Episode Transcript

CARDIFF GARCIA: Kristen, welcome to The New Bazaar.

KRISTEN: Thank you so much for having me.

CARDIFF: In addition to being a pollster yourself and somebody who analyzes what polls and polling averages find, you have a kind of side hustle in polling epistemology, if that’s the right phrasing. You do a lot on the wisdom of what we can and cannot know based on polling. And you wrote something recently that I thought was really interesting. You wrote, “Keep calm and throw it in the average.”

So I want to start there. What does that bit of advice mean?

KRISTEN: A lot of people are very emotionally tied up in the polling averages as we approach Election Day, and that’s somewhat understandable because this is a very high-stakes election. A lot of people feel they have quite a bit at stake personally in who wins or loses. They are gravely concerned about what happens if the candidate for whom they are not voting ultimately wins. And so I find a lot of people who are the real political junkie types, anytime a new poll comes out showing that their preferred candidate is doing worse than perhaps they expected or hoped, they can immediately go into this tailspin of panic, “Oh no, what’s happening?” –

CARDIFF: Or exuberance. They get super excited if they think their person’s doing great. 

KRISTEN: – really excited. “Oh my goodness, my candidate’s going to win in a landslide. This is wonderful.” People get very emotional about it. And my message is that if you are following each individual poll, and letting that dictate your emotional state during election season, that’s a recipe for disaster.

Polls are going to bounce around. Some polls are going to show your candidate doing well. Some might not. That’s sort of the reality of what we do in the polling business. We are grabbing a sample and that sample can bounce around. It is not actually an exact science. And so even a well conducted poll, you should take it with an enormous grain of salt.

So “stay calm, throw it in the average” just means don’t let any one individual poll weigh too heavily on your emotional state in a particular day.

CARDIFF: One of the things about this election specifically, though, is that it seems like the averages of polls have not actually bounced around that much in recent months. So we’ve had these events where a lot of people have interpreted them as having been very meaningful, but you’ve actually written quite a bit about the steadiness of polling averages, certainly the last month, I think maybe even a little bit longer than that, right?

KRISTEN: Right. So one thing that is unique about this year is that while individual polls do have a wide range of individual results, the averages are barely moving. 

So what I mean by that is in 2008, the election between Barack Obama and John McCain, there was a point in time where the polling averages in the last two months showed McCain up by three, there was a point in time right before the election when they showed Obama up by seven. So it’s a 10-point range that you could have in those averages. 

Obama-Romney, somewhat similar, a little bit narrower range, but you still had a point in time where Romney was up by a point or two. And you had Obama up by four or five. It really isn’t until this year, where when you look at all of the polls conducted post-Labor Day, the average of those polls has ranged from Harris being ahead by two to Harris being ahead by 1.4. The range in which that result has traded has been such a narrow band, and it really is remarkable. It’s why I have been skeptical when people say, “Oh, event X, Y, or Z just happened. Is this going to move the numbers?” I can feel pretty confident saying no because nothing seems to move the numbers.

CARDIFF: [Chuckling] Why is that? You’ve written, for example, that it seems like Donald Trump’s brand in particular is quite fixed, that voters have already made up their minds about it. It’s a little different for Kamala Harris, but that seems to be a defining characteristic of certainly at least the last couple of months of this election cycle.

KRISTEN: The idea that Donald Trump’s brand image, and the fact that he has already been president once before, creates a level of stability in the averages because people know what they’re getting with him. They may like it, they may not like it, but nobody can say they don’t know what they’re going to get, “Gee, what do I think a Trump presidency would look like?” 

And so even though a lot of voters are going, “Gee, I don’t know what a Kamala Harris presidency would look like,” for some of them, they’re willing to take that gamble because they know what Trump would mean and they really don’t like it. Or they couldn’t care less what Kamala Harris is offering because they lean against it anyways and they know Trump has got what they want.

CARDIFF: Yeah. You were actually in Phoenix, Arizona a little while back. You were watching football on a Sunday, super relatable, and you stumbled across another theory of the election based on what you saw on television. Can you just tell that story?

KRISTEN: Well, look, anybody who’s listening who lives in one of the battleground states will relate to this. This is not Arizona specific, but it was that if you live in a place that is not a battleground state, you’re probably seeing a fair number of ads for local campaigns. Maybe Congress. Here in the DC area, you’ll see ads for Hogan and Alsobrooks, the race for Senate in Maryland. Maybe you’ll see some stuff for Virginia. But I’m not really seeing a ton of Harris or Trump ads. 

