The NPR Politics Podcast - Here's How Associated Press Determines Who Won The Election

Episode Date: October 31, 2024

David Scott of the Associated Press shares how his organization discerns the winner of the country's elections, including the presidency. NPR relies on the Associated Press for race calls.This episode: voting correspondent Miles Parks and senior political editor and correspondent Domenico Montanaro.The podcast is produced by Jeongyoon Han, Casey Morell and Kelli Wessinger. Our editor is Eric McDaniel. Our executive producer is Muthoni Muturi.Listen to every episode of the NPR Politics Podcast sponsor-free, unlock access to bonus episodes with more from the NPR Politics team, and support public media when you sign up for The NPR Politics Podcast+ at plus.npr.org/politics.Learn more about sponsor message choices: podcastchoices.com/adchoicesNPR Privacy Policy

I'm Elena Moore. I cover new voters for NPR. That means people who've never voted before, especially young people. Their numbers and power are growing. What issues do they care about? How do they feel? What they say can tell us where this election is headed. My job is to bring their voices to you. To help support our work, sign up for NPR Plus. Just go to plus.npr.org. Hey, it's Miles. Real quick before the show, it has been a wild, exciting, oftentimes exhausting election season. And here in the homestretch, we want you to know about a few other ways you can keep up with everything happening each day. First, there's NPR's morning

news podcast Up First. That show is recorded before dawn and out by 7 a.m. Eastern time each weekday. Often you'll hear one of us from this show on that show. It's the morning news podcast that catches all of the news that happened overnight, up first. Second, later in the day, you can find a new episode of Consider This. That's the podcast where NPR covers one big story in depth each weekday evening.

It'll be all over this election and its aftermath as well. And of course, we will be here with you pretty much anytime there's big news with the context and analysis you need to understand it. So up first in the morning, consider this in the evening, and the NPR Politics Podcast every weekday and anytime there's something big happening. This is your election news survival kit from NPR podcasts. Okay, thanks for listening. And here's the show. Hi, this is Joseph from San Antonio.

I'm getting ready to vote early on Halloween. This podcast was recorded at 1.34 PM on Halloween 2024. Things may have changed by the time you hear this, but I will have cast my ballot in my Batman costume. Okay, enjoy the show. There's some kind of joke about like voter ID or... See, I was going to say like you're a superhero of democracy or something.

There you go. I always feel better when I come out of the ballot booth, so let's hope that's the case. Hey there, it's the NPR Politics Podcast. I'm Miles Parks. I cover voting. I'm Domenico Montanaro, senior political editor and correspondent. And David Scott from the Associated Press is here with us. Hi, David.

Hey, guys. Great to have you here, David. NPR relies on the AP for race calls, which means that's you and your team. And I think, honestly, it would be really helpful and kind of important for the listeners and for us to talk a little bit about how you guys do that on election night. So I'm hoping we could just start with, tell us a little bit about what your job is at the AP. Sure. I'm sort of the executive at AP who oversees our election operations in terms

of our vote count, our decision team where we declare winners, sort of everything that goes into the mechanics of elections and election night. And so for us, for our purposes, when anyone goes to npr.org and they see one of those little check marks that says this person has won Kentucky or this person has won Wyoming, that is coming from your guys' team over there. And I guess- Yeah, it's David's fault.

It's not us. David's got the big button, right? But tell us a little bit more about that.. David's got the big button, right? But tell us a little bit more about that. So race calls are different than projections, right? I mean, this is, you guys are completely certain when you make a call on a race that that person has won, right? That's the way we describe it, right?

That we're not making projections, we're making declarations. And it's a little bit of a rhetorical device to remind us the standard that we hold at AP when we declare a winner. We are not saying someone has won a race until we are convinced, we are absolutely certain that the trailing candidates can't catch up to the leader. And so we're looking at data, we're looking at our voter survey, AP Vote Cast, we're looking at the history of a state, we're looking at the advanced vote data, we're really looking at the votes as they're coming in, as they've been counted

and reported by election officials. But that's that standard. We hold ourselves to 100% accuracy. And so we say that we're not making projections. We are saying based on the evidence, this person has won. Soterios Johnson 10.15 Talk a little bit more about the AP Votecast data that you guys are using. So this is, am I correct in thinking this is kind of like an exit poll that you guys are kind of getting data from all over the country in terms of how people are feeling about the electorate or what exactly is that? Yeah, it's kind of like an exit poll, but in a really important, distinct way it's not, which is that we're not relying on interviewing voters as they exit a polling place. AP stopped working with

