RCP moves Michigan from the Obama column to Toss Up

(Update, October 12: A Rasmussen poll reported on October 11 and showing Obama ahead by 7 points has been added to the RCP average for Michigan, increasing Obama’s lead from the 3.7 where it was yesterday to 4.4 Yet RCP still puts Michigan in the Toss Ups. So I guess RPC defines any margin less than five points as a Toss Up.)

Last night I reported that Real Clear Politics showed 217 electoral votes solid or leaning for Obama, 181 solid or leaning for Romney, and 140 Toss Ups. That’s just changed. RCP has moved Michigan with its 16 electoral votes from the leaning Obama category into the Toss Up category, so the projected electoral total is now 201 for Obama, 181 for Romney, and 156 Toss Ups. However, the RCP average of the four most recent polls in Michigan still has Obama 3.7 ahead. I don’t know RCP’s rules, but perhaps any difference less than four percent is called a Toss Up.

More significantly, the 3.7 Obama lead represents a huge comedown from where he was in the second half of September. RCP shows five separate polls done in Michigan between September 14 and September 30, and their average was (get this) a ten point (10.0) Obama lead. Further, the four polls in October, on which the current RCP average is based, were all taken after the debate on October 3. So we’re looking at a 6.3 percent decline in Obama’s margin of support in Michigan that seems to have occurred solely on account of the debate.

I’m sure people who are more knowledgeable than I on these matters can add qualifications to what I’ve said and instruct me on how I’m being fooled by this or that misleading aspect of polls. But to my inexpert eyes this looks pretty spectacular. Offhand can one think of a presidential debate in the last generation that caused such a dramatic shift in the race?

This is fun.

* * *

Furthermore, a Drudge headline today says, “Poll aggregator [meaning RCP] changes 5 states from ‘leans Obama’ to ‘toss up’…” So apparently—maybe just prior to last night when I looked at RCP’s electoral map—RCP has recently moved four other states besides Michigan from lean-Obama to Toss Up. However, I don’t know how to access RCP’s older electoral vote pages to find out which four states they were. (Note: A half hour after I posted this, that Drudge headline was gone. It doesn’t matter; it pointed to the same RCP electoral vote page I’ve linked above.)

- end of initial entry -

A pollster writes:

No, the debate did not cause this big shift, it caused a small shift which, incrementally, was enough to make untenable the continued slanting of the polls to make it look like Obama was easily winning. Therefore they stopped putting their thumbs on the scales and the polls reverted to where they really were all along, plus a small debate-related bump for Romney. All is proceeding as I have foreseen. [LA replies: This is an assertion I do not understand. In what sense were the polling organizations or RCP putting a thumb on the scales for Obama before the debate, but not after the debate?]

At this point I don’t disagree with them on any of the “leaning” states, but I think that they should have counted two tossup states (Pennsylvania and Michigan) as leaning for Obama and two others (Missouri and North Carolina) as leaning for Romney. Of the remaining eight states, Romney has the advantage in Virginia and Florida, Obama [?] in New Hampshire according to my reading of the data.

That brings it to 248 to 241 for Romney—Romney then wins with any three of the remaining five states, or Ohio plus one of Nevada, Iowa, Wisconsin, or Colorado. Obama needs Ohio plus two others, or all four of the others.

Posted by Lawrence Auster at October 11, 2012 07:04 PM | Send