Thanks for writing this; I find approval voting an interesting and intuitively attractive system (and have used it in the past).
I was surprised by the table you show about the impacts of approval voting on the democratic candidate selection. I generally think of approval voting as supporting moderate candidates, as you mention, but here it seems to be favouring the most extreme (Warren, Sanders) over the more moderate (Biden) - though perhaps I have misinterpreted the chart.
This makes me worry that increasing the ‘democraticness’ of elections might lead to worse outcomes. You sort of alude to this concern in the FAQ, but move over it pretty quickly. There are in fact many cases where it is I think relatively common to think that reducing ‘democraticness’ is a good move:
Independent central banks are much more competent, credible and technocratic than having elected politicians control the money supply. I’m not really aware of anyone who thinks this move was a bad idea.
Elected judges are often viewed as far more populist and generally lower quality than appointed ones.
Elected utility commissions are less competent than appointed ones (though perhaps I am biased on this issue).
Referendums and ballot initiatives are often blamed as part of the cause for the decline in Californian governance.
Elected vs Appointed isn’t exactly the same as FPTP vs Approval Voting, but it seems like they have similar aspects.
Garrett Jones has a book on this. It’s only on pre-order at the moment, but you can read Hanson’s comments here.
On the moderate component, it’s important to note a couple things: (1) moderate can change depending on what the population is and (2) sometimes we get a distorted view of what moderate is through the media.
Here, we’re looking at a subgroup—people registered as democrats. So the population is a bit different. One of the platforms that Warren and Sanders are similar on that separate them from Biden is Medicare for All. It tends to poll rather well, particularly among democrats despite it being considered more extreme by the media.
On the democracy not always being good component, I touched on this a little elsewhere with a kind of “what else?” type reply. But perhaps, as you mentioned, there are some positions that are best not elected. I would take less issue with positions best not elected but more issue with positions elected badly (ex// FPTP).
Independent central banks are much more competent, credible and technocratic than having elected politicians control the money supply. I’m not really aware of anyone who thinks this move was a bad idea.
Ron Paul and a large fraction of American internet libertarians believe in the badness of technocratic central banks quite strongly[1], though I agree that this is not a mainstream (or even minority) position among academics or other people I trust on economic issues.
He opposes the Fed, but does he want interest rates set by politicians? My understanding is he wanted a return to the gold standard, where there is less need to directly control the money supply—the only question is whether you insist on 100% backing or accept a lower ratio, but once that is set growth in the money supply is determined by the volume of physical gold.
Similarly bitcoin people oppose the Fed, but not because they want politicians in charge—they have another external rule (e.g. fixed max quantity) that reduces democratic discretion.
Hi Aaron! Thanks for the interesting post. I am very sympathetic to the criticisms of plurality voting you outline, and I agree that alternative voting methods are worth pursuing. This certainly seems like a tractable problem (one which we’re already making headway on) and, like the study of institutional design more generally, I believe it is unduly neglected. With that said, I’d like to sketch some reasons to moderate our confidence in the purported benefits of a transition to something like approval voting.
I take it that part of the motivation for transitioning to approval voting is that our current voting system leads to bad policy and bad outcomes. You write:
“It’s hard to overstate how important our work is. We focus primarily in the US—arguably the most influential country in the world given its GDP and heavy reach over foreign policy. You don’t want a country with this stature to have a broken voting method. Bad policies and irresponsible spending are inevitable, and they affect the rest of the globe.”
Of approval voting, you write that it encourages good candidates to run, as well as favoring “the electorate’s middle”. In conjunction with fostering a political environment conducive to the formation and spread of new ideas, it also promises to benefit future people. Later, when responding to a question about the connection between the implementation of approval voting and the electorate advancing better policies, you rightly point out that if it isn’t the voters advancing policy then it may be some other group who may not have our best interests in mind.
That sounds right. But it doesn’t follow from the truth of the relevant counterfactual that approval voting would lead to better policy and better outcomes. The connection between favoring the electorate’s middle and the creation of better policy is tenuous. Political psychologists have studied levels of voter knowledge for decades, and virtually every study shows that voters are ignorant of even basic political facts. Given standard models of rational voter ignorance (and rational irrationality, etc.), this shouldn’t be surprising. Oversimplifying for a moment, the electorate’s middle are in all likelihood systematically mistaken about the sort of policies that would advance their interests; and when you pair these voters with political leaders who are incentivized to pander, we have a recipe for occasional disaster. I see no reason why this wouldn’t occur in a system with approval voting in the same way that it occurs in our current system.
