This is the latest in a series of weekly corrections of presidential horserace polls, reweighting polls with overcounts of Democrat voters and undercounts of GOP and Independent voters with a partisan participation which more closely resembles that of 2004 and 2010, when the GOP was enthusiastic about the election and the Democrats were not. Note: This is only a rough corrective to show approximately what the polls ought to look like with the Democrat bias removed, not a prediction of the election results.
Here are 6 likely voter polls for this week excluding Rasmussen and Gallup, for which I do not have the partisan weighting:
Monmouth University
Original Poll Results: Obama 46%, Romney 44%
Percentage Democrats outnumber Republicans: +4%
Corrected Results: Romney 46%, Obama 46%
NBC News/WSJ
Original Poll Results: Obama 50%, Romney 45%
Percentage Democrats outnumber Republicans: +7%
Corrected Results: Romney 48.5%, Obama 46.5%
AP/GfK
Original Poll Results: Obama 47%, Romney 46%
Percentage Democrats outnumber Republicans: +11%
Corrected Results: Romney 50.5%, Obama 41.5%
Pew Research Center
Original Poll Results: Obama 51%, Romney 43%
Percentage Democrats outnumber Republicans: +6%
Corrected Results: Obama 47.5%, Romney 46.5%
Reason/Rupe
Original Poll Results: Obama 48%, Romney 43%
Percentage Democrats outnumber Republicans: +8%
Corrected Results: Romney 46.5%, Obama 43.5%
Hartford Courant/U Conn
Original Poll Results: Obama 46%, Romney 43%
Percentage Democrats outnumber Republicans: +8%
Corrected Results: Romney 47%, Obama 42%
Average of Corrected Results: Romney 47.7%, Obama 44.4%
Romney appeared to have slipped a couple points from last week’s 49.6%, but the Obama support is virtually identical to last week’s 44.8%
Of note, the Rupe and U Conn polls appear to have nearly doubled the Hispanic turnout from 2010 to get to their Obama leads.
Unskewedpolls.com is also correcting polls, arriving at more favorable results for Romney than I do.
