Dan Williams made an interesting comment in a note about this essay: “Interesting post. I suspect the driving force is rather simply that people instinctively treat democracy and political egalitarianism as socially enforced sacred values - the rest is post hoc rationalisation.” This made a lot of sense to me, and I’d enjoy learning more—including a conversation between the two of you about this, if you are both so inclined.
This reads like an argument against electoral democracy and I agree.
With the complexity and quantity of political information, it’s almost impossible to ask people to make informed voting decisions.
This is why I advocate for Citizens Assemblies to operate alongside elected bodies to keep them accountable. On the broadest level, replacing the upper house in bicameral systems with a large group of randomly selected citizens would answer both the knowledge and polarization problems we are experiencing now. Gathering people to learn from the same sources, discuss the knowledge face to face in small groups and then recommend and respond to the actions of elected representatives would create a much different set of laws than we have now.
This is just the start for me, I would love to see a four day work week with the fifth being devoted to self-governance and/or civic engagement. Most of our lives are too burdened with work and responsibilities to be able to participate in the active world of politics. As a culture, you can’t grant autonomy if you don’t also grant time. If there is one day a week that people can give to each other to discuss the direction of their common lives we can take control of them back. It would be time for neighbourhood associations, school councils, unions, city planning, sports leagues, crafting guilds. We could have citizens sitting on corporate boards, federal department advisory committees, focus groups. We are asked to give so much to the culture, but the quantity of what we can give is limited by the time we are allowed. It’s no surprise that politics is low on our list of things to care about.
Yes, tests can be manipulated for political advantage as US history shows. Birth certificates are not subject to such manipulation, which is why age criteria don't pose that danger. Of course, setting the appropriate voting age may be a politically contested question.
Great read. I agree wholeheartedly with your final conclusion, though I wasn’t sure that was where you were ultimately headed, which speaks to your even-handedness here.
> If widespread political ignorance threatens democracy, one logical response is to ensure that more knowledgeable citizens have greater influence over political decisions.
How does limiting voting rights to a more educated minority help ensure democracy, if democracy is most simply defined as a system of government by the whole population or all the eligible members of a state?
>On the other hand, they are quick to dismiss epistocratic proposals—such as limiting voting rights to the more informed—on the basis that we cannot accurately identify who is knowledgeable. This contrast is striking. If we can determine who is politically ignorant, why can’t we determine who is politically knowledgeable? Aren’t these just two sides of the same coin?
As you later point out, it seems your students, like me, are skeptical that a system sufficiently safe enough from corruption can be put in place for choosing which people within an ostensible democracy are allowed political agency. A predictable result of such epistocratic proposals is that the political situation will bend toward favoring the highly educated few who meet the voting requirements that they themselves put in place.
> If we accept that infants and young children lack the competence to vote responsibly, then we have already conceded that some threshold for political knowledge exists.
This is true, but we also wouldn’t tax, jail or enlist these people to fight on behalf of the rest of us. The more the state asks of a person, the more say that person should probably have in regard to what the state does.
Thanks for reading the post and for your kind words about it. I think you’re right that restricting voting rights to a more educated minority wouldn’t strengthen democracy, at least not democracy as we know it. That said, Alex Guerrero defends a model called Lottocracy, where a smaller, randomly selected group of citizens is given the time and resources to make decisions on behalf of the population at large. Since he doesn’t see democracy as necessarily requiring rule by the whole population (above a certain age), there’s room to argue that universal suffrage isn’t a strict requirement of democracy. Nevertheless, I generally agree with you that epistocratic suffrage is hard to reconcile with democratic principles.
I also appreciated your point about the relationship between state obligations and political participation: “We also wouldn’t tax, jail, or enlist these people to fight on behalf of the rest of us. The more the state asks of a person, the more say that person should probably have in regard to what the state does.” I completely agree. There are strong reasons—beyond purely competence-based considerations—why certain groups should not be excluded from voting. That said, I do think the epistemic concerns about voter competence remain an important factor here.
>That said, Alex Guerrero defends a model called Lottocracy, where a smaller, randomly selected group of citizens is given the time and resources to make decisions on behalf of the population at large.
I think I’ve read about this, or a similar system, that was instituted in school governments with great success. It would be a wonderful thing to award power to those who don’t seek it at all costs, and it would certainly stifle the influence of entrenched interest groups and solve some other growing problems associated with traditional elections. There’s promise there, though I’m wary of the kind of upheaval that would be needed before such a sea change in governance could take place.
I’d be interested to read more of what you think about such a system, should you post on it.
I may write about lottocracy here soon. There's a chapter about it (and epistocracy) in the book I just finished writing with Elise Woodard, "Political Epistemology: An Introduction." In case you're curious, here's an interview I did with Alex Guerrero on his book: https://www.youtube.com/watch?v=RKPO5uB2YtQ
This essay thoughtfully explores the puzzle of why students recognize political ignorance as problematic yet resist epistocratic solutions. However, I'd like to challenge a fundamental assumption: that the right kind of political ignorance is actually widespread.
There's an important distinction between not knowing certain political facts (like the location of Ukraine or the names of elected officials) and not knowing political information that's relevant to one's interests and lived experience. This distinction relies on Gerd Gigerenzer's work on ecological rationality, which suggests that humans are remarkably good at developing knowledge and making decisions in domains that matter to their specific circumstances.
Gigerenzer and others have suggested that people often employ "fast and frugal heuristics" that are well-adapted to their particular environments. In "Simple Heuristics That Make Us Smart" (1999), they argue about how seemingly limited reasoning strategies can be highly effective when matched to the right environments. Similarly, in "Gut Feelings" (2007), Gigerenzer illustrates how intuitive judgments can outperform complex analysis in certain contexts.
