Abstract
The ability to identify which citizens are democratically competent and which fall beneath the relevant standard of competence bears on numerous questions in democratic theory. These include questions about the distribution of the franchise, the type of civic education that democratic governments should provide to their citizens, and how we might prevent democratic backsliding. In this paper, we aim to identify and defend a criterion of minimal democratic competence. Specifically, we argue that a voter should be regarded as minimally democratically competent with respect to a given election if and only if that voter knows how to vote for the candidates or policies that, if chosen, would not predictably bring about the end of that electoral democracy, and intends to vote that way.
Introduction
Being a democratic citizen—and being a voter, specifically—is a task at which people can be competent to various degrees. Every existing democracy, insofar as they do not allow 5-year-olds to vote, distinguishes between those who are competent enough to warrant bearing the rights and responsibilities of voting and those who are not. They also, generally, tend to provide civic education designed, at least in part, to make their citizens more competent in exercising those rights and responsibilities. This means, of course, that there are at least some standards of democratic competence that are operative, or at least implied, in how democracies function.
Looking more specifically to the views of political philosophers, while most would endorse universal suffrage, few would argue that there are no minimal competence requirements for voting. At a minimum, infants, toddlers, the mentally ill, and those suffering from advanced dementia are generally regarded as falling below the relevant competency threshold. 1 More controversially, some like Brennan (2011) have argued that a large number of enfranchised citizens are incompetent on grounds of ignorance, irrationality, or moral unreasonableness. There is thus a need within democratic theory to get as clear as we can about what the standards of democratic competence should be. Having a clear, defensible standard of democratic competence holds the promise of allowing us to better justify policies regarding extending and restricting the franchise, as well as helping us to establish clear social norms regarding competent voting, to implement more effective schemes of civic education, to set better civic-educational priorities, and to identify when democracies are at particularly high risk of backsliding.
While a number of standards of democratic competence have been proposed, or are at least implied, within previous work in the literature on democratic theory, we will argue that none within that array of currently available standards are fully adequate for the purposes to which democratic theorists would like to put a standard of democratic competence to use. For these reasons, and others that we will articulate below, our aim in this paper is to present and defend a standard of democratic competence.
The standard that we will propose is intended to be a minimal standard, though we mean it to be minimal in a couple of specific senses. By “minimal” here, we mean, first, that it makes as few commitments as possible on issues of substantive disagreement in democratic theory, relying on as minimal a conception of democracy as possible. By this, we mean that we will rely on an account of democracy that deems as few features as possible to be necessary features of democracy. A key part of making the case for our standard will, of course, be to argue that we have good reason to identify a standard that is minimal in these specific ways.
The standard is also minimal in the sense that it lays out a minimal requirement for competence. It will not tell us what a fully civically virtuous democratic citizen looks like. Rather, it provides a clear necessary and sufficient condition for qualifying as a competent democratic citizen, at least on the minimal conception of democracy mentioned above. 2 To meet the standard is to qualify as at least minimally competent as a voter on that minimal conception, and to fail to meet the standard is to fail to be even minimally competent as a voter on that minimal conception.
Note, though, that we do not mean that the standard is minimally demanding in the sense of demanding as little as possible. We expect that a standard of democratic competence is likely to be somewhat morally, cognitively, and educationally demanding even if it makes as few commitments as possible and thereby captures some shared commitments among democratic theorists. We will discuss just how demanding we think our standard is in section “Our proposed standard”.
Our paper proceeds as follows. In the following section, we argue that we should want a standard of democratic competence that is minimal in the ways that we have outlined, and we explain the role that we see our minimal standard playing in the broader conversation about democratic competence. We then present and provide an initial defense of what we take to be the correct minimal standard. In the section “Applying the minimal standard of democratic competence”, we show how the standard can be helpfully applied across different debates to which a standard of democratic competence is relevant, including debates over the scope of the franchise, civic education, and various electoral reforms. Finally, we conclude.
The need for a minimal standard of democratic competence
In this section, we make the case that we need to identify a minimal standard for democratic competence. For a variety of purposes, numerous fundamental debates in democratic theory make implicit or explicit mention of competence. Perhaps most centrally, a clear standard of competence is needed to guide the conversation about who ought properly to count as part of the democratic electorate, particularly in debates about the voting rights of the very young, the very old, and those with severe disabilities. Beyond this, questions about the normative value of democracy relative to other regimes, about voting ethics, about the epistemic and moral quality of democratic decisions, about civic education, and about democratic backsliding all involve some reference to competence. We briefly summarize some of these debates below and we argue that a properly minimal standard of democratic competence can advance the conversation across a number of these fronts. In particular, we argue in section “Applying the minimal standard of democratic competence” that getting clear about the minimal standard of democratic competence will reveal connections between these questions that have gone unappreciated in the existing debates.
As we mentioned in the introduction, the minimal standard that we are looking for should be a necessary standard regardless of one's conception of democracy. There are numerous plausible contenders for what democracy is and ought to be. A non-exhaustive list of the most prominent such alternatives includes representative democracy, direct democracy, polyarchy, aggregative democracy, deliberative democracy, epistemic democracy, realist democracy, and agonistic democracy. Despite their numerous differences, all of these conceptions of democracy share a belief that democracy is preferable to an authoritarian alternative (even if democracy is not an all-things-considered best political regime). Moreover, all of these views include free and fair elections and referenda as part of the fundamental components of democracy, even if most of these views consider free and fair elections to be a necessary, but not sufficient, condition for a normatively desirable democratic system. For example, deliberative democrats would argue for the importance of supplementing elections with deliberative bodies, direct consultation of citizens, and the development of a robust culture of public deliberation. However, we have not encountered any deliberative democrats who argue that we ought to do away with free and fair elections as a component of democracy.
