A year into the Trump Administration, the health and stability of American democracy remain an open question. At a time when almost four in 10 Americans say they are not satisfied with the way democracy is working in the U.S., there is ample reason to ask how committed the American people are to our democracy.
While understanding dissatisfaction with our political system is quite important, knowing the degree to which Americans are open to actual departures from democracy has a deeper, more existential value at a time when populist political leaders and parties have gained significant momentum around the globe. Taking advantage of a battery of questions included in the July 2017 VOTER Survey (Views of the Electorate Research Survey) from the Democracy Fund Voter Study Group, we seek to understand more about these contemporary Americans who are not only questioning democratic norms but also affirmatively endorsing authoritarian alternatives.
Specifically, the Democracy Fund Voter Study Group asked respondents to assess the favorability of three types of political systems:
Our survey then asked respondents how important it is for them to live in a country that is governed democratically, as well as their level of satisfaction with American democracy. We also asked whether democracy is always preferable or whether there are some circumstances in which nondemocratic government can be preferable. Additional questions about checks on executive authority and support for living in a pluralist society were also asked and will be addressed in forthcoming reports.
The VOTER Survey deployed by the Democracy Fund Voter Study Group asked people to assess the broad concept of “democracy” without asking them what they meant by the term. This replicated questions previously asked by the World Values Survey. Unfortunately, this means that we have limited information about how respondents defined democracy and what they had in mind when answering questions about it. Rather, the battery of questions defines democracy only insofar as the VOTER Survey implicitly contrasts it with alternatives such as military rule. Future qualitative research will provide more insight on what respondents mean when they respond to these types of questions. Nevertheless, we take support for a “strong leader” and “army rule” as worrisome signs of sympathy for unchecked executive power.
Among our most important findings are:
If given a direct choice, the overwhelming majority of Americans choose democracy. In fact, on each of the five questions we asked, three quarters or more of all respondents provide at least some support for democracy, and half or more express support for the strongest pro-democratic option. By contrast, depending on the question, between an eighth and a quarter of respondents provide an answer that does not support democracy. Moreover, we find evidence that conflicts with two key findings(i) that have recently raised alarm bells about the state of democracy: (a) We do not find that public support for democracy in the U.S. is declining. (b) Nor do we find higher support among young people for an authoritarian political system.
However, only a slim majority of Americans (54 percent) consistently express a pro-democratic position across all five of our measures. If we look across our battery of questions, almost half of our respondents do not support democracy on at least one of the five survey questions. This pattern indicates that the overall high percentages on each question may mask some deeper softness in support for democracy. Nineteen percent of respondents express one nondemocratic position, 13 percent express two nondemocratic positions, and 15 percent express three or more nondemocratic positions. Notably, 29 percent of respondents show at least some support for either a “strong leader” or “army rule.”
A majority of those who rate democracy poorly or say that it is not essential to live in a democracy still do not support an authoritarian alternative. Among those who rate democracy unfavorably, believe it is not so important to live in a democracy, or do not believe it is always preferable, just over half are opposed to a “strong leader” and “army rule.” This finding clarifies that dissatisfaction with democracy does not necessarily translate into openness to authoritarian appeals, though there is significant overlap among the two views.
Comparing supporters of different candidates in the presidential primaries, the highest level of openness to authoritarian political systems is among those voters who supported Donald Trump in the primaries. Thirty two percent of Trump primary voters support a “strong leader.” The level of support for this option is especially high (45 percent) among those who voted for Barack Obama in 2012 and then switched parties to vote for Donald Trump in 2016. Perhaps more unexpectedly, 20 percent of Hillary Clinton’s primary voters support a “strong leader” unbound by Congress and elections and 15 percent go so far as to support “army rule” — both slightly higher than the levels expressed by the primary supporters of Bernie Sanders, John Kasich, Marco Rubio, or Ted Cruz. (Yet Clinton’s primary supporters were more likely to say that democracy is preferable to any other form of government.)
Viewed through an ideological lens, the highest support for democracy comes from respondents who are either consistently liberal or consistently conservative. In contrast, more than half of those who hold both economically liberal and culturally conservative views support a “strong leader” who does not have to bother with Congress or elections.
The highest levels of support for authoritarian leadership come from those who are disaffected, disengaged from politics, deeply distrustful of experts, culturally conservative, and have negative attitudes toward racial minorities. Those who consume news less frequently are 22 points more likely to support a "strong leader" than frequent news consumers. Similarly, nonvoters are about 10 points more likely to express support for a “strong leader.” Cultural conservatives are 20 points more likely than the culturally liberal to support a “strong leader.” Those who say it is fairly or very important for someone to have European heritage to be an American are 30 points more likely to support a “strong leader” than those who decisively reject this racial conception of national identity. Those who are mistrustful of experts are 25 points more likely to support a “strong leader” compared to those who are trusting of experts.
Many scholars have observed that authoritarian proclivities among the American public are nothing new. Indeed, there is ample reason to believe that a significant number of Americans have always held a collection of nativist, racist, and/or authoritarian views. While this may be true, we are facing several new circumstances in our society that make these questions of public support for our norms and institutions more urgently relevant.
