Are There Demographic Objections to Democracy?

Abstract

Proponents of epistocracy claim that amplifying the political power of politically knowledgeable citizens can mitigate some of the harmful effects of widespread political ignorance, since being politically knowledgeable improves one’s ability to make sound political decisions. But many critics of epistocracy suggest that we have no reason to expect it to make better decisions than democracy, for those who are politically knowledgeable can also possess other attributes that compromise their ability to make sound political decisions. This is one version of the Demographic Objection to epistocracy, and in this paper I argue that the reasoning which underlies it generalizes in such a way that it applies not only to epistocracy but also to democracy. Thus, there are demographic objections to both epistocracy and democracy. I argue that such demographic objections to democracy have important and overlooked implications. Among other things, they imply that we should be skeptical about the arguments of epistemic democrats; they count against proposals to democratize extant epistocratic practices and institutions; and they even count against certain democratic reforms taking place within a wholly democratic framework.

1. Introduction

A significant body of empirical research shows that political ignorance is widespread throughout democracies.¹ Not only is the typical voter unaware of largely uncontroversial findings within politically relevant fields such as economics or political science, but they also often lack knowledge of even basic matters such as the identity of their political representatives, the track record of electoral candidates, the details of competing party platforms, and more. Many economists, philosophers, and political scientists think that such pervasive political ignorance is what causes, at least in part, many of the problems one can find in modern democracies (Caplan, 2007; Somin, 2013; Achen and Bartels, 2016; Brennan, 2016). If they are right, one might naturally wonder what could be done to mitigate these problems.

One proposal that has received attention in recent years is epistocracy, a form of government that makes the possession of some amount of political information a legal requirement for the possession of political power.² Its proponents claim that epistocracy would produce better results than democracy because we can expect political systems to make better decisions as the influence of politically ignorant citizens is reduced.³ However, a number of scholars have noted that implementing an epistocracy would likely yield some seemingly troubling distributions of political power. Given facts about the distribution of political knowledge among a population, implementing some form of epistocracy would result in some demographic groups being disproportionately empowered while other groups are disproportionately disempowered. Newly empowered groups, many of which are already advantaged in various ways, would see their influence even further enhanced in an epistocracy, while newly disempowered groups, many of which are already disadvantaged, would see their influence even further diminished. This disparity between different demographic groups, it is claimed, is an injustice that constitutes a powerful objection to epistocracy.

This objection has come to be known as the Demographic Objection, and there are in fact two versions of it (Estlund, 2008: 215; Brennan, 2018: 54). According to the first, unequal distributions of political power of the sort we would expect to emerge under epistocracy are intrinsically unjust, regardless of whatever consequences such distributions produce. According to the second, we have no reason to think that amplifying the political power of knowledgeable citizens would lead to better results in the way that proponents of epistocracy expect (Ingham and Wiens, 2021). Indeed, empowering such citizens might even lead to worse results, since such citizens may also disproportionately possess traits which reduce the likelihood that they make correct political decisions. Were this so, any motivation to implement epistocracy would be effectively undercut.

Both versions present a serious challenge to proponents of epistocracy, but in what follows I focus on the second version of the Demographic Objection.⁴ I argue that the reasoning which underlies the Demographic Objection generalizes in such a way that there are parallel demographic objections to democracy. If we were considering a transition from epistocracy to democracy – or, more generally, from some epistocratic institution to a more democratic institution – we would have no reason to think that amplifying the political power of less knowledgeable citizens would lead to better results in the way that many proponents of democracy expect. More precisely, we would have no reason to expect better results related to the ability of newly empowered citizens to make better decisions, since they, like those citizens who would be empowered under epistocracy, may disproportionately possess traits which reduce the likelihood that they make correct political decisions. Such demographic objections to democracy, I claim, have important but overlooked implications for various proposed democratic reforms. Among other things, demographic objections count against transitioning from prevailing epistocratic institutions to more democratic alternatives, as well as counting against certain transitions between different kinds of democratic institutions.

