Campus News

Can We Trust the Polls? UK Experts Answer Pressing Questions Ahead of Election Day

of
Getty Image
Photo of Michael Zilis

LEXINGTON, Ky. (Nov. 2, 2020) — A public health crisis, movements against social injustice and the rampant spread of misinformation — 2020 has brought unprecedented challenges. Amid this uncertainty, Americans are casting their ballots in a pivotal presidential election.

In recent weeks, polls have shown President Donald Trump trailing behind Democratic nominee Joe Biden. Does Trump still have a visible path to a second term? Will we receive the highly anticipated results on election night? And could the outcome be contested?

Experts at the University of Kentucky are working to answer those pressing questions.

Stephen Voss, an associate professor in the Department of Political Science in the College of Arts and Sciences, has expertise in elections and voting behavior. Michael Zilis, an assistant professor in the same department, teaches courses in American government, constitutional law and judicial decision-making. 

In the Q&A below, Voss and Zilis provide expert insight to help you — the voter — stay informed.

UKNow: Our country has become increasingly polarized, especially the political dialogue. Based on early voter turnout, do you think that dialogue will translate into action?

Zilis: As of one week prior to Election Day, nearly 70 million Americans had already voted — either in person or by mail. This number dwarfs the number of votes that were cast early in 2016. In fact, the early numbers represent more than half of all votes cast in the 2016 general election. Now, a couple words of caution — early voting has become more readily available due to the COVID-19 pandemic, and we do not know what Election Day turnout will be. Even so, early indications are that interest in this election may be translating into a very high level of turnout compared to previous elections.

Voss: Additionally, this upswing in political engagement does not result from healthy developments in our political system. People are highly mobilized in a negative way — not to get behind a candidate who excites them, but to stop their political enemies. The U.S. is becoming dangerously tribal, with the divisiveness spreading beyond public life into spheres once thought private. We need to worry that such polarization will destabilize the political system.

UKNow: Biden has a lead in the national polls. But in 2016, Clinton also led in the weeks prior. Can we trust the polls? And if they’re wrong, would that be in Trump’s favor?

Voss: The 2016 pre-election polls were not nearly as bad as most people think. Nationally, the final numbers showed Hillary Clinton beating Donald Trump in the popular vote by only one percentage point more than she actually did. And even that small error was not really caused by polling alone, because many Trump voters made their decisions after pollsters had left the field. The polls gave a great picture of where voters stood at the time, and the outcome that emerged fell comfortably within the margin of error.  Only the state polls in the Midwest failed to live up to hopes, and that’s partly because everyone assumed those states would be safely in the democratic column, so they had not invested in running many high-quality polls in the region.

The real problem in 2016 came when people looked at all the polls indicating a narrow Clinton victory and assumed that the consistency of her lead meant that they could be almost certain she would win. If you have five polls all showing the same thing, but they are all skewed in the same way, then you’re not narrowing in on the truth — you’re narrowing in on an error.

Zilis: In conclusion, can we trust the polls? Yes, if we keep a few things in mind. It's better to look at trends across a large number of high-quality polls rather than focus on one attention grabbing result. And while polls often give a good sense of the race, it is not uncommon for them to be off by a few points from the final outcome. Of course, a couple of points can make quite a bit of difference in extremely close elections. It would not surprise me to see Trump slightly over-perform his current polling, but it would also not surprise me to see Biden do so.

UKNow: How big of a role will the handling of the COVID-19 pandemic and movements against social injustice play in who people vote for? How have those issues shaped this election?

Zilis: It would be difficult to understate the impact of COVID-19. It is among the very top issues cited by voters. At least one of the other major issues, the economy, has been dramatically impacted by the pandemic. Plus, the pandemic has shaped the nature of voting itself. For example, states are now much more likely to allow — and voters much more likely to cast — mail-in or early in-person ballots. Kentucky is a great example.

Concerns about racial injustice have a major impact as well, but in a different way. For example, these concerns have sparked a widespread movement, and those implications go far beyond voting. I expect that activism in favor of equality and reform will remain a key aspect of American politics in the years to come.

UKNow: Will a winner be declared on Tuesday night? If not, how long might it take before results are announced?

Voss: If Biden performs as well as the polls suggest, it’s likely the writing will be on the wall. He will not have won in an official sense, but everyone will know that’s where the outcome is headed. If the race ends up being a lot closer than expected, the whole process could end up mired in lawsuits, and we might not know the result for many days.

UKNow: In addition to casting doubt about the election results, the president has refused to commit to a peaceful transfer of power. This is something unheard of to Americans. What are the likely scenarios we could see it the president makes good on that refusal?

Voss: I have little fear that President Trump would refuse to step down if he loses the election. That sort of thing happens in countries where a president can call on some cohesive groups in the society — for example, the military and the corporate elite — whose opposition to the winner is so unifying that they are willing to bring down the system in order to prevent a transfer of power. I fail to see a single institution or politically powerful group in America that would be willing to throw the country into civil unrest in order to prevent a President Biden.

UKNow: The past few years in particular appear to have increased the divide between the two major political parties. Do you believe this election is capable of uniting the country, especially after such a tumultuous year globally? 

Zilis: In terms of ending the trend of extreme polarization, I am doubtful that one election will have this impact. However, I will be watching for whether either candidate can claim a commanding mandate. A big margin may help to lessen partisan discord that would occur in the aftermath of a close, contested election.

Voss: I expect that this election will continue the country’s downward spiral, regardless of who wins the battle for the White House. Another Trump victory likely will lead to an explosive response among disappointed portions of the electorate. A narrow Biden victory likely will lead to accusations of a stolen election. Even if Biden wins as soundly as the polls suggest he might, and enough democrats ride those coattails to take control of the Senate, they will be under severe pressure from the left wing of their party to sweep aside longstanding institutional practices so that they can achieve their policy goals as swiftly as possible.

UK is home to some of the world’s most renowned thought leaders, and they stand ready to answer pressing questions. From epidemiology and virology to constitutional law and political science — faculty and staff expertise spans a broad range of newsworthy topics. Through a comprehensive database, the UK Office of Public Relations and Strategic Communications is working to quickly connect those experts to statewide, regional, national and international media outlets.