Our goal is to provide a standard that rests on as few assumptions regarding controversial issues in democratic theory as possible, which also means that it should rest on the most accommodating conditions for qualifying as a democracy that we can find. Based on the above, we therefore make the minimal normative assumption that democracy is preferable to authoritarian alternatives and the minimal conceptual claim that democracy includes the presence of free and fair elections. Of course, “free and fair elections” require more than the formal spectacle of elections. Candidates need to be free to run for office without fear of government reprisal, they should have the ability to organize into political parties and other associations, and they should have opportunities to present their ideas and policies to the electorate. Similarly, voters should have genuine opportunities to vote without fear of violence or government persecution for voting for or otherwise supporting the opposition. This requires a suite of liberal protections such as freedom of speech and freedom of association in addition to robust civil protections. However, these conditions would not include, for instance, having a robust deliberative culture or having particularly high participation. These may be desirable features of a well-functioning democracy, but they are not strictly necessary in order to have free and fair elections.
Why search for such a minimal standard of democratic competence? We believe that such a standard would be desirable because it would be ecumenical in a way that would allow a greater amount of shared discussion to take place regarding democratic competence and the norms and policies that follow from understanding competence. Understanding the minimal degree of competence that any democratic citizen ought to have would also allow us to set clearer priorities regarding civic education, institutional reforms, and changes to the structure of voting rights in ways that can garner support across a variety of normative perspectives. If achieved, we believe this would constitute significant progress in the conversation.
The need for an ecumenical standard comes in part from the fact that within the current literature, a number of standards of differing stringency are being proposed, often because various thinkers’ understanding of how competent voters need to be depends on various conceptions of democracy that make democratic participation demanding to different degrees. While some of this difference comes from apples-to-apples disagreements about what the standard of competence ought to be, a good deal of it comes from the fact that standards of democratic competence are being proposed in a piecemeal fashion within a number of disparate projects. We begin by surveying some of the discussions of competence in political philosophy, then turn to examples from political scientists aiming to empirically measure competence within given electorates.
There are numerous standards of democratic competence discussed in political philosophy, often without an explicit comparison to other standards operative in the literature. For instance, those who attempt to justify the normative value of democracy on the basis of its capacity to make decisions that are epistemically superior to alternative decision-making systems tend to imply some standard of voter competence related to choosing correctly (Cohen, 1986; Estlund, 2008; Goodin and List, 2001; Goodin and Spiekermann, 2018; Landemore, 2013; Schwartzberg, 2015). This is largely because one of the main results cited by epistemic democrats is the Condorcet Jury Theorem. According to Goodin and Spiekermann (2018: 85): “The Condorcet Jury Theorem shows us how groups can be far more competent than the individuals who comprise them, just so long as the group is composed of a large number of individuals whose votes are minimally (better than random) competent and independent of one another.” The standard for competence that this implies, then, is that a voter is sufficiently competent if they “can be trusted to be right more than half the time on any binary political issue,” where “right” is some combination of getting at the underlying truth of the matter and choosing aptly in light of that truth (though, of course, a good deal of the work here would be done by how “aptly” gets spelled out here) (Goodin and Spiekermann, 2018: 150).
Another source for standards of democratic competence comes from epistocratic critics of democracy. Brennan (2011: 700) defends what he calls “the competence principle,” namely the principle that “citizens have a right that any political power held over them should be exercised by competent people in a competent way.” Brennan uses this principle in order to argue that universal suffrage is unjust and to argue in favor of more restricted suffrage based on the competence principle.
Brennan develops his competence principle in the case of jury decisions, then argues that a similar competence principle applies to decisions made by governments, public officials, and, by extension, to decisions made by voters in selecting elected officials and/or voting on referenda. In the case of jury decisions, Brennan argues that the competence principle translates into two types of requirements. First, there is the requirement that each individual decision-maker who serves on a jury ought to have sufficiently good epistemic and moral character (Brennan, 2011: 704). 3 Second, there is the requirement that the particular decision should itself have been made in an epistemically competent and morally reasonable manner (Brennan, 2011: 704). 4 Brennan claims that the same principle applies to individual voters making democratic decisions. As such, Brennan thus appears to claim that in order to be considered competent, an individual voter should have sufficiently good epistemic and moral character and should make the particular decision about who to vote for in an epistemically competent and morally reasonable manner. In further spelling out the moral reasonability requirement, Brennan simply claims that there are appeals to “moral reasonableness” elsewhere in the literature that he is happy to rely on, and with respect to making decisions in an epistemically competent manner, he appears to mean voting on the basis of evidence and for considerations that actually serve as reasons to vote for one person or proposal rather than another (Brennan, 2011: 708).
Another way that normative theorists think about democratic competence is that it requires fulfilling the relevant functions of citizenship, beyond simply voting, in a competent manner. Much of this discussion has taken place in the context of identifying the right policies for civic education, policies that have been defended variously by Anderson (2007), Satz (2007), Allen (2016), and Gutmann (1999). All of them subscribe to the view that there is a threshold of competence that any democratic citizen should be able to meet, and that we can specify the educational requirements involved in meeting the relevant threshold.
The threshold that people need to meet here consists in being capable of “full membership and inclusion in a democratic society of equal citizens” (Satz, 2007: 636). This involves “a threshold level of knowledge and competence for exercising its associated rights and freedoms,” the ability “to comprehend and apply concepts like ‘reasonable doubt’, ‘negligence’, and ‘probability’, and… to analyze statistical tables and graphs but also to have the capability of responding to the arguments of other jurors during their deliberations” as well as an “understanding of racial discrimination [and] poverty” that comes from “the presence of diverse individuals” and that gives one the requisite “knowledge and perspectives needed in both politics and in the economy” regarding the “interests of the diverse members of that society” (Satz, 2007: 636–638).