I have similar reservations about the purported ability of approval voting to foster an environment in which good, new ideas can spread. In line with the above, I think approval voting guarantees the spread of new ideas at best, not necessarily new and good ideas. Without the flow of good and new ideas, the alleged benefits for future generations are probably overstated too. (On future generations, I favor thinking about possible institutional reforms which *directly* incentivize greater regard for future generations. Tyler John recently posted about this on the EA forum.)
In general, I think the case for approval voting outlined focuses too much on the supply-side of politics and not enough on the demand-side. Now, it is easy to overstate the significance of the demand-side (for reasons I am happy to get into, if you like). Still, I think this is a serious problem, and one that approval voting does not mitigate as far as I can tell.
On future generations, I favor thinking about possible institutional reforms which directly incentivize greater regard for future generations.
I’m curious why you say this given that you earlier noted the problem that a large part of the electorate “are in all likelihood systematically mistaken about the sort of policies that would advance their interests.” Making people give more regard to future generations seems to be of extremely unclear value if they are likely to be systematically mistaken about what would serve the interests of future generations. This seems like a consideration in favour of interventions which aim to improve the quality of decision-making (e.g. via deliberative democracy initiatives) vs those which try to directly make people’s decisions more about the far future (although of course, these needn’t be done in isolation). But perhaps I am simply misunderstanding what you mean by “directly incentivize greater regard for future generations”?
Thanks, David. I should have been clearer. I certainly don’t support interventions that incentivize greater regard for future generations without also attempting to improve the overall quality of decision-making.
In the article above, Aaron claims that a transition to approval voting would benefit future people. For reasons outlined above, I am skeptical of that claim. It is unclear whether greater responsiveness to the preferences of the electorate’s middle would bring about greater regard for future generations. Moreover, the evidence from public polling and political psychology suggests that even if the electorate’s middle had sufficiently high regard for future generations, they would have mistaken beliefs about what sorts of policies would benefit future generations. (Note: I don’t claim that political psychologists have gone out and investigated the degree to which regular voters possess knowledge of future-beneficial institutions and policies. Maybe they have, but I haven’t encountered such research myself. Instead, I claim that given the ignorance of the electorate with regard to even very basic politically relevant facts, we should expect them to be ignorant of future-beneficial institutions and policies.)
Now, we might try something like age-weighted voting. But this strikes me as an intervention that ignores the overall quality of decision-making. At best, it places comparatively more political power in the hands of people who might regard future generations more than their older counterparts. This is the sort of institutional reform that is in tension with my worries about the demand-side of politics.
I favor thinking about ways to both incentivize greater regard for future generations and improving the overall quality of decision-making. I have no settled opinions on what the eventual institutions would look like. Perhaps they would involve independent agencies acting in an advisory capacity, or perhaps they would involve novel governmental bodies tasked specifically with representing the interests of future generations, or perhaps something else entirely. Deliberative reform will probably play some role, but beyond that I don’t know. Institutional design is complicated business and we shouldn’t pretend we know in advance of serious empirical investigation what will work best. Still, I think we currently know enough to know that mere tweaks to the current voting system without improving the overall quality of decision-making will not be enough.
Given standard models of rational voter ignorance (and rational irrationality, etc.), this shouldn’t be surprising. Oversimplifying for a moment, the electorate’s middle are in all likelihood systematically mistaken about the sort of policies that would advance their interests; and when you pair these voters with political leaders who are incentivized to pander, we have a recipe for occasional disaster. I see no reason why this wouldn’t occur in a system with approval voting in the same way that it occurs in our current system.
I can think of one reason: rational ignorance is partially a consequence of the voting procedure used. People have less of an incentive to be ignorant when their votes matter more, as they would with approval voting. I don’t have a strong stance on this, but I think it’s important to recognize that studies about voter ignorance are not yielding evidence of an immutable characteristic of citizens; the situation is actually heavily contingent.
In the first few pages of The Myth of the Rational Voter, Bryan Caplan makes (implicitly) the case that voter ignorance isn’t a huge deal as long as errors are symmetric: ignorant voters on both sides of an issue will cancel each other out, and the election will be decided by informed voters who should be on the “right” side, in expectation. Caplan claims that systematic bias across the population results in “wrong” answers.
