The only real poll was taken in Novemner of 2004. That's the one to govern by. That's exactly what Bush is doing.
<[p>
Next one is slated for November 2008.
It will be interesting to see what the approval numbers do when the tapes of Saddam discussing his WMDs serve up a plate of crow to the "Bush Lied" crowd.
Hard to say since they use what they call 'weighting' to adjust for known biases. If they listed their questions it would be easier to know. The random phone sample tells us nothing of the demographics. How many Dems, Repub, independents and were in the sample? It's probably about a mid 40's approval rating.
Bush's poll numbers would soar to 75% approval if he were to seal the southern border and begin enforcing immigration laws.
Okay which is the real poll and which is the fake? At least Pew gives a breakdown by political party. Rassmussen just says the weight according to the magic sauce...with no specifics given.
The following is all you need to know about the PEW Poll:
PEW RESEARCH
[As Maddie Albright's polling organization, Pew routinely oversamples Democrats. This month it did so by 8 points -- +3 Rs vs Ds and another +5 Independents who lean Democrat vs Republican, for a total of +8. With this in mind, the following polling results are quite POSITIVE for the President]:
Feb 2006 40% approve 52% disapprove
Jan 2006 38% approve 54% disapprove
Dec 2005 38% approve 54% disapprove
Nov 2005 36% approve 55% disapprove
Oct 2005 38% approve 56% disapprove
Sep 2005 40% approve 52% disapprove
While most ratings held steady; the following actually INCREASED in February:
IRAQ: 38% approve
(before Nov 04 election: 37%)
TERRORISM: 53% approve 40% disapprove
(highest since May 05)
IRAQ: 51% right decision 44% wrong decision
(before Nov 04 election: 48% right 41% wrong)
HOW WELL IS GOVERNMENT DOING IN REDUCING THE THREAT OF TERRORISM:
68% (very well/fairly well) . . . Jan 06 66%
WHAT CONCERNS YOU MORE ABOUT THE GOVERNMENT'S ANTI-TERRORISM POLICIES:
33%. . . gone too far in restricting the average person's civil liberties.
50%. . . that they have not gone far enough to adequately protect the country.
DO YOU THINK IT GENERALLY RIGHT OR GENERALLY WRONG FOR THE GOVERNMENT TO MONITOR TELEPHONE AND E-MAIL COMMUNICATIONS OF AMERICANS SUSPECTED OF HAVING TERRORIST TIES WITHOUT FIRST OBTAINING PERMISSION FROM THE COURTS?:
Feb 06 54% right 43% wrong
Jan 06 48% right 47% wrong
[Take that Specter, Leahy, et al!!]
[click on 'topline questionnaire'(ignore leftist analysis)!]
http://people-press.org/reports/display.php3?ReportID=269
All polls are BS, not even because of biased questioning or oversampling, but because of weighting.
Poll 112 people and then weight it assuming America is 31% republican 32% democrat and 37% independant, or whatever 'weight' they want to make it.
So it so totally fugeable, ESPICALLY if a bunch of pollsters 'agree' on a set of weights.
The only 'test' of these weights are elections of course, as others pointed out.
Finally, I think since so many democrats run away from 'liberal', that ex-democrats are describing themselves as 'independant'. This has the effect of both making the 'democrat' percentage more 'pure' and so higher, while also bringing down the previously more conservative 'independant' pool.
If you do not adjust your 'weighting' in that case, it compounds the inaccuracy of the polls.
Of course the is biased questioning and all that, I mean there are a million reasons why polling is generally useless, and BY FAR the most important factor in accuracy is in sample size... thus elections.
I still think the numbers should be better for Bush especially with the good economic numbers...