bq. When candidate support among the different ages of Iowa caucus attendees are applied to the age distribution of the 2004 New Hampshire Democratic primary electorate, support for Obama and John Edwards rises, while support for Clinton actually decreases.
From this piece, by Michael McDonald of George Mason and the Brookings Institution. His simulation buttresses the most recent polls.
(And, yes, I realize that I’m dwelling on the horserace despite my previous post, and that my previous prediction is starting to look, well, potentially very wrong.)