Appalachia Strikes Back

On Election night, my eyes were fixated on Pennsylvania, as the vast, less populated areas sent a tsunami of Republican votes towards Philadelphia. The wave had few boundaries: Republican votes just kept coming in Westmoreland, Erie (ERIE!), Luzerne, even neighboring Lackawanna, with just Allegheny slightly less so. Even in the Philadelphia Collar (which Varad Mehta wrote about earlier today), Clinton’s bump up among the upper class was being muddled by a surge of white working class votes. The Philadelphia volcano of Democratic votes briefly  rose out of this red sea, sputtered at 584,000 votes, and subsided, becoming a Kerguelen Plateau of vanquished Democratic votes.

After looking at the map over and over, I split the state up into two regions, which, I feel, best highlight the dramatic scale of change between the 2012 cycle and 2016: Appalachia as defined by the Appalachian Regional Commission, and the remaining 15 counties, which I’ve informally labeled Acela. The leftover counties include the Lehigh Valley, Pennsylvania Dutch Country, the Philadelphia Suburbs, and Harrisburg, and while diverse in their own right, mostly share the same partisan shift.

Examining the current picture, it’s worth looking back at the 1988 map, the last red one for Republican Presidential contenders in the state:

pa1

 

Back in 1988, Republican Presidential candidate George H. W. Bush was soundly rejected in the Pittsburgh metropolitan area, lost Philadelphia two-to-one, and carried the state, barely, by running up enormous numbers in south central Pennsylvania and the Philly Collar. Democratic candidate Michael Dukakis won the Appalachia region by over 80,000 votes (including a massive six-digit margin out of Pittsburgh), but lost the region I’ve labeled Acela by 185,000.

Republican fortunes waned as time progressed. Democrats enjoyed a changing-of-the-moderate minds in the suburbs, and because the population growth of Acela dramatically outpaced Appalachia, they added over 757,000 votes by the 2012 election cycle to the Republicans’ paltry 88,000 in the region:

pa2
                        Net change in votes from 1988 to 2012, Acela Region

Republicans had made some progress in the rest of the state, building up a sizable advantage in the very same Pittsburgh suburbs that had gone heavily for Dukakis. But their gains were not consistent across the state: Romney performed poorly in Erie, Scranton, and Wilkes-Barre as many non-college white voters still stuck by the Democratic Party:

pa3
                    Net change in votes from 1988 to 2012, Appalachia Region

While Republicans had performed well enough in 2012 to cut their 2008 deficit in half, they were hundreds of thousands of votes shy from taking the state back. It was going to take an unprecedented surge, not just in one metro area (Pittsburgh), but in every conceivable pocket of the state to close the gap.

Enter 2016.

As this cycle progressed, the signs were there something had shifted: Clinton collapsed in cycle-over-cycle primary votes in the white working class pockets of Bensalem, Levittown, Wilkes-Barre, and Scranton. Republican voter registration exploded, and continued to grow after the primary deadline in many areas where earlier growth had been written off as “DINOs changing identities”. Philadelphia voter registration lagged badly behind 2008 and 2012 figures. Still, polling insisted Trump was doomed because of sound suburban rejection, so the draft copies were already being written, I’m sure, about the “Red Pennsylvania fantasy”.

Against a candidate like Donald Trump, you would have expected the southeastern region to have upped its Democratic tallies sizably, as it had done over the previous twenty-four years. But even in Clinton’s best region, it wasn’t to be:

 

pa4
                            Net change in votes in Acela Region, 2012-2016

Clinton added 34,206 votes to Obama’s re-election tally in this region, and won it 1,866,702 votes to 1,419,066- a margin of 447,636 votes. But this margin had contracted from 2012, because Trump added over 70,000 votes here. The Democratic margin in Philadelphia contracted as Republicans piled on elsewhere:

changeinvotesacela20122016

This blunted Democratic edge was still an insurmountable one, had Trump performed as strongly as Romney in Appalachia: Mitt carried the region 1,332,161 to 1,157,598, a respectable 174,563 margin.

Instead, Trump carried Appalachia by nearly half a million, and snuffed Clinton’s changes out cold.

pa6
                  Net change in votes in Appalachia Region, 2012-2016

Outside of Centre County (Penn State) and Allegheny County (Pittsburgh), everything shifted, hard, to the Republican. Luzerne, Erie, and Lackawanna’s were especially massive, considering their 2012 margins:

pa7

Over the preceding six cycles and twenty-four years, Republicans had added a net quarter-million votes to their 1988 performance. In 2016, they added another two hundred thousand, and the Democrats shed over a hundred thousand- the very margin Obama had won the state by four years earlier, 310,000- totally evaporated here.

The 15-county Acela region cast 683,000 more two-party votes than this giant sprawling region, and it didn’t matter. Chester County ran towards Clinton by 24,000 votes. It didn’t matter. Montgomery’s Democratic tally shifted by a five-figure margin. It didn’t matter. Metro Philly, which had turned dark blue over the preceding cycles, pretty much ran out of people to turn. It hit its ceiling, despite a larger population. The less populated, forgotten Pennsylvania jumped.

Appalachia, and Trump, won.

 

Note: the Philadelphia Election Commission updated the county’s tallies today shortly after this post was published. We have edited the figures for the county and the Acela region, including affected charts.