podcast
The Assignment with Audie Cornish
Every Monday on The Assignment, host Audie Cornish explores the animating forces of American politics. It’s not about the horserace, it’s about the larger cultural ideas driving the American electorate. Audie draws on the deep well of CNN reporters, editors, and contributors to examine topics like the nuances of building electoral coalitions, and the role the media plays in modern elections. Every Thursday, Audie pulls listeners out of their digital echo chambers to hear from the people whose lives intersect with the news cycle, as well as deep conversations with people driving the headlines. From astrology’s modern renaissance to the free speech wars on campus, no topic is off the table.
What Polls Do (And Don’t) Tell Us
The Assignment with Audie Cornish
Jun 3, 2024
How will the trial of Donald Trump affect voters? There’s been a lot of partisan messaging from all sides about the numbers and what – if any effect – that’s going to have in November. How do we make sense of all of this? CNN polling and analytics editor Ariel Edwards-Levy joins Audie to discuss the power – and limitations – of surveys and polls.
Read Ariel’s latest: What the polls can’t tell us about the Trump verdict’s effect on the election
Episode Transcript
Audie Cornish
00:00:01
'How will the trial of Donald Trump affect voters? I actually get that question a lot, especially after former President Donald Trump was convicted of 34 felony counts for falsifying business records. And there's been a lot of partisan messaging from all sides about the numbers and about what's going to happen in November. Of course, Trump says he's confident he will still win the election, while Joe Biden and his supporters say the president will get reelected to another term. So we got to make sense of all this. How are we going to do that? To help answer that question. Ariel Edwards-Levy is joining me today. She's the polling and analytics editor for CNN. I'm Audie Cornish. This is The Assignment. And welcome, Ariel. How long have you been doing this work?
'Ariel Edwards-Levy
00:00:45
I have been at CNN for about three years now, but I've been covering polls and politics since the 2012 election, which is how I think of time.
Audie Cornish
00:00:52
So the before times.
'Ariel Edwards-Levy
00:00:53
Yes.
Audie Cornish
00:00:54
Before – were people more or less confident in polls at that time?
'Ariel Edwards-Levy
00:00:58
I think at that point there was more confidence, or at least there was less sort of awareness of what can go wrong with polling and sort of what some of the limitations are.
Audie Cornish
00:01:07
And 2016 was certainly a turning point.
'Ariel Edwards-Levy
00:01:09
Yes
Audie Cornish
00:01:09
Understatement.
'Ariel Edwards-Levy
00:01:10
Yes.
Audie Cornish
00:01:11
Was 2016 – do pollsters look at that as a time they were wrong, or a time when people stopped trusting them?
'Ariel Edwards-Levy
00:01:17
I think 2020 was much more of a concern than 2016
Audie Cornish
00:01:22
How come?
'Ariel Edwards-Levy
00:01:22
Because the actual numbers were further off. You know, I think people look at the sort of binary "this person's ahead versus this person's ahead." But 2020 actually saw some significant, you know, misses in some states, particularly in key swing states. So, you know, but I think seeing the fact that the pattern was the same in 2016 and 2020, that we were missing some of the same kinds of people pointed to an issue that pollsters have tried to fix. And I think that's one reason that you've seen a lot of change in the industry over the past couple of years.
Audie Cornish
00:01:52
All right. So I'm glad that you're bringing this up right away, because then that means we can get to our listeners. Our listeners have questions about polls.
Listener Voicemail
00:02:00
Hello, this is Mary. I'm calling from Kansas City, Missouri. I hear so much about polling, and I don't know why anybody answers the phone anymore. Certainly nobody I knew does. So where are they coming up with these answers on the polling? The voting – yes, no, undecided. I just wonder if it's all phone calls, or mostly focus groups or whatever.
'Ariel Edwards-Levy
00:02:23
That is a great question, and I am so glad to have it, because I think that there's still this image that when people think of a poll, they think of, oh, someone's sitting at home at night and they get a call on their landline phone and they answer and people go correctly. "Well, who's doing that anymore?" And the answer is –
Audie Cornish
00:02:38
Like they put down their TV dinner and the remote and pick up the landline and somehow have a conversation.
'Ariel Edwards-Levy
00:02:43
Yes, and the answer is that is not really how polling is being conducted anymore. I mean, I think you've seen a couple of really big shifts over the past decade or so that even I've been doing this. The first thing is the move from landline to cell phone. Pollsters are calling cell phones at this point, pretty much every major pollster that's doing live interviewers, which means somebody is actually dialing the number is going to be including mostly cell phones.
