not know each other well are more likely to have higher interruption levels (James and Clarke Reference James, Clarke and Tannen1993). By contrast, “friends” interrupt each other less frequently and alternate speaking turns (Bresnahan and Cai Reference Bresnahan and Cai1996; Smith-Lovin and Brody Reference Smith-Lovin and Brody1989). This has been confirmed in Supreme Court oral arguments, where interruptions rise among ideological foes (Johnson, Black, and Wedeking Reference Johnson, Black and Wedeking2009) and when justices’ ideology and gender are different (Jacobi and Schweers Reference Jacobi and Schweers2017).
On the Judiciary Committee, we expect that shared party will lessen the gendered and racial effects of majority group members’ interruptive behavior, while opposite-party affiliation will heighten such effects. The modern Senate is highly partisan, and its committees rely on party-based teamwork (Lee Reference Lee2009). Senators will be more likely to view shared-party nominees as “friends” than those appointed by opposing-party presidents (regardless of race or gender). The same should hold for nominees interacting with senators. With shared party, there will be little hostility to the nomination and less need for aggressiveness in exchanges. By contrast, opposite-party speakers may be seen as “different,” without the mitigating salve of a shared partisanship. Questioning senators are incentivized by partisan loyalties to behave assertively, and nominees will instinctively view cross-party senators as foes. These opposite-party, female and person of color speakers, are thus likely to receive a higher rate of interruptions than their shared party counterparts.
DATA AND METHODS
To study majority group members’ interruption patterns, we examine every public Supreme Court confirmation hearing held before the Senate Judiciary Committee to date (1939–2022).Footnote 2 The unit of analysis is the nominee–senator dyad, with one observation per nominee-senator pair per hearing.Footnote 3 Our main analysis focuses on intrusive interruptions (attempts to take over the conversation) and excludes backchannel interruptions (signs of active conversational listening).Footnote 4 The dependent variables are the percentage of time the senator/nominee intrusively interrupted the nominee/senator they were paired with as a function of all statements they made at each hearing. Our primary analysis is limited, for our gender-centric models, to interruptions made by male senators (or nominees) and, for our race-centric models, to interruptions by white senators (or nominees). We estimate ordinary least squares regression models and cluster the standard errors on the would-be interrupter.
We include independent variables that capture the gender/race of the nominee or senator and their party status relative to the individual they are interacting with. For example, in the model that examines interruptions of nominees by male senators, we include three variables: Female Nominee, Same Party; Female Nominee, Different Party; and Male Nominee, Different Party (with Male Nominee, Same Party as the baseline category). We expect that all three of these variables will be positively signed relative to the baseline, with female nominees from the opposite party of the questioning senator being interrupted more than all other nominee types. We use a similar modeling strategy for models focusing on interruptions by white senators, male nominees, and white nominees. We also include control variables in the models to account for other participant and contextual factors that may affect interruption patterns (detailed in Supplementary Appendix). Our variables’ expected directional effects are summarized in Table 1.
INTERRUPTIONS BY SENATORS
Figure 1 reports the coefficients and 95% confidence intervals for the statistical models focused on senators intrusively interrupting nominees.Footnote 5 The top panel focuses on gender differences and reveals that male senators interrupt female, different-party nominees 4.5 percentage points more often than same-party male nominees (the baseline category). We also find that male senators interrupt male, different-party nominees 3.8 percentage points more than male, same-party nominees, but there is no statistically significant difference between interruptions of female and male, same-party nominees.Footnote 6 These results provide clear evidence that male senators exhibit gendered interruption patterns, interrupting female nominees from the opposite party more often than male, same-party nominees.
Our regression results plotted in Figure 1’s bottom panel similarly reveal strong evidence of racial differences in the interruption behavior of white senators. In particular, white senators interrupt opposite-party nominees of color 6.5 percentage points more than white, same-party nominees. This is, by far, the largest coefficient in the model and one that is statistically significantly larger than every other nominee grouping in the model based on equality of coefficient tests (see Supplementary Appendix).Footnote 7 Compared to white nominees from the opposite-party, nominees of color from the opposite party are interrupted a shocking 108% more frequently.
INTERRUPTIONS BY NOMINEES
We now turn the tables on hearing participants and explore in Figure 2 intrusive interruptions of senators by the nominees. As one would expect in the job interview–like format of a confirmation hearing, interviewees are less interruptive than their interviewers: nominees interrupt senators about half as much as they are interrupted by them (2.1% of statements compared to 4.3%). Those interruptions are not equally distributed, however, with male nominees and white nominees’ interruption patterns revealing sharp gendered and racialized behavior. Compared to their behavior toward same-party, male senators, male nominees interrupt female, different-party senators 6.8 percentage points more (Figure 2, top panel). Equality of coefficients tests reveal that female, different-party senators also receive greater interruptions than other groupings (male, different-party and female, same-party senators). There is therefore clear evidence that male nominees treat female senators from the opposite party differently in their interruptive behavior than all other senators.
