There sure are alot of polls. It seems like everybody and their brother has a Presidential poll showing this, that, or the other trend. There are official polling organizations such as Gallup and Rasmussen, and there are polls run by media outlets such as the Washington Post and Fox News. Most of them don't just select random names out of the phone book and start calling: they have specific procedures for defining their specific sampling criteria. Something that surprised me was that some of these groups actually choose their participants based on their party affiliation from voter registrations. The number of Democrats and Republicans are usually determined by election results from previous years. Some pollsters also sample independents. Others further refine their samples by demographic data such as race or gender. The methodology is not consistent across organizations, so some tend to be more accurate than others. Regardless of who is running them or where they come from, though, I have this simple advice: Ignore them.
You heard me right - forget about them. They will cause you undue stress and frustration (or give you false hope, if you are a liberal). The reason for this is because almost all polls over-sample Democrats. This isn't necessarily intentional, as even Fox News polls reflect this flaw. There are many possible reasons for this:
- There are more self-identified Democrats than Republicans (although this is doubtful, as other non-election data shows the opposite).
- More people *register* as Democrats.
- Many conservatives call themselves independents when they would otherwise vote Republican (Tea Party, small "l" libertarians).
- The contact method (land-line, mobile phone, Internet) is prone to reach more Democrats than Republicans.
Whatever the reason, I don't care. They are hooey... mostly. I use them as a starting point for my projections, and a general guide to trends, but that's about it. If you are a conservative who wishes to take back the country from the Socialist in Chief, take heart! Consider this:
- Based on my observations over the years, polls usually move right after the Democratic Convention bump subsides (I've confirmed this through changes I've seen in electoral vote maps)
- I have also observed that the *actual* election results are almost always at least one point in favor of the Republican vs. the pre-election polls
- Anecdotal evidence shows that the Bradley Effect may come into play this time around. The Bradley Effect is named after a California gubernatorial candidate from 1982 who lost the election after polling data showed he had a commanding lead. Sometimes voters will tell pollster that they are voting for the minority candidate to avoid appearing racist, when they actually end up not voting for that candidate. Obama is unpopular enough now that many people aren't willing to convince themselves of his appeal anymore.
My point to conservatives is this: don't get discouraged. There is a very real chance that Obama will be defeated. Stay strong!
Referenced article by Walter Hickey: http://www.businessinsider.com/polls-skewed-bias-democrats-2012-9