Applied to political knowledge, this perspective suggests that voters may be ignorant of abstract political facts while being quite knowledgeable about issues that directly affect them. A farmer might not name every Supreme Court justice but may deeply understand agricultural policies. A parent might struggle with identifying foreign leaders but comprehend education policies intimately. Herbert Simon's concept of "bounded rationality" is relevant here too—people develop specialized knowledge within the constraints of their cognitive resources and personal stakes.
I wonder if the "ocean of data" on political ignorance adequately controls for this context-specificity. Do these studies test for knowledge that's actually relevant to voters' lives and interests? Or do they primarily measure abstract political knowledge that, while perhaps valuable for academic discourse, may be peripheral to how people actually make voting decisions? I will confess to my ignorance of the literature on whether it addresses the view I've given in this comment.
But even if there is a sea on this ocean of data that does address the view in this comment, empirically examining whether people possess interest-relevant political knowledge presents significant methodological challenges. It's not as straightforward as simply asking people to identify their interests and then testing their knowledge in those domains. People may misidentify their own interests due to limited self-awareness, social desirability bias, or ideological commitments. They might also strategically misreport interests to appear more informed or to align with perceived group norms.
Furthermore, what constitutes "relevant" knowledge may itself contested. A person's objective material interests might diverge from their perceived interests or their identity-based commitments. The factory worker who votes against labor protections because of deep-seated cultural values isn't necessarily ignorant—they may be prioritizing different aspects of their complex set of interests. This makes designing studies that accurately capture interest-relevant political knowledge extremely difficult, as researchers must navigate the thorny issue of defining what knowledge "should" matter to different individuals without imposing their own normative frameworks.
These methodological hurdles suggest we should approach claims about widespread political ignorance with caution. The standard political knowledge batteries used in most studies (at least those I'm aware of) may be measuring the wrong thing entirely if they fail to capture the situated, context-specific knowledge that actually guides voting decisions in practice.
If people are reasonably knowledgeable about political matters that affect their interests, then perhaps what's being labeled as "ignorance" is actually a rational allocation of cognitive resources. This doesn't mean we shouldn't strive for better civic education, but it does suggest we should be cautious about labeling voters as "incompetent" based on standardized knowledge tests that may not capture the situated knowledge that actually guides their political choices.
I close by noting that, again, I could be entirely ignorant of some relevant papers in the literature—if so, I am happy to have someone remedy my ignorance.
I definitely agree that standard political knowledge tests often miss the kind of knowledge that actually matters to voters and political competence. And heuristics can be useful! But I’m skeptical that they’re enough. One reason is that politics isn’t just about personal interests—it’s about navigating trade-offs across many domains. A farmer might know more about agriculture policy, but that won’t help them assess tax policy or judicial appointments, and the trade-offs matter too.
I think the bigger issue is whether voters’ heuristics reliably track political reality. Research on misinformation and partisan bias suggests they often don’t. People regularly misattribute policies to the wrong parties or support politicians who actively undermine their interests. So even if heuristics help, they don’t always lead to informed choices. Some of the heuristics are themselves the result of various biases. Perhaps I'm being too pessimistic about this, but I totally agree with you that a lot of what these political knowledge tests measure is just "trivia" that isn't importantly related to the kind of competence that matters at the polls.
How about candidates who are not ignorant? Dumb voters is one thing—and understandable; they (we) have jobs, family, other distractions. But candidates, they’re saying serving us (me) is all they will do. The voter trusts/believes that the candidate knows what they’re talking about and, as they take that oath, will faithfully execute. Yes, it is our duty to be informed, but, speaking of impacts and effects, it is the duty that much more of the candidate to know what is going on – and then not to lie to us about it all.
I will confess to laughing out loud at seeing the last text on the list: “Harwood, Robin, 1998. More Votes for PhDs.” Though I recognize I am responding to the title, not the article itself (which I would like to get hold of in full and read—so far, I am encountering a paywall), what immediately came to mind was that, if I were the decided on this, some of these folks, let’s say Judith Butler as one example, should be top of the list to be PERMANENTLY disqualified from voting.😎 The discussion here is fascinating, and I look forward to what comes next, though I will remain skeptical of proposed solutions that select out voters as unqualified to vote, even in a single election. Anyone remember literacy tests in the US and who that excluded from voting? (I write this with affection for the project here, just also with the bemused recognition as to why it was that undergrad philosophy classes and I were not the best of friends!)
I really like the larger observations and explanations you offer here, especially the last piece about moral encroachment (and the "meta-encroachment" of corrupted incentives for determining any standard, great point!). Yet something about your initial premise, and the way you set up the dilemma, feels off. I'm not sure that the "asymmetry" you identify between political ignorance and political knowledge is such a mystery (let alone paradox), or even as asymmetric as you say. Is it possible that this puzzle you're wrestling with is mostly an artifact of comparing apples and oranges?
Let's say we have good reason to think (anecdotally, and having seen some polling) that there's a problem of widespread ignorance among voters, and this isn't just our own political bias. This will be based on very obvious cases of ignorance: people saying or doing "ignorant" things, offering dubious explanations, getting most test questions wrong. That doesn't mean we have the ability to discern where the *threshold* is for ignorance, any more than we can discern the threshold for sufficient (or minimal acceptable) political knowledge. It simply means we know egregious ignorance when we see it, the kind that on its face seems problematic for anyone voting. So when you say, your students can determine who is politically ignorant so "why can’t we determine who is politically knowledgeable": the appropriate standard of comparison is not the ability to determine who is knowledgeable, or assess people on that. We merely need to be able to recognize who is *not* grossly ignorant such that the idea of them voting isn't scary. In that sense it really is just two sides of the same coin, since in practice we don't assess ignorance vs. knowledge but ignorance vs. non-ignorance, and we don't assess this at the boundary but closer to the extremes. And this doesn't require any complicated policy or test; it's just common sense. (Of course there's still some inherent asymmetry, by virtue of the general principle that it's much easier to falsify something than to determine that it's true using the right standard - but that's hardly unique to politics, it's a fact of life).
