← Harvard Data Science Review Podcast
Harvard Data Science Review Podcast
Digesting 2024 Election Polls: How the Media Reports and Decodes the Numbers
00:41:04
Loading summary
Transcript61 lines
- [00:02]
Liberty Vittert
Hello and welcome to the Harvard Data Science Review podcast. I'm Liberty Vittert, the feature editor of the Harvard Data Science Review, and along with me is my co host and editor in chief, Shaoli Meng. Often deemed the most important election in recent history, the upcoming US Presidential race arrives at a time when many Americans are grappling with questions about poll accuracy and data reliability. This month we're joined by award winning journalist Leland Vittert, anchor of the primetime show On Balance with Leland Vittert, and Colby hall, founder of Mediaite, for an in depth look at what past elections can teach us about 2024. In a world where data is everywhere, what can we really trust? We'll discuss which data points are essential, how technology is reshaping polling, and which voter demographics could sway the results in very unexpected ways. Join us as we dive into these pressing questions and more on the Harvard Data Science Review podcast. You know, there is so much data now. You know, it's overwhelming. There's so many models, there's so many polls. There's so, I mean, there's just so much information and so much detailed information for both of you. Leon, we'll start with you. How do you decide what to give viewers and what not to?
- [01:23]
Leland Vittert
Thankfully, I don't have to. I have people who are a lot smarter than me, like Scott Tranter at ddhq, who we work with, like a number of the pollsters that we interview who can bring us that. I think the other thing that television news learned out of 2016 is that not all data is created equal. There's good data and there's bad data, but also that the data isn't always right. The polls are only as good as the models, which is something that before 2016, I think a lot of political reporters even really didn't focus on was how important the turnout models were. And then probably the most reliable thing now has become the averages. The DDHQ average in particular is something that we really focus on because it averages for the polls that are intentionally skewed for Democrats, the polls that are intentionally skewed for Republicans, and gives you a better idea, just sort of the direction of things over time rather than just focusing on this snapshot of this one poll at this one moment.
- [02:23]
Liberty Vittert
Colby, what about you?
- [02:24]
Colby Hall
Yeah, Leland summed that up really well. You know, at media, it's a little different because we sort of COVID the media. And to Leland's point, there's so much data out there that to find the signal and all the noise is the challenge. And in fact, there's so much polling data that right now, the stuff that's most interesting is the aggregate. Right. The polling averages, and to that end, the ranking or the assessment of which polls are more accurate historically than others. So a new poll doesn't merit coverage. I think trending is way more interesting data for us than the latest outlier poll. That may end up being meaningless.
- [03:05]
Leland Vittert
I'll give one more thought for Colby, is that for the cable networks that want to have an opinion. Right. It's very easy to find data now that cheerleads one side or cheerleads the other side. And Colby, I know you guys write about that. The thing I find most interesting, and now that we can get to, is the why. Right. Is we now have data on what young black females in Georgia are doing. We now have. We're now at the point of the race where we have data about what young Latino males in Nevada are feeling. And that actually sort of gives you and I think gives the viewer or the reader a much deeper understanding of the race, because those states and those demographics are what decide the election. As fun as the horse race is nationally, to talk about it, it's completely meaningless. Just ask President Hillary Clinton. So the idea now, and I think you're seeing how data is playing such a bigger role in news coverage because we have the data to talk about not just the horse race, but actually the voters that are going to end up deciding the election and how they're trending and how they're voting.
- [04:10]
Colby Hall
Right. To that the data helps you uncover the issues. Right. So it's less about the polling numbers and the issues, to Leland's point. And also partisan outlets, most of cable news these days, shape numbers to help push or pull voters. That helps their narrative and support the candidate they want, which is a real flaw.
- [04:32]
Liberty Vittert
What's been the most egregious that you've seen in media of people choosing numbers that fit their narrative?