My point in bringing this up is just that the existence of large numbers of ignorant voters doesn’t have to be a major issue: large elections are decided by relatively small groups. Different voting procedures have very different ramifications for the composition of these small groups.
Of course I agree that studies of voter ignorance do not yield evidence of some fixed, immutable characteristic of citizens. It is also certainly true that different voting systems provide different incentives to acquire relevant political information. The pertinent question is whether implementing approval voting would incentivize the acquisition of political information to a sufficient degree that we could be confident in claiming that the mere transition to approval voting alone would have all the benefits that Aaron claims it would have. I am still very skeptical of those purported benefits being delivered. Even on minimally demanding normative accounts of what amount of knowledge voters ought to possess (models of retrospective voting, say), the opportunity cost of acquiring the relevant information is, for very many citizens, simply too high. I suspect this will be true even under a scheme of approval voting, and even if more voters were to think their votes matter more. For all we know, the transition to approval voting might simply increase voter turn-out by prompting citizens who previously abstained to go to the polls, now thinking that their votes matter; and since the sort of evidence about voter ignorance alluded to earlier suggests that non-voters are typically less informed than voters, we might end up in an even worse scenario as far as voter ignorance is concerned. Like you, I don’t have a particularly strong stance on this. This is clearly an empirical matter. Still, we shouldn’t rule this scenario out entirely.
In my initial comment, I mentioned that I think it is easy to overstate the significance of the demand-side of politics. This is partly because of the literature on epistemic democracy, which you allude to above in mentioning the so-called “Miracle of Aggregation” (i.e. symmetric errors cancelling each other out). Epistemic democrats would indeed dismiss my concerns about voter ignorance, for they claim that, under certain conditions, collective intelligence can emerge from large groups of individually ignorant agents. If that’s right, then the evidence from political psychology needn’t worry us. However—and with all due respect to the many theorists working on epistemic democracy (it’s nice to say “with all due respect” and mean it) - the claims about the collective intelligence of individually ignorant voters often strike me as pollyannish. To take the case of the miracle of aggregation, I don’t think the required symmetry holds. Without that symmetry, the collective intelligence does not emerge. To take another favored strategy of epistemic democrats (the Condorcet Jury Theorem), I don’t think the necessary conditions for the emergence of collective competence are satisfied.
However, another reason to downplay the significance of the demand-side of politics is worth mentioning, since it is often ignored by various parties to disputes about levels of voter ignorance. The degree to which political leaders respond to voter preferences is often greatly overstated. Policymakers and political representatives enjoy a significant degree of autonomy, and much—if not most—policy-making is done out of the public eye entirely. When you couple this with recent work in political psychology suggesting that voters are often simply happy to toe the party line, this suggests that much of the focus on voters, their knowledge, and politicians responding to their preferences has been misplaced. This is as true of epistocratic worries about voter ignorance as it is about the epistemic democrats who oppose them. I don’t think this means that voter ignorance isn’t a problem (assuming for the moment that it is indeed a problem), but it’s not the only thing we need to be looking at. Many factors other than voter competence contribute to the overall quality of governance, and voting reforms (whether approval voting or otherwise) will not change such factors much, if at all (factors such as the degree to which policymaking can be captured by special interests, levels of corruption, the decision-making methods deployed by legislators and bureaucrats, etc.).
Hey Aaron! Thanks for posting this. I am likely going to include CES in my giving this year as a result of some of the points you’ve made here.
I’ve been researching lobbying recently and I’m curious about this passage:
We will lobby legislators. Because of approval voting’s simplicity, there are opportunities for lobbying elected officials. Normally, this isn’t an option because of the conflict of interest with those elected. But the opportunity presents itself when the party in power suffers because of vote splitting yet wants to avoid implementing a complex method. There are places where RCV is stalled out where we have opportunities. These are typically higher risk but very high reward since they don’t require the same resources as a campaign. Our estimate is that they can be one sixth the expected cost per citizen compared to ballot measures when factoring in their relative probability of success. This also requires funding for a 501(c)4 to do this effectively at scale.
I’m not particularly skeptical about this one-sixth estimate, but I haven’t been able to find anything like it my lit review! Do you have some background on this research?
This estimate is based off of me asking our Director of Campaigns and Advocacy about likelihood of success with lobbying. He gave me an estimate and then I did a correction based off what I’d seen, which lowered the success probability to 5% per effort. The expected cost per person (factoring in this probability) is much lower because when it does work, you can get an entire state to change their voting method. The scale counteracts the cost and probability rather quickly.