None of this speaks to the separate question of how much political competence is a function of being well-informed as opposed to other things, or whether it's even possible to abstract "knowledge" from the capacity and disposition to use it well. Here, I agree with Susan Scheid above that factors like trust and motivation matter much more than knowing the three branches of government, in the scheme of things. Nevertheless, it *is* concerning to see evidence of mass voter ignorance. I think this is not because that particular knowledge is necessarily important, but because widespread ignorance can be a soft sign that other things are amiss, which might track more closely with actual competence.
Thanks for this response. I agree that the general concern abou ignorance vs. the challenge of identifying a specific threshold for political competence are different questions, but as you said, “we know egregious ignorance when we see it, the kind that on its face seems problematic for anyone voting.” So, why not use this to set the minimal boundary for competence? If we can confidently identify cases of extreme ignorance, it seems like we could at least exclude those cases rather than requiring a precise threshold at the boundary to adjudicate hard cases.
You also mention that “the appropriate standard of comparison is not the ability to determine who is knowledgeable, or assess people on that. We merely need to be able to recognize who is not grossly ignorant such that the idea of them voting isn’t scary.” I completely agree—and I tried to make this point in the post, particularly in the section on whether ignorance is easier to spot than knowledge. My argument wasn’t that we need to identify the most politically competent, but rather that we could use clear cases of ignorance to define a lower bound for participation.
Thanks for responding and sorry if I ended up echoing the same points you were already making - we are probably on the same page about most of this. Perhaps I wasn't clear. I was trying to leave aside those thornier questions of policy and determining official standards to focus on the basic tension you outline at the beginning, which to me felt illusory in that it wasn't an apple-to-apples comparison: your students' resistance to epistocracy was based on the severe standard of policy (moral enchroachment, construct validity and the whole nine yards), whereas their confidence to affirm widespread ignorance was based on the much lower standard of informal secondhand evaluation of stuff they see in the media and encounter on the ground (not to mention, preexisting bias). Clearly you understand all this. But the way you were setting it up as a mystery still felt misleading, so it seemed important to push back a little.
With respect to policy: I guess my real issue is not that we can't make global assessments of pronounced ignorance or decide on some minimal standard for stuff people should know. It's rather that "knowledge" (at least, the kind we can test) is hardly the same thing as political *competence*. If anything, that hoary term "critical thinking" - sound judgment, solid reasoning, paying the right kind of attention to the right things - is much closer to the kind of epistemic resources I associate with competence. And even that shouldn't be a blunt screening tool! (I don't think voting age requirements is a good test case; that's applied a priori based on demographics, not performance and assessment).
If each person, including the test authors, is ignorant about a somewhat different range of issues from the areas of ignorance of everyone else, then these tests would seem necessarily to isolate only a subset of these areas, and in so doing could introduce or reinforce blind spots in electoral choices rather than allowing different blind spots as good a chance to cancel each other out by incorporating feedback from more people.
I agree that diversity of perspectives can help eliminate blind spots in decision-making. If ignorance were randomly distributed across the electorate, then aggregating a wide range of viewpoints could help correct for individual gaps in knowledge.
But I think this optimistic picture leaves out an important complication: ignorance can also compound ignorance. Not all blind spots are independent or randomly distributed—some forms of misinformation or misunderstanding tend to cluster within particular social, ideological, or informational environments. In these cases, simply adding more perspectives to the mix doesn’t necessarily correct errors; instead, it can amplify them. If large numbers of people share the same fundamental misconceptions about political reality, incorporating more input from the uninformed could lead to worse, not better, decision-making.
I wonder if a threshold question here might be whether we are even asking the right question. That is, do voters need to be “informed about politics” in order to be competent as voters? While, as most of us likely do, I have taken this as a given without much thought, I now find I am not so sure. First, what does it mean to be “informed about politics”? For example, I know a number of highly educated, politically attuned people who have no idea who their city or state elected officials are. I have been in that situation myself. One tends not to know these things if the people in question don’t seem to offer anything of value to voters. (I would say my state senator is very much like that. I do know her name, but I also experience her as someone who is warming a seat with little to no interest in addressing her constituents’ needs.)
I don’t have a fully formulated thought here (to say the least), but I do wonder whether, in evaluating voter capacity to make good judgments in selecting their representatives, the first place we may need to look is at the severe erosion of trust in our governmental and other institutions on which we have ordinarily relied, at least where I am in the US. We all, necessarily, must look to trusted interlocutors to help us assess what the best approaches are to address the multitude of complex issues we face individually and as a society. If we lose trust to such an extent that we don’t feel we can rely on much of anyone who represents us, is in a position of authority, and/or has certain expertise, then how can it be otherwise than that we will be handicapped in making the political judgments we need to make? No amount of civic education, as one example, can counter that. For one, we would have to trust those doing the civic education.
I think what people do know is what they need from their government, and they have good judgment about whether they are getting what they need or not. If not getting what they need persists over a long period, and also if the society in which they live is under continuous stress, then it is hardly surprising for frustration to boil over. In the US, we have had three back to back spiritual and material crises of enormous scope: 9/11, when we learned the world was much smaller, and we were much more vulnerable than we had thought; the 2008 financial crisis, which ruined the lives of many families and hollowed out whole communities, some of which never recovered; and the pandemic, the shocks and aftershocks of which I think we vastly underestimate and continue to experience today.