Donald Trump and the Emergence of Illiberal Populist Demagoguery
Much has already been written about the trans-Atlantic emergence of populist candidates and parties who have either succeeded in getting elected or have come quite close. Donald Trump and Brexit have perhaps gotten the most attention, but a broad slate of liberal democracies have experienced dramatic electoral gains by right-wing, illiberal parties riding a wave of anti-immigrant populism, from Hungary and Poland to Austria, the Czech Republic, and even (to a lesser but still alarming extent) France, Sweden, and Germany. There is reason to believe that these nativist parties and movements are learning from one another and at times communicating with one another as part of a larger global network.
Countless analyses have documented how leaders such as Donald Trump have actively undermined democratic norms and demonstrated authoritarian tendencies that we long thought were off limits in mainstream American political discourse. Given considerable evidence that the public responds to cues from political leaders, it is important to pay close attention to how these types of messages are resonating with the public and with whom they are resonating.
Rapid Social Change: Race, Immigration, Globalization, and Inequality
Liberal democracies have been shaken by growing objective stresses due to immigration, globalization, rising income inequality and insecurity, and the economic displacement caused by the 2008 financial crisis. There is rising talk not only in the U.S. but also across Europe and even globally that democracy is not working well to address key policy challenges, and this has left voters feeling increasingly alienated. Trust in major institutions has been low and declining for some time now across many advanced industrial democracies.
The U.S. has witnessed rapid demographic change and now has its highest share of foreign-born residents since the 1920s. These demographic changes and other important developments, including the election of the country’s first African-American president, have brought to the surface anxieties and racial animus among some white Americans.
Close Competition Between Two Hyperpartisan Political Parties
Our two dominant parties — more ideologically distant from one another and more geographically sorted than at any time in at least a century — closely compete for control of government and equate losses with threats to the survival of the nation. We see the weakening of key master norms such as “mutual toleration” (in which each party accepts the basic legitimacy of its opponent) and “institutional forbearance” (in which political leaders responsibly wield the power of the institutions they are elected to control).(ii)
In this context, political incentives make it difficult for the few remaining moderate voices to call out extreme elements of their party and to draw bold lines that cannot be crossed by leaders or constituents. Hyperpartisanship — which has turned political party affiliation into something akin to tribal identity — makes political compromise difficult to impossible.
The Disruptive Influence of Technology on Our Democracy
Technological disruption across our society has had positive and negative effects on our political system. Important and concerning context for this inquiry includes:
The enhanced ability of fringe ideas and groups to spread across social media and help otherwise isolated voices to create networks and find allies;
Media and online echo chambers that reinforce pre-existing points of view, deepen divisions, and enable misinformation to travel unchecked and “go viral”;
The decline of gatekeepers within media and political parties that used to play a significant role in keeping illiberal ideas out of the public conversation; and
New tools that enable populist candidates to build an unmediated following, ignite passions, and scapegoat individuals or minority groups.
Foreign Interventions Aimed at Undermining Democracy
Finally, with each news cycle we are learning more about efforts by Russia and its agents to intensify political polarization and undermine confidence in the American political system, particularly through highly orchestrated and automated interventions in social media. This follows a pattern that has been seen in democracies across Western and Eastern Europe over the past decade. In this context, it is important to understand how attributes like low news interest may relate to authoritarian attitudes, as these voters may be especially vulnerable to targeted disinformation via social media.
Data collected for this report by the Democracy Fund Voter Study Group through its 2017 VOTER Survey represent a unique longitudinal sample that offers important insight into the attitudes and behaviors of American voters. By focusing on a group of 5,000 respondents who had previously been interviewed several times over the past six years by the survey firm YouGov, we can reliably track how vote preferences and other attitudes are associated with support for democratic norms and values.
As part of the Cooperative Campaign Analysis Project, more than 35,000 respondents were initially interviewed three times between December 2011 and December 2012. Many of these respondents were reinterviewed by YouGov in July 2016 and then 8,000 were reinterviewed for the Democracy Fund Voter Study Group in November 2016 for its first VOTER Survey. A sample of 5,000 people from this group was then interviewed again in July 2017 for this report providing longitudinal data from six surveys in all.
The size of the 2017 VOTER Survey’s sample offers significant statistical power for this research, while the longitudinal nature of the survey enables us to see how respondents have changed their views over time. However, the longitudinal panel does present some limitations. Most notably, young people are underrepresented in the sample. The youngest people interviewed by YouGov in late 2011 and 2012, when the panel was formed, were in their mid-20s by the time this survey was conducted in July 2017.
I: While each of our five questions draws a pro-democracy response from at least three-quarters of respondents, only half of the public is consistently supportive of democracy across all five indicators.
The July 2017 VOTER Survey specifically asked respondents five questions about their attitudes toward democracy. How would they assess on a four-point scale of “very good” to “very bad”:
Having a strong leader who does not have to bother with Congress and elections?
Having the army rule?
Having a democratic political system?
Respondents were also asked to rate:
How important is it for you to live in a country that is governed democratically? (on a scale of one to 10).
Which of these statements is closest to your view?
Democracy is preferable to any other kind of government.
In some circumstances, a nondemocratic government can be preferable.
For someone like me, it doesn’t matter what kind of government we have.
The latter three of these items probe in different ways respondents’ attitudes toward democracy without defining the term. It is therefore possible for different people to give the same answer while imagining different things. For example, some may have a narrow majoritarian view of democracy as simply people electing a leader, while others consider a democracy to also require strong checks and balances, rule of law, and civil liberties. This may account for the fact (as we explain below) that the first two items — which test people’s willingness to countenance authoritarian forms of rule — are rather independent of the latter three.
