1. Preliminaries

Let us, first of all, spell out a few background details and assumptions that are central to the argument that follows.

First, for the purposes of this paper we will work under a few assumptions concerning political legitimacy. Namely, we will treat democracy and voting as default options, so to speak. At least for the purposes of this work, we will bracket the issue of legitimate alternatives to democracy and proceed on the assumption that political legitimacy requires democratic arrangements. In a similar fashion we will assume that, in terms of public decision-making practices, there are no viable alternatives to standard democratic voting. We will hence not consider recent proposals for alternatives to democratic voting, such as modelled democracy or lottocratic mechanisms (Ahlstrom-Vij Reference Ahlstrom-Vij2020; Guerrero Reference Guerrero, Hannon and de Ridder2021). Correlatedly, while we will discuss the issue of voters’ epistemic competence, we will not consider competence-based restrictions on the right to vote; we will assume that all adult citizens have the right to vote regardless of their level of political competence.Footnote ²

Second, and most importantly, our work here rests on a fundamental premise that needs to be spelled out. Namely, we think the value of the right to vote has an epistemic dimension. This means that if voters vote without a sufficient level of epistemic competence, this represents a problem for democracy in that the value of their right to vote is not fully realized. We understand epistemic competence in voting as the ability to exercise certain core epistemic capacities. We will detail these capacities later, as we defend our view, but it is worth anticipating that we understand them in sufficientarian terms, as a threshold level of capacities. Thus, the ability to exercise these epistemic capacities on a sufficiently high level represents a prerequisite for the full realization of the value of the right to vote.

This key premise can be defended both in instrumentalist and non-instrumentalist terms. Let us begin with instrumentalist considerations, as many others do. In recent years, a significant number of political theorists have called upon the plethora of evidence about the incompetence of average voters precisely in order to cast doubt on the ability of democracy to deliver good political decisions (Caplan Reference Caplan2007; Guerrero Reference Guerrero2014; Brennan Reference Brennan2016; Mulligan Reference Mulligan2018). On these and similar views, voters tend to approach the political choices available to them in ways that betray superficiality and negligence at best or ignorance and irrationality at worst. This, in turn, increases the chances of bad political decisions that are not conducive to anything like justice or the common good and that potentially cause harm to fellow citizens. These widespread instrumentalist arguments seem to rest on the premise of the epistemic dimension of the value of the right to vote. For they seem to argue precisely that, insofar as incompetence undermines the instrumental value of votes as tools for bringing out good political decisions, voting rights are devoid of value if exercised in incompetent fashion.

The instrumentalist argument from voter ignorance, of course, has been challenged. Some political scientists hold that citizens, in the aggregate, possess enough competence, and bad outcomes are better explained by the disproportional influence of affluent social strata on democratic decision-making (Gilens Reference Gilens and Berinsky2019). Epistemic democrats have called upon social epistemology to defend democracy against these criticisms. From their standpoint, democracy is far from an epistemically defective political system. By allowing for a wide diversity of inputs, democracy benefits from mechanisms of collective intelligence that not only compensate for the political incompetence of some citizens but that give rise to unique epistemic qualities, thus making democracy a far more reliable system for good political decision-making than any other alternative (e.g., Landemore Reference Landemore2012; Goodin and Spiekermann Reference Goodin and Spiekermann2018). Even then, however, one may argue that individuals need at least some, not necessarily trivial, level of competence to properly participate in the collective procedures that exhibit the epistemically advantageous properties epistemic democrats discuss. For the purposes of this work, we need not take sides on this debate; it suffices to note that there are instrumentalist arguments for the claim that the value of the right to vote is conditional on voters’ competence, even though these arguments are debated.