In the case of the last two, particularly, the people negatively affected were often in the same demographic and blighted region of the country. Were they helped by their government? Not nearly enough. It is not in the least surprising to me that, after years of this, the frustration boils over and, as a workman once said to me who felt that frustration, “we should just blow the whole thing up.” Civic education will not help with that. Helping lift people up who need our help is all that will. And, just as, once things start spiraling downward, it is hard to stop that, so too, if people do get help, and that continues, it should be possible to rebuild trust. At least I hope so!
Thanks for this really insightful response. I largely agree with everything you say, and I think there are strong criticisms of the very idea that social scientists can adequately measure the kind of knowledge needed to be a competent voter. The erosion of trust in institutions is also a hugely significant factor—if people don’t trust the sources of political information, or the system itself, then even the most well-designed civic education efforts won’t address the deeper issue.
In the post, though, I was hoping to largely set those broader concerns aside and focus on a what seemed like a puzzle: my students clearly do think these political knowledge tests are measuring something important, otherwise they wouldn’t be so concerned about the kind of ignorance these tests reveal. Their attitude is what I find interesting—on one hand, they take the data on political ignorance seriously, but on the other, they resist the idea that we can determine who is sufficiently knowledgeable to vote responsibly. I was trying to make sense of that tension, rather than defending the legitimacy of political knowledge tests in a more fundamental way. But I recognize that this is largely bracketing some of the more interedeting and foundational issues you mentioned in your comment!
You wrote “The issue isn’t that ignorance is easier to identify than knowledge; rather, the consequences of error are far more severe when deciding who is knowledgeable enough to vote responsibly”. I’m not sure about this. It depends on the method of letting a person vote, right?
Maybe the prospective voter has to pass a five-minute quiz at the voting booth, a quiz on all the elections she can vote on *at that moment*. If she passes, she gets to vote. If she doesn’t, then she doesn’t get to vote. But failure doesn’t mean she can’t vote in any other election. So, the consequences of error are pretty limited: she doesn’t get to vote *that day alone*. I wouldn’t call that “severe”. No?
The harder part is writing the questions so that people agree this is a good idea.
Thanks for this—you raise an interesting challenge. You’re right that the severity of the consequences depends on how voter competence is assessed and enforced. If failing a knowledge test only disqualified someone from voting that day—without barring them from future elections—the stakes would indeed be lower than in systems where incompetence results in long-term disenfranchisement.
That said, I still think a version of the concern holds, even in a system like the one you describe. For one, if the quiz is structured in a way that systematically disadvantages certain groups (due to wording, assumptions about relevant knowledge, or implicit biases in question selection), the aggregate impact could still be significant, even if individual failures only apply to a single election. Second, perhaps even a small barrier to voting could create larger effects—e.g., people might avoid showing up at all if they fear public embarrassment or failure, especially in communities with historical voting restrictions.
I also completely agree with your last point—the hardest part isn’t necessarily the idea of a test itself, but designing one that is widely accepted as fair and politically neutral. That, I suspect, is where things get much more difficult.
I was thinking that you take the online quiz in the voting booth, and no one but you knows whether you passed it. If you failed it, you can still leave the booth and lie about it. So, no stigma. I thought the quiz would be multiple-choice, you would have something like three chances to pass it, you'd get ten questions each time, chosen randomly from a database, and so on.
As for systematic disadvantages, that's a problem for the design of the quiz. I really have no idea what the questions should be. Maybe some general ones, like "The US president is head of which branch of government?" and "Which branch of government passes the nation's laws?". But there needs to be ones specific to the candidates I suppose, and I don't know what they would be at all. I think it might be easy for politicians to game the system, by lying about their policies and proposals.
I know that I'd *like* to stop people from voting who thought that Trump really would lower all grocery prices on day 1, like he claimed. That's idiotic. Or people who thought Biden was part of a federal pedo ring involved in sex trafficking. I don't know how to handle that at all.
Then there are the problems about down ballot things, like local races for justices. Virtually no voters know anything about any of those candidates.
Dan Williams made an interesting comment in a note about this essay: “Interesting post. I suspect the driving force is rather simply that people instinctively treat democracy and political egalitarianism as socially enforced sacred values - the rest is post hoc rationalisation.” This made a lot of sense to me, and I’d enjoy learning more—including a conversation between the two of you about this, if you are both so inclined.
This reads like an argument against electoral democracy and I agree.
With the complexity and quantity of political information, it’s almost impossible to ask people to make informed voting decisions.
This is why I advocate for Citizens Assemblies to operate alongside elected bodies to keep them accountable. On the broadest level, replacing the upper house in bicameral systems with a large group of randomly selected citizens would answer both the knowledge and polarization problems we are experiencing now. Gathering people to learn from the same sources, discuss the knowledge face to face in small groups and then recommend and respond to the actions of elected representatives would create a much different set of laws than we have now.
This is just the start for me, I would love to see a four day work week with the fifth being devoted to self-governance and/or civic engagement. Most of our lives are too burdened with work and responsibilities to be able to participate in the active world of politics. As a culture, you can’t grant autonomy if you don’t also grant time. If there is one day a week that people can give to each other to discuss the direction of their common lives we can take control of them back. It would be time for neighbourhood associations, school councils, unions, city planning, sports leagues, crafting guilds. We could have citizens sitting on corporate boards, federal department advisory committees, focus groups. We are asked to give so much to the culture, but the quantity of what we can give is limited by the time we are allowed. It’s no surprise that politics is low on our list of things to care about.
