Are Stealth Democrats Really Committed to Democracy? Process Preferences Revisited

Abstract

Scholarship on “stealth democracy” finds that many citizens want to avoid the debate and conflict that often come with democratic governance. This scholarship has argued that citizens adopt this posture because they are uncomfortable with disagreement and desire a more expedient political process that enables leaders to make decisions without discussion or compromise. We revisit this argument in light of recent political developments that suggest another reason why citizens may desire a more expedient political process. We examine the possibility that some citizens are not merely uncomfortable with disagreement but also want leaders who will aggressively protect them and champion their interests. Using a nationally representative survey, we ask citizens about their preferences for stealth democracy. We also ask questions that tap into their willingness to support leaders who would “bend the rules for supporters” and take aggressive action against political opponents. We find that a substantial component of the electorate continues to prefer a stealth version of democracy. However, we also find that many “stealth democrats” are willing to support leadership practices that would threaten or even undermine democratic norms. We argue that this evidence indicates that, in recent years, many citizens who appear to desire “stealth democracy” pose a threat to democracy itself.

For democracies to remain healthy and survive, citizens must embrace a demanding form of political life. They must embrace decision making that fosters the inclusion of diverse and often conflicting points of view. They must understand the value of debate and recognize that compromise is sometimes necessary to balance competing interests. Just as importantly, they must also support leaders who take this approach to politics.

Of course, in the United States, a significant number of citizens do not live up to this standard. When it comes to governance, many Americans prefer uncompromising political leaders who take decisive action, rather than those who debate issues and are open to finding common ground. Some even express an interest in having unelected experts make decisions about laws and policies (Hibbing and Theiss-Morse 2002; see also Hibbing and Theiss-Morse 2016). Instead of governance that facilitates debate, these citizens prefer to get what they want through a more expedient means of decision making. The implications for democracy warrant careful consideration, especially in our current political moment.

We live at a time when scholars warn about democracies dying from within. This is a warning about political leaders who demonize opponents, strip away rights, and threaten democratic norms and institutions (Hetherington and Weiler 2018, ch. 6; Levitsky and Ziblatt 2018; Mounk 2018). It is also a warning about citizens who allow leaders to act that way. When citizens forsake the value of considering competing perspectives and turn their backs on the value of democratic debate, they can foster conditions that embolden and even empower leaders with antidemocratic tendencies.

For this reason, it becomes important to consider why so many Americans demonstrate this orientation to governance. Two possibilities warrant consideration. The first is that, regardless of what is happening in political life, many citizens find politics unpleasant and are content to rely on their leaders to resolve problems with little need for citizen input or oversight. Some scholarship, in fact, argues that many citizens are “stealth democrats,” who prefer not to see the democratic process in operation (Hibbing and Theiss-Morse 2002). According to this scholarship, such citizens are committed to democracy in principle and want benevolent leaders concerned with the common good. However, in practice, they are uncomfortable with the kind of intense disagreement that often comes with democratic debate. They also think that compromise is unwarranted and that debate amounts to needless bickering. From this perspective, citizens who prefer stealth democracy simply want expedient decisions because they believe that political conflict is unpleasant and unnecessary. They want decisive leaders who will cut through the clutter of political decision making and get things done (Hibbing and Theiss-Morse 2002; see also Hibbing and Theiss-Morse 2016; Medvic 2019; and VanderMolen 2017).

Although this may sound harmless, this desire for an expedient governing process may contribute to the weakening of democracy. When citizens develop a preference for expedient governance to avoid the discomfort of political disagreement, they are, by implication, trying to avoid the competition and deliberation over ideas that a healthy democracy requires. In turn, by devaluing the importance of political debate, these citizens may enable leaders with antidemocratic impulses to act on them. Consequently, even if stealth democrats want benevolent leaders, they may empower leaders who are the opposite.

For anyone concerned about the sustainability of democracy, a widespread preference for stealth democracy would be troubling. Yet, there is a second and even more troubling possibility to consider. When citizens express a desire to avoid debate and compromise, they may be motivated by something other than mere discomfort with political disagreement. And when citizens express an interest in unelected political experts, they may not be looking for benevolent leaders who stand above the fray. Instead, they may be seeking leaders who enter the fray to aggressively protect people like themselves from threats to their in-group or idealized way of life (Galston 2018; Hibbing 2022; Levitsky and Ziblatt 2018; Marchlewska et al. 2018). They may also be looking for leaders who will “bend the rules” if necessary to advance the interests of their in-group (Bartels 2020; Graham and Svolik 2020). Put more bluntly, citizens who at first appear to be stealth democrats may be willing to support leaders who challenge or even undermine democratic norms.

In this study, we explore this very matter. We demonstrate that the citizens most inclined to avoid debate and compromise are not merely trying to avoid the messiness of political competition. Instead, they express a willingness to support rule-bending leaders who will denigrate and punish groups they find threatening and bend the rules for people like themselves. This provides insight into how democracies can be weakened from within. We argue that, at present, the more significant threat is not from stealth democrats who seek to avoid political disagreement. The greater threat comes from citizens who will support leaders who settle disagreements in their favor, even if that comes at the cost of democratic norms.

