Obama can't win an election, even if no one ran against him. Here's proof (?)

Editor's notes: In a Quinnipiac poll making the rounds, it seems that Obama cannot win a election against any unnamed Republican. At least, that is the way this poll will be used. We are profoundly conservative here at Midknight Review, but we need to be careful about our use of information. This is a case in point. There are two polls "out" this week. We have the Quinnipiac poll with the results of a no-name Republican opponent below. A second poll released by FoxNews is name specific. It shows Obama losing to Romney, Huckabee and Gingrich and tied to Palin. We do not believe that "no-name" or generic polls give an accurate view of an approaching election. In this case, the Quinnipiac poll confirms the FoxNews results. Often enough, however, specific polls, name versus name, produce different results than generic polls. We believe that a generic poll shows voter support of party affiliation more than a specific result - hence "generic" poll.

Quinnipiac:
Word format
A year after President Barack Obama's political honeymoon ended, his job approval rating has dropped to a negative 44 - 48 percent, his worst net score ever, and American voters say by a narrow 39 - 36 percent margin that they would vote for an unnamed Republican rather than President Obama in 2012, according to a Quinnipiac University poll released today.

This compares to a 48 - 43 percent approval for Obama in a May 26 national poll by the independent Quinnipiac (KWIN-uh-pe-ack) University and a 57 - 33 percent approval last July, just before the political firestorm created by opposition to his health care plan galvanized political opponents and turned independent voters against him. . . . . . READ MORE >>>

Point of post: to offer a caution as to the use of generic polls. While it may be humorous to argue that Obama couldn't win an election if no one was on the ballot, it is hardly a marker as to how he would do versus an actual opponent. In this case, one poll confirms the other, but, often, that is not the case. In the end, the only poll that really counts is the poll taken on election day at the ballot box. A major media poll on election eve, showed Bob Dole losing to Bill Clinton by 21 points when, in fact he lost but by a margin of only 8%. While the poll predicted the winner, it also influenced the election results, predicting a landslide victory that DID NOT occur. We will never know how many folks would have turned out for Dole if they believed he had a reasonable chance of winning the election. . . . . and with "major media polls," the point of the poll is as much about influencing election results as anything. The ends justify the means is often the motivation of major polls.

In a specific example, we have opposing polls with regard to the Nevada Senatorial election -- Sharron Angle versus Harry Reid. A Gallup poll shows Reid capturing the lead against Angle while the very reliable Rasmussen poll has Angle leading by 6 percentage points. Methodology and bias give us very different results.