Reviewing Caplan’s Reviewers

Economist Bryan Caplan’s Myth of the Rational Voter argues that humans are less rational as voters than as consumers, causing us to support government policies that pursue illusions instead of giving us what we really want.   Since educated folks seem to be more rational (because they tend to agree more with experts), and yet do not vote on substantially differing values, Caplan suggests we weigh educated votes more, or at least stop trying to get out the vote, and that we reduce the scope of government policy. 

Caplan’s book has attracted some high profile reviews, such as in the New Yorker, New York Times, and Economist.  Yet while reviewers seem to grant Caplan’s basic claims, they offer only these lame counter-arguments:

  1. There is no right answer on many big questions, so voters can’t be wrong.
  2. Things would be better if political elites gave voters more guidance.
  3. Experts and the educated sometimes make mistakes or disagree.
  4. Democracy as we have it now achieves important "non-economic" values.
  5. Caplan is wrong: the more educated do often vote on different values. 
  6. Caplan is wrong: the more educated are not in fact right more often.
  7. A narrower franchise would be less legitimate and hence less stable. 

(Lest you think I oversimplify, extensive review quotes are below the fold.) 
The first four arguments are beside the point.  If there are no wrong answers then Caplan’s changes would also not give wrong answers, and the possibility of other improvements doesn’t suggest Caplan’s proposals wouldn’t help.  The possibility of mistakes is also irrelevant; the question is who makes more mistakes.  Also, Caplan’s analysis has nothing to do with "economic" versus "non-economic" values.  He only talks about economics because he knows that best; Caplan claims that democracy is not well-structured to achieve any voter values that differ from voter illusions. 

The last three arguments at least address Caplan’s claims, but reviewers use amazingly weak standards of evidence.  While Caplan offers and cites extensive evidence that the more educated are more accurate but use similar values, reviewers seem content to mention particular cases of errors or divergent values, or to suggest that such cases may appear someday.   

Finally, reviewers offer no evidence suggesting Caplan’s proposals would bring rioting in the streets.  We already tolerate a lot of inequality in political power.  Convicts, children, dogs, foreigners, and those who don’t register cannot vote now, half of those who can vote do not vote, in part due to inconvenient voting times and locations, those with more political knowledge can use their votes more effectively, tricks like gerrymandering and the electoral college give us varying voting powers, and yet we see not a hint of related rebellion.  So why should Caplan’s moderate changes suddenly produce armed rebellion?

I can imagine counter-arguments with more teeth (and will suggest one Thursday), but it is noteworthy that reviewers are not offering them.

P.S.  DC area folks can see Caplan speak on July 17.

(Here are all those quotes I promised.)

New York Times Magazine:

Caplan argues that "voters are worse than ignorant; they are, in a word, irrational – and vote accordingly." …
Caplan has some radical medicine in mind. To encourage greater economic literacy, he suggests tests of voter competence, or "giving extra votes to individuals or groups with greater economic literacy." … Caplan dislikes efforts to increase voter turnout. … If the Supreme Court can strike down laws as unconstitutional, why shouldn’t the Council of Economic Advisers be able to strike down laws as "uneconomical"?

The liberal blogger Ezra Klein wrote: "Obviously I, like most coastal-bred elitists, don’t think voters make terribly good decisions. But I also don’t think economic actors are particularly rational." He might have added that many policy issues cannot be decided on the basis of avowedly rational expert judgment alone. Take immigration, where governments weigh not just economic costs and benefits but also demands of national identity and cosmopolitanism. Or war: it’s very complicated, so should we abandon military planning to the professional generals? …

Caplan’s view of democracy is all about efficiency, not legitimacy. But … A democratic public may not always like – or understand – the government’s policy, but the consent of the governed gives the citizens a reason not to reject the whole system.  Caplan … may underplay the role of political elites in shaping that judgment. Would the public choose badly if it had better guidance? … Maybe the public doesn’t measure up because the politicians are not doing their job properly, not the other way around.

New Yorker:

Caplan is the sort of economist … who engages with the views of non-economists in the way a bulldozer would engage with a picket fence if a bulldozer could express glee. …  Caplan … offers some suggestions for fixing the evils of universal democratic participation (though he does not spend much time elaborating on them, for reasons that may suggest themselves to you when you read them): require voters to pass a test for economic competence; give extra votes to people with greater economic literacy; reduce or eliminate efforts to increase voter turnout; require more economics courses in school. ….

The problem [is] … not a matter of information, or the lack of it; it’s a matter of psychology. Most people do not think politically, and they do not think like economists, either. People exaggerate the risk of loss; they like the status quo and tend to regard it as a norm; they overreact to sensational but unrepresentative information (the shark-attack phenomenon); they will pay extravagantly to punish cheaters, even when there is no benefit to themselves; and they often rank fairness and reciprocity ahead of self-interest. …

Most people, even if you explained to them what the economically rational choice was, would be reluctant to make it, because they value other things – in particular, they want to protect themselves from the downside of change. They would rather feel good about themselves than maximize (even legitimately) their profit, and they would rather not have more of something than run the risk, even if the risk is small by actuarial standards, of having significantly less. …

If all policy decisions were straightforward economic calculations, it might be simpler and better for everyone if only people who had a grasp of economics participated in the political process. But many policy decisions don’t have an optimal answer. They involve values that are deeply contested: when life begins, whether liberty is more important than equality, how racial integration is best achieved (and what would count as genuine integration).

In the end, the group that loses these contests must abide by the outcome, must regard the wishes of the majority as legitimate. The only way it can be expected to do so is if it has been made to feel that it had a voice in the process, even if that voice is, in practical terms, symbolic. A great virtue of democratic polities is stability. The toleration of silly opinions is (to speak like an economist) a small price to pay for it.

In These Times:

The first and most obvious problem with Caplan’s argument is that it quickly leads to some very dark places. … You could instead give more votes to businessmen and university graduates, as Caplan comes close to proposing, or simply require people to "pass a test of economic literacy to vote."