the group that does the exit poll after the 2016 election for a really important reason. Most Americans don't vote in person at a polling place on election day anymore. So if most of the people you're trying to survey aren't available, if you're standing outside a polling place waiting to interview them, we thought that wasn't the right methodology moving forward. So we developed with our partners at NORC, at the University of Chicago, AP Votecast, which is our version of an election survey. We are trying to reach voters as they are casting their ballots, be it in person, be it via mail, early in person. We're in the field for a week before election day, right up until the moment polls close. And we're

trying to meet voters where they are, by text message, by phone, by mail, any way that we can contact them to try to get a sense of what's happening in the electorate as people are casting their ballots. And the thing about those surveys is they're much bigger than the normal kinds of polls that you see in pre-election polling. I mean, in a state, you might get, you know, four to 800 people would be, you know, a normal size that you would get. How many people do you usually have in like, let's say a state like Pennsylvania, maybe three, four thousand? I have no idea why you would pick Pennsylvania for this example. Yeah, four thousand. And so when

you roll up all of what we're doing we're targeting about 120,000 people nationwide with our vote cast survey so in those battleground states Three to four thousand people that we were trying to talk to so it's five to ten times the size of a normal Pre-election survey and then you know on a national level with a sample of a hundred and twenty thousand It's it dwarfs anything else that you might see In a pre-election context. And it's really important that we do that for a couple of reasons. The biggest one is that it allows us to examine the electorate in ways that you can't do with a normal survey.

So, for example, we might want to know this year what Puerto Rican voters thought about some things in the election, as opposed to just Hispanic voters. And with a normal survey, your sample size isn't large enough to break demographic components down into those smaller pieces. And that's something that we can do with Votecast. And that would be because the margin of error would be way too high in something like a national poll to be able to break out those subgroups, right? Exactly. Right. You're just not getting the level of precision that you would need. The

other thing that we do with Votecast that we're really proud of is we just don't do the battleground states. Right. We will have survey data from every state in the union. And it's important to us at AP that that's the case. Every vote matters. Every person who's casting a ballot, their opinion, they're part of the electorate and they're part of the country. And so we want to be able to examine the whole of the electorate, not just the electorate in a place like, let's say, Pennsylvania, that might ultimately be determinative on

the road to 270. Well, I feel like people have been listening to this podcast for a couple minutes and we still have not yet asked David the big question on everyone's mind, which is, when are we going to know who won? David? everyone's mind, which is, when are we gonna know who won? David? It's a great question, and I only need to go talk to those 120,000 people and then listen to 160 million Americans

cast their ballots before I can tell you. Here's what I'll say. It is amazing. I asked really nicely. I know you did. I'm trying to help here. It is truly remarkable to me that as we sit here on this day, Halloween, we are looking

at a race that looks so, so close. The seven states that we think are going to matter, they all appear to be about even. What you're seeing with pre-election polls, everything's within the margin of error, so it's really a coin flip. There's so many factors that are going to go into that race That will be determinative. But if you look back to 2000 when AP didn't call That race that year We've called it a couple of times at 11 o'clock when the big states on the west coast close We've called it a couple of times on Wednesday morning

either very early Wednesday morning when Donald Trump won in 2016 a little bit later on Wednesday when George Bush beat John Kerry. And then the last election, it took until Saturday. So what I've been telling people, not a satisfying answer, I know, sometime next week. Yeah, that's likely. And, you know, it's funny, you mentioned polls close at 11 o'clock on the West Coast, and, you know, you'll call it right as a poll closes. I think for a lot of people, and I've written about this, you know, other times before, but maybe you can explain for people why the AP can call a state like California with zero percent in, you'll see, and then you'll see the AP has made a call.

Yeah, well, I don't know that we'll do that this year. I mean, we'll have to see what the data looks like. What we're able to do, if we look at a state that has a very long history of voting for one party or another, and we look at the advanced vote data, and we look at the registration data, and we look at results from Votecast, our survey of the electorate. And if all of those things point to a candidate winning with an enormous advantage, let's just say hypothetically Wyoming and Donald Trump. He's very likely to win there with a significant advantage. You know, 20, 30 points. That's a race that we can call when polls close. Although I'll be honest with you, I would expect less of that this

year just because of the nature of the race and how close it is. Yeah. All right. Let's take a quick break and more on all of this when we get back. This October, heal your existential dread with Shortwave. We're going to talk about the science of why we die and come out the other side with answers to live better as we live longer. Hear about the innovations when you subscribe now to Shortwave, the science podcast from NPR. We finally made it, election week. It's what this whole never-ending election cycle

has been building up to. And what happens now will dictate the future of the country. Keep up with election news when it matters most with NPR's Consider This podcast. All this week, we are taking major stories from the election to help you make sense of them and what they mean for you in under 15 minutes. Listen now to the Consider This Podcast from NPR. On the embedded podcast, every Marine takes an oath to protect the constitution against all enemies, foreign and domestic. This is the story of a Marine in the Capitol on January 6. This is the story of a Marine in the Capitol on January 6. Did he break his oath? And what does that mean for all of us? Listen to A Good Guy on the Embedded Podcast from NPR.