- [04:38]
Colby Hall
Oh, you know, I think in. Was it 2012 or maybe 2016, there was this trend that was the spirit of the trend was actually accurate, but the practice was wrong. But there was a Twitter account named Unskewed Polls, and this is an individual who actually sort of was right in theory, that the traditional polling outlets had really sort of got their stuff wrong. But I think he overestimated. He flatly predicted that Romney was going to win by a big margin. And this guy really sort of went way out of his way. And, you know, after he was sort of embarrassed, kind of deleted his account, we never heard from him again. But he begat a couple of other sort of non traditional or sort of like outlier pollsters. Trafalgar as an example, which has actually been sort of a perfect balance to what people have. Well, at least in 2016 and 20, polling data that was traditional. But taking a step back, I think so much has changed in the way that we communicate with one another that the polling methodology has been rendered not obsolete, but almost outdated. And we need to look at polling as a snapshot of data that we know about and not a clear predictor of what will actually happen, because it's not a perfect world and things change.
- [05:57]
Shaoli Meng
Kobi, your point is extremely well taken. For Harvard Data Science Review, I think we published a few issues ago, a political scientist wrote about how the polling paradigm needs to be shifted. Right. Because of all these issues you raised. Now, as statisticians, we love to hear people talk about averaging, because that's what we do. But on the other hand, averaging only goes so far. But what if there are something fundamentally biased? The data is a biased. This leads to the question about back in 2016, the majority pools are going one way. No matter how you average, you know you're going to get president. You know, Clinton, right? And at the time there was debates or afterwards, people talk about the phenomenon of, you know, shy Trump voters. So my question for both of you is this year's election has become so unprecedented in many ways. All the surprises. Do you guys see any potential errors that we are just not even anticipating your instinct as a journalist?
- [07:02]
Leland Vittert
I think we're undercounting the young single white women that were not as high of a propensity voter. Certainly in 2016, they stayed home. There weren't a lot who were all that inspired by Hillary Clinton. They came out in droves in 2022 on the abortion issue. I don't know how that's being built into the polling models. I think there's crimes of omission and commission. The polling errors we saw in 2016 I think were mostly of omission, right, because the models were off by just a little bit of high school educated or non college educated whites. And that was really the difference in the three states. So I think the lack of understanding by the viewer that polls are an idea of where the electorate is. They're not precise down to decimal points. We know the polling is going to be wrong. So the groups I would look at, single white females, young African American males are gonna be really important. And then the break in the Latino vote in Nevada and Pennsylvania, there's Something really interesting going on there in their move towards Donald Trump. And we don't know how those groups are gonna offset each other. And that's, I think, where you're seeing at least the fear among Democrats is that they, so far, Colby, and I think you guys have covered this on Mediaite. There's a real fear right now among Democrats that the single white female vote for Harris isn't gonna balance out the Latino vote and the African American vote that's moving to Trump.
- [08:35]
Colby Hall
Yeah, I mean, Leland sort of summed it up wonderfully there. It's interesting you talk about the shy Trump voter and I think that was true especially in 2016 and 2020. I was back in suburban Kansas City for my 40th high school reunion. I was surprised to see a lot more Harris Waltz banners and stuff and very little Trump signage, which sort of goes against the polling that came from that district. I guess what I'm trying to say is that I feel like the extreme Trump flag wearing MAGA set has almost not shrunk, but it's less loud and proud. And there are more Republican voters who may not personally care for Trump, but are going to vote for him regardless and are okay with it. And they say, yeah, he's loudmouth and I don't like what he says and my wife thinks he's a jerk, but I agree with his policies. And I found a lot of my Republican friends were open about it and of course I'm friends with him. So that may be a bias there. To the other issue, I think two things. If your campaign, Harris campaign, you're really hoping that the white women get out to vote based on reproductive rights. That's the thing that they're hoping will sort of end the trend of Trump as seen in polling. If you are the Harris campaign, you're also freaking out about youth votes, especially I think female youth vote surrounding the Israel Hamas war. A good friend of mine was canvassing in Pennsylvania and she was surprised at how many youth who identified as Democrats but were not going to vote for Harris because of the way that they have handled Israel Palestine war. So I really think that especially in Michigan, that the pro Palestinian youth vote could end up really being the thing that takes the Democrats down. And also the Hispanic vote, you know, that's a conservative set of people. If they turn around and go Republican, Katie, bar the door, I think it's game over. And so I think Leland's spot on on that as well.