Although not explicitly framed as a competence standard, Elliott (2023: 115) describes an “ecumenical bare minimum of citizenship expected of any democratic citizen.” Elliot labels this minimum conception as “stand-by citizenship” and describes it as including three components: (1) paying attention to politics, at least periodically (what he calls “critical attention”), (2) the civic skills required to participate in politics, and (3) the capacity to participate actively when necessary (what he calls “upward flexibility”) (2023: 100–106). Elliott (2023: 112–114) argues that this approach can be empirically measured through interest in politics, political knowledge, and turnout.
In addition to various attempts by normative democratic theorists to spell out standards for democratic competence, there are attempts by empirical political scientists to test the degree to which electorates and the individuals that make them up are competent or not. In order to test such a thing, they need to decide on a way of operationalizing competence, and this brings with it, even if only tacitly, some account of democratic competence and what standard should be used to determine whether or not somebody is competent.
One approach here focuses simply on whether or not voters have certain kinds of political knowledge, with political scientists asking people questions about current events, the current state of political affairs, and the rules and procedures of the political system (Barabas et al., 2014; Delli Carpini and Keeter, 1993; for a critique, see Lupia, 2006). The underlying normative assumption of this political knowledge approach seems to be that falling short of a certain amount of such knowledge is sufficient to make one democratically incompetent, with the more general implication, of course, that meeting the standard for democratic competence is a matter of having a sufficient amount of relevant political knowledge.
A second approach focuses on whether citizens vote correctly, where voting correctly means voting for the candidate or political party that most closely represents their own views, interests, or preferences. To implement this approach, researchers can test, for instance, whether or not presenting a voter with more information about the candidates leads the voter to say that they would have voted differently than they did, or whether voters choose the candidates that experts rate as most aptly fulfilling the voters’ stated views, interests or preferences (Lau and Redlawsk, 1997; Lau et al., 2008). On this approach, voters who too often vote “incorrectly” can be considered democratically incompetent, meaning that the underlying normative assumption of this approach is that democratic competence is a capacity to vote in ways that effectively promote one's own (perceived or professed) interests.
A final approach focuses on voter rationality and investigates, for instance, whether voters reward or punish incumbents based on factors that one would consider to be irrelevant or outside of their control (Healy et al., 2010). Achen and Bartels (2016: 116–145) find that voters do, in fact, tend to be influenced by irrelevant factors, which challenges the conception of voter rationality required for voters to hold elected officials accountable. The normative claim presumed by this approach is that democratic competence consists in displaying a certain kind of procedural rationality in one's thinking and decision-making.
This brief look at these various standards should be enough to show how disparate they are not only in terms of how demanding the standards are, but also in terms of how demanding the conceptions of democracy are on which those standards rely. The fact that the conversation appeals to multiple different conceptions of democracy makes a minimal standard of the kind that we are proposing valuable for a couple of reasons. First, offering a minimal standard that aims to be ecumenical should allow participants in these various conversations to have a shared baseline of agreement, a baseline from which they can argue for any additional conditions for a system to qualify as democratic or for related additional conditions for qualifying as a competent voter. Additionally, that shared baseline of agreement might allow empirical political scientists and philosophers working on a range of issues in democratic theory to develop views more relevant to one another, and their views might be more helpful to each other's thinking if they had some shared account of democratic competence. While providing a shared account that can be applied that ecumenically is, of course, a lofty goal, we aim to do so here with the understanding that even if we fall short, that might still leave us with a useful view and an improved conversation.
Our proposed standard
Given the above, we propose the following minimal standard of democratic competence for an individual voter: A voter is considered
This standard includes two components that are both necessary to avoid voting for the end of one's democracy: The epistemic competence requirement refers to the voter's capacity to understand whether particular candidates are likely to advocate policies or make moves that would constitute the end of that electoral democracy. Epistemic competence is not enough on its own, though, to allow a voter to do their part in keeping their democracy alive. This is because a voter might be epistemically competent enough to understand that a particular candidate is advocating an authoritarian takeover, but nonetheless support that move out of an insufficient commitment to avoiding the death of their democracy. This is why the democratic commitment requirement, that is, the requirement that the voter intends to avoid voting for the end of democracy, is also necessary.
We believe that this standard is normatively apt in important ways, that is, that it is fitting for it to serve as the shared, minimal baseline. This is, first, because it presents a necessary enabling condition for any other additional competence conditions that one would want to add. At the very least, in order for voters in democracies to continue making apt policy choices, and to do so through appropriate deliberation procedures, the democracy must be able to continue functioning as such. This is similar to competence in almost any other area. In order to display competence at chess, for instance, one must be able to display the minimal competence involved in not flipping the board over or otherwise making moves that would constitute a failure to continue playing chess. 5
Some standards discussed above do not adequately capture this kind of necessity, if we take them as potential standards meant to preserve democracy on our minimal conception of it as maintaining free and fair elections. For example, the standards described by Anderson, Satz, Allen, and Gutmann describe a citizen that is informed about politics and highly educated. According to Gutmann (1999: 47), for example, a sufficiently competent democratic citizen would need to follow political news, understand economics, and assess complex policy choices such as between public and private provision of various services. While desirable for a good citizen, requiring such skills as necessary for meeting even the lowest acceptable minimum of competence places the bar much higher than what would be required to simply keep the democracy existing as such. For example, one can correctly identify a predictable authoritarian threat even if one would fall short of identifying Pareto optimal policies in organizing mail delivery or transportation services. The same could be argued about the “voting correctly” standard described by epistemic democrats such as Landemore, or the “good epistemic and moral character” standard proposed by Brennan. While, as we have discussed, there is nothing stopping proponents of such views from arguing that we need a more robust conception of democracy than our minimum, and that this requires some of these more demanding standards of democratic competence, we have good reason to reject such standards as too demanding to serve as the minimal standard we seek.
