The way the question is phrased it's almost a generic ballot question-- would you vote for the Democrat or the Republican?--and the 5% spread is the same as in the 2002 mid-terms.

That's another problem with the poll as reported. Although it is implied that this was the first question asked -- which would strengthen your point -- that is not clear.

Actually, considering all the Dean mis-statements as of late and the questions arising about his vunerability going into the Iowa and New Hampshire votes, I would think this poll was probably designed by Karl Rove to shore up any weakening support for Howard among the Democratic undesired.

When you say "surprisingly weak" re-elect numbers for most of this year, you do want to take into account that many presidents have been behind some time in the year before the election. Most notably, the last two presidents to be re-elected to a second term, Presidents Reagan and Clinton, were on the losing sides of such polls for quite a long period time each in the year before re-election.

So by historical comparison, it's not that surprising.

Years ago I heard some old, hard-boiled Washington reporter say that no poll that came out before the World Series meant squat because that's when most people (who have lives after all) start to seriously evaluate the mass of information that's been dropped on them. With Dean's recognition numbers so low, this is almost the same as the Bush vs. Generic Democrat polls we've seen recently. If it helps to keep the Republican base from getting complacent, so much the better.

Why is everyone citing this poll as though it were THE poll, when other polls taken just a week or so earlier show Bush far ahead of Dean. E.g., the ABC News/Washington Post poll from 10 days earlier shows the President with a 55%-37% lead over Dean, which is consistent with his 54%-39% lead in October and 56%-36% lead in September. These numbers are also consistent with the results of a CBS News poll taken a week before the Time/CNN poll, which shows the President with a 55-35% lead over Dean, including an enormous 58%-27% lead among voters who consider themselves "independent."

Says here that the over/under is 62-38. I am taking over.

The fairmodel site (http://fairmodel.econ.yale.edu/)predicts: "The new economic values give a prediction of 58.3 percent of the two-party vote for President Bush rather than 56.7 percent before. This does not, however, change the main story that the equation has been making from the beginning, namely that President Bush is predicted to win by a sizable margin. The margin is just now even larger than before."

Note that this model in the 2000 election: "The actual outcome was that Gore received 50.3 percent of the two party vote. The last prediction on this site (October 27, 2000), which used actual values for all the economic variables, was that Gore would receive 50.8 percent of the two party vote. The error is thus 0.5 percentage points. The standard error of the equation is 2.15 percentage points, and so the actual error is well within one standard error. The presidential vote equation thus did extremely well."

John -- "Weak" is my normative, editorial comment.

Michael -- Everyone, me included, is citing this poll because it shows Dean close to Bush. I might live to regret it, but I would like nothing more than for Dean to be the Democratic nominee, and this poll makes that more likely.

Carter -- I'm relatively skeptical of these predictive model, which are backward looking attempts to fit the equation to the facts. They're right, up until the moment they are wrong. As Professor Fair himself notes, in each election one prediction has to be closest, last time was his turn. I hope its true, as for relatively conservative estimates of GDP growth and inflation (3 and 1.5, respectively), it gives a Bush two party vote of almost 60%, but we certainly can't rely upon it.

Here is a terrific page for fans of polls:
http://www.davidwissing.com/gen2004polls.html

The latest one is from Rasmussen and it shows Bush leading Dean 51-37.

Bzzz -- Good site. My guess is the difference between the Rasmussen poll and the CNN/Time poll is that the Rasmussen number is all respondents while CNN/Time is "likely voters."

So carter are you over or under?