However, even if epistemic democrats are right in arguing that individual incompetence in voting is less of a problem from an instrumentalist standpoint than it might have appeared at first glance, we think that it remains nevertheless a problematic phenomenon from a non-instrumental point of view. To vote is to make a political choice. When we vote for a candidate or platform P over Q, we are making a choice with respect to the two political alternatives on offer. We are choosing P over Q. Now, if we assume that there is some non-instrumental value in having the opportunity to make this choice, such value must encompass an epistemic component. This need not be taken as an excessively controversial point. All we are committing to here is the idea that whenever a person has to make a choice, it is generally better for that person to do so under proper epistemic conditions, such as having proper information, being in the position to properly compare and assess the various options on the table, etc. If one does not know what one chooses, what one's options are, and what consequences one's choice may have, then the value of that choice is severely hindered. If in a restaurant I am given a menu in a language I do not speak, I can still technically choose what I order, but I might as well throw some dice. Under these conditions, the possibility of choosing has lost some of its worth, as I am not in the position to understand how my choice is being employed. The same applies to the value of making political choices. Such value partially derives from being able to make informed judgements about politics. Being able to make informed political judgements (or have informed preferences) enhances both the ability of the voter to form appropriate expectations concerning the outcomes of her choice as well as her ability to make choices that are consistent with her values (Lovett Reference Lovett2020: 15–16).

Of course, even if our choice is uninformed or incompetent, it may be important to us that we make it rather than others making it for us. And, of course, we do not mean that the value of the right to vote, whether instrumental or non-instrumental, is exhausted by this epistemic dimension. But if the right to vote is indeed an opportunity to make a political choice and shape the terms of our social and political coexistence according to one's preference or judgement, making such a choice under conditions of incompetence places the voter under disadvantageous epistemic circumstances in which the meaningful exercise of this opportunity is hindered.Footnote ³ Consequently, something valuable is lost even if mechanisms of collective intelligence were indeed to compensate for this at the collective level as epistemic democrats claim.Footnote ⁴

To sum up, our position is premised on the idea that the value of the right to vote can be fully realized only insofar as the holder can exercise such a right in a competent fashion. This fundamental premise does not commit us to any specific conception of the grounds of political legitimacy, as it is perfectly compatible with both instrumentalist and non-instrumentalist views. On a standard instrumentalist view, competence is a precondition of the value of voting rights, for only exercising voting rights in a competent fashion is conducive to good or correct political choices. On a non-instrumentalist view, competence is a precondition of the value of voting rights, for only exercising voting rights in a competent fashion allows the right holder to exercise a genuine political choice, unhindered by obstacles that might thwart their judgements.

Having clarified this fundamental premise, let us turn to the goals of this paper. Although elections and referenda are collective procedures, voting is an individual action. As a result, most debates on the normative questions of voting have been centred on individual voters and on their responsibility in cultivating the aforementioned epistemic capacities. However, if the value of the right to vote has indeed an epistemic component as we just argued, then we cannot ignore how the state plays a key role in shaping the epistemic circumstances in which most people exercise their right to vote. In what follows, we will focus first on explaining why the state has weighty responsibilities in this regard and then move on to provide an account of these responsibilities. Our analysis, therefore, is meant to widen the scope of the debate on these issues by shifting the focus on what we take to be a neglected topic.

Of course, focusing on the responsibilities of the state should not be understood as entailing that voters have no responsibilities or duties of their own. Our claim is just that the state shares responsibility with citizens for the extent to which they are effectively capable of voting competently. We will return on this in the final section of the paper. For the time being, the point is that the inquiry into the responsibilities of the state should be seen as complementary to the inquiry into the responsibilities of voters, rather than antithetical to it. Consequently, we are convinced that providing an account of the specific responsibilities of the state would also contribute to a correct and balanced view of the responsibilities that can be legitimately placed on voters.

2. Why the state has a duty to foster competence

If voter incompetence is indeed a problem, how are we meant to solve it? The literature on the ethics of voting has suggested some options. For example, Brennan argues that in a democracy, incompetent voters have a duty to abstain to ensure that their epistemically deficient choices cause no harm to others (Brennan Reference Brennan2011: 101). Christiano argues that the adverse effects of voter incompetence should be mitigated by the complex division of epistemic labour between experts and lay voters. Expert knowledge should be transmitted to the general population and to decision-makers by mediators whose own expertise allows them to translate expert findings into a vocabulary intelligible to citizens. “These chains of overlapping intelligibility enable politicians and citizens to have some appreciation of the reasons for and against particular policies” (Christiano Reference Christiano, Parkinson and Mansbridge2012: 39). Others argue that voter incompetence should be mitigated by suspending judgement and deferring to expert opinion (Blake Reference Blake2019).

