This is obviously a terrible abuse of posting privileges to promote something that really ought to be a comment on Harry’s piece below, but whatcha gonna do? I just wanted to add a small note on a technical issue to do with his conclusion about our civic responsibilities:
When you vote, you have a very stringent obligation to deliberate responsibly about the effects of your vote, and about whether those effects are morally justifiable or not. You should deliberate about the moral issues at stake in the elections, and come to have a tentative, but warranted view, about what justice requires, as well as about what the likely effects of policies your candidate is likely to implement (and whether they are morally justified).
That sounds like pretty hard work doesn’t it? However, luckily the Condorcet Jury Theorem comes to our rescue. More or less the same mathematics which ensure that voting is a waste of time also ensure that as long as the average voter has a slightly better than 50% chance of making the right decision and the electorate is large enough, the majority vote will be correct in a two horse race (like a Presidential election; voters in multiparty democracies, do what Harry says). It’s one of those seeming informational free lunches which are the basis of the James Surowiecki’s book.
So, the full advice to potential voters would be that your civic duty is:
1. If you are a reasonably intelligent and responsible citizen, just kind of think for it a bit and make a snap decision, like Malcolm Gladwell says and you’ll probably be right.
2. If you are voting for an essentially completely frivolous reason which has nothing to do with the actual election (like, for example, P Diddy threatened you with death if you didn’t, or you thought it might get you a shag, or you want to commemorate people who died a hundred years ago), then toss a coin; you won’t be bringing the average below 50%.
3. If you’re so stupid that you nearly always cock it up, then follow the Costanza Principle and do the opposite of what you think you should do. Actually, people like that probably can’t be trusted to follow the principle properly, so you lot flip a coin too.
4. If you’re reasonably intelligent, but also a selfish bastard, then stay at home.
So there you go. Voting isn’t actually quite as onerous a social duty as it would seem, at least in two-horse races, so go on, make Stone Cold Steve Austin proud. Or not, as the case may be.
{ 23 comments }
christian h. 03.07.08 at 5:52 pm
So how is flipping a coin different from not voting? In fact, it’s better not to vote than to flip a coin, as the latter method has a non-zero probability of messing things up, while the former doesn’t. Clearly the only reason to vote if you think you belong to one of the coin-flipper categories is if you somehow believe that it makes the system look better if many people vote and for some unfathomable reason you think that’s a good thing.
dsquared 03.07.08 at 6:00 pm
In fact, it’s better not to vote than to flip a coin, as the latter method has a non-zero probability of messing things up, while the former doesn’t
But flipping a coin also has an equal non-zero probability of un-messing things up (ie the case where the ex-you electorate had settled on the wrong choice and you were the swing voter and the coin landed heads) so it cancels out. Personally I agree with you; I’m just trying to make the people who “for some unfathomable reason think that’s a good thing” feel good about themselves, which appears to be something very important to them.
Patrick 03.07.08 at 7:13 pm
Though experiments:
What about honouring people who died last week? Say in the military?
What about voting to provide an example to your kids?
What about voting because you’re a HS civics teacher, and you’d feel like a hypocrite not doing so?
It would seem to me that if the Gladwell principle holds, all of the above would be enough to qualify someone to choose.
Or are all reasons not directly connected with your narrow view of the aims and purposes of voting “frivolous”? If that’s your point, who elected you to issue these pronouncements.
Total 03.07.08 at 7:23 pm
What Patrick said, albeit with a question mark at the end.
Righteous Bubba 03.07.08 at 7:24 pm
What about honouring people who died last week? Say in the military?
I haven’t yet figured out why a religious ritual should be involved. Or shouldn’t be: it might be fun to get wafers and wine involved. Count the uncannibalized portions of your candidate at the end of the day, subtract and there’s your number.
Write-ins don’t work so well.
dsquared 03.07.08 at 7:25 pm
a) if that’s your way of commemorating someone then I’m never going to criticise anyone’s reaction to bereavement but there really isn’t any civic vaue to it at all.
b) somewhat circular; the example you would be providing to your kids would be to encourage them to waste their time voting, unless you also gave them an example of considering the issues.
c) if it feels good do it, but don’t expect the rest of us to care.
Or are all reasons not directly connected with your narrow view of the aims and purposes of voting “frivolous�
basically yes, from a civic point of view. Why the heck should the rest of us get saddled with Gordon Brown just because some mugwump of a high school civics teacher thought that he ought to give an example to his class, eventhough he hadn’t actually thought about the election?
If that’s your point, who elected you to issue these pronouncements.
the great thing about issuing prononcements is that it isn’t actually an elected office (as you have implicitly recognised by issuing your own pronouncements, rather more pompously than I did). Tell you what though, if there actually was an election for the post of Official Pronouncer General of Crookedtimber.org, and it was a two horse race, you against me, with a sufficiently large electorate of voters slightly more than 50% likely to make the right decision, I bet I’d win.
Dan 03.07.08 at 7:38 pm
Actually, people like that probably can’t be trusted to follow the principle properly
I always wondered after that episode why George had stopped the do-the-opposite thing. My guess was either 1. he just wasn’t psychically able to keep up the effort of going against his true nature even despite its overwhelmingly positive effects on his life, or 2. he was doing the opposite so much that his impulses got confused and he couldn’t reasonably discern his opposite reactions from his initial impulses.
Total 03.07.08 at 9:06 pm
rather more pompously than I did
I think I’d disagree with that evaluation.
There’s a certain point at which erudite analyses of “civic virtue” became like reducing a sauce. If you’re not careful, everything scorches and a terrible smell fills the kitchen.
geo 03.07.08 at 9:55 pm
That sounds like pretty hard work doesn’t it?