In Section 2, I outline in more detail the Demographic Objection to epistocracy, while in Section 3 I show that there are parallel demographic objections to democracy. In Section 4, I argue that these demographic objections have previously overlooked implications for various sorts of democratic reform. Lastly, in Section 5, I respond to some objections. First, I address the claim that the intrinsic value of democratic institutions renders demographic objections to democracy irrelevant. I argue on the contrary that where well-functioning epistocratic institutions are concerned, demographic objections count against democratic alternatives. Second, I push back against the claim that the Demographic Objection to epistocracy is more serious than any such concerns about democracy because we have extensive experience with well-functioning democratic political systems while having very little experience with well-functioning epistocratic political systems. In response, I claim that this objection misunderstands the skeptical implications of demographic objections. Even if it is true that we have evidence that democratic political systems in general are well functioning, we still cannot be certain that specific democratic reforms that empower specific groups of people will be similarly well functioning. Consequently, demographic objections to various proposed democratic reforms still have some force.

2. The demographic objection to epistocracy

In epistocratic political systems and institutions, political power ought to be distributed such that the political power of more politically knowledgeable citizens is amplified, with the ultimate aim of mitigating the harmful effects of widespread political ignorance. But there are many complications that can arise when attempting to implement such institutions. For example, the methods by which more knowledgeable citizens are identified could be error-prone or otherwise flawed, failing to amplify the political power of all and only suitably knowledgeable citizens (Kogelmann, 2023). For another, the process by which the political power of relevant citizens is amplified could be abused by self-interested actors looking to further their own ends (Bagg, 2018; Klocksiem, 2019: Somin, 2022; Gibbons, 2025). Importantly, the version of the Demographic Objection with which we are concerned can arise even if those responsible for implementing the relevant epistocratic institutions are both using reliable methods to identify more knowledgeable citizens and doing so in good faith, setting aside any interests they may have in abusing their power. Even under unrealistically favorable conditions, then, amplifying the influence of all and only politically knowledgeable citizens may not lead to better results than what we can expect in a democracy.

Ultimately, this is because one cannot infer that someone is more likely to make a correct political decision than others from the fact that they are better informed than others. Such an inference is a non-sequitur, as Ingham and Wiens observe, because “it ignores the realistic possibility that…the explanation for why some people acquire competence also implies that competence is positively correlated with attributes that decrease the probability of voting correctly” (Ingham and Wiens, 2021: 325).⁵ Roughly put, if whatever explains why some people become better informed than others is also positively correlated with some attribute that decreases the probability of making correct political decisions, then even if citizens are well informed they need not be more likely than others to make correct political decisions. This is true even if being well informed, considered by itself, increases one’s ability to make correct political decisions.

Put abstractly, the non-sequitur highlighted by the Demographic Objection is clear. But what sorts of attributes might in fact make someone more likely to be better informed than others while also making them less likely to make correct political decisions? Some examples include demographic variables such as one’s race or gender, whether one has received third-level education, and so on (Estlund, 2008: 206–22; Ingham and Wiens, 2021: 333–6). Attending a university, for instance, might expose one to plenty of politically relevant information, inculcate in one a desire to deliberately seek out political information, allow one to interact with groups taking a keen interest in politics, and so on. Such attributes, especially in conjunction with one another, might make someone more likely to see their power amplified in an epistocracy than a typical person who has not attended a university.⁶ At the same time, though, such attributes might sometimes make one less likely than others to make correct political decisions. For example, one’s membership of the sort of groups likely to attend university might also lead one to develop biases against certain political parties, even in cases where voting for those parties would be the correct decision (Ingham and Wiens, 2021: 334).

Other examples are related to the sorts of motivations one might possess for acquiring political information (Ibid, 336). As noted at the very beginning of this paper, political ignorance is pervasive throughout democracies. A common explanation of this widespread political ignorance is that such ignorance is rational, a reflection of the fact that acquiring political information is very costly while offering slight benefits in return (Downs, 1957; Somin, 2021).⁷ But if the acquisition of political information is not incentivized in modern democracies, then – as noted by Ingham and Wiens – it is worth asking what sort of thing could motivate somebody to acquire political information (Ingham and Wiens, 2021: 337). One plausible explanation for why some people become politically well informed despite the high costs of doing so is that they have partisan motivations to acquire political information. Evidence from political psychology suggests that the most politically well-informed people tend to be highly partisan (Feldman and Price, 2008; Hetherington, 2009; Kalmoe, 2020). However, the available evidence from political psychology also indicates that politically partisan people are more likely than their less partisan (and less knowledgeable) peers to engage in motivated reasoning, processing political information in highly biased ways (Judd and Brauer, 1995; Box-Steffensmeier and De Boef, 2001; Abramowitz and Saunders, 2008; Joslyn and Haider-Markel, 2014).⁸ Accordingly, well-informed citizens can in some cases be less likely to make correct political decisions than their ill-informed peers, if motivated reasoning is clouding their judgment.