That said, lobbying is a challenging effort. I’m cautious about spending too much time and money here without having a 501(c)4 as it can push away our resources from wins that are more likely. We have a need for momentum at the moment. Still, with the potential cost effectiveness, this is an area we’re considering in the future. It’s likely we’ll be risk averse initially as we gain experience.
If you have particular resources you’ve found on lobbying, feel free to share. Like I said, this would be a new space for us. We had one opportunity here recently, but we decided against it based on the particular circumstances.
I’ll post a summary lit review here on the forum when I’m done with my research. Spoiler alert: political scientists don’t have a great idea of how/why/whether lobbying works and research on its effectiveness is almost strictly limited to trade policy and large publicly traded firms. So you get expressions of effects like “$140 in additional shareholder value for every $1 spent on lobbying.” Interesting, but not particularly generalizable.
It seems like CES’s strategy so far has been to start small, which makes obvious sense. I’m curious to know when/if you make the decision to withdraw from a local advocacy effort that seems like it’s not paying off. It’s not obvious to me that public support is monotonically increasing in dollars spent on advocacy— what’s your stopping rule?
Right now, initiatives are our main path, but with more funds we’d be open to experimenting in lower-risk scenarios. I’d be excited to read what information you come up with.
We use polling as an initial indicator to see how receptive a demographic would be to the initiative. Fortunately, the simplicity helps us with receptivity.
Within a particular target city, we don’t want to add more dollars than necessary to an effort. Winning soundly is important, but we didn’t need to throw $1M directly into Fargo, for example. I suppose if we had, the support may have gone nonmonotinic in relation to the spend and backlashed against us. But I don’t see that as a particularly big risk for us. We’re more efficient than that.
Also, being as early as we are in the game, we’re a little cautious about taking on a city we don’t think we can win. We’re aggressive, but not more aggressive than we think we can get away with.
Q: How would we know approval voting would give better policies? Why would the policies be good? And how do we know the policies would be good rather than merely popular?
A: I’m combining these ideas I pulled out because of their similarity. Approval voting tends to pull out the middle viewpoint (whatever that is for a particular electorate). And because viability is not an issue to gain initial support, it can provide a ramp for new ideas.
Is it possible that the popular opinion is bad? It sure is. But for this to be a real worry under approval voting would mean that the alternative of what we have now as being better. We might find that there is some popular issue that people win on that is not correct or overall good. This is also possible now, but with a poorer voting method. The question is whether approval voting provides a net gain so that popular issues that are good actually move forward at a pace faster than they would otherwise. Because of the quicker feedback and approval voting doing a better job of capturing candidate support, I believe this to likely be true. And to the degree that this is true in enacting better policies, many policies have carryover benefits that go well into the future.
I’m not sure I understood the question on supply and demand-side politics.
P.S. I accidentally added this as a main thread reply. Sorry about that.
Has CES done any analysis of Causus voting? I did not see any reference to it on the CES website.
With the primaries coming up, this might be interesting to explore. A recent On Point podcast interviewed Iowa voters and the process sounded somewhat similar, in that people need to compromise and come to a consensus. I am not suggesting caucus voting as a reform, but I wonder if some analysis could flesh out how it overlaps or differs with other systems like approval voting. If similar, that might provide some comfort to voters that approval voting is not so radical vs. existing systems.
I’d like to see us do much more research and evaluation, but we currently don’t have it in our budget to hire a Director of Research and support staff.
Aaron, thanks much for the update. Did the poll include non-democratic primary voters—e.g. registered republicans, etc.? A broader view of citizen views might offer additional insights.
I agree that more research and evaluation is desirable. In fact, I think that robust evidence is critical for concepts like approval voting to gain traction and support. Could CES organize volunteers to do exit interviews at polling locations during primary and general elections? Capturing plurality, approval, and even other voting preferences during real elections could offer valuable insights. It would also raise the visibility and awareness of different voting systems. And it might also help citizens understand how different systems can affect elections.
The exit polling could also ask people if they would optionally like to receive results (SMS, email.., [not tied to their vote choices]), thus additionally increasing awareness and offering the potential for future polls.
The poll included only those who intended to vote in the Democratic Primary.
It’s very difficult to manage volunteers in this way, particularly given our small staff size. We tend to contract polling out. That said, it takes some expertise to sort through the data. Having staff for research would help us dramatically in both evaluating voting methods and measuring progress within cities that we’ve won in.
