TNR BLOGS

November 20, 2008 | 11:23 AM
November 20, 2008 | 10:57 AM
November 20, 2008 | 10:48 AM

November 20, 2008 | 12:03 PM
November 20, 2008 | 10:45 AM
November 19, 2008 | 11:20 PM

July 26, 2008 | 2:24 PM
July 23, 2008 | 1:55 PM
July 17, 2008 | 3:56 PM

November 20, 2008 | 11:06 AM
November 19, 2008 | 3:17 PM
November 19, 2008 | 2:17 PM
COLUMNISTS
TODAY'S STORIES
19.02.2008
New Ohio Poll: Clinton Up 9; They Love Her in Columbus

SurveyUSA has Clinton leading 52–43 (down from 56–39 a week ago). More interesting, though, are the crosstabs. SurveyUSA has devised a brilliant new way of slicing and dicing the electorate by age: among voters younger than he is, Obama leads 51–45. But he trails Clinton among voters older than he but younger than McCain (54–39), and is getting trounced among the tenth of voters older than McCain (69–24). Remarkably, Obama leads among all voters under the age of 50, and is still losing by nine points. That tells you something about the depth of Clinton's support among the Centrum Silver crowd. There's also, once again, an enormous gender gap (women: Clinton 62–34; men: Obama 55–39).

In addition, there's some strange geographic variation--Obama's winning by seven points in the Cincinnati area, but losing by a full 30 points in the Columbus area (this despite the fact that Cincinnati has only marginally more African-Americans than Columbus, and actually has more senior citizens too). In Cleveland and Dayton it's effectively tied, while Clinton's way ahead in Toledo and rural southeastern Ohio. Perhaps commenters with more knowledge of Ohio political demography than I have might shed some light on what's going on.

--Josh Patashnik 

Posted: Tuesday, February 19, 2008 6:38 PM with 16 comment(s)

Comments

You must be logged-in to comment.

Not a subscriber? Click here to get a digital or print and digital subscription to The New Republic!

primwallflow said:

Does anyone know how a candidate's internal polling is different from the analysis we see in public polls like this one? For example, would all of Obama's past internal polls in Ohio have cross-tabbed by city? It makes sense to me that they would have, but then, why would these public polls not go into as much depth? Is it simply because you need a larger sample size to effectively slice up the data in so many different ways?

February 19, 2008 7:09 PM

virginiacentrist said:

Speaking of Ohio:

We have another potential winner for "worst campaign video of the cycle":

www.youtube.com/watch

Warning - this is painful to watch. Also - the fun stops at about 34 seconds in.

February 19, 2008 7:35 PM

teplukhin2you said:

IIUC Columbus has an enormous lesbian population (seriously). When I first saw the headline of this blog entry my immediate thought was that the author was making a little joke about the Subaru Sisters in Columbus.

February 19, 2008 7:37 PM

blackton said:

why don't these baby boomers just die already?! Or at least just all move down to Florida or Arizona, or wherever all the old hippies go.

Since this is an open primary, is this polling just likely Democratic voters or all of Ohio's voters. If the Republicans come out, then Clinton is sunk because I just can't see many Republicans wasting their vote on McCain and letting Hillary come out on top. Kill her now when they have the chance seems to be the best choice. Drive a stake into whereever it is she might have a heart and put her down so as to not risk her being within the remotest shot of becoming President.

February 19, 2008 7:42 PM

virginiacentrist said:

Tep:

As the saying goes, I guess we'll have to dip him in honey........

February 19, 2008 7:47 PM

prnoonan said:

This is about the opposite of what I would have figured.  Im baffled.  In my likely uninformed opinion, Columbus would be the best of the three cities for Obama -- it has a huge university and the residents (again, my impression) are more educated and affluent (white collar / tech jobs, e.g.).  Cleveland has a big black population, but otherwise it's a lot of catholic white ethnics, union members and seniors.  Plus to the extent there's still a machine there, it's for Clinton.  And Cincinnati I viewed as a quasi-southern city that has more similarities to Louisville than Cleveland.  For him to be tied there is remarkable.

And I don't buy the lesbian theory; I'm sorry there's no way.

February 19, 2008 7:48 PM

adamvaught said:

vc,

I can't believe that video. It's like a Saturday Night Live skit come to life.

February 19, 2008 7:56 PM

RevMom said:

Columbus is the home of Ohio State University, the largest University campus in the country.  The polling there probably does not reflect the potentially huge influence of OSU voters.  

Toledo is my home town.  Obama is still relatively unknown there but the enthusiasm gap is already apparent.  30 people showed up to the grand opening of Clinton's campaign office; on the same day 150 people showed up for the opening of Obama's office.  I'm guessing the tide will turn at least somewhat in Toledo.

February 19, 2008 7:59 PM

Rhubarbs said:

I don't know from Columbus, but a Hillary lead in Toledo would seem to indicate strength among "white ethnics."

February 19, 2008 8:00 PM

virginiacentrist said:

just in case people didn't get my wonderfully PC joke, here's the whole saying:

"Dip me in honey and throw me to the lesbians"

February 19, 2008 8:14 PM

sullydog said:

Sounds like Obama may be leading in delegate-rich regions, but losing the state overall. Anybody have anything on this?

February 19, 2008 8:16 PM

teplukhin2you said:

VA - what's the rest of it? Does this follow, "If I lose my bet with you, then..."?

February 19, 2008 8:21 PM

ackyri said:

My money's on the reverse Bradley effect, which we've seen a lot of this cycle.

March 4th is going to be her last stand.

February 19, 2008 8:33 PM

gqri05 said:

"Does anyone know how a candidate's internal polling is different from the analysis we see in public polls like this one?"

Yes -- the candidates' internal polls are usually much more extensive than what is found in these public polling crosstabs.  For one, they typically ask more questions, both political and demographic.  With a large enough sample size, campaign pollsters can break down support by region, as well as by age and gender within each region, as well as a whole host of other combinations of groups.  

They know exactly where their support is and where they can make up ground.  Public outfits that release crosstabs are providing a useful glimpse as to the basic breakdowns of the race, but that's all -- a glimpse.  

February 19, 2008 8:39 PM

stgla said:

Chance differences.  When you take a sample of 700 likely voters and split it six ways, the margin of error goes way up.  It is somewhat irresponsible to report crosstabs without some interpretive statistics.

February 19, 2008 10:04 PM

benjamin81 said:

Give it a week. Obama has historically improved markedly in the polls once his campaign targets a state seriously. But I agree with stgla - the numbers are just too low to draw a good comparison.

February 20, 2008 11:23 AM