Plain English with Derek Thompson - The Media is Missing Something Big in Biden’s Bad Polling Numbers

Episode Date: November 17, 2023

Today’s episode is about the question of the moment in politics: the meaning of Joe Biden’s terrible polling numbers. Today’s guest is Nate Cohn, chief political analyst at the New York Times, where he does public opinion, polling, demographics, and politics. We talk about the notorious New York Times poll that showed Donald Trump trouncing Biden in the swing states. We talk about why it’s not crazy to take presidential polls seriously right now—even though we're 300 days out from the election. We talk about the multi-layered problems of Biden’s age. And we talk about why the new "engagement gap" might be the most important, under-discussed phenomenon in American politics today. If you have questions, observations, or ideas for future episodes, email us at [e-mail].  Host: Derek Thompson Guest: Nate Cohn Producer: Devon Manze Learn more about your ad choices. Visit podcastchoices.com/adchoices

What would you do if everyone said they heard your trailer a hundred times? You'd probably make a new one. I'm Justin Sales, the host of The Wedding Scammer, the ringer's first ever true crime pod. We've been hunting a con man for a few weeks now, and our hunt is coming to an end. Schemes, Heartbreak, How to Put On a Wire. We've covered all this and more, but there are still a few surprises left. Binge the Wedding Scammer wherever you get your podcasts. Today's episode is about the question of the moment in politics, the meaning of Joe Biden's

terrible polling numbers. Two weeks ago, the New York Times published a poll that sent to chill down the spine of the entire Democratic Party. It showed Joe Biden behind in five out of six major swing states in the 2024 election, behind specifically Donald Trump. But then, less than 100 hours later, Democrats won several off-year elections, including in Kentucky, Ohio, and Virginia. This led the media to wonder,

whether all this hand-wringing about Joe Biden is a bunch of fretting about whole world that has nothing to do with the events of our actual corporeal physical inhabited universe and the actual physical elections that take place in it. Summing up this perspective, MSNBC host Chris Hayes said, quote, the political experience of the Biden era for Democrats is extended periods of intense anxiety about terrible polling occasionally punctuated by strangely positive election results. and then the cycle repeats. End quote.

Now, I think that's more or less right. And there are some tantalizingly obvious explanations of this dynamic. Like here is a simple theory that will appeal to people who like simple theories. Here we go. Republicans are crazy and Joe Biden is old. That's it. That's the theory. Republicans are crazy to anti-abortion, to anti-democracy, to pro-Trump.

And Joe Biden is old. No elaboration necessary on the latter part. I think you can get a decent amount of mileage on this parsimonious theory. And there's always a benefit to having a grand theory of everything that really can fit on a bumper sticker. But I'm not sure that the Republicans crazy Biden old theory gives us the full picture here. So, for example, does Republicans crazy Biden old really explain why young people are so particularly down on Biden? in this moment, and have been down on Biden, not just during the Israel-Palestine debate, but for the last year.

Does the Republicans crazy Biden-old theory explain why Democrats are losing not just the white working class, but are actually bleeding support from non-white working-class voters, even as the media keeps calling Donald Trump an unapologetic racist? I don't think it does. Does the Republicans crazy, Biden old theory, explain why Republicans who have historically dominated in off-year elections like 2023 now seem to lose so much in these off-year elections as they did just a few weeks ago? I don't think it does. In the big picture, I think the media is actually missing something quite subtle but quite important about the shifting American electorate. something you'll hear referenced in this podcast is the engagement gap.

And I think this engagement gap, this growing engagement gap, explains a lot of the mystery of why is Biden polling so badly, even as Democrats, keep winning elections. Today's guest is Nate Cohn. Nate is the chief political analyst at the New York Times, where he does public opinion and polling and demographics and politics and that terrible, terrible needle. We talk about the notorious New York Times poll that showed such awful news for Biden. We talk about why this year in particular it is not crazy to take polls

seriously 300 days before an election, although I think in just about any other election of the last few decades, it probably would be quite crazy. We talk about the multi-layered problem of Joe Biden's age. And we talk about why this engagement gap might be the most important, under-discussed phenomenon in politics today. I'm Derek Thompson. This is plain English. Nate Cohn, welcome to the show. Thanks for having me, Derek.

Before we talk about your now famous, maybe even infamous poll of Biden and Trump in the 24 election, I have some questions about polling more broadly. In 2016, famously, the polls were very wrong. In 2020, less famously, the polls were also quite wrong. What are the biggest challenges of modern presidential polling these days? And how do you, in the New York Times in particular, try to account for those specific challenges? It's a great question.

