if half the people that were polled refused to respond, wouldnt that pretty much invalidate the whole thing?Do the math dumb ass. Less than half of the alleged people who were polled actually responded to the rediculous questions. Push polls are as old as ...well... polls. Select the demographics and tailor the questions to likely response and you can get any result you want.
It's true that over half the people did not answer, but that means the confidence interval will widen to +/- 6-7%. The poll was consistent with other findings at the time, such as the President's popularity numbers, the unpopularity of healthcare reform, etc., so there is no reason to think that the poll was inaccurate.
And I do not believe this is a push poll in the sense that random people clicked online, i.e. ESPN's poll on who is going to win the Monday Night Football game. I believe the people were selected in the same way that they are selected for phone polls.
Track Us - What they're saying about Public Policy Polling