Moreover, while “army rule” is fairly explicit, the strong leader question may elicit seemingly authoritarian responses from some respondents who “only” want to empower an elected leader to override Congressional opposition and get things done. This is a question we hope future qualitative research might help to clarify.
We report the top-line findings for each of these questions below.
In each of these five questions, three quarters or more of all respondents provide at least some support for democracy with a majority expressing support for the strongest pro-democratic option. By contrast, depending on the question, between an eighth and a quarter of respondents give an answer that is not supportive of democracy1—with 29 percent of respondents showing at least some support for either a “strong leader” or “army rule.”
1 For purposes of questions three through five here (opinions on democracy), we consider skipped responses as refusing to give the pro-democracy position, and therefore consider nonresponses as expressing lack of support for democracy. In both cases, only about 1 percent of respondents skipped the question.
These findings closely mirror the findings of a recent survey by the Pew Research Center (see Appendix A). On three of the first four measures, they also represent some improvement over the levels of democracy support found in the two most recent World Values Surveys conducted in the U.S. in 2006 and 2011 (the fifth item was not asked in any previous World Values Survey). As a result, democratic orientations in 2017 have risen closer to the levels previously recorded in such other advanced liberal democracies as Germany and Australia (see Appendix B).2
Interestingly, there is a strong relationship between responses to the two authoritarian questions, but responses to these two questions are more weakly related to views of democracy. Half of respondents who support a “strong leader” also support “army rule” and two thirds of army-rule supporters also prefer a “strong leader.” By contrast, solid majorities of these supporters of authoritarian rule don’t think democracy is a bad system.
Similarly, narrow majorities of respondents who don’t think democracy is a good system are opposed to both a “strong leader” and “army rule.” They might not like what they have now, but they aren’t rushing to support authoritarian leadership. (For full details about the relationship among different responses, see Appendix C.)
If we combine responses across answers, almost half of our respondents fail to be supportive of democracy on at least one of the five survey questions.3 Only a slim majority (54 percent) is consistently supportive of democracy across all five questions.
2 It’s worth noting, of course, that the World Values Surveys and our VOTER Survey were not conducted identically. Most notably, the VOTER Survey used longitudinal data from a group that has been resurveyed several times over the past few years, which has meant that this is a more active/engaged group then was found within the World Values Survey dataset. This makes comparing the data over time more difficult.
3 For the purpose of this report, failing to support democracy is defined by one or more of the following: (a) Belief that a “strong leader” is fairly or very good, (b) belief that “army rule” is fairly or very good, (c) belief that a democratic system is fairly or very bad, (d) rating the importance of living in a democratic system below an eight on a scale of one to 10, and/or (e) belief that democracy is not always preferable to other systems.
This pattern indicates that the overall high percentages on each question may mask some deeper softness in the support for democracy. Indeed, it is troubling to consider how many Americans may be open to illiberal appeals at a time when foreign and domestic actors are actively seeking to subvert our democratic institutions.
For the rest of the analysis, we will narrow our narrative and graphical focus to two of these five questions, the “strong leader” question and the “preference for democracy” question. We do this for three reasons:
First, for sake of narrative ease, it is easier to follow two questions rather than five.
Second, since these two questions have the most discouraging results, we want to understand them better.
Finally, responses for the “strong leader” question and the “army rule” question are highly correlated, probably tapping an underlying authoritarian rule dimension. We go deeper on the “strong leader” question because it is more relevant to our current politics, since army rule seems to be a more remote possibility.
It is also worth noting that our survey instrument also includes a battery of questions related to public support for various checks and balances. We will report on the results of these questions in a separate companion report.
II: Support for a strong leader has declined after a 20-year increase.
The World Values Survey began asking identical questions about public support for alternatives to democracy in the 1995–1998 wave, repeating these questions in subsequent surveys since. Over this 20-year period, Americans showed increasing support for authoritarian alternatives to democracy. While 24 percent supported a “strong leader” in 1995, this steadily increased to 34 percent in 2014. We observe similar increases across other questions from the World Values Surveys assessing democracy and democratic alternatives.
However, the upward trend reversed in 2017 with regard to favorability for a “strong leader” and negative views of democracy (Figure 7). Most notably, support for a “strong leader” in 2017 returned to 1995 levels at 24 percent. The Pew Research Center reached a similar conclusion, finding 22 percent in support of a “strong leader” in its 2017 survey (Appendix A).4
4 An alternative explanation, of course, is that this shift in a two-decade trend is simply an artifact of a different survey methodology and dataset. While we reproduced the same questions used by the World Values Survey, we use a longitudinal dataset that has been reinterviewed over time and is relatively politically engaged. Weighting of the sample helps with this, as does the fact that Pew’s findings are very similar to ours. But as can be seen in the Appendix, our sample is slightly more positive about democracy then Pew’s. At a minimum, this finding calls for further investigation and data collection.