- [10:43]
Leland Vittert
The other thing I would watch for, you talked about the silent shy Trump voter There's an indication that among rural white females, and this is anecdotal, of course, cuz it's very hard to find the shy voter that is a shy Harris voter, especially older rural white females who would typically be seen as in deep red counties. The second point I would add to what Colby said that I think is really smart and we always underestimate trends and then later look back and think about them. But the realignment of the electorate that started in 2016 with rural traditional value Americans who were Democratic voters moving to Republican and white college educated upper income Romney Republicans moving towards Clinton, at least out on the trail, I have seen that trend just keep getting exacerbated over and over and over again and growing. We'll know after this election, I think, how far that realignment's gone. But in terms of what you guys say about the danger in the data, if we underestimate how far that trend has gone, that will skew things pretty significantly.
- [11:57]
Liberty Vittert
The modeling is so difficult and pollsters have really gotten the blame. Everyone talks about how pollsters were so wrong in 2016 and 2020, but pollsters.
- [12:08]
Leland Vittert
Want to get it right. That's how they make money.
- [12:11]
Liberty Vittert
But you have to give it that. The media is also partly to blame here too, because they choose to report what the pollsters are saying. And I see you shaking your head at me.
- [12:22]
Leland Vittert
Lucky we have no blame whatsoever here.
- [12:25]
Liberty Vittert
Right. But after 2016 and 2020, are there any sort of lessons you all learned as journalists as to how you choose to report this information to the public? And Leland, you're not allowed to use the whole, you use people that are smarter than you to tell people things like you all pick the guests that you have come on and you pick which pollsters you have come on. Are there lessons Learned here from 2016 and 2020 for how the media, I mean, people wouldn't know what the pollsters are saying without the media. So are there lessons learned from the media here?
- [12:54]
Colby Hall
Well, I'll start. Nate Silver, he was sort of like the OG polling guru. And the thing that really set him apart was that his formula wisely included, you know, a ranking or grade of past poll corrections. Right. And so it was just another meta layer that sort of weighted polling data based on past predictions. And in fact, what I think is one of the most interesting things that we see, the New York Times latest polling data. They, and I don't know if this is, they're trying to be kind to Kamala Harris or Donald Trump, but they not only tell you where it's trending and where they stand. But they also add, if the polling is wrong, as it was in 2020, it looks like Trump could win in a landslide, you know, at least in terms of the Electoral College. If the polling is wrong in the way that it was in 2020, 20, in the midterms, Kamala Harris wins more closely. And so we see these media outlets that just keep piling on layers of data. Part of it, because I think more data is good, but I also think it's kind of it protects them in some respects. Taking a step back, I think we look at pollsters as telling us confidently what will happen. And really, like any statistician and direct marketer, it's the best reasonable expectation guess that can come in that moment. So I think we as a society need to look, put a little bit less value in polls and maybe see it as like one feature in a larger narrative. I don't think that's going to happen anytime soon. But we get in trouble when we treat polls as gospel, when really it's data that we analyze and make predictions from, from smart people.
- [14:48]
Leland Vittert
The humility factor, I think, is really important, which is hard for anybody on television because humility, as Colby knows from covering people on television, is a difficult quality to find.
- [14:59]
Colby Hall
Yeah.
