The Poll Nobody Knows

With an assist from The Mystery Pollster [http://www.mysterypollster.com/main/] (an old friend, as it happens) I now realize I overstated the number of undecideds (see below). But the number of undecided, uncommitted ("swing") voters is still higher than the polls imply. (See my 10/31/04 Top of the News, Election 2004--Deciding What's Undecided .) For that reason and because of of the "incumbent rule" that says the undecideds break for the challenger[http://www.mysterypollster.com/main/2004/10/the_incumbent_r.html], I still believe that my long-ago, first-in-the-nation prediction of a Kerry victory [http://forbes.com/2003/12/15/sp04_economy_2.html]will will turn out to be correct.

Amazing news from the Pew Research survey: The number of “uncommitted” voters is close to 50% of the electorate and the share of “undecided” is 16%-- not the 3% of 5% that most pollsters claim. Pew says that 27% are “certain” to vote for Kerry; 22% certain to vote for Bush. Another 16% “favor” Bush; 13% “favor” Kerry. The late movers to date are heading towards Kerry.

I believe the difference between these results and the ones we usually see is that most pollsters put leaners in one camp or the other and they insist the respondents express a view.

For details see: http://people-press.org/reports/display.php3?ReportID=231