I think the single biggest challenge is that the response rates to telephone surveys is extraordinarily low now. Only about 1% of attempted telephone calls yield to completed interview nowadays. And that has greatly increased the cost of political surveys. also creates the possibility of considerable biases. The people who don't respond to surveys are different from those who do. Some of those biases are pretty well known to take one that was responsible for the error in 2016. The people who respond to political surveys are much more likely to hold a four-year college degree than the people who do not take a poll.

There are also unknown biases. Even among the people who don't have a college degree who do take a political survey, It's possible that those people without a college degree are better engaged, that they are more trusting of their neighbors, that they like speaking to humans more and are less likely to have a no trespassing sign or to work in a job where they speak to people day in and day out. These are all possible ways that the people who take these polls are different from those who do. When it comes to the way that we of the New York Times try and account for these problems, we really have two different approaches. One is that we can try and sample the groups who we believe are less likely to take political surveys. One of the most important things in that respect is that we spend a lot of money to try and reach people who don't regularly participate in elections. People who participate in both primaries and midterm elections are twice as likely to take a political poll than the people who don't regularly participate in primaries and midterms.

And in practice, that means that we have to call twice as many people who don't participate in primaries and midterms to receive a proportionate number of them. The second thing that we do is we undertake a kind of statistical adjustment that's known as weighting. It's a little arcane, but in practice it just means that we give more weight to people from underrepresented groups so that the overall sample, as we report it, represents those groups in proportion of their share of the population. If we take a poll and there are twice as many people at a college degree as should as there should be, then in our estimates they receive half as much weight. Let's say someone hears this answer, this very considered and fair answer. And they're like, look, Nate, very detailed response. You seem like a nice guy, but I'm sorry. I don't believe polls.

I was alive in 2016. I was alive in 2020. I don't believe the polls anymore, period. How do you, Nate Cohn, tell them that they should trust polls. How do you know that you're successfully accounting for these biases? How do you know if you're doing a good job? Well, as I just said, we don't know. And so I would not encourage people to trust polls.

If by trust polls, they mean, assume that the result of a political survey is going to be right. What I would tell people, though, is that although the polls were wrong in, like, very material and substantive ways in 2016, 2020, it's worth considering all the ways in which even those fairly biased polls were still relatively close to the result. I mean, in 2016, the national poll said Clinton would win by four, and she won by two. In 2020, the national polls had Biden leading by eight, and he won by four and a half. They're not perfect, even in their bad years, but I don't really buy that means that they're useless. So if your interpretation of trust is that you should take it to the bank and assume that what the poll say is going to happen on election day, I will absolutely not try and persuade you of that. But I think that the implication that there's no meaningful information in them is entirely wrong.

And I think that it's also often actionable information, which is to say that even when the surveys may not be predictive of a final election result, I think they still tell us a lot about what's happening in American politics. The polls in 2016 showed Trump was doing way better among white working class voters. The polls in 2020 showed him doing way better among Latino voters. There's a lot to glean from them. And I think that's true of the polls we just conducted, even if we were to stipulate that they're off by the same amount as the polls were in 2020 or 2016. Let's talk about your most recent survey. The headline results of this poll is that Trump leads Biden in the critical swing states of Nevada, Georgia, Arizona, Michigan, and Pennsylvania.

I believe Biden is ahead in Wisconsin. So across those six battlegrounds, you have Trump up in five, Biden up in one. And across all six battlegrounds, the president trails by an average of 48 to 44 percent. Anybody listening to this podcast surely knows that your poll kicked off. a period of extraordinary tooth gnashing and hair pulling and cable news screaming. What do you think everybody got wrong about your poll? I saw lots of people talking about Trump's strength on social media. That is not what I think our poll showed.

I think it showed Trump in a position as weak or weaker than he was three or four years ago. I think the thing that showed was that Biden in particular was far weaker than he used to be. I felt like that subtle distinction may have been missed. And I also thought that the bigger takeaway for me is not like necessarily that these poll results will, you know, be predictive of the next election result. I'm sure we'll talk about how it's a year away and so on. But these results, I think, will really, and not just from our poll, but also because the Biden campaign is going to look at the same numbers that we have. these results are going to shape the way that this what happens over the next year. The whole campaign is going to be organized now around the Biden campaign's effort to appeal

to a group of disaffected young voters, black voters, Latino voters who ordinarily support Democrats. The Trump campaign is going to have the same numbers and they're going to make the same decisions. We're going to have a really different presidential campaign focused on different issues with different messages as a result of these numbers. And it's true that Biden is down. and that's extremely important.