But in Arizona, it was nonstop because that’s one of the seven big battleground states. And I noticed that the Trump ads were all negative and all about Harris. Most of them centered around transgender issues — that was the main focus — and there’s been plenty of reporting that really has been the dominant theme of their ads across the battleground states here in the closing weeks. But the Harris campaign, their ads were a mix: some of them were positive about Harris, some of them were a contrast. Very few of Harris’s ads, though, were just straight-up negative about Trump. 

And so it seemed to me that both campaigns had discerned that the election was going to hinge on: What had voters ultimately concluded about Kamala Harris? 

Trump’s campaign wants them to conclude that she is negative: “She shouldn’t be president.” The Harris campaign wants people to conclude, “She would be great as president; what are you talking about?” 

But the ads were not about litigating, “Is Donald Trump a nice guy?” — that debate is done. The debate now is, “What kind of a president would Kamala Harris be?” And you can see that reflected in the ads on television.

CARDIFF: So the Trump campaign is trying to define Kamala Harris, and the Kamala Harris campaign is trying to define Kamala Harris.

KRISTEN: Well, I now have this weird track record for two weeks in a row on the Chris Wallace show. (I’m a contributor at CNN and every week on that show, where we’re obviously debating what’s going on in 2024.) And I did make two suggestions to the campaigns. 

Two weeks ago, I suggested that the Harris campaign would be wise, if they want to do an ad about Trump and they want to do anything negative about him, the most effective thing to do would be to show clips of him from 2015 and 2016: playing the hits, talking about the issues he likes to talk about — but then contrast those with the way he talks about those issues now, and just letting Trump speak for himself, rather than trying to hit viewers over the head with a message such as it says, “Gee, Donald Trump’s too old and unhinged.” 

Just let the difference speak for itself because it is a little jarring if you watch old clips compared to now, he does seem like a different guy. And sure enough, 48 hours later, they had gone up, at least online, with a before and after sort of ad highlighting those kinds of differences. 

But then for the Trump campaign, what I had recommended last week — as I said for all that this has been so negative, and it’s really taken a dark turn — they would be wise to end on an ad that does at least have some positive stuff about Trump. They would be smart if they went out and said, “Hey, don’t you remember when prices felt low? Don’t you remember when it felt like the streets were safe and the world was safe and the Middle East wasn’t on fire and Russia hadn’t invaded Ukraine and gosh, if you put me back in the White House, we can make that happen again.” And I thought, “That’d be an interesting ad.” 

And sure enough, over the weekend, they announced, hey, here’s an ad that we’re [doing] — I don’t know how much they’ve put it on TV, if at all, but of course, the challenge with that is, do you have a message that is then carried by the candidate? 

And this was paired up against this Madison Square Garden rally, which was not as much about those topics… we’ll leave it at that, but that’s a diplomatic way of saying it’s hard to drive a strategy when the candidate and the campaign seemed to be on slightly different pages about what the message should be.

CARDIFF: You also wrote that one of the strategies for Kamala Harris should be to try to convince voters that she is more moderate than maybe they had originally perceived her to be when she first became the candidate. How do you think she’s doing at that?

KRISTEN: I think it’s been a little bit challenging because while she initially really rallied Democrats to her side, she unified the Democratic Party unbelievably quickly, she still is struggling in terms of brand image with voters who are in the center, who do view her as a continuation of the Biden Administration.

She hasn’t really done the job of breaking with him in a way that is believable to a lot of these voters. I did a focus group for The New York Times of Michigan voters, and many of them were Democrats or Democratic-leaning Independents. They were open to voting for her. They definitely did not like Donald Trump very much, and they were not excited about the prospect of voting for him, but they felt like she had been… “deceptive” was a word that was used by a couple respondents, that they just found this conversion of hers from past progressive candidate for president Kamala Harris to today, to not necessarily be believable because it wasn’t explainable beyond pure political calculation. There was no answer to, “Okay, if you used to believe fracking was bad, now you believe it should be good, tell me why beyond ‘the world is different now’. You have to give me a little more than that.” 

That seems to be what a lot of these voters were saying. They didn’t want to vote for Trump. They just needed some assurance that Harris could be a minimum viable president and were on the fence about whether she has gotten there or not.

CARDIFF: Yeah. You were discussing this poll that found that something like seven percent of Donald Trump voters are what you label “begrudging Donald Trump voters”. That yeah, they’re probably going to vote for Trump, but they could be potentially swayed. And I guess my question was: Is the message that would make them most swayable something that delivered on policy, that was about policy specifics, rather than some of the political messaging that we’ve been hearing in the last few months?

KRISTEN: I think if she was able to make the policy pivot credible, it could have worked to peel some of that seven percent off, but I think right now that 7 percent is simply saying, “We think that she would be too far left. We don’t think she’d be good on policy. And so while I don’t think Donald Trump is a good guy, I do agree with him on policy.”

It’s why I find this pivot here at the end that the Harris campaign has taken to focus on democracy and “Is Donald Trump a fascist?”, those sorts of things, I think it’s an interesting gamble because it’s clearly targeted at that seven percent. It’s clearly targeted at people who are on the fence about considering voting for Trump because they like him on policy and it’s trying to make the case: “Look, no bargain is worth it, right? You think he’s going to be good on taxes are good on this, that, or the other, but he’s going to use his time to exact revenge on his enemies, like you don’t want any part of this, don’t put your name on this, leave the ballot blank at a minimum”. [That] seems to be what that message is more about rather than saying, “Hi, I, as Kamala Harris, am going to be a good president.”

It’s really hoping to peel those wobbly Republicans who have quote-unquote, “come home” to keep them from going home.

CARDIFF: I came across this line in an article in the New York Times yesterday and the quote is: “Never in modern presidential campaigns have so many states been so tight this close to Election Day.” And then it goes on to make the point that if you look at polling averages and the fact that those seven swing states — you mentioned Arizona, it’s also I think, Michigan, Pennsylvania, North Carolina, and three others — if in any one of those states on Election Day, the outperformance of either candidate is half a percentage point higher than what the polling averages are showing now, that’s a victory. That’s the difference between victory and defeat right now. And one of the things that you’ve been hammering home to people is that you can ask me as many times as you want who’s going to win.

But anybody who tells you with any amount of conviction right now is simply selling you something. Is that about right?

KRISTEN: Yes, I don’t think anybody should feel confident. I think the only people who are allowed to project confidence are people who have some data that is not available publicly that they really believe they are confident in. So, for instance, let’s say you’re modeling what every voter in the state of Arizona is probably going to do in terms of their vote, and then you’re watching to see, “Okay, what do we know about which votes are being returned early?”

We think that this is good for us but even that is a little bit of reading the tea leaves at this point, because we don’t yet know, what does it mean if early voting is up or down? Everything from the past has an asterisk next to it because of COVID, and so it’s really hard to know what’s what compared to the past.

The other thing that I would say about the polls is you can have two polls where the top line number tells you the same thing, but underneath the hood, they’re telling you very different things about what this election might be. 

So if the polls are off in North Carolina by a little bit, and Donald Trump wins North Carolina pretty handily by four points, that doesn’t necessarily mean that the battleground polls will all be off in the same way because the demographic makeup of North Carolina is not the same as the demographic makeup of Wisconsin. And so, knowing that, I think it’s not necessarily the case that every swing state is going to correlate purely. 

What I mean by that is there are some polls where, under the hood, their data shows a political realignment is happening. The New York Times Sienna poll is a perfect example of this. They showed the race, and I think it was 48-48. I remember laughing when I saw the results. I was like, “Oh great. This doesn’t give us–”

CARDIFF: [Chuckle]

KRISTEN: “ — anything to work with, 48-48.” If you looked under the hood, you would see things like Trump doing really well with young voters, Trump doing really well with Latino voters. But, at the same time, Harris doing amazingly with women, huge gender gap, and Harris knocking it out of the park with white voters who have a four-year degree, which is a group that used to be a little more open to Republicans, but has since trended a bit away.

So that is a real political realignment, if all of a sudden, divides by race and generation are shrinking, but divides by gender and education are growing. That’s really interesting. But there’s other polls, like CBS’s poll, that also showed the race effectively tied. But under the hood of their poll, they actually didn’t show a huge gender gap, but they did show an absolutely enormous generation gap. They have young voters voting for Harris at nearly Obama-Romney margins and they had older voters voting for Trump at numbers he hasn’t seen with seniors in a while. It had white versus voters of color; big differences there. It had Trump doing really well among white voters with college degrees. That poll suggested this realignment is not happening, [that] the divides that we’re used to are the divides that are coming back, so I call it the reboot instead of the realignment. 

We don’t know which one of those polls is right yet, so even though the top-line number is the same on both, the story underneath is different from poll to poll as well.

CARDIFF: Has polling gotten weirder, or at least methodologically trickier in this election cycle?

KRISTEN: [Chuckling] Yes, yes, extremely. I like that you used “weirder” instead of “worse” because I think the perception out there is that there was some Golden era of polling in the ’90s when it was crazy accurate, and then it’s all just been downhill from here, and that’s not really true. 

If you look back at the history of polls, polls are actually historically more accurate today than they were in the past. We just followed them less obsessively in the past, and people were less tied up in them for their emotional stability and sanity. 

CARDIFF: Yeah. 

KRISTEN: Nowadays there are more polls, and we look at them much more in media than we used to. But it has gotten more challenging to do polling, and that’s because every time there’s a technological advance that makes it easier for me as a pollster to find you, a similar advance makes it easier for you as a respondent to avoid me.

So everybody gets on cell phones, well that’s great, now I can reach people even when they’re not at home, I don’t have to worry about only contacting people at home during dinner time. Except with cell phones, you now have caller ID. You’re not going to pick up a number you don’t know. Back in the dark ages, you were paying for minutes, heaven forbid. (Gen Z-ers listening to this podcast won’t know what that means.) But nowadays, we have to use a variety of different methods. We do a lot of polls that use text messaging to reach some people. We do a lot of polls that use online panels, and these are good panels. 

They’re not just used for politics. They’re used by corporate market researchers around the world to gauge how people feel about brands and products and cultural trends, and you name it. We, just as political pollsters, piggyback on that, but it’s still not perfect, and especially now with the advent of AI and bots, there’s a numerous and growing number of challenges pollsters have to face because every time technology makes it easier for us to do our job, it’s “one step forward, two steps back.”

CARDIFF: It’s interesting what you also say, though, about having followed polls less obsessively in the past, because it seems like that’s of a piece with a larger cultural story of essentially the Gamification or novelization or serialization of American politics, where people are now following it, not just to find out who’s going to be president, what that might mean for policies, but almost as a part of their identity, as a way of following it even as entertainment for a lot of people, I think, and it’s interesting that that also applies to polling.

KRISTEN: Well, think about the way we got news 20 years ago. For the most part, if you were somebody who was really interested in the news, maybe you watched the news in the morning or you read the newspaper while you were eating breakfast and then you got up and you went about your day. And then maybe in the evening you came home and you listened to drive time news on the radio. You turned on the evening news at home. You were maybe dipping into news once or twice a day. 

Where now we carry around news-purveying devices with us all the time and thankfully, your average American is not obsessively reading a Twitter feed and doom scrolling, but you couldn’t even do something like that if you wanted to back in the day. And now, if you do want to be an obsessive, you can — and that’s why this whole market has sprung up where you now have all of these aggregators, all of these forecasters, you have betting markets.

If you look at polls that are historic over a longer time horizon, I believe there was a question that had been asked, I don’t know if it was Pew or Gallup, so I don’t have a good citation for this, but it was something like, “Does whoever wins the presidential election, does that make a big difference in your life?”

And that back in the ’80s, it was less than a third of people said “yes” to that question, and nowadays it’s a very high percentage of Americans who say “yes” to that question.

And there’s a dramatically increased demand for what we do, but at the same time, the demand is almost a little unhealthy. People will get really mad at pollsters and they’ll say, “You guys are just churning out a bunch of data to try to sway the numbers and sway what people think about the election.” 

Which isn’t true. There’s nothing pollsters want to be more than right. We want to be accurate first and foremost. Just think about what our economic incentives are. Nobody wants to hire a pollster that’s 15 points off every time. Everybody wants to be the most accurate pollster, but I think people also don’t understand what polling is used for.

Political pollsters are really the only people in the entire opinion and market research industry that get truly held accountable for the accuracy of their numbers. If I am doing consumer market research for a soda company or a sneaker company, and I go out and I do a study, and I find that 80 percent of people love this soda or love this sneaker, but the real answer is actually a 70 percent, it’s not 80 percent, I’ve overestimated it… It might not be obvious to that company that I, as a pollster, have overestimated this. 

Where in an election, if I say you’re going to win by two and you win by 12, there’s huge backlash that comes my way. 

And so that’s something that I think is interesting about what we do. But people also expect a level of precision that polling is not intended to provide. We are intended to give advice that comes from data that gives us sort of a — We can get somewhat granular, but I’ve used the analogy of, if you’ve ever done baking, I’m not good at baking, I’m not, I don’t follow the instructions close enough, and in baking, you really have to follow things precisely. But I’ve heard that when you are doing any recipe in your baking, you should weigh your ingredients. That’s going to be the most precise. Now, you can have your bathroom scale that you step on and will tell you how many pounds you are, but that’s not going to be precise enough to tell you, “What little amount of baking powder should I put in this recipe?” For that, you need a kitchen scale. Something that is built to measure things at a very fine level. 

Polls are sort of your bathroom scale in this analogy, right? They are equipped to tell you within three to four points one out of — 19 out of 20 times how something might look. But if my poll shows you winning by two and you win by three, that does not mean the poll was off by one and I think it’s that hunger for polling mixed with that misunderstanding of how imprecise polls actually are that leads people to get spun up about our industry and and then frustrated and then they tweet angrily at people like me. [Chuckles]

CARDIFF: It actually reminds me a lot about the way we track economic indicators, with people not realizing how imprecise it is, and also that there’s a very good reason that a lot of economic indicators are updated and revised up or down over time. We’re talking about a massive population of people that we’re trying to follow in real time.

And it’s not so simple, because you have, for example, people who try to forecast whether or not employment went up by 120,000 last month or 140,000. Or was GDP in a given quarter up 2.3 percent annualized as opposed to 2.1 percent annualized. That is completely unknowable in real-time. But you have a massive industry of people, especially on Wall Street, that are really trying to peg it at that level of precision. And we don’t quite realize that that’s not really possible. 

So anyways, I find a lot of the overlapping similarities between polling and tracking the economy to be quite fascinating. 

And you mentioned betting markets. And I’m curious to know, have you followed the controversies over betting markets? I’m curious to know what the relationship is between betting markets and traditional polling, how that’s changed, especially this year.

KRISTEN: So I think betting markets are useful if there is valuable information out there that is not being captured by the other main indicators. So if there’s something going on out there that you think a poll is not accurately capturing, then there’s value in a betting market. 

The challenge about a betting market is that it, at this point, seems relatively able to be influenced by somebody coming in and placing a ton of bets on Donald Trump, and all of a sudden, that swings the market wildly. So ultimately, I do think if you are just looking at the polls, and then taking what you read in the polls and going to the betting markets, I still need to see what other information, what wisdom of the crowds are we actually gaining here from these betting markets, versus this is entertainment.

It’s like people betting on a football game. I don’t actually know whether—well, I think the Pittsburgh Steelers are on a bye this week—but I don’t actually know, are the Pittsburgh Steelers going to win their next game or not? I can make a guess. I can know what Vegas sets the line for, but ultimately, that’s just entertainment. And so that’s how I view the betting markets for now.

CARDIFF: Yeah, it’s interesting because you’ve just drawn a parallel to market inefficiencies. Are betting markets capturing an inefficiency that is not being captured by the polls? And can you take advantage of that inefficiency then, by betting in a betting market? 

But those markets are new, underdeveloped—there’s liquidity issues with some of them. I think you’re not allowed to actually use them inside the United States. So there’s all kinds of oddities related to them. 

I’m not saying that they don’t hold promise for the future. I’m just kind of curious to know, to what extent you see their relationship with polling, but you’ve just answered that question. So that was really interesting to me. 

Every single election also inevitably becomes a referendum on a quote that James Carville gave in 1992, which is: “It’s the economy, stupid.” I’m just curious to know how you see the role of the economy right now in terms of how the candidates are performing, and whether or not people are weighting it more or less than in prior elections.

KRISTEN: So on the one hand, when you ask people: “What is the number one issue in your mind in this election?” It’s the economy. So I’ll give you an example from that New York Times focus group of Michigan voters that I did that we just posted this week. We asked people to tell us, of the various issues, which ones are going to play a major role in deciding for whom they vote. And, the economy—of our 11 voters, seven of them said, “Yes, the economy is going to play a major role.” Which outpaced immigration—we had three people who said it would be a big deal. Democracy—four people said it would be a big deal. Abortion—four people said it would be a big deal. The economy was the big one.

But before we asked about issues, I asked those respondents, “What is this election about?” And none of them mentioned the economy in response to that question. They said it’s about power. It’s about truth. It’s about values and freedoms. It’s about change. It’s about personal responsibility. Nobody specifically said it’s about making milk cheaper. 

Eventually we got to the conversation about: Has cost of living affected you? What would you like to see in terms of the economy? People’s concerns about the American Dream. That was all there. 

But their initial answers to: “What is this election about?”, were about something bigger… about the character and soul of this country. I’ve described it instead as: People feel like the country has lost its mind, and this election is the opportunity for America to try to get its head on straight. 

And that applies to voters on both sides, who think only their candidate is going to help America get its head back on straight. That if America goes the other direction, we’ve lost our mind and it’s irretrievable. But that’s just so much bigger than, “I’m worried about the economy.”

CARDIFF: One last question about the election. You have been suspicious of the idea of an October surprise. And you and I are now talking on October 29th, so we’re running out of time anyways. People will be hearing this podcast on November 2nd, so who knows, maybe in the time between when we talk and when it comes out, something will happen. But you’ve been saying that essentially, that’s not likely in this election. 

And the campaigns themselves are still trying things. But we have not had something big that has actually shifted things meaningfully over the course of roughly the past month.

Can you just say more about why you’re skeptical that anything this late can happen that would actually move things from where they are now?

KRISTEN: Well look, on the one hand, it’s easy to say this isn’t going to make a difference because the polls are so stable and views are so locked in. But the counter argument is that, when things are so close, any little tiny thing could technically matter. 

So let’s take for instance, the comments made by an offensive comedian at the Trump rally at Madison Square Garden—for which the Trump campaign has already tried to distance itself and apologize—where he made comments about Puerto Rico. 

Now this matters for a lot of reasons, but first and foremost, the fact that it is off putting to a group that actually Republicans have been trying to win over and had been doing better with: Latino voters. And specifically, Puerto Rican voters. 

I grew up in central Florida, right near Osceola County, which is a hub for where lots and lots and lots of folks have immigrated from Puerto Rico. All they need to do is establish residency and they can vote. And so you immediately saw Rick Scott come out and say, the Senator from Florida, “I distance myself from these comments. This is not how I feel. I love the Puerto Rican community. Don’t hold this against me.” Because that actually is a big deal, and it may well be a story that is breaking through in a way that most offensive things that happen at rallies or whatever don’t break through.

And it doesn’t have to move hundreds of thousands of voters. If it moves 2000 voters in a key swing state, that could be enough. Again, I came from Florida. In the year 2000, what was it, 500 and some votes decided the whole thing —

CARDIFF: I’m from Tampa. I remember that quite well. I remember that quite well.

KRISTEN: I heard it described: the reason why Florida counts ballots so efficiently now is, it’s like that insurance commercial, we know a thing or two because we’ve seen a thing or two.

CARDIFF: [Chuckling] We’ve been there.

KRISTEN: Florida’s like, “We’re never going to do that again. We’ve got it figured out now.” 

So, I am skeptical of the existence of an October surprise because this has been an election with an unbelievable number of October surprises happening constantly. And yet, the polls have stayed so stable.

The aliens are gonna have to make contact for us to really shake things up. And even then, we might still have a 48-48 race.

CARDIFF: For anything to happen, you need literal extraterrestrial intervention. 

KRISTEN: Hey, we still have time. There’s still, what, two days left in October?

CARDIFF: Right. [Chuckling] Let’s turn now to the survey that Echelon Insights put into the field on behalf of the Economic Innovation Group. The quick background here is that this is part of the American Worker Project that we’ve been doing at EIG. All of the links to the findings in a great slide deck that you put together will be in the show notes.

But essentially, what this survey did was it looked at American workers—so people who are employed or actively seeking work, basically the labor force—rather than a survey of the wider American electorate of all voters or likely voters or registered voters or whatever. 

So let me start with this: What surprised you most about the survey? Which of the findings did you find the most interesting?

KRISTEN: I thought that what was surprising was—this will sound silly—what was surprising was what wasn’t surprising. So there’s been so much talk lately about two phenomena. 

One is that workers are increasingly dissatisfied. They’re fed up. People hate working. And we did not find that in our poll. We found that people generally think if you work hard, you can succeed. We found that people are generally feel satisfied with their overall career. But then we also said, what are the things that you wish were better? And the number one response was, “I want more money. Show me the money. That’s what I need more of.”

And so, in some sense, that’s also not terribly surprising. When you ask people how satisfied you are, how hard you work at your job, your job security, how satisfied are you with the amount of personal time you have outside of work? Those numbers could always be higher, but on average, this was like two-thirds of workers saying that, yes, they were satisfied with those things.

The only thing where you had less than this two-thirds level of satisfaction was the amount of money you are making at this stage of your career. And even there, for college-educated workers, two-thirds of them said that they felt at least somewhat satisfied with how much money they were making at this stage of their career. It was non-college graduate respondents where it was only 50 percent. So there was a bit of a gap there. 

But yeah, show me the money. I mean, we also said, “If you could make one change to your current work situation, what would it be?” And we gave people a huge number of things they could choose from.

We hear a lot of talk about, “I need more childcare. I need more flexibility. I want to be able to work from home more often.” But one of the other options on that list blew everything else away. And that was higher salary. That’s it.

CARDIFF: Yeah, I think all the other ones combined didn’t even come close to a higher salary.

KRISTEN: Right, the chart is kind of comical to look at.

And again, it’s not that those things aren’t important. I’m sure people would love more flexibility. If you’re a parent, you would like childcare. But, if you are an employer, salary seems to be the thing that workers are saying, “Don’t try to give me less in terms of paycheck and try to do this other stuff. Salary is what we’re looking for.”

CARDIFF: Yeah, what I liked about the survey was that it added an interesting nuance to this discussion that a lot of people have been having for the last couple of years, which is, the economy is pretty good, very low unemployment, it certainly recovered a lot faster than the economies of other advanced countries in the aftermath of COVID.

And so, the question was, why are people so dissatisfied? Why are they so nervous about the economy given that it actually has performed a lot better than, during COVID certainly, you might have expected it to? 

And it seems like our survey captured both sides. On the positive side, people do like having jobs, obviously, but maybe less obvious is that they like the jobs they do have.

On the other side, though, they are still nervous that the gains they’ve made in recent years are fragile. That they could be taken away. And that also actually does line up with some of the economic data.

If we look at real wage growth, it has been positive, but not at some super high rate. And that is our best indicator of living standards and how they change over time. 

So it’s just interesting to me that, yes, people are satisfied with their work. They still believe in the American Dream. On the other hand, they’re worried about the factors that affect real wage growth: salary, and they’re still worried about prices. That inflationary episode from the last few years has not left their minds just because inflation, recently, has been better.

KRISTEN: Yeah, and we found this in this survey of workers. I also found this in a recent survey that I did for AARP, which was all about women over the age of 50 and what their concerns were. And a key thing that we were finding is that, yes, cost of living is an issue and people do not necessarily feel like their wages are keeping up with the increases they’re seeing elsewhere. But it’s this thing like retiring comfortably, or being able to save, or being able to just feel secure. That’s where we’re seeing some of this gap. 

So in our survey, when we asked workers about things like, “Are you satisfied with your job?” And then, “Compared to what it was pre-pandemic, are you more or less satisfied?” People are generally more satisfied with their job now than compared to pre-pandemic. Or work-life balance.

But then on something like your prospects to retire comfortably, that’s where things seem to be the worse relative to pre-pandemic. We see this with whether you’re young or old. Whether you are under 50, over 50. Our respondents up and down the age spectrum were worried about their ability to retire.

That’s just one other big thing that sticks out to me when people say, “Well, gosh, the economy’s so good. Why don’t people feel it?” It may not be that people can no longer afford to buy milk and eggs. But they’re really worried that they’re going to have to work three or four more years than they thought they would, because they’re not sure when they’re ever going to be able to get around to having enough extra money to sock away.

So that to me is what’s driving this, “Don’t give me all this other… Just show me the money.” That is the theme.

CARDIFF: Distrust of national leaders—this was another interesting finding. It won’t surprise anybody that a majority of Trump voters said that they could have a better life if we had different national leaders. [But] I was myself pretty surprised that 44 percent, a plurality, of Harris voters also said the same thing. It seems like an implicit rebuke to somebody who is currently a part of the administration at the moment. So were you taken aback by this finding?

KRISTEN: Not necessarily, because people think of their opponents as having more power than their own side does. This could be a whole other 90-minute podcast we do. But I think one of the things that’s driving the polarization in our politics today is, if I talk to Republicans and I say, “Do you feel in charge?”

They will say, “No, are you kidding me? Democrats are much more unified. Their establishment falls in line. They’ve got operators like Nancy Pelosi, and they control all these institutions, and we, Republicans, we are the underdog here fighting for our survival.” But then I talk to Democrats, and they’re like, “Are you kidding me?”

What’s that joke? If I wanted to be part of an organized political party, I wouldn’t be a Democrat. Or it’s, I’ve just botched the joke, but it’s—

CARDIFF: [Laughing] No, that works.

KRISTEN: They’re like, “Republicans, you guys are the ones that rigged the system in your favor. And you have gerrymandering, you have the filibuster, you have dark money from—” And every side in our politics thinks the other side is actually the ones that are powerful and in charge. And I think that’s part of what you’re seeing in that result.

CARDIFF: We asked people whether they regarded the following things as either threats or opportunities: technology, trade, and immigration. Anything about those findings surprise you as well?

KRISTEN: Yeah, I think there’s been a lot of talk about things like AI and automation and to what extent that’s going to be a real threat to people’s jobs. At the moment, it does not seem like that is really punching through. People are much more of the mind that new technology is going to be an opportunity rather than a threat.

And then in terms of something like goods and services coming from foreign countries. We wanted to check in on how people are feeling on trade. So Republicans are no longer the pure free-trade party. Trump voters were really split on whether goods and services coming from overseas create more of an opportunity for their career or threat to their career. Where, for Harris voters, they think it’s generally a good thing.

But you also had a split about legal immigration. So you had, Trump voters who were pretty divided, but actually leaning slightly positive towards saying legal immigrants probably provide more opportunity than are a threat. But for illegal immigration, it really swings the other way. 

And so I think it is interesting. A lot of people lump these topics together. But we see in the data, when you ask people questions about legal immigrants and illegal immigrants, while there is definitely a consensus, especially among Republicans, that illegal immigration poses a threat, that does not mean that they think legal immigrants are a threat. It’s a slice of the Trump electorate, for sure, but it’s not a majority.

CARDIFF: I will say, I was pretty surprised by the overwhelming widespread agreement that new technologies are a good thing.

Maybe a lot of [workers] have experience with working with new technologies to actually help them out. Maybe they think [automation taking jobs is] a problem for other parts of the economy, and that it’s not coming for them. Or maybe, it’s just something that, because it hasn’t affected them directly just yet, they’re extrapolating from current trends, and they don’t think it will ever happen to them. I don’t know. 

But I was pleasantly surprised by that result. If you think technologies over time getting better are going to be a good thing for workers, then you would be pleasantly surprised. But I was surprised.

KRISTEN: Yeah, and I think it’s also valuable to note, even if we disaggregate these numbers by: do you have a college degree? And for whom are you voting? We still find even—so Trump voters are a little more skeptical of new technology than Harris voters, and voters without a four-year degree were skeptical than voters with a four-year degree. But even if you look at Trump 2024 workers who do not have a four-year degree, even they, by a 35 point margin, view new technology as more of an opportunity than a threat. 

So even the group that is the most primed—and frankly, a group that views goods and services from foreign countries as more of a threat, they view even legal immigrants as slightly more of a threat, they viewed everything else we asked about as a threat—but new technology, no, they were pretty significantly all in.

CARDIFF: Did anything about the gender splits that were explored in this survey surprise you?

KRISTEN: In this survey, one of the questions we asked was about progress in the workforce. I think this was a question we asked around the pandemic. We said, “Have things improved for you, pre- versus post-COVID?”

And we found that women were less likely to say that things in their career had improved since before COVID. Which would track with some of my assumptions about women taking more of a hit from having to shoulder the burden of childcare and so on and so forth when all of a sudden the world shut down. Or women making pivots in their careers as a result of this pandemic, perhaps more so than men.

But we also asked, having nothing to do with the pandemic, we said, “Just when it comes to things like workplace safety, benefits, hours, standards of living, do you generally believe that conditions are getting better for Americans over time from one generation to the next?”

And men were much more likely to say yes. Where women were actually slightly more likely to say no, and we don’t know why this is. It’s not a question that then gives an open-ended “well, tell us why you feel that way?” 

But if I had to guess, I would suspect it’s some combination of feeling concerned around… I would bet the Dobbs decision, in the issue of abortion, has something to do with it. We hear that a lot in focus groups of women saying they’re worried that the future is going to look worse than will look better. In some ways, purely because of that issue.

But you’ve also got, I think, men feeling like things in the workplace are on the right track, it’s getting better for everybody, where women may still be feeling certain pressures, challenges, feeling torn between work and family, and just having to balance things in a way that maybe male workers on average are not having to deal with as much. So that’s my hypothesis for why we see women much more pessimistic than men about how things are going over time for workers in terms of workplace safety benefits, hours, standards of living, and so on and so forth.

CARDIFF: Last question, that encompasses everything we just talked about, both polling methodology and also the American worker. What are you most going to be looking for after the election? In other words, what new potential information are we going to get from the results of the election? What do you want to keep studying? Basically, what’s next?

KRISTEN: The topic that I’m the most interested in studying is institutional trust. Why it has been shredded, and can it be rebuilt. We consistently see that people—and it’s not just Republicans, it is across the board—people saying they’re less likely to trust media. Democrats are starting to catch Republicans on not trusting media, at the same time that Republicans are catching up with Democrats on not trusting business.

I have clients of all types that are panicked because they’re saying people don’t trust us, and we’re trying to do a good thing, but people always think that we’ve got the wrong motive. How do we fix that? And I think, regardless of who wins, but maybe even more so if Trump wins, this is going to be an unbelievably hot topic.

What do you do to rebuild trust? Because if we live in a society where nobody trusts anything or any institutions—I think that some skepticism is good. I don’t think it’s bad to be skeptical of business or bad to be skeptical of media—but there comes a point where there’s too much distrust for society to function.

And so, studying that question is something I’m going to be focused on for the next couple of months.

CARDIFF: Fascinating. Well, we’ll be following you. Kristen Soltis Anderson, thanks so much for this chat. This was great.

KRISTEN: Thanks for having me.