Well, the short answer is that the polls I believe are the ones where President Bush wins. Oh, silly, you say, jejeune perhaps? Not so.

We saw the anti-Republican poll bias in 2002, when the GOP in fact cleaned up on election day, we saw it again, hilariously, in the California recall. And we see it from the cheerleading media for Frenchie this time around. Brendan Miniter, Orson Scott Card, and many others write of this in the last days. Card makes the point which can never be stated too often: the lefties in the media absolutely do not see themselves as biased, they are living in a Ptolemaic universe, where the Democrats rule. They refuse to acknowledge as legitimate the real, Copernican universe, where the Presidency, the Senate, the House, and the majority of governorships and state assemblies’ members are Republican. Cover your ears and say “la-la-la-la-la” real loud. So they report these polls — that they create — with a straight face, because they really believe it.

So we get USA Today/CNN/Gallup reporting Kerry 50%, Bush 46%. The question for all polling is: what to make of the methodology. Few polls have the risibility quotient of that LA Times corker where dems were oversampled by ten points or so to produce a scientific result (Kerry Wins!). But let’s take a look at the USA Today disclosures, and see what we get.

Results for likely voters are based on the subsample of 706 survey respondents deemed most likely to vote in the November 2004 General Election, according to a series of questions measuring current voting intentions and past voting behavior. For results based on the total sample of likely voters, one can say with 95% confidence that the margin of sampling error is ±4 percentage points. The likely voter model assumes a turnout of 50% of national adults, consistent with recent presidential elections. The likely voter sample is weighted down to match this assumption.

What are we not told? We are not told the percentages of D’s and R’s in the likely voter sample, and we are not told what the results were before they were “weighted down.” What’s this weighting down anyway? If the likely voters were consistent voters in past elections, where on earth does the pollster get away with tamping down the voting ardor of a committed populace — or at least informing us of what the story was prior to this finagling.

You say: well, it could even be a bigger Kerry win if they told you all that. Perhaps. But, as Taranto points out, Kerry has two NYT reporters as campaign advisers (Sanger and Hoge), so don’t count on it.

Meanwhile, despite their best efforts to conceal, confession oozes from every pore. Exhibit A, an AP/Ipsos poll which is not a horse race poll. Much more interesting. I’ll quote it at length:

(QUESTIONS ASKED OF REGISTERED VOTERS)

* Bush, 50 percent
* Kerry, 44 percent
* Both, 1 percent
* Neither, 4 percent
* Not sure, 1 percent

Kerry’s signal advantage, such as it is, is that he is perceived as more intelligent generally than is Bush. He is, indeed, the king of all adjectives.

Bush’s advantages, on the other hand, come in categories that are, uh, absolutely critical to success for a President and a country: decisiveness and optimism. By 2 to 1 Bush is seen as decisive, where Kerry is evenly split (“I voted for the 87 billion; before I voted against it”). Slam dunk, as they say in the CIA.

But decisiveness alone could be misleading. That’s why the question of who makes you feel more optimistic is so critical; and on this Bush wins 50% to 44%. Plump this up to a two-man race by taking out the wafflers, and you get Bush 53%, Kerry 47%. I would make this my prediction for the general election, but it could be higher for Bush. Consider the methodology:

The Associated Press-Ipsos poll on the personal qualities of President Bush and Democrat John Kerry is based on telephone interviews with 1,000 randomly selected adults, including 804 registered voters, from all states except Alaska and Hawaii. The interviews were conducted July 5-7 by Ipsos-Public Affairs. The results were weighted to represent the population by demographic factors such as age, sex, region and education.

We really don’t know enough about the weighting of D’s and R’s to make a judgement, but I’d predict with confidence that the pollsters did not go out of their way to oversample the R’s.

This entry was posted on Wednesday, July 14th, 2004 at 7:22 am and is filed under General. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.