Chapters
Shownotes Transcript
You're a podcast listener, and this is a podcast ad. Reach great listeners like yourself with podcast advertising from Lipson Ads. Choose from hundreds of top podcasts offering host endorsements, or run a reproduced ad like this one across thousands of shows to reach your target audience with Lipson Ads. Go to LipsonAds.com now. That's L-I-B-S-Y-N-Ads.com. I had something kind of crazy happen to me.
I was swimming in a stream because it was so hot. Something brushes up against me.
I couldn't tell what I was looking at. And so I just screamed and like started going backwards and everyone else started screaming and getting out of the stream. It was an effing water snake, a big ass water snake with a live catfish in its mouth. And it swam up on land. My God. Holy crap. Before it ate it. Before it ate the catfish. Anyway, that was my excitement for the weekend.
Hello and welcome to the FiveThirtyEight Politics Podcast. I'm Galen Druk and welcome to Debate Week. This Thursday, President Biden and former President Trump will go head to head in the earliest presidential debate ever. The debate stands out for a couple other reasons, too. The Commission on Presidential Debates has been excluded from the process. There will be no audience and the candidate who isn't speaking will have his mic muted to avoid interruptions.
This is also, of course, the first contest between a president and former president in more than a century. So given all of that, how much will this debate matter? Who will tune in?
And will historical lessons about debates and voter perceptions actually apply? We're going to talk about it today. We're also going to give a preview of a couple high-profile primary races taking place on Tuesday. Representative from Colorado Lauren Boebert is being challenged by a fellow Republican, and Representative from New York Jamal Bowman is being challenged by a fellow Democrat. And...
And we've got a couple good or bad use of polling examples. Is the universe of persuadable voters in the battleground states really 60 percent of registered voters? And does a new poll from the queen of polling herself, Ann Seltzer, spell trouble for Biden?
Here with me to discuss it all is senior researcher Mary Radcliffe. Welcome back to the podcast, Mary. Good morning, Galen. Thanks for having me. Great to have you. And also here with us is senior elections analyst Jeffrey Skelly. Welcome, Jeffrey. Hey, Galen. Good morning. Let's begin with the debate. And I'm going to have you stake out a position on...
the broad question first, and then we can get into the details. So put the specifics of this race aside for a second. I'm only going to ask you to do that for a second. Do presidential debates matter, period? Mary.
I mean, most of the time I'm pretty skeptical that general election debates change a lot of minds. That's, I think, what we tend to see in research here. There's a 2019 study from some researchers at Harvard Business School that didn't really find any significant impact. So most of the time I would say no, although I mean, this one might be different for a couple of reasons. So we'll we'll see. Yeah, I think I would sort of echo Mary's sentiments, which is maybe like a qualified no.
But there are some unique things about this debate, given the timing, the participants, the fact that they are so not well-liked. I think there's also a fair...
question about how engaged voters are yet. Like the fact that it is such an early debate, will this actually raise sort of voter awareness of the election at an earlier point than you often see in presidential election cycles? Or are people sort of just more informed or thinking about the election? You know, does that have some sort of effect? Or does it mean that people just don't tune in? And I do want to get to that in a second. But first,
I want to ask a little bit more about what we mean when we say it doesn't matter. Because according to an average from 538, looking back to 1976, after the first debate, so just the first debate, not all of the debates necessarily, you see that the incumbent doesn't see much change
at all when it comes to his support in the polls. But you see that on average, the challenger sees an increase in support in the polls of almost two percentage points. So when you say that it doesn't matter, do you mean that, yeah, that's the first debate and by the time people actually vote, that will have...
reverted to whatever the average was anyway, that like it can matter in the short term, it can change perceptions, but that those perceptions won't stay changed or that ultimately two percentage points isn't enough to say that it matters.
I think most of the time, two percentage points isn't enough to change the outcome of an election. Now, this time around, that may be different, right? This is a really close election, much closer than I think what we would see historically going back all the way to 1976. So in this case, maybe two percentage points makes a difference.