What could explain this shift? One possibility is that Donald Trump has personified authoritarian leadership in a way many Americans found distasteful. In the 1995 World Values Survey, Democrats registered slightly higher levels of support for a “strong leader” (23 percent of Democrats compared to 21 percent of Republicans supported a “strong leader”). Democrats continued to show greater support for a “strong leader” over the three subsequent World Values Surveys in the U.S. in 1999, 2006, and 2011. Then, in 2017, the balance shifted. Republicans (especially Donald Trump primary supporters and those who switched from Obama to Trump) were notably more supportive of a “strong leader” than Democrats. In 2017, 19 percent of Democrats said a “strong leader” was a good thing, compared to 30 percent of Republicans.
A similar trend can be found among young voters, who are disproportionately opposed to Donald Trump. Over the four previous World Values Surveys, support for a “strong leader” progressively increased among younger cohorts of voters, a trend that several analysts have seized on with alarm. Our 2017 findings, however, show this trend has reversed. Younger people in 2017 were actually most opposed to a “strong leader” in our survey (although as we explain below, our sample excludes the very youngest voters). Only one in ten of those under 30 years of age support a “strong leader,” compared to roughly a quarter of older cohorts who favor a “strong leader” system (Table 1).
III. Support for democracy is lowest among the least educated and least politically engaged.
One important factor contributing to support for democracy is education. As Figure 6 shows below, more education is associated with reduced support for a “strong leader.” Roughly one in three respondents who haven’t gone to college (32 percent) say a “strong leader” is good, but only one in eight respondents with a B.A. or equivalent four-year college degree (13 percent) want a “strong leader.” By contrast, support for democracy as a system increases only modestly with more education. More than one in six Americans with a BA degree (17 percent) do not express a preference for democracy, as compared to one in four noncollege educated Americans (25 percent) who do not express a preference for democracy.
Another factor related to support for democracy is political interest: Those who are politically engaged and follow the news closely are more likely to express a preference for democracy and are less supportive of a “strong leader” who does not have to bother with Congress and elections.
We measure political interest by asking respondents how much they follow political news. Those who follow the news “most of the time” (53 percent of respondents) exhibit much more pro-democratic attitudes than those who follow the news some of the time (25 percent), only now and then (12 percent), or hardly at all (7 percent).
Those who are not regular consumers of news are less committed to democracy. They are more than twice as likely as frequent news consumers to say democracy is not preferable (31 to 14 percent) and almost three times as likely to support a “strong leader” (35 to 13 percent). Presumably, those who follow political news closely do so because they feel invested in the political system.
We can probe the effect of political engagement more deeply by comparing attitudes of voters and nonvoters. By matching our respondents with voting records, we can estimate who actually voted. In the 2016 election, 56 percent of voting-age citizens cast ballots. In our weighted sample, we estimate that 60 percent of respondents voted.5
Nonvoters hold more negative attitudes about democracy. Nonvoters were nine percentage points more likely than voters to say a “strong leader” is good (29 vs. 20 percent) and to say democracy is not preferable (28 vs. 19 percent).
5 We base this estimate on matches to voter files. For purposes of our analysis, we assume that all nonmatches are also nonvoters.
We also look here at age, since a hotly debated question is whether the young are turning away from democracy, as Roberto Foa and Yascha Mounk have argued.(iii)
Younger respondents are less likely than older ones to say they prefer democracy. But they are also less likely to say they prefer a “strong leader,” as compared to older voters. However, our sample excludes the very young (18–22 years old), which may contribute somewhat to the differences in our findings from the above study.6
6 Because we have panel data going back to 2011, our youngest respondents are now 23. The 23–29 cohort is only 6.5 percent of our weighted sample, with only 93 cases. Using survey weights, the 30–44 cohort makes up 25.9 percent of our sample, the 45–64 cohort 42 percent, and the 65+ cohort 25.9 percent.
Based on this snapshot, we cannot address the larger debate over whether younger voters’ lower preference for democracy is a warning sign or a predictable life-cycle effect, as Erik Voeten has argued.(iv) However, these diverging responses on questions of strong leadership and preference for democracy do help clarify that these two questions may be tapping different underlying sentiments.
IV: Conservatives are less supportive of democracy, especially cultural conservatives.
In looking at ideology, we observe a consistent pattern. Those who identify as liberals are much less likely to say a “strong leader” is good, but only slightly more likely to express a preference for democracy.
First, we break down sentiments about democracy by self-identified ideology. Support for a “strong leader” is more than twice as high among conservatives (30 percent) than it is among liberals (13 percent). However, the nonpreference for democracy is only slightly lower among liberals (17 percent) than conservatives (22 percent) and moderates (also 22 percent).
We also looked at the 6 percent of respondents who did not place themselves on this ideological scale. They are substantially less pro-democratic in their attitudes. This is also evidence that people less engaged by politics are less likely to support democracy.
This finding is also consistent with the 2017 Pew global survey, which found that respondents on the right in 10 advanced democracies were generally two or three times as likely as those on the ideological left to support a “strong leader.”
We can further distinguish between cultural and economic conservatives. In an earlier Democracy Fund Voter Study Group report, “Political Divisions in 2016 and Beyond,” one of us (Drutman) constructed separate economic and cultural (social/identity) indexes from multiple survey questions. Here we break respondents into three equally sized subgroups based on the indexes and label them liberal, moderate, and conservative based on their score. The cultural index combines both social conservatism (views on abortion, gay marriage, and transgender bathrooms) and racial attitudes (specifically toward immigrants, African-Americans, and Muslims).