The Desire for Stealth Democracy, Revisited

In a democracy, political competition is vital. Debate can help citizens think more carefully about ideas and perhaps even find shared interests or notions of the common good (Bernstein 2000; Caspary 2018; Fitch and Loving 2007). More fundamentally, political competition can help prevent any particular group in society from becoming so powerful that it can violate the rights of those it opposes (Allen 2020; Fishkin 2020; Shapiro 2016). From this perspective, democracy has a problem when citizens undervalue competition between rival perspectives and the debate and compromise that generally come with it.

By highlighting two important patterns, scholarship advancing the stealth democracy thesis provides one explanation about why this problem can arise. First, many citizens find the debate and disagreement of democratic politics unpleasant and want to avoid it (Hibbing and Theiss-Morse 2002; VanderMolen 2017). This is what the “stealth” in stealth democracy signifies. On this interpretation, citizens who are uncomfortable with disagreement want decision making that remains “out of sight” so that they can be “free[d] from the need to follow politics” (Hibbing and Theiss-Morse 2002, 131; see also Coffé and Michels 2014; Font, Wojcieszak, and Navarro 2015; Lavezzolo and Ramiro 2018; and VanderMolen 2017).

Second, citizens who seek to avoid disagreement prefer an expedient approach to governance. This desire for expedience is implied by their preferences about the governance process. Citizens described as stealth democrats prefer to avoid debate in favor of decisive action. They dislike compromise and want leaders who stick to their principles. They also express faith in governance by unelected political experts who do not need to work with others to accomplish their political goals. As one study invoking stealth democracy notes, this approach to political decision making can “promote a level of efficiency and effectiveness within the government” that some citizens find attractive (VanderMolen 2017, 687). ¹ Debates between elected officials and efforts to achieve compromise are notoriously time consuming; uncompromising decisiveness, unfettered from debate, is faster. What would make this preference for efficiency a preference for “stealth,” however, hinges on why citizens prefer it. If they prefer efficiency specifically because they want to avoid “the sight” of political disagreement, then a preference for expedient governance is a preference for stealth democracy. Stated more formally,

H1: Citizens with higher levels of discomfort with political disagreement will have stronger preferences for expedient governance.

Importantly, from this perspective, citizens are not “authoritarians, preferring political decisions to be made by a detached dictator” (Hibbing and Theiss-Morse 2002, 239). To the contrary, they simply prefer to avoid the fray of democratic politics, provided that their leaders are “caring, other-regarding, common-good–oriented…empathetic, non-self-interested decision makers” (161). If accurate, this interpretation implies that the threat to democracy is not mainly attributable to people who harbor ill will against their fellow citizens. Instead, it is the people who care about the “common good” but “do not comprehend any legitimate justification for intense disagreement” (Theiss-Morse 2002, 85) who should concern us the most.

There are many reasons to take this interpretation of citizens seriously. A considerable body of research finds that many people feel uncomfortable with political disagreement (Mutz 2006; Ulbig and Funk 1999). These individuals try to avoid disagreement by gravitating toward those who share their political views (Boutyline and Willer 2017; Colleoni, Rozza, and Arvidsson 2014; Huber and Malhotra 2017) and are prone to conformity when they surround themselves with like-minded people (Carlson and Settle 2016; Levitan and Verhulst 2016; Suhay 2015). This appears to be especially true for partisans in a highly polarized political environment (Lyons and Sokhey 2017). These well-documented tendencies suggest the plausibility of the stealth democracy thesis. If citizens believe that everyone agrees about how to address political problems, it seems reasonable to expect that they would see no reason for disagreement and would want leaders who take decisive action. Further, at a time when the vast majority of Americans believe that politics is becoming increasingly uncivil (Hartig 2018; Berry and Sobieraj 2013), it also seems reasonable to suspect that many citizens would prefer “not to see” the debate and conflict inherent to the democratic process.

Even so, the stealth democracy thesis warrants revisiting. Although scholarship advancing this thesis has argued that citizens want “other-regarding, common-good-oriented” political leaders, this claim, to our knowledge, has never been tested directly. It is one thing to know that citizens who are uncomfortable with disagreement prefer unelected experts who do not need to debate or compromise. However, what these citizens want their leaders to do with their political power is another matter. Perhaps it is the case that citizens who are uncomfortable with disagreement are interested in the “common good” but simply do not appreciate the value of hearing from people with different points of view. Yet, given recent political developments, another possibility calls out for examination. It may be the case that citizens who are uncomfortable with disagreement do not want to avoid it but rather want to have someone aggressively resolve the disagreement in their favor. They may see those who disagree with them as enemies from whom they need protection and may see expedient governance by leaders who share their values as a means to this end. It is to this possibility we now turn.

