« MN: Slight Kerry Lead | Main | Electoral College Reform Revisited »
Monday, May 31, 2004
OH: Six-Point Bush Lead
Posted by DavidNYCI think I might be a bit late to the party in reporting on this poll, but the Cleveland Plain Dealer offers us some bad news:
Bush: 47
Kerry: 41
Nader: 3
Undecided: 9
(MoE: ��2.6%)
One unusual thing about this poll is the very large sample size (1,500), which yields, of course, a very low MoE. Other than that, I'm not sure what to say. The most recent prior poll by ARG showed Kerry with a lead of this size, while a trio of polls taken back in March showed a much closer race.
A quick side-note: Some people have asked about alternate empirical prediction methods, to contrast with Chris's GECC. David Wissing (who invaluably compiles all the state polling information in one place, as linked above) uses a very simple method: He awards a state's EVs to the winner of the most recent poll in that state, and where polling is unavailable, he uses the 2000 results. Right now, that gives Bush 296 EVs to Kerry's 242. David, by the way, is a Bush supporter - but his methodology, of course, is purely objective.
UPDATE: Brian in the comments points us to the Ohio Democratic Party's response to this poll. It's nice to see such responsiveness on the official level. Back in November, I asked the New York State Democratic Party if they were going to start a blog. I was told to take a hike. So typical. Sigh.
Posted at 04:07 PM in Ohio | Technorati
Comments
The Ohio poll, as well as the Minnesota poll described in the entry below the Ohio summary, were done by Mason-Dixon.
If you look at the following compilation of state-by-state polls:
http://realclearpolitics.com/bush_vs_kerry_sbys.html
there is some evidence that Mason-Dixon polls come out more favorably to Republicans than do other firms' polls. I do not want to overstate the case. Two cautionary notes are that not all polls within a given state were taken at the same time, making them less than optimally comparable, and that I've also seen other Mason-Dixon polls that match other firms' polls. Having said that, there appear to be at least four states where Mason-Dixon is out of line with other polls:
MINNESOTA -- The 41% reading for Bush in the recent Mason-Dixon poll is consistent with other polls, but the 44% for Kerry is below other polls, resulting in the close 3% margin.
NEVADA -- Again, the potential noncomparability due to when the different surveys were taken must be acknowledged. However, Mason-Dixon's Bush +11 is out of line with the other polls.
OHIO -- The Mason-Dixon poll (Bush +6) is way out of line with two other May polls that have Kerry +4.6 and +7.
FLORIDA -- Every recent poll has had Florida a point or two in either direction. You'll notice a Tampa Tribune poll that had Bush +8. Although the link to get more information on this poll does not appear to work, I've confirmed via a Google search on "Mason-Dixon" "Tampa Tribune" that Mason-Dixon has polled for the Tampa Tribune in the past, and thus is likely to have conducted the Bush +8 poll in question.
Whether Mason-Dixon has a systematic pro-Republican bias is something that will require additional data to document. However, one rule I always follow is that, if a Mason-Dixon poll shows a Democrat with a comfortable lead, such a Democrat must be in GREAT shape.
Posted by: Alan at May 31, 2004 05:11 PM | Permalink | Edit Comment | Delete Comment
Living in Illinois, I am aware of two seperate polls that show far different results. In the presidential race the Chicago Tribune has Kerry with a lead of 53-37. Rasmussen has Kerry ahead by only 48-43 in Illinois. In our Senate race the Tribune had Obama up 52-30 while Rasmussen had Obama up 48-40. In 2000 President Gore carried Illinois by 12 points. It doesn't make sense that national polls show Kerry up by a couple of points but behind in Ohio by 6! President Gore lost Ohio by only 3.6 points.
Posted by: Alan Snipes at May 31, 2004 05:33 PM | Permalink | Edit Comment | Delete Comment
the question for me is which state Kerry can win this year. I am sure Kerry can hold on to Gore's states but Kerry needs to win at least 1 or 2 red states. Which one ? I can't see Kerry win OH honestly. Kerry has better chance in FL than OH then you look at smaller states ie NH, WV, AZ, AR.
Can Kerry win any of those small states ?
Posted by: ben at May 31, 2004 07:39 PM | Permalink | Edit Comment | Delete Comment
No need to panic over the Sunday Plain Dealer's screaming headline: Bush Leads Kerry in Ohio." Its big page one graphic showed Bush with a 6-percent lead over Kerry -- 47 to 41 percent.
