Philip Bump is a correspondent for The Washington Post based in New York. He largely focuses on the numbers behind politics and he is the author of the weekly newsletter, How To Read This Chart.
In this week’s episode of the show, Philip and I talk about his work at the Post, and dealing with all of the haters. We also talk about his work using data in the media and starting his new newsletter.
Be sure to check out my new Winno community! Get great dataviz tips and tricks to your phone every week!
- Can Google searches predict where coronavirus cases will soon emerge?
- At the end of the second year of the pandemic, the effects are more obviously partisan than ever
- Which is easier in your state: Buying a rifle or voting?
New Ways to Support the Show!
With more than 200 guests and eight seasons of episodes, the PolicyViz Podcast is one of the longest-running data visualization podcasts around. You can support the show by downloading and listening, following the work of my guests, and sharing the show with your networks. I’m grateful to everyone who listens and supports the show, and now I’m offering new exciting ways for you to support the show financially. You can check out the special paid version of my newsletter, receive text messages with special data visualization tips, or go to the simplified Patreon platform. Whichever you choose, you’ll be sure to get great content to your inbox or phone every week!
Welcome back to the PolicyViz podcast. I am your host, Jon Schwabish. On this week’s episode of the show, I’m really happy to chat with the Washington Post’s Philip Bump. If you don’t know Philip’s work at the Post, you really should, because it’s a great combination of politics and the economy and, of course, data visualization. And recently, Philip started a new newsletter called How To Read This Chart, which I thought when I first signed up for it, would be about his work at the Post, and maybe some of the more deep dives into some of his columns, and maybe some things that didn’t make it to the Post site, but are sort of on the cutting room floor; but it’s really a deep dive into data visualization, and it’s not necessarily about tools and how to go from step A to step B and build something, but it’s really about his process and how he thinks about creating a visualization that can work and communicate the stories that he’s trying to tell, which, again, is primarily about politics and the economy and other breaking news. So I think it’s a really interesting conversation, it was really fascinating to hear about how he thought about building this newsletter and how he thinks about making arguments and telling stories in his columns and in the newsletter as well. So I am going to pass it off to that interview with Philip Bump, and I hope you enjoy this week’s podcast.
Jon Schwabish: He Philip, good afternoon. Welcome to the podcast.
Philip Bump: Good afternoon. How are things?
JS: Good. How are you?
PB: Good. I mean, it looks like we’re heading into spring here at long last, yeah.
JS: I mean, it does feel like. It wasn’t a bad winter, it just feels like it’s been going on forever.
PB: Yeah, I mean, every winter is inherently bad. I mean, I love snow, don’t get me wrong I’m from the snow belt from Rochester, New York.
JS: Okay, so I’m from Buffalo, so I didn’t even know that.
PB: Nice [inaudible 00:01:52] western New York, okay.
JS: Western New York connection.
PB: So yeah, I mean, you got to love, you know, you get to about October and you’re like, oh, I wish it would snow. And then, by February, you’re like, look, I dealt with this, let’s see that, let’s move on.
JS: And then for us, we still have another four months to deal with it, right.
PB: Right. Then you get those glorious month and a half, and then…
JS: And then you’re back, right back to it, yeah. I think it was a few years ago where we had enough snow where we could build the igloos at the end of the driveway, which my kids aren’t familiar with, but that’s like how I grew up, like, you just have this thing for months that you could just keep building on.
PB: Right, and snow forts, yeah.
JS: Yeah, that’s just the way it is. That’s a good start, the western New York connection. So thanks for coming on the show. I wanted to chat about a variety of things you’ve got going on. I want to start with your basic day job because it seems like kind of the perfect job for someone who’s into politics and the economy and data visualization, and we’ll come to the newsletter a little bit. But is this where you always wanted to be?