Interestingly enough the literature rarely discusses a quite natural suggestion, namely, that to solve the problem of voter incompetence, the state should foster voter competence: the state should create favourable epistemic conditions for citizens to acquire political competence and, if needed, educate and train them to ensure that they use their voting powers in a competent manner. We argue that there are several reasons to believe not only that this would be advantageous, but that in fact the state has a duty to foster voter competence in certain ways. We present two arguments for this claim, one based on an instrumental, one on a non-instrumental conception of democracy. Both accounts are compatible with the assumptions and remarks offered in the previous section.

Let us first discuss the instrumental account which holds that fostering voter competence through state action is potentially a good means to mitigate the problem of voter incompetence. Voter incompetence is often treated as a fact of democratic politics that cannot be changed. But clearly, voter competence greatly depends on the particular epistemic circumstances in which voters find themselves, and these circumstances are, to a large extent, under the control of the state. Voters develop the epistemic capacities (and acquire the political knowledge) necessary for a competent exercise of their role as voters against a background of resources and conditions that is largely determined by the institutions of the political community to which they belong. Most citizens develop much of their basic epistemic competences in public education run by the state. Public and private media, including social media and online platforms, operate in a regulatory environment created by the state. Think about the state's policies on free speech, advertisement, broadcasting, internet service, libel, and so on. They play a momentous role in how political information circulates through the media, whether private or public, in a given political community. One can hardly doubt the importance of the state's role in regulating access to political information and epistemic competence in light of the fact that authoritarian states, such as China or Russia, are often very successful in stifling private media and hindering the spread of political information.

The fact that in developed democratic countries the state most often does not directly interfere with the circulation of information among citizens does not mean that it is not responsible for maintaining the wider institutional framework which greatly determines how this circulation takes place. To be sure, the state should not have dictatorial command over how the media operates or how information circulates. In modern liberal societies media and education should enjoy various protections against state discretion. Moreover, media ventures that operate internationally and within many jurisdictions can function in relative independence of the state. Nonetheless, it remains true that the state's authority to create and enforce a regulatory environment as well as to develop and maintain a society-wide infrastructure for the circulation of information gives it vastly greater power over the epistemic circumstances in which voters make their choices than what private individuals usually have.

Now, if voter incompetence is indeed a crucial political problem to be addressed, then given the state's tremendous power over the institutional background that constitutes much of the epistemic environment in which voters develop competence, it is unclear why the state should not use this power to foster voter competence by creating a better epistemic environment, with robust opportunities for voters to become more competent. From a strictly instrumentalist point of view, it is hard to see why solving the problem of voter incompetence should be a matter of individual behaviour alone. It is not clear why the task of ensuring an effective circulation of relevant information should be entirely up to the individual efforts of experts or ordinary citizens, when the state has a vastly better chance to exert substantial influence on the epistemic environment that shapes the quality of voter's competence.

The claim that fostering voter competence through state action would be advisable from a strictly instrumental point of view is further supported by the fact that the problem of voter incompetence is to a large extent a collective action problem. A major reason why voter incompetence exists is that while it would be socially beneficial for the electorate to be competent, given the incentive structure of individual voters, it is individually rational for each voter to remain incompetent. This is the problem of rational ignorance, as it is sometimes called (Brennan Reference Brennan2016: 35). Now, collective action problems are normally thought of as most effectively solved by central coordination, i.e., by a regulatory body that is able to alter the incentive structure of individuals such that their individually rational choice matches the socially desirable outcome. In the case of the problem of voter incompetence, it is quite natural to think of the democratic state as this regulatory body. This would mean that insofar as such instrumental considerations can generate duties for the state, the state has a duty to foster voter competence which derives from its duty to solve the problem of voter incompetence.

