Over the past few weeks, the daily Gallup poll which is currently sampling adults has had President Obama's approval rating at a lower level, than the daily Rasmussen poll which uses likely voters and is generally understood as Republican leaning.
When Gallup and most other polls for months and years had Obama at much higher levels, Scott Rasmussen claimed it was because his polled sampled likely voters, while others used registered and sometimes "adults". It is widely accepted that Democratic consitutencies are less likely to vote and less likely to be registered voters, so that made sense.
But for the past couple of months Rasmussen has numerous times had the President at 50% and in positive territory (approval - disapproval).
Today Rasmussen has 50% approval 49% disapproval. While Gallup has 45% approval and 48% disapproval.
Could it possibly be that the more likely voters are more in support of the President? That there has been a severe (thanks Romney!) reversal of pattern where Republicans are less involved in the process?
I personally think that Gallup, having received criticism about high poll results for the President changed their methodology and is now offering a false narrative. The gold standard in polling has done strange things before: What's wrong with Gallup?
Conversely, I've suspected that Rasmussen that makes up polls much of the time, so why would he release polls that benefit the President?
At least right now, I think Rasmussen is an honest pollster. It makes the most sense.
But I have noticed that Fox News and the right wing blogs have quoted his polls much less frequently lately.
As a matter of fact, I saw Karl Rove quote Gallup rather than Rasmussen!
Obviously, polling is political.
I'm in favor of having at least one "official pollster" that is paid by we the people through our government that is honest. Sort of like the Congressional Budget Office.
The media and the public use poll numbers to shape and change public opinion, so it is important that they are legitimate from people purporting to be scientific.
Take for example health care. Left leaning pollsters have proven that a majority want to keep the healthcare bill (around 40%) and around 13% want something more liberal. The key is that many of those 13% also would repeal the ACA. So right leaning pollsters can just ask, Repeal or no? and leave the impression that people are against the liberal plan. When in fact, the facts show the opposite. I'm not for censoring pollster and polls, but I am for having people who don't care about Dems or Reps to be fair.