Yes, tests can be manipulated for political advantage as US history shows. Birth certificates are not subject to such manipulation, which is why age criteria don't pose that danger. Of course, setting the appropriate voting age may be a politically contested question.
Great read. I agree wholeheartedly with your final conclusion, though I wasn’t sure that was where you were ultimately headed, which speaks to your even-handedness here.
> If widespread political ignorance threatens democracy, one logical response is to ensure that more knowledgeable citizens have greater influence over political decisions.
How does limiting voting rights to a more educated minority help ensure democracy, if democracy is most simply defined as a system of government by the whole population or all the eligible members of a state?
>On the other hand, they are quick to dismiss epistocratic proposals—such as limiting voting rights to the more informed—on the basis that we cannot accurately identify who is knowledgeable. This contrast is striking. If we can determine who is politically ignorant, why can’t we determine who is politically knowledgeable? Aren’t these just two sides of the same coin?
As you later point out, it seems your students, like me, are skeptical that a system sufficiently safe enough from corruption can be put in place for choosing which people within an ostensible democracy are allowed political agency. A predictable result of such epistocratic proposals is that the political situation will bend toward favoring the highly educated few who meet the voting requirements that they themselves put in place.
> If we accept that infants and young children lack the competence to vote responsibly, then we have already conceded that some threshold for political knowledge exists.
This is true, but we also wouldn’t tax, jail or enlist these people to fight on behalf of the rest of us. The more the state asks of a person, the more say that person should probably have in regard to what the state does.
Hi Woolery,
Thanks for reading the post and for your kind words about it. I think you’re right that restricting voting rights to a more educated minority wouldn’t strengthen democracy, at least not democracy as we know it. That said, Alex Guerrero defends a model called Lottocracy, where a smaller, randomly selected group of citizens is given the time and resources to make decisions on behalf of the population at large. Since he doesn’t see democracy as necessarily requiring rule by the whole population (above a certain age), there’s room to argue that universal suffrage isn’t a strict requirement of democracy. Nevertheless, I generally agree with you that epistocratic suffrage is hard to reconcile with democratic principles.
I also appreciated your point about the relationship between state obligations and political participation: “We also wouldn’t tax, jail, or enlist these people to fight on behalf of the rest of us. The more the state asks of a person, the more say that person should probably have in regard to what the state does.” I completely agree. There are strong reasons—beyond purely competence-based considerations—why certain groups should not be excluded from voting. That said, I do think the epistemic concerns about voter competence remain an important factor here.
>That said, Alex Guerrero defends a model called Lottocracy, where a smaller, randomly selected group of citizens is given the time and resources to make decisions on behalf of the population at large.
I think I’ve read about this, or a similar system, that was instituted in school governments with great success. It would be a wonderful thing to award power to those who don’t seek it at all costs, and it would certainly stifle the influence of entrenched interest groups and solve some other growing problems associated with traditional elections. There’s promise there, though I’m wary of the kind of upheaval that would be needed before such a sea change in governance could take place.
I’d be interested to read more of what you think about such a system, should you post on it.
I may write about lottocracy here soon. There's a chapter about it (and epistocracy) in the book I just finished writing with Elise Woodard, "Political Epistemology: An Introduction." In case you're curious, here's an interview I did with Alex Guerrero on his book: https://www.youtube.com/watch?v=RKPO5uB2YtQ
Beautifully argued, and I very much like your “bottom line.”
This essay thoughtfully explores the puzzle of why students recognize political ignorance as problematic yet resist epistocratic solutions. However, I'd like to challenge a fundamental assumption: that the right kind of political ignorance is actually widespread.
There's an important distinction between not knowing certain political facts (like the location of Ukraine or the names of elected officials) and not knowing political information that's relevant to one's interests and lived experience. This distinction relies on Gerd Gigerenzer's work on ecological rationality, which suggests that humans are remarkably good at developing knowledge and making decisions in domains that matter to their specific circumstances.
Gigerenzer and others have suggested that people often employ "fast and frugal heuristics" that are well-adapted to their particular environments. In "Simple Heuristics That Make Us Smart" (1999), they argue about how seemingly limited reasoning strategies can be highly effective when matched to the right environments. Similarly, in "Gut Feelings" (2007), Gigerenzer illustrates how intuitive judgments can outperform complex analysis in certain contexts.
Applied to political knowledge, this perspective suggests that voters may be ignorant of abstract political facts while being quite knowledgeable about issues that directly affect them. A farmer might not name every Supreme Court justice but may deeply understand agricultural policies. A parent might struggle with identifying foreign leaders but comprehend education policies intimately. Herbert Simon's concept of "bounded rationality" is relevant here too—people develop specialized knowledge within the constraints of their cognitive resources and personal stakes.
I wonder if the "ocean of data" on political ignorance adequately controls for this context-specificity. Do these studies test for knowledge that's actually relevant to voters' lives and interests? Or do they primarily measure abstract political knowledge that, while perhaps valuable for academic discourse, may be peripheral to how people actually make voting decisions? I will confess to my ignorance of the literature on whether it addresses the view I've given in this comment.
But even if there is a sea on this ocean of data that does address the view in this comment, empirically examining whether people possess interest-relevant political knowledge presents significant methodological challenges. It's not as straightforward as simply asking people to identify their interests and then testing their knowledge in those domains. People may misidentify their own interests due to limited self-awareness, social desirability bias, or ideological commitments. They might also strategically misreport interests to appear more informed or to align with perceived group norms.