- [15:00]
Leland Vittert
I would add one more. I think it is made all the more important the actual reporting, because it's so easy to look at the data and have data, to talk about data, data, data, data, data. And it almost comes at the price. And I say that because it's cheaper as a news organization to talk about data than it is to send crews out into the field and interview people and go get a sense of what's happening in black barber shops and go get a sense of what's happening in coffee shops where soccer moms all come after they drop their kids off to school. And what's happening in union halls and going to trivia night at bars or pool halls and figure out what the steel workers are talking about in various towns. And that's like the third layer of reporting. It's the most difficult to do and it requires the most nuance. And the thing that matters the most is going out and talking to voters individually to get a feel for the way the wins are. And there were a lot of reporters who traveled with Trump in 2016 and who were at our rallies. And Colby, I think you remember some of them who would come back and say there is something different going on in 2016 with Trump.
- [16:13]
Colby Hall
Right.
- [16:14]
Leland Vittert
And they did that because they were on the trail they were out talking to swing state voters. They weren't in Washington or New York or in their home studios in the Hamptons pontificating about, well, this poll says this and this data says that and this subset of that says this.
- [16:28]
Colby Hall
Yeah, I think a lot of media personalities, especially in the partisan opinion media realm, tend to look at viewers and voters as numbers and subgroups. In fact, an old friend of mine used to be head of research at HBO in the 90s and early aughts and he posited from his market research group that you could sum up demographic makeup, psychographic makeup into 12 distinct groups. And this was late 90s early aughts. And I just thought of that now because that seems like a really antiquated idea. I think one of the things that's happened with like the digital revolution and this sort of targeted marketing and one to one and you know, siloed entertainment, it's not just 12Americans anymore. So Leland's point of like going to black barber shops or you know, Hispanics or union halls, that's. But those are but three examples of the dozens of demographic makeups that you can probably subset together. But again, it goes to like, you know, the statistical polling model, and this is where you all are the experts, seems to have not necessarily come up with how Americans are living differently and identify differently psychographically, demographically. We're in a brave new world.
- [17:48]
Shaoli Meng
Well, I guess, you know, obviously it's most important to collect data firsthand. Right? By going to these places, you see yourself instead of getting data which being processed in many ways, then it has all kinds of bias embedded into it. So again, the data quality is so important. I want to ask both of you another question about reporting these early votings. As you all know, this issue become kind of increasingly important, at least on people's mind. And as far as I know that by now is over at least like 20 million voters already have voted, probably even more. You guys know much more up to date stat. When you think about reported early votings, do you worry about the impact of reporting itself? Is there some kind of self fulfilling prophecy here? And do people feel like, oh, a lot of people vote, I don't need to go out and vote, or a lot of people vote, I should have voted. More from your perspective, your experience, is there anything you worry about in kind of reporting those early votings?
- [18:50]
Leland Vittert
Absolutely not. And to your point, almost a third of the number of Georgians that are going to vote have already voted. And if we learned One thing from 2020 it was the need to really explain to viewers and to the public what was happening and why it was happening. And the more transparency you can give, the better. So I think explaining the numbers as they're happening in real time and pulling back the curtain for the viewer in what is happening provides a lot more faith in our elections, and especially in an election where there is so little faith and there is so much interest and self interest in not fighting for an outcome, but questioning the result. And I say this from both sides, the more transparency there can be, the better. And you know what? If somebody decides because so many people have voted, I'm not going to go vote. Well, shame on you, then. Stop complaining about the result.
- [19:54]
Colby Hall
Yeah, I agree with Leland. I will say it's an interesting ethical question of how you specifically report on early voting. I think total number of votes is totally fair game, but I think there's a question whether or not outlets should report that, say, you know, a third of the all voters in Georgia voted, but we've learned that 60% of them are registered Republicans. That could have an impact.
- [20:17]
Leland Vittert
It's on the Secretary of State's website. Colby, what? Are you supposed to not report publicly available data?
- [20:22]
Colby Hall
Well, I'm just saying that maybe the Secretary of State shouldn't put that out because I think that it. It raises an ethical question. Does it keep people from voting? I'm pro transparency. I think we should be open with everything. And if you're not going to vote because of what you saw or heard somewhere else, you should vote anyway.
