Gallowglass

Strategic and sincere voting

A small debate has been going on about whether lefties should have voted for Nader or Gore in the comments section to this Patrick Nielsen Hayden post. On the one hand, Darren Madigan says

I will continue to live on dream time and cast my vote for whoever I personally think will make the best President. Just like I did in Florida in 2000 for Ralph Nader.

On the other, Xopher argues

I cannot understand how any intelligent person could possibly believe Nader was sincere in wanting to shift the country to the left. His actions had the opposite effect, and he’s a smart enough man to know what effect they would have.

This illustrates nicely the distinction that political scientists draw between “sincere” voting and “strategic” voting. And it’s an important distinction. When someone votes, they presumably have a preferred end-result in mind; they would prefer to see candidate x rather than candidate y or candidate z win the election. Simplifying wildly, a “sincere” voter is somebody who will vote for his preferred candidate or outcome, x regardless of what everyone else who is voting does. A “strategic” voter will instead try to weigh up what everyone else is likely to do before she makes her final decision.

So far, so abstract. What does this mean in practice? Well, imagine that a hypothetical voter in the 2000 US presidential election knows that his vote counts, and faces three possible outcomes - either Nader wins (N), Gore wins (G), or Bush wins (B). He’s a leftist, so his ideal outcome is that Nader wins. However, if it’s a choice between Gore and Bush, he very definitely prefers Gore to win. His least preferred outcome is that Bush wins. In game theoretic parlance, the payoffs from N are greater than the payoffs from G, which are in turn greater than the payoffs from B.

Now, if he’s a sincere voter, what does he do? The answer is obvious - he votes for Nader (N). This is his preferred outcome in an ideal world, and he votes for it. Very straightforward.

If he’s a strategic voter, the question is much more complex. He tries to weigh up what everyone else is likely to do, before making his final decision. For example, if he knows that Nader voters are in a small minority, so that Nader has no chance of being elected, he will very likely vote for Gore instead, as G has a better payoff than B, which is his worst possible outcome.

This suggests that people like Madigan are motivated by a fundamentally different logic than people like Xopher. Those who voted for Nader are more likely to be “sincere voters” - that is, they see voting as a means of expressing their identity and political convictions, no matter what the outcome is. People like Xopher, in contrast, are “strategic” - they are interested in outcomes, and thus will vote for a second best candidate if this staves off the worst possible scenario: Bush becoming president. And indeed, this paper provides some interesting evidence to suggest that Gore and Nader understood the difference between strategic and sincere voting quite well. Gore made strategic appeals to likely Nader supporters, asking them to consider the likely alternative if he lost, while Nader asked people to vote their consciences, no matter what result this had.

But things get more complicated. It isn’t necessarily strategically irrational to vote for Nader rather than Gore, even if you know that this might mean that Gore will lose the election. Nader supporters may credibly worry that they will always lose out if they vote for more centrist candidates. Not only will Nader never win, but the mainstream of the Democratic party won’t ever take them seriously. It will know that it can always take their support for granted, because Naderites (a) know that they have no chance of winning themselves, and (b) will always prefer the Democratic candidate to the Republican. If a political party can count on your support, no matter what it does, it won’t ever need to respond to you, say by shifting its position to the left. Instead, it will always move towards the center to hoover up as many votes as possible (this is predicted by the so-called Median Voter Theorem).

Thus, it may be rational for strategic voters to vote for Nader, even if they know that this will result in a right wing President being elected, provided that they are more interested in the long term battle for control of the left wing agenda, than in the outcome of the specific election in question. They’re sending a costly signal - showing that they’re prepared to see the election lost rather than sacrifice their policy aims. Their presumed hope is that the mainstream Democratic party won’t be able to take them for granted in the future, and will thus have to run to the left.

Still, to say that it’s rational isn’t to say that it’s smart. I’m not American - I’m not even resident in the US - but it seems clear to me that the American left will lose enormously over the longer run if Bush is re-elected. Voting for Nader in 2004 won’t do anything to push American politics leftwards in the short term, and, given what’s happening to institutions such as the courts, will likely be profoundly unhelpful in the long term too. Digby’s take on this, which prompted this debate in the first place exaggerates a little. But not much. If I had the vote next time around, like Digby, I’d be a yellow dog Democrat - 2004 is going to be more about damage control than ideal outcomes.

Update - have changed the post slightly to make it clear that the Gore-Nader thing was not in Patrick’s post, but in the comments thereto.

Posted by Henry at June 26, 2003 06:34 PM | TrackBack