http://www.tnr.com/blog/campaignjournal?pid=1243....
Of these three second-tier candidates I think Edwards is best positioned for a late surge for several reasons:
1) Undecideds, about 15 percent of the voters at the moment, usually don't break for the incumbent, and Kerry and Dean--neighboring politicians well-known to New Hampshire voters--are effectively the incumbents in this race. As for Clark, he has had weeks to himself to make his case and he is losing voters, not winning them.
2) Edwards, as Slate's Mickey Kaus points out, is the top second-choice candidate of New Hampshire voters, according to the WMUR poll.
3) Edwards supporters are more "sticky" than Clark's or Dean's, according to the ARG poll:
As of the tracking ending January 23, Clark's strong support is 63% of his ballot preference, Dean's strong support is 47% of his ballot preference, Edwards's strong support is 77% of his ballot preference, and Kerry's strong support is 74% of his ballot preference. As a result, a total of 29 percentage points overall could switch before Tuesday, not including the 12% undecided.
So not only are Edwards's voters more loyal to him than his oppoents' voters are to them, but when his opponents' wishy-washy supporters start to abandon them, they're most likely to shift into the Edwards camp (see point number 2 ).
4) Finally, New Hampshire voters have a more favorable opinion of Edwards than of Dean and Clark. In The Boston Globe poll, there is a 44-point difference between Edwards's favorable and unfavorable ratings. The spread is 37 points for Clark, and just 10 (!) points for Dean. Dean's unfavorable rating went through the roof after his Iowa scream speech, and it still hasn't climbed down. In fact, in the Globe's poll it actually went up even after Thursday night's emergency interview with Dean and his wife.