It’s not all that hard, people. There are four issues, three of which are fundamental.
1. Polls are retrospective not predictive. With comments that were coming out – especially of Obama helping to nationalize the election – opinions can change very quickly.
2. Push polls vs. opinion polls. Far too many of the polls I saw seemed more interested in shaping opinion rather than discovering opinion. Sometimes this was due to the biases of the poll’s designer subconsciously influencing the questions. Sometimes it was by intent to use the poll as a marketing tool.
3. Error in interpreting the meaning. This was a biggie. It seems almost everyone interpreted dissatisfaction or dislike of Republicans as something that benefited the Democrat. The reality is that much of the anger at Republicans is that they have not done enough oppose and reverse Obama’s agenda. Outside of a few moderate centers, people are sick and tired of Republicans being “Democrat Lite,” and their bait-and-switch tactic.
4. People responding lied. In some cases it was out of fear and distrust. I mean, when the Democrats are sending out warnings that they’ll know if you don’t vote, can you really blame that? The other reason that people lied is to intentionally screw up the results. Browse around online, you’ll find plenty of people bragging of just that. This is especially the case in libertarian circles and with people who are fed up with the constant spam phone calls.
Add these in to the inherently chancy nature of predicting future based on a poll and you have the recipe for what we’ve seen here. There is one interpretation that should be easy to draw from these poll results. The political class – including these pollsters – are completely out of touch with the American people. Want better polls? Check yourself for the logical errors in presumption and then address the four points above.