Making sense of the election

Penn political scientists helped a virtual audience process polling, voter turnout, litigation, and a chaotic presidential election.

A mail worker sorts through a large pile of mail-in ballots.
A panel of Penn political scientists helped a virtual audience make sense of the presidential election and what comes next.

From poor poll predictions to the Trump administration’s litigation efforts to Latino voting trends, a panel of Penn political scientists helped a virtual audience process the chaotic presidential election and the events of the past week, discussing what was expected, trying to make sense of the unexpected, and grappling with what is yet to come.

Moderated by Michele Margolis, the Nov. 9 discussion included Daniel Hopkins, Diana Mutz, Michael Jones-Correa, and Rogers Smith.

Close race, voter turnout

“This was obviously a very, very close race, and the idea that the geographic distribution of a tiny fraction of voters is so impactful in who controls the American presidency is very important,” Hopkins said. He noted that the country is highly nationalized right now. “One of the reasons that we’ve all been so glued to our televisions of computers is precisely because we care deeply about the presidency. Very few people care about their state’s governor with the same level of passion that we care about the presidency,” he said.

Pennsylvania was, as predicted, the linchpin of the election, but Philadelphia’s turnout was about what it was in 2016, at a time when turnout was up across the state, Hopkins said. It’s unclear if the fact that fewer college students are in the city due to the pandemic played into that, he said. But on the flip side, the suburban counties just outside of Philadelphia put up record margins for President-elect Joseph Biden and helped him overcome the fact that in the most rural parts of the state of Pennsylvania President Donald Trump outperformed his margins for 2016.

“In some ways it’s not surprising that this election came down to Pennsylvania because Pennsylvania actually looks a fair amount like the nation as a whole,” he said.

Mutz added that in an election cycle “where almost nothing seemed like business as usual we had record-breaking turnout during the height of a pandemic. Who would have predicted that?”

There was GOP concern about the legitimacy of the outcome of the election before the first ballot was even cast, there was the steepest quarterly decline in GDP in modern history, and there were widespread protests about racial injustice, she noted.

“Everything seemed tumultuous and totally in turmoil, and yet data suggest that, when it comes to voter behavior, things were not at all unique,” Mutz said.

She cautioned against equating the percentage of Americans who take the time to vote as a sign of the health of our democracy. Her research among voters has shown it differently.

“Their reason for voting is not that they’re thrilled with the state of democracy; it’s that they’re mad as hell. And that was true both for people who were supporting Trump as well as people who were supporting Biden,” Mutz said. “While it’s easy to look at voter turnout and say, ‘Wasn’t it terrible we don’t have higher levels?’ in some ways that’s a luxury of a population that isn’t really unhappy with the state of affairs.”

Screen shot of a Zoom call with a woman and two men in the top row and a woman and a man in the bottom row
Penn political scientists shared their thoughts on the election at a virtual event on Nov. 9, 2020. They included moderator Michele Margolis (top left), Rogers Smith (top center), Micheal Jones-Correa (top right), Diana Mutz (bottom left), and Daniel Hopkins (bottom right).


“It’s clear that the public polling was incorrect,” Hopkins said. “The critical question is what happened? A key thing to know about polling nowadays is it is very hard.”

He shared an anecdote from last month when he tried to survey infrequent voters in Pennsylvania via Facebook. He started with a list of 10,000 registered voters, successfully served ads to more than 1,000, inviting them to take the poll. Forty-eight people clicked the link, and six people completed the survey, he said.

“The central challenge with polling nowadays is that there are many Americans who vote but don’t respond to polls,” he said, adding that he and Margolis have worked exit polls together and people just walk by and won’t talk.

“It’s not that they’re shy about their allegiances; there were people wearing Trump hats saying, ‘My vote is private.’”

As for those who claimed the 2016 results were off, Mutz said the polls perfectly predicted the popular vote, but there was much less focus on state polls in 2016.

It’s a disjuncture between national polling results and the Electoral College results, Mutz said.