Which brings us to the second problem: what constitutes economic consensus. Caplan spends considerable time attempting to persuade the reader that if experts and the general public disagree, the experts are right and the public wrong. That may often be the case, but it’s not a static proposition: What experts believe evolves over time, and the same is true of the public. In 1996, the public thought taxes were too high, but recent polling suggests that’s no longer the case. … Caplan’s book wouldn’t have made much sense 40 years ago, which prompts the question: Will it make much sense in the future? Caplan thinks he’s describing the fundamentals about human nature, but he might just be elaborating on the contingencies of an era.

What’s more, sometimes the public is right and the experts are wrong. Economic experts used to believe in price controls. …  Finally, Caplan over-interprets the degree of economic consensus. … when it comes to policies the disagreement [among economists] is tremendous. Caplan thinks the minimum wage borders on quackery, but last year more than 500 economists, including a half-dozen Nobel laureates, signed a petition in favor of raising it.

Caplan wants to grant a presumptive authority to the consensus view of economists, but the consensus view of economists is that voters are rational, which is, of course, precisely the position he wants to convince us is wrong.

CATO Unbound:

Average citizens are not sober-minded assessors of evidence bearing on policy determinations. As Caplan observes, they vent their prejudices. … Citizen passions are, then, incentives to pandering. But they also are a gateway to accountability. Voters have precious little control over any particular political platform. What they can do, though, is "throw the rascals out." … This is far from an ideal way to exercise governance: rule by philosopher-kings it’s not. For Caplan, as for Plato, that is a damning indictment of democracy. I’m not so sure. Give me an airtight guarantee that those advertised as the Best and the Brightest are the genuine article and that, in addition, they are indelibly committed to serving the public good, and I’ll sign up with the Plato/Caplan antidemocrats. … Caplan is partial to the policy judgments of economists. Because he himself is one, he should know better.  [Loren Lomasky]

If we know which survey questions measure voter competence, then we could presumably devise an exam to determine how many votes each person will be allowed. … On the factual and technical parts of the test economists and others with university degrees would do better than many others. But would they do better on the whole test? … How to slow inflation, or to mobilize people for war, or to raise taxes for social programs is one sort of thing; when these should be done is another. … Consider the demographics of the economics profession, and of college education in America in, say, the decades just before 1949 … the need for racial justice was urgent, and the idea that an even smaller, whiter, more male, more Protestant electorate would have performed better on the morally most important matters is far from obvious. … the demographics of expertise do not necessarily mirror the population in all other respects while simply adding on the benefits of factual and technical expertise. … there is no entitlement to rule others based simply on the fact that you know what is best. Democracy, on the other hand, can effectively bring intelligence to bear on public problems in a way that avoids these controversial invidious comparisons. That, I think, is its claim to authority. … Voters and market actors are the same people, so we should expect the charges of ignorance and irrationality to be leveled against people in both guises. [David Estlund]

Caplan is right to think that what the public lacks is expertise in economics. He is wrong to think that expertise should be conflated with educational credentials, even for heuristic purposes; or, as a corollary, that public wisdom should be equated to public agreement with credentialed "experts." If Caplan were in a discipline (such as any of the other social sciences) in which he did not happen to agree with his colleagues, or if he found himself writing about the views of economists trained in the Soviet Union or the Islamic Republic of Iran, he would not so readily overlook the fact that when an educational institution credentials an "expert," it is merely certifying that this person has been indoctrinated with whatever is thought by the faculty of the institution to be the truth – which has no necessary correlation with what is actually the truth. … there is a perverse tradeoff between ignorance and ideology: less of the first leads to more of the second. So if I had to choose, I’d pick rule by the masses over rule by elites. That’s not because I entertain illusions about "the wisdom of crowds." It’s because I’d rather be ruled by open-minded ignoramuses than by doctrinaire ideologues. [Jeffrey Friedman]

GD Star Rating
Tagged as:
Trackback URL:
  • Hopefully Anonymous

    First, really great post. Maybe the best post I’ve read on this blog yet.
    Second, I’ll be buying Caplan’s book, thanks to your review of it here.
    Third, I’m partial both to your critique of the criticisms Caplan’s theories have received, and to his prescriptions for a more rational voting/decision-making process. For example, the idea of a political literacy test (presumably keyed to determine who has a threshhold level of rationality behind their decision-making process) as a prerequisite for voting.

    It would be great if you, he or, others expanded this system critique to the judicial process, particularly our arcane, arguably irrational use of juries as finders of fact instead of the use of experts like most developed nations. It seems to me all the checks against bias (adversarial lawyers, appeals, open process) could be maintained, without turning random members of a community into the people with the final word on ballistics, dna, and probability theory in determining guilt or liability.

  • conchis

    Dani Rodrik presents arguments and evidence for 5 and 6 in the case of trade, and is one review that I haven’t seen Bryan respond to (although I asked his opinion on it in comments at econlog).

    I also have qualms about your dismissal of (a slightly modified version of) 4. Bryan’s argument does have to do with outcome values as opposed to procedural values. There seems to be ample evidence that people value (the illusion of?) control, and not simply for instrumental reasons (as an example see e.g. Frey and Stutzer on the effect of greater direct democracy on subjective well-being, but there’s a lot in the psych literature too). Now, you might think that outcome values should trump or outweigh such procedural values, but (a) I don’t think you can simply dismiss such concerns; and (b) why should you or I get to decide whether outcome values trump procedural values?

    Relatedly, on 7, I think some people are simply pushing the argument that “a narrower franchise would be less legitimate”, which they view as a bad thing independently of any actual effect on stability (though I personally doubt that the stability argument is as weak as you suggest).

  • Robin Hanson

    Conchis, Rodrik just says that he disagrees with economists, and hence the educated, about free trade, and that he sees some pocketbook voting on trade. Caplan admits experts/educated can make mistakes, and that there is some rare pocketbook voting. Again the issue is the overall trends, not particular cases. Also, Bryan argues that democracy should choose illusory procedural values over real ones, just as it should choose illusory outcome values over real ones.