The Desire for Stealth Democracy, Reconsidered

When scholarship on stealth democracy first emerged in the early 2000s, there was little reason to suspect that citizens who wanted to avoid the messiness of democratic governance also wanted leaders who would deviate from democratic norms. Open-ended interviews with citizens, in fact, revealed evidence of apathy and a distaste for the contentiousness of politics, rather than a desire for leaders who would privilege some groups while oppressing others (Hibbing and Theiss-Morse 2002; Theiss-Morse 2002). Although this early research did not ask about citizens’ willingness to support such leaders, there seemed no pressing reason to ask such questions.

In our present political moment, that is no longer the case. Two possibilities call out for attention, and each suggests the existence of tendencies that are different and darker than a discomfort with disagreement. The first possibility is suggested by scholarship on authoritarian attitudes, which finds that some citizens are more inclined toward obedience and deference to social hierarchies (Feldman 2003; MacWilliams 2016; Stenner 2005). Of course, not all the citizens with these tendencies may want their preferred leaders to use their authority in antidemocratic ways. They may simply place their trust in the abilities of those whom they believe have sound judgment and are qualified to be in positions of authority, which is exactly what expedient governance would accomplish.

However, a considerable body of research finds that citizens with stronger authoritarian attitudes also display a distinctive willingness to support leaders who will protect them and even impose order (Cohen and Smith 2016), particularly if they feel threatened (Hetherington and Suhay 2011). Citizens who are threatened and inclined to defer to authority, in fact, are among those “most willing to want to use force…and limit civil liberties” (Feldman, quoted in Taub 2016; see also Altemeyer 1996; Feldman 2003; and Stenner 2005). Further, Cohen and Smith (2016, 6) argue that “authoritarians—defined by parenting attitudes—do tend to vote for authoritarian candidates—defined by disregard for civil liberties and democratic checks and balances.” Citizens with these tendencies may want expedient governance, but not for the reasons anticipated by stealth democracy scholarship. This suggests the following hypothesis:

H2: Citizens who express stronger authoritarian attitudes will have stronger preferences for expedient governance.

Yet a second possibility warrants consideration. It may be the case that citizens with authoritarian attitudes are not the only subset of citizens who express a preference for expedient governance and a willingness to support leaders with antidemocratic qualities. Citizens who do not have strong authoritarian attitudes may also have the same tendencies, but for a different reason. Many political observers have noted that, at least since 2015, many citizens have expressed a desire for leaders who will protect “people like themselves” from perceived threats to their values and way of life (Galston 2018; Hibbing 2022; Levitsky and Ziblatt 2018; Marchlewska et al. 2018). Perceived threats not only include those from outsider groups that may not have much political power, but also threats from those seen as corrupted political elites, who may have considerable power. This willingness to support politicians who challenge authority figures or political elites is notable. In some cases, citizens may blame elites for enabling threats from outside groups (Galston 2018, ch. 3). Although citizens with authoritarian attitudes tend to follow any form of established authority (Altemeyer 1996; see also Jost 2020), other citizens may only support authorities who promise them protection. This may have an appeal for citizens who have less deference or perhaps no deference to authority (Hibbing 2020; see also Enders and Uscinski 2021; Marchlewska et al. 2018), ² but who want the threats they perceive addressed immediately. Such citizens may not have authoritarian attitudes or be looking for an authoritarian leader in a strict or literal sense, but rather are seeking a warrior for their own in-group.

Recent scholarship has proposed that citizens who are looking for this type of in-group warrior are most accurately described as “securitarians” who are motivated, above all, by a desire to be protected from threats to their own in-group and idealized way of life ³ (Hibbing 2020, 2022). This insight about the desire for protection is corroborated by experimental research that finds that citizens with strong partisan attachments, in particular, demonstrate an affinity for leaders who appear to fulfill this desire. One study finds that affective partisans are willing to award disproportionately high fines to protestors from their out-group and relatively minimal punishment to corrupt members of their in-group (Lelkes and Westwood 2017). There is also evidence that both Republicans and Democrats exhibit a willingness to vote for politicians who belong to their own party, even when those politicians indicate that they will violate democratic norms, such as by prosecuting journalists, banning protests by political rivals, and undermining fair electoral competition (Graham and Svolik 2020).

Importantly, although this may signal that some citizens have a penchant for leaders who assert dominance against an out-group they dislike (Ho et al. 2015, 2012; Kugler, Cooper, and Nosek 2010; Womick et al. 2019), citizens may not be motivated by dominance over others but are simply willing to accept this outcome if it benefits their in-group (Lelkes and Westwood 2017, 496). Whatever the specific motivation, this pattern suggests that some citizens are not only inclined to support leaders who marginalize out-groups but are also willing to support leaders who defy rules and norms for the benefit of their own in-group (Filindra and Harbridge-Yong 2022; Graham and Svolik 2020; Lelkes and Westwood 2017). Expedient governance offers a means to this end, but for a different reason than avoiding discomfort with disagreement. Protecting one’s in-group and its values from out-group threats may provide a stronger motivation for preferring leaders who avoid debate and compromise in favor of decisive action. This suggests the following hypothesis:

H3: Citizens who express stronger preferences for protective leaders will have stronger preferences for expedient governance.

