I mentioned in an earlier thread that I've been voting for the Greens consistently since 2003 (both at the federal and provincial level). Prior to that, I rarely voted at all because I believed it was irrational. What lead this economist to change his mind?
- In one sense, voting is almost always an irrational act. Each one of us only has one vote and even a close race has a gap between the two candidates of hundreds or thousands of votes. Take the riding of London West, which my parents moved to in 1988. Since then there have been 7 elections with the following results:
- 1988 – Progressive Conservatives over Liberals by 5 071 votes.
- 1993 – Liberals over Progressive Conservatives by 15 915 votes.
- 1997 – Liberals over Progressive Conservatives by 13 752 votes.
- 2000 – Liberals over Canadian Alliance by 13 632 votes.
- 2004 – Liberals over Conservatives by 7 726 votes.
- 2006 – Liberals over Conservatives by 1 329 votes.
- 2008 – Conservatives over Liberals by 2 121 votes.
In the 2008 election, where the seat changed hands, the election was decided by over 2000 votes. Unless you really care if the Conservatives win by 2 122 votes (or 2 120 votes) instead of 2 121 votes, why bother voting? Sure, you don't know ex ante what the outcome of the election will be, but you can be certain it won't be decided by a single vote, so why bother?
And please don't suggest that you should vote because our grandfathers died for the democratic right to vote, etc. etc. If that's really the case (and it's always suggested without evidence), they were also dying for the democratic right to be a candidate. I've never understood why we have some sort of moral obligation to vote, but we don't have the moral obligation to also put our name on the ballot.
(An aside: One reason you might want to bother voting is the fact that by voting you provide your chosen party with a roughly $2/yr of funding thanks to the current funding system. That didn't exist when I was a non-voter, so I'm excluding it).
Since voting has an opportunity cost and since the chance of you affecting the outcome is near zero, voting must be irrational. Or at least for people who are very bad at math. That's how I used to think.
I had a change of heart around the year 2000, thanks, to of all things, a Montreal Expos game. We were in Montreal, all wearing our 'Spos uniforms and hats and watching the Expos blow another game to someone (I believe it was the Diamondbacks, but I don't fully recall). The umpire made a (in my view) particularly bad call, to which I instinctively yelled out "Keep your head in the game, Blue!".
A friend sitting next to me asked "Do you think that really helps?"
I responded "Of course not. But it makes me feel better."
Voting is a lot like going to a baseball game. When I went to an Expos game to cheer on Vladi and Vidro and Javy, when we yelled at the umpires and heckled Randy Johnson, we weren't doing it because we were expecting to change the outcome of the game. We were they to show support to our team, to have a good time and to be a part of something larger than ourselves.
Voting is only irrational if you're doing it for the sole purpose to change the outcome of the election. You can still vote and be a card carrying member of the rational actors club.
Worthwhile Canadian Initiative 
I have a feeling a lot of you will respond “well, yeah, duh”. Took me until the age of 23 to figure out!
I don’t think you’re there yet. The equivalent to yelling at the umpire would be getting on twitter and calling Tony Clement a bone-head, etc… It’s public, low cost and immediately gratifying. Voting is more like walking down to the field at the end of the game and trying to get the ump’s attention. We call people who do that crazy (or drunk).
I’m sure there’s a rational reason to vote, but I think it has to do more with some crazy prisoner’s dilemma type game than it does emotion.
I figure voting buys me the right to complain about the government later. If you don’t vote, you should not have the right to complain about what you get. If you do, you may have had teh correct idea, but everyone else was fooled, or you may have voted in someone you now hate and were fooled, but either way you have the right to complain.
I used to think it was irrational to vote, because the chance of being a decisive voter is so small. Then I read an economics(?) paper that showed the counter-argument.