Furthermore, what constitutes "relevant" knowledge may itself contested. A person's objective material interests might diverge from their perceived interests or their identity-based commitments. The factory worker who votes against labor protections because of deep-seated cultural values isn't necessarily ignorant—they may be prioritizing different aspects of their complex set of interests. This makes designing studies that accurately capture interest-relevant political knowledge extremely difficult, as researchers must navigate the thorny issue of defining what knowledge "should" matter to different individuals without imposing their own normative frameworks.
These methodological hurdles suggest we should approach claims about widespread political ignorance with caution. The standard political knowledge batteries used in most studies (at least those I'm aware of) may be measuring the wrong thing entirely if they fail to capture the situated, context-specific knowledge that actually guides voting decisions in practice.
If people are reasonably knowledgeable about political matters that affect their interests, then perhaps what's being labeled as "ignorance" is actually a rational allocation of cognitive resources. This doesn't mean we shouldn't strive for better civic education, but it does suggest we should be cautious about labeling voters as "incompetent" based on standardized knowledge tests that may not capture the situated knowledge that actually guides their political choices.
I close by noting that, again, I could be entirely ignorant of some relevant papers in the literature—if so, I am happy to have someone remedy my ignorance.
Hi Alex!
I definitely agree that standard political knowledge tests often miss the kind of knowledge that actually matters to voters and political competence. And heuristics can be useful! But I’m skeptical that they’re enough. One reason is that politics isn’t just about personal interests—it’s about navigating trade-offs across many domains. A farmer might know more about agriculture policy, but that won’t help them assess tax policy or judicial appointments, and the trade-offs matter too.
I think the bigger issue is whether voters’ heuristics reliably track political reality. Research on misinformation and partisan bias suggests they often don’t. People regularly misattribute policies to the wrong parties or support politicians who actively undermine their interests. So even if heuristics help, they don’t always lead to informed choices. Some of the heuristics are themselves the result of various biases. Perhaps I'm being too pessimistic about this, but I totally agree with you that a lot of what these political knowledge tests measure is just "trivia" that isn't importantly related to the kind of competence that matters at the polls.
How about candidates who are not ignorant? Dumb voters is one thing—and understandable; they (we) have jobs, family, other distractions. But candidates, they’re saying serving us (me) is all they will do. The voter trusts/believes that the candidate knows what they’re talking about and, as they take that oath, will faithfully execute. Yes, it is our duty to be informed, but, speaking of impacts and effects, it is the duty that much more of the candidate to know what is going on – and then not to lie to us about it all.
Would you share a link to your syllabus? I'd love to see it!
Here you go!
Topic 1. Truth and Politics
Hannon, Michael and Elise Woodard. Politics and Truth. In Political Epistemology: An Introduction
Arendt, Hannah. 1967. Truth and Politics. The New Yorker
Topic 2. The Epistemology of Democracy
Anderson, Elizabeth. 2006. The Epistemology of Democracy. Episteme
Landemore, Hélène and David Estlund. 2018. The Epistemic Value of Democratic Deliberation. In The Oxford Handbook of Deliberative Democracy
Topic 3. Democracy, Empathy, and Understanding
Hannon, Michael. 2020. Empathetic Understanding & Deliberative Democracy. Philosophy and Phenomenological Research
Cassam, Quassim. 2023. Sensemaking, Empathy, and Democracy. In The Epistemology of Democracy
Topic 4. Political Disagreement
Feldman, Richard. 2007. Reasonable Religious Disagreements. In Philosophers without Gods
de Ridder, Jeroen. 2021. Deep Disagreement and Political Polarization. In Political Epistemology
Topic 5. Badmouthing, Belief, and Partisanship
Hannon, Michael. 2021. Disagreement or Badmouthing? The Role of Expressive Discourse in Politics. In Political Epistemology
Williams, Dan. 2020. Socially Adaptive Belief. Mind & Language
Topic 6. Voter Ignorance
Somin, Ilya. 2021. Is Political Ignorance Rational? In The Routledge Handbook of Political Epistemology
Brennan, Jason. 2009. Polluting the Polls: When Citizens Should Not Vote. Australasian Journal of Philosophy
Topic 7. Irrationality and Bias
Brennan, Jason. 2016. Against Democracy, Chapters 1 & 2
Huemer, Michael. 2016. Why People Are Irrational About Politics. In Philosophy, Politics, and Economics: An Anthology
Topic 8. Democracy, Social Media, and Fake News
Rini, Regina. 2021. Weaponized Skepticism. In Political Epistemology
Rini, Regina. 2017. Fake News and Partisan Epistemology. Kennedy Institute of Ethics Journal
Topic 9. Utopias and Utopophobia
Estlund, David. 2014. Utopophobia. Philosophy and Public Affairs
Ancell, Aaron. 2019. Political Irrationality, Utopianism, and Democratic Theory. Politics, Philosophy, and Economics
Topic 10. Epistocracy
Brennan, Jason. 2016. The Rule of the Knowers. In Against Democracy
Harwood, Robin, 1998. More Votes for PhDs. Journal of Social Philosophy
I will confess to laughing out loud at seeing the last text on the list: “Harwood, Robin, 1998. More Votes for PhDs.” Though I recognize I am responding to the title, not the article itself (which I would like to get hold of in full and read—so far, I am encountering a paywall), what immediately came to mind was that, if I were the decided on this, some of these folks, let’s say Judith Butler as one example, should be top of the list to be PERMANENTLY disqualified from voting.😎 The discussion here is fascinating, and I look forward to what comes next, though I will remain skeptical of proposed solutions that select out voters as unqualified to vote, even in a single election. Anyone remember literacy tests in the US and who that excluded from voting? (I write this with affection for the project here, just also with the bemused recognition as to why it was that undergrad philosophy classes and I were not the best of friends!)