Audie Cornish
00:03:06
Right.
'Ariel Edwards-Levy
00:03:06
But then, yeah, a lot of people don't answer their phone. You know, you screen your calls, you see a number you don't recognize, you just sort of hit reject and move on with your life. Or sometimes it wouldn't even get to your phone. So pollsters are trying other things. One thing that we've done that I think is really great because it's sort of going very far back, is mail, which is.
Audie Cornish
00:03:25
So we at CNN, you mean.
'Ariel Edwards-Levy
00:03:28
Yes, one of the things that we do for our sort of major benchmarking surveys is we send letters or postcards out to a random sample of addresses all across the U.S. and say, "we'd like to hear your opinion on things." And people take the survey online, there's a link to it, or they can call in, but people still look at their mail because it's not sort of flooded the same way. And then there are other things. There are text messages, you know, there are instead of just randomly dialing phone numbers, you can look at voter files, which gives you a little bit more information about who's not answering your calls.
Audie Cornish
00:03:58
Which makes sense, because when I think about my own phone, I get a ton of fundraising or political kind of text spam. So obviously I'm findable.
'Ariel Edwards-Levy
00:04:09
Yeah. And so, you know, people's how you vote is private, absolutely private. But whether you have voted in elections is public. And so we can use some of that information to sort of understand what kind of people are answering surveys. Who are we missing? What kind of places do we need to sort of redouble our efforts? Because the ultimate goal of polls is to make sure that we are actually talking to a representative sample of people all over the U.S. And, you know, in this day and age, I think that's difficult to say it's absolutely everybody because there are some kinds of people who won't answer polls, ever. But for the most part, people who don't answer polls in many ways that count and that matter in terms of politics and in terms of attitudes and in terms of their lives, look like the people who do answer. And so we still can get something of a sense.
Audie Cornish
00:04:58
'Let me follow up on a couple of things you said there. First, you talked about sending cards out in the mail, etc., and that that's randomized. At the same time, are you trying to get a cross-section of Americans in terms of race, class, education? I don't know how much you can tell us about the recipe, but how does it work?
'Ariel Edwards-Levy
00:05:15
Yes, absolutely. And so what you are is you're starting with a sample where people have sort of an equal chance of getting picked, but we're also using weighting methods and sampling methods, which basically means we look at the responses we get. We know what the United States population or the population of a specific state looks like in terms of demographics. And we want to make sure that the responses we're getting actually line up with that. For instance, there's sort of a known thing, and you were talking about 2016. One of the things that we learned from 2016 is that people who have a higher level of formal education are more likely than people with a lower level to answer polls. And that's sort of always been the case. But it didn't really matter for election polling specifically because it wasn't very correlated with how you voted until 2016, when there opened up this big educational split. So in that case, if you're only getting people who have a higher level of education and you're not correcting for that, that's going to bias your sample. So that's sort of one of the types of things that people are looking for.
Audie Cornish
00:06:16
What I hear in our listener's question is a lack of trust. And I sort of understand it because that sense of like, "no one I know has answered a poll." I feel the same way about TV ratings. I'm like, "I don't know anyone with a Nielsen box." How do you how how should our listeners be thinking about this?
'Ariel Edwards-Levy
00:06:33
You know, I think it's a really understandable sort of sentiment. And, I mean, one thing I will say is that people do answer polls. It's not something that everyone is going to be sort of shouting from the rooftops that, "oh, I got surveyed" because I don't think to anybody but a pollster, it is the most exciting event of the week to be included in the survey. But, you know, one of the things that I love most about doing surveys is looking at the responses and sometimes will ask open ended questions where people can sort of write their own responses. Or sometimes I'll call back to talk to people who have participated in a survey. And every data point you see is the reflection of hundreds or thousands of people and people's individual views that are sort of being pushed into that. But every single survey is based on the views of normal people living in the US. And I think a lot of times those views don't get heard, except for the sort of loudest voices that either have a platform or are, you know, very active on social media, which doesn't represent everybody.
Audie Cornish
00:07:31
Social media is not real life?
'Ariel Edwards-Levy
00:07:32
Shockingly. I mean, I think it is. I mean, I think it is real life. But we also do know the kind of people who talk about particularly politics on social media are not representative, because a lot of people who don't feel that strongly or that polarized aren't going to be the ones posting. And so you're missing a lot of voices that I think are so crucial, especially when we're sort of talking about, you know, democracy and what we want to be as a country.