Spotting the close connection between levels of political knowledge and political partisanship, some critics of epistocracy have argued that it would likely produce worse results than what we can expect under democracy (Friedman, 2019; Gunn, 2019; Hannon, 2022). But it is important to note that in its most persuasive form, the Demographic Objection implies a weaker, skeptical conclusion – namely, that given the very real possibility that being politically well informed (an attribute that by itself improves one’s ability to make correct political decisions) correlates with attributes that reduce one’s ability to make correct political decisions, we have no reason to expect those who would be empowered under epistocracy to be more likely than those empowered under democracy to make correct political decisions. And since we have no reason to expect more reliably correct political decisions in an epistocracy than in a democracy, we have no reason “to seriously entertain epistocracy as a practical proposal” (Ingham and Wiens, 2021: 326).

3. Demographic objections to democracy

The Demographic Objection severs the purported link between being politically well informed and the ability to reliably make correct political decisions. It highlights the possibility of attributes that correlate with being well informed while at the same time compromising one’s ability to make correct political decisions. Without being able to rule out such “confounders,” we have no reason to expect an epistocracy to make better decisions – and hence produce better results – than a democracy (Ingham and Wiens, 2021: 331). If so, we have no compelling reason to transition from a democracy to an epistocracy. More precisely, we have no compelling reason related to epistocracy’s ability to generate better decisions and better results to attempt such a transition.⁹

Proponents of epistocracy might be tempted to respond to the Demographic Objection by trying to find some way to identify people who are both suitably knowledgeable and free of whatever attributes might cause one to otherwise have a reduced ability to make correct political decisions. For example, one could propose the use of “refined selection mechanisms” that can identify people who are both sufficiently knowledgeable about politics and less prone to the sorts of biases and politically motivated reasoning just discussed in the previous section (Gibbons, 2022: 266–70). But I wish to set aside this sort of response in what follows, granting that it is infeasible to design selection mechanisms that can identify such individuals. Instead, I argue that the very same sort of demographic considerations that cast doubt upon the ability of epistocracy to generate better results than democracy can also be used to cast doubt upon the ability of democracy to generate better results than epistocracy.

At its core, the Demographic Objection shows that attributes that plausibly improve one’s ability to make correct political decisions (such as being politically knowledgeable) can correlate with other attributes that can reduce one’s ability to make correct political decisions (such as partisanship). But it is easy to overlook how widely this form of reasoning generalizes. First, such reasoning can be used to show that the sort of people who would be empowered under political systems other than epistocracy possess attributes that can diminish their ability to make correct political decisions. Second, although standard presentations of the Demographic Objection concern the decision-making ability of individuals, similar objections can be formulated about groups (such as electorates). So, an entire electorate might possess attributes that plausibly increase its collective ability to generate correct political decisions while also possessing other attributes that detract from this ability. Taken together, these points suggest that there can be demographic objections, both at the level of individuals and at the level of the electorate, against non-epistocratic political systems, including democracy.

That there can be demographic objections pitched against the sort of people who would be empowered in democratic political systems should be obvious. Indeed, standard arguments for epistocracy can be viewed as invoking demographic considerations to cast doubt upon the ability of democratic voters to reliably make correct political decisions. Suppose, for example, that one endorses the view that each person is the best judge of their own interests, and that having regard for one’s own interests can be regarded as an attribute that increases one’s ability to make correct political decisions. If so, this counts in favor of democratic distributions of political power granting universal suffrage.¹⁰ However, as already mentioned, epistocrats can respond to such claims by pointing to a vast empirical literature documenting persistently high levels of political ignorance among democratic electorates. One might initially think that people are in general the best judges of their own interests and that they will vote accordingly. But without knowing which of the available options really are in one’s best interests – and, of course, to know this requires one to know much about the available candidates, their policies, their likely efficacy, and more – it isn’t clear that merely having regard for one’s own interests can help one make the correct decision. In short, while having regard for one’s own interests can be viewed as an attribute that increases one’s ability to make correct political decisions, being politically ignorant is an attribute that diminishes such an ability.

