September 12th, 2004
As I write this, the election is just 52 days away. The polls all say that President Bush is leading John Kerry by a significant margin. Political commentators urge Kerry to get moving and close the gap. There’s a widespread feeling that Bush is pulling ahead of Kerry.
Balderdash and falderol! Kerry has the election locked up; I expect him to win by something short of a landslide. The reason for my bucking the common wisdom is that polls don’t elect a President; elections do. And there’s a big difference between answering the phone & stating preferences and actually going to the polling station to cast your vote. This country has a voting participation rate averaging about 50%. This election will be decided by participation rates, not polls.
Let’s walk through some numbers. Suppose that 50 million potential votes support Bush and only 46 million support Kerry. Suppose further that, in keeping with past trends, only 50% of Bush’s supporters actually cast their votes, but that an additional 10% of Kerry’s supporters cast their votes. The final result will be 25 million votes for Bush and 27.6 million votes for Kerry – a decisive victory.
How do we determine relative voter participation rates, then? It’s pretty hard to predict how many actually will go to the trouble to cast their votes. There are also plenty of secondary issues, such as the fact that Kerry gets a lot of his support from blacks, who are more likely to have their votes rejected on procedural grounds.
However, there are two factors that bear on this question. The first is voter anger. It’s an oddity of American politics that getting one person mad at you is much worse than getting one person to like you. Angry voters go to a lot of trouble to oust the objects of their anger. They have high participation rates. So we must ask ourselves, who is angrier about the state of this country: Bush supporters or Kerry supporters? The answer, I think, is obvious. The Bush-haters are really, really mad at Bush. There aren’t so many genuine Kerry-haters; most of his detractors are really just Bush-lovers going through the motions.
But I also have some empirical evidence to offer. For the last few weeks, I’ve been keeping track of the letters to the editor in my local newspaper. Now, I live in a rural, conservative area. Republicans dominate most elected offices. Although Oregon voted narrowly for Gore in the 2000 election, most of Gore’s support came from the Portland area while rural areas like mine voted overwhelmingly for Bush. So let’s start off with the observation that the people writing letters to the editor come from a population that’s strongly Republican.
So far, I have counted 81 letters supporting Kerry or attacking Bush, against 34 letters supporting Bush or attacking Kerry. That’s more than 2-1 in favor of Kerry! If letters to the editor were the only votes, then Kerry would win by a landslide in this conservative area.
Of course, letters to the editor are not votes. But they clearly indicate just how great the difference in voter participation rates will be. The pollsters are in for an embarrassing November.
Note added December 13: 2005. Well, golly gee, I turned out to be wrong. I’d be willing to admit my error with a smile but for the fact that there were serious electoral anomalies in Ohio and Florida. I’m not playing some sort of tin-hat conspiracy game here; instead, I will simply challenge the reader with the question, “Can you prove that the election anomalies in Ohio and Florida did not affect the outcome of the election?” I’m not claiming that the election was stolen; I’m saying that we can’t be sure that it was an honest and fair vote count. Nevertheless, until more evidence comes in, I have to admit that my prediction appears to be way off. Even if we accept that the election was stolen, I was predicting a big margin for Kerry, and that certainly didn’t show up. So I am certainly wrong on that point.