We also believe that meeting the epistemic competence and democratic commitment requirements is sufficient for voters to do their part in preserving free and fair elections. In other words, to meet these two requirements is to adequately meet the minimal expectation that one does one's part in keeping one's democracy alive. Some standards discussed above are not sufficient for this purpose, since citizens can fulfill them, even universally so, and still end up voting for the predictable end of their democracy and thus failing to do their part. For instance, stand-by citizenship as described by Elliott would not be sufficient for democratic competence under our minimal conception of democracy. Although a citizen would need to pay attention to politics in order to make a competent democratic decision, “critical attention” together with capacity and willingness to participate in politics is not enough. Stand-by citizenship would count as competent those citizens who intend to vote for candidates who would predictably end democracy and replace it with an authoritarian regime. Provided that they were interested in politics, knowledgeable about the candidates and willing to turn out, voters for Chávez in Venezuela or for Hitler in Weimar Germany would count as democratically competent on such a standard. This illustrates the possibility of setting the floor of democratic competence too low.
Is it possible that our two conditions set the bar too high, and are therefore not genuinely necessary for voters to fulfil their role in preserving democracy? Is it possible that they set the bar too low, and are therefore not genuinely sufficient? To the potential charge of setting the bar too high, it might be argued that a significant portion of voters can fail to meet our requirements without the electorate thereby voting to end the democracy, so it is not strictly necessary that everyone meet it. For example, in a majoritarian electoral system with only two candidates for the presidency, only half-plus-one of voters would need to be democratically competent for the electorate to vote competently. However, it certainly is necessary that we need enough voters to meet those requirements in order for the electorate to avoid voting to end their democracy and it is true that the more voters meet the minimal standard, the more likely the electorate is to do their part in preserving democracy. If it is something that we definitely need enough voters to be able to do, and we want to have a uniform conception of voter competence, then it is something we should expect every citizen to be able to do. 6
To the potential charge of setting the bar too low, it might be argued that voters meeting our two conditions might still fail to stop an authoritarian takeover, because such a thing can happen even when the voters meet our conditions. To this, we would argue that meeting our two conditions is at least sufficient for voters to avoid predictably empowering an authoritarian. While we would say that this is sufficient for voters to do their part in preserving their democracy, we would not say that voters doing their part is, itself, sufficient for the preservation of democracy. We believe that the voters doing their part can only ever be one component in a sufficient, comprehensive defense against an authoritarian takeover of a given democracy. Another way of putting this is that while meeting our two requirements is not enough to stop a democracy from veering toward the authoritarian ditch, meeting them is sufficient for voters to avoid being the ones jerking the wheel in that direction. It would overestimate the power of voters to say that doing their job in keeping democracy alive requires more than this.
That our standard serves as a necessary and sufficient enabling condition for voters to fulfill their basic role in preserving democracy also means that the ability to meet our minimal standard can serve as a normatively appropriate top priority as we think about civic education and what we need citizens to learn in order for them to be minimally competent to participate in electoral democracy. While there may be good reasons for them to develop an understanding of the nuances of tax policy and for them to develop a robust set of civic virtues, they will only be able to meaningfully do so if enough of them can first meet the minimal standard of being able to avoid voting for the end of their democracy. This also serves to make a proposed civic education more politically ecumenical, since the only normative commitments or political judgments built into the standard of competence are those that ought to be accepted by any citizen committed to their democratic project.
To further understand what this standard entails and why it might be apt, we can ask how demanding it actually is in practical terms. One initial thing we can see here is that the standard is clearly less demanding than some other conceptions of democratic competence at which we have looked, since the knowledge required to avoid voting for (or otherwise bringing about) the predictable end of democracy is a mere subset of the knowledge that it takes to engage in voting behavior in sufficient accordance with procedural rationality, or to vote for apt policy choices that positively impact the wellbeing of one's fellow citizens, or to correctly vote in one's own interests. The standard being less demanding in these ways helps it to serve the ecumenical purpose that we want it to serve. It is better positioned than those other standards to provide a condition for democratic competence that everyone, even those with more demanding standards of democratic competence, should accept as necessary. 7
This does not yet mean, of course, that our standard is undemanding enough that most or even some citizens within our democracies can meet it. We should want our standard to be one that at least some substantial portion of a population can meet. This is because part of what would make such a standard valuable is that it serves to genuinely distinguish between people in our world, to separate real, competent voters from others. This also means, of course, that we do not want a standard that every citizen trivially meets. So, what reason do we have to think that our standard meets this aim of being both demanding and undemanding enough to be useful in genuinely distinguishing between people in our world?
On the way to answering this question, one important thing to notice about the criterion that we propose is that it does not render a global judgment about a voter's capacity to make democratically competent decisions across all possible configurations of candidates and/or electoral institutions. For the moment, we remain agnostic about whether such a global judgment of competence is even possible given the uncertainties involved in the types of candidates and policies that might emerge in any future electoral contest. However, even if one could construct such a global standard, one would still need to have a sense of its application to any particular election in ways that the standard above provides.
