Earlier in the week, we wrote about a recent SurveyUSA poll of Alabama’s 2nd Congressional District that showed Republican Jay Love leading Democrat Bobby Bright by an unexpectedly wide margin: 56%-39%.
We wrote at the time that the poll, commissioned by Roll Call, didn’t really pass the sniff test. Despite the 2nd District having an African-American population that makes up between 29 and 30% of the district, SurveyUSA pegged the black vote at only 16% of their voter screen. Other pollsters, including the reputable Democratic firm Anzalone-Liszt, peg the African-American vote at 10 points higher — and consider that a conservative estimate given the historic nature of this year.
It now turns out that Roll Call itself was dissatisfied with SUSA’s work:
The poll, according to SurveyUSA CEO Jay Leve, calculated black turnout at 16 percent. According to exit polling from the 2004 general election, black turnout in Alabama’s 2nd Congressional District was 24 percent. Leve said black turnout was calculated at 16 percent because “that’s what we got when we conducted the survey.”
After the poll results were released by Roll Call on Tuesday, Roll Call asked SurveyUSA pollster Jay Leve to “re-weight” the results based on greater black turnout.
According to Leve, Love’s lead shrinks to 50-44 if black turnout equals 2004 levels. It shrinks even more to 49-46 – within the 4 percent margin of error – if black turnout increases to 27 percent.
“I do not consider the Democrat concerns anything other than fairly voiced,” Leve said. “It’s important to understand that there is no way to know which number is closest to the truth, because the truth is hard to know. It is safe to say that depending on what the black turnout is, the Republican will either win by quite a bit, or the race may be very close.”
“Democrat concerns”. Heh, nice one.
As I said on Tuesday, it’s very telling that even Jay Love’s internal polling has shown this race to be neck-and-neck. Alabama pollsters with a long track record in statewide politics like Anzalone and the Capital Survey Research Center have both shown Bright leading by ten. Perhaps the race isn’t quite as favorable for Bright as it was in early August (though there’s no real reason for it not to be), but it’s pretty clear that SurveyUSA popped out a dud — plain and simple.
SSP currently rates this race as a Tossup.
(Hat-tip: Left in Alabama)
Jay Love can wipe that smirk off his face now.
Here he is talking about the Texas Primary polling he did:
SUSA can get it wrong, and has in the past, but their refusal to make forced assumptions makes their polling much more credible IMO.
but they messed up big time on this one.
It’s not that their numbers are wrong, it’s that their crosstabs are.
This race is a tossup. This poll makes that clear when it’s crosstabs are adjusted to fit reality.
The pollster’s name is Jay Leve and the candidate’s name is Jay Love? Crazy freakin’ coincidence.