But I think also, you know, this time around, we think about incumbent and challenger. I mean, you may see an effect on the challengers vote share because people don't really know who they are. That's not the situation we're in right now. I would be skeptical that this time around we'll see an impact like that. Any impact we might see would be more voters starting to tune in a little earlier than they normally do. Yeah.
So when we think about maybe the limited way the debates have impacted the race historically, what should we think about? Are there specific moments that change things? Is it the first debate can matter, but none of the other debates matter? What should we be looking for if this debate is going to matter?
I think the first debate question is kind of an interesting one. How empirical it is, I think, is debatable, but there is some thought that incumbent presidents have struggled in first debates.
Trump in 2020, for instance, was viewed as having done a very poor job. Barack Obama in 2012 seemed to lose to Mitt Romney. Lose, of course, here is like, you know, sort of in the narrative after the debate and people's reactions to what was going on. Ronald Reagan in 1984, which might even be a more pertinent example, given the concerns about age of both of these candidates, was seen as having kind
kind of meandered and not performed well and seemed to not have a huge grasp of some statistics. And this caused a lot of coverage after the debate to talk about, well, maybe his age is a concern. At the time, he was 73. He was the oldest president in
in the country's history at that point. But then in the second debate, he had a much stronger performance. It even had kind of a funny moment where he tweaked his opponent, Walter Mondale, the former vice president, who was in his mid-50s and said, I will not make age an issue of this campaign. I'm not going to exploit, for political purposes, my opponent's youth and inexperience. So, you know, sort of putting the age thing to rest. But after that first debate, Reagan did see a drop in his support. Mondale's support rose in the polls.
Was it ephemeral? Tough to say, obviously, because the narrative around the second debate was that Reagan had sort of recovered his footing. I think the first debate question is interesting, especially given that this one is super early compared, you know, instead of it being like mid to late September.
We have a debate in late June, which, yeah, as we've said, is unprecedented. Yeah, but I mean, I think in some ways we sort of have to temper our expectations for the like polarized environment that we're in now. You know, Jeffrey was talking about that first debate in 2020 where the media narrative was like that Trump did terribly. Biden absolutely destroyed Trump.
But if you look at the 538 forecast from the time in the two weeks following that first debate, the popular vote share in our forecast only changed by one and a half percentage points. Despite the fact that there was pretty universal agreement that Trump kind of fell down on the on the job on that debate, it didn't really impact the polling that much because everything is so polarized. Because there's even a difference between saying, I think this or that person won and
This or that person won, and therefore I'm going to change my support because 538 even did polling with Ipsos before and after that presidential debate in 2020. 66% of viewers rated Trump's debate performance as poor, so two thirds, and 57% rated Biden's performance as good.
So even though it seems that's only one poll, but yes, the media also maybe emphasized it, but people did actually think that Trump didn't do so well and maybe Biden did better. But we didn't see a dramatic change as a result of those views. We've already mentioned here that there are some things that are different about this time. So I want to be specific about how it could change things. So
One, this is earlier than ever before to the presidential commission on debates is not involved. So there's no audience. The mics will be muted. We have a president and former president debating. And I don't know if we want to add this to the list, but both are actually also uniquely old. Does all of that come together to mean that this matters differently? Yeah.
I certainly think there's potential for it to matter. I mean, look, any live event in politics has some degree of uncertainty, right? In most election cycles, like a presidential primary debate where you're choosing – you have voters who are choosing among a set of candidates from the same party. They might like more than one candidate. Their opinions can move around a lot.
In a general election, as Mary was saying, like, especially in our more polarized environment where there are a lot fewer voters who could be considered swing voters out there, there are just not that many true swing voters.
You wouldn't expect things to move very much, but again, if it moves a little bit, that could still have consequences. But I will say also at the same time that we're having this very early debate, and that leaves a lot of time for future events in the presidential cycle to change things. And I think one of the – actually, one of the difficulties about trying to get a read on this is that most of the analysis, of course, of prior debates is based on debates that happened in the fall when you're sort of at the height of the campaign.