  • Plinius

    Could it be that centuries of power being concentrated in the hands of the elites have spoiled the masses to think that opinions don’t matter and hence one can afford all kinds of wacky theories? If so, an argument pro-democracy would then be that the masses are learning and the more direct power you gave them the more they can learn from their mistakes. On the other hand, if you insulate them from power they will never learn.

  • Hopefully Anonymous

    Plinius, I don’t think Caplan is advocating rule of elite. More rule of the threshold competent. I’m actually more partial to rule by expert with checks against abuse of power than rule of threshold competent, but I think rule by threshold competent would be a vast improvement on the state of the union.

  • Andrew

    In foreign policy, there is some evidence that policy leaders favor military-oriented policies that give them more freedom of action, while Americans, on average, are more positively disposed toward peaceful policies and international cooperation. For example, from a paper by Benjamin Page, “large majorities favor multilateral uses of U.S. troops for peacekeeping and humanitarian purposes, but majorities oppose most major unilateral engagements.” See here:

    One “overcoming bias” reason for the policymakers’ preferences has to do with who is actually implementing the policy. Experts may be biased in favor of options that given U.S. policymakers more power because they can imagine themselves, or people like themselves, implementing these policies.

  • Zenkat

    Caplan thinks we should “weigh educated votes more”?

    We’ve tried this before. Literacy tests were widely required throughout the south until the 60’s. They disenfranchised millions of African Americans and ensured white minority rule for decades.

    Interestingly enough, this period (ie, the antebellum south) is not widely considered a paragon of “rational decisionmaking”, enlightened governance, or economic vitality.

    While requiring “voter competency” may theoretically lead to more rational decisions, in reality it will necessarily bias political decisionmaking to favor the interests of groups that are, for whatever reason, overrepresented by the competency test. Even worse, the temptation to game the rules of the system to lock in political power and further disenfranchise the outgroups would likely prove irresistable.

  • Zenkat

    Whoops. That should be “postbellum”, not “antebellum”. Pardon, it’s early on this coast.

  • conchis


    Obviously you’re right that the issue is general trends. Rodrik’s argument is merely a data point to the contrary that didn’t strike me as being based on “amazingly weak standards of evidence”, and which highlighted the more general issue of how we assess whether experts are more often right in a way that isn’t merely tautological.

    I think you’re stretching more on the second point. It’s not really an issue of voters choosing illusory procedural values; it’s one of whether the process of allowing everyone’s votes to count equally embodies particular procedural values. And a lot of experts and educated people seem to be claiming that it does.

    Even if it were an issue of voters choosing values, where’s the evidence that voters favour illusory procedural values over real ones? To be honest I’m not even sure what illusory values would be, or how you could establish that someone was choosing the wrong values (as opposed to the wrong means by which to realize those values).

  • Robin Hanson

    Plinius, maybe the educated would also learn faster if we gave them more power.

    Andrew, you may have identified another example where the educated would vote on different preferences. But again the question is how common such examples are on average.

    Conchis, my “amazingly weak” phrase referred to the idea that pointing to a single contrary example was enough to reject Caplan’s thesis. And of course particular democratic procedures reflect particular procedural values; the question again is whether those are our actual values, or the illusory values resulting from irrationality.

  • Hopefully Anonymous

    I think your comparison to voting literacy tests in the antebellum south is important and should be instructive shouldn’t make consideration of threshhold competency tests unconsiderable. The salient criticism of voter literacy tests in the early 20th century South, in my opinion, is that the tests were selectively applied and not narrowly (or even competently keyed) to testing information a voter should know prior to being able to make a rational decision in the voting booth.

    As for whether threshhold competency requirements for voting, whether it will irresistably lead to further advantage of special interest groups and disenfranchisement of outgroups (I take it in a way that’s bad for us?) is an empirical question, there’s no reason for us to proclaim it a foregone conclusion one way or another. Ideally it’s the sort of thing that states could engage in experimentation in. and we can wait until we get some clear data before implementing this type of thing on the national level.

    I doubt threshhold competency requirements for voting will result in policies more harmful to disenfranchised population -I suspect it will result in more rational policies that will cause them greater benefit- but I’m willing to cede this is an empirical question, rather than propagandize for my personal inclination on the topic.

  • Hopefully Anonymous

    Whoops, made the same mistake as Zenkat. Early 20th century south, not antebellum.

  • Michael Sullivan

    Indeed. Zenkat’s argument is directly on point. Any proponent of voter competency tests must be prepared to explain how we will avoid test bias that is worse than the bias it is trying to overcome, using early 20th century history as a guide for what can (and will) happen.

  • Hopefully Anonymous

    One useful aspect of threshhold competency voting that would lead to the enfranchisement of a class of people is that we could abandon minimum age requirements for voting.

  • Hopefully Anonymous

    An important task, but I think it can be done. To be pragmatic, I think policy advocates should be prepared to accept the following right out the gate:
    (1) blanket exemptions for populations in voting rights districts
    (2) blanket exemption for people who are whatever age makes you eligible for the AARP
    (3) A “knowledge of minorities and history of oppression” component to a threshhold competency test -specifically covering knowledge of the following minorities and their oppression, in this heirarchical (and advocacy power) order: (1) women, (2) african americans, (3) latinos.

    To sell voter competency tests, I think women and women’s rights groups should be heavily targeted, since it will probably lead to even greater relative enfranchisement for women. The test could be structured in such a way that at least women won’t do worse on it (such as more heavily weighted in reading comprehension).

    I think if we do all this, we’ll still manage to exclude a huge number of demonstrably incompetent (and to Caplan’s point, systematically incompetent) voters and end up with more rational national policy. I think this type of approach can also neutralize a significant amount of criticism.

  • Eliezer Yudkowsky

    I first confess that I haven’t read Caplan’s book, which potentially invalidates all comments below.


