Two reasons to question the poll numbers
[1] As others have argued, the polls that have Bush leading by double digits tend to overestimate the number of Republicans likely to vote vis-a-vis Democrats.
From talking with Republican friends of mine (yes, I have some), I have the impression that the Republican turn-out will end up being disappointing. One friend of mine who voted for Bush in 2000 will not be voting for anybody in 2004. Others who were strongly pro-Bush in 2000 and for the first two years of his term have so soured on him that I wonder whether they too will end up not bothering to vote on election day.
[2] The AP reports that young people (those under 30) are registering in large numbers for this election, especially in such battleground states as Michigan, Wisconsin, New Mexico, Ohio and Florida.
Since young voters are mobile and tend not to vote, polls often ignore them in estimating "likely voters." We shall see whether the factors motivating them to register in larger numbers translate into actual votes on election day.
[Thanks to EA]
9:17:00 PM 
|