Suppose, however, that one finds instrumental considerations unsatisfactory or implausible. There are still reasons to think that the state has a duty to foster voter competence. If one subscribes to a non-instrumentalist conception of the value of democracy, one probably also believes that the state has a duty to ensure that citizens are included in the democratic decision-making process. For non-instrumentalists, the imperative of democratic inclusion is likely on a par with the state's duty to ensure justice or the protection of basic human rights. Clearly, democratic inclusion requires more than universal franchise in the merely formal sense. It also requires doing away with informal obstacles to meaningful participation in democratic politics, i.e., of guaranteeing what Rawls calls the fair value of political liberty (Rawls Reference Rawls1999: 197–8).

Our view is that excessive voter incompetence may preclude meaningful participation and hence have exclusionary effects. Recall our remarks in the previous section concerning the epistemic dimension of the value of the right to vote. To vote is to choose. In voting, citizens make consequential choices on political matters. If our considerations above are correct, the value of such choices partially depends on how competently one can make them. It depends on having, at least to a certain extent, the ability to understand the nature and implications of one's choice and how it is to be compared with alternative options. If, as we argued, the meaningful exercise of voting rights depends on this epistemic dimension, those who find themselves incapable of mastering sufficient epistemic competence are hindered in participating in voting practices in a meaningful way, to the point that this might have exclusionary effects.

What we mean by this is that although voters continue to possess the formal right to vote, if poor public education, difficult access to quality reporting on public matters, the lack of academic freedom, the overflow of fake news and political propaganda make it extremely demanding for them to access and evaluate relevant information on the basis of which they could vote in an informed and independent manner, then they are effectively forced to shoot in the dark. Their right to vote is stripped from a central component of its value and meaning. Such epistemic circumstances effectively hinder the meaningful exercise of voting rights. And thus if, as non-instrumentalist accounts of democracy hold, the state has a duty to ensure that voters are not excluded from meaningful participation in democratic decision-making, neither by discriminatory election laws nor by informal obstacles such as a bad epistemic environment, then the state also has a duty to foster epistemic competence so that citizens reach the threshold necessary for a meaningful exercise of voting rights.

Again, this does not rule out that voters can be held responsible and accountable for their own incompetence. If they never read the news and do not try to stay up to date on political events, then what precludes the meaningful exercise of their voting rights is their own carelessness and epistemic negligence. But if voters constantly have to fight an uphill battle for attaining reliable information about politics because the state fails to set up proper regulatory environments, large-scale technological infrastructures and institutional background conditions that enable the circulation of information and know-how within society, then incompetence cannot exclusively be a matter of voters’ negligence.

In those cases, it seems to us, those who accept the state's duty to guarantee not only formal, but substantive democratic inclusion, must also accept that the state owes it to citizens to rectify the situation and help them attain competence. In our view, this suffices to show that the state does in fact have a duty to foster the epistemic competence of voters, which can be established both on instrumentalist and non-instrumentalist grounds. In both cases, of course, this duty is construed as a pro tanto duty, defeasible by some conflicting duties and countervailing considerations. As we will see, the state faces specific difficulties and risks in fostering voter competence and when those become excessive, the state may not try to foster voter competence. However, our argument does not imply that only the state has responsibility to foster voter competence. Non-state actors, from individual citizens to civil society organizations, may also have such responsibilities, and such actors could sometimes be as well or even better positioned to address the issue of voter competence as the state. We will not further discuss the role of these actors, however, for here we are only interested in the state's duty to foster voter competence. In the next section, we focus on the precise content of this duty as well as on the limits to which it might be subjected.