    Does Caplan take into account Tetlock’s distinction between real and fake expertise? If you measure correctness by agreement with experts, then of course educated people will often be more “correct”. But when there is no higher criterion by which the experts themselves are measured, the expertise often turns out to be fake.

    Economics contains real truths but it is incomplete; it cannot carry through to the complete analysis of complicated real-life systems. Caplan himself, in proposing his governmental changes, would try to fix one particular problem with an ad-hoc patch that would have a vast number of other effects beyond the pleasant one he envisions. What are the psychological effects of being told that you can’t vote because you didn’t pass a threshold test for economic literacy? What happens when every interest group has an enormous incentive to game the test? What happens when the Voters start to think of themselves as an in-group distinct from the hoi polloi outgroup? According to Tetlock, no talking-head economist would do better at predicting these effects than a non-economist – though I suspect that an experimental social psychologist or political historian could take a stab at it.

    Does Caplan even realize how many effects his proposed patch would have? How little of it economics could even claim to predict? So many times throughout history (long before Plato’s Republic) the “experts” in Field #823 have proposed that political power should be concentrated into their hands. For obvious reasons of evolutionary psychology, everyone thinks they should be in charge. Perhaps we should put only evolutionary psychologists in charge… It’s always the same old trap. This is wisdom?

    Caplan should have stuck to identifying problems – it seems likely enough those are real – and not proposed solutions, or at least, not proposed huge systemic political changes. The problem with putting experts in field #823 in charge is that they won’t realize how much they don’t know.

  • Matthew C

    Great comment Eliezer, you’ve nicely summarized my own concerns with Caplan’s proposal.

  • Michael Sullivan

    There is a central problem with competency examinations by experts, and that is framing. Often a lay person appears “irrational” (as opposed to merely ignorant) to the expert, only because they do not understand the relevant framing expected by the expert. I’ve noticed this in discussion between me and my wife, where we each have a strong incentive to get to the bottom of disagreement and not write each other off. Where we each are experts, we often assume a bunch of framing and basic assumptions that the other does not. Judging the layperson from the expert’s frame of reference, we may decide they are “incompetent” but if we can figure out exactly what those frames are and explain them, it’s not uncommon for the layperson to then have interesting information which undercuts the framing for some particular application.

    The problem with designing tests is that the easiest tests to design merely test for someone’s ability to parrot the frame. It’s very much harder to test for the ability to make logical decisions within the frame, and far harder still to test for the ability to see circumstances which partially or completely invalidate the frame.

    I trust a person with rounded education and a lot of general skill at the latter two processes who has little or no education about a particular discipline, more than I trust some random graduate in the discipline. I trust the random graduate more on knowing facts in the field of course, but not necessarily as much on determining expected consequences or designing policy.

  • conchis

    “the question again is whether those are our actual values, or the illusory values resulting from irrationality.”

    OK. And I’m arguing that (a) I’m not even sure the idea of “illusory values” even makes sense (we value what we value, don’t we?), and (b) even if it does, as far as I can tell, neither you nor Bryan has done anything to suggest that these values are actually illusory.

  • tc

    Robin says: “So why should Caplan’s moderate changes suddenly produce armed rebellion?” I would hardly call a change that explicitly ends the principle of one man, one vote moderate. Anyways, we just had an example of how things would go differently: the recent immigration bill was supported by many of the elite but was ultimately defeated by a groundswell of populist reaction. I know that Caplan’s book would call this “anti-foreign bias”, but it seems to me this is an true instance of a difference in actual values.

  • Robin Hanson

    Eliezer, every change to our political system, including the changes that have widened the franchise, are subject to exactly the same criticisms you offer of Caplan’s proposed changes. To the extend one is concerned about unintended effects, one might prefer to start with small changes in a few locations and them monitor to watch for unintended effects.

    Tc, “one man one vote” is now only a slogan, with only a rough correspondence to actual practice. We could start with small changes which wouldn’t much change that rough correspondence.

  • Roger Rainey

    Wow, there are some truly scary, elitist and even fascist views expressed in this comment section. If this is representative of progressive thinking today, then we should all be worried. Why aren’t you focussing on educating common voters instead of disenfranchising them? I have not read Caplan’s book, but it seems to me the irrational voters take positions that are consistently trumpeted by our mass media – the economy is bad, free trade is bad, immigration is bad, price controls are good, etc. These are all positions that the 6 pm news and most newspapers reflexively and consistently stake out. More educated and interested voters are able to balance this with more reliable information, but we can’t expect everyone to do this. If anything is failing in our society, it is our mass media and its inability to present accurate information.

  • Hopefully Anonymous

    Roger, I think focusing on “educating common voters” is worthy of attention, in addition to disenfranchising incompetent/uninformed voters. It’s possible to consider common voters and incompetent/uninformed voters to be two different populations. “Common” implies majoritarian elements to a population, so I think you’d agree that the 25% least competent people that actually vote aren’t representative of common voters? If so, a threshhold competency voting test could reserve voting rights for (1) everyone who passes an objective, expert-set threshhold for informed voting competency, OR (2) the top 75% performers on that test, whichever population size is LARGER.

    I think that would significantly address your concern that “common voters” would be disenfranchised.

  • Plinius


    point taken. This somehow reminds me of the famous debate between Luther and Erasmus whether it was safe to let the common people handle the bible or was it better to leave that task to an educated minority. Luther seems to have won the debate, but I’m not a good enough historian to determine if this has been good or not.

  • Paul Gowder

    Why would our procedural values be illusory and due to irrationality? Here’s the other story: I have an overriding interest in having a (fair, equal) say in the conduct of my own society, an interest that outweighs my interest in having slightly better policy come out.

    That’s why I think we’d object to a wise and benevolent dictator, even if we assumed that the dictator was so wise that he was more likely to reach a correct decision than an arbitrarily large number of experts. Objection to dictatorship of the smart, extra votes for the smart, etc., follows by induction.

