Polling Literalism

For the record, I'm not spending my Friday night being lame -- should be off to the Fiery Furnaces show pretty soon. That said, let's blog some. The main thing that troubled me about Ezra's "fuck the center" post from this morning wasn't really my quasi-accurate invocation of the median voter theorem. Rather, I worry that trouble's brewing when I see political strategy based on this sort of data:

Because, surprise surprise, the nation doesn't quite rest in the magic center either. They like class warfare, soaking the rich, government-run health care, preserving the environment, and participating in all manner of international treaties.

Call it "polling literalism" and call it a huge mistake. Polls show all kinds of stuff, and the main thing you can learn from studying polling about voters' stances on the issues is that the average person doesn't really understand what he's talking about.

I've seen polls showing, pretty consistently, that people favor a flat tax. Polls also consistently show that people think the rich should pay more taxes. And that spending should be increased in virtually every area. And that spending on the whole is too high. Back in 1984, 19 percent of the electorate, a plurality, said that the budget deficit was the most important issue. 64 percent of those people voted for Ronald Reagan, the man who'd created the deficits, over Walter Mondale the man who (accurately) predicted that either candidate would wind up raising taxes to close the deficit but that Reagan would lie about it and pretend that he wouldn't.

Back during the 2004 campaign, John Kerry stepped right into the gaping polling literalism void by going on and on about how important it was to be "respected in the world" because he had polls showing this was popular. Well, it was popular. But not granting "permission slips" was popular, too. And so was not flip-flopping. It was hard to say "respected in the world" without sounding like you meant "permission slip" or else falling into a French-sounding, flip-floppy nuanced explanation of what you meant. The dynamics of a campaign aren't like the dynamics of a single poll question. More recently, the White House has stepped right in it on Social Security, apparently emboldened by a huge amount of polling indicating that "people should be allowed to put a portion of their payroll taxes in private investment accounts" (or something like that) was popular. The polls were real. Not popular, though, are cuts in guaranteed benefits or large amounts of new borrowing. If you call it "borrowing from China and Japan" the borrowing gets a lot less popular. The administration didn't consider that. Nor did they consider that the main focus of public support for privatization was among a demographic (young people) who already deeply distrusted the President and could be easily (and, as it happens, accurately, but the accuracy is largely incidental) that Bush was lying about Social Security. Nor did they consider that while few Republicans want to lose their seats over privatization, most liberals are perfectly happy to make this the hill we die on.

One could go on. The death of Social Security privatization may make for an interesting book someday, just as the death of Bill Clinton's superficially-popular health care plan has already made for a lot of fascinating post hoc reading. The point, however, is that you've got to be careful with these polls. They're snapshots into the confused, confused mind of the American electorate and need to be handled with a great deal of care.