3. The content of the state's duty

Having established that the state has a duty to foster voter competence, we turn to the content of this duty and what it entails precisely. Note, first, that the state's duty to foster competence is subject to various moral constraints, such as liberty and fairness. Even if the forced detention of voters in re-education camps would make them incredibly competent, which is of course dubious, this would not be a permissible way to foster competence, as it would violate voters’ liberty. Or suppose that aggregate competence could be best promoted by fostering competence only for a specific subsection of the electorate. However epistemically beneficial such a policy would be all things considered, it would likely be unfair to the rest of voters to leave them behind, as it were, in terms of competence. Finally, we must keep in mind that the state's duty to foster voter competence is a pro tanto duty and may come into conflict with other duties of the state that must take precedence, e.g., protecting basic rights or ensuring the just distribution of the benefits and burdens of social coexistence. When such conflicts arise, the state may have to decide not to foster voter competence but attend to more important duties.

Given these constraints, what should be the precise goal of the state's efforts to foster voter competence? In other words, what is precisely the content of the state's duty? We believe that the state should aim at ensuring that voters are competent enough to have substantive opportunity to attain further levels of political competence. According to the way we understand the concept, a person has substantive opportunity to X when she is in a good position to deliberate about choosing X over some alternative Y or vice versa. As Scanlon notes in his discussion on substantive opportunity:

Applied to the case of epistemic competence itself, Scanlon's principle supports the following account of the conditions for substantive opportunity to attain epistemic competence. Voters, on this view, have substantive opportunity to attain epistemic competence in political decision-making when they are in a good position to deliberate about choosing to improve their competence rather than engaging in some other activity that they deem worthwhile. This requires, on the one hand, creating favourable epistemic circumstances in which the costs of attaining competence are not prohibitively high, and, on the other, improving voters’ competence directly. This does not simply mean feeding citizens political information, or directly imparting political knowledge to them, but also building and developing voters’ epistemic capabilities. That is, building those capabilities that would make it possible for voters to further expand their competence and to become even more informed should they wish to do so. This includes capabilities to find and evaluate information on politically relevant matters, to understand and assess politically relevant arguments and policy proposals, and so on. All this involves some amount of knowing-that, but also knowing-how: for example, the intellectual skills and habits required for proper engagement in certain kinds of analytic thinking or argumentative activity.

Once again, there are both instrumentalists and non-instrumentalist reasons to accept this view concerning the content of the state's duty to foster voters’ competence. Let us start with the non-instrumentalist argument this time. Recall that from a non-instrumentalist point of view, the main worry about voter incompetence is that it undermines the value of choice in voting, thereby precluding the meaningful exercise of the right to vote. What triggers the duty of the state to foster voters’ competence is, from this standpoint, a commitment to guarantee that the meaningful exercise of the right to vote is not hampered. Consequently, the level of competence the state has a duty to guarantee is determined in function of this commitment. This, coupled with the understanding that the state and voters share responsibility for epistemic competence in voting, provides a plausible case for the broadly sufficientarian account of the content of the state's duty offered above. Voter incompetence has the kind of exclusionary effect that triggers the state's duty to intervene insofar as it is the result of factors beyond the responsibility of the voter, such as the excessive costs or the absence of the epistemic capabilities needed for attaining and further cultivating one's epistemic competence. An implication of this is that state's intervention in this regard would not guarantee a meaningful exercise of voting choices on part of all voters, as some of them may still deliberately decide to remain incompetent or neglect to improve their competence. But while self-imposed incompetence of this kind may undermine the value of one's choice in voting, this loss of value does not have the same exclusionary effect that externally imposed incompetence does, and which would compel the state to intervene.Footnote ⁵ We believe that voters with substantive opportunity to attain competence are put in a sufficiently good epistemic position to make informed and meaningful choices about the exercise of their voting rights, and therefore have no further claims on the state in terms of fostering their competence.Footnote ⁶

From an instrumentalist point of view, substantive opportunity for competence does not seem to be a sufficiently ambitious goal. For instrumentalists, the main problem of voter incompetence concerns the quality of political decisions. Hence, one may argue that the state's goal in fostering voter competence should be ensuring that voters reach a maximum level of competence; the level which is most conducive to high quality political outcomes. Keep in mind, however, that even an instrumentalist argument must observe the constraints on the state's duty to foster competence mentioned above, i.e., liberty, fairness, and competing duties. Liberty dictates, for example, that voters retain some discretion as to how competent they wish to become, and therefore the state cannot permissibly force them to always be as competent as possible. It seems, therefore, that even instrumentalists should conceive of the state's duty to foster competence in a sufficientarian manner, as being aimed at establishing a threshold level of competence within the electorate.