Thanks for writing this; I find approval voting an interesting and intuitively attractive system (and have used it in the past).
I was surprised by the table you show about the impacts of approval voting on the democratic candidate selection. I generally think of approval voting as supporting moderate candidates, as you mention, but here it seems to be favouring the most extreme (Warren, Sanders) over the more moderate (Biden) - though perhaps I have misinterpreted the chart.
This makes me worry that increasing the ‘democraticness’ of elections might lead to worse outcomes. You sort of alude to this concern in the FAQ, but move over it pretty quickly. There are in fact many cases where it is I think relatively common to think that reducing ‘democraticness’ is a good move:
Independent central banks are much more competent, credible and technocratic than having elected politicians control the money supply. I’m not really aware of anyone who thinks this move was a bad idea.
Elected judges are often viewed as far more populist and generally lower quality than appointed ones.
Elected utility commissions are less competent than appointed ones (though perhaps I am biased on this issue).
Referendums and ballot initiatives are often blamed as part of the cause for the decline in Californian governance.
Elected vs Appointed isn’t exactly the same as FPTP vs Approval Voting, but it seems like they have similar aspects.
Garrett Jones has a book on this. It’s only on pre-order at the moment, but you can read Hanson’s comments here.
Hi Larks,
On the moderate component, it’s important to note a couple things: (1) moderate can change depending on what the population is and (2) sometimes we get a distorted view of what moderate is through the media.
Here, we’re looking at a subgroup—people registered as democrats. So the population is a bit different. One of the platforms that Warren and Sanders are similar on that separate them from Biden is Medicare for All. It tends to poll rather well, particularly among democrats despite it being considered more extreme by the media.
On the democracy not always being good component, I touched on this a little elsewhere with a kind of “what else?” type reply. But perhaps, as you mentioned, there are some positions that are best not elected. I would take less issue with positions best not elected but more issue with positions elected badly (ex// FPTP).
Ron Paul and a large fraction of American internet libertarians believe in the badness of technocratic central banks quite strongly[1], though I agree that this is not a mainstream (or even minority) position among academics or other people I trust on economic issues.
[1] https://en.wikipedia.org/wiki/Ron_Paul#Economic
He opposes the Fed, but does he want interest rates set by politicians? My understanding is he wanted a return to the gold standard, where there is less need to directly control the money supply—the only question is whether you insist on 100% backing or accept a lower ratio, but once that is set growth in the money supply is determined by the volume of physical gold.
Similarly bitcoin people oppose the Fed, but not because they want politicians in charge—they have another external rule (e.g. fixed max quantity) that reduces democratic discretion.
Maybe this discussion is a bit tangential to The Center for Election Science’s fund-raising.
Hi Aaron! Thanks for the interesting post. I am very sympathetic to the criticisms of plurality voting you outline, and I agree that alternative voting methods are worth pursuing. This certainly seems like a tractable problem (one which we’re already making headway on) and, like the study of institutional design more generally, I believe it is unduly neglected. With that said, I’d like to sketch some reasons to moderate our confidence in the purported benefits of a transition to something like approval voting.
I take it that part of the motivation for transitioning to approval voting is that our current voting system leads to bad policy and bad outcomes. You write:
“It’s hard to overstate how important our work is. We focus primarily in the US—arguably the most influential country in the world given its GDP and heavy reach over foreign policy. You don’t want a country with this stature to have a broken voting method. Bad policies and irresponsible spending are inevitable, and they affect the rest of the globe.”
Of approval voting, you write that it encourages good candidates to run, as well as favoring “the electorate’s middle”. In conjunction with fostering a political environment conducive to the formation and spread of new ideas, it also promises to benefit future people. Later, when responding to a question about the connection between the implementation of approval voting and the electorate advancing better policies, you rightly point out that if it isn’t the voters advancing policy then it may be some other group who may not have our best interests in mind.
That sounds right. But it doesn’t follow from the truth of the relevant counterfactual that approval voting would lead to better policy and better outcomes. The connection between favoring the electorate’s middle and the creation of better policy is tenuous. Political psychologists have studied levels of voter knowledge for decades, and virtually every study shows that voters are ignorant of even basic political facts. Given standard models of rational voter ignorance (and rational irrationality, etc.), this shouldn’t be surprising. Oversimplifying for a moment, the electorate’s middle are in all likelihood systematically mistaken about the sort of policies that would advance their interests; and when you pair these voters with political leaders who are incentivized to pander, we have a recipe for occasional disaster. I see no reason why this wouldn’t occur in a system with approval voting in the same way that it occurs in our current system.
