Thread by @lara_putnam

So I tweeted earlier today about Trump being uniquely impactful vis a vis two groups of voters👇. But one could as easily argue, looking at PA county results, that the remarkable thing is how un-unique 2016 & 2020 look. [quick thread]

https://twitter.com/lara_putnam/status/1324382649864069123

Here's the GOP share of presidential votes from Bill Clinton's first race to today. (The counties are non-metro western PA, w/ a few large counties from the non-Philly-suburb east.) They're arrayed from least rural to most rural. Trump's two runs are in blue ('16) and green ('20)

If any candidate looks consistently unusual it's Barack Obama, who in quite a few places (inc. some very white & rural ones) momentarly clawed back a trend of steady GOP gain. Trump 2016 largely just jumped GOP totals back to where you wouldve thought, pre-Obama, they were headed

There's a consistent, tho not very large, rural vs less-rural divide in terms of 2016-20 shift: counties less rural than Mercer (44% rural pop) saw Trump's vote share hold steady or edge down, while from Mercer on out, counties saw Trump slightly improve, tho no faster than trend

Here's Dem 2-party-vote margin instead. NEPA & Erie had been mixed, w/strong Dem ys under Obama. But all of SWPA outside Allegheny saw Dem collapse start after Bill Clinton's 1st term & never look back. Well, until 2020: which saw small improvmnt (=big trend change) in less-rural

Mainly I look at this & marvel at how long Greene and Fayette stayed loyally Dem when onetime strongholds like Armstrong & Westmoreland were long gone. That's the United Mineworkers, & the enduring impact of firsthand experience of labor struggle on a generation that's now passed

It's also striking that every less-rural county up thru Washington (=30% rural pop) saw Dems gain at the presidential level in 2020, though nowhere by much. Still, in Westmoreland Beaver & Washington it was the first time in ~30 ys that had happened

Given downballot results this week, I'd say the small-but-opposite shifts btwn 2016-20 likely reflect the varying balance, across counties, of two relatively small subgroups: Republicans & Independents who don't like Trump, and previous non-voters who do

https://twitter.com/lara_putnam/status/1324382649864069123

Looking at county-level results treats counties as monoliths: but of course, the dynamics driving these changes over time have everything to do with the ways they are not. They encompass diverse people, & groups in motion in opposite directions. democracyjournal.org/magazine/57/ru…

Here's a different way of thinkng about shifts from 2016 to 2020: why doesn't rural Pennsylvania look like rural Ohio? Given how many new voters Trump came up w/ in rural counties, how did Dems in PA so consistently come up w/ nearly as many, to counter them & hold margins close?

As a reminder, here's what gubernatorial vote shifts in PA looked like from 2014-2018 (graphic from a fabulous analysis by @Elaijuh you should go back and re-read now). So a Dem rural hold from 2016-20 was in no way a foregone conclusion: all the contrary inquirer.com/politics/inq/p…

Bottom line, as far as I see: Democrats found new voters willing to come out & vote against Trump in an unexpected range of places, includng both upscale+less wealthy suburbs+rural regions. Ticket splitting, tho, suggests only a very partial embrace of Democrats' agendas. Mnwhle:

Donald Trump *also* found new voters, also from a wider-than-expected array of places, incl suburbs & rural areas one might have thought he'd maxed out + some urban communities. More of these voters did less ticket splitting.
Their future engagement & loyalties are open questions

To me, if these results don't have you tearing up your pre-write on 2020— & pausing to question your interpretation of 2018— you're doing it wrong