    Also, see Josiah Ober. 2007. “Natural Capacities and Democracy as a Good-in-Itself.” Philosophical Studies 132:59.

  • Robin Hanson

    Those of you who are strongly against any narrowing of the franchise, are you equally strongly in favor of widening the franchise? Shouldn’t we lower the voting age, and let convicts and foreigners vote, even if we give them a lower voting weight? Why would the status quo be so optimal?

    Roger, again, the fact that there might be something else we could do to improve things does not argue against Caplan’s proposed improvements.

  • Paul Gowder

    Robin: the status quo is a pretty good compromise between instrumental values (not letting kids vote because they don’t know anything, are subject to undue influence, etc.) and intrinsic ones. That being said, of the three categories you list, I’d support widening the franchise to include convicts. Convicts have a legitimate interest in the conduct of government and some kind of autonomy that we ought to respect. Also, felony disenfranchisement, especially of people who have already served their sentence, can create major distortions in the political system, especially given disparate patterns of criminal enforcement against (e.g.) young black men.

    (Foreigners aren’t members of the political community — although there are conceivable arguments for letting them vote too, they’re less compelling.)

  • Hopefully Anonymous

    Paul Gowder, we seem to have a neat opposite symmetry on this topic. I like the idea of kids of any age being able to vote that demonstrate threshhold competency. I think you’re performing a rather significant status quo bias, and your puportive reasons for why the status quo is either great or just don’t seem that coherent to me. For example, you’re concerned that disenfranchising convicts disparately impacts young black men -but disenfranchising kids also disparately impacts young black men, because black males form a significantly higher percentage of the under 18 population than they do of the over 18 population, due to both birth and death rates.

    I like the idea of enfranchising both felons and people under age 18, provided they pass the same voter competency/literacy test as the rest of the population.

  • savagehenry

    Paul, I think the point being made by Caplan and others is that in terms of knowledge of how the political process works, knowledge of basic economic principles, and the ability to understand policy implications many adults are just as ignorant and subject to undue influence as minors. Also, minors have just as much interest in the conduct of government as convicts, so excluding them just because they are young with no evidence any one individual minor is incompetent seems to me like a bias against younger people. Disclaimer: I’m only 21 myself so I am probably likely to be more biased towards siding with young people on that particular aspect of the issue.

    I think Zenkat’s concern is a legitimate one, but I think that if what Caplan is suggesting was implemented on a local (city or county) scale and then evaluated over a period of time before being implemented on a state or national scale many of those concerns could be eliminated because you could work out any procedural biases that would disenfranchise a minority such as blacks but would still disenfranchise people who don’t even know what it is the Senate actually does.

    The main problem I see with focusing on voter education is it has been my experience that your average person, even people inclined to be more politically aware/active, don’t want to be educated. They simply don’t have the time to devote to getting familiar with current issues.

  • Eliezer Yudkowsky

    Eliezer, every change to our political system, including the changes that have widened the franchise, are subject to exactly the same criticisms you offer of Caplan’s proposed changes.

    First of all, historically speaking, the US was an anomalous success case of political changes – consider Revolutionary France as a contrast – and then the US was imitated. So to cite the example of the US alone, as if all sweeping changes had been successful, is hindsight bias. Furthermore note that political change in the US was far more incremental than in Revolutionary France – the states federated first, then introduced weak centralization, then strong centralization. Big fast political changes are scary. You’re living in the one country that happened to survive its big ideas. Lots of other ideas that sounded just as reasonable, at the time, as the women’s vote sounded at the time, didn’t work out so well for the countries that implemented them, especially the ones that implemented quickly. You could have made a damned good case, historically speaking, for starting out by giving women voting rights in one state at a time – no matter how bizarre this sounds to us, today, when we grew up already knowing that it worked out excellently.

    “Big fast and unknown” was only part of my problem with Caplan. A much greater sin than proposing something huge is proposing something without being aware of how huge it is, thinking as if actions have only their intended effects and no others. In fact, I’d bet that’s one of the chief sins of which Caplan accuses governmental folks – but you get it just as strongly from disciplinary myopia. Why test folks on economics rather than social psychology, evolutionary psychology, Bayesian reasoning, statistics, general scientific literacy, historical knowledge, historical knowledge of the late Roman Empire…

  • Robin Hanson

    Eliezer, I don’t understand why you don’t think we could not start out with small incremental versions of Caplan’s proposals to “weigh educated votes more, or at least stop trying to get out the vote, and that we reduce the scope of government policy.” Which direction to explore policy changes is a different issue from how big and fast to move. Unless you think we should never make any government changes, why are Caplan’s changes more of a problem?

  • Dan

    I think Caplan’s argument makes much more sense as a critique of direct democracy. Establishing trade policy etc. by ballot measure would be quite unwise. However, the irrational voters aren’t actually making policy (and to a significant extent are not even aware of the policies that do get made.) Further, unless the cutoff point was fairly drastic, I don’t think the elected leaders would adopt policies that were much different from what they do now, and certainly not different enough to allay the legitimate concerns expressed by zenkat and Roger Rainey.

  • michael vassar

    When I think of voter competency tests I tend to think of the Chinese Mandarinate more than of the postbellum south.
    Very Very stable, but Very Very stagnant.

    I’m sure that a well-informed committee of social scientists with contemporary knowledge could invent a constitution FAR superior to that of the current US, but how to pick the experts. One way would be for the US federal government or some other nation with lots of territory (Canada?) to sell sovereignty (probably limited by some restrictions and possibly subject to polity-level tribute requirements) to those who were confident in their systems. In addition to the immediate revenues, they would derive beta-testing of new political possibilities. It seems stupid to test new political theories on a polity the size of the US. Just the possibility of doing so creates HUGE rent-seeking among the many entities who imagine that they should be the ones to have their theories so tested, and usually be given huge pork payments while they are at it to compensate them for the rent-seeking expenses incurred.