I have similar reservations about the purported ability of approval voting to foster an environment in which good, new ideas can spread. In line with the above, I think approval voting guarantees the spread of new ideas at best, not necessarily new and good ideas. Without the flow of good and new ideas, the alleged benefits for future generations are probably overstated too. (On future generations, I favor thinking about possible institutional reforms which *directly* incentivize greater regard for future generations. Tyler John recently posted about this on the EA forum.)
In general, I think the case for approval voting outlined focuses too much on the supply-side of politics and not enough on the demand-side. Now, it is easy to overstate the significance of the demand-side (for reasons I am happy to get into, if you like). Still, I think this is a serious problem, and one that approval voting does not mitigate as far as I can tell.
I’m curious to hear your thoughts on this!
I’m curious why you say this given that you earlier noted the problem that a large part of the electorate “are in all likelihood systematically mistaken about the sort of policies that would advance their interests.” Making people give more regard to future generations seems to be of extremely unclear value if they are likely to be systematically mistaken about what would serve the interests of future generations. This seems like a consideration in favour of interventions which aim to improve the quality of decision-making (e.g. via deliberative democracy initiatives) vs those which try to directly make people’s decisions more about the far future (although of course, these needn’t be done in isolation). But perhaps I am simply misunderstanding what you mean by “directly incentivize greater regard for future generations”?
Thanks, David. I should have been clearer. I certainly don’t support interventions that incentivize greater regard for future generations without also attempting to improve the overall quality of decision-making.
In the article above, Aaron claims that a transition to approval voting would benefit future people. For reasons outlined above, I am skeptical of that claim. It is unclear whether greater responsiveness to the preferences of the electorate’s middle would bring about greater regard for future generations. Moreover, the evidence from public polling and political psychology suggests that even if the electorate’s middle had sufficiently high regard for future generations, they would have mistaken beliefs about what sorts of policies would benefit future generations. (Note: I don’t claim that political psychologists have gone out and investigated the degree to which regular voters possess knowledge of future-beneficial institutions and policies. Maybe they have, but I haven’t encountered such research myself. Instead, I claim that given the ignorance of the electorate with regard to even very basic politically relevant facts, we should expect them to be ignorant of future-beneficial institutions and policies.)
Now, we might try something like age-weighted voting. But this strikes me as an intervention that ignores the overall quality of decision-making. At best, it places comparatively more political power in the hands of people who might regard future generations more than their older counterparts. This is the sort of institutional reform that is in tension with my worries about the demand-side of politics.
I favor thinking about ways to both incentivize greater regard for future generations and improving the overall quality of decision-making. I have no settled opinions on what the eventual institutions would look like. Perhaps they would involve independent agencies acting in an advisory capacity, or perhaps they would involve novel governmental bodies tasked specifically with representing the interests of future generations, or perhaps something else entirely. Deliberative reform will probably play some role, but beyond that I don’t know. Institutional design is complicated business and we shouldn’t pretend we know in advance of serious empirical investigation what will work best. Still, I think we currently know enough to know that mere tweaks to the current voting system without improving the overall quality of decision-making will not be enough.
Thanks for the clarification, I strongly agree with the position described in this comment.
I can think of one reason: rational ignorance is partially a consequence of the voting procedure used. People have less of an incentive to be ignorant when their votes matter more, as they would with approval voting. I don’t have a strong stance on this, but I think it’s important to recognize that studies about voter ignorance are not yielding evidence of an immutable characteristic of citizens; the situation is actually heavily contingent.
In the first few pages of The Myth of the Rational Voter, Bryan Caplan makes (implicitly) the case that voter ignorance isn’t a huge deal as long as errors are symmetric: ignorant voters on both sides of an issue will cancel each other out, and the election will be decided by informed voters who should be on the “right” side, in expectation. Caplan claims that systematic bias across the population results in “wrong” answers.
My point in bringing this up is just that the existence of large numbers of ignorant voters doesn’t have to be a major issue: large elections are decided by relatively small groups. Different voting procedures have very different ramifications for the composition of these small groups.