Not in the US, it isn’t. Just vote against the Republicans, no matter who the individual candidates are. A Republican congressional majority is the ultimate disaster.
lemuel pitkin 03.07.08 at 10:55 pm
if there actually was an election for the post of Official Pronouncer General of Crookedtimber.org, and it was a two horse race, you against me, with a sufficiently large electorate of voters slightly more than 50% likely to make the right decision, I bet I’d win.
Best. Putdown. Ever.
Total 03.07.08 at 11:09 pm
Best. Putdown. Ever.
Yeah, it’s right up there with Churchill’s “A modest man, who has much to be modest about.”
Like the original post, the putdown was overwritten.
qb 03.08.08 at 12:28 am
i suppose the implication of Case 4 is that reasonably intelligent but selfish bastards tend to vote incorrectly and might thereby undermine Condorcet if there were enough of them? we are so totally screwed.
Andrew Bartlett 03.08.08 at 1:40 am
What is the ‘right decision’?
qb 03.08.08 at 2:30 am
andrew: i believe the theorem supposes the vote is to determine an objective matter of fact and not merely a procedure to aggregate preferences. hence “jury” theorem, where the objective is to determine guilt or innocence. obviously this cannot simply be plugged into a theory of voting for political representatives as is, but mutatis mutandis it seems suggestive for this case as well. the ‘right decision’ here is the one that puts the ‘best’ candidate in office on some fixed interpretation of ‘best.’
Scott Hughes 03.08.08 at 3:01 am
I consider myself reasonably intelligent, but I guess I’m also a self bastard because I do boycott the vote. :)
Anyway, that is a funny post!
Sortition 03.08.08 at 4:19 am
Daniel wrote:
For this to be true an additional crucial assumption must be made: the votes cannot be very highly correlated. If voters are prone to make mistakes, and make them in the same direction at the same time (because, for example, voters are swayed by clever TV ads and they all watch the same TV ads), then the errors do not cancel each other, and there is a good chance that the “wrong” candidate gets elected.
Thus, the difficult part is not getting to close to 100% chance of choosing the right candidate (50% + ε will do). The difficult part is making your mind in an independent way that is not manipulated in the same way that all the other voters are manipulated. Only then can you rely on the law of large numbers to do the job for you.
Tom Hurka 03.08.08 at 2:29 pm
Voting is only a waste of time if you’re motivated just by self-interest. If you care about other people, then although your vote has only a tiny chance of making a difference, it would, if it did, make an enormous difference — e.g. Gore rather than Bush might have saved tens of thousands of lives. And a small probability of an enormous difference can be worth going for.
Ray 03.08.08 at 4:33 pm
Buy a lottery ticket, and give your winnings to charity.
Patrick 03.08.08 at 6:42 pm
For a start, Katha Pollit on the Washington post “women are stupid column” is a much better putdown.
Dsqared, of course, you’d get more than half the votes if it were a two person field. For one thing, I’d vote for you over me in a heartbeat. Of course, I wouldn’t think too much about it in that heartbeat, but I bet we’d all count the vote.
I read the blog regularly. Comment too. It’s not just Berube. If it were a three person race, he’d probably win, you’d come second, and we could bar anyone who voted for me from ever voting (for anything) again.
I’m not sure about the pompous thing, though. And just so you know, it stung.
The point I was trying to make is that there are all sorts of non-frivolous reasons for voting, and to me it sounds like you’re defining “non-frivolous” in a way that sounds suspiciously like, “accords with my notion of seriousness, which is, by the way, the right notion.” As a veteran of the canon wars in literature, where I heard an English professor in 1975 claim that there were no “serious” literary artists in the United States and never had been, that sort of thing gets my back up.
And my concern is exactly this: I tend to think that people who will vote for Ron Paul or Ralph Nader aren’t being “serious.” But I don’t think I’m the one who gets to decide.
dan 03.08.08 at 8:24 pm
as you have implicitly recognised by issuing your own pronouncements
But he didn’t issue any pronouncements. He asked questions and made one qualified statement.
Jeff 03.09.08 at 10:14 pm
The Condorcet Jury Theorem does not apply in this context. It presupposes that the voters all share the same goals (e.g. convicting the guilty and acquitting the innocent). In that circumstance, if every individual is more likely to be right than wrong, then majority rule should be accurate for binary choices in arbitrarily large groups.
The Condorcet Jury Theorem is also undermined by strategic behavior — which, in this context, refers to rational individual use of available information. For the CJT to work, individuals have to vote according to their own beliefs about the truth. The theorem is then just a result of the law of large numbers.
Voters do not share identical preferences over outcomes in most elections. “More likely to be right than wrong” is meaningless when voters have genuine disagreements about what is “right.”
Moreover, voters are aware of what large numbers of other voters think. If I think that x is true, but I know that a majority of other voters think y is true, and I know that people are more likely to be right than wrong, then I may vote for y. This is perfectly rational from my perspective, but the CDT doesn’t work when individuals aggregate information this way before voting. At least, that’s my understanding of the recent work by Fedderson and Pesendorfer.
Martin Wisse 03.10.08 at 11:23 am
In the end, voting makes barely any difference, so what does it matter if you vote correctly or not, or not at all.
Michael Mouse 03.10.08 at 12:04 pm
if there actually was an election for the post of Official Pronouncer General of Crookedtimber.org, and it was a two horse race, you against me, with a sufficiently large electorate of voters slightly more than 50% likely to make the right decision, I bet I’d win.
But bet at what odds?
Obviously, we must set up a toy electoral prediction market. And then save ourselves the bother of actually running the election.
(I bet nobody bothers.)
Comments on this entry are closed.