What is happening in Pennsylvania? A trend, or back to normal?

[wbrocks67]

Here to get everyone's thoughts on perpetually purple (though I guess you could argue besides 2016, Tilt D) Pennsylvania.

Obviously, Trump happened in 2016. He won in PA 48.2% - 47.5%. 3rd party candidates picked up 4.4%. It would appear that HRC was just a bad candidate for PA (or Trump was a good one), but it appeared it was more than just the two of them -- Pat Toomey also beat Katie McGinty by a little more than Trump's margin in the senate race. So even in a non Hillary-Trump race, Toomey performed generally the same statewide (with better margins in the suburbs a bit).

HOWEVER, in 2018, obviously things swung back to the Dems favor. Wolf won by 17% and Casey won by 13%. PA actually had one of the biggest swings from 2016, bigger than Michigan or Wisconsin.

Obviously this had to do some with candidate quality (Wagner and Barletta were not good), though I think the Senate race was pretty close to Generic D vs. Generic R. Casey obv has name recognition, but most people don't particularly feel one way or the other about him. While Barletta had no scandals, and despite his super-right past in Hazelton, I don't think that was common knowledge for most folks.

Not to mention, Conor Lamb's huge early 2018 swing in a Trump district. Also, Wolf and Casey simply demolished Barletta and Wagner in the suburbs.

So what is happening? Realignment? Back to PA's Tilt D past? Is this a super-charged reaction to Trumpism?

To me, it's a combo of all 3, but if you're looking for a state that took a chance on Trump the most and now has repudiated Trumpism the most, I think PA is the best example.

[TML]

As I mentioned elsewhere, the decisive region in 2016 was northeastern PA (Scranton/Wilkes-Barre and surrounding areas). Both Wolf and Casey performed better than Clinton and McGinty in this region. They also improved upon Clinton's performance in the Philadelphia suburbs and in rural PA.

In the long run, I think PA's growth trends are more favorable toward Democrats, since its fastest growing region is the southeast (Philadelphia and surrounding areas) while the western region is declining in population.

[Orser67]

I feel pretty good about PA for Democrats. The suburban Philly counties have had a pretty strong reaction against Trump in the last couple of years, and I think Montgomery, Chester, and Bucks county (and possibly Delaware County) will all probably vote more Democratic in 2016. There's room to grow in Philadelphia, too.

I also expect some bounceback towards Democrats outside of the Philly and Pittsburgh metro areas. For example, Obama won 63% in Lackawanna County (home to Scranton) in 2012, whereas Clinton won just 50%. Another one is Luzerne; Obama won 52% in 2012, while Clinton won just 39%. I don't think the Democrat will be winning by Obama margins, and in the long term these will probably be GOP areas, but I think Democrats have a strong chance of improving there if they nominate the right candidate.

[Zaybay]

I personally believe a key feature of whats happening in PA is threefold.

1. The Suburbs of Philly have almost fully converted to the Ds.

2. The Ds bled in the North West, but they stopped most of it, and gained a bit in the Scranton area, accounting for the wave.

3. The west of the state, specifically Erie, PA, went hard back to the Ds, and was part of a shift that we saw throughout 2018, of urban/suburban areas that voted for Trump in 2016, with D ancestry and a more WWC background, going hard Democratic. In the South of the West, a new effect was seen, a spillage. The Blue Vote from Pittsburgh and its suburbs was starting to spill over into the old D stomping grounds, causing a shift. These spillage areas are the only areas growing in Pop. to, so its especially deadly to Rs.

PA is definitely losing its Midwestern identity and embracing its more Northeastern persona. If I were the D presidential candidate, I would be extravagant about the PA results in 2018.

[Comrade Funk]

PA is definitely losing its Midwestern identity and embracing its more Northeastern persona. If I were the D presidential candidate, I would be extravagant about the PA results in 2018.

Thank you for noticing this. Don't understand why some pundits can't.

[lfromnj]

Eh afaik PA is really incumbent friendly. Id definetely PA is the most likely to flip back of all the states.
Anyway looks like the new bellwether is Erie county.. Matching both the 2016 SEnate and president along with the 2018 gov and senate within a few points.