But the kind of, you know, it's worth pausing to think about how much campaigns not only affect election results, but also affect the decisions that politicians make in government. And there will be a lot of consequences to a campaign focused on these groups instead of, say, white working class voters like four or eight years ago.

To ask the question that has been asked 10,000 times, and so I apologize for anyone who has heard the answer 9,99 times. What's the deal with Biden's approval rating? It's not just this poll. It's not just all the polls last week. The core dynamic here, as I see it, is that if you take the average of Joe Biden's approval rating,

it is now stuck at about 1,020 days post-inauguration below every single post-war president not named Jimmy Carter. It's below all of them. he is an astonishingly unpopular incumbent. Now, you could say he should be more popular because look at the unemployment rate, he should be more popular to look at GDP. To a certain extent, that may or may not be true,

but it's also kind of like cope. The ultimate fact is that the approval rate is what it is, and it's very low, even for an incumbent in the 21st century where Americans just tend to dislike all of their politicians. What is your answer to the question of why voters don't seem to approve of Joe Biden? So I think there are two broad theories that I'll offer you. And I will give you the unsatisfying answer that it's so hard to tell which of these things is more important

because almost all the people in question believe both of them. And we're trying to entangle them even though everyone thinks the same thing. But the two things are it's just the state of the country and the economy. and the other thing is his age. And it's very difficult to my mind to untangle the two because I think it's entirely possible that Joe Biden's age has a huge effect on the way people see the state of the country

and his handling of it. But there's no doubt that an overwhelming majority of voters think the economy is bad and think the country is on the wrong direction, or heading in the wrong direction. We would almost always expect an incumbent president to suffer so long as that's true.

And it's also true that, 70% of voters think he's too old to be an effective president. And exactly the way cause and effect here runs, what way cause and effect runs is sort of hard to say. But they're definitely related. And I think that no matter what you think of Joe Biden, you would have to agree that Joe Biden has not been very good at or at times even attempted to do those more superficial aspects

of the presidency that could play an important role in building people's confidence in their leadership and in the direction the country is headed. you mentioned what I think are the two most popular theories. One, it's the economy, two, it's his age. I want to actually dig into those words, economy and age, because they are specific reasons, in a way, but in a larger sense, they are buckets. The economy is a million different things. The perception of Joe Biden's age is many different things. So let's talk about the economy first. I am very sensitive to an argument that goes something like this. It's all about wages and prices. And the

problem that Joe Biden faces is that even as inflation is coming down, we don't have deflation. We have disinflation. And that means that if the price of, let's just say, apples doubled between 2021 and 2022 and is only grown by 2% between 2022 and 2023, well, that means the price of apples has more than doubled since 2021. And people can see that every single time they buy an apple. So disinflation is not a medicine for inflation. It actually just means more inflation. And the only way to make prices go down is broad deflation across the economy. And that almost only happens in the case of a recession. And so Biden is kind of stuck. It's just really hard to have a popular message about the direction of prices when you've had about inflation. I do buy that idea. I also think it's possible

that people's opinion about the economy is broadly shaped by media messages, and the media has generally, and especially partisan media, has generally been very critical about the economy in a way that has obfuscated even the good news about,

let's say, real wage growth or productivity or unemployment or overall GDP growth. When you dig into this sort of bucket of the economy, what do you pull out as being the most important sort of, I'm mixing metaphors here

because I don't know if you pull weights out of a bucket. But what are the most important weights on Joe Biden's approval within this category of the economy? I look at it from a historical perspective. And one thing that makes it very hard for me to answer this question is that the kinds of things that are happening in the economy today are not things that have really happened during the record of modern public opinion research. We have seen, we did see inflation in 1970s. But most of the data that we have is more recent than 1980. And there are other elements of the economy over the last few years. that aren't simply about inflation.

We have huge increases in interest rates. We have a flat stock market. We have supply chain shortages. We have such a low employment rate that businesses are looking for work and labor that they can't find. And so I have tried to zoom out farther than the public opinion data

to try and understand what dynamics might be in play. And I can't help but land on 1920 and 1946 and 1947 as being the most analogous economic moments over the last century. are the moments after the end of World War I and World War II when the wartime economy was demobilized and there was an immediate bout of inflation as the normal economy returned. It was not transitory in the way that pundits meant it two years ago, but it was transitory in the sense that it lasted only a couple of years. And these were moments of tremendous political upheaval

in our country. You know the 1920 election most famously is the return to normalcy election when the Republicans won by an extraordinary margin. And 1946 brought the biggest Republican sweep in some amount of time that I'd have to look up before I, you know, risk saying something that was wrong. But it was a decisive Republican victory in 1946. And it was a volatile period as well. You may remember that 1948 was famously the Trumie, the Dewey beats Truman election where Dewey led in the polls before Truman mounted a late comeback. And so even, and it's also worth noting that, employment was not high in 1946, was not high, you know, in the early 1920s. And in fact,