Other examples more closely parallel the sorts of attributes that feature in the Demographic Objection to epistocracy. Earlier we noted that membership in certain groups (such as those who have received a third-level education) might also lead one to develop biases against certain political parties, thereby causing one to vote against such parties even when doing so is the right decision. Even if attending university is likely to leave one better informed than those who have not attended university, one may not always be more likely to make correct political decisions. But nothing about such dynamics, where membership in a certain group or class causes one to develop certain biases, is unique to the sort of people who would see their power amplified in an epistocracy. As Estlund writes when outlining the Demographic Objection to epistocracy, “people are inevitably biased by their race, class, and gender” (Estlund, 2008: 215). If this is true, it is as true of those groups who would not see their power amplified in an epistocracy as it is of those groups who would see their power amplified. So, even if it were true that politically ignorant people lack the particular biases that the politically knowledgeable are burdened with, they may possess their own biases against certain groups, parties, or viewpoints, and these biases will sometimes lead them to make incorrect political decisions.¹¹

Consider one final example, this time at the level of the electorate rather than at the level of the individual. Proponents of epistemic democracy contend that its “tendency to make correct decisions provides an important defense of democracy” (Schwartzberg, 2015: 188).¹² Drawing upon mathematical theorems such as the Condorcet Jury Theorem and the Hong-Page Theorem, epistemic democrats variously claim that by diversifying the electorate, or by increasing the number of people who can vote, we can increase its ability to make correct political decisions.¹³ Thus, under the right conditions, democratic decision-making can be superior to epistocratic decision-making. Even if the former involves individually ignorant people while the latter involves politically knowledgeable people, the collective performance of the former can be superior once the group possesses the relevant attributes. But as with attributes that can improve an individual’s ability to make correct political decisions, attributes of groups that are conducive to reliable decision-making can be counterbalanced by other attributes that detract from reliable decision-making. For instance, even if a sufficient amount of diversity is an attribute that increases a group’s ability to correctly make political decisions, diverse democratic electorates can also possess attributes such as being polarized or being ignorant that can nullify any advantages that diversity might otherwise yield.¹⁴

The Demographic Objection to epistocracy states that since we have no reason to expect more reliably correct political decisions in an epistocracy when compared to a democracy, we have no reason to entertain epistocracy as a serious practical proposal. It does this by showing that attributes that plausibly increase ability to make correct political decisions can co-occur with other attributes that detract from such an ability. But now we have seen that the very same sort of considerations can be used to show that we have no reason to expect more reliably correct political decisions in a democracy when compared to an epistocracy. Whatever attributes one might think that voters in a democracy (or democratic electorates) possess which enable them to reliably make correct political decisions can be counterbalanced by other attributes. Thus, demographic considerations count against the ability of both epistocracy and democracy to reliably make correct political decisions.

4. Demographic objections and democratic reforms

Some of the most prominent proponents of the Demographic Objection have recognized that the sort of considerations it appeals to apply equally to democracy and epistocracy. For instance, David Estlund writes that “[neither] equal voting nor departures from it can be defended…on the basis of invidious epistemic comparisons” (Estlund, 2008: 221). And Ingham and Wiens, making explicit the skeptical implications of the Demographic Objection, acknowledge that “it does not provide a reason to endorse universal suffrage” (Ingham and Wiens, 2021: 325). But unlike in the case of epistocracy, one might reasonably wonder whether such conclusions have any further implications for democracy. While the Demographic Objection seemingly rules out epistocracy as a serious practical alternative to democracy, demographic objections to democracy don’t entail that we ought not to maintain existing democratic institutions. Demographic objections to democracy perhaps close off one potential argument in its favor, but otherwise they have no further implications.

However, such a conclusion would be premature, and demographic objections to democracy have implications that should not be overlooked even if it does not follow from them that we ought not at least maintain existing democratic institutions. If we are focusing only on their comparative ability to make correct political decisions, demographic objections would also rule out democracy as a serious practical alternative to epistocracy if, counterfactually, we were in an epistocracy rather than a democracy. Moreover, they also rule out various sorts of democratic reforms within democracy which are sometimes defended, at least in part, on the grounds that they would generate better results than other democratic institutions.