[IceSpear]

Dems clearly have a higher ceiling and floor than the GOP does in PA. I can't remember the last time the GOP has won a statewide race by double digits (Specter in 2004?), but the Dems have a bunch of times. GOP wins tend to be narrow, as Toomey only won by <2 points twice and Trump by <1 point. Ironically, Corbett's 9 point in 2010 stands out as one of the best GOP performances.

[Southern Senator North Carolina Yankee]

What are we up to on the GCB? D+9?


PA was  basically tied at D+2 in 2016. In a uniform swing environment thus the state would be expected to vote about D+7 or D+8 based on that.  The Governor's race would have a bigger differential than the Senate race obviously and the same thing occurred in 2006.

Casey outperformed by about 5% to 6%, which based on his name id, popularity in the NE and Lehigh Valley and yes record performances in the cities and the suburbs, makes a lot of sense.

I think PA is a close to even state with the national average, it is the front lines between the two demographic blocks of each party and it is a tendency to favor incumbents as someone said. There certainly was a turnout problem for the GOP as the electorate that showed up in the exit poll was Clinton by 6. While we cannot be certain as to the accuracy, there is reason to believe it might be higher than that based on the turnout data posted above but it is certainly in the ballpark.

Still for a state that turned out in a manner hostile to the President, it still was split close to even on trade and immigration policy. A plurality said Trump's trade policies help more than they hurt IIRC and the split on immigration between "too tough" and "Just Right/Not Tough Enough"  was 52%-48%. this number appeared elsewhere in the exit polls as well.

[Dukakisite1988]

Dems clearly have a higher ceiling and floor than the GOP does in PA. I can't remember the last time the GOP has won a statewide race by double digits (Specter in 2004?), but the Dems have a bunch of times. GOP wins tend to be narrow, as Toomey only won by <2 points twice and Trump by <1 point. Ironically, Corbett's 9 point in 2010 stands out as one of the best GOP performances.

1998: Tom Ridge won re-election as governor by 57-31.......and that's with the Constitution Party's nominee Peg Luksik receiving over 10%. Arlen Specter was re-elected by 61-35. Since then, yes, only Specter's 2004 win exceeded a 10% margin in a Senate/Gubernatorial race.

[Starry Eyed Jagaloon]

There are two inverse trends happening at the same time. Non-SEPA is shrinking and getting more Republican, whereas SEPA is growing and getting more Democrat. The two trendlines crossed around 2016, with non-SEPA just winning out, but now everything is uphill for Democrats with a growing SEPA propelling them to ever larger victories. A similar trend could be seen in Virginia, if it were one state with WV. WV was ancestrally Democratic and would have once carried the state with Arlington and some downstate areas, but is shrinking and trending hard R. Meanwhile, NOVA is growing and trending hard D. The trends would have meant the GOP would win most elections in this combined state clear through 2016, but from 2018 on, growing NOVA will more than cancel out the rest of the fake state.

[ElectionsGuy]

We need to be careful about reading too much into this. Democrats did very, very well in the 2018 midterms in this state, but they also had popular Democratic incumbents and very weak Republican challengers. A few years down the road, I see this year being an exception to Pennsylvania's purple status.

[Zaybay]

We need to be careful about reading too much into this. Democrats did very, very well in the 2018 midterms in this state, but they also had popular Democratic incumbents and very weak Republican challengers. A few years down the road, I see this year being an exception to Pennsylvania's purple status.

Wasnt just that, we have the congressional vote as well, and a similar overpreformance was seen there as well, not to mention state legislature races.

[ElectionsGuy]

We need to be careful about reading too much into this. Democrats did very, very well in the 2018 midterms in this state, but they also had popular Democratic incumbents and very weak Republican challengers. A few years down the road, I see this year being an exception to Pennsylvania's purple status.

Wasnt just that, we have the congressional vote as well, and a similar overpreformance was seen there as well, not to mention state legislature races.

Yeah but there's a top-down effect generally. This can be seen in cases like Kansas and Oklahoma, with D's doing very well in OK-05, KS-03, and KS-02 with overperforming Democrats at the gubernatorial level. Another example is Texas where many house races (and the house vote overall in that state) were much closer than expected due to Cruz's close race.

