Obama Is Ahead -- Five to Seven Points

A Short Note on Reading Polls and Media Interpretation:

I wrote you last week that you should just ignore the tracking polls, which suffer from poorer execution than the one-time national polls that appear every week or so. Over the past two weeks, the onetime polls – CBS, CNN, Time, Pew, etc. – show Barack Obama with a very stable and consistent 5 to 7 point lead nationally over John McCain.

The tracking polls by contrast have consistently had the race between no points and a couple of points up for Obama over exactly the same time period, and the tracking polls show so-called “movement” that is not really movement at all, but simply shifts that are occurring because of the way the polls are executed.

The tracking polls receive more media attention because the so-called “closeness” and “movement” is interesting, even if it is not real at all. The so-called “movement” gives the talking heads something to theorize about – endless speculation that cannot be confirmed or denied by anyone, but sounds at least plausible. The talking heads, particularly those who wear Republican clothing, prefer the tracking polls because a close race is more newsworthy than a five to seven point stable lead.

My reading of this election at this point in August, weeks before the conventions, is as follows:

1. Senator Barack Obama enjoys a small, but quite stable lead over Senator John McCain in the range of 5 to 7 points, but his support still hovers below 50 percent with a significant undecided factor that certainly could decide the final outcome.
2. The so-called “movement” – and all interpretations of that movement – is sheer nonsense based on poorly executed tracking polls that have a harder time reaching people who are more likely to support Obama – the young under 35, African Americans, and Latinos.
3. There is a consistent “media bias” to over-interpret polling that shows the race close as opposed to polls that show the race favoring Obama. This is not a bias toward McCain per se, but a bias in favor of a race that is close because it is more news worthy and interesting.
4. Many of the so-called “explanations” for changes in the polling numbers are without any real foundation at all – sheer speculation lacking in substances and based on changes that are themselves suspect.
5. The regular, one time polls, with one notable exception from Gallup, are all fairly close and well within the normal tolerances of repeated polling. When a single poll comes out that is substantially at variance with most of the rest, you should just disregard it because consistently applied commonly used techniques produce results that are usually very consistent, within one to three points at most.

This race will continue to be close, and it is never going to be a blow out for Obama, despite the overwhelming antagonism toward the Bush Administration. I will send a longer blog on this subject tomorrow.

Just my opinion,

Gordon Black