I really like the larger observations and explanations you offer here, especially the last piece about moral encroachment (and the "meta-encroachment" of corrupted incentives for determining any standard, great point!). Yet something about your initial premise, and the way you set up the dilemma, feels off. I'm not sure that the "asymmetry" you identify between political ignorance and political knowledge is such a mystery (let alone paradox), or even as asymmetric as you say. Is it possible that this puzzle you're wrestling with is mostly an artifact of comparing apples and oranges?
Let's say we have good reason to think (anecdotally, and having seen some polling) that there's a problem of widespread ignorance among voters, and this isn't just our own political bias. This will be based on very obvious cases of ignorance: people saying or doing "ignorant" things, offering dubious explanations, getting most test questions wrong. That doesn't mean we have the ability to discern where the *threshold* is for ignorance, any more than we can discern the threshold for sufficient (or minimal acceptable) political knowledge. It simply means we know egregious ignorance when we see it, the kind that on its face seems problematic for anyone voting. So when you say, your students can determine who is politically ignorant so "why can’t we determine who is politically knowledgeable": the appropriate standard of comparison is not the ability to determine who is knowledgeable, or assess people on that. We merely need to be able to recognize who is *not* grossly ignorant such that the idea of them voting isn't scary. In that sense it really is just two sides of the same coin, since in practice we don't assess ignorance vs. knowledge but ignorance vs. non-ignorance, and we don't assess this at the boundary but closer to the extremes. And this doesn't require any complicated policy or test; it's just common sense. (Of course there's still some inherent asymmetry, by virtue of the general principle that it's much easier to falsify something than to determine that it's true using the right standard - but that's hardly unique to politics, it's a fact of life).
None of this speaks to the separate question of how much political competence is a function of being well-informed as opposed to other things, or whether it's even possible to abstract "knowledge" from the capacity and disposition to use it well. Here, I agree with Susan Scheid above that factors like trust and motivation matter much more than knowing the three branches of government, in the scheme of things. Nevertheless, it *is* concerning to see evidence of mass voter ignorance. I think this is not because that particular knowledge is necessarily important, but because widespread ignorance can be a soft sign that other things are amiss, which might track more closely with actual competence.
Hi Chris,
Thanks for this response. I agree that the general concern abou ignorance vs. the challenge of identifying a specific threshold for political competence are different questions, but as you said, “we know egregious ignorance when we see it, the kind that on its face seems problematic for anyone voting.” So, why not use this to set the minimal boundary for competence? If we can confidently identify cases of extreme ignorance, it seems like we could at least exclude those cases rather than requiring a precise threshold at the boundary to adjudicate hard cases.
You also mention that “the appropriate standard of comparison is not the ability to determine who is knowledgeable, or assess people on that. We merely need to be able to recognize who is not grossly ignorant such that the idea of them voting isn’t scary.” I completely agree—and I tried to make this point in the post, particularly in the section on whether ignorance is easier to spot than knowledge. My argument wasn’t that we need to identify the most politically competent, but rather that we could use clear cases of ignorance to define a lower bound for participation.
Thanks for responding and sorry if I ended up echoing the same points you were already making - we are probably on the same page about most of this. Perhaps I wasn't clear. I was trying to leave aside those thornier questions of policy and determining official standards to focus on the basic tension you outline at the beginning, which to me felt illusory in that it wasn't an apple-to-apples comparison: your students' resistance to epistocracy was based on the severe standard of policy (moral enchroachment, construct validity and the whole nine yards), whereas their confidence to affirm widespread ignorance was based on the much lower standard of informal secondhand evaluation of stuff they see in the media and encounter on the ground (not to mention, preexisting bias). Clearly you understand all this. But the way you were setting it up as a mystery still felt misleading, so it seemed important to push back a little.
With respect to policy: I guess my real issue is not that we can't make global assessments of pronounced ignorance or decide on some minimal standard for stuff people should know. It's rather that "knowledge" (at least, the kind we can test) is hardly the same thing as political *competence*. If anything, that hoary term "critical thinking" - sound judgment, solid reasoning, paying the right kind of attention to the right things - is much closer to the kind of epistemic resources I associate with competence. And even that shouldn't be a blunt screening tool! (I don't think voting age requirements is a good test case; that's applied a priori based on demographics, not performance and assessment).
If each person, including the test authors, is ignorant about a somewhat different range of issues from the areas of ignorance of everyone else, then these tests would seem necessarily to isolate only a subset of these areas, and in so doing could introduce or reinforce blind spots in electoral choices rather than allowing different blind spots as good a chance to cancel each other out by incorporating feedback from more people.
I agree that diversity of perspectives can help eliminate blind spots in decision-making. If ignorance were randomly distributed across the electorate, then aggregating a wide range of viewpoints could help correct for individual gaps in knowledge.
But I think this optimistic picture leaves out an important complication: ignorance can also compound ignorance. Not all blind spots are independent or randomly distributed—some forms of misinformation or misunderstanding tend to cluster within particular social, ideological, or informational environments. In these cases, simply adding more perspectives to the mix doesn’t necessarily correct errors; instead, it can amplify them. If large numbers of people share the same fundamental misconceptions about political reality, incorporating more input from the uninformed could lead to worse, not better, decision-making.
