28 November 2011

Rank 20 / 32: Pennsylvania


The Battleground state from 2008 that never was really a Battleground“


The solid DEM win in 2008 that may be a huge Battleground in 2012“


Table1: Overview of the last 6 cycles

Year Rank. Winning % % Margin % Margin over National Notes
2008 20 / 32

2004 18 / 34

2000 16 / 36

1996 23 / 29
Closely paralleled national stats.
1992 16 / 37

1988 13 / 39

Pennsylvania margin average, 1988-2008 (6 cycles): DEM+5.48%

Note: the margin average for these cycles for PA is practically identical to Wisconsin: DEM+5.47%

Note: It seems counter-intutive that PA has actually moved down 2 places in the liberal rankings from 2000 to 2004 and again from 2004 to 2008, but the DEM percentage in this state has gone up continually since 1988. That coupled with the fact of a landslide +10.31% win in PA in 2008, but with the state „only“ in 20thplace, means that 19 other states had larger landslides in 2008 than PA, 2 of them being DEM pick-ups from 2004: NM and NV. So, the slippage of PA in the partisan rankings, but themuch higher margin is once again statistical proof of the solidity ofthe Democratic Electoral column from 2008.

PA county-by-county EXCEL spreadsheet
(raw totals for 2008 and 2004, margins, swings, % of state PV, county growth rate)


The Partisan Rankings over 44 years

The partisan rankings for Ranking 20 (Pennsylvania) and Ranking 26 (Ohio) compared to each other from 2008 backwards in history to 1964 in Table-format (PA in yellow, OH in green):

Rank2008Margin '082004Margin - 042000Margin '001996Margin '961992Margin '921988Margin '88
11 – 41CT22,37%HI8,74%CA11,80%AR16,94%MN11,63%WA1,59%
12 – 40ME17,32%DE7,59%VT9,94%MN16,14%WA11,44%IL2,08%
13 - 39WA17,08%WA7,18%WA5,58%MD15,99%HI11,40%PA2,32%
14 - 38MI16,44%NJ6,68%MI5,13%DE15,25%MO10,15%MD2,91%
15 - 37OR16,35%OR4,16%ME5,11%WV14,75%OR9,95%VT3,52%
16 - 36NJ15,53%MN3,48%PA4,17%MI13,21%PA9,02%CA3,57%
17 - 35NM15,13%MI3,42%MN2,40%CA12,89%NM8,56%MO3,98%
18 - 34WI13,90%PA2,50%OR0,44%WA12,54%ME8,33%NM4,96%
19 - 33NV12,49%NH1,37%IA0,31%LA12,07%DE8,20%CT5,10%
20 - 32PA10,31%WI0,38%WI0,22%IA10,34%MI7,40%MT5,87%
21 - 31MN10,24%IA0,67%NM0,06%WI10,33%CT6,43%SD6,34%
22 -30NH9,61%NM0,79%FL0,01%NH9,95%IA6,01%CO7,78%
23 - 29IA9,53%OH2,11%NH1,27%PA9,20%TN4,65%MI7,90%
24 - 28CO8,95%NV2,59%MO3,34%OR8,09%LA4,61%LA10,21%
25 - 27VA6,30%CO4,67%OH3,51%NM7,33%WI4,35%OH10,85%
26 - 26OH4,58%FL5,01%NV3,55%OH6,36%CO4,26%ME11,45%
27 - 25FL2,81%MO7,20%TN3,86%MO6,30%KY3,21%KY11,64%
28 - 24IN1,03%VA8,20%AR5,44%FL5,70%NV2,63%DE12,40%
29 - 23NC0,33%AR9,76%AZ6,28%TN2,41%MT2,51%TX12,60%
30 - 22MO0,13%AZ10,47%WV6,32%AZ2,22%NJ2,37%ND13,06%
31 - 21MT2,38%NC12,43%LA7,68%NV1,02%OH1,83%KS13,23%
32 - 20GA5,20%WV12,86%VA8,04%KY0,96%NH1,22%NJ13,64%

