The stakes of the 2024 election are existential for the future of American democracy. On one side is Donald Trump, a man who has demonstrated contempt for the Constitution and the rule of law. He has also repeatedly shown himself to be a racist, a misogynist, and both a likely and convicted criminal who is obsessed with violence. In total, Donald Trump is extremely dangerous to the United States and the world. On the other side is Kamala Harris, a defender of American multiracial democracy and the rule of law. Her candidacy is truly historic. She is the first Black woman to become the presidential nominee for a major political party in the United States. Trump’s vision of America’s present and future is a dystopian hellscape. Harris believes in the American project, our goodness and ability to improve.
In a healthy society, the choice between the two candidates would be clear. Instead, public opinion polls show an election that is basically tied and one of the closest in modern American history. Harris has the momentum and is now ahead nationally. However, Trump and Harris are very close in the key battleground states that will decide the outcome of the Electoral College.
In an attempt to make better sense of the 2024 election and what the public opinion polls currently indicate (or not) about the state of the race for the White House, I recently spoke with public opinion and politics experts Elizabeth Jarosz and Clifford Young from the global leading independent market research firm Ipsos. Jarosz is an Ipsos market research expert. She is a trusted advisor to Fortune 500 companies, known for her ability to decode audience behavior with unparalleled accuracy. Young is Ipsos President of U.S. Public Affairs and leads the global election and political polling risk practice. He has conducted polling for over 100 elections globally, including 25 U.S. midterm races in 2010 and various elections in Nigeria, Canada, Egypt, and Kuwait.
In this conversation, Jarosz and Young explain the fundamentals, limits and potential of public opinion polling. They also reflect on the public’s mistrust and cynicism about public opinion polls following the surprise outcome of the 2016 election.
Notably, Jarosz and Young also share their insights on the phenomenon of so-called shy Trump voters, who are believed to be systematically undercounted in polling. They conclude that the 2024 election is a battle between Harris’ likeability and Trump’s perceived strengths on the economy and immigration — and that Harris has an easier path to victory than Trump.
This interview has been lightly edited for clarity and length. The written responses represent both interviewees:
Public opinion has famously been described as a type of chorus. What is that chorus telling us now with 30 days or so until the election?
America is in a transitory place today. The country is stuck between looking to the future and wanting to go back to the past. This is the primary rift today and it differentiates and defines the two candidates. While the election rests on a razor’s edge, our qualitative interviews reveal that deep down, America is still quietly hoping for someone they can wholeheartedly believe in and get behind: “America’s Quiet Hope: A Leader Worthy Of My Wholehearted Support.” How this plays out of course we will have to see.
What are some of the basic concepts of public opinion research?
Perhaps it is best to start with a definition of the pollster. A pollster measures or analyzes public opinion. Ultimately, we believe that polls are the linchpins of democracy because they allow for immediate feedback to those in power; elections are feedback mechanisms as well but slower in their effect. As we see it, public opinion is the most important societal stakeholder today as it gives legitimacy and license to those in power. Again, the pollster and the poll are the conduit for that dialogue.
The most important thing is transparency. Many different types of methodologies are employed today (mail, telephone, online), all with decent track records.
What is actually going on with the polling and the state of the presidential contest?
It is always best to take the average of the polls instead of looking at any one poll. Applying that approach we see an extremely close presidential race. It looks like Harris has a slight lead in the rustbelt states and Nevada. In the end, we believe it will come down to the relative ability of each campaign to win the framing war. “Throw the Bums Out” because of inflation versus future economic opportunity for all Americans. Harris is more effective with the latter narrative frame.
What do you think the mainstream news media does not correctly understand about public opinion polling and what it can tell us or not?
Polls are only fuzzy forecasts at best; they are a snapshot of a moment in time. Responses to polling are influenced by media effects. As such, any given poll or event can make public opinion seem capricious when it is not in the aggregate.
Therefore, we should be very focused on the issues. In the 2024 election, the major tension is between “personality” and “policy.” Donald Trump wins on policy (2 of the 3 top issues are the economy and immigration where he holds strong leads). But he’s not the clear frontrunner because Harris wins on personality.
One of the recurring themes since at least 2016 is how the polls are so wrong, inaccurate, and should not be believed.
The industry has taken a lot of steps to understand the consistent Trump undercount. We have found this in other countries as well such as France, Brazil, and Turkey among others. Support for right-wing candidates is consistently undercounted by the polls.
Leading pollsters in the United States have employed adjustments to correct for this problem, especially taking into consideration the past voting behavior of the mass public. We think we have a solution, but we will see.
What is the approach that Ipsos uses in its modeling and overall approach to public opinion polling?
We use a probabilistic approach to sampling and modeling. This is the gold standard in the industry. We recruit panelists offline. We also interview them online. At Ipsos we are also focused a lot on context and meaning. We proceed from questions such as what does the election all mean? How does it stack up against past elections? What is the mood and spirit in America today?