For a lot of voters, they're not that tuned in yet, and we have this much earlier debate
How useful the past analysis is for this situation is unclear to me. I've mentioned in the past that I think maybe this could be a convention bounce kind of thing. We have to be we have to be modest that we don't really know. Open minded. Yeah, yeah, I think that's right. I think, you know, because it's so early in the election cycle, you know, if you look at like 538 polling average or recent polls, unless the pollsters force voters to choose between Trump or Biden, you end up with something like 20 percent of the electorate that doesn't want to choose either of them.
And that's the set of voters that I think this debate could speak to, which are mostly, you know, what are sometimes called low propensity or low information voters that aren't really paying attention yet, haven't made up their mind yet. You know, we have YouGov polling from last week that shows most voters that say they plan to vote for Trump or Biden don't think that the debate will change their mind yet.
But there's that, you know, some 20 percent that are not saying either Trump or Biden right now. I think those are the voters that that might be impacted here. I would be really interested to see not just as one of the two candidates get a bump from this, but does the share of undecided voters go down after this debate? I think that's that'll be an interesting question, because 20 percent undecided in June isn't that unusual, but it is unusual, I think, given the choice of candidates here as like universally known.
And the other question would be, does RFK Jr.'s absence from the stage sort of de-emphasize him as an option? Because perhaps of those 20% who really don't want to put their support behind Biden or Trump at this point, maybe some of them are RFK Jr. supporters. I want to do a mini good or bad use of polling here, which is that in the recent YouGov poll that you mentioned, where something like 80% of Trump and Biden supporters say that the debate will have no bearing on their vote, they also...
Asked registered voters, have you ever changed your mind about whom to vote for as a result of a presidential debate? 17% of registered voters said they had that they had ever changed their mind as a result of a debate.
You also mentioned earlier on some academic research that has suggested general election debates do very little to change voters' minds. It was something like three and a half percent of general election viewers switched from one candidate to another versus the 35 percent who will switch during a primary election debate.
Where does that put us in terms of Americans self-reporting on changing their minds and what the academic research suggests about Americans' propensity to change their mind due to a general election debate?
Yeah, I mean, people aren't that good at remembering what they did in the past. People aren't always good at remembering who they even voted for, let alone why they voted for them. There's some similar research from Pew Research Center that looked back at their post-election surveys over a couple of years and asked people when they decided who they were going to vote for, not what factors influenced who they were going to vote for, but when they decided. It
It's something like 10% had decided after the beginning of the debates who they were going to vote for. So by the time you got to the general election debates in previous years, most people had already made up their minds of who they were going to vote for. This year, obviously, is a little bit different since this debate is coming so early. We shouldn't expect that many people have to have made up their mind at this point. So maybe this debate will be more impactful as a result. At the end of the day, most Americans are not going to watch the debate. And most of the people who do watch the debate
are going to be rooting for one side or the other already.
So that's another reason why I think we have to automatically be cautious, especially in our very polarized period here to say, like, this is going to really shake things up. Yeah. I can't tell you how many folks in my life. Do you want to get dinner on Thursday? I'm like, no, I have to work. There's a presidential debate. Everyone's like, wait, there's a presidential debate. What? So that's a notch in favor of your argument that people may not even know that it's happening, may not watch it Thursday night, but we'll see the TikToks and the tweets and the
Instagrams and whatever. And that's how they will form their opinions. And we are going to be staying up late on Thursday watching the whole debate and recording a reaction podcast. So make sure to check your feeds either super late Thursday night or early Friday morning for some analysis after the fact. But for now, let's move on to our good or bad use of polling examples. But first, a break.
Today's podcast is brought to you by BetterHelp. This year has been going by quickly. It's already June. So what's something that you want to accomplish this year? When life goes fast, it's important to take a moment to celebrate your wins and make adjustments for the rest of the year. Therapy can help you take stock of your progress and set achievable goals for the next six months or so.
If you're thinking of starting therapy, give BetterHelp a try. It's entirely online and designed to be convenient, flexible, and suited to your schedule. You just fill out a brief questionnaire to get matched with a licensed therapist and switch therapists anytime for no additional charge.
Take a moment, visit betterhelp.com slash 538. That's the numbers, not the letters. Visit betterhelp.com slash 538 today to get 10% off your first month. That's betterhelp, H-E-L-P, dot com slash 538. Today's podcast is brought to you by Shopify. Ready to make the smartest choice for your business? Say hello to Shopify, the global commerce platform that makes selling a breeze.
Whether you're starting your online shop, opening your first physical store, or hitting a million orders, Shopify is your growth partner. Sell everywhere with Shopify's all-in-one e-commerce platform and in-person POS system. Turn browsers into buyers with Shopify's best converting checkout, 36% better than other platforms. Effortlessly sell more with Shopify Magic, your AI-powered all-star.
Did you know Shopify powers 10% of all e-commerce in the U.S. and supports global brands like Allbirds, Rothy's, and Brooklinen? Join millions of successful entrepreneurs across 175 countries, backed by Shopify's extensive support and help resources.
Because businesses that grow, grow with Shopify. Start your success story today. Sign up for a $1 per month trial period at shopify.com slash 538. That's the numbers, not the letters. Shopify.com slash 538.
Today's podcast is brought to you by GiveWell. You're a details person. You want to understand how things really work. So when you're giving to charity, you should look at GiveWell, an independent resource for rigorous, transparent research about great giving opportunities whose website will leave even the most detail-oriented reader stunned.
Busy. GiveWell has now spent over 17 years researching charitable organizations and only directs funding to a few of the highest impact opportunities they've found. Over 100,000 donors have used GiveWell to donate more than 2 billion dollars.
Rigorous evidence suggests that these donations will save over 200,000 lives and improve the lives of millions more. GiveWell wants as many donors as possible to make informed decisions about high-impact giving. You can find all their research and recommendations on their site for free, and you can make tax-deductible donations to their recommended funds or charities, and GiveWell doesn't take a cut.
Again, that's givewell.org to donate or find out more.
Last week, a couple of polls came out that caught our attention, and it's been a minute since we've done this segment, so we thought it was time for a double feature of good or bad use of polling. So the first example is a poll of registered voters in six key battleground states from the Washington Post and Schar School of Policy at George Mason. They focused mainly on what they call, quote, the deciders, or people who will determine the outcome of the election in those swing states.
A registered voter counts as a decider, according to this analysis, if, quote, they voted in only one of the last two presidential elections, are between the ages of 18 and 25, registered to vote since 2022, did not definitely plan to vote for either Biden or Trump this year, or switched their support between 2016 and 2020. So they don't have to be all of those things. They just have to be one of those things.
They suggest that although the common narrative is that the election will be decided by just tens of thousands of voters in a handful of states, that the universe of voters whose behavior is not predictable is actually fairly large. They concluded that 61% of voters in those six states can be called deciders. Is this a good or bad use of polling?
I mean, overall, I would say good use of polling here. The methodology they use here is really comprehensive to try to make sure that they have a good sample of the kinds of voters they're trying to reach.
They did sample recruitment in two phases, one phase to establish response rates so that they could make sure they were reaching out to enough voters across the political spectrum, across different kinds of voting behavior, you know, really tailored their sampling approach to the kinds of voters they were trying to reach and
how frequently those kinds of voters answer polls. They set up an incentive system to encourage responses. They spent six weeks reaching out to voters. All of this together, they ended up with a response rate of 8.5%. So the representativeness here is probably...
really, really good. They really worked hard to make sure that this is a good sample. Snaps, snaps, snaps all around. Also, Mary, your first good or bad use of polling segment ever, and you're coming out hot calling it a good use of polling, which doesn't happen all that often on this podcast.
You know, I can quibble a little bit with their definition of deciders, right? I think voters who say they're probably going to vote for a candidate, I don't know if those really should be in the persuadable universe, because if they are probably going to vote for a candidate, my guess is they...