Now hold on one sec....
The PD's Mason Dixon poll, with a 2.6 percent +/- error factor, was taken May 20-25.
In contrast, the Zogby poll of Ohio likely voters from May 18-23 gave Kerry a 4.6 percent lead with an error factor in the same range as the margin.
Could things have switched so dramatically last Monday and Tuesday, May 24 and 25?
Nope, it's just that horse race polls can be up or down in such a dead-heat race -- with events careening around and Kerry still working to define himself to most Americans.
"Recent polling in Ohio has fluctuated and it is still very early to rely on any one particular poll," Jennifer Palmieri, Kerry spokeswoman in Ohio, told the PD. Further proof of polling fluctuation: Kerry led Bush in Ohio by 49 to 42 percent in a poll taken May 10-12 by American Research Group.
This Dog Blog of May 14 told you of national polls in early May that showed Kerry leads like this:
CNN Time -- Kerry 51 to Bush 46
Pew Research Center -- Kerry 50 to Bush 45
Rasmussen Reports -- Kerry 47 Bush 45
A week ago, John Zogby was saying his numbers show "it's now Kerry's race to lose." No, it's probably more like what Scott Rasmussen of the Rasmussen Report keeps saying. It's a dead heat with the lead changing hands back and forth for 10 weeks in the daily Rasmussen polling -- all within the Rasmussen margin of error.
The main message from many polls is that Bush approval ratings are in free fall, but Kerry still needs to make the case for himself. The PD's poll showed that 62 percent of those voting for Kerry cited as their No. 1 reason their dissatisfaction with Bush. Other recent national polls have shown double digit percentages for voters "undecided" about approval or a favorable opinion of Kerry. Those "undecided" percentages on Kerry are:
Zogby -- 21
Quinnipiac -- 16
Newsweek-- 14
Annenberg -- 21
CBS News -- 35
As Ms. Palmieri told the PD today, "These numbers reflect an overall dissatisfaction with Bush and an opportunity for John Kerry. We have to work hard to make sure people know who John Kerry is."
To attract on-the-fence voters, Kerry is launching in June a new round of biographical TV ads to introduce himself in Ohio and other battleground states. Look for Bush to keep running negative ads to try to define Kerry his way.
The Wall Street Journal's John Harwood and national Blogger Taegan Goddard (www.politicalwire.com) are speculating that 2004 is beginning to resemble the election of 1980. Wrote Goodard:
"In many ways, the presidential race mirrors the 1980 campaign when incumbent Jimmy Carter faced declining approval rates yet the public was unsure of Ronald Reagan, a candidate viewed by many on the ideological fringes of his own party. Reagan's challenge then -- to make himself more appealing to the average voter -- is the same as Kerry's today."
Well, we all know how 1980 turned out.
(From Blue Dog Rising at www.ohiodems.org)
Posted by: Brian Usher at May 31, 2004 07:40 PM | Permalink | Edit Comment | Delete Comment
Brian,
I didn't know Ronald Reagan.
Ronald Reagan was not a friend of mine.
But Brian...
John Kerry is no Ronald Reagan!!!
:)
Posted by: Jeremy at May 31, 2004 08:11 PM | Permalink | Edit Comment | Delete Comment
No, Kerry is no Ronald Reagan, and that's a damn good thing ;)
Posted by: Greg Pratt at May 31, 2004 09:01 PM | Permalink | Edit Comment | Delete Comment
The 4/1 Florida poll is indeed by Mason Dixon, as confirmed by this dKos entry:
http://www.dailykos.com/story/2004/4/5/18340/88380
Posted by: Ed Fitzgerald (unfutz) at May 31, 2004 11:11 PM | Permalink | Edit Comment | Delete Comment
I am sure Kerry can hold on to Gore's states but Kerry needs to win at least 1 or 2 red states. Which one ? I can't see Kerry win OH honestly. Kerry has better chance in FL than OH then you look at smaller states ie NH, WV, AZ, AR.
I wouldn't give up on Ohio so easily. Kerry was ahead of Bush (by inside margins) in the two most recent polls before the Mason-Dixon. He can also pick up NH (Zogby has him ahead by an outside margin of 9.6, the Granite State poll on 4/26 had him ahead by 4 points, just inside the margin of error, and Rasmussen had him ahead by 2 points on 4/21). He's extremely unlikely to pick up WV, AZ or AR, and *may* have trouble holding on to IA and NM.