- [20:41]
Liberty Vittert
Yeah. And I mean, Lucky, to Colby's point, it's a secret ballot in the United States, so there's a reason they're not reporting. We have X amount of votes for Trump and X amount of votes for Harris. I mean, technically speaking, they could already be counting that and reporting that from the Secretary of State's office. So why do they even report how many early votes or how many Republicans or Democrat votes they have?
- [21:02]
Leland Vittert
Right, But I think there's a nuance in the question. I think the job of the media is to report. There's very, very few things that if we know it, we should not report it. Right. I'm old enough to remember, barely the quaintness, okay, of when you didn't talk about polls on election day because God forbid you influence. Okay, please, we're past that. But if you want to say we're only gonna report aggregate number of ballots returned or requested in early voting, that's a political decision. That's not a decision for the media. To make.
- [21:42]
Colby Hall
Yes, but exit polling doesn't get reported until polls closed. And it may be quaint in the way that you just said, maybe that goes away. I think early voting raises questions that we're still trying to suss out. And I agree with everything that Leland just says, that the media's job is to report the knowledge that's out there, other than, like secrets that put American troops in harm's way or elected officials doing secret trips like that obviously is a whole different thing. More transparency is better. But if the media report that Harris is up big in early voting in Pennsylvania a week before the election day, I think a lot of people would be upset about that, and understandably so. So they're not doing that because Pennsylvania doesn't vote early votes until all the votes are cast. But again, it's an interesting ethical question.
- [22:36]
Shaoli Meng
Just a quick follow up question. I got intrigued when you said it's not a media question, it's a political question. Who make these decisions for your station?
- [22:47]
Leland Vittert
It's a great question. You know, publicly available information is fair game to report, right? I guess at some level, there's an editorial decision made on it without getting too in the weeds. Like in any organization, the bigger the decision, the higher it goes. So, for example, if I have a scoop with some really salacious information or explosive information, I don't just go on TV and say it. It goes through my executive producer, and he and I make a decision about, hey, do we need to talk to management about this? And it sort of goes up the ranks, obviously trying to avoid what 60 Minutes did in 2004 when they falsely reported the George W. Bush letter. And I see Colby smiling because he remembers that. So that's always what you want to do, is you want to make sure you get it right. That's number one. Number two, there are cases in which, as Colby said, and it's happened to me, where we had information that was gonna put lives at risk or put American national security at risk. And then that's a really delicate and important decision that you make with editors and executives and depending on how big of a story it is, how high it goes up. I think what's interesting in this context is the political decision of what numbers to release. Right. Because We've learned from 2020 that some states start counting before election Day. Other states don't open anything until after the polls have all closed across the state, so early votes aren't tallied. And those are all political decisions made at the state level. Politically.
- [24:25]
Colby Hall
I mean, one of the Most interesting examples of this was in 2020 at Fox News, which was very controversial because they were the first to call Arizona. And as it turns out, they got the call right at the time. It looked like they got the call early. And, you know, to Fox News credit at that time, the decision desk there was given the autonomy to make that call. And a lot of stuff has emerged that there was a lot of teeth gnashing and saying that this isn't smart. But ultimately, the data scientists at Fox News decision desk ran the numbers and they were confident and they were proven to be true by a number of 10,000. You could argue maybe that wasn't statistically enough to go with it. Every network is different. But those big decisions, there are nods of ascension to make sure that there's not legal liability because you make a mistake. 60 Minutes made a mistake in how they chose to promote a Kamala Harris interview by airing a clip, an answer that didn't air in the actual show. And as a result, their reputation is pretty damaged right now. So the reason you go through the nods of attention in the management is that you have other people that can sign off on another set of eyes to make sure that you're not making a big mistake that's going to do reputational harm.
