I meant the newspaper endorsements in Pennsylvania - as in the whole state.

Small towns and all.

o.k., can we at least agree that if zogby is as horrible in PA as it has been for the earlier primaries that they should officially be banned from TPM?

Zogby...enough said. Clinton by 20%+ in PA.

That's two now, though. PPP also has Obama ahead.

The problem with Zogby this cycle has been that he, like many mainstream pollsters, re-weights his samples in accordance with a predicted turnout model. In this poll, for example, he purchased voter registration lists (and it's reasonable to ask whether those included the most recent registrants). Then he randomly sampled phone numbers from area codes and telephone exchanges, in proportion to the population. The selected phone numbers were called up to six times, and the interviews were conducted live.

It's the next step that's typically tripped him up - he took the 601 people he'd reached, and re-weighted them by region, age, race and gender. Most years, that's a prudent decision - it's not a huge sample, and even a little random variation can otherwise skew results. The problem is that this cycle, turnout has been wildly unpredictable. Thus, Zogby went badly astray in California by assuming that black turnout would be moderately up, and that Latino turnout would hold roughly constant. In fact, Latino turnout showed an enormous surge, and black turnout did not.

That's why SurveyUSA has often outperformed other polling outfits this cycle. They just call around, and don't reweight their results. If many more women than men indicate that they're going to the polls, SurveyUSA assumes it's a genuine gender gap, and reports it.

So we have to take Zogby's topline numbers with more than a little bit of caution, because he neither discloses his raw data nor tells us what assumptions have been employed to reweight it. But that doesn't mean we should ignore the poll entirely. In particular, we might pay attention to the subgroups he pulls out in his unusually expansive discussion. There are some surprising findings there - like people who've lost jobs backing Obama by an 18% margin. Or that voters agred 35-54 have swung from 45-40 Clinton to 47-41 Obama. Or that Hillary maintains double-digit leads among whites and hispanics. We don't know if all those subgroups include significant sample sizes, but re-weighting is less likely to skew these internals.

W and Obama are strikingly similar in one respect. I remember back in 2000 arguing with Republicans, and even some Democrats, stating the obvious fact that W is a moron. And sometimes the retort would be, well, he went to Yale and HBS, he can't be *that* dumb, huh? And people overlooked the fact that W had never, ever done anything substative in life other being governor of Texas for a few years. Well, same with Obama. Granted he's not as dumb as W, but that's a pretty low bar to cross. But like w he's really never done much of anything. And being President of the HLR, the thing that put him on the map. Elected position without line editorial responsibility. Getting on HLR? If you're black, no need to grade on, you're on automatically. Getting into HLS? Again, not so hard, an LSAT that would not get you into Vanderbilt is sufficient for HLS if you're AA.

Hey good luck, US (of the KKK of A!)

Zogby!

Ay dios mio!

Zogby....lolololololololo

oh, I implicitly recognized in my post that W was a legacy admit. But what I find funny about defenders of affirmative action is the argument, well, 10% of admits to Harvard are dimwits with old money family connections, so what's wrong with admitting 10% black dimwits. Well, then you end up with 20% of elite skools' graduates being dimwits with tony degrees, and you make it all the harder for smart, motivated non-AA lower income students to get in.

Yeah, Zogby is pretty pathetic this season but to someone's point he got Iowa and Missouri right (SUSA had HRC winning by 11pts in MO).

Of course no matter what the polls say, I am expecting a 20pt win for HRC so I am pleasantly surprised by any improvement Obama can make on that.

The thing is - and is that most of the polls that are coming out of PA that include data AFTER the weekend have Obama higher in actual % than previous polls. Up until this point he made a quick gain when he first started campaigning and then hovered around the 40% mark for weeks. It has been HRC's numbers that have bounced around a lot. (ranging from 56-36%!)

However these polls are showing him in the mid-40's. Not a substantial gain but a notable one.
Consider this too- HRC has been running these negative ads and nothing positive. The media is all over this non-troversy when polls show most people could give a rat's-ass. I think there is some backlash going on. I am not from PA, but I am from WI which also has a lot of conservative blue-collar types. This stuff went over like a lead balloon there. According to my bro-in-law who works a factory job in WI...

"I am making crap wages with shi**y health care, paying $4/gal for gas while my friends and neighbors are getting shipped off to Iraq - I couldn't care less if Obama called me a red-necked, gun-toting zealot and he was BFF with Osama-bin-Laden himself. If he could fix some of this crap, I would vote for him."
And yes, he is supporting Obama (and he's a republican)

I don't care if Zogby got a number of polls wrong, they're always so positive about him, they brighten my days before the vote. Sure I get disappointed sometimes, as with California, but it's not Zogby's fault that the voters didn't go for Obama. The vote is going the way it will be going. It's like the Shrodinger's Cat paradox. Zogby doesn't hurt. They're my "pollbama", if that makes me a Pollyanna, so what? I *heart* Zogby.

Zogby ...kiss of death

Every time I've mentioned a Zogby poll to my brother, the response is "How is President Kerry's favorite pollster doing"