ENOUGH WITH THE INDIES! It occurs to me that what we really need isn't yet another study showing how this or that set of independents/office park dads/populist parents/reactionary retirees thinks, votes, or behaves. I don't trust the data on these voters, and so far as I can tell, any pollster with access to a thesaurus and a printer can make a compelling case that the relevant swing group just happens to be the one that demands precise fealty to the pollster -- or their organization's -- policy demands. What I'd like to see is an analysis going in the opposite direction: Which types of campaigns work best? Assuming -- and this is a big if -- you could create some blunt-but-semi-accurate metric for distinguishing base motivation campaigns from efforts directed towards winning the center, you could possibly get some data on what's more effective at actually reaching voters, not what's most theoretically in line with a momentary snapshot of their preferences. I'm not saying it'd be methodologically straightforward, but it seems at least as enlightening as these constant attempts at demographic sub-slicing. --Ezra Klein
