Yet More On That South Carolina Poll

A veteran observer of Democratic politics points out:

** Four of the last six polls gave Sen. Hillary Clinton a lead of at least 15 points in SC.
** If you increase the percentage of black voters in the sample, since Clinton has a bigger lead among blacks than among whites, her margin would be greater.
** The 2004 exit poll shows that whites made up about 51% of the Democratic primary electorate.

Recent SC polling I've found:

This latest one, with HRC at 43%, Obama at 27%, Edwards at 17%.

ARG polled June 26-30 and had Hillary at 37%, Edwards at 22 percent and Obama at 21%.

Mason Dixon, which polled from June 13 to June 15, had Obama with a 9 point lead over Clinton, 34% to 25%.

A May 31 PPP poll gave Obama a lead within the margin of error.

And a May 16 Winthrop/ETV poll gave Clinton an eight point lead over Obama.

ARG's May sample gave Clinton a 16 point lead over Obama, with Edwards in second place.

Comments

Hmmm, those numbers seem all over the place. Seems like numbers have been a little more stable in other early states.

What's intriguing to me, though, is how Mason-Dixon could have Obama up by nine a month ago. If these polls are even accurate within their margins of error, seems like Obama lost ground for some reason, and conversely Clinton gained ground.

Posted by PLM | July 20, 2007 2:21 PM

...or they have different standards for likely voters

Posted by Michael | July 20, 2007 3:15 PM

You should point out that Opinion Research is owned by Clinton backer Vinod Gupta. Info USA is the parent company of Opinion Research, and Bill Clinton is a paid consultant for them. He also earned $200,000 for a speech in front of them on 7/27/2001. (This is part of the $41 million that he has made speaking in front of various industry and PACs since 2000 -- according to Hillary's financial reports on opensecrets.org.)

Posted by George | July 20, 2007 3:15 PM

as George pointed out, both Opinion research and Arg are owned by Mr. Gupta. Their past polls have always been outliers.

we need to sunlight this clinton ties to the pollster.

Posted by danny | July 20, 2007 3:34 PM

where are the 4 polls that gave a lead of at least 15 points to clinton

Posted by moi | July 20, 2007 3:37 PM

Wait a minute, the OR poll was 380 people selected from a sample of over 1000!?? That is pretty fishy to begin with.

To keep your sanity in polling situations, ignore the percentages and look at the actual numbers. If OR polled 380 people, this is what the results are:

148 people for Clinton
95 people for Obama
57 for Edwards
38 for Gore
11 for the other candidates

Not a tremendously impressive sample in a state of over four million. This is the moral equivalent of standing outside of a Starbucks, except you have the added bonus of throwing away 2/3 of the responses that you get.

It gets even more insane when you try to break it down by race and gender. I haven't gone into it, but the differences must be about 10 people.

Posted by George | July 20, 2007 4:48 PM

What are the track records for pollsters in SC? Does Mason-Dixon on average poll SC better than other pollsters? If so, then, Obama's team may have a point.

Edwards' team likes to point to his lead in IA. Some polls have shown that he's losing his lead in the state or his leads are within MOE... but the one poll (I think conducted by Des Moines Register) that everyone was saying "that pollster knows how to poll IA" helped keep Edwards in the top tier (in the eyes of bloggers at least), given his lead in that poll.

Posted by Newsie8200 | July 20, 2007 5:09 PM

xtuneydq wyxtfjl heylrxqo jtbxyik loam snvefplr empz

Posted by sfbcvw wbgyaermc | August 10, 2007 11:39 AM

alternative buy viagra

Posted by buy online purchase viagra | November 12, 2007 7:59 PM

want delete your site from spam bases? mail your domain name - andydelay[at]gmail.com.

Posted by xyujbc | January 26, 2008 6:05 PM