Let’s begin with the former point. If you’re reading this paper, you are likely reading it while living in a democracy. For you, democracy is the norm, while epistocracy is a controversial and largely untested alternative. But things could have been different. You could have been living in a more epistocratic political system, perhaps one with greater restrictions placed on the franchise, perhaps with more demanding qualifications required for political office, or perhaps some combination of both. You might occasionally encounter articles or books (as well as political activists and the movements of which they are a part) arguing that we should transition to a more democratic political system. Some of these arguments would appeal to the requirements of justice itself, treating the likely consequences of democratic rule as of secondary importance. But others would appeal directly to the supposed instrumental benefits of democratic rule – democracy is not only required by justice, but it also would be “smarter” than epistocracy and produce better results.

However, the problem for such arguments should now be clear, for we now know that they face demographic objections. If, counterfactually, we lived in an epistocratic political system and democratic rule was a controversial and largely untested alternative, we would have no reason related to its ability to make correct political decisions to entertain it as a serious practical alternative. Arguments in favor of democracy that appeal directly to the requirements of justice might ultimately persuade us, and there could even be instrumental considerations in its favor that are unrelated to its ability to make correct political decisions.¹⁵ Nonetheless, an important class of arguments would be closed off to us. In such a counterfactual scenario, the arguments offered by epistemic democrats and others would be dismissed as mere conjecture, relying upon “heroic assumptions about how all the myriad influences on voting behavior are jointly distributed within a population” (Ingham and Wiens, 2021: 325–6).

One might view this as a result that we can safely ignore since the relevant counterfactual is so far removed from our actual situation. But that isn’t quite correct, for at least two reasons. First, pervasive democratic rule and universal suffrage are relatively recent phenomena. In the not-too-distant past, states all around the world disenfranchised approximately half of their adult population by denying women the right to vote. Eventually, arguments in favor of extending the franchise to women won the day. But it is notable that demographic objections threaten to undercut the subset of those arguments which appealed to the greater ability of systems with universal suffrage to make correct political decisions and thereby generate better results. John Stuart Mill, for example, famously defended the political rights of women in his essay The Subjection of Women. Many of his most powerful arguments highlighted the flagrant injustice of denying women the right to vote. But it is also clear that he thought expanding the franchise would lead to better results, characterizing the then status quo as “one of the chief hindrances to human improvement.” Moreover, he expends a considerable amount of effort addressing criticisms to the effect that women possess certain attributes that make them unfit for the franchise. Such criticisms can be viewed as offering demographic objections to universal suffrage, and a troubling implication of our discussion so far is that these sorts of demographic objections are difficult to decisively rebut.¹⁶ That is true of the Demographic Objection to epistocracy, but it is also true of calls to expand the franchise in ways that most would accept. As critics of epistocracy concede, demographic considerations cut in both directions.

Second, we need not look to the past or to counterfactual scenarios for less democratic political systems, for all existing political systems have epistocratic features, including paradigmatically democratic political systems. A notable example of an epistocratic practice found within modern democracies is the electoral exclusion of children. “All democracies are weakly epistocratic,” writes Jason Brennan, pointing to the fact that “all modern democracies exclude children from voting and holding office, on the ground that children are incompetent” (Brennan, 2011: 701). Other epistocratic enclaves embedded within more broadly democratic systems include independent central banks, independent judiciaries, and the civil service (Jones, 2020). Whether such institutions ought to be more or less democratic is a difficult question. But what is relevant for our purposes is that arguments to the effect that democratizing such institutions would allow them to more reliably make correct political decisions are ruled out by demographic arguments. The same is true of arguments that lowering the voting age would not diminish the ability of electorates to make correct political decisions.¹⁷ At least some epistocratic institutions and practices, then, are indirectly supported by demographic arguments.

Something similar is true of certain democratic reforms that are widely viewed as taking place within a democratic framework. For example, consider recent debates about how to improve democratic decision-making regarding the long-term future. Many philosophers and political scientists have noted with dismay the tendency of democratic institutions to focus on delivering short-term results while neglecting or even ignoring long-term problems (Caney, 2016; González-Ricoy and Gosseries, 2016; John and MacAskill, 2021). And perhaps unsurprisingly, many democratic reforms have been defended on the grounds that they would improve the ability of democracy to address long-term problems. Some of the most prominent and well-known such proposals involve amplifying the political power of younger generations, whether by more heavily weighting their votes, imposing youth quotas on legislative bodies, or by some other means.¹⁸ As should be clear by now, though, such proposals face demographic objections: one might view younger people as possessing attributes (having a greater stake in the future, for example) which make them more likely than older people to make correct political decisions involving the long-term future, but these attributes might be nullified or outweighed by other attributes (such as immaturity) which make younger people less likely than older people to make such decisions. For all we know, then, implementing such reforms might decrease the ability of democracy to address long-term problems. Demographic objections suggest that we should be skeptical either way.

