Rasmussen (1/4, likely voters):
Martha Coakley (D): 50
Scott Brown (R): 41
Some other: 1
Not sure: 7
(MoE: ±4.5%)
Rasmussen Reports sees a nine-point race in the special election contest between Democratic AG Martha Coakley and Republican state Sen. Scott Brown to replace Ted Kennedy. At first glance, that may seem closer than expected, since people are generally used to Democrats winning federal races in Massachusetts with at least 65% of the vote. Frankly, I'm not especially surprised, though, given that a) it's Rasmussen, whose likely voter model seems tailored to project the Dems' worst case scenario, b) there were pervasive rumors yesterday of an unreleased private poll that had a Coakley lead of 50-39, and c) this is the first Senate race in, well, more than a decade where the Republicans have bothered fielding an appealing, somewhat-well-known candidate instead of the usual unknown sacrificial lamb. Recall that while Ted Kennedy and John Kerry won their last elections against nobodies 69-31 and 66-31, respectively, Kennedy beat Mitt Romney 58-41 in 1994 and Kerry beat William Weld 52-45 in 1996. If Rasmussen's numbers project out to, say, a 54-45 Coakley win, then, well, that certainly fits within the scope of those latter races.
Needless to say, this poll has led to a fair amount of press hyperventilating today, wondering if Brown could somehow pull it off. Chris Cillizza remains highly skeptical, saying that Coakley's name rec and fundraising advantages are "close to determinative." Coakley was, as of Nov. 18, sitting on $1.9 million (after raising $4.2 million) compared to Brown's $258K, which will certainly set the tempo for the last few weeks' worth of TV advertising, especially since Brown seems unlikely to receive any NRSC financial help (which would have arrived long ago if they had the data to support the idea that there was a competitive race here). Cillizza also cites Democratic sources saying that their own polling hasn't seen Brown emerge out of the mid-30s.
Nate Silver is also skeptical, if mostly of Rasmussen's sample composition; it's heavy on conservatives, compared with exit polls, and light on registered independents, who make up a big chunk of the Massachusetts electorate. In the end, he throws up his hands, though, saying that in a super-low-turnout election (as this one is poised to be), almost anything is possible, from a Coakley landslide to a nailbiter.
http://www.swingstateproject.com/diary/6132/masen-rasmussen-sees-coakley-leading-brown-by-9