- [25:43]
Leland Vittert
Colby brought up a great point that I'll just sort of let your listeners in on the decision to call elections this time around because there was so much focus on Fox in 2020 and because for the first time in recent memory, we are going backwards in how we're doing things in America. You know, we went from the Pony Express to trains to airmail to Federal Express to Amazon prime in two hours. And we now are going backwards in how we count ballots in the time it takes to get accurate and clear information. In elections, that's a different issue. But it's going to be fascinating to watch the networks and hear the reporting afterwards on when they make race calls, because you're going to have some networks calling some states and other networks calling other states. I think this is the first time, Colby, we're going to see a huge deviation among the networks of when they're calling things and when they're comfortable calling things.
- [26:42]
Colby Hall
I think I saw somewhere that the over under on the the winner will be declared is Saturday. And that's what happened last time. Biden wasn't declared winner until Saturday. I think it might be later because there will be an abundance of caution and people get frustrated because there was a time where you could wake up the next morning or you could stay up till midnight and find out who was your next president. I think it's more important that these networks get it right. And the way that the voting is going to looks like. And especially Pennsylvania, Michigan, Arizona, I mean, it's. We're talking about tens of thousands of votes. And so it'll be interesting.
- [27:20]
Liberty Vittert
Colby, you already kind of alluded to it, that news organizations have. I mean, you put it in a nicer way. I'm going to put it in the way of they're protecting themselves by layering on caveats of data of, well, if it goes this way in 2020 or this way in 2022, just sort of protect themselves. Have there been other things that you have seen really change? I mean, you've covered the media, you've broken all the stories. Have you seen other things really change, especially in sort of election reporting, things that have really changed over the last decade and how news organizations have done it? And I know Lockie just alluded to the idea that we're gonna see different news organizations report on different states at different times, but what else have you sort of seen really change for the better or for the worse?
- [28:05]
Colby Hall
The coverage of polling data in campaigns, the way that's changed has been a symptom of an enormous shift in the last 15, 20 years. And that is sort of a rise of partisan media, opinion media and the traditional accepted standards of journalism is almost a quaint idea now for a variety of reasons. A, it's not really economically viable. It's really hard to be a centrist and to be fact based when people want point of view. And as it comes to political coverage, regardless of the outlet or the outcome, rather, I'd say 50% of Americans are going to be gobsmacked by the result because they've been tuning into outlets that have led them to believe that the impossible actually happened. And I think that's really dangerous. And no one's talking about that. People are tuning into what they want to hear and being given the information solely because they're being monetized. And I don't think we've really figured out. I don't think it's something you can regulate. I think it's something we're going to have to work through and it's probably going to get worse before it gets better. Sorry to Pratt along, but it's a complicated thing.
- [29:18]
Shaoli Meng
I certainly would say that is something that we all worry. We should worry because we need a fact, right? Especially for those who do data science. I mean, the whole idea of Data science is, you have evidence, use the data. It's not just all your opinions or your ideology. But unfortunately these days those things are just so strong, interesting for this conversation, that we haven't talked about generative AI yet, because that's a topic, obviously everybody's mind as well. I want to ask you a question along that line. We understand that lots of your reporting comes from your knowledge, your common sense, human understanding, your relationship with others based on data. And now these days, these generative AIs can help at least something I've seen. It was kind of gathering information faster and helping you to summarize factors faster. And sometimes they're hallucinating. And so how are you navigating that world? How do you feel like those technologies, well, or maybe already have impact on your reporting for, for better or for worse. And as a journalist, how this advanced technology that will help you or in some sense might be hurting you?
- [30:34]
Leland Vittert
You know, Colby covers this a lot more because he covers the media and the times. The media falls for things that are fake. And it's a great point because not only do you have now opinion journalists, you never had to before ask effectively, is this real? If you saw a video or a picture or a headline or whatever. And I'm thinking about a lot of right leaning journalists who fell for a fake headline from the Atlantic that they all tweeted about and everybody got upset about and even politicians got upset about. It turns out not to be true. Fine. So I think that's number one, which is you now have to start asking yourself before you report something, is what I'm seeing even real? Is this photograph or is this video even real? That's number one. And number two, I think the more there is a gender of AI and the more we have to step back and it's hard in a 24 hour instantaneous news environment. You know, we used to talk about the news cycle as a 24 hour environment. Now it's a 24 minute environment because of social media. And I think it's all the more important for credibility's sake just to step back, you know, I mean, I can't tell you how many times at noon we think this is gonna be the major story of the day. And by 7pm it's history, it's gone. And Colby knows this better than anything because he watches way too much cable news.
