Gallup: Big Kerry mo'! (or CYA?)
by kos
Mon Nov 1st, 2004 at 05:39:04 GMT
Two weeks ago, Gallup had Bush with an 8-point lead. Today, in their final poll, they have it tied. I suppose the story should be "Big Kerry Mo'". But I no longer trust Gallup. I smell a CYA attempt to bring their numbers closer to what other pollsters are reporting.
Gallup. 10/29-31. MoE 3%. (10/22-24 results)
Likely voters, allocating undecideds
Kerry 49 (46)
Bush 49 (51)
Registered voters
Kerry 48 (47)
Bush 46 (49)
Meanwhile, the final round of battleground states also show
big momentum in Kerry's direction. Either that, or some more CYA. (Previous poll results in parenthesis.)
RV's LV's
FLORIDA
Kerry 49 (42) 49 (43)
Bush 45 (51) 46 (51)
IOWA
Kerry 46 (48) 46 (46)
Bush 47 (47) 48 (50)
MINNESOTA
Kerry 51 (46) 52 (46)
Bush 43 (47) 44 (48)
OHIO
Kerry 51 (50) 50 (48)
Bush 44 (44) 46 (47)
PENSYLVANIA
Kerry 49 (49) 46 (50)
Bush 47 (46) 50 (47)
WISCONSIN
Kerry 46 (43) 44 (44)
Bush 49 (51) 52 (50)
Kerry wins Florida and Ohio, but not Pennsylvania? I sincerely doubt it. Using their LV screen with the head-to-heads (ignoring the incumbent rule for a moment), plus New Hampshire, plus the Gore states, Kerry wins 273-265. Give Bush NH, and we get the dreaded 269-269. Using the RV numbers, plus NH, plus the Gore states, and Kerry wins 294-244, with PA going Blue. Throw in the incumbent rule (Bush loses roughly 80 percent of undecideds), and Kerry breaks 300 EVs.
Personally, I think we'll win all six of these states. I don't buy Gallup's "likely voter" screen, and Bush isn't above 50 percent in any of these with registered voters.
As for Gallup, it will have to account for itself after the election. Its polls trumpeted big Bush leads on CNN and USA Today headlines for months, even as other polls consistently showed a tight race. Even if their final "prediction" is more in line with the final results, they have been irresponsible this entire election cycle. We're not going to forget.