It makes sense for instrumentalists to identify this threshold as the level of substantive opportunity, for this fosters the efficient distribution of competence within the electorate without burdening the state with the exceedingly complex task of centrally planning the whole of the epistemic economy of society. Voters with substantive opportunity for competence are in a good position to ascertain the benefits of becoming more competent, the dangers of failing to do so and hence possess the capacity to act upon their assessment without excessive costs. They can be trusted to make a good decision about whether or not they should attain more competence. Without substantive opportunity to competence, this decision would likely be based on partial information and the use of insufficient epistemic capabilities. Aiming at a higher level of competence, on the other hand, would risk misallocating resources, as it would compel some to spend time in acquiring further competence for whom it would have been better to spend that time otherwise.

Let us now turn to what granting voters substantive opportunity for competence entails. While here we cannot present a detailed policy proposal, we believe that it is worth outlining some possible measures by means of which the state may discharge its duty to foster voter competence. As mentioned already, fostering competence involves, on the one hand, creating good epistemic conditions for the acquisition of competence. This likely requires regulatory measures against the unrestrained spread of disinformation and propaganda, the use of manipulatory strategies in electioneering, and so on. On the other hand, fostering voter competence so that voters have substantive opportunity to attain further competence also involves improving their knowledge and epistemic capabilities. Concrete measures to this effect may be public education reform by way of strengthening civic education at the primary and secondary level. While this may be one way in which the groundwork of epistemic capacity building may be laid, we believe that the sufficient development of epistemic capabilities likely has to involve continuous exercise and training well into adulthood. Furthermore, keeping up with current events requires continuous exposure to political information, which is why in discharging its duty to foster voter competence the state must engage not only school children, but also adult citizens.

But how can the state ensure that voters exercise their epistemic capabilities and receive sufficient political information without compromising their liberty, i.e., without re-education camps and the like? We believe that there are benign solutions that avoid the threat of unfreedom while allowing voters to attain substantive opportunity to competence. A few examples of potential arrangements can be drawn from proposals associated with deliberative approaches to democratic legitimacy. For instance, citizens could be given the chance of developing and refine their political views through mini-publics or other educative or deliberative forums (Fung Reference Fung2003: 340–1; Goodin and Dryzek Reference Goodin and Dryzek2006). Another option could be offering informative classes on issues relevant to voting and public decision-making or upcoming ballots. Such classes might even include decision-making exercises where citizens could gain insight into the moral and political complexity of voting through the simulation of hypothetical voting scenarios. The state could subsidize and incentivize participation in these ‘training programmes’ without coercing anyone to partake. These are, of course, merely some examples that indicate the general outlines of how our proposal may be implemented. Determining what particular arrangement and policy solutions best suit any given community is bound to be a contingent matter. We doubt that a ‘one size fits all’ approach is plausible or even desirable in this sense.

We also recognize that, under some circumstances, our view might be difficult or too costly to implement. Once again, we recognize that the duty to foster epistemic competence is a pro tanto duty. As previously noted, considerations about the scarcity of resources or conflicting duties of greater import may defeat it. But recall that fostering competence serves valuable social goals; on the non-instrumentalist view, it is meant to ensure the meaningful exercise of the right to vote, while according to instrumentalism, it is supposed to contribute to better political decision-making. Both goals are worth the resources spent on them. Furthermore, we do not believe that fostering voter competence would always be exceedingly costly, for it would essentially build on already existing practices which the state seems capable of sustaining. Our proposal calls for the stricter regulation of societally detrimental misinformation, a greater supply to the public of political information and enabling the public to better process this information through public education and potentially some form of state-sponsored adult learning, all of which exist today in some form or another. If states can already bear these costs, we don't see any principled reason why it could not bear the cost for what concerns the cultivation of a sufficient level of political competence.