Statewide polls are of lesser quality, as well as have smaller sample sizes than national polls, so it is not surprising they predict less well, she said.

“It is very difficult to predict Electoral College outcomes with a collection of state-by-state polls because if they’re not battleground states there’s hardly any polling at all,” she said. “Getting decent response rates in polls either requires huge budgets and other things that the kinds of organizations that tend to sponsor statewide polls just don’t have.”

Latino vote

Jones-Correa noted the increasing importance of Latino voters in every election, saying there are about 800,000 new Latino voters every year.

Despite that, the pattern of how Latinos vote has been consistent since the 1960s: Two-thirds of the Latino vote goes to the Democratic Party and a third goes to Republicans, he said. Issues do not appeal to these voters equally. For example, first-generation Latinos, born in other countries themselves, tend to identify with immigration issues, while those who have been in the country for three generations or more may not be connected to those issues.

“Generation really matters,” he said.

Jones-Correa criticized how Latinos are always lumped together, saying there’s no such thing as a singular Latino vote. The term is artificial and was created by the government, the media, advertisers, and those who want to create a market. “These are artificial labels, and you have to keep in mind there are differences underneath,” he said.

He gave the example of Florida to illustrate his point regarding a heterogenous group. “The shorthand for explaining Florida is often that it’s all about Cubans, and that is just not right,” he said. In Florida, the Latino electorate is about 28% Cuban American, but Puerto Ricans are very close behind. They are moving in different directions: Puerto Ricans are more tilting towards the Democratic Party and Cuban Americans somewhat more toward the Republican Party, he said. Then the electorate breaks down further with people whose origins are in Mexico, the Dominican Republic, Colombia, Venezuela, and Nicaragua. “You have the representation of the entire hemisphere,” he said.

The issue for the Democratic Party is they have done a “terrible” job at outreach to Latino voters in Florida, he said. “You can see this in the 2018 elections when there was a Senate seat and a governor’s seat up for up for grabs that year and the Democratic candidates didn’t have a bilingual website and had almost zero bilingual outreach.”

The Republican Party and Trump administration targeted South Florida, and they did it successfully, he said. But he added that doesn’t mean the Latino vote in that area is strongly partisan.


The Electoral College votes aren’t official until Jan. 6, when the president of the Senate, Vice President Mike Pence, opens the Electoral College tallies with Congress, and then the election is finally resolved.

“There’s a lot that can happen between now and then, particularly this year because there is a lot of litigation about this election, challenging the legitimacy of some of the returns that have been reported,” Smith said.

Republicans have every right to raise legal challenges to any part of the electoral process they think has been done illegally, but whether or not that’s a good thing to do for American democracy is a separate question, he said.

“The litigation has virtually no chance of overturning the result of this election unless real evidence of election fraud comes to light,” he said.

So, what is the point of all this litigation?

“There is a hope that evidence of massive fraud will appear and that might result in reversing the election outcome. There’s also a scenario that Rudy Giuliani and others have suggested several times, that the massive litigation leaves it uncertain whether the elections were fairly conducted or not and that creates at least enough public perception of illegitimacy to call the election into question,” Smith said.

That could open up state legislatures deciding who the states’ Electoral College delegations will be. “But the prevailing understanding is that the state legislature has to choose the results of the state’s popular vote winner; it would be pretty outrageous to do otherwise,” he said. “This is definitely a Hail Mary approach.”

He adds that the courts have shown that they’re trying to resolve the cases around the country as quickly as possible, and they’re pushing hard to make sure that litigation isn’t prolonged without any evidence.

“The law is very clear: Donald Trump’s term expires on Jan. 20, when he is no longer president of the United States,” Smith said. “Unless the Electoral College results are certified in his favor on Jan. 6, that simply can’t be altered. What the president can do is make the transitional process less successful by not authorizing the General Services Administration to work with Biden to prepare for the next administration, and that’s a real danger. But there is no way that Donald Trump can legally stay in office after Jan. 20.”

A video of the virtual event is available to view here.