Rank1984Margin '841980Margin '801976Margin '761972Margin '721968Margin '681964Margin '64
06 – 46PA7,35%MD2,96%AL13,11%WI9,67%ME12,23%NY37,25%
07 – 45IA7,39%HI1,90%SC13,04%OR10,12%MS40.44%WV35,87%
08 – 44NY8,01%MA0,15%TN13,00%CA13,46%WV8,82%CT35,72%
09 – 43WI9,18%TN0,29%MN12,87%MI14,39%MI6,73%MI33,61%
10 – 42WV10,51%AR0,61%RI11,28%IA17,13%NY5,46%VT32,61%
11 – 41HI11,28%AL1,30%NC11,05%NY17,34%CT5,16%AK31,82%
12 – 40OR12,17%MS1,32%KY7,19%WA18,28%LA20.11%NJ31,75%
13 - 39IL12,88%KY1,46%MD6,07%CT18,44%AL47.13%MD30,94%
14 - 38WA12,97%SC1,53%LA5,78%IL18,52%PA3,57%PA30,22%
15 - 37CA16,25%NC2,12%DE5,41%PA19,98%WA2,11%KY28,36%
16 - 36TN16,27%DE2,33%FL5,28%MT20,08%MD1,64%MO28,10%
17 - 35VT17,11%NY2,67%NY4,43%DE20,41%TX1,27%MN27,76%
18 - 34OH18,76%ME3,36%MO3,63%OH21,56%AR7.64%OR27,75%
19 - 33MI18,99%WI4,72%TX3,17%ME22,98%MO1,13%NH27,28%
20 - 32DE19,85%LA5,45%PA2,66%AK23,51%NJ2,13%TX26,82%
21 - 31MO20,05%VT5,96%HI2,53%MD23,90%OH2,28%OH25,89%
22 -30GA20,39%MI6,49%MS1,88%NM24,49%AK2,64%WA24,59%
23 - 29NM20,48%MO6,81%WI1,68%MO24,59%IL2,92%WI24,35%
24 - 28KY20,66%PA7,11%OH0,27%NJ24,80%CA3,08%IA23,97%
25 - 27NJ20,89%IL7,93%OR0,17%HI24,96%DE3,51%CO23,07%
26 - 26CT21,90%CT9,63%ME0,84%VT26,20%WI3,62%DE22,17%
27 - 25ME22,05%OR9,66%IA1,01%ND26,28%GA12.43%NM18,98%
28 - 24AR22,18%OH10,60%OK1,21%WV27,22%OR6,05%IL18,94%
29 - 23AL22,26%WA12,34%VA1,34%NV27,36%KY6,14%MT18,38%

I will be putting out a statistical study of PA-OH regarding margins, trends and placement in the partisan rankings once all 51 state bios are finished.

Helpful Info Links Helpful Election Links
PA VR link
PA population 2008: 12,448,279 PA VR Current Stats
PA Population Density: 107.2 persons per square Km.
PA VR Changes
Electoral Vote Density: 592,775 persons per EV. PA VR Archives


PA Electoral Development (electors through history): 10(1789), 15 (1792-1800), 20 (1804-1808), 25 (1812-1820), 28 (1824-1828), 30 (1832-1840), 26 (1844-1848), 27 (1852-1860), 26(1864-1868), 29 (1872-1880), 30 (1884-1888), 32 (1892-1900), 34(1904-1908),38(1912-1928),36 (1932-1940), 35 (1944-1948), 32 (1952-1960), 29 (1964-1968), 27 (1972-1980), 25 (1984-1988), 23 (1992-2000), 21(2004-2008), 20 EV (2012- ) 