Probably will. What would that get us down to? So instead of 60% of registered voters, it's something like... Yeah, I actually looked at this. I mean, it actually roughly cuts it by about half. Oh, okay. There's some qualifications in there because they use the term sporadic, and there's some overlap among the uncommitted and sporadic.
But if you just include, regardless of how sporadic they are, if you include all the uncommitted voters, or at least how they categorize them as uncommitted, who said that they would probably vote for Trump or Biden, it was like 30 of the 44% roughly. So that knocks you down to like a third of the electorate in the decider category instead of six, three out of five, which is a pretty big deal. And one of my main quibbles actually was because, to Mary's point,
People who lean toward a candidate almost always end up voting for that candidate, or at least a very large percentage of them do. So I don't think you should include them. I find them to be like in a different category from the other ones that they used for this for this particular analysis. Jeff's right here. I think there's really three categories. There's the decided voters, the leaning voters and then the persuadable universe.
Well, hold right on here, because I interpreted the deciders to also include folks who may or may not vote, period. So if we just focus on, you know, this person said they'll probably vote for Trump, and if they vote, they will vote for Trump. We should also be thinking about those people as maybe they won't turn out to vote at all. And so it's not just...
you know, will you vote for Trump or Biden? It's will you be motivated enough come November to get out and vote? In which case, I think it's fair to call those people deciders, especially in an environment where both candidates are really unpopular. People find politics distasteful. And we I don't know that we expect this to be a low turnout election. In fact,
I think that's pretty unlikely, but probably not the 70 percent or close to 70 percent. I think it was two thirds of registered voters or eligible voters voted in 2020. Yeah, but I mean, I think they're capturing those voters that may or may not turn out to the extent that it's possible by looking at history. Right. I wouldn't think that you could capture in a poll those voters who mostly always vote, but sit this one out. I think that would be pretty difficult to determine who that would be.
So what you're saying is if you voted in the past four elections and you say you're probably going to vote for Trump or Biden, you should be taken out of the deciders. But what if you only voted in one of the past two elections and you say you're probably going to vote for Trump or Biden? Because that's one of the categories of the deciders. You only voted in one of the past two elections. Yeah, I'd keep those in the deciders. They're deciding between Trump or Biden, whichever they say they prefer, and the couch vote.
I have one more follow up question here, which was that they come to the conclusion from this analysis that the people in the deciders camp, as they say, skew more in the direction of Biden's natural coalition than Trump's, which is to say younger and more non-white than the electorate at large.
They say that basically it's not a good thing that this decider group is part of Biden's coalition. You hear other people make the argument that actually it is good that this part of the electorate is more naturally Biden's coalition because that means he can win them back, basically, and increase his support between now and Election Day. So which is it?
I think this is the kind of question that might be easier to answer after the first debate if there is any shift in the polls, right? In the sense that if Biden is able to win back maybe some voters who would tend to lean Democratic but don't really like Biden all that much or are at best ambivalent about Biden.
but are not necessarily like obvious gets for Trump, you know, does that does that end up improving Biden's support? Is this debate an opportunity for him to to sort of reenergize or at least pull back in some of the voters who formed part of what was really an anti-Trump coalition in his 2020 election victory? That's like the big question there. And so it's sort of chicken or egg, maybe. I don't know.
Okay, so Mary, we got a good use of polling with some quibbles. Jeffrey, are you coming down as good or bad use of polling on this? I have no quibbles with sort of the polling aspect of this. Maybe it's a bad use of analysis. I'm often very...
skeptical of any attempt to try to make the case that there is a large portion of the American electorate that is somehow in play. It depends on how you're sort of framing your narrative and how you're writing about it and the data you're using. But I do think that all the evidence would suggest that that is just not true. Like there's just not much of the electorate that is in play. What is in play to the point of this analysis is that some people will show up and some won't.
We know that like some of those sporadic voters, they show up in one election, they might not show up in the next. And so that was actually I think where maybe there's the most benefit, like to say that, you know, the good use of analysis in this was to sort of write about those sporadic voters. Because we know that there was like a 2015 article from Michigan State political scientist Corwin Schmidt. He basically pointed out that like around a fifth of the electorate over time has been what he called surge and decline voters who voted in one election, but not necessarily the next or the other way around.