it was lower than it was often throughout the previous decades. It was much lower in 1946 than it was during the Great Depression. It was lower in the early 1920s than it was during the early 1900s. And yet, voters were still deeply dissatisfied. I think it's easy to rationalize why this kind of economy would be unpopular. And I think if we had a, if we had public opinion data and robust economic data going back further, I think, there's a good chance we would see that conditions like these to the extent they can be quantified are associated with a lot of political love people. I think that's a really solid answer. Let's talk about Biden's age. So I think there are two layers here. One, it's not just that

Biden is old, as in has been alive for many years. Donald Trump is old. Nancy Pelosi is old. Robert De Niro is old. The problem is that Biden seems old. And it's weird to try to distinguish old from seems old, but actually, I think in this case, is very important. And the fact that Joe Biden seems old on television creates a second problem, which is that the White House knows that he looks old. And that, I think, is why he's had much less press availability than the typical president in the TV age. The White House has at its disposal this tool, the bully pulpit. And they leave it relatively empty because I think they know that when Joe Biden talks, he often looks old.

And so there's this two-layer problem here. Both that Biden's TV appearances serve as a reminder of the thing voters don't like about him and that the White House, therefore, can't make use of the bully pulpit and the presidency in a typical way because the president's appearances often risk hurting more than they promise helping. Is it really any more complicated than that? I think that when voters have seen Joe Biden, they think he looks old. And therefore, it's not necessarily in the White House's interest to have him out there more. I mean, I think that's the that's the likeliest explanation. I think that's a real problem for the White House, because it's

not a problem that you can fix. Like, age only age goes in one direction. I mean, is there anything else to even say here? This is in a way the most important thing. Look, I talk to people, even like smart Democrats all the time who I'm always blown away that I feel like there's something to explain here. Like I had a conversation recently where they just didn't understand why voters would think Joe Biden is too old, but not Donald Trump when they're only three years apart. It's like, well, they're not judging it based on their birth certificates. They're watching them and they have reached a judgment. That's whatever you think of it based on what they see and they don't think that this is a guy that's sharp and capable of handling the White House and the way that

they would want. And that he is absent only adds to that. And he's absent in part because, as you've noted, it wouldn't necessarily serve his interests to be out there more. And when the world seems chaotic and when things aren't quite going right, that does not wind up reflecting well on the president, especially one who can't perform the superficial aspects of the job. To respond to what might have been one of the most common criticisms of the reaction to your poll, let's talk about the panic meter and the degree to which one poll 300 days out says anything about the election. It seems to me like there is at least some historical evidence that polls that try to predict presidential elections a full year out aren't very good at anticipating the future.

That said, to me it seems like 2024 is different. Not only do we have an incumbent on the Democratic side, we also have a kind of incumbent on the Republican side. It seems to me almost impossible that Donald Trump will not win the Republican primary, in which case we will have the first presidential rematch in about 70 years. This is not your typical election in the history of presidential rematches. This is an election that almost everyone knows a year out who the two candidates are going to be. And those two candidates are two of the most famous politicians in the world. There's a high degree of familiarity baked in on both sides. That is my, I guess, kind defense of why people should pay very, you know, closer than average attention to your poll a year out.

How do you feel about this question of thinking responsibly about a presidential poll 12 months before the actual election? So first, I broadly agree. I think that there's no doubt in my mind that polls a year out are much more meaningful and perhaps and also predictive. I'm I'm putting that in quotes, of the final result. Right now in the modern era of polarized and highly informed politics, then they have been over the long history of presidential polling since the 1930s. I think that we have a rematch between Joe Biden and Donald Trump makes it all the likelier that these polls are plausibly, again, in quotes, predicted.

That said, I would urge people looking at the poll. at this stage to think a little less about whether they're predictive and instead more about like, for lack of better word, whether they're serious, whether they're based on real and genuinely held attitudes or whether they're based on voters offering a response to a pollster that, you know, maybe they haven't thought very much about before. And it's possible for a poll like this one a year out to be extremely serious, which is to say based on genuinely and firmly held views about Joe Biden and Donald Trump and not be predictive. In this particular poll with Joe Biden and Donald Trump, like, I take these numbers extremely

seriously. These numbers are coming from voters who know all about these two guys. Some of the most famous people in the world, as you noted. And that does not necessarily mean that they will be predictive, in part because the results to the poll show that there are so many young and non-werectuary. white voters who are disaffected with Joe Biden, who will now have to agonize about whether to vote for Trump, return to Biden, or not vote as this campaign gets underway. So there's a ton of possibilities, many possibilities for voters to change their vote preference over the next year. But that still

doesn't take away that the attitudes that they're expressing with these candidates are very real and they're informed and they're actionable if you're a member of the Biden or the Trump campaign and they're worth taking seriously if you're a reader. If you dig into this poll and others, there are a few important trends that I want to get your mind on. The first is that Hispanic and black voters seem to be shifting away from the Democratic Party and toward the Republican Party.