Note:in 1912, as PA reached its electoral peak of 38 Electors, this represented 7.3% of the 531 Electors from the Electoral College of that time. Together, PA and NY (45) had 83 Electors in 1912 and more often than not voted in unison for the same candidate. Their combined strength equalled 15.9% of the Electoral college in 1912, or 31.2% of the electors needed to win the Electoral College. The Republican Party pretty much had a „lock“ on these two states for a long time, making it easier for the GOP to win Presidential elections.Today, with 20 EV and at the waning phase of its electoral development (PA last had just 20 EV from 1804-1808), PA now represents 3.7% of the Electoral College. In 1804, those same 20 EV represented 11.4% of the Electoral College of that decade. In other words, PA had more than 3 times as much electoral „firepower“ in 1804 as it has today. Essentially, PA of 1804 was like our CA of today. PA had 1more elector than NY in 1804. That is food for thought over the theory and history of „Electoral Shift“.


Pennsylvania is the 20th most liberal state and the 32nd most conservative state, with a Democratic winning margin of +10.31% and having voted 3.05% more Democratic than the national margin in 2008.

 It was the 18th most liberal state in 2004, with a Democratic winning margin of +2.50% and having voted 4.96% more Democratic than the national margin in that year.

 It was the 16th most liberal state in 2000, with a Democratic winning margin of +4.17% and having voted 3.65%moreDemocratic than the national margin in that year. 

From 1904-2008, Pennsylvania went for the GOP 
14 times, for the DEMS 12 times and 1 timefor an Independent candidate (Roosevelt, 1912).

Since 1948 Pennsylvania went for the GOP 7 times, for the DEMS 9 times.

With the inclusion of the Republican Party in the Electoral College in 1856, PA was a GOP bastion, save 1856 itself: in the 3-way race between Buchanan (D), Fremont (R) and former President Milllard Fillmore (Whig – American), Buchanan romped with+18.12% over Fremont. Only, the Whigs were the predecessors to the Republican Party, meaning that Fillmore split the newly developing Republican vote in Pennsylvania, a phenomenon we would see again in 1912 with Taft and Roosevelt. This was proven as Abraham Lincoln (R) won theKeystone State with almost exactly the same margin for the Republicans (+18.71%) in 1860, a+36.83% shift in the state. From 1860-1908, PA went reliably Republican 13 cycles in a row. It should however be noted that 8 of these 13 wins were single-digit wins. Single-digit wins are a characteristic trait of PA.

PA loved Teddy Roosevelt and gave him 
68.00% (+40.67) of the vote in 1904, but then broke the Republican winning streak in1912 by again choosing Teddy Roosevelt as the „Bull-Moose“candidate (once again, a former President who campaigned under a 3rd party banner, just as in 1856) , who won with +4.04%of the vote in a three-way race against Incumbent Taft (R) and Wilson(D). From 1916-1932 (6 cycles), PA remained reliably Republican. In 1932, PA was one of only six states to resist the first FDR landslide, but in 1936, after 80 years of Democrats not winning PA, FDR easily carried the Keystone State for the Democrats, with +16.04%, and held the state in both 1940 and 1944, albeit with single-digit margins, both under the national average.

Harry Truman was unable to hold the state for the Democratic Party; Dewey flipped PA for the Republicans in 1948, where it also chose IKE in the 1950s by varying margins. Interestingly enough, we see the building blocks for a major Democratic coalition in PA starting in 1952: in spite of a massive national landslide in 1952, Ike carried 3 Northeastern states with low and middle single-digit margins: RI, DE and PA. Up to this time, PA had the most Republican voting record of these three. The reason for this, historians agree, was the massive shift blue-collar voters toward the Democrats with FDR, a good deal of whom stayed with the Democrats in the 1950s and grew through the1960s.

PA went easily for LBJ in 1964, with a Democratic record-setting marginof 
+30.22%. It went against Richard Nixon twice, giving Kennedy and Humphrey the vote in 1960 and 1968, respectively.

Nixon did finally capture PA in 1972 with his re-election, but Carter flipped the state again in 1976, with a margin very close to his national winning margin: +2.66% (national +2.06%).

Reagan flipped the state again for the Republicans in 1980, where it stayed in the GOP column through 1988. However, PA trended very strongly against the Republican party in 1988: Bush, Sr. +2.32%, nationally +7.73%.This will be explained in the chart below.

