« How will the financial crisis affect the economics profession? | Main | The story of Gary Gorton and AIG »
Voting Videos
Here's a great little video from PBS (!) featuring Gordon Tullock on why he doesn't vote and why you shouldn't either. (Andrew Gelman and Noah Kaplan beg to differ in this article, but their theory applies only to altruists - not to Gordon!).
And from The Teaching Company here is a free video on voting theory, i.e. Arrow's theorem, the Borda count and all that other good stuff.
Posted by Alex Tabarrok on November 3, 2008 at 10:20 AM in The Arts | Permalink
Comments
Great video, but I wonder what effect that mindset has on outcomes. For instance, I've heard this idea expressed by quite a few libertarians and anectdotally, I believe they're less likely to vote. Does this mean that we have less libertarian influence in our government because libertarians (or any other group) tend to vote less than average?
Posted by: Team John L at Nov 3, 2008 11:31:37 AM
Before seeing this, if I had to assign a probability to PBS making a well-made video explaining the rational voting model with Gordon Tullock, then I would say that would be roughly equal to the probability of my voting being decisive.
Wow. Kudos PBS.
Posted by: Scott Wentland at Nov 3, 2008 11:48:57 AM
I've been asking the Teaching Company for several years now to run a course on Game Theory. This is a brand new offering and I just ordered the DVD. Thanks for the notice!
Posted by: Bob Kozman at Nov 3, 2008 12:09:21 PM
If you care about other people, then even a very small chance of making a difference in the lives of million of people can make it rational to vote.
However, if you think people should only act in their own narrow self-interest, then it is probably irrational to vote.
Corollary: Does that mean that it is irrational or hypocritical for most libertarians to vote?
Posted by: a student of economics at Nov 3, 2008 12:33:55 PM
My rational argument in favor of voting is as follows: While I have an insignificant chance of being the deciding vote in favor of the presidential candidate, I have a small but measurable chance of deciding the local city council election (1300 people or so in my district). I do have a preference in that election, and it's worth my time to walk 2 minutes down to the polling station to vote in that race. While I'm there, the extra effort to register my preferences in other elections is practically zero (measured in fractions of a second), so I might as well. If the presidential election were the only race on the ballot, I wouldn't vote either.
Posted by: C L at Nov 3, 2008 12:52:15 PM
As Tullock himself says in the video, there are other "benefits" to voting besides casting a deciding ballot.
I maximize my preference function when I vote, but it ain't cause of instrumentality.
Posted by: Jeff H. at Nov 3, 2008 12:58:17 PM
We have the worst prospects for government in my memory, and I am the least motivated to vote ever. One would think a good system would have feedback loops that cause bad government to be followed by better government. I don't think this is the case with the way our current system operates. Bad government motivates the extremists. So, a poorly run statist administration is likely to be followed by a very vigorous leftist regime.
There is another problem. People think their vote "helps" and they are wrong.
Posted by: Andrew at Nov 3, 2008 1:03:56 PM
I vote because it's fun, not because I think my vote has any influence.
Posted by: Andy at Nov 3, 2008 1:13:41 PM
libertarians care about other people, student.
Posted by: josh at Nov 3, 2008 1:28:37 PM
If the argument is that a single vote doesn't have a material impact, and we’re agreed that election outcomes matter, surely the logical decision is to find ways to influence large groups of voters. For example, join or form a political party, create non-partisan campaigns around key issues, start a blog, etc. There’s ample evidence that coordinated efforts can shape election outcomes.
Yes, you end up with a consensus candidate, but at least you can ensure that some of your interests are represented. Choosing not to vote only indicates laziness.
Posted by: In Check at Nov 3, 2008 1:54:47 PM
Politicians care about other people too, but they do it with my money.
http://finance.yahoo.com/banking-budgeting/article/106065/Special-Coverage:-Race-to-the-White-House
Posted by: Andrew at Nov 3, 2008 1:55:46 PM
Student - even if you're altruistic, it doesn't change the outcome. The model is simple: vote iff pV>C, where V is how much you'd be willing to pay for your candidate to win, p is the probability, and C is the total cost of voting (incl. opportunity costs). Gelman, in another paper, estimated the p at b/w 0.5x10^(-7) and 1.5x10^(-7). So let's say the opportunity cost of your time is a penny, and you take the higher p. Voting implies $66,667. That should include any of your altruism. The higher the opportunity cost, which is obviously the case, the larger V has to be to make sense. Of course, I think most people vote on the absolute value of V, and not on the expected value of V, which is why this always seems so weird to people.