    As a practical matter, it would probably be necessary to enable these new countries to bar entry but not to restrain exit (or cause injury?). Probably an absolute ban of this sort would be necessary, implying banishment as a possibility but not prison. The recipient polity (the US or Canada probably) would probably charge a standard fee for each immigrant absorbed as a result of banishment.

  • Hopefully Anonymous

    You write “Why test folks on economics rather than social psychology, evolutionary psychology, Bayesian reasoning, statistics, general scientific literacy, historical knowledge, historical knowledge of the late Roman Empire”.

    My response: of course. It doesn’t make much sense in my opinion to just test on economics (although if we could only test on one thing -which is not the case- it would be a pretty good sole subject). However the general concept of voter competency/literacy testing is a promising one in my opinion, and we don’t have to discard it just because there may be some Caplan bathwater around the baby of optimizing effective governance.

    I don’t think the US is as anomolous as you say, particularly post-world war II. But either way, I think state level goverment experimentation is one of the more underutilized advantages of our federal republic. I’d like to see much wider state level experimentation, particularly on economics-related policy, and ways of optimizing population behavior. It might help if the government gave states more slack in temporarily restricting intra-state immigration and witholding full faith and credit for the laws and judgments of other states for the explicit purpose of running such experiments.

  • michael vassar

    Robin: In practice I probably *do* think that as far as possible we should never make any government changes. The actual dynamics that seem likely to generate changes seem to me to overwhelmingly favor unfortunate over desirable changes. I can see all sorts of changes that I think would be desirable, but given the power of the status quo as a Schilling Point I don’t see any that I think offer a better expected marginal reward per unit of political effort than does the prevention of possible harmful changes.

    I’m not saying this as a Burkean conservative. I don’t think that there has been enough time for evolution, which proceeds through testing to destruction, to build much algorithmic information into existing political institutions, so in so far as they work I think it’s mostly due to rational design, but I’m also not at all sure that they work very well at all. It’s just that the information input mechanisms that could be brought to bear to redesign them seem to dramatically favor insanity over rational design. Progress may in practice be the result of improvements in human knowledge as expressed in culture producing a historical trend of better people with worse institutions over time.

    In this framework, I suppose that one way of thinking about colonialism is to model it as “primitive” societies having excellent institutions like weak government which enabled bearers of more advanced culture from societies with worse institutions to come in and create massively more value than they could at home.

  • Hopefully Anonymous

    Michael, could you elaborate more on why “the actual dynamics that seem likely to generate changes seem to me to overwhelmingly favor unfortunate over desirable changes” -because I don’t see that … if anything, it still seems to me to be a very open empirical question. If it is (an open empirical question) I think that would encourage diversified experimentation rather than avoiding governmental change as much as possible.

  • michael vassar

    Heavily for the sort of reason that Bryan discusses in his book.
    Efforts to change the rules suffer from the same problems that other policy decisions within democracy suffer from.

    I agree that diversified experimentation would be VERY powerful and helpful, but I’m skeptical as to it’s probability.

  • Aaron Davies

    Um, we *did* phase in women’s suffrage one state (or territory) at a time, at least to start. See History of women’s suffrage in the United States.

  • Eliezer Yudkowsky

    Eliezer, I don’t understand why you don’t think we could not start out with small incremental versions of Caplan’s proposals to “weigh educated votes more, or at least stop trying to get out the vote, and that we reduce the scope of government policy.” Which direction to explore policy changes is a different issue from how big and fast to move. Unless you think we should never make any government changes, why are Caplan’s changes more of a problem?

    Robin, I’d have much less of a problem if Caplan had suggested trying this in Louisiana first. I wouldn’t expect it to work because of group polarization and false expertise and compartmentalization, but I wouldn’t try to stop him from trying.

    Aaron, I stand corrected.

  • Hopefully Anonymous

    “false expertise” and “compartmentalization” seem to me to be problems that can perhaps better be managed by reducing the impact of incompetent voters, who would be less adept and screening out those type systematic problems in administrative governance. I’m curious why you apparently disagree?

  • Robin Hanson

    Eliezer, Caplan mainly just suggested a direction to move – he didn’t specify what scale on which to start experimentation.

  • Doug S.

    I propose that we adopt the political system that allocates the most resources to myself. 😉

  • Hopefully Anonymous

    Doug S.,
    I agree, as long as you will devote those resources to maximizing my personal odds of persistence. =D

  • Keith Elis

    US Voters rarely vote directly on policy, instead they seem to choose well-educated people from among themselves to vote on policy for them. The system has evolved such that rational donations of wealth decide who will be available as a viable candidate. This rational money flows according to each candidate’s values and odds of winning. The resulting pool of viable candidates usually includes bright, well-educated, well-connected people any of whom could be expected to be more rational than average. So no matter how voters make their ballot decisions, values or illusions, the candidate that eventually wins and finds him or herself in a position to vote on policy will probably do so rationally.

    There is no guarantee that this decision-maker will make decisions that correspond to the aggregate of voters’ values *or* illusions. But this is a different problem; not the one with which Caplan seems to be wrestling.

  • Hopefully Anonymous

    I think your description of how voter irrationality is mitigated by more rational stakeholders in American public policy has a lot of merit. However, there’s also a functional vs. optimal element to this. The American system may have developed some functional ways of screening out the systematic biases of its least rational voters. However, I think there is still likely value in diversified experimentation to look for ways to optimize the policy decision making process further. Voter competency screening looks like an approach worth testing for that purpose, in my opinion.

  • TGGP

    Keith Elis, I’m not sure campaign funding matters all that much. There’s a section in Freakonomics on it.

    Speaking of Freakonomics and female suffrage, what do you guys think of John Lott’s theory in Freadomnomics that <>giving women the vote resulted in expansion of the state?

  • Hopefully Anonymous

    TGGP, sounds plausible. Not an earth-shattering insight, but a plausible one.

  • Matthew C

    TGGP, I’ve heard that theory since long before Freedomnomics came out. There certainly exists some evidence for it.