Audie Cornish
00:07:57
One more question before the break: I was reading this New Yorker article where they were interviewing the Democratic strategist, Simon Rosenberg about polls. And he was pushing back a lot in the conversation. And he was basically saying, and this is a quote, that there's a tendency in recent years among commentators to overestimate the strength of Republicans and to underestimate Democrats. But he was also saying that that goes to the polling as well, that there's slightly more attention being paid to Republican polling firms versus Democratic ones. And I was wondering if you could talk about this world of polling that comes from partisan corners?
'Ariel Edwards-Levy
00:08:40
Yeah, absolutely. And I mean, I think we generally tend to stay away from covering polls that are sponsored by a particular advocacy group or specific partisan side. And there's a reason for that. It's not that partisan polls aren't good because many of them are excellent. Their incentives are to provide accurate information for their clients like everybody else, and a lot of them are doing excellent work. But when a partisan poll is being released, it's because they are trying to drive a message with that poll. And that message is not necessarily "I want to inform the public." It's
Audie Cornish
00:09:12
Like I give everyone a talking point with data for the next couple of days.
'Ariel Edwards-Levy
00:09:16
Precisely. And it's trying to sort of bolster your point with data that supports what you're trying to say.
Audie Cornish
00:09:21
But if you think about RealClearPolitics or any of these sites, not not just them, but sometimes people will say, this is a poll of polls and they'll take quote unquote, all the polls, put them together and yield some sort of number. Doesn't that also include polling from consultants that are have partisan leanings?
'Ariel Edwards-Levy
00:09:40
So it varies. Ours, you know, at CNN, we do not include those in our poll of polls. A lot of the aggregation does and I think –
Audie Cornish
00:09:48
Aggregation, meaning when we just see a number out on the internet.
'Ariel Edwards-Levy
00:09:51
Yeah. If you're looking at something that's saying, like, we're rounding up every single poll that's released and people treat this differently, there's a lot of ways of doing this. On one hand, I think it's really good to look at polls together, because a lot of times when you're seeing agreement in the polls on something or when you're seeing a lot of variation, even just that itself tells you something about sort of the strength of the findings that you're seeing. But it is important to sometimes sort of pick apart what you're looking at and say, okay, here's what this polling is coming from, and just be aware of what those numbers are coming from, whether a lot of them are coming from one side or the other, because that can have an effect.
Audie Cornish
00:10:29
'We're going to take a short break. When we come back, Ariel Edwards-Levy is going to talk to us more about the questions pollsters will be asking over the next few months. Welcome back. I'm talking with CNN polling and analytics editor Ariel Edwards-Levy. So let's talk about the question that's on everyone's minds, which is what are pollsters going to be asking in the next couple of weeks in the aftermath of the conviction of Donald J. Trump in New York? So before we get into why it's hard, first of all, what what is the question you craft and how do you think about the question you craft?
'Ariel Edwards-Levy
00:11:12
So there's a lot of ways of approaching that. And I think, you know, the first thing to say is that there is an importance to the trial beyond just whether it changes people's votes, because I think sometimes that gets lost, that the political impact in terms of sort of the electoral outcome is not the only factor in play here. So I think that a lot of pollsters are asking questions about how people view the verdict, how people view the trial, whether they were paying attention, whether they think it was fair. And those sort of, I think, form a sort of baseline of what people are taking away from this. And, you know, hopefully we can sort of inform people about A) what the actual facts of the trial were, and then also see how people are responding to that. And then there are a lot of different ways to sort of try to measure what the effect is. And that's just a really tricky thing to do.
Audie Cornish
00:11:58
So tricky that you wrote about it, which felt like a little bit of a warning sign when the pollster is like, tippity tap, tap tap, um, guess what, we don't know. But you brought up a couple of things, which is first, you said it's notoriously difficult to interpret something like this, meaning an event, a debate performance, or in this case, a conviction. Why?
'Ariel Edwards-Levy
00:12:21
Yeah. So there's a question that pollsters have sometimes asked that drives me a little bit insane. And I have been on this warpath for a little while now. But they say, okay, this happened or this might happen. Does that make you more likely to vote for this person? Less likely to vote for this person? Or not change your mind? And that is an impossible question to answer because that's not how people think. You know, if you are a strong partisan, are you going, "okay, well, I had a 90% chance of voting for Trump before, but now it's only an 83% chance. "
Audie Cornish
00:12:53
It's asking people to make this sort of like formal decision tree, also devoid of any other context.