The highest levels of negative sentiment about democracy come from those who express more culturally conservative views. Strong liberals express the most consistently antiauthoritarian leader views.7 In particular, both cultural conservatives (31 percent) and cultural moderates (29 percent) are nearly three times as likely as cultural liberals to favor a “strong leader.” However, we do not observe any meaningful pattern with regard to economic liberalism/conservatism, so we do not report those results here.
These patterns are consistent with long-standing findings that those who hold the most progressive values on social issues tend to also support the kinds of self-expression values that underlie pro-democratic attitudes. By contrast, those who hold more traditionalist values tend to hold authority and hierarchy in higher regard, often in opposition to prodemocratic values of self-expression.(v)
7 These attitudes appear to be driven by both moral and racial attitudes (in regression analysis, both are independently significant). However, since moral and racial attitudes are highly correlated, it is difficult to disentangle them entirely.
V: Those with negative views toward racial minorities and cultural diversity are more supportive of a “strong leader” and less likely to prefer democracy.
We also compare sentiments based on responses to a few telling questions about attitudes toward racial minorities and cultural diversity. First, we asked respondents how important being of European heritage is to being American. While most respondents think it is not at all important (47 percent) or not very important (27 percent), 9 percent say it is fairly important and 8 percent say it is very important. An additional 9 percent say they “don’t know.”
More than 40 percent of those who think European heritage is very or fairly important to being an American are supportive of a “strong leader” — fully four times the proportion who say European heritage is “not at all important.” The pattern holds in somewhat less striking fashion with regard to democracy support: Those who think European heritage is very or fairly important are more than half again as likely (28–30 percent) to not prefer democracy as those who think such heritage is not at all or not very important (18 percent).
There is a similar pattern related to cultural diversity. When asked, 56 percent of respondents say that it is better to have “many different cultures with different values and traditions that people believe in” while 31 percent say we’d be better off with “one primary culture with traditions and values that most everyone believes in.” Those who would prefer one primary culture are almost twice as likely to say democracy is not preferable (27 percent) as those who think it’s better to have many cultures. But the highest rates of those responding negatively about democracy are among the 14 percent of the respondents who say they “don’t know” which is better. Many of those giving the “don’t know” response to this question may prefer one primary culture but feel that this is a socially undesirable response.
Respondents were also asked if there should be increased “surveillance of mosques and other places where Muslims may congregate” and whether we should “target Muslims at U.S. airport screenings to ensure the safety of flights.”
In both cases, those who expressed higher levels of distrust of Muslims were considerably more likely to support an authoritarian leader and were also less likely to prefer democracy. Most notably, those who wanted to increase surveillance on mosques (Figure 14) were more than twice as likely to support a “strong leader” than those who did not want to increase surveillance on mosques (and the ratio was about three to one compared to those who strongly opposed increased surveillance).
Other measures of attitudes toward racial minorities produce similar patterns. Regardless of the measure, those who express more negative attitudes toward racial minorities are consistently more supportive of a “strong leader” and less likely to prefer democracy.
We also investigate the relationship between authoritarian parenting style and negative sentiment about democracy. During the rise of Donald Trump, there was considerable commentary noting that “authoritarian” attitudes (an indicator based on a four-question battery about parenting styles) were particularly high among Trump supporters.(vi)
Respondents who score high on the authoritarian parenting score are twice as likely (31 percent) to support a “strong leader” than those who score low on the authoritarian parenting score (16 percent). However, “authoritarian” parenting attitudes certainly do not indicate support for “authoritarian” leadership, since more than two thirds of respondents with authoritarian parenting attitudes reject a “strong leader.”
VI: Support for democracy is weakest among those whose views are not ideologically consistent.
Combining the cultural and economic indexes described above yields an important insight. The highest support for democracy comes from respondents who are either consistently liberal or consistently conservative.
Among consistent liberals, only 4 percent want a “strong leader,” while 13 percent decline to say democracy is preferable. Consistent liberals, thus have the most consistently prodemocratic attitudes of all subgroups. Consistent conservatives are notably more proauthoritarian (19 percent want a “strong leader”) and slightly more unsupportive of democracy (17 percent don’t prefer democracy).
By contrast, the lowest support for democracy comes from those who are in the “off-dimension” position — especially those who are economically liberal and culturally conservative. This group formed the core of the Obama to Trump voters, and they are also a hotbed of sentiments that are unsupportive of democracy. A remarkable 52 percent of these voters say a “strong leader” would be good and 40 percent do not prefer democracy. Those who are economically conservative but socially liberal (“libertarians”) also say they don’t prefer democracy at very high rates (37 percent), though they are closer to the average on support for a “strong leader” (28 percent). However, as Drutman’s scatter plot of the electorate in the “Political Divisions” report showed, only a small percent of the electorate fits into the economically conservative/culturally liberal (“libertarian”) quadrant.
It is also notable that consistent moderates harbor higher than average sentiments that are unsupportive of democracy, with 34 percent saying they want a “strong leader” and 30 percent indicating that democracy is not preferable.
