Wednesday, November 3, 2010

Occam's Elections

On election day, two worlds collide: the tiny, Plutonian rock made up of people who care about issues, politics, policy, parties and what happens in Washington smacks straight into the Jupiterian gas giant of actual voters who just don't know that much, and whose behavior is often fairly straightforward to predict.

The first dirty little secret of campaigns is that campaigns don't actually matter. By extension, all the political posturing surrounding them - "should we stage a vote on this bill, in order to force the other party to defend it in November? Is the election a big defeat for Democrats, a big victory for Republicans, or both?" - is also meaningless. This renders the lives and occupations of thousands of influential pundits, politicians, consultants and staffers irrelevant - which is why you never hear this point of view in the media. The media is heavily invested in the idea that political coverage is "serious" and so they take great pride in it, but the reality is that nobody watches it (look at FOX or MSNBC's ratings if you disagree - FOX was the 5th most-watched cable network a few weeks before an election; MSNBC, the second most popular news network, came in 24th, 12 places behind SYFY).

Sure, campaigns matter in individual races - Angle lost in NV even though a standard-issue Republican almost certainly would have won, and I'm sure you can find instances of a particularly well-run campaign winning a close race that they should have lost. But as Matt Yglesias points out, American elections (particularly Presidentials, but also the broad outcomes of midterms (instead of individual matchups)) are really pretty predictable. Typically, even if your staff is pretty smart and your tactics shrewd, the other side has a pretty smart staff and pretty shrewd tactics as well.

Think of it like baseball's Value Over Replacement Player (VORP): in a competition, all that matters is the difference between you and your opponent. At the highest levels, both of you are pretty formidable competitors, so most of your individual greatness cancels out. So the significance of, say, that amazing get-out-the-vote drive you're so proud of isn't that you registered 2 million voters; it's that you registered 100,000 more than would have otherwise done so.

The second dirty little secret of campaigns is that most voters don't know much. Voters, in the aggregate, are morons. Consider that 2/3 of voters thought either bankers or the GOP were responsible for the sorry shape of the economy, and yet Republicans won more House seats than they have since at least the 1940s. Those who said bankers were to blame voted 11% more for Republicans - the party of, by and for bankers. In California, Democrats ran the table by more than 10 points in every major statewide race - except Attorney General, where Steve Cooley and Kamala Harris are less than 1% apart. That means there had to have been hundreds of thousands of voters who voted largely Democratic, except for Cooley - a strident conservative. This doesn't make any sense at all, unless you assume that voters are stupid.

This might sound harsh, but it's been proven to be true in every single election ever: here's some more recent evidence. California's proposition system is also submitted for your consideration in this regard.

Most voters (whether or not they know it) are low-information, whose voting behavior is highly predictable in the aggregate and therefore more or less entirely independent of anything Democrats or Republicans actually do. That's why the second half of last week's This American Life is so painful to listen to - this guy is getting so worked up over the kind of small-bore tactics that just don't matter, and an election that was essentially decided the day after Obama was elected. Keep that in mind as you watch pundits, Republicans and conservative Democrats draw "lessons" from the results of last night.

EDIT: Yglesias feels me.

No comments: