On Tue May 4, Doug Henwood quoted an AAPORite who cited:
>See Nonvoters: America's No Shows (Doppelt, Shearer), Sage Publications,
>1999? They indicate nonvoters aren't so different - if they would have
>voted they wouldn't change outcomes.
I've sent out for some of these texts to take a look at them, but I'm curious whether the AAPOR-list went into the mechanics of this argument. I'm under the impression that
1) Nonvoters have a clearly distinct demographic profile from voters: they're poorer, blacker, more hispanic, younger, and less educated; and
2) AAPORites believe that demographics align with voting behavior. Some believe there's a very complicated relation, but they all believe there's some relation; that's why they're in this business.
Are either of these propositions wrong? Because if not, it would seem that if you argue that nonvoter results would be indistinguishable from voter results, you have to argue one of three things:
either (1) or (2) above are false; or
(3) there is a complex way in which the differences cancel each other out.
And for that last argument to fly, it would have to be explicated in propositions that would have to hold for, and could be tested against, the active electorate.
It seems to me that any of these finding would be an earthquake in AAPOR circles, the first two for obvious reasons, and the last because, on the face of it, the demographics of the nonvoter pool would seems to point to a democratic bias.
Did the AAPOR list discussion make clear which of these tacks the "nonvoters are the same" proponents take?
Michael