« Playing the Race Card for Fun and Profit | Main | Rethinking the Draft I »

March 09, 2005

The Dangers of Polls

Demonstrating why I get annoyed when he resorts to cheap shortcuts, Kevin also notes an interesting poll that suggests that Americans would cut defense heavily while increasing spending on education and deficit reduction if they controlled goverment spending. Such a poll is probably maddening to many liberals who wonder how it is they can lose when they're calling for just those things, but Kevin rightly notes that the poll simply can't be accurate: if John Kerry had called for a $100 billion cut in defense spending, he would have been clobbered in November and people would be talking about the Bush landslide.

I don't know why it is that people tell pollsters one thing and vote another way, but it is clear that they do. Polls that ask whether or not people voted in the last election, for example, consistently come up with more voters than the actual tally of votes. Presumably many people tell pollsters what they think they should say, rather than what they honestly believe. But the end result is that polls can do as much harm as good, because they often reflect reality as well as a funhouse mirror.

A case in point is the very one Kevin makes: Democrats point to polls like this as proof that people really do want liberal policies, and so the Democrats don't need to change. (I'm reminded of Principal Skinner's reaction to his inability to locate Bart Simpson when he skips school: "Why, there are no children here at the 4H club, either! Am I so out of touch? No, it's the children who are wrong.") Which dovetails neatly with my own recent arguments regarding the Democrats and national security. There are polls that suggest that the Democrats are on the more popular side of numerous issues, yet they can't seem to break the magic 50% barrier (quick, name the last Democrat to win 50% of the national vote. Yes, that's right, it was Jimmy Carter with a whopping 50.08%.) Now the Democrats can continue down Principal Skinner's road and insist that the voters are wrong; from the perspective of principle, there's nothing wrong with that. I often think the voters are wrong, because their decisions conflict with my own. But that path also leads to continued defeat at the polls.

The science of polling is a fascinating one. But like any other science, it is only as good as its results. Due perhaps to deception on the part of the polled or perhaps because more and more people simply don't respond to pollsters or perhaps for some combination of reasons the polls simply don't provide all that accurate a reflection of the country. The only polls that count are the ones also known as elections, and their results have not trended in the Democrats' favor. It appears that the Democrats are unlikely to improve their standing in the polls that count as long as they're relying on public opinion polls that bolster their core beliefs.