Do you have a definitive explanation for why we're seeing this shift? You asked for explanations. Honestly, we could spend a whole podcast listing out and evaluating the merits of the different theories. And I'll skip sort of to the, end of that by saying that it's extremely hard to tell which of those theories is most important. And I note a decade later, we're still arguing over why Donald Trump did so well among white working

class voters in 2016. Was it racism? Was it the economy? Was it Hillary Clinton? Was it immigration? I mean, there are so many theories that are so poorly tested that even now with the benefit of the final election results and the final post-election studies, we still don't have a definitive answer to one of the biggest shifts in modern electoral history with a much larger demographic group, by the way, than black and Latino voters. We're not going to have a definitive answer to this question in this election either. The broader pattern, though, that I see, not just among black and Latino voters, but also among white working class voters in 2016, as I just mentioned, is that every time

we've had an election over the last decade with Donald Trump, the sort of the older political attachments forged in earlier era of politics seem to weaken. If you are some, you know, white working class voters are the most obvious examples, right? I mean, these are people who, you know, they're not liberal on social issues necessarily, but they voted for Democrats and the belief that they were best for working families and would fight for, you know, working people over big business and so on. And that was that, those those views of the two parties were forged during a totally different era of American history at a time when we had an industrial society with big labor and big business. And

that's just not the country we live in anymore. And so those, you know, the legacy of as, as we move farther away from fights that forged those political allegiances between white working class Democrats and the party have moved farther in their rearview mirror, they get weaker and eventually there's an opportunity for something to break through them. And I think something similar is probably true for black and Latino voters as well. The all but unanimous support for Democrats among black voters was forged between 1960 and 1964. I mean, before the Civil Rights Act, black voters didn't give Democrats 95% of the vote, nothing like it, even in northern black city, even in northern, you know, predominantly black areas.

So the, we're now 50 years after the, after the Voting Rights Act, or 60 years, in fact. And it doesn't surprise me that the bonds forged during that period are weaker. And Donald Trump has been such a huge part of American politics, such a well-known figure, that as we polarize around attitudes about Donald Trump, this one human, we wind up learning that the attitudes about Donald Trump, this one idiosyncratic human do not line up with the way political divides were drawn 50 or 100 years ago. And so we see a weakening of those longstanding political patterns. And the fact that Donald Trump has a populist brand of politics with appeal to blue collar and less engaged

workers makes that even easier to happen because all of the groups that I've been talking to here, black, Latino, and white working class voters are groups that are not necessarily, I ideologically consistent liberal groups, but they're less affluent. They have lower on, they have lower average educational attainment. And so there is a large number of voters from these groups who have been voting for Democrats, even though they don't hold the Democratic Party's view on all the issues and are also sort of naturally receptive to the message of a populist outsider. And so when you put all that together, I think you have like a unifying theory of this populist outsider conservative named Donald.

Trump dominating our political life for 10 years and by bit, by bit, eroding political allegiances that were forged long, long ago. In September, you published an analysis of the non-white electorate, and you found that just about every cohort among non-white voters in America are moving to the right. Black voters are shifting right. Hispanic voters are shifting right. Other non-white voters like Asian Americans, some of them are moving right as well. This is happening for men and for women.

It's happening for old and for young. Democratic support declined among non-white Americans for every gender and age and education and income level except for one. And that is higher income non-white voters. With incomes over $100,000, this is the only place among non-white voters where democratic support seems to be holding even. And this seems like a huge deal to me. right the Democrats in the last decade have been partied to this trade where they've picked up a lot more rich voters, a lot more college graduates. They become a more high-income cohort, but they have lost a lot of support among the working class, both the white working class and as you show in this article, the non-white working class. Do you want to quickly comment on this little wrinkle in the Democrat non-white cohort before we move on to the next thing I want to talk to you about, which is young voters?