- [31:55]
Colby Hall
No, not enough. I love it. To that point, I've said a lot that using the parlance of some of the younger staff, when every story is lit, no stories are lit, meaning if everything is a bombshell and if everything's over the top and this is the game changer, this is the big thing, then nothing is. And you have a media ecosystem that really sort of looks for stuff that is unique and scarce and something that is like game changing content, and that's what gets serviced. But back to the AI question, you know, getting to Leland's point, it sort of, I think that underlines what I'd said earlier, that, like, we're increasingly untethered from reality because stuff. You consume news, and your first question is, can I believe my eyes? Right? And that's just sort of the deep fake or cheap fake or whatever it may be. Personally, I really sort of stay away from AI because I feel like I'm a writer. The last thing I want to do is cede any of my sort of thoughts and talents to a supercomputer that's crawling a ton of info and giving me what I should know. I don't even use it as research. I use Grammarly, which is effectively AI that will spot grammar problems, that doesn't change it for me, but highlights them. But other than that, I stay away because I feel that that is a slippery slope. And I know that there are a lot of content farms and that's the new thing that you worry about, the singularity. There's a lot of outlets that are not using writers, they're just using editors to have ChatGPT and AI produce content, which short term may save you some cash, long term may have much greater expense. And I think at the end of the day, what's valuable in the thing that you're really selling is a point of view and something that's unique. And it's tough for AI to at this point, really nail a unique human voice. Give it a minute and it'll get there. But AI is something that we're. You know, I don't think it really has affected this election too much. I think it's a future concern that we need to get our heads around.
- [34:09]
Leland Vittert
I think it makes what Colby does and what I try to do all the more valuable because when every. When, when there's content farms just turned churning out content, having a perspective that's been thought about and kicked around and that the reader or viewer goes, I hadn't thought of that, or I didn't know that, or I hadn't looked at it in that way, AI can't do that. That's still unique to the human brain. So the plethora of content makes good content Even more valuable.
- [34:39]
Colby Hall
Well said.
- [34:40]
Liberty Vittert
To that point, I had written an op ed and for a different reason. I needed to sort of go more deeply into it. And it was a niche topic. And I asked ChatGPT to go more deeply into it. And I was reading what ChatGPT had produced, and one line I was like, oh, wow, that's a really good. Like, that's a really smart line. Another line I was like, that's really cheesy and stupid. And all of a sudden I was like, both those lines look, like, really familiar. And I looked back in my bed and copied and pasted it. So clearly I had some good stuff and some bad stuff for my own person.
- [35:13]
Colby Hall
There it goes. That's hilarious.
- [35:16]
Shaoli Meng
That's a good one. That's a very good one.
- [35:19]
Liberty Vittert
So we have our magic wand question, and it's a two parter this time for both of you. The first part of the magic wand question is if you could wave your magic wand and you could convince anyone still alive today to run in the next presidential election for one for Democrat and one for Republican so that they could win not who you personally would want to win, but so that they could win, what would be the matchup from each party? And the second part of the magic wand question is, if you could have anyone from history run for president, who would it be? Colby, I'll start with you.