'Ariel Edwards-Levy
00:13:00
'Yes. And just like foregrounding this one thing. So that overstates the importance. And then also people who are sort of political are going to use that to sort of reaffirm their support, because, you know, if you're somebody who is very anti-Trump and is never going to vote for him, you're not going to go, oh, well, it's not it's not going to have any effect, even if that's probably the case, because you were never going to vote for him in the first place, because you see this and you're thinking, this shows that he's not fit for the presidency, and you're going to say less likely.
Audie Cornish
00:13:28
But reaffirm to who? The pollster?
'Ariel Edwards-Levy
00:13:30
Yes. And that's –
Audie Cornish
00:13:31
Like people want to impress the pollster?
'Ariel Edwards-Levy
00:13:33
People – there's this sort of phenomenon that's called "partisan cheerleading," which is basically the same way that you sort of root on a sports team. When you're taking a poll, you want to answer in a way that sort of supports the side that you're on, which makes sense. But we see this pop up to some extent in everything from these sorts of questions to even questions about how people feel about the economy, where people's answers are A) genuinely informed by their politics. But also there's sort of a tendency to answer in line with your politics sometimes.
Audie Cornish
00:14:02
Because they want it to show up.
'Ariel Edwards-Levy
00:14:04
Yes. And but, you know, even just, you know, it's not sort of, you know, people are like, purposefully like making things up. It's, you know, if you're a Trump supporter and you're outraged and you're saying, well, this is going to make me more likely to support him, you're sort of reacting in a way where you're doubling down on your support, but that's not something that's going to show up as a change in the polls.
Audie Cornish
00:14:26
One interesting thing about this whole Trump trial, this one in particular, is that not as many people as we in the media think were actually even paying attention to it. Can you give me like, the numbers or sense of that?
'Ariel Edwards-Levy
00:14:40
Yeah. So I think, you know, we've consistently seen I mean, most people are aware there was some polling that came out at the end of last week that I think 80% of people said that they realized that there was a verdict. So it's not that this is, you know, something that nobody's paying –
Audie Cornish
00:14:53
That part of our job was achieved.
'Ariel Edwards-Levy
00:14:55
Yes.
Audie Cornish
00:14:55
They know there was a verdict
'Ariel Edwards-Levy
00:14:56
But, you know, it's closer to probably like a third of people who say they're paying very close attention. And there's still if you give people the option in a poll to say they're not sure how they feel about something or they haven't, they don't really have an opinion. You know, I think probably somewhere like a quarter or 30% of people are going to say, "oh, I haven't really decided how I feel about this verdict." And a lot of the people who aren't paying super close attention right now are exactly the kind of people whose opinions are less set. It's going to be younger people. It's going to be political independents who aren't so tied into politics. It's going to be the exact kind of people whose votes might theoretically be in play here. So, you know, we're starting to see the first wave of polling that actually is asking questions that can tell us a little bit more that are asking. There was one poll, that just came out over the weekend, where they actually went back and asked the same people who they'd surveyed earlier how they felt, which is a smart way to sort of measure the actual impact of this to some extent, but you're still not going to see an immediate effect necessarily. It's going to really matter how this plays out in the weeks and months to come.
Audie Cornish
00:16:03
Can you talk a little bit more about those people who aren't locked into a decision? Because the cynic in me is like, there's no more swing voters. There's no more people to be convinced. We're all just partisans in sheep's clothing making a decision. And it's like a numbers game at the end. I repeat, that's Audie's cynical position. It's it's completely meaningless and not based in any kind of fact.
'Ariel Edwards-Levy
00:16:28
This is sort of the paradox of modern politics and modern polling, I think, is that people are extremely polarized. Many people have their opinions made up. I think that's particularly the case in a scenario where both of the candidates are known quantities, and this election has played out in one form before. On the other hand, the margins may be so close that the small share of people who don't fall into that category matter. And one of the political divides I think we don't think or talk enough about sometimes is not, you know, Republican versus Democrat, but it's highly engaged versus less engaged because there are a lot of people who just don't engage in politics in the way that you or I, or probably people listening to this podcast engage with politics where they're just it's something that sort of exists on the periphery of their life. They're busy, they have other things going on, and it's just sort of a fundamentally different way of how much attention you're paying to a campaign very far before you vote and how much you're sort of thinking about that.
