The story over the weekend in the Des Moines Register: "Iowa's Independent Voters Turning Away From Obama" certainly turned heads. The Iowa Poll is really good at what it does, and so the news that 38% of self-described independents approve of Obama is dispiriting. (It's down from 48% when Obama was elected.)
But the story here is, as it usually is, much more complicated than it seems. The basic truth about independent voters is that they are not really independent. They like to think they are, and that cognitive dissonance has been the source of endless confusion. As John Sides points out, it's one thing to register as an independent, and to tell pollsters that you perceive yourself as an independent, but identification and behavior differ. And independents, most of them, tend to lean so far in the direction of one party that their voting patterns and ideologies are virtually indistinguishable from the median of the party itself. Where independents of this sort, as opposed to the roughly 10% of the presidential electorate who are truly non-affiliated -- let me start that again -- where the partisan independents do matter is that they tend to vote less frequently than those registered or self-identified with a particular party. Even this is overstated; about 75% of independent leaners are as partisan as self-identified partisans in behavior.
When people say that "independent voters are likely to swing Republican for the midterms," they're confusing several concepts. What's actually happening is that conservative-leaning independents are simply more likely to vote than liberal-leaning independents when the momentum and enthusiasm are on the conservative side.
What's happening with Obama? His approval ratings are down across the board, but the drop is sharpest among conservatives and Republicans -- Republicans and Republican leaners. It is much more pronounced, Sides notes, among Republican leaners than it is among pure independents (a fraction of the electorate). So the idea that Obama is hemmoraging independent voters tells us very little. The contribution to his ratings decline from Democrats is probably a result of a turning away by blue collar independent Democrats in the Midwest -- Reagan Dems -- many of whom identify as Republican but reliably vote Democratic.
Here's a chart by Mark Blumenthal that illustrates this to some degree:
This article available online at: