Last night gpclay's diary alerted us to an excellent analysis yesterday in Allentown's Morning Call newspaper by Pennsylvania election and polling experts Terry Madonna and Michael Young, in which they describe a realistic roadmap of how Obama could actually pull an upset win in the Keystone State. Ironically, the roadmap is based on the same strategy that was used successfully by Gov. Ed Rendell, Clinton's chief surrogate in Pennsylvania, during his own 2002 primary fight for the governor's seat. I won't repeat gpclay's diary, but would like to focus on something not covered specfically in that diary: Madonna's and Young's list of the three keys of Rendell's 2002 victory that they say Obama must follow closely if he is going to beat Hillary in Pennsylvania, or at least hold down her margin of victory.
As a brief background, Terry Madonna is professor of public affairs at Franklin and Marshall College in Lancaster, PA, and Michael Young is managing partner of Michael Young Strategic Research in Harrisburg. In yesterday's diary, gpclay noted that they are viewed in Pennsylvania as "fair, nonpartisan and independent."
Madonna and Young acknowledge that Clinton's advantages in Pennsylvania are "substantial." But they draw parallels between the Clinton-Obama contest and Rendell's Democratic primary fight against Bob Casey, with Casey in the Clinton role as the "prohibitive favorite with deep roots in state politics, a political brand name second to none in Pennsylvania with widespread party establishment support" and Rendell as a "brash upstart" and a "politician of uncertain outline to a majority of voters."
Rendell defied expectations and beat Casey by about 150,000 votes. Madonna and Young say Obama could also pull an upset win, or at least hold down Hillary's margin of victory, by following the same three steps that were key to Rendell's win:
1. Massive concentration of effort:
Pennsylvania has 67 counties, and Rendell managed to lose 57 of them. But the counties he won were the big counties, and he won them big. Key was the impressive percentages he recorded in Philadelphia and the four suburban counties. He rolled up 75 percent in Philadelphia and more than 80 percent in the suburbs. Rendell ended with 300,000 votes in the Southeast, twice his statewide margin.
I had always known that the Philly region would be important to Obama, but I didn't realize just how critical it would be until reading this analysis.
2. Iron control of the swing vote:
Rendell topped off Philly and the suburbs by winning the Lehigh Valley and much of the southcentral part of the state, capturing 60 percent or more in many of those counties. Rendell's regional dominance in eastern Pennsylvania was critical. Nine of the 10 counties he won were east of the Susquehanna.
For those of you unfamiliar with Pennsylvania geography, the Susquehanna runs north-south, right through Harrisburg, and divides the eastern one-third of the state from the central and western two-thirds of the state. As shown in the map below (Rendell in purple), the only county Rendell won outside of the southeast region was Centre County (home of State College/Penn State):
3. Managed statewide turnout:
Finally Rendell also was able to turnout a higher percentage of Democratic voters in the Southeast than voted in the Southwest. This was accomplished in part by increasing Democratic registration, including luring some Republicans across party lines, but mostly by exciting the voters in the Philadelphia TV market that overwhelmingly favored his candidacy.
I had always assumed that the Pittsburgh region in the southwest would be important for Obama, and it certainly is, but I was struck at how much Madonna and Young stressed the overaching importance of the southeast region. They repeat this point for emphasis in their summary of the keys to Obama's prospects in the state:
Measured against Rendell's 2002 template, Obama's chances aren't unpromising. He is positioned to emulate Rendell to a greater extent than is generally recognized. Essentially he must win the same 10 counties Rendell won in 2002, while reducing Clinton's margins in her strongholds. Currently he leads in Philly and will likely win there decisively, making the suburbs a major battleground. The Democratic voters there largely mirror the upscale, affluent voters Obama has been attracting nationally: They are the most liberal in the state, strongly oppose the Iraq War, with a low regard for President Bush. For insurance Obama needs to join the Philadelphia suburbs to the two pivotal swing areas, the Lehigh Valley and Southcentral, where the Democrats are moderate to liberal and where he currently polls well against Sen. Clinton. Beyond this, managing turnout will be crucial for him.
Rendell in 2002 was able to turnout a higher percentage of Democratic voters in the Southeast than voted in the Southwest. This is central to any Obama victory. He must do the same. Here enthusiasm and Obama volunteers could make a big difference.
Given the importance to Obama of the Philadelphia suburbs and the southcentral region, and of turnout in these regions, it is perhaps encouraging that voter registration in these areas has been particularly high:
The nine counties with the biggest percentage increases in Democratic enrollment since last fall — more than 5 percent — are mainly in two tightly contested areas — the Philadelphia suburbs and the state's vast central region.
It's clear, then, that Obama's GOTV efforts in Philadelphia and the surrounding suburbs, along with Lehigh Valley and southcentral, will be critical to the race. So Obama volunteers have a clear roadmap for where we need to be focusing our energies. Volunteers already played a critical role in the record-breaking voter registration in these areas. Now it's on to canvassing and phonebanking - targeting the regions outlined in this diary - and, of course, donations. This map by Pennsylvania for Change nicely outlines where we need to focus our efforts, essentially reflecting the recommendations of Madonna and Young (with the exception of adding in the Pittsburgh region):
(click this link for a larger image of the map)
Yes We Can!