Audie Cornish
00:17:27
But the key is what you're saying is those people still do vote, maybe not as consistently, but it's not like they're completely out of the system.
'Ariel Edwards-Levy
00:17:33
Exactly. And that's something that is another sort of complicating factor is these people are less likely to vote. They're not necessarily going to turn out, but a lot of them will still cast a ballot.
Audie Cornish
00:17:45
So how confident would you be in any voting about the effect of the conviction on November?
'Ariel Edwards-Levy
00:17:54
I have no confidence that I could I could say right now, you know, there's this sort of idiom that gets trotted out a lot that pollsters love, which is that "it's a snapshot, not a prediction." And I think that's really true. Polls can tell us what people are thinking right now in sort of the broad contours of public opinion, which I think is deeply important to understanding how people are viewing the election what matters to people, how people think about politics. But polls can't tell us how people who haven't decided are going to decide. We can't poll for –
Audie Cornish
00:18:27
Which is literally what we're looking for them to do.
'Ariel Edwards-Levy
00:18:29
Yes, it's – we can only report what people's opinions are and –
Audie Cornish
00:18:34
In the moment
'Ariel Edwards-Levy
00:18:35
Exactly.
Audie Cornish
00:18:36
The last thing I want to tackle is understanding the state of the presidential race right now, because we hear that it's close. We hear that the Republican nominee to be, Donald Trump, is ahead by one point or several points. But in some states, but not in all states. When you have to explain this to your friends and family, what do you say?
'Ariel Edwards-Levy
00:18:59
The first thing I'm going to say is that the phrase "ahead by one point" should be completely knocked out of everybody's vocabulary.
Audie Cornish
00:19:05
Thank you! Is that not within the margin of error?
'Ariel Edwards-Levy
00:19:07
It is so far within the margin of error.
Audie Cornish
00:19:09
It's so weird to hear that. Like someone's up by 1 or 2, I'm like uh, okay?
'Ariel Edwards-Levy
00:19:13
Up by one does not mean anything. You know, in one poll, changed by 1% between two polls does not mean very much. You know, these things just – polls are not a precision instrument. That's not what they're good at. What they're good at is sort of saying, let me step outside the people who I talk to and the voices I'm hearing and get a broader sense of sort of what people think about things, but it is not a precise estimate. And what I would say about the state of the race right now is it's close. It looks like it's close.
Audie Cornish
00:19:41
No Ariel, not good enough. What do we say here at CNN? What's the official like figure we're saying, and how are we deriving it?
'Ariel Edwards-Levy
00:19:48
What we are saying about the state of the race nationally is that there is no clear leader. And we are deriving that by looking at a lot of the polls, which have all consistently shown it very close.
Audie Cornish
00:19:59
But do you look at certain swing states and say, "well, that's where it looks like someone has an advantage?"
'Ariel Edwards-Levy
00:20:04
Yeah, we've looked at swing states. A lot of the swing states also look very close. And, you know, especially some of the sort of, you know, Wisconsin, Michigan, Pennsylvania – I'd say more, you know, Wisconsin and Pennsylvania – they look they look close. And I think that's –
Audie Cornish
00:20:18
Because I hear not infrequently "Trump is in the lead." Not that not just that he's competitive.
'Ariel Edwards-Levy
00:20:24
Yeah.
Audie Cornish
00:20:24
But that he is leading.
'Ariel Edwards-Levy
00:20:25
And I mean, there are some states where polls are showing Trump ahead.
Audie Cornish
00:20:28
By one or by six.
'Ariel Edwards-Levy
00:20:30
By more, by enough of a margin that it's a little bit more than –
Audie Cornish
00:20:33
That you respect.
'Ariel Edwards-Levy
00:20:34
Yes.
Audie Cornish
00:20:35
Okay.
'Ariel Edwards-Levy
00:20:35
But I you know, that's a couple of states and things – there's so much room for things to change. I think that the meaningful thing right now is it looks like a close race. And I think polling. And election polling – the more you understand what it can't tell you, the more that what it does tell you is meaningful. And right now, what it's telling us is that it's a close race. It's telling us that there's sort of a meaningful pocket of people – it's a minority – but there's a meaningful pocket of people who are not super happy with either candidate, who have maybe not been tuning in super closely.
Audie Cornish
00:21:07
And that's the combination that's most important, right? Not paying attention as closely and the like unhappy with both.