So, far from having no implications for democracy, demographic objections to democracy count against certain types of proposed democratic reforms. In the same way that epistocracy is ruled out as a serious practical alternative to democracy, demographic objections to democracy rule out transitions to more democratic forms of existing epistocratic institutions, as well as ruling out transitions from our current democratic institutions to different types of democratic institutions. Additionally, demographic objections highlight the perhaps uncomfortable fact that which political systems we have reason to entertain as serious practical proposals is hostage to contingent facts about whichever political systems happen to prevail at any given time. If we were living in largely epistocratic political systems, we would be viewing at least some prominent arguments for democracy and universal suffrage with suspicion. At the very least, we would remain skeptical about arguments for democracy that made claims about how it is epistemically superior to epistocracy.

5. Objections and replies

There are demographic objections to both epistocracy and democracy and, as we have just seen, demographic objections to the latter are not without important implications. But one might still think that there is nonetheless a way in which the Demographic Objection to epistocracy is more serious than any demographic objection to democracy. First, one might argue that demographic objections to democracy are irrelevant since democratic political institutions are intrinsically valuable. Second, one might argue that the Demographic Objection to epistocracy is more serious than any demographic objection to democracy because, while we have extensive experience with well-functioning democratic political systems, we have very little experience with well-functioning epistocratic political systems.

5.1. Demographic objections to democracy and the intrinsic value of democracy

We noted earlier that there are in fact two types of demographic objection to epistocracy. One kind, the Demographic Objection that is the main focus of this paper, contends that we have no reason to expect the sort of people who would be empowered in an epistocracy to be reliably better at making correct political decisions than the sort of people who would be empowered in a democracy. If so, we have no reason to entertain epistocracy as a serious practical alternative to democracy. However, we have seen that the very same reasoning applies to democracy in such a way that wherever epistocratic institutions and practices prevail, we have no reason to entertain democratic alternatives as serious practical alternatives. At the very least, we have no such reason related to the purported ability of these alternatives to more reliably make correct political decisions than their epistocratic counterparts. But recall that the other kind of demographic objection to epistocracy focuses rather on the fact that the unequal distributions of power that would likely emerge under epistocracy are intrinsically unjust. So, even if it is true that considerations raised by the Demographic Objection apply equally to both epistocracy and democracy, there is arguably an important asymmetry between epistocracy and democracy when it comes to the other kind of demographic objection – specifically, epistocracy is intrinsically unjust, while democracy is intrinsically valuable. And one might think that this asymmetry renders demographic objections to democracy more-or-less irrelevant. Perhaps it is true that where epistocratic institutions and practices prevail, we cannot be sure that democratic alternatives could match their ability to reliably make correct political decisions. All the same, the intrinsic injustice of the former, coupled with the intrinsic value of the latter, is enough reason for us to replace epistocratic institutions with democratic alternatives.

Suppose it is true that democratic institutions are intrinsically valuable in a way that epistocratic institutions are not.¹⁹ Perhaps democratic institutions are intrinsically valuable because of their equality, perhaps they are intrinsically valuable because they provide a fair way to make collective decisions in conditions of deep disagreement, or perhaps something else.²⁰ Suppose also that epistocratic institutions can have at most instrumental value insofar as they enable us to secure good outcomes (or at least better outcomes than available alternatives). Even so, it does not follow that we all-things-considered ought to transition from any epistocratic institutions or practices to more democratic alternatives. That something has intrinsic value does not mean that such value outweighs any amount of instrumental value. And if certain epistocratic institutions and practices have instrumental value then, at the very least, we should be cautious about transitioning to democratic alternatives when we are uncertain about whether such alternatives will be able to match the performance of those epistocratic institutions they are supposed to replace.