[IceSpear]

We need to be careful about reading too much into this. Democrats did very, very well in the 2018 midterms in this state, but they also had popular Democratic incumbents and very weak Republican challengers. A few years down the road, I see this year being an exception to Pennsylvania's purple status.

At the very least, it should dampen that hot take which was going around about PA being more likely to vote for Trump than WI.

[Don Vito Corleone]

We need to be careful about reading too much into this. Democrats did very, very well in the 2018 midterms in this state, but they also had popular Democratic incumbents and very weak Republican challengers. A few years down the road, I see this year being an exception to Pennsylvania's purple status.

At the very least, it should dampen that hot take which was going around about PA being more likely to vote for Trump than WI.

A) People say that?

B) I'm curious as to what their rationale was.

[Skill and Chance]

What are we up to on the GCB? D+9?


PA was  basically tied at D+2 in 2016. In a uniform swing environment thus the state would be expected to vote about D+7 or D+8 based on that.  The Governor's race would have a bigger differential than the Senate race obviously and the same thing occurred in 2006.

Casey outperformed by about 5% to 6%, which based on his name id, popularity in the NE and Lehigh Valley and yes record performances in the cities and the suburbs, makes a lot of sense.

I think PA is a close to even state with the national average, it is the front lines between the two demographic blocks of each party and it is a tendency to favor incumbents as someone said. There certainly was a turnout problem for the GOP as the electorate that showed up in the exit poll was Clinton by 6. While we cannot be certain as to the accuracy, there is reason to believe it might be higher than that based on the turnout data posted above but it is certainly in the ballpark.

Still for a state that turned out in a manner hostile to the President, it still was split close to even on trade and immigration policy. A plurality said Trump's trade policies help more than they hurt IIRC and the split on immigration between "too tough" and "Just Right/Not Tough Enough"  was 52%-48%. this number appeared elsewhere in the exit polls as well.

PA looks a bit like the Dem version of NC after this year.  It swings to narrowly support the new president, but the party that seems to be doomed demographically is able to fight back a lot better than expected because their base areas haven't been maxed out yet.  It's almost certainly the best medium/long term Dem prospect of the Obama-Trump states.

[IceSpear]

We need to be careful about reading too much into this. Democrats did very, very well in the 2018 midterms in this state, but they also had popular Democratic incumbents and very weak Republican challengers. A few years down the road, I see this year being an exception to Pennsylvania's purple status.

At the very least, it should dampen that hot take which was going around about PA being more likely to vote for Trump than WI.

A) People say that?

B) I'm curious as to what their rationale was.

Yeah, it got quite a lot of votes in this poll.

https://uselectionatlas.org/FORUM/index.php?topic=278854.0

[Boss_Rahm]

Two things have to happen for PA to go red: Republicans have to be enthusiastic and Democrats have to not be. That's what happened in 2016, and with the right/wrong set of candidates it could happen again. I also think it's wrong to suggest that Republicans have "maxed out" their rural vote. But for the near-/medium-term future, there will be more Democratic voters in PA than Republican voters. So PA will be a blueish state that can still go red in a good Republican environment.

[Southern Senator North Carolina Yankee]

We need to be careful about reading too much into this. Democrats did very, very well in the 2018 midterms in this state, but they also had popular Democratic incumbents and very weak Republican challengers. A few years down the road, I see this year being an exception to Pennsylvania's purple status.

At the very least, it should dampen that hot take which was going around about PA being more likely to vote for Trump than WI.

A) People say that?

B) I'm curious as to what their rationale was.

Yeah, it got quite a lot of votes in this poll.

https://uselectionatlas.org/FORUM/index.php?topic=278854.0

I say it all the time and I will keep it saying it for this reason.

Trump only won Wisconsin because of lower turnout. He got less votes than Romney did. If Milwaukee turns out in 2020, Trump would need to rebound in WoW to RoJo levels to compensate.

The same is not true for PA.

Two things have to happen for PA to go red: Republicans have to be enthusiastic and Democrats have to not be. That's what happened in 2016,

That is not what happened in 2016. In MI and WI you can say that because Trump won them while getting less votes than Romney.