What could explain these patterns? For one thing, those who are most consistently liberal or consistently conservative are also generally the most politically engaged, and this tends to correspond to higher support for democracy. People hold consistent political views because they consume more political information and know what they “should” think. A longstanding finding in political science is that the most ideologically “consistent” voters are the most highly engaged voters and also therefore the most polarized. And generally, those who are most affiliated with the parties tend to be most supportive of the overall political system.vii Indeed, political scientists have long viewed highly engaged and informed partisan activists to be the strongest supporters of democratic ideals.8 By contrast, voters who follow politics less closely are less ideologically “consistent.” They know less about what they “should” think on the issues.
Something similar may also be going on with moderates, since there is some evidence that individuals who give more middle-of-the-road responses in surveys tend to be less informed politically.(vii) And as we showed earlier in this report, those who are less informed politically tend to have more doubts about democracy.
It is also quite possible that voters who hold “off-dimensional” views have checked out of the political process because they feel that neither party represents them particularly well, given that their views do not cleanly fit within either party. This may also be the case with “moderate” voters whose moderation results from lack of political engagement. If these two groups of voters feel as though our democratic political system doesn’t represent them well, it may seem reasonable to conclude that democracy isn’t such a great system.
We can observe some related patterns by comparing support for democracy among different types of primary voters. Here, we use our longitudinal survey, which asked respondents in June 2016 who they voted for in the presidential primaries.
Comparing supporters of different presidential primary candidates, the highest levels of skepticism about democracy are found among Trump primary supporters. Twenty-three percent of Trump primary voters say democracy is not preferable and 32 percent say a “strong leader” is good.9 Core Trump supporters tended to be less educated and more politically disengaged, and they also tended to hold economically liberal but culturally conservative views, which have not been well-represented by the two-party system.
8 For example, McCloskey et al. (1960) describe partisan activists as distinct in “their strong approval of democratic ideas, their greater tolerance and regard for proper procedures and citizen rights, their superior understanding and acceptance of the rules of the game and their more affirmative attitudes toward the political system in general . . . the evidence suggests that it is the articulate classes rather than the public who serve as the major repositories of the public conscience and as the carriers of the Creed. Responsibility for keeping the system going, hence, falls most heavily upon them.” (Our italics.) Herbert McClosky, Paul J. Hoffmann, and Rosemary O’Hara, “Issue Conflict and Consensus Among Party Leaders and Followers,” The American Political Science Review, vol. 54, no. 2, June 1960, pp. 406–27, Accessed January 17, 2018. Available at: https://www.cambridge.org/core/journals/american-political-science-review/article/issue-conflict-and-consensus-among-party-leaders-and-followers1/A9FE35C4BD8175AB607EE8C3B75719ED.
9 This pattern bears a striking resemblance to the 2017 Pew Research Center findings that support for a strong leader “who can make decisions without interference from parliament and the courts” is, in Britain, much higher among supporters of the anti-immigrant UKIP party (42 percent), and in Italy, much higher among supporters of former Prime Minister Berlusconi’s Forza Italia (43 percent) than it is for other parties. Source: http://www.pewglobal.org/2017/10/16/democracy-widely-supported-little-backing-for-rule-by-strong-leader-or-military/.
More surprising perhaps is the observation that one in five supporters of Hillary Clinton from the Democratic primary express support for the “strong leader” option. This was somewhat higher than was found from the supporters of Marco Rubio, Ted Cruz, John Kasich, and Bernie Sanders (although her supporters are generally more inclined to say that democracy is preferable).
It is interesting to note that while both attitudes about democracy (pro-strong leader and democracy not preferable) are distinguishing characteristics of the anti-establishment candidate in the Republican primary (e.g., Donald Trump), only the democracy-not preferable option is characteristic of supporters of the anti-establishment (e.g., Bernie Sanders), who seem to be quite suspicious of democracy but not interested in authoritarian alternatives.
The highest levels of negative sentiment about democracy come from respondents who cannot remember who they voted for, did not vote in the primary, or favored a candidate who did not make it far enough to be included in the list of top primary performers. These make up a small percentage of voters, but again, it shows that those least connected to the political system hold the strongest negative attitudes.
We can also look at respondents based on their 2012 and 2016 vote choices. Here party switchers who voted for Obama in 2012 and then voted for Trump in 2016 stand out as very pro-authoritarian (45 percent support a “strong leader”) and very unsupportive of democracy (45 percent declined to express a preference for democracy). Notably, the consistent Democrats and consistent Republicans express the highest levels of support for democracy, in keeping with the pattern that those most closely affiliated with the party system are most supportive of democracy.
VII: Dissatisfaction with democracy and distrust of elites are related to lack of support for democracy, but are not the same thing.
Often, we confuse satisfaction with democracy with support for democracy. Yet, it is quite possible to support democracy as the best system of government but be dissatisfied with the way that it is working. We can see this clearly below. We asked respondents to describe their level of satisfaction with democracy in the U.S. and got the following results: Eight percent of respondents were not at all satisfied with democracy, 31 percent were not very satisfied, 50 percent were fairly satisfied, and 11 percent were very satisfied.
Not surprisingly, those who are satisfied with democracy are considerably more likely to prefer democracy than those who are not satisfied with democracy. But even a clear majority of respondents (63 percent) who are “not at all satisfied” with democracy in the U.S. still prefer democracy, despite being deeply dissatisfied.
Interestingly, there is no relationship between satisfaction with democracy and support for a “strong leader”. Within all levels of satisfaction with democracy, the same percentage (23–24 percent) want a “strong leader.” Clearly, something else explains the support for a “strong leader.” It seems likely that many supporters of this option favor an illiberal and highly majoritarian version of democracy, where a president, once elected, could govern with few if any restraints.