In 1992, PA re-entered the Democratic column and has voted reliably Democratic for 5 cycles in a row.

From 1976-2004, 8 consecutive cycles, PA was a „single-digit state“and therefore considered competitive for both sides.

Obama's 2008 landslide in PA, with +10.31% (just slightly larger than his landslide in perennially Democratic Minnesota), broke that single digit-trend and is:

-the largest Democratic win since 1964.
-the fourth largest Democratic win in PA's history (after 1964, 1856 and 1936).
-the largest win overall in PA since 1972 (36 years). 
-the longest Democratic voting record in PA's history.
- a statewide partisan shift statistically identical to Florida (a 0.01% variance is nothing more than statistical noise), but somewhat greater than the partisan shift in neighboring Ohio:

PA: 10.31%- 2.50% = 7.81% partisan shift within the state.
FL: 2.81%- (-5.81%) = 7.82% partisan shift within the state.
OH: 4.58%- (-2.11%) = 6.69% partisan shift within the state.
We will come back to PA-OH comparisons later.

An interesting side-note going into 2012: PA has never really had a true „sqeaker“ election like MO just had in 2008 (McCain+0.13%) or Wisconsin in 2004 ( Kerry +0.38%) or Florida in 2000 (Bush, Jr. +0.01%). There is a tie for the closest race in PA: both Kennedy and Bush, Sr. won the Quaker State by +2.32%, respectively. In other words, in thetypical sinus-wave of contraction and expansion in margins as a state is undergoing electoral change, PA has never really gone into the „nail-biter zone“, in contrast to neighboring Ohio.

Not only, that, +2.32% margin is only somewhat under the standard margin of error in polling (+/- 3). We have had 7 presidential cycles where the national winner in the PV had a margin of LESS than +2.32%: 2000, 1976, 1968, 1960,again in 18881884and1880. This means that statistically, it is very, very unlikely that the winner in PA will win by less than 2 points in 2012. It is just not part of the make-up of this state. Should polling show Obama or the GOP candidate ahead by only 1 point, it is pretty sure that the side that untimately wins will win by at least 2 points..

PA and Incumbents

RI and MI, Pennsylvania has rewarded incumbents who won re-election in percentage, but the „State minus Natl's“ (Partisan Value, or TREND) is all over the board. For both incumbents who LOST re-election in recent history (Bush 41, Carter), the Partisan Value (or TREND against the National) was actually in favor of the losing party:

President - PA Year / Margin Year / Margin State Shift National Shift: State minus Nat'l
1932 / +5.51
1936 / +16.04
1936 / +16.04
1940 / +6.89
1940 / +6.89
1944 / +2.78
1952 / +5.88
1956 / +13.19
Nixon 1968 / +3.57
1972 / +19.98
1976 / +2.66
1980 / +7.11
1980 / +7.11
1984 / +7.35
Bush 41
1988 / +2.32
1992 / +9.02
1992 / +9.02
1996 / +9.20
Bush 43
2000 / -4.17
2004 / -2.50
2008 / +27.81
2012 / ???

Redshading = GOP pick-up
Blueshading = DEM pick-up

To explain this gobbledy gook, the most important value to understand is the “State minus Nation” value, in the right-most column, shaded in grey. 
If it is a positive number, then this is good for that incumbent.Unlike RI and MI, where most of the “State minus National”numbers are positive, PA is a mixed-bag of results.

From the table above (in posting number 1), we see two very popular incumbents who were re-elected, where the actual margin in PA barely shifted upward:

Reagan went from 
+7.11% in 1980 to +7.35% in 1984 (a shift of just +0.24%), while his national margin SOARED from +9.74% to18.22% (a shift of +8.48%). For this reason, a negative value (-8.24).

Very similarly, Bill Clinton went from +9.02% in 1992 to +9.20% in 1992 (a shift of just +0.18%), while his national margin rose +2.96%. Again, his trend value is a negative one, but much smaller than that for Reagan, simply because Reagans NPV margin was much more impressive in 1984.

