I'm one of those rare people who actually reads the small print at the end of advertisements. By examining the print most people can't even see, I can tell the difference between how the deal is being framed by the advertiser and the reality of the deal.
I take this same care with polls and while doing my research. Earlier this week, I happened to see a headline on the gallup.com website that captured my attention: "2012 Electorate Looks Like 2008." I clicked on the link and read the article as well as the data behind the article. http://www.gallup.com/poll/158399/2012-electorate-looks-like-2008.aspx
As I read the article, I noticed that the demographics of likely voters in the 2012 election are expected to be pretty much the same as they were in 2008. The proportion of men to women, ethnicity and age are expected to be very close to what they were in 2008, according to Gallup's polls. The proportion of voters in different areas of the country, as well as the voters' educational levels are also expected to be very close to 2008.
But then I noticed some data almost buried at the bottom of the page that showed a significant shift in likely voters. Unless you took the time to read the entire article, you wouldn't have even noticed it. The headline didn't even hint at this shift. The fact is that the percentage of "likely voters" in 2012 has shifted significantly to the Republican side, when compared to 2008 voters.
In 2008, voters identified as being Democrats were 39% of the electorate, with independents making up 31% and Republicans coming in at 29%. In 2012, the numbers of likely voters have shifted to 35% Democrat, 29% independent and 36% Republican. In other words, based on Gallup's polling Democrats are expected to vote at a rate four percentage points lower than in 2008, but Republicans are expected to vote at a rate seven percentage points higher than in 2008. Interesting. That's an 11-point swing.
If you look at the projected likely voters in terms of those who are leaning Democrat compared to those leaning Republicans, Republicans are expected to vote in higher nunbers than Democrats, 49% to 46%. In 2008, the electorate was 54% Democrat or leaning Democrat, vs. 42% Republican or leaning Republican. That's a 15-point swing to the Republican side.
What does all of this mean?
According to Gallup's numbers, all of the polls that are oversampling Democrats by 5% or even more are dead wrong. In a poll showing President Obama winning by 5% with 5% oversampling of Democrats, Mitt Romney should win by 3%. In polls with 5% oversampling of Democrats that show the President winning by 3%, Romney should win by 5%. To put this simply, a few of those states that are leaning Obama may actually turn out to Romney wins, and those states that show a dead heat may actually turn out to be Romney wins of 8%. This election could turn out to be a rout in Romney's favor.
Polls of early voting by the same Gallup group seem to reinforce the notion that Republicans are going to out vote Democrats in 2012. Through Sunday, 15% of those surveyed said that they had already cast a ballot in person or via absenstee ballot. Of those surveyed who had already voted, 52% of those polled said that they had voted for Romney, while 46% said that they had voted for President Obama. Of the 63% surveyed who plan to vote on Election Day similarly plan to support Romney by a 51% to 45% margin.
http://online.wsj.com/article/SB10001424052970204846304578090820229096046.html?mod=googlenews_wsj
I'm glad I actually take the time to read the fine print. That's the only way I would know that the vast majority of polls are so biased in favor of President Obama that they are bordering on the ridiculous.
No comments:
Post a Comment