National political podcast host Jody Avirgan and his wife recently visited Juneau, marking Alaska as the 50th state to which he’s traveled.
Avirgan, who hosts the FiveThirtyEight sports, culture and politics podcast called “What’s the Point,” accepted an invitation from Andrew Kitchenman, KTOO and Alaska Public Radio Network politics reporter, to discuss the Alaskan political climate on Friday, August 19.
Andrew Kitchenman: I’m Andrew Kitchenman, the state government and politics reporter for KTOO and the Alaska Public Radio Network, in the KTOO studio today with Jody Avirgan, who does the podcast “What’s the Point” for FiveThirtyEight, which is a sports, politics and culture —
Jody Avirgan: Yeah, it’s kind of weird mix of data journalism.
Kitchenman: — data-focused journalism site, owned by ESPN. We’re interested in Jody’s thoughts on how data is being used in politics and his thoughts on the relevance for politics here in Alaska. And “What’s the Point” addresses how data is used in politics and through that, he’s done a series of podcasts on the history of data in politics, where he’s talked to a mix of academics and journalists. So, Jody, could you tell me a little about what your podcast and what you do?
Avirgan: Sure, so I work at FiveThirtyEight, which as you said, is a data journalism site. So, the show “What’s the Point” is about how data affects our lives. Data affects almost every part of our lives. We also have an elections podcast, which I host as well. And so that’s really honed in on what’s happening in the election right now. So between those two shows, and then obviously all of the writing that happens on the site – especially by my boss Nate Silver, who a lot of people may know – we try to take a little bit more of an analytical approach to all kinds of journalism – as you mentioned, sports, culture and politics – but obviously right now, the election. It’s a whole thing, so we’re pretty focused on that.
Kitchenman: So through this series of podcasts you’ve done on the history of the use of data in politics, what’s been most striking to you?
Avirgan: Well, there’s two things. One is the notion of the way that data’s been used by politicians to maybe get elected for president, goes much further back than I kind of expected. I think there’s this notion out there that Obama in 2008, was this data revolution, and he had this huge data team. And certainly, that campaign made huge advances, but in reporting this out, I learned about efforts in the 1860s to get voter rolls, target voters, gather information about specific voters – and that’s the other kind of thing that really comes through for me – and I think is very relevant to this election – which is that campaigns need to be efficient and one of the ways to be efficient about who they’re going to target is through the use of data. And so that’s ultimately the goal, is, you know, there’s a lot of money in elections but not that much. And so campaigns need to use their resources efficiently and, in order to figure out, OK, who are the voters who are going to actually tip this election – and there’s a million reasons why only certain voters get a chance to tip an election — but once you’ve identified them, you know, how do you use data to identify them and then target them and get them on your side?
Kitchenman: You’ve asked some pointed questions of your guests about some potential dangers that, by such a narrow focus on these specific demographics or subsections of the electorate, that the politics or campaigns are maybe ignoring other people who aren’t engaged in the political process.
Avirgan: Yeah, I mean, we’d like to think that every voter in America matters equally, right? But the fact of the matter is, that kind of every level of our system, starting from, you know, the Electoral College, or the primary process, all the way up through the way money gets allocated and the way that swing states kind of have this outsized proportion – it just inevitably means that certain voters are more powerful than others. You know, we’re sitting here in Alaska – I don’t know that much about Alaska politics, but I do know that no one is fighting over Alaska in this presidential campaign. And that’s for, you know, a bunch of sort of baked-in reasons. And you’d like to think that a voter in Alaska matters as much as a voter in Ohio, or Florida, or North Carolina, but the fact of the matter is, it isn’t.
Kitchenman: Nate Silver has shown in the last couple presidential election cycles that polling data can be very accurate, when you use it correctly, in projecting the outcomes of elections. And maybe campaign professionals may have known that for a long time, but I think he’s really helped the general public and journalists understand that better, so there’s been a huge focus on polling. But then there’s some states where we don’t benefit from a huge amount of polling. What are your thoughts about a state like Alaska?