PB: No, it’s interesting, so my dad is a sports writer, and so, I grew up both going to a lot of games, Rochester Red Wings, Buffalo Bisons on occasion as well, back in the day; so he was a journalist, but I never sort of foresaw myself going into journalism. I came to it actually, just to a large extent through technology. When I was in college, I was in college in the early 90s, right at the advent of the web, I taught myself HTML, I made websites, I taught myself Perl. I learned how to do dynamic websites. And then I discovered that it would be fun to have my own blog, and so, I made my own blog, and so, it was me sort of riffing on what’s happening in the news. But then also, I worked for a while as a designer at Adobe, and so, I also knew how to do design stuff. And so, there’s a lot of sort of, like, the data side, the math side, manifested in the design work to a large extent, and in what I’m sort of interested in, in being able to take things that were in the news and see them, which is just sort of how I see things often just through this lens with data. And so then, I was, you know, it took a while, it wasn’t until I was in my 30s that I ended up working in media specifically, because I found myself, I was just writing for websites, like I worked for media all the time, and it’s just sort of like on the side from my regular job, and I was like, why don’t I just do this, and so, that’s what I started doing.
JS: Well, so you didn’t go through journalism school as to be a journalist, got you.
PB: [inaudible 00:04:22] as a philosophy major.
JS: That’s turned out pretty well, right? So things have obviously changed and moved around with the Post, and there was Wonkblog for a while, it was sort of like the kickoff to this more data, I don’t want to call it data driven, because that’s not really fair, but like more in the weeds data, I guess, than traditionally. But you write so much about politics so much, I want to ask, like, how do you keep your sanity, like, especially today? And I’m sure, the amount of hate mail that you get is nothing that I want to look at, but so how do you keep your sanity, how do you maintain your, I don’t know, your confidence, you know, how do you not have panic attacks every day, basically?
PB: It’s an interesting question, and, I mean, there’s a very honest answer, which is that there’s a lot of stress, it’s very stressful for everybody, it’s been a stressful six years, seven years. I have two kids that were born in that meantime, so layer on top dealing with toddlers. And the fact that no matter what you write, there are going to be some people who are very angry about it, and how do I deal with it, I don’t know, maybe I’m just a New York [inaudible 00:05:42] by now. But yeah, it can be a drag, I mean, I’ll say this, if you read something that you think is good from someone, take a moment and send them a note if you can, because it’s useful, because usually the only people who are motivated are the people who are furious. So if you can do that, that’s nice. But yeah, I mean, there’s no, you know, look, I’m not going to pretend that life is harder for me than it is for a lot of Americans, I have a job, I maintained a job through the pandemic. I was able to work from home, I have a lot of advantages a lot of people don’t have. And so, it’s difficult for everyone, and I think everyone would like to get back to a time of normalcy and stability, and uninteresting times as the [inaudible 00:06:28].
JS: Right. Do you feel as, I mean, I’m sure you get a lot of the hate mail as sort of like off the deep end, but do you feel, because your stuff is so heavily data driven, and a lot of it, at least, my reading of it is it’s very much like, here’s the data, here are the facts, do you feel like when you get that mail, that you’re just like, I’m just presenting the facts here? And if you hate that, then fine, but like, do you ever feel like, well, show me your facts that make the argument, as opposed to I just hate you, because I just hate you?
PB: No, I mean, it depends on what I’m dealing with. I mean, obviously, there are ways to present data that influence perceptions. Right? That’s obviously the case. I’m not going to pretend that every single time I drop a number to an article I am doing so expecting it to be treated as entirely objective and not in support of an argument. Right? That’s how it works. The way in which what you just described manifests most frequently is on the voter fraud stuff, which I write about a lot, because it really is a data story, it is about evidentiary processes, and it’s about, you know, it’s trying to – there’s this huge group of people that’s trying to use statistics to prove fraud occur. They don’t ever actually prove fraud, they just say, look at this weirdness in the numbers, therefore, a fraud must have happened, which is this disconnect that they usually don’t recognize. But this is a black and white issue, and so, it is absolutely the case that there are times in which I will write something about how there has not been any rampant voter fraud in the 2020 election, and get feedback that’s just like, oh yes, there is, you just ignore the evidence. And it’s like, look, by now I am better versed in the evidence for voter fraud than probably 99% of Americans, there’s nothing that I have not seen and considered, and I really have to say, look, if I could discover this [inaudible 00:08:16] from 2020 election, I would win a Pulitzer Prize.