And sort of in a similar vein, though, a slightly different stat here because you have new people who are eligible to vote. Nate Cohn recently noted that around a quarter of presidential election voters do not have a record of voting in the previous presidential election. You know, that does suggest that like the sporadic or new voters like they could be important. And we do. And sort of getting back to your question about.
You know, is this good or bad for Biden? We do have some polling that might suggest that new voters, people who hadn't previously voted, are more of a mixed bag for him than you might expect, given like the demographics of that group. They tend to be younger, more likely to be nonwhite. But I know that at least there was like Pew Research polling that showed that people who didn't vote in 2020 were like a 50-50 split.
Biden and Trump. Yeah. Yeah. Time Sienna had something similar. So that's the thing is that it's it's not a given that like passed his prologue here. Wait, so can I do some really back of the envelope math here? I think generally across a
a bunch of different polling outfits, we see that somewhere in the range of 10% of Americans are truly undecided, as in they want to vote or they may have intention to vote, but they really don't know where they fall. And they could potentially be persuaded by one side or the other. And you're saying that somewhere in the range of 20 to 25% of folks will choose between voting or just staying home. If you want to term the deciders,
as folks who will decide between two candidates and whether to vote at all, it's maybe a third of registered voters? Or is that still... That still seems high. Some percentage of registered voters will not show up. Is that 25%?
I don't know. Maybe if two thirds of eligible voters showed up in the 2020 election, maybe three quarters of registered voters did. And obviously, there's a lot of state to state variation there. A lot of northern states have higher turnout. A lot of southern states have lower turnout. Jeff is just saying, as we all know, it all comes down to turnout. Brilliant analysis there. Yeah.
Doing my job here, really, really, really creating a lot of clarity. It all came down to turnout. All right. Our next example comes courtesy of a lot of different online commentators. We're going to look at The Guardian for these purposes, but you could really look at a bunch of different things being said over the past week. So last week, Ann Seltzer released a poll with the Des Moines Register showing Trump with an 18-point lead over Biden in Iowa.
50% of respondents said they support Trump compared to just 32% for Biden. And this spawned a lot of takes, including this one from The Guardian that reads, quote, The former president's big lead could be a bad sign for Biden's support in other Midwestern states.
he must win in order to secure a second term. While Iowa is not considered a swing state, there has in the past been some correlation between Trump's lead there and his support in Pennsylvania, Wisconsin, and Michigan, all states that Biden is targeting to win. They go on to say that this is still early in the game, as we all know. So Ann Seltzer is one of the best pollsters in America. And I am old enough to remember that
Late in the 2020 cycle, she released a poll of Iowa showing Trump leading by seven points when the average showed Trump and Biden tied in Iowa. And that was an early sign that Biden would underperform his polls across the Midwest. So is this takeaway from Seltzer's latest Iowa poll a good or bad use of polling, Mary? Bad use of polling.
OK, all right. That was now. Now we're back on track with what we expect from a good or bad use of polling segment. So, I mean, Galen, you mentioned this this October poll in 2020. Do you happen to know what her September poll in 2020 said in Iowa? Yes, they were. They were tied. They were tied. They were tied two months before the election. Early polls are early. Everyone needs to just calm down.
That's all you have to say. Early polls are early. Just calm down. Two months is well in the territory that we usually think of polls being more predictive and right post Labor Day is what we usually tell folks is when to pay attention to the polls. I
I've talked to Anselzer actually about what happened there. Why did you show Trump and Biden tied two months before the election, which was in line with the average and then this late in the game shift towards Trump? I think, you know, it's hard to put your finger on what exactly it was, but some actual late in the game movement, also the differential between Democrats voting early and by mail and Republicans sort of turning out kind of at the last minute, which means election day.
getting that get out to vote message just later than the Democrats were getting because they were voting for months or even just actual changes in the cycle. You know, I hesitate to say this doesn't matter at all. You know, like we also say that polls won't change all that much between now and Election Day because the universe of persuadable voters isn't that big. I mean, I think it's also it's one poll. No, it's the most important poll. Yeah.