Moving beyond general dissatisfaction with democracy, we go deeper and examine two of the major current critiques of democracy: (a) the system is rigged and (b) the people running the system are all corrupt.
Following the work of Eric Oliver and Wendy Rahn,(viii) we construct two indexes, a “rigged system” index and a “distrust of experts” index.10
Our “rigged system” index consists of three evaluative questions:
People like me don’t have any say in what the government does.
Elites in this country don’t understand the problems I am facing.
Elections today don’t matter; things stay the same no matter who we vote in.
The “distrust of experts” index consists of four evaluative questions:
I’d rather put my trust in the wisdom of ordinary people than the opinions of experts and intellectuals.
When it comes to the really important questions, scientific facts don’t help very much.
Ordinary people can really use the help of experts to understand complicated things like science and health.
Politics is ultimately a struggle between good and evil.
Distrust of elites is also part of the democratic deconsolidation narrative, which goes something like this: “Democratically anointed leaders failed us; therefore democracy isn’t such a great system.” This is simplistic, but it may reflect a broader distrust in institutions that contributes to doubts about democracy and support for authoritarianism.
10 Oliver and Rahn use the term “anti-elitism” instead of “rigged system.”
Both anti-system attitudes have some explanatory power. Those who think the system is rigged are only a little more likely to support a “strong leader” than those who think the system is fair (26 percent vs. 20 percent), but those who think the system is rigged are notably more likely to refuse a preference for democracy (26 percent) than those who think the system is fair (17 percent).
We get more explanatory purchase on the “strong leader” question. Among those who distrust experts, 38 percent want a “strong leader” — nearly three times the proportion wanting a “strong leader” (13 percent) among those who value expert opinion. And 29 percent of those who distrust experts don’t think democracy is preferable, compared to 19 percent who trust experts.
The rigged-system and distrust-of-elites indexes are only weakly correlated with each other.11 However, distrust of experts is highly correlated with cultural conservatism.12 This analysis suggests dangerous consequences to President Trump’s penchant for undermining the value of neutral expertise by casting any evidence he doesn’t like as “fake.” The more that expertise is distrusted, the more support for democracy may weaken.
11 The Pearson’s correlation between these two indexes is 0.15.
12 The Pearson’s correlation between these two indexes is 0.61.
VIII. Perception of community disorder makes people more inclined toward a strong leader — financial insecurity doesn’t.
Another set of possible explanations for any turn away from democracy involves life circumstances. One popular theory is that economic anxiety is driving negative democracy sentiments. Another possibility is that Trump’s “American carnage” approach is designed to stoke fear, which might lead to more support for authoritarian leadership.
We have put together two indexes to test these theories, one asking people about their own personal finances (whether they were satisfied with their income, their savings, and their level of debt), and a second asking people about the extent to which they think their community is falling apart (whether crime, drugs, and alcohol are problems in their community). We then break these measures into three groups, based on the responses.
Let’s start with community disorder.
Not surprisingly, those who live in communities of high perceived disorder are notably more proauthoritarian, with 32 percent saying they would support a “strong leader”. Interestingly, it is the middle category that is least pro-authoritarian, with only 16 percent saying they want a “strong leader.”
We don’t see much relationship between perceptions of community disorder and preference for democracy, though we do observe the same pattern — that it is the middle category that expresses the highest support for democracy.
We also look at personal financial circumstances to test the “economic anxiety” hypothesis. Personal financial circumstances don’t appear to matter much. There are no differences among the subgroups, at least as measured at this level. We find no evidence at this level of analysis that people are turning against democracy or embracing strong leadership because of their personal financial troubles.
Our analysis has looked closely at both support for a “strong leader” and willingness to support nondemocratic alternatives. Though these sentiments sometimes overlap, they are also distinct. Turning away from democracy does not mean embracing authoritarian rule and some respondents may simultaneously support democracy but have authoritarian sympathies.
The good news is that overall levels of support for democracy remain quite high at the aggregate level and we have seen a reverse in the long-term trends expressing negative attitudes about democracy. However, there are worrying signs of weakness and softness. And while American levels of support for democracy are generally in line with levels in major European democracies and in Canada, support for authoritarian options in the U.S. — both military rule and a “strong leader” — are higher than in many other advanced democracies. In particular, while the level of support for a “strong leader” in our survey (24 percent) is significantly lower than the 34 percent observed in the 2011 World Values Survey, it is still twice as high as what the 2017 Pew survey found in France and four times the level in Germany (for a fuller discussion of the comparative evidence and trends, see Appendix B).
Our analysis finds that the highest levels of support for authoritarian leadership come from those who are disaffected, disengaged from politics and the party system, culturally conservative, and particularly those who express negative attitudes toward racial minorities and are deeply distrustful of experts. Not surprisingly, these are the kinds of voters who have been most supportive of Donald Trump.
Though we are encouraged by the top-line good news (support for democracy remains high), the correlates of negative attitudes about democracy involve many hallmarks of our current political environment, with its racialized culture-war politics, its relentless attacks on elites and institutions, and its bitter but narrow two-party, hyperpartisan competition.
