Skip to Main Content

UML Pollster Josh Dyck Takes Pulse of Divided Nation

Center for Public Opinion Director Offers Insight into 2020 Election

Joshua Dyck, director of the UML Center for Public Opinion
Despite deep division and a pandemic, Josh Dyck, director of the Center for Public Opinion at UMass Lowell, predicts plenty of voters will turn out Nov. 3.

10/02/2020
By David Perry

It’s the busy season for Josh Dyck, director of the Center for Public Opinion at UMass Lowell. An associate professor of political science, Dyck oversees the center’s expanding polling operation, which includes surveys in battleground states this election cycle. 

Center for Public Opinion polls are routinely used far and wide by news organizations to show momentum in a variety of races, including the presidential contest. 

Through polling, Dyck takes the public pulse and gauges the political temperature of a deeply divided American public. This fall, there are added twists – a pandemic, deep racial and political fissures, more mail-in ballots than ever before, and a president casting doubt on the process. 

Despite the chaos and uncertainty surrounding the presidential campaign, Dyck believes we’ll know on Nov. 3 who will hold the nation’s top office. We caught up with him as the presidential election entered the home stretch and as he waited to sift through a batch of fresh data. 

Q: Voting has always seemed like a Norman Rockwell moment, a sacred American duty. Has it now been weaponized by those who cast doubt on the voting process and on the legitimacy of the results? 

A: There are three critical parts to having a functioning democracy. The first is that the choices have to be real. You can’t have a sham election where you coronate a dictator. The second is the voting system and the results have to be fair and legitimate, and they must be seen as fair and legitimate. The third is that the losers have to consent to be governed once it’s all over. When a candidate, politician or political party decide as a matter of strategy to cast doubt on election results before the election has even begun as a means to hedge their bets, it does unbelievable harm to the democratic fabric. The shame in all of this is that our voting process works very well and empirical political scientists have found no evidence of significant fraud or abuse of the voting process. 

Q: We hear a lot about mailed ballots being a potential problem. Is voting by mail any riskier than voting in person, in terms of ballot security? 

A: I’m going to give a two-part answer to this question. So, the first part of this answer is that there are many states in this country that run very successful all-mail elections and have for years, decades even. This includes Oregon, Washington and Colorado. From a ballot security standpoint, there’s not much of an issue with fraud or abuse. 

However, there is some concern that local election officials have considerable power to decertify ballots that are not completed correctly. For instance, many people who have never voted by mail forget to sign their ballot or use the security envelope provided. I recommend that those voting by mail double check all the instructions, mail their ballots back at least two weeks ahead of time, or, better yet, hand deliver your ballot to your local election official. 

Q: Are you seeing any trends or marked differences from 2016 or 2018 in your pre-election polling? 

A: You’ll have to wait and see, but we will be very busy this October. One thing I can say is that I expect turnout to be very high. This is a very enthusiastic electorate. 

Q: Do you think it will come down to one state deciding the election? If so, which state? 

A: No, I don’t. I expect that we will know the final result before midnight on election night. If there is a single decisive state, it is likely to be Pennsylvania. 

Q: Do you think the election may be decided by a demographic group? 

A: Generally, I think too much attention if often paid to demographic changes from election to election or to pinning electoral fortunes to a single group. This is a retrospective election on Donald Trump and it will ultimately be decided by his approval rating, his handling of COVID-19, his handling of police violence and protests, and his handling of the economy. Those are the issues of this election. However, I am very interested in how the growing Latino electorate affects the ultimate disposition of the vote in two states: Arizona and Texas. If Biden wins either of these states, it’s because the demographics of the electorates in those states have been shifting year over year. 

Q: Many states like Massachusetts have adopted expanded mail-in/absentee voting and early voting laws this year, in response to COVID-19. What do you expect the impact of those changes will have on voter turnout? 

A: This answer will likely surprise some people, but the research on early/absentee voting is that it largely conveniences people who would vote anyhow and doesn’t do much to increase turnout. In a pandemic, it makes a lot of sense to give people greater flexibility in how they vote, but I expect this, like 2016, to be a high turnout election, somewhere around 60-plus percent of the eligible voter population. Early/mail-in/absentee voting will help keep voting rates high by keeping those who might be scared off by COVID-19 in the electorate, but they won’t help them go any higher. 

Q: As we grow more and more technologically savvy, we’re taking longer to determine a winner for president. When do you think we will actually learn who wins the 2020 presidential race? 

A: It’s taking longer because we are closely divided, and because we have worked to convenience voters at the expense of expediency. But I still think we’re going to know the winner on election night before midnight, Eastern time. I know it’s possible for protracted counting to go on, but I don’t think the election will be decided by a single state, and the news networks, using exit polls and available returns, are likely to be able to call enough states on election night to project the winner. If you’re reading between the lines here, another way to read this is that I’m not sure this election is that close. 

Q: What is your biggest nightmare scenario for Nov. 3? 

A: That this election is not won decisively by the votes of the American public and that we end up within what legal scholar Rick Hasen has called “the margin of litigation.” The best outcome for our country is for Joe Biden or Donald Trump to win the election decisively, based on the votes of the people. Any other outcome is likely to have very bad implications. 

Q: What do you view as the most challenging aspect of polling this year? 

A: The Center for Public Opinion has expanded its polling operation in 2020. In 2016, we were a regional pollster doing swing state polling in a single state. In 2020, we will be focusing on several swing states, polling them multiple times, and also doing national polling work. We (Associate Director John Cluverius and I) are very, very busy. 

Q: If the election were held today, who do you think would win the presidency? 

A: If the election were held today, Biden would win. He’s in a better position today than (Hillary) Clinton was vis-à-vis Trump in 2016. Of course, Clinton was a 3-to-1 favorite at this point in 2016. Looking at the polls, I think Biden is closer to a 6-to-1 favorite today. Part of (my thinking) is based on looking at polls and part of this is my looking at Trump. He’s an incumbent running for re-election without ever having a positive approval rating. In 2016, he ran against a candidate who also had a net negative favorability rating. In 2020, he’s running against a candidate with a net positive favorability rating. Trump gets poor marks for how he has handled COVID-19 and how he has responded to protests about police violence in this country. He gets better marks on his handling of the economy, but the second quarter of 2020 saw minus 32 percent growth and the current unemployment rate is 8.4 percent. If he were re-elected under these circumstances, it would be truly remarkable, unprecedented even.