'Ariel Edwards-Levy
00:21:16
Yes. And you know, I think there's a lot of unhappiness right now. It's whether people are looking at the last four years and whether they looking back at the Trump administration and whether what they're looking forward to and what they understand as the promises that each candidate has made and what it's going to look like if Biden has a second term or if Trump comes back into power. And I think just looking at that and polls can tell us sort of how people are seeing this race. But I don't think polling right now can tell us this one candidate is in, you know, absolute position where if the election were held today, this person would win.
Audie Cornish
00:21:52
'In the end, what do you think are the lessons that pollsters came away with from 2016 and 2020 and even 2022, right, when the so-called red wave didn't really materialize of Republicans just sort of overtaking everything in the midterm elections?
'Ariel Edwards-Levy
00:22:09
I mean, I will say, just quickly, that our polling in 2022, and a lot of the public polling in 2022 was quite good.
Audie Cornish
00:22:16
Say more.
'Ariel Edwards-Levy
00:22:17
You know, our polling, our margins –
Audie Cornish
00:22:19
So was it just bad punditing? What was happening?
'Ariel Edwards-Levy
00:22:21
You know, I think a lot of people got gun shy where they had seen two presidential elections that ended up being more Republican than the polling, you know, may have shown. And so they assumed – but there was polling that showed, you know, Democratic senate candidates doing pretty well, and that ended up materializing. And, you know, it showed –
Audie Cornish
00:22:42
So it sounds like you're saying we shouldn't conflate the marketing of information with the information itself, because
'Ariel Edwards-Levy
00:22:47
Yes
Audie Cornish
00:22:48
Everywhere you heard this idea that Republicans were going to do so well, so much that had it had earned a preemptive nickname.
'Ariel Edwards-Levy
00:22:54
Yes. And I think what polling also tells us is the sort of example that I go back to that's not just about the sort of margin between candidates is the polling on Roe is something where a lot of times there's huge variation.
Audie Cornish
00:23:06
This is the abortion ruling.
'Ariel Edwards-Levy
00:23:07
Yes. The polling on Roe has been so consistent since the overturn. You know, around 60% of the country was opposed to the decision to overturn Roe versus Wade. And that has been consistent across polls, across time. And I don't think our public discourse would be better if we didn't have a way to point to say, "this is what the public feels about this," and this is the energy that has persisted, and that helps us explain what happened in 2022. And I think that's sort of the value of what polling can tell us is, how are people thinking about politics? What are the sort of broad contours of public opinion? And that goes beyond just, you know, who might be ahead in one particular race?
Audie Cornish
00:23:49
Okay, you gave me the good footnote about being technically right, but what were the lessons?
'Ariel Edwards-Levy
00:23:53
I think the main lessons were just that we need to meet people where they are in terms of polls, and we need to expand what we're doing. And there is not there used to be this sort of idea of a gold standard poll. This is the right way of doing things. And I think also, you know, I hope that a lot of us knew this lesson beforehand, but I think it made it even more important to communicate the limitations of what polling can tell us, which is that, you know, I see polls reported out to a 10th of a point sometimes, and it drives me a little insane because we don't have that level of precision. Polls are not meant to tell us exactly to a 10th of a percent what people think. And I think the more we can communicate that, especially when you're looking at events in the future among a population of voters, that doesn't exist yet, a little humility is good. And sort of thinking about the scope of what you can learn is really important.
Audie Cornish
00:24:47
'Well, Ariel Edwards-Levy, thank you so much for hashing this out with us.
'Ariel Edwards-Levy
00:24:51
Thank you so much for having me.
Audie Cornish
00:24:52
Can we have you back later to harass you about, various polls as they come out?
'Ariel Edwards-Levy
00:24:56
Please do.
Audie Cornish
00:24:57
'Okay. That's all for today, and we especially want to thank our listener Mary for calling in with this question, which launched this assignment. We appreciate you, and it's a good time to mention that we are looking for you guys to give us more assignments. You can call. You can text. You can send a voice memo. The number is (202) 854-8802. It makes us and the show better. This episode of The Assignment, a production of CNN Audio, was produced by Carla Javier. Matt Martinez is our senior producer and the executive producer of CNN Audio is Steve Lickteig. We got support from Haley Thomas, Alex Manasseri, Robert Mathers, Jon Dianora, Leni Steinhardt, Jamus Andrest, Nichole Pesaru, and Lisa Namerow. Special thanks, as always to Katie Hinman, I'm Audie Cornish, and thank you for listening.