4. Objection: anti-pluralism and condescension

Now we turn to what we take to be the most serious objection against the view we have developed in the paper, namely, that our argument rests on patronizing and disrespectful assumptions about ordinary citizens’ capacity for political judgement. On the one hand, state policies to foster epistemic competence presuppose that, if left untrained, citizens would not be able to cultivate the relevant epistemic capacities on their own and hence that they would not be capable of exercising the right to vote in any meaningful way. Furthermore, such policies would have to treat as settled political matters that are subject to reasonable disagreement and controversy in a pluralist society. After all, what state agency could authoritatively decide what citizens have to know in order to participate in political decision-making meaningfully?

Does our argument have anti-pluralist and patronizing assumptions and implications? We do not think so. Let us start by addressing its alleged anti-pluralist implications. Note, first, that state policies regularly take sides regarding contested issues, for example, in regulating public education, journalism, hate speech, and so on. What should be taught at schools or what should count as hate speech is no less controversial than what level of competence citizens ought to have in order for the meaningful exercise of the right to vote to be possible. Yet if the state is to develop and enforce any effective policy in these areas, it needs to treat these issues as settled at least to some extent.

Of course, any such policymaking procedure is fallible, and may erroneously estimate the level to which certain contested issues can be treated as settled. Perhaps state agencies and policymakers wrongly assess the level of scientific consensus in deciding what should and should not be part of school curricula in public education. Perhaps because of certain biases they wrongly estimate the adverse effects of certain forms of speech in designing legislation on hate speech. It is also true that any such measure creates incentives for political actors to try to capture and use them to advance partisan interests. The risk of this happening, e.g., the risk of certain parties using institutions of fostering competence for the purpose of propaganda and manipulation, should be taken into account in weighing the costs and difficulties of implementing these institutions.Footnote ⁷ Nonetheless, as with other institutions requiring the state to take sides on contested issues, the risk of fallibility and partisan capture can usually be reasonably minimized by establishing transparency and accountability in the decision-making process, by subjecting it to democratic control, and ensuring that it can and will be regularly reviewed and revised so as to continuously improve it and guard it from illegitimate interference.

Similarly, in designing state policies aimed at fostering competence, pluralism and neutrality are best respected by ensuring that those in charge of designing these policies operate in a transparent and accountable manner, that the decision-making process is subject to input from the wider population and that it is regularly reviewed and revised so as to counteract bias and human fallibility that is characteristic of any kind of policymaking process. Possible practical solutions include allowing political parties, civil society organizations, and other associations to send their own representatives or experts of their choosing to these programmes and talk to voters in a properly moderated setting, appoint institutional watchdogs tasked with securing the neutrality of the programmes, emphasize dialogue and interactive exercises and so forth.

One may argue that, perhaps precisely because of the contested nature of these matters, the state may only make very minimalistic commitments about them. For example, the state should play no role in determining the school curriculum at any level and should only designate as hate speech instances where speech is most clearly and directly connected with acts of physical violence. Similarly, when it comes to fostering voter competence, the state should play little to no role in it, given the controversial nature of any proposed training course or other kind of policy. However, pursuing such a minimalistic route itself counts as taking sides on controversial issues. Whether the state's involvement with these matters should be minimal in light of their contested nature is itself a matter on which reasonable disagreement exists. Committing to this type of minimalism on a principled basis without future prospect for review and revision appears to be a far less neutral and pluralist option compared with our proposal of a transparent and accountable decision-making process checked by political parties, civil society organizations, and so forth. Such a process may at one point settle on a more minimalistic interpretation of what it means to provide citizens with substantive opportunity for competence, or it may not. But the outcome is subjected to a decision-making procedure that while aiming at addressing a pressing matter for democratic decision-making, takes seriously the challenge of bias and human fallibility posed by it. Thus, we believe that the objection that our argument entails anti-pluralist implications can be countered.