Wehave seen Pennsylvania flip in re-election campaigns and also in open elections, but there are very important patterns to see,we just need to train the eye to see them. In the last 100 years, Pennsylvania has „flipped“, or been a pick-up state, 10 times.

Pattern 1: In 6 of 10 cycles in which PA flipped to the other side, the losing side won the state by single digits in the cycle before (1916, 1936, 1948, 1972, 1980, 1992, highlighted in yellow):
1908: Taft (R-inc) +23.43%, flipped to Teddy Roosevelt (I) +4.04% in 1912.
1912: Teddy Roosevelt (I) +4.04%, flipped to Wilson (D-inc) +14.03% in 1916.
1916: Wilson (D-inc) +14.03%, flipped to Harding (R) over Cox (D)+38.56% in 1920.
1932: Hoover (R-inc) +5.51%, flipped to Roosevelt (D) +16.04 in 1936.
1944: Roosevelt (D) +2.78%, flipped to Dewey (R) against Truman (D-inc) +4.01% in 1948.
1956: Eisenhower (R) +13.19% flipped to Kennedy (D) aginst Nixon (R) by +2.32% in 1960.
1968: Humphrey (D) +3.57%, flipped to Nixon (R-inc) +19.98% in 1972.
1972: Nixon (R) +19.98% , flipped to Carter against Ford (R-inc) +2.66% in 1976.
1976: Carter (D-inc) +2.66%, flipped to Reagan +7.11% in 1980.
1988: Bush, Sr. (R-inc) +2.32%, flipped to Clinton +9.02% in 1992.

Pattern 2: in 6 cycles, 3 + 3, the state flipped three times in a row, back and forth, with at least one „loner“ flip cycle many years apart from the rest:

1936 (16 years apart)
1948 (12 years apart, both back and forth)
1960 (12 years apart, both back and front)
1992 (12 years apart)

As of 2012, 20 years will have passed since PA last „flipped“. Logically, if this pattern of historical precedent were to have held, then historically, PA should have flipped in 2004 or 2008, but this was not the case. WE can also consider that, at some point in time,when PA does „flip“, it will most likely flip 3 times in a row, as it did twice in the last century.

Third pattern: you have to go back 100 years in history, back to 1912, to find a case where an incumbent President won Pennsylvania by double digits in his first election and lost the state in his re-election, namely, Taft 1908-1912. Obama won PA with a double digit margin in 2008. The difference is that Taft lost his re-election in PA and nationally in 1912 in good part due to a challenge within his own party from his predecessor, former President Theodore Roosevelt, a man with unbelievable popularity. When the intra-party challenge failed, Roosevelt launched a third party bid and ended up being the second party ticket in the election. Most historians agree, had Roosevelt not challenged Taft via third party run, then the GOP may still have lost the election in 1912, who knows for sure, but would most likely have retained many of its core states like PA. This is not the case with Obama in 2012: there is no challenge from within the Democratic party to President Obama and at least as of this time, there is no third party on the horizon for 2012.

Based on its voting record, PA is not a bellwether state, having missed the winner in 7 of the last 26 cycles, or 100 years, and more recently, it has missed the winner 3 times since 1960.

Why was PA called a „battleground“ in 2008?

In the election of 2008, as his own campaign manager admitted afterwards, John McCain made two major „Hail-Mary“ passes: he picked Sarah Palin to be his Vice-Presidential candidate and hedecided to make a hard play for PA, seeing that he could likely lose some Bush 43 states from 2004. And honestly, the statistical data from 2004 was a good support for his argument. According to the polling information collected by RealClearPolitics, the final polling average in PA was: Kerry +0.9%. This means that John Kerry outperformed the final polling averages by 3.4% in PA,while at the same time, losing by 2.46% nationally to President Bush in the 2004 GE. Seeing a trend of reduced margins from 2000 to 2004 and very close polling in 2004, McCain made the decision to sink money and time into PA and MN (which is a very, very reliable Democratic state). Only, the polling in 2008 never once showed this state to really be competitive.Starting on September 23, 2008,I published nightly „BATTLEGROUND STATE REPORT“s. Already, as of 10/02, I indicated that, based on the average margin of +6.6 or more for Obama, that PA and MN were both going to leave the BATTLEGROUND REPORTS. There was no statistical reason to keep them there. Feel free to go through all the battleground reports, I will be doing the same thing starting in the middle of September, 2012. As it turned out, the final polling average for PA for 2008 was: Obama +7.3, so he out-performed the composite end polling by 3 points.