Avirgan: Yeah, so, we have this forecast on our site, right? You go to the forecast. You can see all the states we project, which states we think are going to go in which direction. I mean, obviously the election’s still about three months away. In some states – the Ohios, the Floridas, the North Carolinas, there’s five polls a week, going on. Right now, on our forecast, the last poll in Alaska, and the only poll that gives us any information about Alaska, was conducted in January (a second poll was conducted in June and just released). It’s just a state that is not – doesn’t have a heavy amount of polling about it. One, for the reasons that we were just discussing – that it’s not considered a swing state, and pollsters are the same way. Channelling them: ‘Why would I go to a state that doesn’t quote-unquote matter, and invest all sorts of resources to do a poll, when I would rather do a fifth poll in Ohio than a first poll in Alaska?’ And the other thing is that polling is often – good polling is often done by universities, news organizations, that a well-funded, because this stuff is expensive. And so, when you have a state that has less of a population, fewer, sort of population-density centers, fewer huge universities, you’re just going to end up with less of that infrastructure.
Kitchenman: So, if you go to this forecast, for the last couple weeks, Alaska’s been running at about a 75 percent chance that Donald Trump is going to win, under the polls-only forecast. So, considering there’s just that one poll, can you give us a sense as to what else goes into that formula?
Avirgan: Yeah, so the other stuff that goes into that is historical data, I mean, Alaska is a state that has consistently been red, and so you can make some assumptions there. But, historical data, precedent, is, like, something that has basically been thrown out the window dozens of times this election so far. So it is interesting to see what happens. And if you think about a state like Alaska, the thing that I think FiveThirtyEight will when we look at Alaska, we’ll start to think about is the role of Gary Johnson, right? Because it is an idiosyncratic state. You know this better than I do, it’s a state that has all sorts of different demographics and interest groups and someone like Gary Johnson could make a big impact in this state. Montana is another state that I think is sort of analogous to Alaska in many way, from an electoral perspective. Montana is the state Gary Johnson did the best in, other than his home state of New Mexico in 2012. You could see that happening in Alaska this year, and then all of a sudden, I mean, if we’re talking about a Clinton landslide, a scenario in which she wins just an unprecedented number of states – including somewhere like Alaska – it’s probably because Gary Johnson is pulling like 10, 15, maybe more, percentage of the vote.
Kitchenman: Anything for this election cycle state out as particularly interesting, from a data perspective?
Avirgan: I guess there’s this big question of, like, how effective are the polls, right? Because we missed the Donald Trump phenomenon, as did almost everyone else. And that was because we were thinking in these other – you know, the polling is just one element. You use polling alongside historical data, kind of what you know about the way that parties operate, about the way that elections operate, and then you kind of make your best guest. If you had looked just at the polls, and blindly looked just at the polls, Donald Trump led in the polls from basically the day he announced all the way through the time he got the nomination in the primary. But there’s other factors that go into it that led us to be skeptical of Donald Trump. So that to me is one of the big questions of the many big and weird questions is, like, what is the state of polling, you know? Because there’s been a couple good cycles, these last few. But it’s a really hard enterprise. And especially as fewer people are answering their phones. They haven’t figured out how to do good polling online – it’s just becoming harder and harder to know how good is polling, and I think this election will kind of teach a lot of lessons towards that.
Kitchenman: It seems like at the very time the consciousness of how accurate polling has been – and maybe it went through a period that it was underplayed by the media almost that, not individual polls, the media’s always jumped on individual polls –
Kitchenman: But the sort of aggregation of polls that Nate Silver does and FiveThirtyEight does – it seems like at the very time that the media and the public are becoming conscious of it — there are these new dangers that polling can’t be done as accurately as it traditionally has been.
Avirgan: Yeah, polling is definitely a part of the conversation in a way that wasn’t the case 10 years ago and frankly I think Nate is partly responsible for that. Donald Trump loves to talk about polls – I mean, certainly, when he’s up in the primary, he talked about polls nonstop. And now, the other interesting phenomenon is that he’s down in the polls, you’re seeing a lot of questions about polling, questioning the methodology, this talking of skewing polls and so forth. And, you know, I think that’s sort of dangerous, because when polls are done right, they’re a really good way, and empirical way, to get a sense of the electorate. And the more you undercut that, the more it just takes away the notion that we can be rational about this and really get a sense for what people are thinking.
In the audio version, there’s more about what brought Avirgan to Alaska, the 50th state he has visited. And he described the experience he and his wife had getting stuck while hiking and being rescued.