Whether any epistocratic institutions in fact have greater instrumental value than their democratic alternatives is something I return to in the following section. For now, some clarifications about this response are in order. First, why be uncertain about whether democratic alternatives can match the performance of their epistocratic counterparts? The simple answer, of course, is that demographic objections of the sort we have heretofore been concerned with induce uncertainty about the performance of alternatives to prevailing institutions, whether epistocratic or democratic. Demographic objections have skeptical implications. Thus, if there are demographic objections to certain democratic institutions, then we should be skeptical about whether such institutions would have the same instrumental value as those they are intended to replace.

Second, why think that the intrinsic value of democratic institutions could be outweighed by the instrumental value of epistocratic institutions, supposing that the latter can in fact possess greater instrumental value? Generally speaking, the view that any amount of intrinsic value can outweigh any amount of instrumental value has deeply implausible implications. Such a view would entail the result that a decision-making procedure with intrinsically valuable features that reliably produces catastrophically bad consequences is more valuable than a decision-making procedure that lacks intrinsically valuable features but reliably produces good consequences. It should therefore be conceded that instrumental value can outweigh intrinsic value, in at least some cases. If so, then epistocratic institutions with only instrumental value could in principle be more valuable all-things-considered than intrinsically valuable democratic institutions.

Admittedly, even setting aside such an implausible view, there is likely to be reasonable disagreement about how to trade off intrinsic and instrumental value in cases where they conflict, such as when choosing between intrinsically valuable institutions that are less instrumentally valuable than their less intrinsically valuable competitors. Perhaps in some cases, intrinsically valuable institutions are all-things-considered preferable to more instrumentally valuable institutions that lack intrinsic value. Still, it is worth noting in this vein that almost all theories of democracy, including theories which accord democracy intrinsic value, are willing to permit the use of undemocratic decision-making procedures provided that doing so is required to prevent sufficiently bad outcomes.²¹ At least in cases where it is necessary to use epistocratic institutions to prevent sufficiently bad outcomes, then the instrumental value of such institutions outweighs even the intrinsic value of their democratic alternatives.

Putting these two points together – that it is uncertain whether democratic alternatives can match the instrumental value of prevailing epistocratic institutions, and that the instrumental value of epistocratic institutions can outweigh the intrinsic value of democratic institutions – it seems clear that we should be cautious about replacing extant epistocratic institutions with democratic alternatives when we cannot rule out the possibility that the former are necessary to avert the sufficiently bad outcomes that the latter could bring.²² The intrinsic value of democratic institutions, then, does not render demographic objections to democracy irrelevant.

5.2. On the skeptical implications of the demographic objection

If the Demographic Objection which is the central focus of this paper is successful, it has skeptical implications. It establishes that we have no reason to treat epistocratic counterparts of democratic institutions as serious practical alternatives, but it also establishes the same of democratic counterparts of extant epistocratic institutions, as well as various widely defended democratic reforms. However, one might think that such skeptical implications are more pronounced for epistocratic institutions than for democratic institutions. After all, is it not the case that we have considerable evidence that democratic institutions function at least tolerably well while having little to no evidence that epistocratic institutions can function similarly well? Indeed, one might think that the force of the Demographic Objection stems not just from the fact that we cannot rule out the possibility that epistocratic institutions cannot make correct political decisions as reliably as democratic institutions but also from the fact that epistocratic institutions are largely untested proposals for which we lack evidence either way. If this is right, we have yet another asymmetry between democracy and epistocracy which seems to make the Demographic Objection more serious for the latter than for the former.

The first part of this objection – that we have much evidence for the claim that democratic institutions function reasonably well – is difficult to deny. In fact, this is something that even critics of democracy accept. In the preface to Against Democracy, for example, Jason Brennan writes that “democracy is positively correlated with a number of important outcomes,” citing the fact that democracies “do a better job of protecting economic and civil liberties than non-democracies,” as well as the fact that they tend to be wealthier than non-democracies (Brennan, 2016: ix). Others claim that democracies are better at mitigating climate change than non-democracies, that democracies experience fewer famines than non-democracies, and more.²³

But even though it is true that there is much evidence for the claim that democracies in general function well, this objection fails. First, while there is indeed much support for the claim that democracies function well, the other part of the argument for the claim that the Demographic Objection is more serious for epistocracy than for democracy – namely, that epistocratic institutions are largely untested institutions for which we lack evidence – is not as plausible. Much like we have evidence for the superior performance of democracy over its non-democratic alternatives, we have evidence for the superior performance of at least some epistocratic institutions over their non-epistocratic alternatives. Second, and more fundamentally, this objection misunderstands the force of the skeptical implications of demographic objections. Even if there is evidence that democratic institutions in general function well, demographic objections imply skepticism about specific democratic reforms that would empower a specific group of people (in a specific decision-making context, and so on).