JS: Right. Yeah.
PB: But there’s just not, there’s just not there, and so, that’s very frustrating to be like, look, this is objectively not true, I can show you how it is not true, and how you’re being misled; and half you’ll be like, you’re a liar, and you’re fake news.
JS: Right. Yeah. So when you’re doing, you know, voter fraud is a good example, so you’ve probably been like neck deep in the data, because there’s a lot of data that you can parse from all the election places around the world or around the country. What makes you happiest in the process of doing your work, is it being deep in the data, is it writing the article, is it making the graph, like, which part do you really get excited about?
PB: It depends on what I’m writing about, I mean, I enjoy making a nice, complicated graph. It’s probably too complicated at times, absolutely, admittedly. But sometimes I also, just like I like to debunk hucksters, like, yeah, this guy, Douglas Frank, who’s this math teacher from Ohio, who puts together all this total nonsense seven [inaudible 00:09:25] a month saying this podcast total nonsense, about voter fraud, and it’s fun to pick that apart and be like, look, this is stupid, and here’s why it’s stupid. And without being like, you’re stupid for blaming it, but there’s no reason to take this at face value, and here’s why. That sort of thing is fun, because it’s, I mean, I’d mentioned that I’ve been a philosophy major, and the value of that is in being able to walk through an argument and assess things logically, like, that’s the side of philosophy that always really appealed to me. And I feel like that’s something that I’m fairly in debt to that particular in regards to these nonsensical claims.
JS: So in addition to your column at the Post, you recently started a new newsletter called How To Read This Chart, and I’ve really been enjoying it, and what was kind of surprising to me when you launched it is that it really is like a DataViz in the weeds type of newsletter. It’s not like the post that you wrote on the Cutting Room Floor. It’s really like here’s how I made this chart, or here’s the varieties of that chart. Can you talk a little bit of how you decided to go that route, rather than any of the millions of other ways you could have gone?
PB: Sure. So one of the things that I have learned over the course of my career and experience on the internet, which is longer than it might be, you know, people tend to think I’m younger than I actually am, but I’ve been around a long time, and one of the things I’ve learned is that there’s always – if there’s something I think is interesting, there’s an audience for it. Right? The example I love to use is furries, they are this group of people, they dress like animals, and they probably thought they were weird outcasts, and then all of a sudden, they get online, they discover they’re all these other people that they can have a convention every six months. If you are into something, there’s going to be an audience for that, and so, the thing that I’m into is just looking at data visualizations and presenting them and figuring out how they work and why they’re interesting, when they don’t work, and when people try and mislead through numbers. And so, I wouldn’t say that I’m in the weeds in the sense that I’m like, open up R, enter this code, and I’m like not quite there.
JS: Right. But still, it’s often not necessary a step by step, and like how you built it, but it’s a step by step on how you sort of thought through it.
PB: Yeah-no, right, because part of the goal here is to have people be more confident in figuring out how do they want to actually present that. So I don’t want to just be here’s a cool chart from the Financial Times that you can never ever replicate, but rather, it’s like, here’s why this works, and even if I don’t say this is precisely how you can make this chart, although I have at times offered tools [inaudible 00:11:47] nice tool that you’ve made, that allow you to sort of replicate some of this stuff, I do think it’s valuable, nonetheless to say, here is a different way of thinking about how you can present this data, and the thing that I go back to, and in every single newsletter, someone’s like, oh, you should be [inaudible 00:12:04] which I have, and I’ve been to one of your seminars, and I actually put that one in the newsletter, like, hey everyone, I’ve been to, I’ve seen him, I’m familiar with him. And, of course, people didn’t read that one, or they signed up later, and like, oh… Anyway, the thing that he did in his seminar, which was really, really eye opening for me, it’s smart in a way that you might expect. I assume everyone who’s listening this knows who he is.
