As Tuesday night’s results of the U.S. presidential election came trickling in, many residents were shocked by what they saw. Businessman Donald Trump pulled off one of the biggest upsets in the country’s history by defeating the heavily favored Secretary of State Hillary Clinton. About 61 percent of Americans viewed Trump unfavorably entering this week and, conversely, Clinton comfortably led in recent political polling for the past month. We talked to Goucher College’s Mileah Kromer, the director of the Sarah T. Hughes Field Politics Center, which conducts the famous Goucher Poll, about why the polls appeared to be so off, what swayed voters in the end, and where disenchanted constituents can go from here.
First, as a polling expert, how did every big national poll miss this so badly?
After it all shakes out, it looks like we will end up with around a 4-point miss. The final nationwide polls had Clinton around 49 percent and Trump around 45 percent. They ended up with 48 and 47 percent, respectively. This difference is within a typical margin of error.
Pollsters rely on folks telling us two central truths: Their likelihood of turnout and their real voter preference. We will certainly spend a lot of time between now and the next election trying to figure out why Trump support was systematically underestimated. Did Trump voters actively mislead pollsters about their true preferences? Did Trump voters just refuse to answer surveys? Did Democrats overestimate their likelihood of voter turnout? I’m leaning toward some non-response bias from Trump supporters. It goes hand-in-hand with their distrust of the mainstream media. They heard Trump bash the polls, they saw polls discussed by media they hate, so I have to wonder if that lead to lots Trump supporters simply opting themselves out of polls.
Is polling accuracy overrated?
No, it’s not. But, it also isn’t a flawless predictor of electoral outcome. Many of the polls leading up to the election showed narrow Clinton leads in those key swing states. Polls showed a thin margin in states like Florida or North Carolina. And, in a lot of cases the results fell within the margin of error. Keep in mind that the margin of error refers to plus or minus each number.
For example, Trump’s support in Pennsylvania in the last polls was around 45 percent. A typical margin of error and the margin of error is 4 percent, thus the expected range will be anywhere from 41 to 49 percent. Trump ended up with 49 percent of the vote in Pennsylvania. Another example is Florida, where Trump ended with 49 percent and Clinton 48 percent of the vote. The last polls before the election showed Clinton at 48 and Trump at 47, a difference within the margin of error.
However, none of this matters come Election Day. Whether pollsters predicted the winner is the only thing people remember. And, this election cycle, pollsters failed to do so. We are modeling the behavior of humans who are inherently unpredictable at the individual level, at the same time there recognizable patterns of group behavior.
Similarly, the entire pundit class missed this. Is the so-called Capitol Hill elite and media that out of touch with middle America? Why?
I think that the pundit class missed the message that was resonating behind the bluster and inappropriateness of Trump’s rhetoric. What working class voters heard was a tough stance on ISIS, economic policies that centered on protecting manufacturing jobs, and harsh criticism of the Washington establishment. While the talking heads were focusing on how offensive Trump’s comments were, large parts of the country simply weren’t moved by them.
Why did Hillary Clinton under-perform in some of the generally blue states like Pennsylvania?
It’s the result of poor messaging to voters outside of the urban centers of the state. This cycle had some cringe-worthy descriptions on blue-collar Pennsylvania—like Bill Clinton’s “coal people” or when Hillary Clinton talked about bringing renewable energy to coal country but mentioned that “we’re going to put a lot of coal miners and coal companies out of business.” This sort of rhetoric, even though the Democrats have fought hard to protect the union jobs of mineworkers in the region for decades, drives the narrative that establishment politicians don’t know or care about the needs of blue collar voters. Trust was an issue, too. It just wasn’t an issue unique to Pennsylvania.
This is probably an impossible question to parse, but if you could try. Does this presidential result represent a strong repudiation by many Americans, of say, the typically more liberal social values of the Democratic Party—given shifting demographics and issues around gender, race, and ethnicity?
I will say this, we need a better understanding of the growing Latino vote. They are far from a monolithic group and political parties still aren’t sure how to fully engage them in electoral politics. Even with Trump’s hardline on immigration, Latino turnout remained flat from 2012. Moreover, nearly 30 percent of Latinos voted for Trump.
Or, did economic issues remain the core driving force this year?
Yes. Pocketbook votes matter. And, economic recovery only matters if it reaches all parts of the US. There are places that are still really struggling to adjust to a changing economy. These regions broke for Trump. There is the unbelievably underlying conventional truth to this election: Republicans voted for the Republican candidate and Democrats voted for the Democratic candidate. The problem for Clinton is not enough of the Democratic coalition showed up—and white women, who she was counting on to flip, went Republican.
How much did Trump’s stunning numbers influence the key Republican U.S. Senate races?
A wave of Democratic voting to stop Trump never came to be. People don’t vote to stop someone, they vote because they believe in the message. Many of us, myself certainly included, thought that Trump’s comments on women and minorities would have prevented him from winning. And, that his downfall would take a few Republican seats with him. This was clearly not the case.
Has Trump, for all intents and purposes, put an end to traditional presidential campaigning? In the age of social media and the celebrity-news complex—do traditional endorsements, coalition building, TV ad buys, and get-out-the-efforts matter any longer?
The men and now woman who make it to the big show all leave their mark in some ways. I think all candidates will be more open to the media now. Gone are the days of the scheduled press conference. It will be all about doing the circuit of cable and national news as often as possible. However, let’s see how Trump does before we talk grand shifts in political campaigns. If things go the way many Democrats expect it to, I doubt the American voter will gamble on an outsider for a while.
Does this result potentially help or hurt Gov. Larry Hogan in his re-election bid? He famously wrote-in his own father’s name rather than vote for Trump.
It is way too early to tell. The Democrats in this state will be fired up for the midterm elections. If Trump does something unpredictable and wildly unpopular, it could drag Hogan down by virtue of shared party identity. However, Governor Hogan has proven thus far that he can focus on Maryland and ignore the national political noise.
Bernie Sanders beat Hillary Clinton in the Wisconsin and Michigan primaries and polled about 10 points higher than her versus Donald Trump by the end of the primaries. Would he have beaten the president-elect?
We’ve never had a chance to see how Republicans would have messaged against a “socialist from Vermont.” And, frankly, it doesn’t matter. Disaffected voters get this week, but next week all the Monday morning quarterbacking should stop. If you are a voter distraught over the outcome, focus your energy on the midterm elections, rather than the “what ifs.”