One important implication of focusing on individual elections is that voters who might be minimally competent in some elections would fail to meet the necessary threshold in others. In elections in which none of the major candidates would predictably bring about the end of democracy, all of the voters who choose one of these candidates would meet the necessary threshold of competence. 8 In contrast to these “easy elections,” some elections present a genuine danger of empowering an eventual authoritarian takeover, with major candidates who have given numerous signs of intending to bring about the end of democratic elections, such as expressing the need for anti-democratic measures such as suspending elections or suspending the constitution, defending foreign leaders who undermined democratic institutions in their home country, or declaring an unwillingness to abide by the results of the upcoming election in the case of electoral defeat. In such elections, it is more difficult to meet the standard of democratic competence, if only because voters will not simply meet the standard by default.
Of course, it is not certain that when an anti-democratic candidate is elected, they would succeed in bringing about the end of democracy. Any number of things could keep that from happening, from the democratic institutions proving to be sufficiently resilient, to the would-be dictator changing their mind, to the would-be dictator being struck by lightning on the golf course. This panoply of possible outcomes bears on whether or not a vote for the anti-democratic candidate is a vote that would predictably bring about the end of the electoral democracy and thereby violate our minimal standard. As such, we need to get a little clearer on what it means for an end of electoral democracy to be predictable.
For the purposes of our standard, an end of democracy is predictable if a candidate has given clear indication that they would make certain moves or support certain policies that would constitute an end to democracy. If an anti-democratic candidate, once elected, were to succeed (either individually or as part of a collective) in, for instance, suspending elections or in modifying the electoral rules to de facto eliminate the opposition after clearly signaling some intent or even willingness to do so, we would consider the anti-democratic outcome to be a predictable one. You can think of this as similar to the situation where a bully threatens to beat you up if you don’t surrender your lunch money. While it is possible that the bully was merely bluffing or that a third party will stop the bully before they carry out their threat, the threat itself predicts the beating in a way that renders that outcome “predictable” and the outcome of getting away unscathed as “surprising,” “fortunate,” or “unexpected.”
This makes the end of democracy predictable in at least some circumstances. Arguably, those circumstances of predictability actually obtain in most elections that might bring about the end of a given democracy, since would-be authoritarians tend to run on recognizably authoritarian platforms rather than suddenly pivoting to an authoritarian takeover mid-term. This all should head off the potential concern that, since we cannot ever be certain that an authoritarian takeover will succeed, the end of democracy is never predictable, that one can therefore never vote to predictably bring about the end of democracy, and that our standard is therefore too weak to judge any voters to be incompetent. 9
Since, in these elections in which the end of democracy is a predicable result of voting in a particular way (at least in our sense of “predictable”), there are going to be those who have the sufficient epistemic competence and democratic commitment to avoid voting that way and those who lack it, our standard should be able to draw genuine distinctions within various existing democratic populations, at least for particular elections. While a standard that renders more global judgments of competence would, of course, be more straightforward than one that renders judgments of competence for a particular election, the latter also serves to highlight features of democracies that may help to make it easier for voters to be minimally competent. By allowing us to draw this distinction between elections in which it is easy to be competent and elections in which it is more difficult to be competent, we also enable ourselves to make claims about different democratic systems, and to use our minimal standard to recommend systems, institutions, and so forth that help to make competence easier by making an authoritarian takeover less likely. Just as getting people to have the appropriate epistemic competence and democratic commitment to avoid ending democracy can be seen as an important priority in discussions of civic education for democratic competence, designing institutions that make it easier for more voters to meet the minimal competence standard more of the time can be seen as an important priority in discussions of democratic institutional design. We will discuss in greater detail, in the next section, the upshots of applying our minimal standard within discussions of those issues.
Applying the minimal standard of democratic competence
In this section, we consider three ways in which we might apply the standard of competence we defend: civic education, electoral reforms, and the distribution of the franchise. Because both the success of civic education and the design of electoral institutions can influence how competent individual voters are as judged by our standard – the former through getting citizens over the minimal standard and the latter through making it easier for citizens to meet the standard – and because they then influence how we should approach the distribution of the franchise, we will discuss those first.
Civic education and civic virtue
We mentioned in section “Our proposed standard” that our minimal standard, in focusing specifically on the competence that enables voters to keep a democracy alive, can help to clarify our immediate priorities when it comes to issues such as civic education. We are not saying that civic education should only be aimed at meeting our minimal standard, since there may be some clear benefits to getting citizens to develop much higher levels of understanding of their political and legal systems and to develop much higher levels of epistemic, moral, and civic virtue than our minimal standard calls for. However, we believe that our minimal standard helps to clarify what it is absolutely essential to impart to citizens within their civic education. The most important first steps on the road to a fuller civic virtue, we believe, should be getting citizens to meet both the epistemic competence and democratic commitment conditions that appear in our minimal standard for democratic competence.
Given that we want more voters to be able to meet the epistemic component of democratic competence, it is important to ensure that the largest possible subset of current and future voters understand the current electoral system and the constitution of their democratic state. In the US, this includes knowing how votes are counted, whether the states or the federal government control the organization of elections, what candidates and parties are allowed to do during an election, and how the government works after an election, including knowing what the role of the executive is, what limits there currently are on executive power and why those limits exist, knowing what the role of Congress is and how a bill becomes a law, and knowing what the role of the Supreme Court is and how judges are appointed. It is not surprising that this fundamental knowledge about one's government has always been part of civics courses. Our proposed standard does not lead us to argue that basic civic education of the traditional kind is unhelpful — if anything, it supports the view that more investments should be made in this type of education and that more states should include it as part of their mandatory K-12 curriculum. Moreover, more opportunities should be made available for adults to receive free civic education refreshers and to attend publicly sponsored civic education events.