I can’t remember the paper, or the precise details of the argument, but it went something like this: suppose there’s 1,000 voters. The probability that you will be decisive is 1/1,000 (say). But if you are decisive, your vote will affect the outcome of 1,000 people. Therefore, on an expected benefit analysis, the expected benefits of your voting are (1/1,000)x1,000x effects of outcome on representative agent. So it’s exactly the same as if you were a single voter voting for a policy that would affect you only.
Further, with a secret ballot you don’t know if your vote is one of the “deciding” votes or not.
I was a Poll Clerk last federal election. It really is an exercise in applied micro/macro transition.
The polls close at 9pm in Ontario. The Poll Clerk and Deputy Returning Officer then open the box and count the ballots. There is a huge amount of cross-checking of serial numbers and such to make sure that the vote is honest. We complete our sheets and hand them in to the Poll Supervisor. She then calls in the preliminary results to the Riding Office. I had to help our supervisor take the ballot boxes to our Riding Office and while there I came across a bank of computers. That’s where the party reps and the media sit to collect the individual poll information. They then use those computers to relay them to the CBC and CTV media centres, the party HQ’s, etc.
That’s how Peter Mansbridge can report those poll results so quickly, broken down by neighbourhood and locality. That’s the line from you making an X on your ballot to my hands and then back to your television. That’s how the magic happens.
Expanding on that. suppose for instance, that this election comes down to one vote, in one riding, that decides the election between a Liberal minority and Conservative majority. Now, the deciding vote would probably have to come from only from of the up-for-grabs riding, so let’s say you have a 1 in 10 million chance of being the deciding vote in the upcoming federal election. Now, let’s also say that we live in a world where the are actually significant policy differences between the Liberals and the Conservatives, such that the value of the difference that their governments would make is some crazy figure like $50 billion. So, in this ridiculous counter-factual, the expected benefit of voting is $5000.
Obviously, this doesn’t reflect the world we live in the slightest, but it’s not inconceivable that the value of your vote in a close riding might exceed your opportunity cost.
“I figure voting buys me the right to complain about the government later. If you don’t vote, you should not have the right to complain about what you get. If you do, you may have had teh correct idea, but everyone else was fooled, or you may have voted in someone you now hate and were fooled, but either way you have the right to complain.”
This assumes that democracy is some competition where we have to “beat” the irrational voters who support the “other guys” who have terrible platforms. But the people that you disagree with are just as convinced that they are doing the right thing as you are. Its only bad to you because others are forcing you to support programs you disagree with, but that somehow becomes something to be proud of when you are the winner. Why not support a move away from this winner take all system, instead of creating conflict we should be working together to support a system where people get to support the programs they believe in.
“Further, with a secret ballot you don’t know if your vote is one of the “deciding” votes or not. ”
There’s no such thing as a single ‘deciding’ vote if the person wins by a 2+ vote margin.
I understand in some countries, the poor tend not to vote. Therefore, politicians don’t pay much attention to the poor. It seems worthwhile voting if by doing so I help sustain a universal norm of voting.
RE: Jesse’s argument. I like George Carlin’s take:
“They say, ‘If you don’t vote, you have no right to complain,’ but where’s the logic in that? If you vote, and you elect dishonest, incompetent politicians, and they get into office and screw everything up, you are responsible for what they have done. You voted them in. You caused the problem. You have no right to complain. I, on the other hand, who did not vote – who did not even leave the house on Election Day – am in no way responsible for that these politicians have done and have every right to complain about the mess that you created.”
I was implying that regardless of wheter you voted for the person who gets in or not, sooner or later you are going to be disapointed. If you don’t, then you didn’t even try to take part in the process, and shouldn’t have the (social) right to complain about the results.
awesome
What if I dont support democracy as a political system, am I allowed to complain?
I was referring to the 2000 people who presumably make up the margin in your original example, Mike.
However as an exercise in information gathering, those 2000 deciding votes have no meaning outside of relation to all other votes cast. One more vote means one more piece of information available. You can’t have a 2000 vote winning margin without somebody to compare that against.