I wonder if a threshold question here might be whether we are even asking the right question. That is, do voters need to be “informed about politics” in order to be competent as voters? While, as most of us likely do, I have taken this as a given without much thought, I now find I am not so sure. First, what does it mean to be “informed about politics”? For example, I know a number of highly educated, politically attuned people who have no idea who their city or state elected officials are. I have been in that situation myself. One tends not to know these things if the people in question don’t seem to offer anything of value to voters. (I would say my state senator is very much like that. I do know her name, but I also experience her as someone who is warming a seat with little to no interest in addressing her constituents’ needs.)
I don’t have a fully formulated thought here (to say the least), but I do wonder whether, in evaluating voter capacity to make good judgments in selecting their representatives, the first place we may need to look is at the severe erosion of trust in our governmental and other institutions on which we have ordinarily relied, at least where I am in the US. We all, necessarily, must look to trusted interlocutors to help us assess what the best approaches are to address the multitude of complex issues we face individually and as a society. If we lose trust to such an extent that we don’t feel we can rely on much of anyone who represents us, is in a position of authority, and/or has certain expertise, then how can it be otherwise than that we will be handicapped in making the political judgments we need to make? No amount of civic education, as one example, can counter that. For one, we would have to trust those doing the civic education.
I think what people do know is what they need from their government, and they have good judgment about whether they are getting what they need or not. If not getting what they need persists over a long period, and also if the society in which they live is under continuous stress, then it is hardly surprising for frustration to boil over. In the US, we have had three back to back spiritual and material crises of enormous scope: 9/11, when we learned the world was much smaller, and we were much more vulnerable than we had thought; the 2008 financial crisis, which ruined the lives of many families and hollowed out whole communities, some of which never recovered; and the pandemic, the shocks and aftershocks of which I think we vastly underestimate and continue to experience today.
In the case of the last two, particularly, the people negatively affected were often in the same demographic and blighted region of the country. Were they helped by their government? Not nearly enough. It is not in the least surprising to me that, after years of this, the frustration boils over and, as a workman once said to me who felt that frustration, “we should just blow the whole thing up.” Civic education will not help with that. Helping lift people up who need our help is all that will. And, just as, once things start spiraling downward, it is hard to stop that, so too, if people do get help, and that continues, it should be possible to rebuild trust. At least I hope so!
Hi Susan,
Thanks for this really insightful response. I largely agree with everything you say, and I think there are strong criticisms of the very idea that social scientists can adequately measure the kind of knowledge needed to be a competent voter. The erosion of trust in institutions is also a hugely significant factor—if people don’t trust the sources of political information, or the system itself, then even the most well-designed civic education efforts won’t address the deeper issue.
In the post, though, I was hoping to largely set those broader concerns aside and focus on a what seemed like a puzzle: my students clearly do think these political knowledge tests are measuring something important, otherwise they wouldn’t be so concerned about the kind of ignorance these tests reveal. Their attitude is what I find interesting—on one hand, they take the data on political ignorance seriously, but on the other, they resist the idea that we can determine who is sufficiently knowledgeable to vote responsibly. I was trying to make sense of that tension, rather than defending the legitimacy of political knowledge tests in a more fundamental way. But I recognize that this is largely bracketing some of the more interedeting and foundational issues you mentioned in your comment!
You wrote “The issue isn’t that ignorance is easier to identify than knowledge; rather, the consequences of error are far more severe when deciding who is knowledgeable enough to vote responsibly”. I’m not sure about this. It depends on the method of letting a person vote, right?
Maybe the prospective voter has to pass a five-minute quiz at the voting booth, a quiz on all the elections she can vote on *at that moment*. If she passes, she gets to vote. If she doesn’t, then she doesn’t get to vote. But failure doesn’t mean she can’t vote in any other election. So, the consequences of error are pretty limited: she doesn’t get to vote *that day alone*. I wouldn’t call that “severe”. No?
The harder part is writing the questions so that people agree this is a good idea.
Hi Bryan,
Thanks for this—you raise an interesting challenge. You’re right that the severity of the consequences depends on how voter competence is assessed and enforced. If failing a knowledge test only disqualified someone from voting that day—without barring them from future elections—the stakes would indeed be lower than in systems where incompetence results in long-term disenfranchisement.
That said, I still think a version of the concern holds, even in a system like the one you describe. For one, if the quiz is structured in a way that systematically disadvantages certain groups (due to wording, assumptions about relevant knowledge, or implicit biases in question selection), the aggregate impact could still be significant, even if individual failures only apply to a single election. Second, perhaps even a small barrier to voting could create larger effects—e.g., people might avoid showing up at all if they fear public embarrassment or failure, especially in communities with historical voting restrictions.
I also completely agree with your last point—the hardest part isn’t necessarily the idea of a test itself, but designing one that is widely accepted as fair and politically neutral. That, I suspect, is where things get much more difficult.
I was thinking that you take the online quiz in the voting booth, and no one but you knows whether you passed it. If you failed it, you can still leave the booth and lie about it. So, no stigma. I thought the quiz would be multiple-choice, you would have something like three chances to pass it, you'd get ten questions each time, chosen randomly from a database, and so on.
As for systematic disadvantages, that's a problem for the design of the quiz. I really have no idea what the questions should be. Maybe some general ones, like "The US president is head of which branch of government?" and "Which branch of government passes the nation's laws?". But there needs to be ones specific to the candidates I suppose, and I don't know what they would be at all. I think it might be easy for politicians to game the system, by lying about their policies and proposals.
I know that I'd *like* to stop people from voting who thought that Trump really would lower all grocery prices on day 1, like he claimed. That's idiotic. Or people who thought Biden was part of a federal pedo ring involved in sex trafficking. I don't know how to handle that at all.
Then there are the problems about down ballot things, like local races for justices. Virtually no voters know anything about any of those candidates.