In PA, Trump got almost 300,000 more votes than Romney while Clinton only lost 70,000 from Obama. In Philadelphia, Clinton was only under Obama by 4,000 votes, but Trump ran 12,000 ahead of Romney.

Trump won Pennsylvania because he got higher rural turnout than Romney and because NE, NW and SW PA reacted positively to his message on trade and immigration, which meant that a lot of people flipped from Obama to Trump in those areas. Even in 2018, the exit polls showed a plurality with a positive view on Trump's trade policies and underwater by just 4% on immigration. This with an electorate that claimed to have voted for Clinton by 6% and voted to reelect Wolf and Casey by double digits.

I think there is a lot of faulty analysis based off the 2018 results in PA, that yes ignore incumbency, the quality of the GOP candidates, the unwinding of an overextend House delegation based on a gerrymander and other factors as well as drawing on a lingering faulty narrative about the nature of Trump's win in 2016 in PA.

[Zaybay]

We need to be careful about reading too much into this. Democrats did very, very well in the 2018 midterms in this state, but they also had popular Democratic incumbents and very weak Republican challengers. A few years down the road, I see this year being an exception to Pennsylvania's purple status.

At the very least, it should dampen that hot take which was going around about PA being more likely to vote for Trump than WI.

A) People say that?

B) I'm curious as to what their rationale was.

Yeah, it got quite a lot of votes in this poll.

https://uselectionatlas.org/FORUM/index.php?topic=278854.0

I say it all the time and I will keep it saying it for this reason.

Trump only won Wisconsin because of lower turnout. He got less votes than Romney did. If Milwaukee turns out in 2020, Trump would need to rebound in WoW to RoJo levels to compensate.

The same is not true for PA.

I dont know where this idea that WI is going to become the most R state of the three when, as you point out, Trump didnt do better than Romney in raw voter totals. At least in PA and MI, Trump made gains.

[Southern Senator North Carolina Yankee]

We need to be careful about reading too much into this. Democrats did very, very well in the 2018 midterms in this state, but they also had popular Democratic incumbents and very weak Republican challengers. A few years down the road, I see this year being an exception to Pennsylvania's purple status.

At the very least, it should dampen that hot take which was going around about PA being more likely to vote for Trump than WI.

A) People say that?

B) I'm curious as to what their rationale was.

Yeah, it got quite a lot of votes in this poll.

https://uselectionatlas.org/FORUM/index.php?topic=278854.0

I say it all the time and I will keep it saying it for this reason.

Trump only won Wisconsin because of lower turnout. He got less votes than Romney did. If Milwaukee turns out in 2020, Trump would need to rebound in WoW to RoJo levels to compensate.

The same is not true for PA.

I dont know where this idea that WI is going to become the most R state of the three when, as you point out, Trump didnt do better than Romney in raw voter totals. At least in PA and MI, Trump made gains.

Even in MI Trump's gains were smaller than Clinton's losses. Clinton lost about 250,000 votes while Trump only gained about 100,000.  I think you can safely say that WI and MI were decided by low AA and Democratic turnout in key Democratic strongholds in both states. Now while I didn't examine Scranton or Pittsburgh, the Philly numbers seem indicative that Democrats actually did turnout in PA, because if they didn't Clinton would have lost far more votes relative to Obama and Trump would have won the state by 5% or 6%.

In large measure this is probably due to the local strength of the Philly Democratic operation, which is well known for getting the job done. In 2010 for instance, it is said that when Toomey saw the Philly numbers, he expected to lose.

[Zaybay]

We need to be careful about reading too much into this. Democrats did very, very well in the 2018 midterms in this state, but they also had popular Democratic incumbents and very weak Republican challengers. A few years down the road, I see this year being an exception to Pennsylvania's purple status.

At the very least, it should dampen that hot take which was going around about PA being more likely to vote for Trump than WI.

A) People say that?

B) I'm curious as to what their rationale was.

Yeah, it got quite a lot of votes in this poll.

https://uselectionatlas.org/FORUM/index.php?topic=278854.0

I say it all the time and I will keep it saying it for this reason.

Trump only won Wisconsin because of lower turnout. He got less votes than Romney did. If Milwaukee turns out in 2020, Trump would need to rebound in WoW to RoJo levels to compensate.

The same is not true for PA.