Regarding the first point, we earlier took note of various “epistocratic enclaves,” epistocratic institutions which can be found embedded within more broadly democratic political systems. Recall that such institutions include independent central banks, independent judiciaries, and so on. To focus on one example, central banks are typically independent, epistocratic institutions insofar as they are staffed by experts and insulated from the vicissitudes of democratic politics. But they could be more democratic, without a legislative guarantee of independence, and more accountable to elected officials. How do such independent central banks compare to their more democratic alternatives? As Garett Jones persuasively argues, it appears that independent central banks are more effective than their less independent, more democratic counterparts, delivering lower rates of inflation, steadier economic growth, and more (Jones, 2020: 41–62). Something similar is true, he argues, of other epistocratic institutions such as appointed city treasurers and appointed judges, both of which are plausibly viewed as more epistocratic than available democratic alternatives (Ibid, 63–94). Such evidence belies the claim that epistocratic institutions are untested proposals for which we lack any reason to think they can outperform their democratic alternatives.²⁴

Regarding the second, more fundamental point, it is important to be clear about the force of the skepticism implied by demographic objections. Such objections foreground the fact that decision-making in political contexts is influenced by a potentially large number of attributes, how these attributes are distributed throughout a population, and how they interact. Considered in isolation, some attributes (such as being well-informed or being diverse) may improve the ability of individuals or groups to make correct political decisions. But for any novel institutional proposal, we cannot rule out the possibility that these attributes will be counterbalanced or completely outweighed by other attributes (such as being ignorant or being polarized) that diminish the ability to make correct political decisions. What is most important for our purposes is that we cannot always tell ex ante which attributes will most heavily influence decision-making. When considering a specific institutional reform, we would need evidence that empowering the relevant group of people in that particular decision-making context would lead to improved decision-making, and this is something that appeals to other institutions empowering different groups of people (each with different attributes, in different contexts, and so on) cannot provide. To be sure, we may learn ex post whether the reform in question produces better results than its predecessor. But this is as true of epistocratic proposals as it is of proposals to reform institutions in a democratic manner. So, the Demographic Objection, if sound, still cuts against both epistocratic and democratic reforms.²⁵

6. Conclusion

The Demographic Objection to epistocracy says that since we have no reason to expect epistocracy to be better able to correctly make political decisions than democracy, we therefore lack any reason to take epistocracy as a serious practical alternative to democracy. It does this by highlighting the fact that one can be politically well informed, an attribute that by itself increases one’s ability to correctly make political decisions, while also possessing other attributes that diminish one’s ability to make such decisions. However, this form of reasoning generalizes to other attributes and, more importantly, to other institutions and political systems. Political systems such as democracy are held by many of their proponents to possess certain attributes that make them better than available alternatives at making correct political decisions. But demographic objections, once applied to democracy, say that such attributes can be counterbalanced by other attributes that detract from this ability. Thus, demographic objections would suggest that if we were considering a transition from an epistocratic institution to a democratic institution, we would be unable to rule out the possibility that the latter simply cannot function as well as the former. Ultimately, we would lack any reason related to the latter’s ability to make correct political decisions to treat it as a serious practical alternative to whatever epistocratic institution it is supposed to replace.

Initially, one might think that such demographic objections are less serious for democracy than they are for epistocracy. For one thing, democratic institutions possess intrinsically valuable features that count in their favor, while epistocratic institutions can have at most instrumental value. For another, we have much evidence that, generally speaking, democratic institutions can perform reasonably well, while epistocratic proposals are largely untested. But such considerations do not diminish the force of demographic objections to democracy. First, the intrinsic value of democratic institutions can in some cases be outweighed by the instrumental value of epistocratic alternatives. Second, the skeptical concerns raised by demographic objections cannot be undercut by such general evidence. What is instead required is evidence that empowering certain groups in certain contexts will lead to improved decision-making and thereby improved outcomes. And this is something we typically lack, both for proposed epistocratic reforms and proposed democratic reforms. So, there are demographic objections to both epistocracy and democracy.²⁶