Why national polls don't matter 90 days away from the first votes

[zulchzulu]

Today marks 90 days until the Iowa caucus if you think it will be still on January 14th, 2008. It will probably move to January 4th or 5th.

Some people are obsessing about national polls and spiriting the idea that it is all but "inevitable" for Hillary Clinton to win the nomination. Yet, if you look at the history of front runners this far out of the race's first votes, you see how historically it is not actually a good position for her to be in.

CBS News Poll, December 14-16, 2003
Dean - 23%
Clark - 10%
Lieberman - 10%
Gephardt - 6%
Sharpton - 5%
Kerry - 4%
Edwards - 2%

As you can see, Al Sharpton was beating John Kerry in the national polls in December.

A week before the Iowa caucus occurred, the polls suggested with Opinion Dynamics:
Dean - 20%
Clark - 13%
Lieberman - 8%
Gephardt - 7%
Kerry - 7%

After Kerry won Iowa, the national polls changed drastically overnight:
Kerry - 29%
Dean - 17%
Edwards - 13%
Clark -11%
Lieberman - 5%

Just to illustrate what national polls do after the first four states vote, here's Time/CNN poll breakdown:

John Kerry
2/5-6/04 - 43% (Dean 8%) (Not sure 8%)
1/14-15/04 - 9% (the day the Iowa caucus happened) (Dean 19%) (Not sure 17%)
1/1/04 - 10% (Dean 22%) (Not sure 23%)
11/18-19/03 - 9% (Dean 14%) (Not sure 23%)

If people want to take solace in their candidate's lead in national polls at this point and do nothing on a grassroots level, then they are surely going to help out other candidates who have a great ground game in the first four states.

http://www.pollingreport.com/wh04dem.htm
http://www.fpc.edu/pages/institutes/poll/poll_03_1208.pdf