I dont know where this idea that WI is going to become the most R state of the three when, as you point out, Trump didnt do better than Romney in raw voter totals. At least in PA and MI, Trump made gains.

Even in MI Trump's gains were smaller than Clinton's losses. Clinton lost about 250,000 votes while Trump only gained about 100,000.  I think you can safely say that WI and MI were decided by low AA and Democratic turnout in key Democratic strongholds in both states. Now while I didn't examine Scranton or Pittsburgh, the Philly numbers seem indicative that Democrats actually did turnout in PA, because if they didn't Clinton would have lost far more votes relative to Obama and Trump would have won the state by 5% or 6%.

In large measure this is probably due to the local strength of the Philly Democratic operation, which is well known for getting the job done. In 2010 for instance, it is said that when Toomey saw the Philly numbers, he expected to lose.

Only WI was decided by low AA turnout.

[Amenhotep Bakari-Sellers]

PA usually tries something new after 2 terms, like they vote out incumbent party govs after two terms. So, they tried Trump, and he slash entitlements without raising cap on the wealthy on Social Security or build the wall or immigration. So, the trend went back Democratic

[Mikestone8]

Luzerne County (Wilkes-Barre) - Comparison to 2016-Pres counts in parenthesis
2018 Gov
D: 55,654 (+6%)
R: 50,642 (-36%)
T: 107,681 (-21%)

2018 Sen
D: 49,076 (-6%)
R: 57,965 (-26%)
T: 108,092 (-20%)

2016 Pres
D: 52,451   
R: 78,688
T: 135,901

Beaver County (Pittsburgh MSA)
2018 Gov
D: 36,141 (+11%)
R: 30,571 (-37%)
T: 67,829 (-20%

2018 Sen
D: 34,403 (+6%)
R: 31,895 (-34%)
T: 67,579 (-20%)

2016 Pres
D: 32,531
R: 48,167
T: 84,462

Here are the statewide figures for comparison:

Statewide

2018 Gov
D: 2,850,210 (-3%)
R: 2,015,266 (-32%)
T: 4,941,392 (-20%)

2018 Sen
D: 2,745,018 (-6%)
R: 2,109,023 (-29%)
T: 4,935,119 (-20%)

2016 Pres
D: 2,926,441
R: 2,970,733
T: 6,166,729


It would be better to compare a few counties, but that's tedious with either PA's official data or the newly castrated non-members version of the Atlas.

Wolf did well in the Northwest, but you would need precinct level numbers to determine how much Trump-Wolf voters had to do with this. Casey's massive under-performance here provides some sense, as he lost about one in five of Wolf's voters. Clearly Trump's coalition here includes more than its share of civic illiterates who don't bother turning out in midterm years. (Much like that subset of Obama voters who couldn't be bothered when their TV President wasn't on the ballot.)

The story in the Southwest looks similar, but with a smaller gap between Casey and Wolf and a larger swing back toward Democrats. Conor Lamb, who now represents the entire county, obviously has something to do with this. Again, it's striking that numbers up and down the ballot look similar except for the massive difference in the Republican total.

If anyone had precinct level results up and down the ballot in both years, I wonder what kind of an effect you could show in voter turnout based on competitive Congressional seats alone.

Looks to me as if a lot of normally Republican voters abstained (turned off by Trump?) but mostly didn't vote Democratic. So the Republicans have had a bad year, but it's probably not irreversible.

[pbrower2a]

If there is any pattern, it is with one Party creating disillusionment with its agenda and creating an opening for the other (2016), and the winning Party in that election (2016) creating much disillusionment with it in the next election and creating openings for the other Party (2018). This is more like a "Plexiglass" effect in baseball statistics in which a pitcher who has an unusually-high ERA gets shut down and effectively benched, gets more rest, and the next year goes back to his old norm (think of Anibal Sanchez with the Detroit Tigers in 2017 and with the Atlanta Braves in 2018) in a different environment. The Tigers couldn't wait to get him off their roster, and the Braves got to the playoffs while the Tigers went 64-98. Or think of the slugger who goes from .280 with 38 home runs and 92 RBI to .245 with 14 home runs and 67 RBI who gets traded away for next to nothing and comes back hitting .278 with 22 home runs and 87 RBI.