While’s I’m at it, there hasn’t been an acclamation in a federal election in Canada since the 1960’s.
I think the best argument for not voting is this: “How the hell should I know, better than the next person, which party would make the best government?” There does seem to be something irrational, not to mention arrogant, in each person thinking he knows better than the average person.
Doesn’t voting contradict Aumann’s thingummy theorem, anybody?
http://en.wikipedia.org/wiki/Aumann%27s_agreement_theorem
Voting is irrational on selfish grounds, since the probability of being decisive is so small.
Voting could be rational on altruistic utilitarian grounds, since the very small odds of being decisive are offset by the very large number of people affected by the outcome.
If I were a better judge than the median voter, I should vote. If I am a worse judge than the median voter, I should not vote. (nearly) half of all voters will be worse judges than the median, so should not vote.
Therefore, only one person should vote. QED. 😉
[Monty Python}
That’s ridiculous Nick.
We are a Democracy. We are all equally citizens of Canada. The law does not hold one person’s opinion any higher than another person’s when conducting an election for our representatives. Our interest in this country is presumed to be equal because we live here. (Or used to live here and will again, let’s not get technical.)
That one person’s interest was greater than someone else’s was an idea that was thrown out with the Reform Acts in the UK in the 19th Century. It was stabbed in the chest in 1832 and shot to death in 1918.
Determinant is right. The logical principle of the identity of indistinguishables applies to secret ballots. Even if the difference in votes is one, there is no such thing as the deciding vote.
It’s surely a good rule of thumb that if a widespread practice seems to be irrational, the fault is in the definition of rationality. Usually it is too narrow, and usually it is something to do with identity, as I’m sure you found out last time you were in a rush and offered someone $2 to move ahead of them in the supermarket queue.
My personal motivation is close to Jesse’s (although I’d also like to feel I can celebrate a victory if my party wins).
Some years ago I was at a polling station in the lobby of an apartment building. There was another identical apartment building next door, which shared the polling station. The voting rate was about twice as high in the building with the station. So I suspect there is a realistic cost model on fairly narrow economic grounds somewhere.
Let me expand on that. Suppose that the final tally is 1000 – 999. Which vote among the 1000 is the deciding vote? It is impossible to say.
OTOH, take votes in legislatures. They are decidedly not secret, and in the bargaining that goes on, the deciding vote can be quite valuable.
Determinant: But, suppose I decide that I am a worse judge than the average voter of what party is best. I decide not to vote, because my voting will reduce the probability that the best party will get elected. Morally, I have done the right thing by not voting.
But, rationally, half of any group of voters should believe they are worse judges than the average voter. Therefore, half of all voters should decide it is immoral for them to vote.
Therefore, if all people are rational and moral, all people except one will decide not to vote.
Sorry Nick, I don’t agree with any of your argument. With all due respect, it’s completely wrong.
First, there is NO such thing as being a “worse” or “better” judge of which party is best for all people. One individual can’t have that knowledge, it’s practically impossible.
You can only say which party you think is best and in your opinion is best for the country. Your judgment is just as valid as anybody else’s. It is only when we add up everyone’s opinion that we can say what the consensus is. In order to do that, we need everyone’s vote. That is what enables the process of discovery to happen. Elections are a process that enables us to discover a consensus.
Politics, like physics, rejects an absolute frame of reference, that is an objective, a priori knowable truth. My opnion isn’t yours, and my opinion cannot replace yours. You and I have different lives, different experiences, and different expectations. Both our our opinions are required for the political to find a consensus.
In politics all decisions are relative.
Or put another way, Mr. Perfect Average Voter doesn’t exist. He never did. He’s a vain thing, fondly imagined.
And he was probably a she anyway.
I’m not sure how rational it is, but I figure that by voting I’m increasing the likelihood that other people on my side will also vote, while by not voting I would be increasing the opposite likelihood. Since I am the mindless automaton of the gestalt, I can change the mindless gestalt by changing my actions as its automaton.