JS: Yeah, I am sure everybody, yeah, and everybody listening to this, simultaneously has gone to his seminars, and [inaudible 00:12:31]
PB: Okay, so anyway, the thing that I thought was really smart about it is he starts out in the morning, and he goes through all this really fascinating stuff, the stuff with the O rings, and yada, yada, yada. And it’s really a great assessment of it, and, well, at least, when I was there, which is probably 20 years ago in San Jose, you come back, you go to lunch, and you come back, and then, it’s just like, oh, how do you do, nice charts of PowerPoint. And at the time, I was like, that’s so hokey. But then I’m like, oh, but of course, all these business is going to pay their people to go and take this.
PB: So they get this, but it’s also valuable, right? It’s valuable to people who will be like, oh okay, that’s cool, he did that thing. And like, so I want to sort of both inculcate an appreciation for data visualization by making it so it’s not intimidating, but at the same time also offer something of use so that people actually can take away something that they can actually apply in their real life.
JS: Right. Do you try to watch the DataViz field in terms of just the general conversations that are happening on Twitter, or you’re just kind of like more on the media side, the conversations on database maybe on the media side of things?
PB: Yeah-no, just on the media side. I mean, I’ll come across stuff usually, I mean, and people are – I absolutely encourage people to send me either good or bad data visualizations for the newsletter, just because it’s useful. But I mean, I come across interesting stuff every week, and I’m like, okay, this is what I’m focused on. That’s what [inaudible 00:13:49] I mean, it is, in part because I came to graphic design, data visualization I’ve been self-taught. I’ve never been sort of part of communities in that way, which is probably short sighted, but I think that’s probably the rationalization point.
JS: Well, I mean, I’ve had what, more than 200 episodes of this podcast, and I don’t think I’ve had two people come to the DataViz field from the same spot. I mean, you’re from philosophy, I’ve had astronomers on the show, I’ve had all sorts of different people. So, yeah, I mean, it’s interesting, and it’s interesting how people think about their process of working their way through the process of visualization. I’m curious when you saw Tufte 20 years ago, did he have the white gloves with the Galileo book, or did he get that book later?
PB: Oh God, I don’t remember. Is that how you say it, I always thought it was tough, I never…
JS: I think we’ve always had Tufte, but I’m sure he would be happy to tell us what [inaudible 00:14:46]
PB: He’d give us a little chart about why the pronunciations are very [inaudible 00:14:49].
JS: Right. So I know you don’t talk about this in the newsletter, or at least, haven’t yet, although I feel like I saw maybe a tweet from you the other day about the toolset that you use. And so, maybe you could talk about, I know you started at Adobe, so I’m guessing you’re deep into the Adobe Suite, but are you like a coder on top of that? Are you an Excel guy on top of that? And then, how does that intersect or not with what’s going on, on the graphics desk [inaudible 00:15:14]
PB: Great question. So it is a little embarrassing, because I’m like it’s all like ad hoc and bespoke really. I mean, I said, I taught myself Perl, I still use Perl a lot for static, because I’m just familiar with it.
JS: Right, yeah, sure.
JS: And is that for the – do you use raw graphs for more of those – I mean, a lot of RAWGraphs as more of those bespoke sort of off the standard thing, they’ve got the [inaudible 00:16:38].
JS: So are you using the charting engine in Illustrator, and do you hate it as much as everybody else?
PB: Sure, yeah, absolutely. No, I do, yes. What I use raw graphs for is if I want to do like a scale scatterplot, because it’s super easy to do that in RAWGraphs, and you can’t do it in Illustrator very easily. I just, you know how when you – it’s like having an old crappy car, and you know how to make it run really fast, and you know how to maintain it, you know what to do and what not to do. That’s how Illustrator is making graphs, and I’ve just – I’ve had this crappy car a long time. It’s just I know how to make it work.
JS: Yeah-no, I hear you. I’ve been trying to do some Tableau learning this year, and it’s like, all these little things where I’m like, this makes no sense, and then, I’ll talk to someone who’s great at Tableau. They’re like, well, it’s an obvious fix, but if you’ve never driven that car before, you’re like, where’s the turn signal.
