Over a week ago I published a diary illustrating a potential scenario for the PA primary, based on Al Giordano's predictions at The Field and jlkenney's predictions here at Daily Kos.
Since that time there has been some movement in the polls suggesting that Obama's campaign in Western and Central PA has had some success in attracting voters. In particular, the Survey USA pollreleased on April 1st suggests that Clinton will not score landslide victories in a number of districts which formed her core support in early to mid March.
For a map, analysis and poll, see below the fold.
The current situation, interpreted moderately optimistically for Obama, looks to be as shown below:
This schema shows the 19 CDs in PA with their size in proportion to their actual number of delegates, not their geographic size. This gives a better picture of the balance of strength.
Blue is Obama, Pink is Clinton; Lavender is probable ties. (Apologies to those who find the color choice sexist; just remember that pink was considered a boy's color in the Victorian period because it was so passionate, and therefore unsuitable for girls, so they thought, who would do better with a kind of cool and passive blue.)
Diagonal shading in a square indicates a predicted delegate win for Obama. (But the location of the square within the CD has no significance here.)
The Survey USA results actually bring the map more into accord with jlkenney's March 6th predictions and away from Giordano's, which were based on a comparison of the PA demographics with the OH demographics and used the OH primary results to predict PA. A lot has changed since then.
Significant changes include the following:
- Pittsburgh is very much up in the air. It could go +1 for Clinton or +1 for Obama. Hoping that Obama has moved ahead owing to various endorsements and his campaign stop there.
- CDs 18, 4, 3, 12, formerly "strong Clinton", have been moved to "lean Clinton", giving Obama 4 more delegates than before, although for these CDs Clinton still holds a net advantage +4. In reality it is probably +5 since Obama may not manage 2 delegates in all four of these CDs.
- I've moved CD 5 (which includes State College) into a split. I'm hoping that the massive rally at Penn State shows that this is not unreasonable.
- Despite his surprise stop in Altoona, Obama is still very unlikely to win CD 9, so this, along with CD 10, remain Clinton territory.
- Survey USA polling shows Obama doing poorly in the far Northeast (Scranton-Wilkes-Barre). This remains strong Clinton.
- It's difficult to gauge the southeast at the moment. Remember that CDs 17, 6, 19 and 16 are all even-numbered districts. Obama would have to pull off a major victory to move these away from splits. So they are stable for now.
- I'm beginning to think a 7/2 split of CD 2 (West Philadelphia) is not unreasonable. I live here and have seen nary a Clinton sign in weeks. If there are Clinton supporters around here they are hiding themselves.
- Philadelphia East (CD 1) is leaning Obama, but as yet I hesitate to predict more than 4/3. A 5/2 for Obama would be very helpful, but not yet plausible.
- The Philly suburbs (CDs 13, 8, 7), that is, Delco, Montco and Bucks, are all still leaning Obama at 4/3.
Summary:
1 Phila East: +1 O
2 Phila West: +5 O
3 Erie: +1 C
4 New Castle: +1 C
5 State College: tie
6 Reading: tie
7 Delco: +1 O
8 Bucks: +1 O
9 Altoona: +1 C
10 Williamsport: +2 C
11 Scranton Wilkes-Barre: +3 C
12 Johnstown: +1 C
13 Montco: +1 O
14 Pittsburgh: +1 O
15 Allentown: +1 C
16 Lancaster: tie
17 Harrisburg: tie
18 Bethel Park: +1 or +2 C
19 York: tie
Total: +1 or +3 Clinton. Not bad.
For now, Obama needs to:
(1) Hold a distinct advantage in the southeast, including all Philly suburbs.
(2) Win Pittsburgh.
(3) Hold Harrisburg, Reading, Lancaster and York to ties.
(4) Consolidate gains made in the central and southwest regions.
(5) And if Philadelphia East (CD 1) goes 5/3 or Obama can hold Clinton to 3/2 in the four western PA districts (excluding Pittsburgh) we are now talking about a near TIE in the delegate count.