I think the point I'd like to make is I would question whether it was really the Democrats who made the trade. I think the Republicans made the trade. I think that Donald Trump embraced a much more populist and in many respects more moderate brand of conservatism that undermined the old basis for Democrats to appeal to these voters on socioeconomic grounds. In 2016, it was Hillary Clinton who was on the defense on trade for supporting TPP. Remember that back in 2004 and 2008, it was Obama and Kerry who got to campaign on outsourcing and the way that Republicans would outsource middle class jobs and so on. That was now Trump's issue in 2016. In 2012, Democrats could campaign on how the Republicans would take away Medicare and Social Security. Donald Trump took that issue away from them. I mean, we can go down the list, but the old basis for

many of the, not the whole old basis, but many of the arguments that Democrats used to make to working class people were co-opted by Donald Trump in 2016. And we're than co-opted because Hillary Clinton so happened to be a candidate who was especially vulnerable to a critique of neoliberalism, so to speak. And it's worth pausing that the Democrat, the other half of the trade that Democrats then make to campaign against Donald Trump as a threat to democracy, to campaign against Donald Trump as being racist and to campaign for something like stronger, you know, on a message like stronger together. I mean, that was the only card Hillary Clinton had to play in that election. And as long as the Republicans are,

this sort of brand of populist politics that has as much appeal to working class voters as it does, I mean, I don't know that they have some other option to campaign on a different set of issues. This is their only path to winning right now if you stipulate, if you're willing to stipulate that there is an authentically popular element of Trump's brand of conservat. And there's a very, very fair interpretation. In your polling and others, Biden's support among young people who have four, the last many generations moved toward the Democratic Party has been very weak. What's the smart way to interpret Biden's weak support among young people?

Do you think that it's young people who are really considering voting for Republicans just as likely as they're voting for Democrats? Or do you think there's a kind of approval boycott of Biden among people who are overall clearly to the left of center? So one, I think you're absolutely right to wonder about how much, this is the sort of arcane survey word for it, expressive responding is going on here in a poll taken a year ahead of an election. For some of these people, they haven't begun to think about whether they're going to vote for Biden or Trump yet, but they, you know, right now Biden is the one on their mind. They don't like him. They don't want with Trump.

I think that's entirely plausible. And as the campaign gets underway, if that theory is right, we could see a steady movement toward Joe Biden, a moment. among young voters. One twist, though, that I think really shapes, and one, sorry, one twist that really shapes my framing on this and my thinking on this is that there's a huge split among young voters and non-white voters as well, depending on whether they are regular voters in American politics. The regular voters who participate in primaries and you participate in midterms, they look much more, and putting in quotes, normal in terms of whether the breakdown of Democratic or Republican support looks like what we're accustomed to in recent cycles. Joe Biden is

winning among the people who voted in 2020, among the young voters who turned out in 2022 by a wide margin. He's winning by a much more typical margin among the black and Latino voters who turned out in 2022 and so on. That is not true among the less engaged voters, where Donald Trump can sometimes even leads among less engaged black, Hispanic, and white younger voters. Again, I don't know how seriously to take that. from the standpoint of like whether it is predictive,

I think there's a distinct chance that many of those people simply don't vote. That if you're someone who is a low, historically low turnout voter, you're relatively liberal, you don't like either of the candidates, you may just not vote. And their willingness to support for Donald, to indicate support for Donald Trump could be an easy harbinger of that. So at the moment, if the election were held today, my first guess would be that youth, black, and Latino turnout is quite low

and that Biden does better among those groups than he does in our polls. If turnout is high and the election were held today, I do think that a lot of this expressive responding would become a real protest vote. I want to put a pin in something you just said because I think it's subtly mind-blowing to anyone who is used to the politics of the 2000s and early 2010s. It's Democrats who are historically the party of get out the vote. It's celebrities on MTV telling viewers go vote or celebrities on Instagram, making up songs to get young people to vote. And the calculus there was, the more people, the more young people we can bring into the ballot box, the better off Democrats will be as a result.

And of course, I want to bracket this and actually kind of scream this, that as a civic matter, I absolutely agree with the idea that it is good for people to vote, period. Voting is good. But what you just said that I think is true and counterintuitive is that a high turnout election in 2024 could, I ironically benefit Donald Trump because he seems to do particularly well among people, including young people, who are less educated and less likely to vote. So this is a good segue to the recent off-year elections that we had a few Tuesdays ago. Democrats won the governor's race in Kentucky. They won legislative elections in Virginia. They won on abortion in Ohio. And this is just hours after your poll suggested that Joe Biden was in deep trouble. What lessons do you take from

the 2023 elections. I don't think there are too many lessons to take from it. I know that's not an answer that a lot of the people responding to me on Twitter seem to want to hear, but these are very unusual and atypical races. We're talking about a Kentucky governor's race where the Democrat, regardless of what happened, was going to run 15 points plus better than Joe Biden did because he has such broad support among voters who would never consider voting for a Democrat for president. We're talking about an abortion referendum, which isn't even a partisan vote between the two sides.

