An Embarrassing Performance
[I bumped this for an update.]
Yes, I'm talking about the Pennsylvania primary. No, I'm not talking about Barrack Obama.
I've had running battles (on other sites) about what I viewed as godawful polling practices by an outfit out of North Carolina called Public Policy Polling. Anyone who has had political science training could look at the crosstabs for their polls and see countless red flags. For starters, the weighting that they gave to black voters and to the Philadelphia area seemed all out of proportion to the actual demographics of the state or any historical precedent. Also, the fact they kept the raw numbers hidden from scrutiny was also troubling for anyone trying to gauge them from the perspective of the social sciences. There is nothing wrong with weighting your polls, but you should give people the opportunity to scrutinize them without having to engage in reverse engineering.
There were 39 polls released about the Pennsylvania primary since April 1st. A grand total of three of these polls had Obama in the lead, Clinton led in 35 of the others with one saying it was a tie. All three of the polls that had Obama in front were the handiwork of PPP (last of the three, a massive survey of 2338 likely Democratic voters with a +/- of only 2% -supposedly- here.) In a sense this isn't surprising after you look at the client list for PPP. I conducted searches to see who the majority of PPP clients had endorsed for the Democratic nomination. There were copious Obama endorsements, but not a single Clinton endorsement. Now, this might just be a hell of a coincidence, but given the tightness of the nomination fight that seems a little unlikely. It seems more likely that PPP as an outfit are not simply Democratic pollsters, but pro-Obama (or anti-Clinton) pollsters.
Such a view is confirmed by the actual results in Pennsylvania tonight (at this moment Clinton ahead 10% points with 96% reporting.) PPP's polls, which had Obama ahead by 3%, were not just wrong they were terrible, if the purpose of a poll is to measure a larger population accurately. Of course, if accurate measurement was not the goal of these "polls", well then maybe it is mission accomplished.
Crossposted at The Iconic Midwest
UPDATED: PPP responds to the obvious. Where Did We Go Wrong?
The networks have called it for Hillary and I think after Florida in 2000 they're not going to make that mistake again, so obviously our polls were wrong.
First off, please do not call us or e-mail us and tell us we suck! We are well aware, and it does not feel good.
….
It's pretty easy, based on the exit polls, to see where we went wrong. We had the black vote at 18% when it turned out to be 14-15%, and we had the under 45 vote at 41% when it turned out to be 31%.
….
Looks like the two areas where we were particularly off, as might have been expected, were with white women and senior citizens. We showed the race too close with both of those groups in our final poll, and Hillary ended up winning them by the kind of large margin that would have been expected.
See, they only misweighted the black vote, the youth vote, the white woman vote and the senior citizen vote.
Otherwise they were solid.
Politics | Rich Horton Wednesday, 23 April , 2008
By Uncle Pinky, Tuesday, 22 April , 2008 @ 11:16 pm
We’ve seen this before and we’ll see it again.
sigh
Thanks for the research.
By Neo, Wednesday, 23 April , 2008 @ 10:33 am
This leaves the argument, that the exit polls showed Kerry winning Ohio in 2004, in tatters.
By SteveB, Thursday, 24 April , 2008 @ 12:25 am
So their polling for a Democratic primary was accurate for white men only then. Brilliant.