Additionally, the reasons why citizens seek protection in our current political moment—and the type of protection they desire—warrant attention. Political discourse emphasizing protection of an idealized way of life has been especially pervasive among leading figures on the American political Right, with Donald Trump’s “America First” rhetoric serving as the most prominent example (Hibbing 2020, 2022; Jenne, Hawkins, and Castanho Silva 2021; Rowland 2019). As the very notion of “America First” implies, and as Trump’s broader campaign rhetoric also suggests, promises to protect an idealized way of life can be expressed in nationalistic language. Such nationalistic language can also reflect exclusionary impulses grounded in racial and cultural identities, particularly for members of historically dominant groups (Bartels 2020; Jardina 2019; Theiss-Morse 2009; Wong 2010). Although these undercurrents of American nationalism are important in their own right, our focus here is on the kind of leadership and governance citizens are seeking. ⁴ For this reason, we examine whether citizens who desire the protection of their idealized notion of the American way of life are also more inclined to prefer leaders who avoid deliberation and compromise as a means of gaining this type of protection.

Data and Method

To test our hypotheses, we conducted a survey of 1,500 respondents, fielded online in November 2018, using Qualtrics data collection services. In this study, quotas were used to ensure that the sample would be nationally representative with respect to race, gender, income, and education. ⁵ This survey contains measures about citizens’ preferences about governance and a measure of citizens’ comfort with disagreement (Hibbing and Theiss-Morse 2002). It also contains a battery of questions designed to tap into authoritarianism and another one to discern whether respondents would endorse leaders who promise to favor their own group while marginalizing others.

Measuring Preferences about Governance

To understand whether citizens who prefer expedient governance are motivated by a desire to avoid disagreement, authoritarian attitudes, or a willingness to support aggressively protective leaders, we relied on a measure established by the scholars who advanced the “stealth democracy thesis” (Hibbing and Theiss-Morse 2002). This measure includes the following three items. ⁶

1. 1. Our government would be run better if decisions were left up to non-elected, independent experts rather than politicians or the people.

2. 2. What people call “compromise” in politics is really just selling out one’s principles.

3. 3. Elected officials would help the country more if they would stop talking and just take action.

Each item in this battery of questions is designed to assess whether citizens would prefer autonomous leaders making decisions with little, if any, visible debate or disagreement among competing views. For each item, respondents could select whether they “strongly agree,” “agree,” “disagree,” or “strongly disagree.” We then constructed an index, Preference for Expedient Governance, ⁷ using responses to each of these three questions and scaled the variable to range between 0 and 1. The higher the score, the stronger the preferences for expedient governance. ⁸

Measuring Discomfort with Disagreement

For the stealth democracy thesis to be supported, a preference for expedient governance would primarily reflect a desire to avoid disagreement, rather than authoritarian tendencies or a willingness to support protective, and potentially antidemocratic, leaders. For this reason, our analysis used a measure included in previous scholarship about governance preferences (Hibbing and Theiss-Morse 2002): “When people argue about politics you feel uneasy and uncomfortable.”

Respondents indicated whether they “strongly agree,” “agree,” “disagree,” or “strongly disagree” with this item. We constructed the variable such that respondents who strongly agreed (i.e., respondents who are most uncomfortable with disagreement) received the highest scores and scaled the variable between 0 and 1.

Measuring Authoritarian Tendencies

To examine whether and to what extent citizens express authoritarian attitudes, we employed a widely used and well-established measure (Cizmar et al. 2014; Feldman 2003; Feldman and Stenner 1997; Hetherington and Suhay 2011; Stenner 2005). This measure contains four items that ask respondents to assess which of two desirable traits is more important for a child to have: “respect for elders” versus “independence”; “obedience” versus “self-reliance”; “curiosity” versus “good manners”; and “being considerate” versus “being well-behaved.” This measure captures authoritarians’ emphasis on order and control, conformity, and obedience. The authoritarian response in each pair was scored 1 and the non-authoritarian response 0. We then constructed an additive index, scaled between 0 and 1.

Measuring Support for Protective Leaders

To discern whether a preference for expedient governance was motivated by a willingness to support protective, rule-bending leaders, we developed a five-item index (Cronbach’s alpha = 0.81) that examines whether citizens express support for leadership with these characteristics. More specifically, this measure includes references to several actions that previous scholarship suggests protective leaders do, particularly those who threaten democratic norms: attacking and punishing outgroups that some citizens may dislike (Norris and Inglehart 2019), using nationalist rhetoric (Carothers and O’Donohue 2019; Galston 2018; Jenne, Hawkins, and Castanho Silva 2021; Marchlewska et al. 2018), cracking down on independent news media (Ragragio 2021), and bending the rules for their supporters (Bartels 2020; Crimston, Selvanathan, and Jetten 2022). Although using aggressive language toward out-groups and nationalistic language may not necessarily damage democracy itself, such rhetorical tactics are often used by leaders who do attempt to undermine democratic norms (Carothers and O’Donohue 2019; Galston 2018; Jenne, Hawkins, and Castanho Silva 2021; Marchlewska et al. 2018; Norris and Inglehart 2019). Meanwhile, actions like shutting down news organizations and bending rules for supporters constitute departures from a commitment to democracy. To the extent that citizens embrace most or all the practices included in our measure, they embrace a form of protective leadership that can erode democracy.

