What is Going on? Reuters/C-SPAN/Zogby Poll: Obama by a Landslide. AP Poll: They're Tied!!

What is going on with this election? A Reuters/C-SPAN/Zogby poll says that Obama will win by a landslide 51.6% to 42%.

But as reported by the Drudge Report, there is an AP poll that says they're tied 44%-43%.

The Investors Business Daily says:

Contrary to other polls, some of which show Obama ahead by double digits, the IBD/TIPP Poll shows a sudden tightening of Obama's lead to 3.7 from 6.0. McCain has picked up 3 points in the West and with independents, married women and those with some college. He's also gaining momentum in the suburbs, where he's gone from dead even a week ago to a 20-point lead. Obama padded gains in urban areas and with lower-class households, but he slipped 4 points with parents.

The Wall Street Journal has a great article on why polls might be incorrect here. In part they say:

[T]he exit polls in almost all the primaries this year showed Mr. Obama doing better than he actually did. The same respondent bias -- the greater willingness of Obama voters to be polled -- which apparently occurred on primary days could also occur in the exit poll on Election Day, and in the phone polls of early and absentee voters that Edison/Mitofsky will conduct to supplement it.


Under the WSJ theory, Obama supporters are very excited to be polled and declare their support for him, while McCain supporters don't want to be bothered with answering a pollster's questions on the phone in the evening. (I certainly understand. I have turned down survey people most of the time when they call.)

The Financial Times of London offers this explanation for the difference in polls:

For a poll to be useful, its sample must match as closely as possible the demographic make-up of the voting population. This year, however, the traditional rules have been thrown into confusion by Mr Obama’s efforts to mobilise two groups that have been under-represented in previous elections: young people and African-Americans.

An aggressive registration drive by the Obama campaign and its allies has added millions to the electoral roll, but it remains uncertain how many will take part in the election.

Those polls giving Mr Obama his biggest lead, such as the CBS/New York Times polls that have shown him with a 13-14 point margin over the past two weeks, tend to be based on the most generous assumptions of higher Democratic turnout.

Polls that depict a tighter race, such as Reuters/Zogby and Rasmussen, which have consistently put Mr Obama’s lead in the low to mid-single digits, tend to be based on more conservative models that assume less change from previous elections.

Gallup, one of the most respected pollsters, has acknowledged the uncertainty by offering two separate estimates of likely voters in its daily tracking poll.

One takes into account past voting behaviour; the other uses estimates of an “expanded” turnout, so adding to Mr Obama’s lead.

One thing is sure. Polls don't decide who wins an election, voters do. That's why they play the game! Get out and vote.