That is how Nate Silver, who went from strictly dealing in baseball stats went to discussing politics. He made the transition very well because of some of the similarities between models of sporting events and elections.

It is easy to characterize the settlement patterns of Pennsylvania. Pennsylvania used to be an R-leaning state because the Democrats did not have a strong machine in Philadelphia or Pittsburgh as they did in such places as Boston, Chicago, New York, Memphis, or New Orleans. Pennsylvania agriculture did reasonably well where it was (southeastern Pennsylvania is practically a farmer's paradise with adequate rainfall, a long and reliable growing season, and good terrain -- and the rest of the state is horrible for farming), so the rural distress did not foster 'prairie populism'. Pennsylvania voted for Hoover in 1932, and any state that gives its vote to someone who wins 59 electoral votes is obviously very partisan or giving a protest vote.


Because Pennsylvania is urban, its suburbs could harbor plenty of 'Rockefeller Republicans', people liberal on social issues but conservative on defense and economics. It could elect people like the late Arlen Specter to the Senate as a Republican. The D-R division used to be slight, but that is over.

So Pennsylvania is

(1) Philadelphia (enough said about its current politics)
(2) Philadelphia suburbs (which have been drifting from R to D like suburbs of all othe rNorthern states except Indiana)
(3) Greater Pittsburgh (but a no-growth area -- think of Cleveland)
(4) northeastern Pennsylvania (Scranton, Wilkes-Barre, Allentown -- in decline and economic distress, and willing to listen to any demagogue who promises a return to the Good Old Days)
(5) south-central Pennsylvania (Harrisburg, Lancaster, York, Reading -- much like western Michigan as a no-growth area)
(6) Erie (small and no growth)

-- and --

(7) everything else. Appalachia, which has tended to go to the Right with contempt for anything modern, exotic, or liberal. I have seen Appalachian Pennsylvania described as if "Pennsyltucky" or "Pennsylbama". It is the Backwoods, hostile to any people who might be seen in any way exotic in appearance of values.  

Appalachia south of the New York-Pennsylvania state line is very right-wing and Republican. It used to be decidedly Democratic, which might explain how Carter won the state in 1976. But whites in the Mountain South (and central Pennsylvania is in the Mountain South politically and culturally even if it is one of the chilliest areas in America climatically) were receptive to the Moral Majority and the politics that went with it. It made a sharp turn to the Right and went for Ronald Reagan, carrying an otherwise-close state in Presidential elections to the GOP in 1980, 1984, and 1988. Democrats from Clinton to Obama got some edge in Pennsylvania as Bob Dole, Dubya, McCain, and Romney were less effective in appealing to Pennsylvania voters, but Trump sang just the right tune for Appalachia with his economic populism and self-righteous expressions of ethnic and religious bigotry. Will that work in 2020? Just look at how well the Republicans did in Pennsylvania in 2018!

Approval polls suggest that Donald Trump is highly unpopular in Pennsylvania, with disapproval numbers in the high 50s and low 60s before the 2018 election.

A trend or a rebound? It depends on whether the President has the skill set of Reagan or Obama on the one side (their skill sets are much the same)  or Donald Trump. I see Trump as a catastrophic failure as President, but that is my opinion. In view of the statewide elections for high state offices and the blowout win for the Democratic Senator , and the dominance in Democratic voting for House seats (Pennsylvania is heavily gerrymandered, so the House delegation does not reflect the statewide vote), I have cause to see Trump losing Pennsylvania decisively in 2020.

Effective politicians push a trend. Ineffective ones get a rebound. People usually can't admit that their vote for an incumbent was a mistake -- unless their views change (which explains part of how Carter did so badly in 1980 -- the rise of the Religious Right while he was President) or the economy goes into the toilet (Hoover going from the most promising President in American history to one of the biggest goats). With Trump I see plenty of stuff going wrong, and I cannot say what will explain his nearly-certain failure in 2020. Sure, I despised him in 2016 -- in ways in which I did not despise John McCain, Mitt Romney, or the elder Bush -- so I can't hide my bias. But I can as easily hit Trump from the Right as from the Left. Say what you want about Reagan, but at least nobody had any question of his patriotism!