To implement this measure, we asked respondents whether and to what degree they agreed with each of the following statements:

1. 1. The only way our country can solve its current problems is by supporting tough leaders who will crack down on those who undermine American values.

2. 2. Political leaders must sometimes use rough language to criticize entire groups who refuse to work hard and contribute to our country.

3. 3. To protect the interests of the people they represent, political leaders must sometimes attack the reputations of their political rivals.

4. 4. If political leaders believe that a news organization is attempting to undermine American values, they should take action to shut down that news organization.

5. 5. To protect the interests of people like you, political leaders must sometimes bend the rules to get things done.

For each item, respondents indicated whether they “strongly agree,” “agree,” “disagree,” or “strongly disagree.” We then created an additive index, scaled between 0 and 1, where higher scores indicate stronger levels of support for protective leaders.

Although some may wonder whether using terms like “American values” introduces a potential confound, we argue that it reflects the empirical reality of the leadership preferences we aim to study. Existing scholarship identifies that leaders who promise protection often make nationalistic appeals and that some citizens are especially inclined to see themselves and like-minded others as the “true people” of their nation (Galston 2018; Müller 2016). These appeals mirror other examples of “negative othering” that emphasize threats to national values (Moss 2018). Further, scale reliability is highest when we included all five items.

However, to address concerns that including references to “American values” in our measure introduces a potential confound, we also included a measure of support for protective leaders that includes only two of the items mentioned earlier (Chronbach’s alpha = 0.69):

1. 1. To protect the interests of the people they represent, political leaders must sometimes attack the reputations of their political rivals.

2. 2. To protect the interests of people like you, political leaders must sometimes bend the rules to get things done.

Importantly, both items begin with the phrase “to protect the interests of people like you,” which directly reflects the concept of protection, and neither item includes any reference to country or “American values.” The items are also highly correlated, suggesting that they are tapping into the same underlying construct.

The Preference for Stealth Governance, Reexamined

Democracy, by its very nature, requires citizens to engage in or at least be exposed to debate and disagreement. Of course, some citizens have little appetite for this kind of experience. As past research has demonstrated, they would prefer an expedient form of governance that does not require their oversight, much less their involvement (Hibbing and Theiss-Morse 2002, 2016). The question is whether this merely reflects a desire to avoid the messiness of political competition or whether it reflects something more troubling; namely, a willingness to support practices that undermine democratic norms.

As a preliminary matter, we begin by examining how many citizens express a preference for the kind of expedient governance that scholars advancing the stealth democracy thesis have considered. Following this previous research, we classify citizens as having the strongest preference for expedient governance if they strongly agree (1) that “government would work best if it were run by non-elected experts,” (2) that compromise is just “selling out,” and (3) that elected officials should “stop talking and just take action.” Consistent with the findings of previous research (Hibbing and Theiss-Morse 2002, 2016), Table 1 reports that nearly one in four respondents fit this description: they have both a strong distaste for political conflict and a strong willingness to defer to unelected leaders. The question is why these citizens hold this preference.

Table 1 Prevalence of Stealth Democratic Traits

Source: 2018 Center for Political Participation Survey.

According to scholars who developed the measure of governance preferences used in this study, the preference for expedient governance signaled a desire for stealth democracy or to have a democracy without having to see the messiness of political disagreement among elected leaders. To test this proposition, these scholars examined whether citizens who expressed discomfort with political disagreement were more likely to prefer expedient governance. We replicate this analysis using the same model, with the same construction of the model variables, that appears in the study that originated the concept of stealth democracy. We then compare it to models that include a measure of authoritarian attitudes and our measure of whether citizens support protective, rule-bending leaders.

Table 2 reports a series of ordinary least square regression models that predict support for expedient governance practices. Our dependent variable ranges from 0 to 1, with higher scores indicating stronger preferences for expedient governance. The first model in the table includes a measure for how uncomfortable citizens are with disagreement—the original explanation provided by Hibbing and Theiss-Morse—as well as standard demographic variables. ⁹ The second model adds a measure of authoritarianism, and the third one adds our protective leadership scale. The fourth model includes an interaction between support for protective leadership and discomfort with disagreement.

Table 2 Who Prefers Expedient Governance?

Source: 2018 Center for Political Participation Survey.