Hi Matt,
Of course I agree that studies of voter ignorance do not yield evidence of some fixed, immutable characteristic of citizens. It is also certainly true that different voting systems provide different incentives to acquire relevant political information. The pertinent question is whether implementing approval voting would incentivize the acquisition of political information to a sufficient degree that we could be confident in claiming that the mere transition to approval voting alone would have all the benefits that Aaron claims it would have. I am still very skeptical of those purported benefits being delivered. Even on minimally demanding normative accounts of what amount of knowledge voters ought to possess (models of retrospective voting, say), the opportunity cost of acquiring the relevant information is, for very many citizens, simply too high. I suspect this will be true even under a scheme of approval voting, and even if more voters were to think their votes matter more. For all we know, the transition to approval voting might simply increase voter turn-out by prompting citizens who previously abstained to go to the polls, now thinking that their votes matter; and since the sort of evidence about voter ignorance alluded to earlier suggests that non-voters are typically less informed than voters, we might end up in an even worse scenario as far as voter ignorance is concerned. Like you, I don’t have a particularly strong stance on this. This is clearly an empirical matter. Still, we shouldn’t rule this scenario out entirely.
In my initial comment, I mentioned that I think it is easy to overstate the significance of the demand-side of politics. This is partly because of the literature on epistemic democracy, which you allude to above in mentioning the so-called “Miracle of Aggregation” (i.e. symmetric errors cancelling each other out). Epistemic democrats would indeed dismiss my concerns about voter ignorance, for they claim that, under certain conditions, collective intelligence can emerge from large groups of individually ignorant agents. If that’s right, then the evidence from political psychology needn’t worry us. However—and with all due respect to the many theorists working on epistemic democracy (it’s nice to say “with all due respect” and mean it) - the claims about the collective intelligence of individually ignorant voters often strike me as pollyannish. To take the case of the miracle of aggregation, I don’t think the required symmetry holds. Without that symmetry, the collective intelligence does not emerge. To take another favored strategy of epistemic democrats (the Condorcet Jury Theorem), I don’t think the necessary conditions for the emergence of collective competence are satisfied.
However, another reason to downplay the significance of the demand-side of politics is worth mentioning, since it is often ignored by various parties to disputes about levels of voter ignorance. The degree to which political leaders respond to voter preferences is often greatly overstated. Policymakers and political representatives enjoy a significant degree of autonomy, and much—if not most—policy-making is done out of the public eye entirely. When you couple this with recent work in political psychology suggesting that voters are often simply happy to toe the party line, this suggests that much of the focus on voters, their knowledge, and politicians responding to their preferences has been misplaced. This is as true of epistocratic worries about voter ignorance as it is about the epistemic democrats who oppose them. I don’t think this means that voter ignorance isn’t a problem (assuming for the moment that it is indeed a problem), but it’s not the only thing we need to be looking at. Many factors other than voter competence contribute to the overall quality of governance, and voting reforms (whether approval voting or otherwise) will not change such factors much, if at all (factors such as the degree to which policymaking can be captured by special interests, levels of corruption, the decision-making methods deployed by legislators and bureaucrats, etc.).
Actually, upon re-reading my comment, I see that I somewhat inaccurately represented my own views. Instead of:
“In general, I think the case for approval voting outlined focuses too much on the supply-side of politics and not enough on the demand-side.”
I should have said that there is a combination of focusing too much on the supply-side of politics and overlooking problems on the demand-side.
Replied at the bottom by accident. Starts with, “Hi, Adam.”
Hey Aaron! Thanks for posting this. I am likely going to include CES in my giving this year as a result of some of the points you’ve made here.
I’ve been researching lobbying recently and I’m curious about this passage:
I’m not particularly skeptical about this one-sixth estimate, but I haven’t been able to find anything like it my lit review! Do you have some background on this research?
Hi, Matt! Thanks for the question.
This estimate is based off of me asking our Director of Campaigns and Advocacy about likelihood of success with lobbying. He gave me an estimate and then I did a correction based off what I’d seen, which lowered the success probability to 5% per effort. The expected cost per person (factoring in this probability) is much lower because when it does work, you can get an entire state to change their voting method. The scale counteracts the cost and probability rather quickly.
That said, lobbying is a challenging effort. I’m cautious about spending too much time and money here without having a 501(c)4 as it can push away our resources from wins that are more likely. We have a need for momentum at the moment. Still, with the potential cost effectiveness, this is an area we’re considering in the future. It’s likely we’ll be risk averse initially as we gain experience.
If you have particular resources you’ve found on lobbying, feel free to share. Like I said, this would be a new space for us. We had one opportunity here recently, but we decided against it based on the particular circumstances.