In addition to reaffirming the importance of traditional civics, however, our standard also invites one to incorporate a more systematic education in how to spot a would-be authoritarian. This can be done through the study of literature, history, political science, and other humanities and social science courses in which K-12 students can review pivotal moments when democratic citizens domestically and abroad failed to prevent would-be authoritarians from grabbing power and ending democracy. Studying democratic backsliding and democratic death should be a key component of civic education, as should studying what we might call the “authoritarian playbook” and the warning signs that a regime might be heading into an authoritarian direction. This also involves teaching students about rhetoric, persuasion, and manipulation and the ways in which they might defend themselves against falling for charismatic would-be authoritarians. Just like civic education in the constitutional essentials should be more easily available beyond K-12, we believe that governments should also do more to sponsor the kind of education described in this paragraph.
Beyond the epistemic component, our standard of democratic competence also invites attention to students’ and citizens’ levels of democratic commitment. A democratically competent voter is one that prioritizes the survival of democracy even if that means potentially renouncing cherished policy goals (when they are championed by a potential authoritarian) or potentially endorsing candidates one finds less appealing on other grounds. This type of commitment to democracy is difficult to foster and one would need further empirical research to find examples of programs or curricula that foster such a commitment.
In seeking to foster this democratic commitment as part of a general civic virtue, our minimal standard also creates a new opportunity to foster a clear, universal social norm around being a competent democratic voter. After all, people should and generally do tend to care about being a competent voter, just like they care about being a competent driver or a competent friend. That competence is often taken, though, to be a more maximalist capacity to understand and correctly judge the aptness of particular policies, and this lends itself to judging one's co-partisans as competent and one's counter-partisans as incompetent. Having a norm that is tied merely to the survival of democracy itself should allow societal competence judgments to be more politically ecumenical, which should in turn allow for a broader acceptance of the norm, and for the norm to thus exert more force in keeping people from voting in anti-democratic ways.
Additionally, in recent years, political scientists have increasingly documented the dangers of political polarization, particularly affective polarization, for democratic survival (Binder, 2003; Graham and Svolik, 2020; Levitsky and Ziblatt, 2018; McCarty et al., 2006; Svolik, 2019, 2020). Affective polarization refers to the tendency of people to view opposing partisans negatively and co-partisans positively (Iyengar and Westwood, 2015; Iyengar et al., 2019). The growing negative feelings toward opposing partisans and opposing political parties or candidates significantly increase the costs of making a democratically competent decision. This is partly because political polarization raises the bar in terms of how democratically committed one has to be in order to vote for a hated politician in order to avoid an electoral victory for the anti-democratic candidate. 10 To the extent that civic education can foster more positive attitudes toward people who disagree with us politically, it can make it easier for larger numbers of voters to meet the democratic competence standard. As we mentioned above, this is all the more likely if we accept a norm that says that competent voters are those who vote to keep democracy running, even if they disagree with us across the board on matters of tax policy, foreign affairs, and so forth. Moreover, we suspect that fostering democratic commitment would involve more hands-on community engagement activities where students learn the value of incorporating diverse perspectives and learn the importance of cooperation with people with whom one disagrees.
Democratic reforms
In addition to setting priorities when it comes to civic education, our minimal standard of democratic competence can also help in setting priorities in reforming democratic institutions, particularly electoral institutions. We consider two ways in which this is the case: (1) in determining which democracies are the most precarious in terms of likelihood to survive as democracies and (2) in mitigating the risk of democratic backsliding through democratic reforms, particularly reforms of the electoral system.
First, our minimal standard can serve as an important criterion for assessing democratic resilience in the face of threats of democratic transitions to authoritarianism. The larger the proportion of the electorate that meets the minimal democratic competence standard, the more resilient the democracy in question is, all other things being equal. The larger the percentage of incompetent voters, the more at risk the democracy is, all other things being equal. This can allow us to rate democracies in terms of democratic resilience and to compare different institutions based on their contributions to bolstering or weakening this important form of resilience. We can even develop a threshold of collective incompetence above which a democracy would be considered in “critical condition” or at the highest levels of risk of authoritarian takeover. This can allow citizens and observers to dedicate higher attention to the risk of democratic death and to triage which democracies need the highest levels of democratic aid. Similarly, domestic investments in civic education or in democratic reforms might be, all other things equal, more justified in situations where a democracy is critically at risk than in situations where the democracy rates high in resilience.
Using voter competence as a measure of resilience is complicated, however, by the fact that different institutions can make it easier or more difficult for citizens to meet the competence standard in the first place. They do this through design features that make it more or less likely for an anti-democratic candidate to run for high-ranking offices or for an anti-democratic referendum to be put in front of voters. This indicates that there are some important normative implications for democratic system design that come from adopting our minimal standard of democratic competence. Variables in the organization of an electoral system such as the rules for translating votes into seats, whether the system is majoritarian or involves proportional representation, how much power the executive has, the strength of the judiciary, whether parties are weak or strong, what rules govern campaign finance, and so on all shape how an electoral democracy functions and can, under certain circumstances, influence the choices placed before voters in ways that bear on competence as we have defined it. Since each of these features of the design of electoral systems can make it easier or harder for voters to vote in a competent manner, the fact that it is easier for citizens within a system with certain features to meet the minimal standard of democratic competence can serve as an important consideration in favor of adopting a system with those features or reforming electoral institutions in order to lower the risk of democratic decline.
Much of the conversation around democratic competence holds a particular electoral system constant and discusses educational policies that can increase the level of competence of the electorate. Our point is that we can also hold fixed the level of education in a given electorate and focus on electoral reforms that would make it easier for voters to meet the minimal standard of competence. Once we acknowledge that we can increase competence by either increasing education levels or lowering the demands of competent voting, we can also design policies that use both mechanisms to maximize the number of citizens who are democratically competent.