Posted by: Ed Fitzgerald (unfutz) at May 31, 2004 11:19 PM | Permalink | Edit Comment | Delete Comment
David Wissing (who invaluably compiles all the state polling information in one place, as linked above) uses a very simple method: He awards a state's EVs to the winner of the most recent poll in that state, and where polling is unavailable, he uses the 2000 results. Right now, that gives Bush 296 EVs to Kerry's 242. David, by the way, is a Bush supporter - but his methodology, of course, is purely objective.
He doesn't use all available polling data, though, since I don't see the publicly available polls comissioned by partisan organizations (Dem and GOP), or the recent battleground polls by Zogby. His e.v. totals would be quite different if he did.
Posted by: Ed Fitzgerald (unfutz) at May 31, 2004 11:39 PM | Permalink | Edit Comment | Delete Comment
Ed is right about not giving up on Ohio so easily -- and not just because of the conflicting polls. Ohio Dems are very hungry and showing it in several ways -- including primary turnout and record $ at a statewide dinner in May with John Edwards speaking. Also, the state is in bad shape and the long-dominant GOP is cracking up with internal feuds among officeholders leading to feds investigating from Justice and IRS. Kerry just can't (and won't) do what Gore did in 2000 -- pull out of Ohio, drop ads 5 weeks out from the general election.
Posted by: Brian Usher at May 31, 2004 11:58 PM | Permalink | Edit Comment | Delete Comment
Ed - all I can say is you'll have to take it up with Wissing. I can understand a reluctance to use partisan polls. I prefer not to blog about them on this site. And serious questions have been raised about Zogby battleground poll's methodology. (Which was apparently done over the Internet.)
Posted by: DavidNYC at June 1, 2004 12:21 AM | Permalink | Edit Comment | Delete Comment
Ed - all I can say is you'll have to take it up with Wissing. I can understand a reluctance to use partisan polls. I prefer not to blog about them on this site. And serious questions have been raised about Zogby battleground poll's methodology. (Which was apparently done over the Internet.)
Yeah, not using partisan polls is understandable, but Zogby's standing behind the Internet polling, so if they're willing to put their name on it, how does it differ from their other polls?
Posted by: Ed Fitzgerald (unfutz) at June 1, 2004 01:35 AM | Permalink | Edit Comment | Delete Comment
I don't worry about IA or NM because of Richardson and Vilsack.
In order to win, Kerry must win 2 small red states or maybe a big one OH or FL. NH and AR are possible if Kerry spends money there. I am dumbdounded as to why Kerry wanted to spend his money in VA. or maybe Kerry already conceded that he can't win other red states so VA is another option based on Warner's advice.
Posted by: ben at June 1, 2004 02:24 AM | Permalink | Edit Comment | Delete Comment
VA is winnable, but a long shot. Ohio is not a lost cause by any measure. The jobs situation is Ohio is bad and getting worse. And end to hostilities in Iraq won't help that sitatution out at all. Ohio went for Clinton and Gore narrowly lost it, so we know it's within reach. Kerry will probably wind up winning Ohio in the end. Also, I think Kerry will win Florida as he's been holding up well there and Dems will be sure to turn out in force in November. AR, NC, WV, TN, GA, MO are all probably out of reach unless Kerry picks Edwards or Clark for VP and they really put on a show with the southern folk and get them excited about voting Dem this fall. It will be a close race, but in the end the numbers indicate Kerry will win this fall.
Posted by: Rock_nj at June 1, 2004 03:04 PM | Permalink | Edit Comment | Delete Comment
Don't be fooled by those "Margin of Error" numbers. That's assuming that you truly do have an unbiased sample of the population you're trying to measure, and few pollers do. The wonderful Prof. Pollkatz has, among his other goodies, this overview of poll biases.
http://www.pollkatz.homestead.com/files/bushindexprobushtable.htm
I don't know what Ipsos-Reid is or why Pollkatz set his zero there. That may or may not equate to "unbiased". But it's interesting to see the size of the spread and the consistency of the bias.
If you're a real data geek, you'll love the scatter plot that shows every one of these polls vs. time.
http://www.pollkatz.homestead.com/files/pollkatzmainGRAPHICS_10331_image001.gif
Posted by: rppa at June 1, 2004 05:36 PM | Permalink | Edit Comment | Delete Comment