How is success measured when evaluating a public opinion poll?
That is a difficult question. From an election standpoint, how accurate is the poll in picking the right winner and doing so within the margin of error? But most importantly, it should be about how well a given poll helps us to better understand public opinion. A horserace poll by itself says little about the state of the nation.
What do we know about Trump and his “silent” and/or “shy” voters? Is it actually true that Donald Trump systematically underperforms in the polls or is that just speculation and conjecture that has solidified into conventional wisdom among the news media and punditry?
We know that politics today is wreaking havoc on our methods. We have seen a rise of anti-establishment sentiment around the world which in turn has led to strong anti-establishment political leaders and brands such as Trump, Le Pen, and Bolsonaro. They in turn attract voters who are hostile to normal politics and civil society. So, the pollster's central challenge is to ensure that such individuals are captured in our polls. Yes, the “shy Trump voter” or the “Trump disconnected voter” keeps us up at night. We think we have solutions as an industry. We find out how correct we are on Election Day in a few weeks.
You use sophisticated tools to evaluate how voters and the mass public respond to a given candidate. Can you share how that works?
The pollster is a multidisciplinary profession that demands an understanding of statistics, political science, sociology, survey methodology, and cognitive psychology among other disciplines. We employ models from social and cognitive psychology to understand how attitudes are formed and the role emotions play.
Very simply attitudes are a function of the importance people place on the issue and how well they think a candidate performs on that issue. Inflation is the most important issue today and Trump is stronger on it than Harris. The question is how can Harris change that perception? She will do this by leaning into emotion and the credibility of her message. To that point, Harris is now emphasizing “price gouging” and “future economic opportunities for all Americans.” That language conjures up strong emotional reactions. The first is populist and directed at big business and the second is a fundamental American value. Harris’ messaging is an attempt to break through already existing attitudes. Will this be successful? We will see.
Technically, polls are the “what;” Qualitative research is the “why.” At Ipsos we use focus groups, in-depth interviews, ethnography, and neuroscience tools to uncover why voters feel the way they do. We give voice to the numbers. The role of psychologists and sociologists is important, as it gives us innovative ways to ask questions and analyze responses (like linking decisions to deep human motivations). We use experts in behavioral science, facial coding, galvanic skin response, and other subjects and research methods as well.
What do we know about the relative likeability of Harris and Trump?
If you look at the United States as a whole, Trump is underwater on favorability, whereas Harris is in the positive. That indicates that on a national level, Harris is more likeable. However, when we interview Trump supporters and loyalists, they are diehard and committed. When we interview Harris supporters, they are not as enthusiastic and dedicated because they are still getting to know her. We live in highly polarized times. Negative partisanship and “the anti-vote” are much more important in our minds than straight-up favorability.
How is the public responding to Tim Walz and JD Vance?
Voters see Walz and Vance through a highly partisan lens. If you’re a strong Democrat who watches liberal and/or progressive media (MSNBC, etc.) you think Vance is the devil. If you’re a strong Republican watching conservative media (Fox News, NewsMax, OAN, etc.) you think that Waltz lies and avoided going to Iraq.
These are unprecedented times here in America. How does that uniqueness challenge data-driven political analysis such as polling?
Pollsters should be humble right now. As a country given all these recent events, and this era more generally, pollsters need to pause, sit back, and reflect on the state of our political culture. With Biden stepping aside as the nominee, the contrast between two older white men has been overwritten by Kamala Harris. The past might just not be prologue this year.
What are you focusing on in the vice presidential debate tonight?
Vice presidential debates — and this is true of presidential debates as well — do not historically determine the winner. We will look at tonight’s debate as a metaphor for who (and which campaign) is more on message.
Looking at the polls and other data, who would you rather be right now Kamala Harris or Donald Trump?
The map looks easier for her. Her campaign is also firing on all cylinders. But to reiterate, the presidential race is so close which means that any number of things can happen in the remaining weeks that will determine the final outcome.
Both sides understand that the game will be won or lost around the issue of inflation. Right now, Harris is more effective at this than Trump. But there still is a month until Election Day.
Qualitatively, we are finding that gender and race may be impacting underlying decision-making more than voters are aware of (and are articulating in polls). There is a great deal of quiet sexism in America. Racism also exists, even if it is subconscious. Both show up when voters (primarily men) say Kamala Harris “is not qualified” or they “can’t recall one thing she did as VP” or that she’s “a DEI hire." When given her credentials, or when I ask in panels and focus groups, "What did Pence do as VP” these respondents may deflect and then pivot back to “but she slept her way to the top.” This is not something often said about men.
In addition, there continues to be an inability to hold complex thoughts on most issues. For example, abortion is not about the number of weeks. Immigration reform is multifaceted and not just about building a wall, or deporting people. Most voters are not truly 100% aligned with the Democratic Party or the Republican Party, and yet we act as if we are all on one side or the other and there is no gray.