I’ll post a summary lit review here on the forum when I’m done with my research. Spoiler alert: political scientists don’t have a great idea of how/why/whether lobbying works and research on its effectiveness is almost strictly limited to trade policy and large publicly traded firms. So you get expressions of effects like “$140 in additional shareholder value for every $1 spent on lobbying.” Interesting, but not particularly generalizable.
It seems like CES’s strategy so far has been to start small, which makes obvious sense. I’m curious to know when/if you make the decision to withdraw from a local advocacy effort that seems like it’s not paying off. It’s not obvious to me that public support is monotonically increasing in dollars spent on advocacy— what’s your stopping rule?
Right now, initiatives are our main path, but with more funds we’d be open to experimenting in lower-risk scenarios. I’d be excited to read what information you come up with.
We use polling as an initial indicator to see how receptive a demographic would be to the initiative. Fortunately, the simplicity helps us with receptivity.
Within a particular target city, we don’t want to add more dollars than necessary to an effort. Winning soundly is important, but we didn’t need to throw $1M directly into Fargo, for example. I suppose if we had, the support may have gone nonmonotinic in relation to the spend and backlashed against us. But I don’t see that as a particularly big risk for us. We’re more efficient than that.
Also, being as early as we are in the game, we’re a little cautious about taking on a city we don’t think we can win. We’re aggressive, but not more aggressive than we think we can get away with.
Hi Adam,
Q: How would we know approval voting would give better policies? Why would the policies be good? And how do we know the policies would be good rather than merely popular?
A: I’m combining these ideas I pulled out because of their similarity. Approval voting tends to pull out the middle viewpoint (whatever that is for a particular electorate). And because viability is not an issue to gain initial support, it can provide a ramp for new ideas.
Is it possible that the popular opinion is bad? It sure is. But for this to be a real worry under approval voting would mean that the alternative of what we have now as being better. We might find that there is some popular issue that people win on that is not correct or overall good. This is also possible now, but with a poorer voting method. The question is whether approval voting provides a net gain so that popular issues that are good actually move forward at a pace faster than they would otherwise. Because of the quicker feedback and approval voting doing a better job of capturing candidate support, I believe this to likely be true. And to the degree that this is true in enacting better policies, many policies have carryover benefits that go well into the future.
I’m not sure I understood the question on supply and demand-side politics.
P.S. I accidentally added this as a main thread reply. Sorry about that.
Has CES done any analysis of Causus voting? I did not see any reference to it on the CES website.
With the primaries coming up, this might be interesting to explore. A recent On Point podcast interviewed Iowa voters and the process sounded somewhat similar, in that people need to compromise and come to a consensus. I am not suggesting caucus voting as a reform, but I wonder if some analysis could flesh out how it overlaps or differs with other systems like approval voting. If similar, that might provide some comfort to voters that approval voting is not so radical vs. existing systems.
https://podcasts.apple.com/us/podcast/the-on-point-iowa-caucus-where-do-voters-stand/id121534955?i=1000462627942
Caucus voting still has vote splitting as voters aren’t able to support multiple candidates simultaneously. With approval voting, you can support multiple candidates simultaneously. We haven’t analyzed caucus voting. We did do this poll, however, on the democratic primaries: https://www.electionscience.org/press-releases/new-poll-74-of-democratic-primary-voters-would-support-warren-for-president/
I’d like to see us do much more research and evaluation, but we currently don’t have it in our budget to hire a Director of Research and support staff.
Aaron, thanks much for the update. Did the poll include non-democratic primary voters—e.g. registered republicans, etc.? A broader view of citizen views might offer additional insights.
I agree that more research and evaluation is desirable. In fact, I think that robust evidence is critical for concepts like approval voting to gain traction and support. Could CES organize volunteers to do exit interviews at polling locations during primary and general elections? Capturing plurality, approval, and even other voting preferences during real elections could offer valuable insights. It would also raise the visibility and awareness of different voting systems. And it might also help citizens understand how different systems can affect elections.
The exit polling could also ask people if they would optionally like to receive results (SMS, email.., [not tied to their vote choices]), thus additionally increasing awareness and offering the potential for future polls.
The poll included only those who intended to vote in the Democratic Primary.
It’s very difficult to manage volunteers in this way, particularly given our small staff size. We tend to contract polling out. That said, it takes some expertise to sort through the data. Having staff for research would help us dramatically in both evaluating voting methods and measuring progress within cities that we’ve won in.