Interestingly enough, the election results from PA and MN – the two states John McCain picked to open as major battlegrounds, were almost carbon copies of each other:

Total Vote
New Hampshire

So, no, PA was never really a battleground in 2008, ever.

Can PA become a battleground in 2012?

Yes, definitely. Any state can suddenly become a battleground, if conditions are right for such a shift, and all polling of PA thus far shows President Obama in a statistical tie with Mitt Romney almost every time, or just slightly ahead. Against all other GOP hopefuls, he wins PA with secure margins. Pundits have often jokingly described Pennsylvania as „New York on the edges, Alabama in the middle“, and there is some political validity to this. With overwhelming margins from Philadelphia and secure margins from Pittsburg, a Democrat can just hold this state. If other blue-collar bastions like Scranton go heavily for the Democrats, then a landslide happens. If they do not and Pittsburgh turns against the Democrat, then the Republican wins. That is how the game works in PA.

There is one more indicator that most do not know: at the moment, we have a quandary. For the first time I can ever remember, we have a Democratic incumbent polling better in Ohio than in Pennsylvania, but the electoral history of these two states has had Pennsylvania consistently higher in the Partisan Rankings since 1964 and – I believe, all the way back to 1952. 1948 would be the last time that Ohio had a higher place in the Dem vs. GOP Partisan Rankings than PA. So, were the GOP candidate to win PA in 2012, but President Obama to hold OH in 2012, that would be a major electoral tectonic shift.

State Superlatives

PA Superlatives
Winning %
Roosevelt, T.
Roosevelt, T.
IND Victory
Winning Margin %
All-time “squeaker”
Bush, Sr.

In Pennsylvania, the Governor, Lt. Governor and 1 Senator are Republicans. The other Senator is Democratic. The US House delegation from PA is comprised of 12 Republicans and 7 Democrats, a shift toward the GOP since 2008. In the Pennsylvania General Assembly, the Republican Party has a strong 60 to 40 majority in the Senate and a 55-45 majority in the House. 

Facit: In2007, I wrote: The blue wave from 2006 swept through PA and will most likely strengthen DEM margins in 2008. PA is a “solid” DEM state, actually more solid than MI. Additionally, PA is one of the three states in the so-called “Trifecta”: FLA, OH and PA. According to conventional wisdom, whoever wins two of these three states in the GE has, with a statistically very high probability, won the election. For this reason, these three states are hard fought each cycle. However, with OH and PA both losing electors as of 2012, this trifecta can change in the next cycles.

Pretty prophetic words. The DEM margins in 2008 were indeed higher, and then reversed on the congressional level in 2010. The real surprise was the unsung massive landslide in MI, something that very few people saw coming in 2008. The emergence of VA and NC (Mid-Atlantic Seaboard) as major swing states changes the so-called „Trifecta“ to a „Quintifecta“, and we will see in 2012 if this holds. Pennsylvania is, however, still a Democratic state, but could become THE battleground of 2012, depending on the GOP candidate. In 2000, we held our breath for Florida. In 2004, we held our breath for Ohio. Will be hold our breath for Pennyslvania in 2012? One thing is for sure: PA was the most polled state in 2008, in both the Primaries and in the GE and will probably be the most polled state in 2012 as well.

No comments:

Post a Comment

Constructive comments and critique are always welcome. Please keep it polite and respectful.