While the specific elements of democratic system design that might influence how easy or difficult a democracy makes it for citizens to meet our minimal standard of competence are too numerous to elaborate and explain here, we can at least briefly discuss a few of them. For instance, in a democracy where party leadership has a decisive say in nominating candidates for higher office such as the presidency, parties can act as gatekeepers and keep potentially anti-democratic figures off the agenda. Levitsky and Ziblatt (2018) give numerous examples of the major US parties refusing to nominate potentially anti-democratic candidates prior to the implementation of the more democratic primary election system. In a democracy where party nominations are done through democratic primaries, however, political parties have a much harder (if not impossible) time keeping anti-democratic candidates from receiving the party nomination—and the substantial amount of democratic popularity and legitimacy that comes along with such a nomination. This means that electoral systems with democratic primaries will be more likely to have national elections in which one of the candidates is anti-democratic. This will make it harder to meet the competence threshold for voting in the national election.
Another set of institutional reforms that may reduce the demands on voters are institutional reforms that limit or reduce polarization. Given the role of polarization in increasing the risk of democratic backsliding, institutional reforms that drive political parties toward the center and that reduce ideological distance between parties can also reduce the competence demands on individual voters. Going back to the Graham and Svolik (2020) model in the section “Civic education and civic virtue,” the costs of supporting a democratic candidate in a highly polarized system is higher than in a system with low levels of polarization. This is because it is less costly to punish a would-be authoritarian by voting for the opposition in systems where you find the opposition party to be an attractive option (i.e. where you have positive or at least neutral affect toward opposition politicians and where you see the ideological distances as smaller). Examples of policies that have been argued to reduce polarization include the introduction of ranked choice ballots or approval voting, the introduction of compulsory voting, transitions from first-past-the-post electoral systems to systems of proportional representation with moderate to high thresholds for gaining seats. 11
There are numerous other ways in which electoral institutions make it easier or harder for individual voters and for the electorate as a whole to meet the required standard of democratic competence. An important task where normative political theorists and empirical political scientists can fruitfully collaborate involves examining the ways in which different electoral institutions increase or decrease the likelihood that voters will face “easy elections,” where none of the candidates, if elected, would predictably bring about the end of democracy versus “hard elections,” where at least some of the candidates, if elected, would predictably bring about the end of democracy.
In addition to setting priorities in terms of which democracies have the highest number of citizens falling below the minimal threshold of competence (and therefore being at the highest risk of democratic backsliding) and in terms of which democratic reforms would allow the largest percentage of voters to meet the minimal competence threshold, we believe that our discussion of competence reveals an important normative reason to prefer democracies that make it easier for voters to meet the relevant competence threshold. In this regard, competence assessments can themselves ground arguments in favor of certain electoral regimes over others. All other things equal, we should generally prefer democracies that make it easier for voters to meet the requisite standards of competence. 12 There are multiple reasons why this might be preferable.
First, democracies where a large percentage of the electorate is democratically competent have a much higher likelihood of surviving and flourishing. Provided that one agrees that democracy is better than authoritarian non-democracy, reforms that facilitate democratic survival should have priority over those that make democratic survival less likely.
Second, democracies that makes fewer demands on voters in order for them to be democratically competent make it easier to include a larger number of voters and therefore create more egalitarian and inclusive distributions of the franchise. To the extent that equality and inclusion are important democratic values, this further bolsters the case for what we might term “easier democracies.”
Finally, democracies where it is easier to be democratically competent economize on the time, energy, and resources of voters that would otherwise be spent investing in democratic competence. This could have important benefits — voters could take some of the time that would otherwise have had to be dedicated to the essential work of ensuring democratic survival and spend it on building meaningful relationships with friends and family, educating their children, contributing to economic growth, or otherwise contributing to the collective project of leading good lives in whatever way they deem best. Even if all the voters do with the extra time is relax more, one would consider this to be an important improvement in a world that makes so many demands on people's time already.
The distribution of the franchise
In addition to the points made above regarding civic virtue and the virtues of various design elements of electoral systems, our minimal standard can also be brought to bear on the question of who, within a given democracy, ought to have the franchise. While the question of who ought to have the franchise raises the specter of disenfranchisement, particularly disenfranchisement for anti-democratic or partisan ends, the primary benefit of our minimal standard in its application to this question, we believe, is that it aims to be ecumenical. In contrast to more maximalist or more demanding standards, which would implicate substantive disagreements about the nature of civic virtue, the proper ends of policy making, and so forth, our minimal standard should be able to gain purchase more easily both among more of the electorate and among more democratic theorists. It is a standard that is intended to protect democracy, full stop, as opposed to protecting any of a number of other goods that democracy might be useful in promoting. As such, it should be a standard that democratic citizens find easier to support than other standards of democratic competence.
The minimal standard also, as mentioned in the discussion of civic education, gives us a clearer and less unwieldy account of the kinds of capacities that citizens need in order to be competent voters in all, or at least most, of the elections that they are likely to participate in during their lives. Just as the standard can provide us with some clear priorities for civic education to help ensure that citizens can attain competence, it also provides us with a clearer way to assess competence than we might otherwise have, given the acceptance of some other, more maximalist standard of competence. This is important, given the gravity of decisions over who to enfranchise, and the need for those decisions to be made on the basis of clear and compelling evidence. This, combined with the minimal standard's being more ecumenical and normatively apt, should help it to be more applicable for the purpose of determining the proper scope of the franchise.