Entries are ordinary least squares regression coefficients with standard errors in parentheses. ***p<.001, **p<.01, *p<.05, #p<0.10

Looking first to model 1, citizens who are uncomfortable with disagreement are more likely to prefer expedient governance. The coefficient for the discomfort variable is positive and significant (p < 0.001). It indicates that moving from being the least uncomfortable with disagreement to the most uncomfortable with disagreement is associated with a 0.08 increase (on a 0–1 scale) in support for expedient governance. This is consistent with the size and direction of the result reported by Hibbing and Theiss-Morse (2002, 146). We also find that Republicans are more likely to prefer expedient governance and that those who are white, female, older, and more educated are less likely to favor expedient governance.

Taken on its own, the pattern emerging from this first model suggests that discomfort with disagreement motivates citizens to prefer expedient governance because they simply want to avoid the messiness of political competition. The other models in table 2 complicate this claim. In the second model, we include the measure of authoritarianism. We find that this attitude also predicts support for expedient governance (p < 0.001). Importantly, including a measure of authoritarianism does not diminish the influence of discomfort with disagreement but instead suggests a different pathway to supporting expedient governance. The substantive effect is similar to that of discomfort with disagreement. Moving from the lowest levels of authoritarianism to the highest corresponds with a 0.06 increase in support for expedient governance.

The third model presented in table 2 provides evidence for another, more powerful explanation of citizens’ governance preferences. When we include the measure of support for protective leaders, we find that it is highly predictive of a preference for expedient governance. Although much of the rest of model 3 findings are similar to those of models 1 and 2, it is worth noting that partisanship changes from having a positive relationship with expedient governance to having a negative one. ¹⁰ Additionally, including the protective leader scale diminishes the influence of discomfort with disagreement and of authoritarianism. In model 3, the statistical significance of support for protective leaders is notably stronger (p < 0.001) than discomfort with disagreement (p < 0.05) and authoritarianism (p < 0.10).

More importantly, citizens who indicate the strongest preference for protective leaders exhibit the strongest support for expedient governance. Moving from the lowest level of support for protective leaders to the highest level of support for protective leaders increases support for expedient governance by 0.38. The coefficient for protective leadership is more than four times the size of the effect of discomfort with disagreement in model 1 and six times greater than the effect of authoritarianism in model 2. It is also noteworthy that the model that includes support for protective leaders explains more overall variation in governance preferences (model 3, R² = 0.27) than the models that only include discomfort with disagreement and authoritarianism (model 1, R² = 0.06; model 2, R² = 0.07). We argue that this increased fit indicates that preferences for expedient governance are not merely about avoiding disagreement, as the stealth democracy thesis maintains, or about holding authoritarian attitudes, but rather about doing whatever is necessary to protect yourself and your group.

Additional evidence for this interpretation emerges from the fourth model presented in table 2. This model examines whether support for protective leadership conditions the effect of discomfort with disagreement on governance preferences. The stealth democracy thesis maintains that citizens who are more comfortable with disagreement will be more inclined toward a political process that involves deliberation and compromise and will be less inclined toward decision making by unelected experts. In other words, those who are more comfortable with disagreement will prioritize a deliberative (even conflictual) governance process over an expedient one. Just as importantly, the stealth democracy thesis maintains that citizens who are most uncomfortable with disagreement will prefer expedient governance as a means of avoiding discomfort and not because they are willing to support leaders who will bend the rules or attack groups they find threatening. However, evidence that discomfort with disagreement matters most for citizens who prefer such leaders would undermine the stealth democracy thesis. It would indicate that citizens who are uncomfortable with disagreement are looking for a leader who will protect them by aggressively resolving disagreements in their favor.

The fourth model in table 2 includes an interaction term to examine this possibility. This interaction is both positive and statistically significant (p < 0.001). Given the challenge of interpreting interaction terms (Brambor, Clark, and Golder 2006), we represent this relationship graphically. Figure 1 plots the marginal effect of discomfort with disagreement as a function of protective leadership preference.

Figure 1 Expedient Governance Practices

Note: Data from the CPP 2018 Perspectives on Politics Survey. Plot shows marginal effect of discomfort with disagreement on expedient governance preferences (based on table 2, model 4). Shaded areas denote the 95% confidence interval around each prediction.

As figure 1 illustrates, for those opposed to aggressively protective, rule-bending leadership, discomfort with disagreement does not lead to a stronger preference for expedient governance. According to the stealth democracy thesis, these citizens are expected to prefer expedient governance as a means of avoiding discomfort and not because they are looking for leaders who will give them advantages while attacking rival out-groups. However, contrary to that expectation, when citizens who are uncomfortable with disagreement are not inclined toward such leaders, they prefer governance that is more deliberative and open to compromise.

