02 November 2006

Of Polls And Men

According to the polls, the Democrats have already won the mid-term election. All in all, there are 13 House seats that are already lost to the Republicans, and 20 seats more that are all-but-lost. At this point, holding the Democrats to a 20 point gain would be a moral victory. According to the same polls, the Democrats are likely to take the Senate, although that is not locked in place.

The narrative of this election has already been locked down, and nothing -- not John Kerry, not capturing Osama, not a stirred up Republican base -- is going to change it at this late date.

If the polls are right.

Now, I'm not one of those conservatives who believes that the fix is in for the Democrats in the polls. I don't believe that the polling companies or the media companies they work for are trying to present a story of Democratic dominance in order to keep conservatives at home. But I do believe that polls tacitly favor Democrats by one to two points, and that they are becoming less reliable over time for reasons that have nothing to do with the pollster's political bias. If the Republicans win on Tuesday by keeping control of both houses, the age of polls will have passed. That would make Tuesday a great day in history.

Polls have been bad for politics. The belief that every day brings immediate feedback, that every little decision is accurately judged as it happens against public opinion, is debilitating to representative government. Democracy is, at its heart, a magic show; the belief that the opinion of the mass of mankind, agglomerated, contains wisdom is based on faith, not reason. The moral case for democracy is strong but not so strong that it trumps the moral case for good government.* The wisdom and skill that the people bring to elections is not in knowing which policy will work best, but in judging which candidate is a better person. This has been particularly important in presidential elections, where we simply don't know what issue will suddenly become most pressing over the next four years. Choosing the person we most trust to face that issue, and craft the response, is what the people are best at.
_______________________________________________
* But what about Jefferson, you ask. Jefferson's "consent of the governed" was given to the government as a whole, and is the opposite of rebellion. The form of government need not be democratic, so long as it has the actual consent of the governed.

4 comments:

Bret said...

"the belief that the opinion of the mass of mankind, agglomerated, contains wisdom is based on faith, not reason"

There's a fair amount of empirical evidence regarding the Wisdom of Crowds so I disagree that it's not based at all on reason.

Brit said...

This has been particularly important in presidential elections, where we simply don't know what issue will suddenly become most pressing over the next four years. Choosing the person we most trust to face that issue, and craft the response, is what the people are best at.

Very, very true.

It's the same in the UK. British politicians have now openly abandoned any pretence of running on policy. Blair started the process by killing ideological politics of left and right with the invention of New Labour; and already David Cameron has taken it to its logical conclusion by explicitly stating that he is focussed on 'direction' (ie. image and personality), rather than policy.

It also makes the famous American refusal to vote for an atheist understandable, but disturbing.

One problem with this system is that the politicians are very aware of it, so you get sickening negative personal campaigning, and excruciatingly pathetic attempts to manipulate image. Eg, all British politicians have to be both 'traditional' in their values, but also 'cool' and 'down with the kids'. And US presidential candidates all have to pretend to be bona fide God-botherers.

Another problem is that the 'choice' for voters is so limited, and is determined by inscrutable party machinations.

I know as an American engaged in the day-to-day battle of moonbats v conservatives you can't possibly admit it now, but surely Bush v Kerry was a 'choice' like "do you want a punch in the face or the kidneys?" is a choice.

David said...

Bret: I, too, am a big wisdom of crowds/efficient market kind of guy. An efficient market, in the absence of transaction costs (two ways of saying the same thing) can treat a fact known to one participant as if it were known to all participants. But voting is not an efficient market, and the transaction costs in sufficiently educating the voters are immense. What voters get right is what a majority of people can get right. More or less, it's which guy would you trust with something important that you can't foresee.

House elections are a little different, because with a two year term you've got a better chance of knowing what's going to be important, and a quicker opportunity to replace a candidate that turns out badly. Two year terms are like a reduction in transaction costs.

But some problems simply can't be given to the electorate. Global warming is a good example. I consider myself (you'll never have guessed) to be smarter than the average bear, but after really trying to come to terms with, for example, the hockey stick controversy, I've just had to throw up my hands. Forgive my lack of humility, but if I can't figure it out, my working assumption is that the average voter isn't going to do much better.

Brit: I'm satisfied with the choices we get. Sometimes we get elections that are absolutely clear: Reagan v. Mondale is a good example. But we hardly ever (actually, I can't think of one) get an election between two superlative candidates. That suits me as I'm not a man-on-a-white-horse can of guy.

David said...

Bret: OJ is providing us with a really nice test of the limits of polling v. the wisdom of crowds. In his election prognostication contest, the consensus is that the Rs hold on to both houses easily.

Of course, his audience skews right wing and a pollster would object that not every crowd is wise. A crowd assembled to go to a Barry Manilow concert might overestimate Mr. Manilow's contributions to music.