Final Update: Pennsylvania New Democratic Registration Map

First-Time	Switchers	Total

Statewide since January first, there have been 35 previously unregistered voters who registered as Democrats for every 1,000 votes for John Kerry, and 47 party-switchers to the Democratic party for every 1000 votes for George Bush, for a total of 41 new Democrats for every 1000 votes in 2004. That's an astonishing figure; if every one of those voters votes D in 2008, and the rest of the state votes the same way, the final result will be 54-46 instead of 51-49.

Eight counties had a Democratic increase of more than 5.0%:

• Dauphin County (Harrisburg-Carlisle): 7.3%
• Monroe County: 6.6%
• Cumberland County (Harrisburg-Carlisle)
• Lehigh (Allentown): 5.8%
• Delaware (Philly's inner suburb): 5.7%
• Philadelphia: 5.6%
• York: 5.2%
• Chester (the most Republican Philly suburb): 5.0%
   

The New Registrations were fairly uniform, but most prevalent in the Clinton stronghold of Northeastern Pennsylvania. The switching was heaviest in Philly and its suburbs. On balance, unless the Obama team was able to register Obama-leaning voters even in Clinton country, the new voters will give him little to no boost.

Comments

[Benjamin]

Hum ... excuse me but isn't Philly and its suburbs HIS strongholds ?

[Nicholas Beaudrot]

Yes. But the big 5 -- Philadelphia, Montgomery, Bucks, Chester, and Delaware -- account for only 94,098 out of 234,801 new voters. even if you throw in Allegeney, Centre, Dauphin, Monroe, Lehigh, and Northampton counties it only gets him just over 50% of the new voters. For this to be a net win for O, you have to believe that there's something fundamentally different about the new voters in Clearfield County or the party switchers in Lackawanna County that makes them more likely to vote for Obama than the rest of the voters in their vicinity.

[low-tech cyclist]

On another subject, is it time to put Public Policy Polling (PPP) on the 'ignore' list? I honestly trust them less than ARG at this point.

They had Obama jumping from a 1-point lead in NC to a 21-point lead from one week to the next. And over a 2-week interval, they had Hillary's lead going from 26 points to negative (a 2% Obama lead).

Nobody thinks there was a 20-point swing in NC in a week, or a 28-point swing in PA in two weeks. Or anywhere close. Them's crazy numbers.

[Nicholas Beaudrot]

I think PPP has changed their voting universe. They've started to include people who voted in a general election in 2006, on the assumption that that's more in line with turnout (in Ohio, Clinton an Obama got 50,000 more votes than Sherrod Brown in '06), which explains some of the shift in NC. Presumably they did the same thing in PA. But I don't believe it at all.

[ikl]

PPP's new NC numbers seem plausible (their old poll their seemed off so that would explain part of the swing). The PA numbers do not.

[John]

es. But the big 5 -- Philadelphia, Montgomery, Bucks, Chester, and Delaware -- account for only 94,098 out of 234,801 new voters. even if you throw in Allegeney, Centre, Dauphin, Monroe, Lehigh, and Northampton counties it only gets him just over 50% of the new voters.

Berks, Lancaster, and York (but probably not Monroe) would probably be wise to look into.

Also, that's 40% of new voters in the Philly area, as compared to those counties comprising 31% of the total population of the state. That means new registration occurred disproportionately in the Philly area, doesn't it?

[John]

The internal numbers on the PPP, by the way, don't seem all that insane. I won't say it's necessarily right, but I don't know that it should be dismissed out of hand. Quinnipiac's numbers (Clinton +9) seem closest to what I'd suspect the "real" numbers are, but Obama is clearly starting to surge in PA.

[Nicholas Beaudrot]

Yes, but historically the Philly area represents 40% of Democratic voters. They ended up representing 45% of new voters, but only 41% of switchers.

I don't see why new voters in York, Lancaster, and Berks—or even Northampton and Lehigh—would be Obama voters. Demographically those areas favor Clinton -- almost entirely white, education levels below the statewide average. Now, maybe it's true that all new voters and party-switchers everywhere favor Obama but I just think that's very doubtful.