We have known in the polls for years that there are millions of Republicans who support abortion rights, and they have shown that in every ballot initiative that we've had since the Supreme Court overturned Roe v. Wade. Of the elections, the Virginia election is the one that is closest to a typical D versus our national political matchup. That was also the worst for Democrats. They did retain control. They did win control of the state legislature narrowly. But, I mean, this is a state Biden won by 10. It was not, if you, like, looked at the popular vote.

It would be a significant underperformance. I'm not telling you or anyone to read into the Virginia result either. It's also an idiosyncratic state election shaped by issues that are not the same as what will decide a presidential election. But, I mean, again, it's just very hard for me to read anything in particular into it. And if I did, it would probably be on net something good to read in, um, just, the Republicans. One final point I'll make on this is that, just broadly speaking, the polling showed Democrats poised to win Kentucky, and they showed abortion referendum and marijuana referendum leading in Ohio, and they showed close race in the Virginia state legislative contest as

well. And those same polls also show Joe Biden is extremely unpopular. So there's no necessary contradiction between Democratic strength in these elections. And at times, very good Democratic results. One interpretation of Tuesday that I thought was really interesting is the observation that since the 1980s, 1990s, Republicans tended to be the party that overall performed better in midterm and off-year elections than it did in presidential years. And that's because it tended to have more active and motivated voters in those years. It had a larger share of college-educated and high-income voters, people who were more conscientious about participating in local elections that often avoid the national spotlight. But the trade that we've been discussing, whether it was the Democrats,

the Republicans that are responsible for the trade, in which Republicans have become a more diverse and more low-income, more low-education coalition, means, ipso facto in a two-party country, the Democrats have shifted toward a more high-income, more high-education party that is better positioned to take advantage of lower salience, local and state elections. And to a certain extent, that is a part of what we saw on Tuesday. To what extent do you think that interpretation holds any water? Well, it's exactly what our polling says. Now, as we started the show, it's, as we said at the beginning of the show, polling is imperfect. Maybe it's wrong. But our polling shows Joe Biden leading among regular voters who vote in midterms and who vote in primaries and

trailing among everyone else. And by the way, fearing especially poorly among the subset of people who have only voted in the 2016 or 2016 presidential elections, but not in any of the other ones. So there are, to my mind, three things going on there. One is what you mentioned. There's this demographic shift where Democrats do better among white voters, college educated voters. And also, by the way, older voters than they did a decade ago, I think that's overlooked. These electorates are extremely old, and the polls show Democrats very well among seniors and sometimes winning them them even now. The second thing, though, is that I think that there's an additional turnout advantage for Democrats that I see in the data that goes beyond demographics. And I'm just going to call it

like the resistance effect. Since Donald Trump's election, Democrats have been uniquely motivated to turn out in defense of abortion rights and in defense of democracy. And the way that I sort of see that in the data is the difference between more engaged Democrats who vote in elections and those who don't. And our polling, the people who, the Democrats who turn out in the midterm are extraordinarily Democratic Democrats. And there's nothing like that on the other side for Republicans. High turnout and low turnout Republicans are basically equally likely to support Donald Trump. So as a consequence, you might have a low turnout election with an equal number of Democrats and Republicans, but where Democrats win because the Democrats who are showing up are extraordinary. ordinarily loyal. And you can also then have a higher turnout election where the partisan balance

remains equal, but you've drawn into the electorate, these much more, these less anti-Trump voters on the Democratic side, and now the Republicans when even though the partisan balance looks fairly similar. And then the third thing I'll note is that in our most recent polling, Joe Biden is doing really, really badly among these less engaged voters. So if Biden fixed his problem among young voters and, you know, and black and Latino voters, then they wouldn't be at a relative disadvantage in higher turnout races anymore. It would go back to something like what it was in 2020, where we didn't really see a huge difference between the voters who turned out that election, those voters who turned out in midterms. We have this sort of artificially

inflated gap between the highly engaged electorate and the rest, because the rest is almost entirely where Joe Biden's weakness is concentrated. I just want to make sure I'm correctly here. one implication of your argument. It seems like Democrats who are used to making the case that high turnout helps Democrats. The more people vote, the better things are for Democrats.