More research is needed on these questions, and special attention must be paid to replication in question wording since variation in wording can make comparison over time (and across countries) hazardous. Future research must look not just at political cynicism and alienation in the U.S. — themes that survey researchers here have periodically measured for decades — but also levels of explicit and implicit support for democracy vs. authoritarian options. It also may be time to examine these questions through qualitative research to more definitively understand underlying motivations for authoritarian attitudes that cannot be pulled out through quantitative surveys.
We have assumed for a long time that public support for democracy isn’t an interesting topic to look at in the U.S. Unfortunately, it now is.
On October 16, 2017, the Pew Research Center released a report based on global research about attitudes toward democracy. “Globally, Broad Support for Representative and Direct Democracy: But Many Also Endorse Nondemocratic Alternatives” by Richard Wike, Katie Simmons, Bruce Stokes, and Janell Fetterolf finds results from the U.S. that mirror our own findings. For more, see: http://www.pewglobal.org/2017/10/16/globally-broad-support-for-representative- and-direct-democracy.
In our survey, 61 percent say they are either satisfied (50 percent) or very satisfied (11 percent) with the way democracy is working in the U.S. This is identical to the overall proportion in 2017 saying they are satisfied with democracy in the United Kingdom, and roughly comparable to Germany (68 percent) and France (57 percent) in the Eurobarometer surveys. Roughly the same proportion (62 percent) said they were satisfied in Japan, Korea, and Taiwan (when surveyed between 2014 and 2016) in the Asia Barometer surveys.13 The 2017 Pew survey, however, found lower levels of satisfaction in the U.S. (46 percent), U.K. (52 percent), Japan (50 percent), and France (34 percent), and higher levels in Germany (73 percent) and Canada (70 percent).(ix)
Levels of satisfaction with democracy tend to be more volatile than underlying value orientations toward democracy, as they shift over time in response to economic and political circumstances. On the one hand, we take heart in the fact that Americans do not express sweeping disaffection with the state of their democracy; in fact, only 8 percent say they are “very dissatisfied.” On the other hand, only 11 percent of our respondents are very satisfied, and the lower figures in the Pew survey (and in a recent Americas Barometer survey) do not suggest complacency. Our finding that four in 10 Americans are not satisfied with the way democracy is working may be taken as a warning sign, and possibly a lower boundary of the disaffected population. A recent Pew survey found that only half of Americans currently trust the national government “to do what is right,” compared to 69 percent of Germans and 67 percent of Canadians.(x)
Support for a “strong leader” who doesn’t have to bother with Congress (Parliament) or elections has declined from the peak level found in the World Values Survey in 2011. The 2017 U.S. level (24 percent) is now more in line with what was found in Australia, Sweden, the Netherlands, and Japan in the previous World Values Survey (Figure B1). In the 2017 Pew global survey, U.S. support for a “strong leader” (with a slightly different question wording) was found to be at 22 percent, lower than Italy (29 percent) and the U.K. (26 percent) but higher than Canada (17 percent), Sweden (9 percent), and Germany (6 percent).
13 The figures from Europe are from the annual Eurobarometer, conducted in May 2017, and the figures from Japan, Korea, and Taiwan are from the Asian Barometer.
The U.S. has become an even more extreme outlier in terms of the percentage of people who support “army rule” (18 percent), exceeding levels previously recorded by World Values Survey in other liberal democracies (Figure B2). However, the 2017 Pew survey of global democracy attitudes showed a narrowing of differences, with 17 percent supporting military government in the U.S., France, and Italy, 15 percent in the U.K., but only 10 percent in Canada and 4 percent in Germany and Sweden.
The proportion of Americans who rate a democratic political system as “good” or “very good” has risen in 2017 to the earlier recorded peak of about 87 percent. This places it at about the level of Australia previously, but still well below Germany, where the response was 94 percent in the previous World Values Survey (Figure B3). A more explicit form of this question (specifying the meaning of democracy as elected representatives deciding what becomes law) garnered about the same level of U.S. support in the 2017 Pew global survey, comparable to Canada, slightly above the U.K. and France, and somewhat below Sweden (92 percent) and Germany (90 percent).
The percentage of Americans who say living in a democracy is very important, which we take to be a placement of eight to 10 on the one to 10 scale, has increased since the 2011 World Values Survey in the U.S., from 73 to 83 percent. This puts the U.S. more in line with the figures then recorded in Germany, the Netherlands, and Australia (Figure B4).
Finally, we do not have trend data on the preference for democracy, which was not asked in the World Values Survey. The 2014 Americas Barometer (Latin American Public Opinion Project survey) found the percentage of the U.S. sample saying “democracy is preferable” to be 81 percent, or just slightly higher than the percentage in our survey (78 percent). By comparison, only 68 percent of Canadians in the 2017 Latin American Public Opinion Project survey and only 68 percent of Japanese in a 2016 Asia Barometer survey said, “democracy is preferable to any other kind of government.”
Among the “democracy not preferable” respondents, 49 percent are consistent anti-authoritarians (oppose both “strong leader” and “army rule”)
Among the “democracy not essential” respondents, 44 percent are consistent anti-authoritarians
Among the “democracy is a bad system” respondents, 49 percent are consistent anti-authoritarians
“Follow the Leader” was originally published on March 13, 2018, and was updated March 21, 2018.
Subscribe to our mailing list for updates on new reports, survey data releases, and other upcoming events.