  • Peter McCluskey

    Michael, you imply that Burkeans think political institutions evolve in a Darwinian fashion. I suggest that Lamarckian evolution is a better description might be a better interpretation of what Burke meant (my recollection of Burke is pretty rusty, but he did write well before Darwin).
    Is there any way to analyze how much information Lamarckian evolution could have built into our political institutions?

  • Hopefully Anonymous

    An interesting article, from a very cool blog (someone to invite to be an overcomingbias contributor?). An inverse correlation between voter turnout and schizophrenia in neighborhoods indicates a possible reason to allow arguably incompetent voters to vote:

  • mnuez

    You gotta love economists (almost as much as you gotta love lawyers, politicians and casino owners). They have full and complete faith in the ability of the individual human being to engage in his most responsible, self-benefiting behavior when he engages in the market (despite the few measley trillions of dollars spent in psychological/marketing studies designed to fool people into spending money in ways that do not serve their best self-interests) yet they have zero trust in the same small human unit when it comes to his knowing how to vote his self-interests.

    Wonderful human beings, these economists. Let’s put them in charge. (I’m joking of course, they already are in charge.)

    Update: Wow. I wrote the above when only about a third of the way through the New Yorker article when I wrote the above. Having a fascination with the science of economics though I continued reading the piece in the assumption that – though I might disagree with and dislike the beliefs and goals of the economists of the piece (the economist written about and the economic-minded fellow doing the writing) – I would probably learn something interesting nonetheless.

    Well, I’ve read further and the learning has yet to have taken place. What I have come across however are patronizingly disgusting biases disguised as honest reporting.

    Here for example, the article’s author offers a few theories he considers to be silly and naive as to why allowing everyone to vote might be something less than a tragedy. Among these naive theories that silly people might grasp at so as to rationalize the allowance of universal voting, the author offers…

    “Then, there is the theory that people vote the same way that they act in the marketplace: they pursue their self-interest. In the market, selfish behavior conduces to the general good, and the same should be true for elections.”

    But of course that isn’t the case. In the market, you see, every guy who buys a fifty thousand dollar car because well-designed ads have convinced his subconscious that he needs this cars lest he be emasculated – is of course benefiting and serving his truest “self-interest” with this debt-inducing purchase. With regards to voting, by contrast, people might stupidly support policies that well respected Objectivists know to be contrary to Randian philosophy.

    And here’s the article’s offering of the “four main areas” where people have terribly “irrational” misunderstandings regarding economic policies that “differ from [those of] the economic expert”.

    “The average person, he says, has four biases about economics—four main areas in which he or she differs from the economic expert. The typical noneconomist does not understand or appreciate the way markets work (and thus favors regulation and is suspicious of the profit motive), dislikes foreigners (and thus tends to be protectionist), equates prosperity with employment rather than with production (and thus overvalues the preservation of existing jobs), and usually thinks that economic conditions are getting worse (and thus favors government intervention in the economy).”

    So in short, being suspicious of the profit motive is silly, being employed is less important than ensuring that good “productivity” is taking place and strawmen (xenophobia and alarmist fears as having anything to do with people’s interest in some protectionism and governmental intervention) are always fun.

    Then there’s some more mockery of people’s silly interest in survival (“people really believe that the country would be better off if profits were regulated, if foreign goods were taxed, and if companies were prevented from downsizing”), some school-mistressly scare-mongering of the silly little boys (“politicians who pander to these beliefs are more likely to be elected, and the special interests that lobby for protectionism and anticompetitive legislation are the beneficiaries—not the public. The result, over time, is a decline in the standard of living”) and the inevitable coup de grace of suggesting as reasonable that none but Economics Orthodoxists be allowed to vote in elections.

    But really now, why concern ourselves with matters such as these when somewhere in Wyoming there may be a crazy on the loose?

  • Hopefully Anonymous

    mnuez, I sense a little polarization bias? I think there are good reasons to consider voter threshhold competency requirements that have little to do with the ideological bathwater that Caplan added to this baby. Why not focus on the best arguments, rather than take an approach “Caplan said this stupid/ideological baiting thing as a reason for concept X, I have an opposite ideological perspective and he has thus effectively baited me, therefore I oppose concept X”. Your and Caplan’s polarization dance party is getting in the way of my desire for more rational and effective governance, in my opinion.

  • TGGP

    mnuez, I noticed you said you take have “a fascination with the science of economics”. Which economist or school of economics do you find the most interesting, or has had the greatest effect on your views? Do you read much public choice, rational choice or behavioral economics (these would be among the most relevant for Caplan’s argument)? Finally, I think you might like to read more of Caplan’s research (which can be found at this page and discussed at econlog), which would give you a better picture than the New Yorker article.

  • Hopefully Anonymous

    Thanks TGGP. Here’s a 2006 article I found on the site:

    “The Gender Gap of Economics: Why Do Men Think More Like Economists? Evidence from the Survey of Americans and Economists on the Economy”

    This gem will sink viability of Caplan’s proposal unless economic literacy is counter-weighted with something women outperform men on (reading comprehension of passages about candidates or policies?) on a voter threshhold competency test. Because regardless of whether men are more economically literate than women or not, women vote in greater proportion than men, and I don’t see them giving up that democratic advantage.

  • mnuez

    Hopefully ~

    You’re correct in noting that I didn’t address myself to the issue of whether there should be any threshold of competance required for voting but you’re incorrect in assuming that this was the case owing to the fact that I dislike Caplan’s economics.

    I didn’t address myself to the issue of whether all should be allowed to vote because it wasn’t the most important issue of the article. The most important issue was the fact that Caplan and the article’s author were portraying as a given that Randian econimics were “correct” and that left leaning economics (which they presented as the beliefs of the average ignoramous) were “irrational”. That sort of coup could not be allowed to pass unnoticed.