She's a great pollster and no one's disputing that. I just think that we should never go off of one poll ever. So in that sense, I think it's a bad use of polling. Now, if Biden is down 18 in our polling average in Iowa, I would take that as – come September, I would take that as like a very bad sign for his chances in a state that he lost by what, eight in 2020? So that would be a definite –
At the same time, if you look at this poll, I mean, you had 9% of people saying they were voting for Kennedy. Where does that number go, for instance? And what did the Kennedy supporters look like? Does Kennedy end up at like 4%, 5%, which is sort of my ballpark expectation.
It might be that it's in Iowa that the Kennedy supporters are a little more Democratic leaning at the moment. I or in this particular poll, it was like that. So there's just a lot of a lot of variables to consider, to be sure. Yeah. I mean, I was also going to say, you know, Seltzer has a history of showing late movement, um,
In a lot of these races, you know, we mentioned 2020. She showed late movement between the September and poll in the October poll. The same thing happened in 2022. Her final poll of the Senate race had Grassley up by 12. He ended up winning by 13. But the poll a month prior to the race, she only had Grassley up by three. So I think, you know, Ann Seltzer's superpower is really detecting those late shifts in the race. And I think that's where she really shines. But we're not there yet.
So I do want to be sympathetic, though, to maybe some skepticism that we're seeing out there, because we saw in 2016 and 2020 that...
the Republican candidate, Trump, overperformed his polls across the Midwest. We're talking about Iowa here, but Wisconsin is also a prime example where the polls were really off, especially in 2020. And so when you see a high-quality pollster putting out a poll that, if it's true, seems like it would not be in line with the kinds of numbers we're seeing of a close race in Wisconsin, Michigan...
Pennsylvania, then maybe that tips people off a little bit. You know, so far we have been saying that Biden's numbers have held up amongst white voters significantly better than they have held up amongst voters of color. And his support amongst older voters has also held up much better than his support amongst young voters. But if Trump's
2024 is a replay of what we saw in 2020 and 2016, which is that polls are actually just not doing a good job of representing the views of white voters, particularly maybe non-college educated white voters, maybe older white voters or what have you. Then,
It's not that there's a difference between white voters and voters of color. It's just that we are not detecting what's happening with white voters as well. I am not saying that that's what's happening at all. I'm just saying I'm sympathetic to the skepticism that is out there. I think if you look historically, you don't see correlation between the ways in which polls are off from one cycle to the next. So it happens to be the case that they were off in 2016 and 2020, both in a way that favored Republicans. But just if you look over the
the broad course of history, it doesn't really mean anything necessarily for the future, right? Like the direction of polling error to some extent looks more like a random event than something like we can predict in that way. I don't mean to say that it is random. It occurs because of like changes in the electorate and sampling strategies and the ways that pollsters conduct polling.
polls, but it doesn't correlate one election to the next. So I would be pretty skeptical about reading too much forward for 2024 based on the ways that polls were off in 2016 and 2020.
Mary, let's say that this poll is correct. Iowa is a state that's roughly 90% white. Wisconsin is a bit less, and Michigan and Pennsylvania are quite a bit less than that. But, you know, the majority of voters in all three states are very clearly white. Can you apply lessons from...
Iowa to Wisconsin, Michigan and Pennsylvania in the way that people are doing. So put aside accuracy. Let's consider them totally accurate. Can you apply those lessons across those three states? Sure. I mean.
States are correlated and the closer they are demographically, generally the closer the correlations are. And our forecast takes us into account. Right. So when this Iowa poll hits the forecast, well, when it did hit the forecast, it also impacts our forecast in those states, as you mentioned, you know, Michigan, Wisconsin and Pennsylvania.
