I'm sure Poblano will be all over this, but on the ARG blog now is this assertion from Dick Bennett:
Clinton, and not Obama, tends to win in the states where turnouts are higher. The relationship is strong enough that Democratic primary turnout can be predicted using Clinton’s share of the vote.
The proof?:
In Mississippi, Obama received 61% of the primary vote when 24.4% of all registered voters in the state voted in the Democratic primary.
on the one hand, vs.:
In Ohio, Clinton received 54% of the primary vote when 30.5% of all registered voters in the state voted in the primary.
The first problem that jumps out at me is that he's comparing the percentage of the entire electorate, but in a Democratic primary. (Is he accounting for open v. closed primaries? I doubt it, but I don't think it matters much because the numbers aren't even useful to begin with.) The states he highlights for Obama are: Georgia, Virginia, and Mississippi. For Clinton: West Virginia, Pennsylvania, Ohio. My hunch is that the Democratic Party registration in the latter three states is a lot higher than the former three.
Also, there's no comparison of turnout v. last year or a presidential year. You can't say that someone is "bringing out more voters" unless you have a baseline.
It is quite odd that ARG is both bizarrely anti-Poblano, and now apparently in the bag for Clinton. It seems clear now that it is a right-wing outfit, finding itself aligned with Clinton just like Limbaugh does [Update from the comments: dcatalin rightly notes that I'm making a pretty unsubstantiated leap here]. (Also, see here, where Poblano notes that ARG always overestimates Clinton's strength by an average of 4.5 points.)
I guess an actual rational conclusion to draw from this data (assuming that he's right that you can predict turnout from Clinton's share) is that Clinton does better in strongly Democratic states. I don't think that's a surprise. But I think it's already been debunked that this means Obama is in trouble in the fall.
cross posted at yazilikaya