problem with that is many pollsters are independent of news services. To them it's all about accuracy and credibility. Gallup gains nothing by skewing a poll in one way or other or to keep it a tie.
That's what's confounding me, because I truly do not see 50% of Americans ready to change leaders in a time of war and even worse... to an ultra-liberal elitist from Boston...America has NEVER voted that way.
another bit of anecdotal evidence that backs my assertion:
Internals don't match externals:
-Women Support WAY up for Bush compared to 2000.
-Black Support, doubled for Bush compared to 2000.
-Jewish support up
-Catholic support up
YET IT'S A VIRTUAL TIE, (JUST LIKE 2000??????) HUH?
Polling is a business. Somebody pays the polling firms for their services.
Gallup gains nothing...
(1) Gallup has a long-standing business relationship with big media. (2) When polls prove inaccurate when compared to actual election results, few people seem to notice and/or care. There is always some excuse; some reason peddled as the cause. You'll see those excuses parrotted on TV and on forums such as this without so much as a heartbeat's pause to reflect on the obvious:
There is no objective way to evaluate the accuracy of all polls except those taken immediately prior to an election. The election results provide actuals against which poll estimates can be compared.
Far more often than not, poll estimates do not compare favorably to actual results. When the miss is signficant enough, some excuse is always found to explain why the election "broke" one way or the other at the last minute. Did it really "break" or were the polls just wrong throughout the election season? (I vote for the latter.)