The effect of discomfort with disagreement changes, however, for those who are willing to support leaders who bend the rules and attack their rivals. Citizens who prefer this type of leadership are much more likely to support expedient governance practices. Support for expedient governance is highest, in fact, among citizens who have a high degree of support for protective leadership and who express discomfort with political disagreement. Ultimately, these individuals want to avoid disagreement by having a leader who would do what it takes to protect them, even if it means bending the rules and acting aggressively toward out-groups they deem threatening. It is also worth remembering two of the practices that citizens who support protective leaders generally endorse: (1) shutting down media outlets deemed to “undermine American values” and (2) “bending the rules” to “protect the interests of people like themselves.” These particular practices reflect a willingness to support not only protective leaders but also protective leaders with antidemocratic tendencies. When individuals who are willing to support such practices are uncomfortable with disagreement, they are especially inclined to favor expedient governance.

To this point, our findings have indicated that although discomfort with disagreement and authoritarian tendencies contribute to a preference for expedient governance, a willingness to support aggressively protective, rule-bending leaders provides a stronger explanation. Further, citizens who desire protective leadership and who are uncomfortable with disagreement are especially inclined toward expedient governance. Of course, as mentioned earlier, a skeptic might raise questions about whether the “American values” discussed in the protective leader measure are driving these results. To address that possibility, we created a second version of our protective leadership scale using the two items that do not invoke “American values.”

In table 2, models 5 and 6 replicate models 3 and 4, swapping this two-item protective leader measure in place of the five-item scale. In model 5, we find that, even using a two-item protective leader measure, support for a protective leader is still large and statistically significant (p < 0.001). Interestingly, the negative effect of partisanship we found in models 3 and 4 is eliminated by using this two-item measure, which suggests that this more limited measure is less subject to the confounding influence of partisanship. However, the general importance of protective leadership is still clear in model 5. The magnitude of the effect is lower (the coefficient in model 5 is 0.28 compared to 0.38 in model 3), as is the variance explained (adjusted R² = 0.23 compared to 0.27). Additionally, in model 5, we find that the coefficient for authoritarianism is twice as large as in model 3. This might suggest that some of the items tapping into a commitment to “American values” may also be supported by those with authoritarian attitudes. However, the strong and independent influence of our two-item protective leader scale suggests that the protective leader scale tapped into something different than an authoritarian tendency. Instead, some citizens are looking for a leader who will protect their interests even if it comes at a cost to democratic norms.

In model 6, we see a similar pattern. Once again, the interaction is still large and statistically significant (p < 0.001). And, as with model 5, the magnitude of the effect is smaller (the coefficient on the interaction is 0.17 compared to 0.20), and the variance explained is less (adjusted R² = 0.23 compared to 0.28). That said, the overall pattern is the same as in model 4. For those with lower levels of support for protective leadership, disagreement is unrelated to their preference for expedient governance. But those who prefer protective leaders and dislike disagreement are among the most likely to support expedient governance. ¹¹ Both models 5 and 6 make clear that this explanation is not merely reducible to a commitment to a notion of “American values.” Instead, those who seek protection from leaders by asking them to bend the rules and attack rivals strongly prefer expedient governance.

Taken altogether then, the findings presented in this study challenge the notion that citizens who prefer expedient governance are stealth democrats who simply want to avoid the disagreement that generally accompanies political competition. Citizens who have the least support for protective leaders do not demonstrate an inclination toward expedient governance, even when they are very uncomfortable with disagreement. They favor deliberation and compromise, even when the idea of political disagreement makes them personally uncomfortable. Meanwhile, the citizens most likely to favor expedient governance are both uncomfortable with disagreement and broadly supportive of a leader who offers them aggressive protection, even if it comes at a cost to democracy. These citizens may not be stealth democrats at all because they demonstrate support for characteristics in a leader that are at odds with democratic norms and outcomes.

The Desire for Protection and the Undermining of Democracy

Concerns about the health of American democracy have become both widespread and deeply felt. Scholarship on “stealth democracy” suggests one account of how this problem can emerge. If many citizens want to avoid the debate and disagreement that come with democratic politics—if they want a democracy that they do not have to see—it can undermine the conditions necessary for democratic accountability. If enough citizens prefer not to see ardent disagreements that often characterize political debates and if they prefer expedient decision making, they may forsake the reflexes necessary to hold representatives accountable if their expedience leads to antidemocratic outcomes.

Avoiding disagreement, however, is not the only motivation or even the primary one that can threaten democratic norms and governance. Our findings show that there is another significant motivation underlying the apparent desire for stealth democracy: a desire for leaders who will aggressively protect their idealized way of life and even bend the rules for people like themselves. Notably, this pattern is not limited to citizens who have authoritarian attitudes. Although such citizens do have a tendency to prefer expedient government, our measure of support for protective leaders is a stronger predictor of this preference. This stems from the fact that many citizens who lack authoritarian tendencies are still interested in leaders who bend the rules for them and act aggressively toward out-groups that they believe are threatening.

This points to a different explanation for why citizens may prefer the efficiency of expedient governance. Many citizens who want expedient governance do not appear to be invested in enabling “other-regarding, common-good-oriented…empathetic, non-self-interested decision makers” (Hibbing and Theiss-Morse 2002, 61). Instead, they appear willing to enable leaders who will protect their own in-group and advance its interests, even if that means bending the rules and cracking down on rivals.