@Determinant
Does this represent the ideal voter or the common, typical, less patriotic voter:
“You can only say which party you think is best and in your opinion is best for the country.” as opposed to “you think is best and in your opinion is best for you and your family”.
Odd not to see any references to GWB’s first election where this grand exercise of voting was put aside for the consideration of a few judges…who were prolly thinking “why vote?” too.
Nothing like that happens in Canada maybe.
More often than not, the importance of this grand exercise is to ensure that you have some control over your representative if he or she “under performs” –one threatens them with “voting them out”, yes?
So one should look upon those recent turnouts as an indication that Canadian politicians are tolerable. Or maybe the electorate is just that tolerant they’ll accept just about anything.
Your family, your country. It doesn’t matter, you deal with both in the same act. Whichever a voter chooses to emphasize is their choice.
Secondly, yes, as this is a Canadian blog I can happily ignore GWB. It doesn’t work like that up here. We use paper ballots with 150 people to a poll, counted manually. None of this electing everyone down to the dogcatcher, just the candidates to be the Member of Parliament for your riding.
After being a Poll Clerk last time I can say that we have a very, very efficient system with lots of checks for errors, very simple and clear rules (and ballots!) and a culture among poll staff that bends over backwards within the rules to help people vote. They are easy rules to follow and the training is good, clear and there is always the training book and supervisor on hand to help you if you get into difficulty
We vote for MP’s, the leader of the party that commands the majority gets to be Prime Minister and name a Cabinet.
I won’t bore you with details on what happens when no party has a majority in the House of Commons but the Prime Minister has to maintain the confidence of the House through passage of spending bills and winning confidence motions to stay in power. That didn’t happen a few weeks ago which is why we have a federal election right now.
Up here with four parties in the House it’s also “which party do I like best to represent me?” or “which party can extract the most favours from the governing coalition for my riding or province?”
@Mike: Your argument originates with Brennan and Buchanan, somewhere around 1982. Then taken up by Brennan and Lomasky (1986). Culminating in Caplan’s stuff on rational irrationality. Changing the vote from instrumental to expressive keeps it individually rational, but we lose all the good normative properties of median or mean voter equilibria. I love your bit on duty to vote vs duty to be a candidate though – don’t think I’ve seen it before.
I’m a conscientious non-voter.
Nick: Your argument requires altruism and Aumann’s theorem (also sufficient conditions for a representative agent economy). But Aumann’s theorem also requires identical priors, the absurdity of which has been has been argued here at length (at least by me :-). Anyways, despite all that, I fully agree with you. People who don’t vote like me are ignorant egotistical jerks. And I’m hereby letting you all know that since I plan to vote on May 2, you won’t have to. You’re welcome.
Nick,
What if I’m significantly worse than the median voter at evaluating my ability to vote? Should I flip a coin, and then vote if it’s heads, and stay home if it’s tails?
This might be slightly tangential to the rationality/irrationality of voting. Even if you think your vote doesnt count for much but you support the idea of democracy then you should go out to vote. I think this skirts the whole issue of the negative aspects of a democratic system as a system of conflict resolution.
I dont vote because I dont believe a majority has the right to impose their beliefs on any minority, including the belief that democracy is the best system of conflict resolution. The idea that we enter into a competition where the winner gets to dictate the spending priorities of all Canadians ignores possibilities of spending discrimination that would reduce conflict in the long run.
For example, assume that we have an election where the main election issue is the war in Afghanistan. We could have a vote, where the winner dictates the spending priority, forcing the losing side to follow the outcome of the vote, or we could simply say those that want to support the war in Afghanistan can send their money to the Canadian Expeditionary Force that is over in the middle east and those that are against the war can spend their money on the programs they support. The inevitable result of this setup is that no one would support the war in Afghanistan because it is very unlikely that the supporters of the war in Afghanistan really want to bear the full cost of the war. The people that currently support the war in Afghanistan do not support the war in Afghanistan, they support a war in Afghanistan that is being supported by other people. They require a monopoly on spending to have their demands for a war in Afghanistan met.
