It’s been a few days since the stunning results of the presidential election, and we’ve all had some time to process. There’s a lot that reporters and political consultants and pollsters and, hell, even the campaigns themselves got wrong about this election.

The jury is still out on why. Exit polls are still coming in. But we can definitively say that there were some predictions we made and wrote about that ended up being wrong in the end. So we decided to sort of fact-check ourselves.

Here are six things we got wrong in the lead-up to the presidential election in Pennsylvania and the math behind what ended up happening:

1. ‘Any candidate who wants to win a statewide election in Pennsylvania has to do well in the Philadelphia suburbs’

This came from an August poll that showed Clinton was up 40 points in the Philadelphia suburbs. Besides that prediction being wildly incorrect, the idea that any candidate who wants to win Pa. has to do well in the Philadelphia suburbs wasn’t exactly correct.

We always thought that Pennsylvania was won and lost in the Philadelphia suburbs. A third of the state’s registered voters live in Philadelphia and its suburbs. The four counties surrounding Philly — Delaware, Chester, Bucks and Montgomery — had become battleground areas as voter registration has shifted over the last decade or so.

This year, those four counties actually voted more Democratic than they did in 2012 and voter turnout was similar. Trump got a smaller percentage of the vote there than Romney, McCain or Bush.

But that didn’t swing the state. Instead, Trump flipped several previously blue counties and cleaned up in southwestern Pennsylvania, putting him over the edge in the state.

In addition, this story documented how Trump’s poll numbers in the Philadelphia suburbs had “gone from bad to worse.” It said Clinton was up in the suburbs by 28 points. In reality, she won the suburbs by just over 10 points.

We have to note that these were polls. They were not an election, and they were taking the pulse of a sample of people at the time. Feelings change, people vote differently than they might think. Since the election, Politico wrote about pollsters reflecting on what might have gone wrong: “The potential causes for error run the gamut from a secret army of Trump voters lying to pollsters, to a late-breaking wave of voters flocking to Trump after most of the polling had concluded.”

2. ‘The Republican presidential candidate is staring down what some say could be a landslide loss’

Oof. This story documented what it was like for Trump supporters at a rally in Newtown, Pa. toward the end of October. The article showed that even though Trump was plummeting in the polls both in Pennsylvania and the nation in general, his fans were still supporting him. “In denial” is actually how we phrased it.

The polls showed Clinton leading by so much in Pennsylvania that we were all pretty sure she was going to win. Here’s the photo we planned to use on our post about how Clinton won PA:

Balloons fall as Hillary Clinton and Tim Kaine stand on stage with their families after Clinton spoke during the 2016 Democratic National Convention at the Wells Fargo Center.
Balloons fall as Hillary Clinton and Tim Kaine stand on stage with their families after Clinton spoke during the 2016 Democratic National Convention at the Wells Fargo Center. Credit: Jack Gruber-USA TODAY NETWORK

Clearly, that turned out to be wrong! Though polls at the time were showing Trump could have lost in a landslide, he didn’t. The candidate won in Pennsylvania — even winning some wards in Philadelphia — and he, of course, won the White House.

His supporters predicted this, at least publicly. The day before Election Day, PA Congressman Scott Perry, a Trump supporter, said this in a statement: “Pennsylvania will lead the way in sending Donald Trump to victory on Election Day. As the electoral map continues to shrink, it’s clear the American people are ready for change. Voters across the Keystone State have heard Donald Trump’s message of change and are coming home in droves.”

3. ‘Trump isn’t that popular among PA whites’

Well, this only turned out to be kind of wrong. A CNN poll that came out at the end of September showed “about 48 percent of white CNN poll respondents said they planned to vote for him (and 41 percent for Clinton).” That’s less than the share of the white vote the Republican presidential candidate got the last three elections in Pennsylvania.

Exit polls, according to The (Allentown) Morning Call, showed that white voters made 80 percent of the electorate in Pennsylvania (slightly more than in 2012, when they made up 78 percent). White women were split on the candidates — in 2012 54 percent of white women voted for Mitt Romney — and Trump won votes from more than 60 percent of white men. When it comes to white men without college degrees, that support was even higher.

4. ‘She’s dominating in the Keystone State when it comes to other factors that point to strong campaigns’

Everybody laughed about Donald Trump’s lone Philly outpost on South Street. And around the state he didn’t have much else of a presence. We wrote about his lack of offices, cash and employees in the state and noted how Clinton was “dominating” Trump in all of it.

She had 56 campaign offices compared to 12 and had raised nearly $8 million in the state compared to about $2 million for Trump: “As the latest polls show the Pennsylvania presidential race widening to Hillary Clinton’s advantage, her supporters have another reason to be confident: she’s dominating in the Keystone State when it comes to other factors that point to strong campaigns.”

Trump’s PA senior advisor, David Urban, told us back then, “We will have all the offices, staff and resources we need to win in November, and any attempt to measure our strength by outdated metrics such as this one simply overlook the real state of play in the race.”

He was right.

5. ‘The one week Pennsylvania really liked Trump’

The headline on this story turned out to be pretty astute: “Pennsylvania voters unfazed by the Access Hollywood tape, polls show.” But there was a section in the story about “the one week Pennsylvania really liked Trump.” It referenced a time in late June when polls were closest.

There weren’t many polls that had this race looking close after that, save for one from the right-leaning Trafalgar Group that showed Trump up by a point in the days leading up to the election.

The candidates also turned out to be close in another week: The second week of November. Trump ended up winning by about 64,000 votes out of just under six million total votes cast. That’s a margin of about a point.

6. ‘This was the year of the suburban Philly woman’

Both candidates were courting women in the Philadelphia suburbs. We also wrote: “In the year of Trump versus potentially the first woman president in history, women across the country have become a vital voting bloc for both sides. In the Philadelphia suburbs where educated, white women make up a large chunk of the population, that’s even more true.”

Both candidates really were focused on this group of voters. But it was white men — not Philadelphia suburban women — who put Trump over the top in Pennsylvania.

Bonus: What we — or this Main Line mom —  got right

At an October event with Ivanka Trump in the suburbs, one woman had what was maybe the most correct observation of the election cycle.

“Polls,” she said, “are for skiers and strippers.”


Mark Dent is a reporter/curator at BillyPenn. He previously worked for the Pittsburgh Post-Gazette, where he covered the Jerry Sandusky scandal, Penn State football and the Penn State administration. His...