Truth, independence, fairness, transparency, respect, excellence. This is NPR. And we're back with the Associated Press's David Scott. And David, I'm hoping you can talk a little bit about how this election compares to 2020 when it comes to your processes. I think when I talk to election officials, it's becoming clearer and clearer that 2020 looks like an outlier election in a bunch of different ways. I'm wondering how all of the changes to our how people actually vote impact your guys' ability to make race calls. Four years ago, so different. I mean, we were in the middle of a

pandemic. I mean, I was looking at some pictures from our election night coverage just the other day and we were all wearing masks and face shields. And so think about how much of the moment that we were in as a country as we were trying to conduct an election that year. I mean, the candidates, they weren't even out campaigning in the way that they are today. Some states have totally changed how their votes are cast and what their system of voting is. New Jersey went to an all-male election in 2020 and they've since reverted back to a multi-mode election system like they had prior to the pandemic. So I think the biggest thing that

we've done compared to four years ago is to try to build on what we started then, which is to be so much more transparent about what goes into our vote count and what goes into our race calls. When I started doing this work, you know, an AP race call was a headline. And that was enough. And our audience sort of accepted our headline. And now our audience, your audience, the public at large, they want to know more information. They want to know about the mechanics of voting. They want to know how the process works.

There's a level of distrust that's been fostered by misinformation and deliberate bad actors, but there's this hunger for information about what goes into the process of voting and what goes into a race call or why someone has won. And so we've just had to restructure our news report so that our race calls don't have to be trusted, but they can be accepted as facts because we've shown our work, we've shown the data, we've shown the information that has yielded why our analysis has determined that this is the winner. One of the things, obviously people look at is

when the people who make calls get it wrong or it gets close, right? I mean, we had congressional hearings after the 2000 election. How much does something like that, experience of 2016 the experience of 2020? Sort of shape what you guys do because you know a lot of the other networks you know say that they make projections that often means that they're faster than you guys and making a Determination on who they think has won and maybe you can talk a little bit about why that might be the case and

You know what you guys have learned from the last couple of elections in one and maybe you can talk a little bit about why that might be the case and you know what you guys have learned from the last couple of elections in patients well it comes back to what I was saying earlier about you know our standard being a hundred percent in the primaries we got two race calls wrong both were state legislative races one was because election officials reported the wrong results and that one that we got wrong based on our analysis, it still gnaws at me. We really hold ourselves to that standard that we expect our race calls to be 100% accurate. We do think about being swift, and the reason for that is we think misinformation flows into a vacuum.

And so if we've not called a race, it gives an opportunity for bad actors to say, well, why hasn't AP done this? And so we want to be calling winners and declaring winners as soon as we're able to do so, but never at the compromise of accuracy, never at the compromise of making sure that those trailing candidates don't have a path to victory. Aaron Ross Powell You guys came under attack in 2020 because you made a quicker call for Arizona for Biden and the race got tighter and tighter and tighter and tighter. We all watched and we were wondering, is this going to flip the other direction?

It didn't wind up doing that. But maybe you can just talk about Arizona for a minute and what you learned from that, especially how close this election seems like it might be. You know, what we try to do when we're looking at whether the trailing candidates can catch up is to give them every benefit of the doubt. If everything broke their way in the votes left to count, would they have the ability to catch up? And in that year, our analysis concluded that if that happened for Donald Trump, he couldn't

catch up. He would fall about 10,000 votes short of Joe Biden in Arizona. And you know, everything broke his way in the days after Election Day. And he lost by 10,000 votes short of Joe Biden in Arizona. And, you know, everything broke his way and the days after election day, and he lost by 10,000 votes. So we're actually, in some ways, really proud of that analysis, because it held up, it showed that, you know,

the work that we were doing yielded the correct race call. I will tell you though, that we have spent more time probably looking at that race call than any others to make sure that we're never missing an opportunity to find a data point. We're never missing an opportunity to analyze data in a new way to make sure that we remain holding to our standard of 100% accuracy. Soterios Johnson 20.10.15

Domenico, can I ask you, you know, and NPR uses the AP race calls in our coverage, you're going to be on the air a lot next week explaining to voters how they should be watching election results. What do you have any advice you can kind of pass along? Well, I mean, I think David hinted at the idea of patience and really trying to get out front to explain to people what is going to be likely to happen so that they understand that just because you see the trend in one direction for one candidate doesn't necessarily mean that that's what's going to happen, to explain to people that we're not likely to get a result on election night itself.