None of this gets at the real policy goal, though. We want people to make informed votes, not simply vote. What is the right policy solution to being an ignorant voter? And is it right to see voting as a collective action problem? Can we tax people to become informed voters (assuming monitoring costs were zero), or does that not really address the source of this failure?
Posted by: jason voorhees at Nov 3, 2008 1:56:25 PM
Also, one more thing. It's a mixed strategy nash equilibrium to vote. I think that was what Tullock was getting at towards the end when he said he'd vote when no one else did. Of course, the point is that in the real world, since those probabilities are so tiny, that is likely not what is motivating people either.
Posted by: jason voorhees at Nov 3, 2008 1:58:15 PM
Josh: I did hedge a bit by saying "most". My reasoning is that many libertarians (at least in my experience) justify their views by appealing to Ayn Rand-style arguments about the virtue of selfishness. For instance, on this blog, I occasionally hear such views.
Jason: It's not clear that $67K is an unreasonably large value for V. There are millions of people who volunteer to risk their lives in Iraq, or who devote their lives to helping the disabled, or who make other choices with costs much greater than $67K because they think they are helping others. It seems plausible that the difference between candidate x and candidate y will have overall impacts greater than these other volunteer choices.
Posted by: a student of economics at Nov 3, 2008 2:19:52 PM
Student - No, it's not. BUt one penny for your opportunity cost is an unreasonably small value. My point was to make it as generous as possible. If you set the opportunity cost of your time equal to $5 - the price of a matinee movie ticket - then redo the math. It comes out to $33 million. And that is several orders of magnitudes higher than what I've seen from the risk literature on the subjective value of life.
Point being. Realistic values for p and C are going to cause V to get huge. Too huge to really make sense.
Posted by: jason voorhees at Nov 3, 2008 2:41:07 PM
When I vote the nice ladies give me a cookie! Self-interest resolved.
Posted by: odogrph at Nov 3, 2008 4:09:04 PM
Does this mean that we have less libertarian influence in our government because libertarians (or any other group) tend to vote less than average?
Yes but we also have happier libertarians. They might even do something useful like criticize Tyler's latest post, rather than wasting time in a meaningless process.
Posted by: Bob Murphy at Nov 3, 2008 4:10:13 PM
Gelman and Kaplan need to explain why they aren't using Brennan and Lomasky's formulae.
Posted by: Jason at Nov 3, 2008 4:37:56 PM
The side effects of groups of people not voting is an externality. We often internalize it by making it a moral issue.
Posted by: Jorge Landivar at Nov 3, 2008 4:55:58 PM
The rational voter model is wrong. It's expected that economists will try to save it but it's wrong. ( I'm an economist)
Tullock is right, people vote because they like voting. Why? because it's in our genes.
Posted by: Bruno at Nov 3, 2008 5:31:16 PM
How is voting helpful to the rest of society? How is not voting an insular activity? I do not understand the logic behind those assumptions. If we all agree that politics, and ultimately our decision to vote, is the culmination to an irrational acceptance of some bizarre political narrative, specifically designed to aid the cause and career of one candidate and his ideas, ideas that could be possibly cataclysmal, then how can we ALWAYS say that voting is a selfless "good" act. In many ways it could be an irrational sanctioning of awful acts.
Posted by: John Pertz at Nov 3, 2008 5:56:07 PM
One should also consider that if you are part of a demographic that regularly votes (say those collecting social security) instead of a demographic that historically does not vote (college kids), then you are more likely to gain some benefit from the elected official even when your vote doesn't put the candidate in office.
Posted by: David S at Nov 3, 2008 6:02:15 PM
Tyler/Alex - Take a look at Richard Tuck's latest book on "Free Riding" that addresses the paradox of voting as a canonical example. There is an excellent review at the London Review of Books:
http://www.lrb.co.uk/v30/n19/runc01_.html
The review itself is quite dense, but it is a subtle argument and definitely worth reading.
Posted by: kxm at Nov 3, 2008 7:32:39 PM
Jason Vorhees: Don't forget that your analysis ignores the fact that tomorrow some people will die in vehicle accidents, because they decided to go vote.
Posted by: Jay at Nov 3, 2008 7:41:52 PM
Didn't we just have this conversation a few days ago?
http://www.marginalrevolution.com/marginalrevolution/2008/10/why-vote.html#comments
I referenced Tullock - his logic breaks down because he has no idea how many people will go out and vote. Therefore he cannot accurately say how much weight his vote has had until after the fact.
Yes, we all know that in the real world the # of voters will be >2, so his vote will more than likely be diminished by some orders of magnitude. However, look what happened in Florida in 2000. Nonvoters in Florida had a chance to influence 8 years of history for millions - more like billions - of people.
Posted by: meter at Nov 3, 2008 8:36:11 PM