That in the 2024 election, it's at least possible that that's not true, that in fact, a high turnout election will turn out a group of voters that are, as you put it, less anti-Trump,

in other words, more anti-Biden, less likely to be reflective of the co-hotice, were the voters, the electorate that just helped Democrats win in a bunch of elections last Tuesday and more likely to help swing an election to Donald Trump. Is that conceivable? I mean, if the election were held today, that's what our polling suggests. I mean, even this most recent wave of battleground state polls, Biden's doing better among, quote, likely voters than

all registered voters, which is a reversal of what we're accustomed to seeing. The thing that, the only thing I'll give, that gives me pause about this, though, is that all of this, as I said, depends on democratic weakness. And if Joe Biden were to, you know, sort of re-energize his support among young black and Latino voters, not only would he gain in the polls, but he would also then probably stand to gain less from this lower turnout scenario that you're talking about. It feels like we're circling and circling this idea that I want to make sure I remember because it really seems like such a big deal.

I'm going to try to think of it as like the engagement gap or the engagement riddle of 2024. What makes this election different is that less engaged young voters typically lean wide left, but today less engaged young voters are telling pollsters like you that they're especially alert to and allergic to Biden's age. Is that it? I think they're two easily reconciled facts that are both contained in the poll end today. Joe Biden is relatively strong among these highly engaged voters who have been motivated to defend. democracy and abortion. That's not his problem in the polling. His problem is among less engaged voters who don't vote in elections like the ones we had on Tuesday. They're just not in this electorate.

And besides, many of them are liberal and might have supported abortion if they showed up, even though they have serious reservations about Joe Biden himself. And the thing that's going to decide this election is whether over the next year, these less engaged voters who we haven't been seeing in special elections, who we haven't been seeing in Nafier generals, who we haven't been seeing in midterms, whether they can be persuaded to return to Joe Biden's side, There are reasons to think that's possible, like their relative liberalism on an issue like abortion, the fact that they don't like Donald Trump. There's also this big question about age and the economy that potentially stands in the way of that. And I don't know which will prevail over the end.

The polling at this stage can't answer it. But what I'm sure about is that that question was not put to the test in a low turnout election about the Kentucky governor. And how seriously do you take this theory, given what you can see in the polling data? I take it very seriously. I think it makes intuitive sense to me that voters who are not highly engaged and who are not especially ideological and who judge their politicians on a more superficial basis than many other people are judging Joe Biden in no small part based on his appearances. And as a consequence, he is far weaker than he would be otherwise. Again, I keep coming back to this engagement gap because I think it potentially makes sense of a lot of things. Like, who are, I mean, of course more ideological voters still stick by Biden, while the sort of voters who don't care about policy are moving away from him and would be, what would naturally make sense of that?

Well, something superficial like age is exactly the sort of thing that helps make sense of that. Now, the economy does two to be fair. But in the past, when the economy is weak, I don't think we see these kind of, you know, huge differences between regular and irregular voters. I think a democratic strategist listening into our conversation might say something. like this. You're both missing something huge here. The thing you're missing is that right now polls are a referendum on Joe Biden, and the entire media keeps talking about should we replace him, which by the way isn't going to happen, and oh my God, he looks so old, and we're committing many of the fallacies or the mistakes or problems that this Democratic strategist is typically against.

But next year, this Democrat is saying, the election is not going to be just about Joe Biden. It's going to be a referendum on Donald Trump and his crimes and his weird authoritarian statements and his general clownishness. So, Nate, why should or shouldn't Democrats all just relax? You know, one thing that strikes me is that Joe Biden's approval rating hasn't improved at all this year. Typically, after a midterm, the president's approval rating begins to go back up a little bit. They benefit from the contrast with the Republican with the other side. Donald Trump himself started to do better in the polls by this time in 2019. I don't see any of that happening for Joe Biden.

And one explanation for that is his age. And the other explanation, though, is the debate about Joe Biden because it prevents Democrats from sort of getting behind him and beginning that process of, you know, beginning to rally behind their own candidate and thinking about the election mainly in terms of Joe Biden versus Donald Trump, which might be a favorable comparison for Joe Biden. But instead, many voters are thinking about this in terms of just should it be Joe Biden. And until the debate about whether Joe Biden's Democratic nominee is over, I don't see how we can be surprised that many voters are thinking about their thinking about Joe Biden primarily in those terms and not with respect to the Republican alternative. Nate Cohn. Thank you very much.

Thanks for having me, Derek. Thank you so much for listening. Plain English is hosted by me, Derek Thompson, and produced by Devin Manzi. Some great news for you all. As you probably know, we are returning, have returned back to our normal schedule. of two pods a week, so be on the lookout for new episodes every Tuesday and Friday. If you like our podcast, please rate. Give us five stars. Subscribe wherever you listen, and I'll see you later.