MyDD :: New Polling Developments

In this post I will try to cover several issues concerning the horse race numbers. First let look at Ohio. Rasmussen has McCain ahead of Obama by 10. The number does not run against Rasmussen other Ohio polling. Without pressing, the lead is 6. Prior monthly polls have the lead at 1 or 2. So essentially the shift is a 4 point shift - pretty much between the MOE range and Obama having lost some of his luster. Nevertheless, has anybody noticed how Obama can increase his popularity with a notch at any given time? BTW- Rasmussen is the only outfit that shows McCain consistently ahead of Obama so his weighing is obviously different than others. An intesting question would be why leaners trend towards McCain- an obvious answer is that while the Republican party has suffered loses in affiliation, the new independents lean conservative and McCain is a good fit for them. I think polls this year are inaccurate because of the Bradley effect and the new voters surge; however, it is the only thing we have - the RCP average is specially helpful as it incorporates all the adjustmensts done by the different polling outfits in this difficult polling season. I think Ohio would ultimately go McCain becasue of Appalachia unless 1) it is a landslide (which is a big possibility) or 2) Hillary is in the ticket (my best guess is yes).
Virginia is true a tossup (I live here) but I think McCain will prevail because of the Bradley effect since it clearly happened once before(Wilder).
Florida is in play only if Hillary is part of the ticket. Rasmussen now has Obama ahead by 1.
Ultimately, the election will be decided in Colorado where Obama has a strong advantage in structure, money and demographics. I do not foresee Michigan flipping at all.
The party that wins election would be the party that has won the most the last 100 years. Right now both parties are tied 48-48.
One more observation, after 7-8 months of running one would have thought the McCain operation would have been smoother. There is no reason to think they will improve in the next three months and will probably continue campaigning negatively.

Permalink :: 19 Comments :: Post a Comment

Re: New Polling Developments (none / 0)

I also live in Virginia. The Wilder election was 19 years ago, and I think the Bradley effect has faded. It certainly didn't exist in the Virginia primary (where it would have been clearly apparent).

Ohio, Florida, Virginia, Indiana and/or Missouri would be gravy. If Obama takes the Kerry states plus Iowa, Colorado, New Mexico, he's elected (even if he were to lose New Hampshire). And he's well ahead in all of them. Just in case, though, one of those five states would be nice...

by feynman on Wed Jul 23, 2008 at 02:21:21 PM EST

Not automatically (none / 0)

The Kerry states plus CO, IA and NM but minus NH would result in a 269-269 tie.

by conspiracy on Wed Jul 23, 2008 at 05:15:00 PM EST
[ Parent ]

Ohio is bizarre (none / 0)

I don't get it. What happened to provoke the change? You can't say Appalachia because Appalachia was there when Obama was more popular there. So why the downward trend?

by catfish2 on Wed Jul 23, 2008 at 02:30:16 PM EST

Independents (none / 0)

who supported Democrats but are now turning back to McCain on the war and the media's constant barraging of "Who is Obama?" without ever helping explain who he is.

The downward trend leads me to believe that Clinton would've faced the same problem.

The American people; they were for the war before they were against it.

by nrafter530 on Wed Jul 23, 2008 at 03:05:31 PM EST
[ Parent ]

They know who Clinton is (none / 0)

She is pretty well-defined. They would have tried, and probably would have used different tactics.

by catfish2 on Wed Jul 23, 2008 at 03:59:40 PM EST
[ Parent ]

yeah (none / 0)

and her one problem was that many thought she was untrustworthy. McCain and the Republicans would be slamming her left and right on the trust issue.

I don't think she would've been able to hold up. She would've also amended her Iraq war position and probably been hit as a flip-flopper. It would've stuck to her too.

The American people; they were for the war before they were against it.

by nrafter530 on Wed Jul 23, 2008 at 07:11:13 PM EST
[ Parent ]

Re: yeah (none / 0)

She still polls better against McCain.

by killjoy on Wed Jul 23, 2008 at 08:45:37 PM EST
[ Parent ]

Speculative (none / 0)

she hasn't been polled in over a month.

The American people; they were for the war before they were against it.

by nrafter530 on Thu Jul 24, 2008 at 09:59:51 AM EST
[ Parent ]

Re: Independents (none / 0)

Not so. She outflanked Independents with support to their right and to their left. That's why she polls better than Obama against McCain.

by killjoy on Wed Jul 23, 2008 at 05:26:26 PM EST
[ Parent ]

Re: New Polling Developments (none / 0)

Sorry for the embarassing typos - one look is not enough. On Ohio- I though Hillary won there? No?Besides-the obvious point - different electorates between the primary and the general.

by RAULC on Wed Jul 23, 2008 at 02:48:51 PM EST

Rasmussen's bias? (none / 0)

I've heard it said often that Rasmussen has about a 3%-5% Republican bias in their polling schemes. I don't know this first-hand, but it seems accurate from what I've noticed over time.

Regardless, one poll does not a trend make. I wouldn't go off the handle and base any major decisions on this.

In this avalanche, the pebbles get to vote.

That One/Another Fella '08

by Dracomicron on Wed Jul 23, 2008 at 02:50:50 PM EST

Re: Rasmussen's bias? (2.00 / 1)

One place to see this is in their job approval numbers for Bush which tend to be about 5 points above the average of other polls.

Yes, they do have a Republican bias.