And you can go take a look and see, you know, there was some movement in our forecast at that time. But it is just one poll. Right. So we have to bear that in mind. The other thing is, I think, you know, while the demographics may be similar, there are also some differences between Iowa and these other closer Midwest states that are not just about race.
but about other aspects of the electorate. So, for example, Iowa is significantly more rural than those other upper Midwest swing states. So we would want to take those kinds of factors into consideration as well. There's certainly going to be correlation there, right? You do have a lot of rural white people in Michigan, Wisconsin and Pennsylvania, as you do in Iowa. But the more diverse electorate, not just by race, but also by like urbanicity and other factors, I think is also going to make a difference here.
Also education because we obviously know that that is the most notable split among white voters. But obviously white voters without a four-year college degree tend to be much more Republican and white voters with a four-year college degree now may lean slightly Democratic overall nationally. These are states that have large white populations without a four-year college degree, but Iowa's is larger.
As the person who tracks all the polls at 538, I went back to the polls to take a look at this question of white voters. We track crosstabs internally. So I pulled crosstabs in all the national polls, head to head polls over the last four months just to get a look at what white voters and see what we're seeing. And I don't think we see any significant movement really among that demographic category.
at all. So back in March, you see Trump is ahead roughly 13 or 14 points on average with white voters and the head to head nationally. Today, it's maybe 11 or 12 points on average. I mean, we shouldn't be too invested in these numbers because they're crosstabs and crosstabs can be noisy and subject to a significant amount of sampling error. But like,
Just to have a sense of it, there's maybe one or two points of movement in the last four months among white voters, which is about the same as we see nationally among all voters. So I don't see reason in the broader set of data to suspect that we're seeing big changes among the white electorate. I don't know that I think it's a good use of polling.
But I, like I said, have sympathy for people who are skeptical of things like this. And it's our job ultimately to try to make it make more sense. And there is lots of reason to be skeptical about polls. And that's because we should expect error. And so we just have to approach all of these things with uncertainty. But to pretend we know which direction the error will go and Mary, you're correct, we don't. So we are just going to have to sit in our uncertainty.
Before I let you all go, as I mentioned at the top, there are a couple high-profile primary races on the ballot in New York and Colorado on Tuesday. Jeff, what are you going to be watching? We actually are watching 15 different primaries in New York, South Carolina, Colorado, and Utah. The most high-profile is almost certainly the 16th District Democratic primary in New York between Democratic Representative Jamal Bowman,
and Westchester County Executive George Latimer. It apparently is now the most expensive House primary on record, according to Ad Impact. And there's been a ton of outside spending by basically a super PAC affiliated with AIPAC, the American Israel Public Affairs Committee, pro-Israel group.
basically targeting Bowman or helping Latimer. Again, getting back to sort of the Democratic divisions over Israel and the conflict in Gaza. This is a place where that is playing out, although I will note that Bowman's weakness is maybe not entirely due to that, but also he had previous weakness. He hadn't had that impressive of a reelection showing in his primary in 2022. And there were some other stories about him, like pulling a fire alarm and things like that that have
maybe not gone over well in terms of securing his position. And then in Colorado, we've got Lauren Boebert, who is pulling the unusual move in a non-redistricting year for Colorado of switching districts. So she currently represents the third district, which takes up most of West and Southern Colorado.
She's switching over to the fourth district, which takes up most of the eastern part of Colorado, which is a much more Republican district after she nearly lost in her current seat in 2022. So she's running for the Republican nomination there. And it looks like she's probably going to win the nomination there. She has Trump support. She has the backing of a lot of other people. But it's just an unusual switch. And obviously, she's a high profile individual within the Republican Party. So that's another thing we're definitely watching.
All right, we're going to leave it there for today. Thank you so much, Mary and Jeff. Thank you, Galen. Thanks so much, Galen. My name is Galen Druk. Our producers are Shane McKeon and Cameron Chertavian, and our intern is Jayla Everett. You can get in touch by emailing us at podcast at 538.com. You can also, of course, tweet at us with any questions or comments. If you're a fan of the show, leave us a rating or review in the Apple Podcast Store or tell someone about us. Thanks for listening, and we will see you soon.