So why do these types of people support democracy? If they are only winning this vote on average 50% of the time wouldnt they still end up paying the same amount, getting subsidized when they win but having to pay more when they lose? I believe that they enter into this democratic contest because when they enter into democracy with other like minded monopolists they can gain the required surplus if the vote between monopolists is being subsidized by a third party, namely nonvoters.
This is why I dont vote, I dont believe that I have the right to subsidize my preferred spending priorites at the expense of others. Even if you think nonvoters are lazy, ignorant or irrational, coercing the ignorant is not a legitimate moral principle to base a political system on.
I don’t vote anymore. I broke my philosophical commitment in a recent local election and voted, but on reflection realized, with embarrassment, that it was uninformed and purely emotional decision, and that I ended up voting for someone I disliked the least. Looking back, that was a waste of my time.
I think the mathematical arguments in favor of voting are overblown, as are some of the anti-voting math arguments. Ultimately, I don’t vote because it is a choice, and my choice should reflect substantial support or agreement. I don’t eat vegetables that I slightly like but mostly dislike, and I don’t wear clothes that are two sizes too small or too large. I do not believe that the party victorious in this election will be any different from what the past few years have offered, just as the CPC has effectively repeated the Martin/Chretien years, scandals included.
Nick,
The paper you have in mind is probably Edlin, Gelman, and Kaplan, “Voting as a Rational Choice”:
http://www.stat.columbia.edu/~gelman/research/published/rational_final6.pdf
Here is a blog post by Gelman explaining the idea:
http://www.stat.columbia.edu/~cook/movabletype/archives/2004/10/why_its_rationa.html
I don’t buy it.
Vladimir: Yes. That looks like the paper I was thinking of.
“Each one of us only has one vote and even a close race has a gap between the two candidates of hundreds or thousands of votes.”
Tell that to “landslide Annie,” who was known for winning elections by single digits. Until she eventually lost.
At any rate, the problem with opting to stay home is that each individual choice impacts a much larger whole. I’m pretty sure you economists have a term for this.
I have been wondering if there’s any research about the theoretical vote distribution of those who stay home. Presumably some stay home because they really don’t care, but particularly in “safe” ridings, I’ve been wondering if the people who vote are representative of the whole electorate. Here in Alberta, do Cons stay home in greater numbers because they’ll win anyway; or do ABCs (Anyone But Conservatives) stay home in greater numbers because the whole affair is just so discouraging.
I keep voting because I hope that given some an encouraging “close” election, maybe more of my fellow ABC voters will turn up next time.
Voting is one part of a much longer process–the maintenance and nurturing of democratic institutions–that are inevitably dynamic and interactive, even if that means just screaming at the radio.
Irrational voting can be viewed as a social ritual that signals a firm commitment to a particular set of resource allocation rules.
To me this is proof that rationality is often a bad motivator for activity. (Dan Ariely has a neat study about people who are primed to be rational and then shown a story of some needing help are less charitable, not more.)
You are assuming that your decision vote is uncorrelated with others. It is not. There are huge collective effects that do matter. If someone feels that voting doesn’t matter then what? Others are likely to also. An engaged population is less likely to be hoodwinked (assuming the engagement doesn’t get too cultish.) and more likely to insist on good government. Their feelings (apathy or engagement) may extend to other forms of civic engagement.
Also, voting at least reminds people every few years to remember what they want in a society. Do you want lower taxes and less government? Do you want more government and more taxes? Do you want the tooth fairy to provide more government and less tax?
All these issues matter: cheeky reductionist arguments that voting (this goes for you too, Nick) are simply the college-educated version of “all politicians are the same; they are in it for themselves; it doesn’t matter.”
If there were a rational argument to NOT vote, then only irrational people would vote, leading to a government that is itself irrational.
Sadly, it may be more than just a hypothetical.