One reason is that they do not do call backs. That means that they only include phone numbers where someone is home the first time they call. They don't retry those same numbers other days. This means that the polls skew toward the sorts of people who are home more often.

Also, they take their results and then weight them so that there is a fixed percentage of Republican/Democrats/Independents. Since the number of Republicans has been declining, this overrepresents Republicans.

We care about politics because we know politics matters for people's lives and opportunities.

by politicsmatters on Wed Jul 23, 2008 at 02:57:51 PM EST
[ Parent ]

Re: New Polling Developments (none / 0)

There tend to be three possible roads to a Democratic Presidential win. New Hampshire is not in play, btw.

1) Ohio

Florida
Wisconsin-Iowa-New Mexico-Nevada and/or Colorado

Missouri, West Virginia, and Virginia are very close but imho still on the wrong side of 48/51 and 49/51 splits. So is Arizona, though the margin there depends on amount of effort- the state Democratic Party and its turnout machinery just aren't solid afaict.

Just look at it in a big picture way, the "landslide" a lot of Obama boosters were crowing about not three or four months ago is not going to happen. McCain may not be much, but the Republican coalition is holding together. The conservative leaners, moderate Republicans, social conservative Christians, and the militarist/opportunist sorts have decided Obama isn't getting their vote.

So the bad news is that Obama is operating under the 52% national ceiling of Democrats and liberal/Democratic leaners; he has lost or is losing what little support he has outside that 52%. (Clinton maxes it out because of diminished opposition of people on the other side.) The good news is that that ceiling is about 4% higher than it was in 2004.

So what to make of the three routes to the Presidency....

-Florida. First, Obama neglected to do what it would take for him to win the state pre-election. That was to put hard political effort into the reenfranchisement of the many (hundreds of thousands of) black men that can't vote due to felony conviction. His subsequent black, young, and male identity politicking against Clinton has now cost him the state, which has a mere 51%ish ceiling for Democrats this year.

Florida would be a squeaker win in the best case and writing it off is not fatal. But it's a demonstration of the Axelrod math and Obama's northern Middle American domestic political judgment having some built-in misassessments. It's a case where being less centrist and less identity politicking probably would mean a better result.

-Ohio. I think the Obama identity politicking is likewise showing a backlash. He might recover from it, though. It would take a lot more emphasis on real leadership in managing economic affairs and implementing good policies. And probably making public pledges that answer the white blue collar worker suspicion or objection to him which is at this point unspoken- just in what fashions and to what extents he will pay back his black supporters.

-Wisconsin/Iowa and Colorado/New Mexico/Nevada. I think Obama's political sense is calibrated for the first two states, and his Illinois organization and collaborations and money there will ensure that those are no problem for him. Though the margins will not be that great.

The hard part will be how to persuade the swing voters on which all depends in Las Vegas, Phoenix (somewhat), Albuquerque, and Denver and Pueblo. I'll guess that a negative, anti-McCain and anti-Republican approach might be the only thing that works, and that just barely.

I think Axelrod & Co. will decide to go two ways at once- secure Iowa and Wisconsin, focus the campaign on Colorado and Ohio, and keep the desert Southwest as a second tier. Missouri, Virginia, and North Carolina will be kept in play for as long as possible to keep McCain in trouble, and Florida and some other small states as long as it doesn't cost much to do so.

by killjoy on Wed Jul 23, 2008 at 05:21:19 PM EST

A tad too pessimistic I think (none / 0)

I think we can forget about West Virginia and Arizona what with the latter being the home of our opponent (though I think AZ will closer than the CV presumes). I don't think Wisconsin and Iowa will be particularly close. I think New Hampshire might be but that it goes to Obama in the end. People are reading too much into one poll in Ohio while the numbers coming out of Florida seem to be consistently more encouraging than was expected. I think there will be enough resources to compete wherever the campaign believes they can be competitive. Virginia is more likely than Missouri. Nevada looks promising with all those new registrations but it isn't enough alone to get over the top. The firewall (Kerry states plus IA, NM, CO) though looks to be holding.

by conspiracy on Wed Jul 23, 2008 at 06:34:39 PM EST
[ Parent ]

Virginia is the crucial state (none / 0)

In my opinion. If the senator does not win Virginia, I doubt he will carry Ohio. Although he may still be able to build a majority by winning Colorado, Indiana and Montana.

by ann0nymous on Wed Jul 23, 2008 at 05:34:09 PM EST

Re: New Polling Developments (none / 0)

This won't be close. Stop worrying so much.

by QTG on Wed Jul 23, 2008 at 06:03:59 PM EST

Possibly (none / 0)

But anything could yet happen. July polling is historically a poor prognosticator. I think what we can say however is that if Obama does not win either Colorado or Virginia or Ohio (at least one of these) McCain will be president.

by conspiracy on Wed Jul 23, 2008 at 06:25:35 PM EST
[ Parent ]

Re: Possibly (none / 0)

New Mexico, Nevada, Iowa plus Kerry states gives 269 w/o Colo, Va or OH. According to 538.com a 269 result would lead to Obama.

by RAULC on Wed Jul 23, 2008 at 06:56:42 PM EST
[ Parent ]

Theoretically yes (none / 0)

But do you trust the House to do the right thing? Never mind the issue of a mandate.

by conspiracy on Wed Jul 23, 2008 at 07:25:00 PM EST
[ Parent ]