October 14, 2004
Here are the links to the post debate survey results. ABC and Gallup once again re-interviewed previously contacted respondents immediately after the debates. CBS and Democracy Corps did the same using a nationally representative Internet panel (the Democratic aligned pollster put out a release & questionnaire), though the CBS sample was limited to "uncommitted voters" (bonus: a white paper on the Internet panel methodology from its vendor).
Although the initial coverage may not reflect it, the surveys were consistent: Controlling for initial partisanship, viewers thought Kerry was a better debater than Bush in all of the surveys. On both the Gallup and ABC surveys, Democrats judged Kerry the winner in larger numbers (86% for Gallup, 81% for ABC) than Republicans favored Bush (73% for both Gallup and ABC). Gallup also reported that Kerry supporters gave their man higher marks (89%) than Bush supporters gave Bush (80%).
The uncommitted voters in the CBS survey judged Kerry the winner 39% to 25%. Once again, the ABC, Gallup and Democracy Corps surveys showed Kerry with narrow but nonetheless consistent advantages among Independents:
- Gallup: Kerry "did a better job" 54% to 34%
- ABC: Kerry "won the debate" 42% to 35%
- Democracy Corps: Kerry "won the debate" among Independents "by six points"
Despite the small sample sizes, the consistency across surveys increases my confidence that Kerry's advantage among independent debate viewers was real. On the other hand, ABC noted intheir press release (not yet posted to the web) that "the independents who tuned in were more Democratic-leaning in vote preferences, breaking 52-43 percent, Kerry-Bush."
Of course, with the exception of the Democracy Corps survey, none showed significant changes in voter preferences, but that is the usual pattern. Obviously, since the first debate, Kerry has gained ground, but there is no telling whether that pattern will continue following the third debate. As always, the debate coverage is most important.
Related Entries - Debates
It seems to me it matters where they started. If ABC's sample accurately reflects the electorate, for example, then it doesn't matter much if Kerry wins the independents 42-35--he can't win if the electorate is 38% Republican. If Bush has managed to pull Republican party ID up that high, then the remaining independents are likely to lean Democratic.
Posted by: Thomas | Oct 14, 2004 4:22:05 PM
Zogby's daily tracking has Bush up 4. When you consider the party id weighting that seems to heavily favor Democrats (39-35) vs what many 2004 polls show (Republican edge), he must have gotten a boffo Bush number on most recent day.
Posted by: Eric | Oct 15, 2004 8:28:32 AM
"Controlling for initial partisanship, viewers thought Kerry was a better debater than Bush in all of the surveys."
Um, OK. Heck, I'm a conservative Republican and *I* think Kerry was a better "debater" (i.e. was more glib, facile, etc.) than GWB.
But I'm not sure that's very germane, since I don't find the "debates" a very useful means of determining who would make a better President.
In fact, I think the obsession with debates and debate performance is largely a /deformation professionelle/ on the part of the commentariat: since words and their manipulation are their stock-in-trade, they tend to focus on that as the determinant of Presidential effectiveness.
Here's what I see: after a non-trivial but hardly overwhelming surge toward Kerry, the tracking polls seem to have reverted pretty much to where they were. So I think the American people have digested the debates and found they learned nothing new about the candidates.
I would have been much more impressed with the notion that the debates "changed the dynamic" in any meaningful way if kerry were out campaigning (or at least buying ads) in Red-State America: as I recall, he pulled the plug on Arizona, Colorado, Louisiana, Missouri and a couple of others that had seemed at one time to be "in play", and--at least insofar as I know--has not reinstated his media buys there.
Conversely, GWB is playing in Kerry's space, contesting such states as Minnesota, Wisconsin and New Mexico which supported Gore in 2000. Seems to me that suggests who's winning far more than any number of polls! :-)
Posted by: David Hecht | Oct 15, 2004 4:49:50 PM
It is not correct to say that the debates did not change the dyanmic. Prior to that, in several polls, including Gallup, WashPost, Kerry was down 9-10 points. Now he's even or a point or 2 down in most polls.
Yes, GWB is conteting MN, WI, NM. Together those have only slightly more EV votes than Ohio, which Kerry is contesting. The game bolis down to OH, PA, FL. Whoever wins 2 of those states will very likely win the WH. Kerry has lead in most PA polls for a while. OH and FL are more up for grabs, maybe with a slight Bush edge. In any case, the election is still very much open.
Posted by: erg | Oct 15, 2004 7:11:13 PM
Erg, the most recent national polls show Bush up by 3, 1, 6, 2, and 4. The RealClearPolitics average over the past few days is Bush by 3.4. The Bush average lead has been rising since the last debate.
It's not sure what is happening in the states, save that state polls tend to lag the national polls by several days at least.
Looking at the RealPolitics summary of state polls the overall summation seems to be that Bush has a narrow lead in Florida. Kerry has a narrow lead in Pennsylvania and a knife-thin lead in Ohio. In all three states the most recent polls show Bush gaining in Ohio and Florida and closing on Kerry in Pennsylvania. Bush is looking decent in Wisconsin and Colorado (and the referendum to split Colorado's electoral votes may be trailing).
Kerry looks out of the woods in New Jersey and Michigan and decent in Minnesota. Kerry's leading by a hair in Iowa.
what does this mean? Who knows? The margin of error means either man could be leading or trailing. even so if I were Kerry I'd be disturbed by the recent trend. Were I'd Bush I'd be mildly encouraged.
Posted by: Don | Oct 16, 2004 10:56:56 PM
The comments to this entry are closed.