The bottom panel in Figure 2 shows a similar pattern for white nominees as interrupters. Compared to white same-party senators, white nominees interrupt senators of color from the opposite party 5.6 percentage points more. And, per equality of coefficients tests, opposite-party senators of color are interrupted more than same-party senators of color.Footnote 8 As with the interruption patterns for other nominees and senators, these are substantively large effects revealing that white nominees intrusively interrupt opposite-party senators of color almost three times as much as is average for nominees.
CONCLUSIONS
Interrupters hold power to assert dominance, take the conversational floor, and disorganize the speech of another. Despite the salience and stakes attached to Supreme Court confirmation hearings, they are not immune to interruptive behavior by participants. Interruptions are not only frequent in this setting, but they are also unevenly dispersed. As our study reveals, women and person of color speakers, when paired with a male or white speech partner during the hearings,Footnote 9 face a disproportionate number of interruptions relative to other speakers—especially when the speech pair does not share the same political party.
Viewed as a whole, our analyses present multiple contributions of note to the study of American politics, Congress, the legal system, and race and gender bias in society.Footnote 10 First, gender and racial biased interruptive behavior is not isolated to private conversations, in nonpublic arguments at the Supreme Court, or in sparsely attended congressional committee hearings; rather, it also happens at what is likely the most prominent of congressional hearings. Second, this interruptive bias is rampant, extending from members of Congress who hold the upper hand in power and have the “questioner” status during these hearings to nominees themselves in ways that largely mirror what we see from interrupting senators. It is also not limited to gender. While racial bias in interruptions has been largely unexamined in political and legal settings, our findings indicate it is just as potent as is gender bias. And finally, like with so much in politics, shared party and political worldview can mitigate the emergence of interruptive biases, while the lack of this common political perspective among a speaking pair leaves a gulf of space for biases to fill. And that is exactly what we see during the Senate Judiciary Committee’s Supreme Court confirmation hearings.
SUPPLEMENTARY MATERIAL
To view supplementary material for this article, please visit https://doi.org/10.1017/S0003055424000145.
DATA AVAILABILITY STATEMENT
Research documentation and data that support the findings of this study are openly available at the American Political Science Review Dataverse: https://doi.org/10.7910/DVN/C2IATB.
ACKNOWLEDGMENTS
A previous version of this project was presented at the 2023 Annual Meeting of the New England Political Science Association, 2022 Conference for Empirical Legal Studies, and faculty workshops at the University of Virginia and George Mason University law schools. We are grateful to our colleagues for their feedback on this research, including Scott Blinder, Jennifer Bowie, Roberto Carlos, James Cooper, Pam Corley, Josh Fischman, Justin Gross, Susan Haire, Rebecca Hamlin, Morgan Hazelton, Rachael Hinkle, Alex Hindman, Gbemende Johnson, Sally Kenney, Ray La Raja, Monika Lindbekk, Wendy Martinek, Lauren McCarthy, Alison Merrill, Greg Mitchell, Paul Musgrave, Michael Nelson, Tatishe Nteta, Rachel Potter, Jesse Rhodes, Meredith Rolfe, Jamie Rowen, Geoffrey Sheagley, Barbara Spellman, Amy Steigerwalt, Susan Sterett, Art Ward, and Leah Wing. We received invaluable research assistance from the following students: Madison Alvis, Matt Baker, Juliana Bird, Brie Bristol, Rosa Brown, Haley Carman, Rayani Chonmany, Jenah Clarkson, Madison Conkel, Mitchell Director, Jordan Dyer, Gillian Gaines, Caroline Greenlaw, Riley Grube, Grace Hayes, Ryan Heilmann, Julianna Hightower, Emma Hupp, Telly Jacobs, Eshaa Joshi, Nadeen Jumai’an, Caitlyn Kinard, Kate Kostel, Chloe Levy, Elisabeth Lopez, Bryce McManus, Maggie Mendoza, Alaina Moore, Brianna Owen, Moksha Padmaraju, Haley Patel, Katie Pitner, Kiana Powers, Elise Puschett, Emma Rowr, Sam Rusling-Flynn, Adam Rutkowski, Kyle Venit, Allison Vick, Nora Vonmoltke-Simms, Lizzy Walker, Veronica Walsh, Megan Wan, Emily Willard, and Jessie Zacune.
CONFLICT OF INTEREST
The authors declare no ethical issues or conflicts of interest in this research.
ETHICAL STANDARDS
The authors affirm this research did not involve human subjects.
Comments
No Comments have been published for this article.