So you know, watching these states looking at where the results are coming in, but also understanding what's left and having some patience to understand that things can change. I mean, you know, just because you see a big early vote lead, for example, for Democrats, doesn't mean that the Democrats are going to win by a lot because Republicans vote so overwhelmingly on election day. And just because that changes doesn't mean that it might not change more because mail voting comes in after that. And that might mean in the past, Democrats have done better on some of those mail votes. And that has seen a change, especially in races this close. Take a step back, take a breath, and be prepared for the fact that it could go on sometime.

David, just to take a devil's advocate perspective here for just one second, I want you to kind of explain to people who look at the idea of media outlet at all calling races and just say, why don't you get out of this business? You know, media, trust in media right now. We all hear about a headline every single day that it's at an all time low. Why is, do you feel like it's still important to be in the business of calling races as opposed to just letting election officials spend their days and weeks counting votes and then releasing the kind of certified results at the end of that process?

Well, I think a couple of things. First, I'll say when we're counting the vote, what we're doing is reporting back to voters decisions that really they've already made, right? Those facts, those vote counts are their votes. And we think it's important that they have a trusted source, a nonpartisan source like the Associated Press, to get that information as soon as it's available. Our system of democracy is decentralized. James Madison, Alexander Hamilton, they weren't thinking about 300 years into the future where

we are today. And a presidential election in the United States is really a series of elections contested at the local level at thousands of counties and towns and cities in New England parishes in Louisiana. And if it wasn't for the Associated Press doing this work of counting the vote, we wouldn't have a source, an unbiased source of information on election night and in the days after about what all those vote totals were. We would literally have to wait until the end of the process in January when those electoral college votes are open. To take a step back, just to think about the breadth of what the AP does, David, maybe you can just sort of shed a little bit of light on how you

guys gather vote data from different precincts and how many people do you have going out and doing that? Yeah, it is an effort that is both high tech where we are taking in data programmatically from counties and states that offer it that way. We're scraping websites, so we're writing code that pulls the data from the websites of counties and towns as they are posted, but it's also a manual process. We will have 4,000, what we call vote count reporters in county and city and town and parish election offices next Tuesday. And they will be there to literally get from election officials the votes as they are counted those reports in real time.

They call them in to AP. We still use the phone. We like that better than an app because there's no no fat finger mistakes. And we're able to quality control check that data. 800 people will be taking in those calls. We'll be looking at data. We'll be doing quality control checks. We look at multiple sources. So what do the, what are the results from those vote count reporters say versus what's posted on the website? If there's a discrepancy we go back to the election officials and say hey these numbers don't match up. What are the real numbers? And we

do that for weeks on end until we get to the point where all the vote is certified. But it is thousands of people, more than 5,000 people. We call it the single biggest act of journalism there is. Well, it does bring up kind of a final point I was curious about, David. Like everyone, I spend my days talking to everyone involved in elections and it feels like everyone involved in elections is feeling a lot of pressure right now Are you feeling that pressure? I mean now a few days away from Tuesday I don't like to say that we feel pressure here. We we honor the responsibility that we have I mean the way that the country's democracy has evolved back from those days and, you know, when the founders wrote the Constitution and us getting into this business of counting the vote and declaring winners

literally two years after the AP was founded in 1848 to today, we honor the responsibility that we have. It's really important for us, regardless of who wins, because it's not about us. And as I like to say, once polls close, it's no longer about the candidates. It's about what the voters have decided. And it's really just a big simple math problem. Who got the most votes? Let's add them up and see who won. The AP is David Scott. Thank you so much for joining us. Glad to be here, guys. I'm Miles Parks. I cover voting.

I'm Domenico Montanaro, senior political editor and correspondent. And thank you for listening to the MPR Politics Podcast. are just some of the podcasts you can enjoy sponsor free with NPR Plus. Get all sorts of perks across more than 20 podcasts with the bundle option. Learn more at plus.npr.org. You care about what's happening in the world. Let State of the World from NPR keep you informed. Each day we transport you to a different point on the globe and introduce you to the people living world events. We don't just tell you world news, we take you there, and you can make this journey while you're doing the dishes or driving your car. State of the World podcast from NPR.

Vital international stories every day.