- [36:00]
Colby Hall
I'll go first. Yeah. I am a big fan historically of Nelson Rockefeller. I love the idea of a Rockefeller Republican. Perhaps that's because my father was a proud Rockefeller Republican. And, you know, I guess if he were to be in the race against someone who I thought was similarly kind of commonsensical, but different enough to make it sort of interesting, this is going to be a deep cut. Paul Tsongas, because I thought Paul Tsongas was not a good candidate because he wasn't terribly electable, because he thought it was a good idea to be videotaped swimming butterfly in a Speedo. But other than that, Paul Songis was kind of like a normal, like, old guy. So my magic wand race would be Nelson Rockefeller versus Paul Saugus. And that's probably weird enough to lose a few viewers.
- [36:54]
Leland Vittert
Yeah, I'm thinking, I'm hearing you now and thinking about it later.
- [36:58]
Liberty Vittert
And who's the one person from history?
- [37:01]
Colby Hall
You know, it's gonna be cliche, but I grew up a dorkily enormous Abe Lincoln fan. And Abe Lincoln was to me, like the major cool dude because he freed the slaves. And I was like, so, you know, I'm sure There's better answers, but put on the spot. If I had to be, Abe Lincoln would be the safe bet there, because universally he was a major dude.
- [37:27]
Leland Vittert
I'm still thinking about Tsongas and Ospito, so I may have forgotten exactly the nuances of the questions.
- [37:32]
Liberty Vittert
Liberty, briefly, person from history that you'd want to have be president, and then two people who are alive today, one who would run in the next race is a Republican, one is a Democrat who you think would make.
- [37:45]
Leland Vittert
Rockefeller and Tsongas are both dead. I know Rockefeller's dead.
- [37:48]
Colby Hall
Colby, if they have to be alive today, then I would go with Youngblood. I would go with Pete Buttigieg and Nikki Haley, because they both seem to be smart, they both seem to be happy in their own, comfortable in their own skin. More importantly, they're young and they represent a new point of view. And I desperately think we need to get rid of the Gerontocracy. So I would go with Mayor Pete and Nikki Haley kind of top line, but live today, that's my answer.
- [38:16]
Leland Vittert
Yeah, I think those are probably pretty good representations of more centrist folks. I think, from a Democratic perspective, I think it'd be really interesting to see Nikki Haley and Gina Raimondo, of probably equal intellectual weight, run from history. I would say Eisenhower, not just because our grandfather liberties and my grandfather worked for him and admired him greatly and was close to him. And rarely did our grandfather speak so highly of someone as he spoke of Eisenhower, But I think he had something that almost no politician has, which is the willingness to do things and share or give away the credit so long as they got done, which is a quality completely missing in American politics. Hailey Raimondo. I think it'd be really interesting. I think Marco Rubio, I'd also add to that list of somebody who's deeply principled, but also pragmatic, that you could have. The question I think you're asking, Liberty, is who could become president? That most of the country would say, I'm not necessarily sure I love their policies, but at the end of the four years, this person will have done their best to make the country better for everyone. The closer you get to the center, the more people feel that.
- [39:37]
Shaoli Meng
Thank you. Thank you. And, well, we will follow up with you four years from now to see if those people are actually running. But seriously, thank you for this great conversation. You have given our listeners a lot to think about, and not just about election, but about reporting, about facts, about media. There are lots of things I know on many, many people's mind but at this moment, probably the most important thing for all of us to emphasize. I'm going to quote a green line from a Secretary of State from Minnesota. We did an interview with him and his call for the general public is to ensure there will be high turnout but low drama. And I think that is probably very important for all of us.
- [40:24]
Colby Hall
Well said.
- [40:25]
Shaoli Meng
Yeah, that's well said.
- [40:29]
Liberty Vittert
Thank you for listening to this week's episode of the Harvard Data Science Review podcast. To stay updated with all things HDSR, you can visit our website at HDSR mitpress mit.edu or follow us on x instagramhdsr. A special thanks to our executive producer Rebecca McLeod and producers Tina, Toby Mack, Ariane Frank, Gavin Yang and Belle Riley. If you like this podcast, please leave us a review on Spotify, Apple or wherever you get your podcasts. This has been the Harvard Data Science Review. Everything Data Science and Data Science for Everyone.