Among the issues most commonly discussed are individuality, the rights of the individual, the limits of legitimate government, morality, history, economics, government policy, science, business, education, health care, energy, and man-made global warming evaluations. My posts are aimed at intelligent and rational individuals, whose comments are very welcome.

"No matter how vast your knowledge or how modest, it is your own mind that has to acquire it." Ayn Rand

"Observe that the 'haves' are those who have freedom, and that it is freedom that the 'have-nots' have not." Ayn Rand

"The virtue involved in helping those one loves is not 'selflessness' or 'sacrifice', but integrity." Ayn Rand

For "a human being, the question 'to be or not to be,' is the question 'to think or not to think.'" Ayn Rand

07 October 2012

Making Sense of the Rasmussen Presidential Tracking Poll

According to a three-day average over 3 - 5 October, the daily Rasmussen Presidential Poll reports that Romney was leading by 49% to 47%.  However, Rasmussen also says that Romney has the vote of 89% of Republicans and Obama has the vote of 88% of Democrats, while Romney is up 16% on Obama among those not affiliated with the Republican or Democrat Parties.  The report I have seen does not say how many Republicans are planning to vote for Obama or how many Democrats plan to vote for Romney.  The report also does not say what percentage of Republicans, Democrats, and unaffiliated voters were sampled.

The Rasmussen party affiliation survey for September is now available.  According to it party affiliation is as follows:

Republicans, 36.8%
Democrats, 34.2%
Other, 29.0%

Democrats have increased by 0.9% since August, Republicans have decreased by 0.6%, and Others have decreased by 0.2%.  I expect the media claims that the Romney campaign was losing and Romney did nothing but make gaffes had pushed down the Republican affiliations and boosted the Democrat affiliation.  Romney's debate performance may reverse that trend in October.

I will assume that 4% of the 29.0% may vote for the micro-parties such as Libertarian, Green, Constitution, and Socialist Workers.  This leaves 25.0% of non-affiliated voters to pick either Romney or Obama and Romney is up by +16, so this implies that 58% will vote for Romney and 42% will vote for Obama.  The last two Washington Post/ABC News polls I adjusted for party affiliation had about 7% of each party voting for the candidate of the other party, so I will assume that here, to the advantage of Obama.  So, how does the adjusted vote turn out then:

Romney:  (0.89)(36.8%) + (0.07)(34.2%) + (0.58)(25.0%) = 49.6%

Obama:   (0.07)(36.8%) + (0.88)(34.2%) + (0.42)(25.0%) = 43.2%

Thus, Romney had a 6.4% advantage in a poll including the day of the debate and the two days after the debate.  How on Earth Rasmussen manages to reduce this to a 2% Romney advantage, I have no idea.

If the pollsters are thinking that Republicans and anti-Obama unaffiliated voters are not going to vote in this election, I think they are out of their minds.  What is more, Rasmussen reported that this poll gave Romney a double-digit advantage with voters over 40 years of age and these voters do turn up at the polls.  Obama on the other hand had a double-digit advantage with voters under 40.  That part of his pool under 25 are represented among the unemployed in large numbers.  Many of them cannot afford the high Obama gasoline costs to fuel their overly expensive clunkers (thanks to the Cash for Clunkers Obama program that made old cars scarce) to get to the polling places.  In general, the younger voters do not vote in the same percentages as older voters, so my adjusted results probably understate the extent of the Romney advantage.

In my lifetime, no presidential candidate has won the electoral vote when behind in the popular vote by more than a few tenths of a percent.  Even before the debate, there was considerable evidence that the polls of the swing states were heavily over-sampling Democrats and the reported leads by Obama in many of those states were fallacious.  An evaluation of polls for the states of Florida, Ohio, and Virginia made before the debate indicated that they were highly misleading and that Romney was really ahead in each of them.  I think Romney is going to win the election despite the best efforts of most of the media to throw the election to Obama.

I am quite disappointed in Scott Rasmussen and his polls.  I thought he usually did a good job polling, but the evidence here is that he is not doing so now.

3 comments:

L.A.Henry said...

The country has had enough of the Obama experiment. Hope and change only benefited those who don't contribute to the tax rolls. this isn't redistrabution, this is payback. It's time for a real change. As far as the Rasmussen poll,, I agree with you that the lead for Romney is way above 2 percent.

Harry Dale Huffman said...

Good. I only wish this were the narrative of even a portion of the mainstream media. But even Fox just trusts Rasmussen, instead of doing this important little bit of investigating and reporting. I think, just as with climate science, the "experts" are not what they are unquestioningly assumed to be.

Charles R. Anderson, Ph.D. said...

L. A. Henry and Harry,

Thanks for your comments. Too often, reporters seem to take the easy route. They report the unemployment number, but only occasionally note that the percentage of the population employed dropped even as the unemployment percentage dropped. They fail to note what kind of jobs people may have created for themselves, such as starting their own uncaptitalized business because they could not find work in the Obama economy.

Looking deeper into the polls and reporting the problems is similarly lacking. Of course, there is the occasional mention that some are claiming a problem, but not enough emphasis is given to acknowledging that to counter the impressions that most voters have. There is no strong impetus to see that voters know and understand what is going on.

Yes, Harry, we have seen the same problems with respect to global warming alarmist claims. The problems there are probably harder to explain though. The employment and poll problems are relatively simple to explain, however.