    Whether there should be some threshold of study required before a citizen is allowed to vote is a seperate (and far less relevant) mattter. The real issue here is that an elitist asshole (whom I’m calling such based only his repulsive economic tastes) and his reviewer are attempting to take control of the intellectual atmosphere regarding economic policies by presenting their own Darwinesque tastes as “correct” (and that’s quoted because that’s the exact word that they used) and as more left-leaning Judeo-Christian economic tastes as “irrational”.


  • Hopefully Anonymous

    Mnuez, I still think you’re over-feeding off of polarization bias. I think it might be more fruitful to sidestep a whole left-right approach to Caplan and to focus on what policies (economic and otherwise) actually seem best to achieve particular goals (such as one I’d suggest, to maximize your and my mutual odds of persistence).

  • TGGP

    mnuez, Rand was not an economist. There is no “Randian economics” (that’s philosophy/literature), nor “Darwinian” (biology) or “Judeo-Christian” (religion). Caplan seems to like Rand’s writings (he compares her to Victor Hugo), but is not an Objectivist. In his “Why I Am Not an Austrian” essay he refer’s to Huemer’s “Why I Am Not an Objectivist” as representing his views since graduate school. Furthermore, he does not necessarily attack “left-leaning economics”, if that includes economists who would consider themselves to the left of the center. He says that on average economists are more accurate than the general public AND (this is important) the average economist is left of center! Anti-foreign bias, in particular, is not at all a stranger to the political right. I myself (I’m a paleolibertarian and I despise both the representative left and right in this country, though I occasionally find myself in agreement with both anarcho-socialists and monarchists!) strongly disagree with his position on immigration, which he would attribute to that bias.

    If you would like to show that your disagreement with Caplan is on rational grounds, you could use the arguments that are numbered 1 through 7 above, or come up with some new ones. Referring to Caplan as an “asshole” (even if he is one, I personally do not know but presumably Robin does) is not likely to be an effective tactic and on the contrary would likely distract people from the points you are trying to make.

  • Jim Outen

    You embarrass yourself with this comment:

    “[Economists]have full and complete faith in the ability of the individual human being to engage in his most responsible, self-benefiting behavior when he engages in the market (despite the few measley trillions of dollars spent in psychological/marketing studies designed to fool people into spending money in ways that do not serve their best self-interests) yet they have zero trust in the same small human unit when it comes to his knowing how to vote his self-interests.”

    Read Caplan’s book or related articles, as he addresses this issue thoroughly. He details how voters do not vote to maximize their financial self-interest, and instead entertain various fantasies and delusions. If I recall, he calls this the “near-neoclassical” demand curve, and claims that voting mechanism makes the expected value of a vote for your self-interest very low (the converse being that the price of irrational beliefs is low, thus we happily indulge in them).

    Reading before ranting is generally a superior strategy.

  • mnuez

    It just seems so unlikely. Would a blog that calls itself “overcoming bias” engage in censorship?

    I blame typepad. Until an experiment has been replicated Occam’s Razor would demand that we assume that a blog called Overcoming Bias would never delete non-spam comments.

    I’ll attempt to replicate that experiment now with a re-post.


    P.S. In the unlikely event that my recent comment was purposely deleted and that the subsequent one will be deleted as well but that for some reason THIS comment will remain, I direct anyone interested in reading the deleted comment to my blog where it’s been reprinted. But again, deleting comments does not quite seem to be in accordance with the supposed intellectually honest values of this blog so I’ll have to assume that never took place. ~ m

  • Robin Hanson

    Mnuez, we don’t allow arbitrarily long comments here. I sent you an email explaining that your 1000 word comment was way too long, noting that you had a previous comment (still up) on this post that was also too long, and inviting you to put a comment here with a link to your longer comment.

  • mnuez


    First of all I should note that I enjoy your posts and that this is not my first visit to Overcoming Bias and that I’ve also noted that I enjoy this blog on my own blog.

    Lest you think that I’m saying this in order to flatter you I’ll continue with the following 🙂

    Aside from the fact that I never received any email from you, I should also note that an explanation for censorship of ideas in the form of letting me know that you don’t allow “arbitrarily long comments” is quite silly. I’m not entirely certain as to what an “arbitrarily” long comment would be but I’d find it hard to imagine that it would be a detailed response that speaks directly to comments made in response to an earlier comment of my own.

    Perhaps I’m crazy but it would appear to me that the more true reason for the wholesale deletion of my comment would be the more impolitic explanation to the effect that it might offend some people.

    And if that isn’t bias NON-overcome I’m not sure what is.

  • mnuez

    CORRECTION: Your email was indeed sent and received as you said. For mysterious reasons Gmail had spammed it (dunno why).

    Of course the rest of my comment stands. My clear and detailed response to three comments (made on a previous comment of mine) was not in any way “an arbitrarily long comment” and appears to have been excised lest it offend – something that would appear to fly in the face of what the principles of Overcoming Bias claim to be.


  • Hopefully Anonymous

    Just post a summarized version here with a link to the longer version on your blog.

  • mnuez

    Summary: I responded to the comments made by Hopefully Anonymous, TGGP and Jim Outen.

    Body: Available here

    Conclusion: The opinions I expresssed were censored. Allowing a link to elsewhere is in no way the equivalent of not deleting the possibly offending comments from the very conversation where they are relevant.

  • Hopefully Anonymous

    I just read your replies and I don’t think you were censored. Also, you could probably repost the deleted comment in the Open Thread here at overcoming bias without it being “censored”.

    I think your contrasting between “judeo/christian” vs. “lassiez faire/darwinian/nazi” economic public policy is unhelpful for enlightenment or for best policy formation. But I do think you’re being baited into that framework by folks like Caplan -so I think it’s an emergent problem from what I called the “polarization bias dance party” between Left/Right types. Both groups seem to be satisfying mutual psychological needs by reducing policy discussion to these two polarized points, but I think it’s far from optimal in terms of actually determining best policy (to maximize my (or our) persistence odds).

  • Pingback: Overcoming Bias : Voting Is A Far Fest