All that said, we are not advocating, here, for any particular way of setting the rules about how one might earn, retain, or lose voting rights. While the competence of any given voter is an important factor in such decisions, they should probably be based on further considerations besides competence. For instance, it may be the case that the average 16-year-old in the U.S. is capable of meeting our minimal standard, though we might nonetheless have other reasons for not extending the franchise to them, such as protecting them from the negative effects of targeted political advertising. It might also be the case that a significant portion of voters in certain other categories fail to meet this competence standard, and yet we have other good reasons not to disenfranchise them. It may be the case that, as we mentioned above, certain types of civic education or electoral reforms can make it easier for such people to become competent. It may also be the case that a system can handle some sub-critical mass of incompetent voters.
With respect to this latter point, we propose using the collective competence of an electorate as a measure of “democratic slack,” or how much tolerance a particular system has for allowing incompetent voters to vote without bringing about the end of that democracy. In an election where 95% of the electorate is competent, there is quite a bit of democratic slack, and therefore one would face little risk from allowing the 5% of incompetent voters to cast their ballot. In contrast, in an election where 51% of the electorate is competent, there is much less democratic slack, and one would incur higher levels of risk in allowing incompetent voters to cast their ballots. This would not decisively indicate that the latter democracy automatically should disenfranchise voters, but it does provide a clearer assessment of the tradeoffs and the risks involved in enfranchising the incompetent, something that is facilitated because of the direct connection that exists between citizens’ capacity to meet our minimal standard and the risk to a given democratic system.
Democracies with a lot of slack and a high tolerance for risk may, for instance, decide to allow all citizens to vote and only disenfranchise in cases where a particular voter has failed to meet the requisite standard in 100% of elections over a period of 20 years. Democracies with very little slack and a lower tolerance for risk may decide to impose stricter restrictions. At the same time, many of these decisions will depend on the percentage of the population that falls into any particular category. For instance, it is very different to have a democracy where 20% of the population is under 18 compared to a democracy where 80% of the population is under 18. 13 The question of enfranchising children who are less likely to be competent across a range of elections will look very differently in each of these circumstances. The same will be the case when it comes to other groups with potential shortfalls in minimal democratic competence.
At this point, some critics may worry that any attempt to identify which citizens are democratically competent and which fall below the relevant threshold is necessarily counterproductive because it summons the specter of competence tests and voter disenfranchisement for partisan political ends. As discussed above, we agree that any procedure for expanding or limiting the franchise should be subject to the highest levels of scrutiny and that, in practice, the best decision may be to avoid disenfranchising incompetent voters regardless of the democratic costs. Even if this is the case, we believe that one cannot properly make this determination without a clear consideration of all the relevant costs, including the costs of a properly targeted policy of civic education that can assist voters who fall below the relevant threshold or who risk falling below this threshold in the future.
Additionally, some might worry that focusing on individual competence wrongly assigns responsibility for maintaining democracy to individual voters when the main bulwarks against authoritarianism are political institutions, parties, courts, the media, and other powerful actors with higher levels of control over the future of democracy. While we also agree that institutions matter a great deal and that any plan to preserve democracy requires numerous actors to work in tandem to resist the threat of democratic backsliding, we believe that it is helpful to clarify what bulwarks exist, how they each function, and what it takes for them to function effectively. In the case of institutions like parties, media, and others that shape the choices that eventually face voters, our standard of democratic competence allows us to see their importance in a new light and to rank ideas for their reform based on their ability to expand or contract democratic competence. In the case of institutions like checks and balances and courts that are designed to stop a would-be authoritarian after they have been chosen by voters, while we believe that these are indeed important bulwarks, focusing exclusively on such institutions can also lead to the illusion that such mechanisms are self-enforcing and require little effort on the part of voters. Unfortunately, there is no political system that can long survive as democratic when more than 50% of voters support an authoritarian candidate. The competence of voters will thus always itself comprise an important bulwark against an authoritarian takeover, and we thus ought to do what we can to make it the most effective bulwark it can be, while doing the same to our various institutions.
Conclusion: A democracy, if you can keep it
In this paper, we argued in favor of developing a minimal standard of democratic competence. We then proposed and defended what we take to be the standard that best fits that bill, and we discussed some of the upshots for normative democratic theory that we believe should follow from accepting our proposed standard. Having made that case, we shall conclude here with the following few thoughts.
In 1787, a lady asked Benjamin Franklin whether the United States was a republic or a monarchy. Franklin's answer, as it was preserved in the notes of a Maryland delegate to the Constitutional Convention, was “A republic. If you can keep it.” Franklin's remark highlights the important reality that liberal democracies are enterprises that, in ordure to endure, must be self-sustaining and self-protective. To continue being liberal democracies, they must be actively kept liberal democracies. A democracy is only a democracy, and democratic citizens are only democratic citizens, insofar as they do not collectively decide to throw that away.
The standard of democratic competence that we have presented here speaks to both the centrality and the gravity of Franklin's concern. By identifying minimal competence with avoiding voting for the predictable end of one's democracy, we highlight the idea that the first duty of a democratic citizen is to remain a democratic citizen. While, as we have clarified, we do not believe that citizens discharging this duty serves as the only, or even the most effective bulwark against a democracy's demise, and while we have argued that institutions can aide citizens’ competence in a variety of ways, we also hope that taking our standard of competence seriously can also lead individual democratic citizens to be a more effective bulwark than they otherwise might have been. This, hopefully, can make the prospect of keeping our democracies intact a little less iffy.
Footnotes
Acknowledgements
The authors would like to thank the journal Editor Professor David Wiens as well as the anonymous reviewers for their helpful comments and guidance. They would also like to thank Susan Bickford, Keith Dowding, Jeff Spinner-Halev, Sam Schmitt, and audiences at the Brown Center for Politics, Philosophy, and Economics and at the 2023 Meeting of the PPE Society for helpful comments on earlier versions of this paper.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