Yet one might still suspect that our argument expresses a patronizing and disrespectful attitude towards citizens. By assuming that citizens’ competence must be fostered, it denies them the standing of autonomous agents capable of exercising their rights responsibly, and instead treats them as agents who are in need of being helped out. To see why this objection is also not convincing, consider, first, that part of the state's duty to foster competence is to eliminate those obstacles to the acquisition of competence that individual voters have neither power nor authority to do away with on their own, e.g., inadequate education or media policy. Such measures hardly express any condescension as they do not presuppose anything about citizens’ innate ability to attain competence. Rather, they rest on an assessment of the epistemic environment in which voters acquire and exercise these capacities and of the institutions that shape it.

Second, any argument regarding the role of personal autonomy in voting needs to take proper account of the fact that as democratic citizens we act in a specific social role wielding specific powers not only over our own individual lives but also over the lives of others. When people occupy such roles, for example, if someone is acting in a specific role such as teacher, juror, or professor, it is hardly patronizing or disrespectful to express concern about their competence. After all, they influence the lives of others in profound ways. Saying that concern about citizens’ competence is inherently patronizing or disrespectful presupposes that voting is not the exercise of a public role but rather a private or personal matter. Our view could be regarded as inherently disrespectful only on condition that we accept a conception of voting powers that portrays them primarily as personal resources. While some philosophers seem to hold such a view (Christiano Reference Christiano1995; Goodin and Tanasoca Reference Goodin and Tanasoca2014), such conception has been questioned quite extensively (Gaus Reference Gaus1996: 248–51; Wall Reference Wall2007).

Admittedly, voting does have a personal dimension – for example, it can be viewed as implicating personal autonomy and empowerment. Through the vote, a citizen can contribute to shaping the terms of the social and political coexistence of the community according to their political views and values, and thereby uphold them. But for the same reason, voting is an exercise of political power that cannot be regarded purely as a personal resource. Voting is not just a personal choice. It is an input to a decision-making procedure that is meant to shape the terms of our social and political coexistence as a community and that contributes – albeit indirectly – to alter the nature, scope and features of the coercive laws by which we will all have to live. Voting is, therefore, also the exercise of a public responsibility.

Now, appealing to the public responsibilities associated with voting might be seen as a problematic move. The worry might be, more precisely, that such a move is in tension with our emphasis on the state's role in controlling the epistemic environment and background conditions in which voters exercise their rights. How can citizens have responsibility to vote well, if their ability to do so depends on the state? The tension is, however, only apparent. First, the fact that the state shapes the background conditions of voting practices only commits us to the idea that the responsibilities associated with the role of voter must be constrained according to this fact and proportionated to the limited control that voters have over the epistemic circumstances in which they are acting. Limited control entails limited responsibilities, but it by no means entails no responsibility at all. This is consistent with our claim, put forward in the first section of the paper, that the state and voters share responsibility for their competence.

Second, appealing to the public responsibilities associated with voting is something that strengthens our account rather than anything else, as it provides yet another argument for the state's obligation to foster voters’ epistemic competence. Let us go back to the idea that the value of voting should be understood as analogous to the value of a choice. Our claim was that the value of such choice is meaningful only insofar as citizens are indeed in the position to make the choice under proper epistemic conditions. The idea that citizens at the same time have certain responsibilities of their own for what concerns how they make such a choice does not threaten our previous claim. It just means that bad epistemic conditions, where it is extremely costly for citizens to attain an appropriate level of epistemic competence, have a further problematic feature. Namely, not only do they put voters in a position where they cannot make a meaningful choice; they also create conditions under which a voter cannot properly fulfil the public responsibilities associated with such choice. Not even abstaining is safe for them, for under bad epistemic conditions they are not able – unless they assume extreme costs – to ascertain whether they do more harm than good by abstaining. This puts them in a very precarious position from the moral point of view. We believe the state has a duty to alleviate this moral precarity, for only the state has the appropriate resources and power to improve the epistemic background conditions in which voters operate. And a significant way to do this is to foster their epistemic competence and create substantive opportunity for a meaningful exercise of democratic decision-making powers. Of course, there is much more to explore about this supplementary argument from moral precarity, but further discussion of it would exceed the scope of this paper.