« Fun With (Medicare) Math! | Main | Has Clark Got His Mojo Working? »

The Iraq Situation and the 2004 Vote

The latest NPR poll, conducted by Greenberg Quinlan Rosner/Public Opinion Strategies, has a few interesting findings that deserve to be highlighted. First, the generic presidential ballot–Bush vs. our trusty unnamed Democrat–has changed dramatically since their poll in late May. At that point, Bush was leading by 15 points (50 percent to 35 percent); now he’s leading by just 3 points (44 percent to 41 percent).

That’s consistent with trend on most other public polls. But what’ s interesting here is that they broke down the late May and current poll samples by those in states Bush won by 5 percent or more, those in swing states and those in states Gore won by 5 percent and more. This exercise shows that all of the move toward the Democrats over this period has been in swing states (from +19 for Bush to dead-even) and in Gore states (from dead-even to +13 for the Democrats). The Bush states haven’t budged (+22 for Bush in May, +23 for Bush today).

One reason for the pro-Democratic shift over this time period is the rise in salience of the situation in Iraq. Just in the last couple of months, the number citing the situation in Iraq as one of the two top issues that will influence their presidential vote in ‘04 has doubled (from 14 percent to 28 percent). And those citing Iraq favor the Democrats in the generic presidential ballot by 29 points.

Those citing a number of other areas also favor the Democrats: education (by 25 points); affordable health care (by 21 points); the federal deficit (by 20 points); Social Security/Medicare (by 14 points); and the economy and jobs (by 12 points).

But it seems clear that the higher the voter salience of the Iraq situation, the better the Democrats are likely to do in November, 2004. Not exactly what Rove and co. had in mind (see DR’s post on “Plan A Falls Apart” for more discussion).


Holy cow, did they ever undersample Democrats!

Look at page 19, in which they refer to respondents' votes in 2000 to correlate with their attitudes today. Of their respondents, 51% voted for Bush and 37% for Gore. That's a pretty significant variance from reality!

Otherwise, this is one of the more informative polls I've seen. Thanks, DR!

OK, for those of us who aren't pollsters, which states are considered swing, Dem, or Bush? It would make a useful sidebar. Thanks

Check out http://swingstateproject.typead.com/ for a rundown of what the swing states are.

This is very interesting. It demonstrates, once again, that Bush is a very polarizing figure. But the polarization is very negative for Bush because it indicates that those who were on the fence about him in 2000 are generally falling away from him as time goes by.

He's uniting the swing voters with the Gore voters. Good job Dubya!

Here's the correct address for the Swing State Project (previous comment was missing a 'p'):


So what areas do Bush-supporting voters cite? Tax cuts? War-on-terrur?

Ian: to my knowledge, yes, war on terror. And don't forget likeability factor. He's a person of integrity, honesty, blah, blah, blah.

Gotta like those numbers, but I can't shake the very solid feeling that national security is still going to be a problem unless we really get our act together and present a strong, consistent message on all the foreign policy issues, including Iraq. We simply can't afford to re-fight the Iraq war in any way, with the "it could be OK without so much dishonesty and incompetence" camp pitted against the "it was never going to work no matter who did it" camp (I'm in the former one). Let's all focus on what we have in common on national security, which is a much broader set of ideals and policy preferences than the ones that divide us. To take just one example, witness the similar stances on N. Korea for both Dean and Lieberman. Heck, aside from the (admittedly very large) Iraq issue, is there any significant foreign policy issue where there's a really noticeable level of dissension within our ranks? I don't think so. Let's pull together and get people trusting us to take care of world affairs, so that they'll be less likely to be swayed by the constant demagogic appeals to the reactionary hypocrisy that passes for Republican foreign policy.

That's an interesting pickup, Chris.

One thing that may explain it is that people often misremember who they voted for in the last election - in particular, it commonly occurs that some number of people who claim to have voted for the winner of the last election in fact must have voted for the other candidate. We know this because the gap you noticed frequently appears in survey results.

Now whether the gap in this poll can be explained that way I'm not sure; best you could do would be to study the poll's methodology and see whether there's an obvious bias/gap. But the order of magnitude suggests there may be no underlying bias.

Which is itself interesting; does Ruy or anyone know whether pollsters believe people only overreport voting for the winner of the last election - or whether people sometimes underreport voting for the last winner, perhaps because they've decided the current pres is a bum?

Do these numbers mean much given the electoral situation? Until these start getting broken down by states, it just seem like wasted energy following these kinds of numbers.

Just a thought. I could be wrong, of course.