It is also telling that the strongest inclination toward expedient governance is expressed by individuals who want to avoid disagreement and who also have the strongest preferences for aggressively protective leaders. Those who feel uncomfortable with disagreement may not want to “see” the messiness of political competition, as scholarship on stealth governance preferences has long argued. However, our findings point to the conclusion that these citizens want not only to avoid disagreement but also want a leader who will end the disagreement in a way that benefits them and disadvantages others who they find threatening.

Further, these citizens express a willingness to support leaders who would embrace antidemocratic governance practices. In this study, we do not only ask people whether they would be willing to support leaders who aggressively criticize political rivals and groups that “undermine American values.” We also ask whether people will support leaders who shut down media outlets, which by implication, violates the democratic right to freedom of political expression. Additionally, we ask about whether people will support leaders who will “bend the rules” to “protect the interests of people like you.” This indicates a willingness to permit departures from the rule of law, which should also concern proponents of democracy. Citizens seeking protective leaders generally support both these practices, and ultimately, these citizens are also among the most supportive of an expedient governing process, particularly when they are uncomfortable with disagreement. This signals a different challenge to democracy than previous scholarship on stealth democracy emphasized. Although citizens who support aggressively protective leadership may not consider themselves antidemocratic, they may nonetheless be susceptible to supporting practices that can undermine democratic norms. In this way, the desire for protection—which appears to be a strong and recurrent desire in political life (Hibbing 2020; Smith 2018)—can pose a threat to the health of democracy.

Two other matters, however, remain unresolved. First, leadership that undermines democratic norms might come in many different forms. This study examines citizens who seek protection for their version of “American values” because prominent politicians have frequently used such language in recent years. Of course, this sort of nationalistic language is associated with politicians on the political Right (Bartels 2020; Hibbing 2020, 2022; Jenne, Hawkins, and Castanho Silva 2021; Rowland 2019). Yet recent research emphasizes that the political Right does not have a monopoly on authoritarianism (Conway et al. 2018; Wronski et al. 2018), and more attention should be devoted to the circumstances under which those on the Left might be likely to support leaders with antidemocratic tendencies (but see Graham and Svolik 2020). Our data do not allow us to explore this possibility, but it is an important area of future work.

Second, our findings emerge in data from 2018. The original studies about preferences for stealth democracy relied on data from the late 1990s (Hibbing and Theiss-Morse 2002). For this reason, we cannot know whether citizens who were classified as “stealth democrats” in the 1990s were really citizens who would have embraced leaders with antidemocratic tendencies. It may be the case that they were, but it may also be the case that our findings from 2018 reflect more recent political developments.

Since 2018, we have witnessed a sitting US president—Donald Trump—attempt to undermine confidence in a free and fair election by making explicit and baseless assertions of systemic election fraud (Eggers, Garro, and Grimmer 2021). In the immediate aftermath of the 2020 presidential election, elected officials of the president’s party actively supported efforts to overturn the results of this election (Beavers and Wu 2021; Corasanti, Epstein, and Rutenberg 2020). On January 6, 2021, we witnessed an insurrection at the Capitol as Trump supporters marched from an event where President Trump addressed them and stormed the chamber in which members of Congress were counting electoral votes. The president never condemned them and called them patriots (Bump 2021; Cohen 2021). Even months later, polls continued to find that majorities of Trump supporters and Republicans more generally believed that the election was stolen, and roughly half of Republicans viewed the pro-Trump insurrection that attempted to overturn the 2020 presidential election as a “legitimate protest” (Monmouth 2021; PRRI 2021). A full 30% of Republicans described those who participated in the insurrection as “patriots” on the very day the insurrection occurred (Smith, Ballard, and Sanders 2021).

In hindsight, the patterns we report in this study—based on data we collected in 2018—foreshadow these events. They are also consistent with other scholarship that has chronicled the foment that led to the ascendance of Trump as a presidential candidate in 2015 and 2016 (Enders and Uscinski 2021; Rudolph 2021). The attitudes that helped fuel Trump’s rise were not obvious when path-breaking scholarship on stealth democracy was first conducted in the 1990s under very different political conditions. Perhaps if political conditions in the future begin to resemble those of the 1990s, a preference for expedient governance would once again reflect a desire for stealth democracy. Yet under present conditions, it seems imprudent not to view the preference for expedient governance as reflecting something different and darker.

No matter whether enduring or new, these patterns are cause for concern and cast the desire for expedient governance in a new light. The promise of democracy is to enable a free and fair competition over ideas not only during elections, but also during the governing process. Yet some citizens are willing to trade the promise of democracy for the promise of protection without even recognizing what has been lost. Simply put, a desire for expedient governance may not signal that citizens are “stealth democrats,” but rather that citizens are not sufficiently committed to democracy itself.