PB: But this also goes back to your question about the graphics, so I actually don’t work with the graphics desk at the Post, which is absolutely at times been a frustration to them, and if any of them are watching, I apologize for having been in frustration. But, I mean, it is – we’re talking about a very specific product here. And so, I tend to do mostly just static images, because, A, that’s obviously the Post’s heritage. But it’s just, it’s easier, you know, I don’t have to get in, and I don’t have to figure out, you know, I don’t have to figure out scaling based on devices, and all that nonsense.
JS: Right. And so, are the various parts then of the Post broad enough where they’re not so, I’m assuming that’s where the graphic has to be maybe concerned about the branding and the look that everything is sort of consistent just because there’s so many different verticals?
PB: Yeah, no, they are, I mean, one of the things that I hope happens over the long term is that we have more people who are able to do basic coding and basic graphing, and we have a chartable tool that’s internal and [inaudible 00:18:30] in the proper setting. I use a template with graphics teammates, so I make sure I’m using the right typefaces and colors and so on and so forth. I tend, you know, it tends to still sort of be [inaudible 00:18:41] just by the nature of what I’m making, which I’m sure is frustrating to them. And so, it’s not the total Wild West, but it’s sort of like there’s now this one town that’s connected by one small railroad, and that’s my [inaudible 00:18:54] and I get to sit on it.
JS: Right. So you’ve got the desk that’s doing a lot of the custom interactive stuff, but you had mentioned earlier that you are a little more fascinated with more complex graphs. So is that what gets you excited is like a cool whatever, custom connected scatterplot or Sankey diagram or something? And then, when you are building those, are you thinking about how to educate your reader and how to read that not just like, here’s the graph, sort of, like, you can go figure it out?
PB: I try to, I [inaudible 00:19:37] perfect at it. And part of the challenge is obviously that my day job is about responding to the news, and so, I don’t have, you know, I’m not usually plotting things, you know, in the morning, I’m like, hey, I should do this, and then, by the afternoon, I’ve done it, and so it’s I will often not spend as much time as I ought on some more complicated things, in part out of confidence or either in part out of the oversight on my part, part of the goal of how to read this chart, honestly, was to get people who read my stuff a lot to feel more familiar with the sorts of things that I do. So yeah, I am not as good at it as I should be, but also, all of us have blind spots, right? Something that’s intuitive to me is not necessarily intuitive to others, and I may not be able to predict what is not intuitive to others. And so, I will get feedback about graphs, and then, I will later go back and update Post and add things, and be like, oh, here’s how you read this, you know, just to make it more clear. In terms of what really interests me, the things that really interest me are when data really locks into place, even if it’s a simple graph. And so, there are two examples, where one is that I endeavored a couple of years ago to try and figure out how presidential vote results related to the number of licensed gun dealers in a zip code or county [inaudible 00:20:48] and I did this plot that was adjusted for population, and it was just this absolutely perfect, just perfect correlation, which is just really sort of rewarding. You know what it revealed, there’s another time when someone had mentioned on Twitter that there had been – they were seeing that there were increases in the number of searches for loss of smell and loss of taste that preceded bumps in COVID cases. And so, I did a per capita analysis by state and then a Google Trends value for those searches, and it was, I mean, it wasn’t just that they overlapped, it was that they overlapped the per capita value was literally the same in those states as the scale from 0 to 100 search values in Google, which I think was just a weird artifact of the data. But it was they overlap literally, exactly, and it was just sort of fascinating, I don’t know if it still holds, and I’ve been meaning to go back and look at it, but it’s just, you know, those sorts of things where things just overlap and fit exactly when you wouldn’t necessarily expect them to, those are the sorts of things that I find kind of thrilling.
JS: To this point you just made about going back, do you think about going back to things and like, you know, we had all these conversations about COVID vaccinations or COVID infections and the vote share, do you go back, and do you just look out of a sense of personal curiosity, or do you end up, if it makes an interesting story, you’re like, well, now I can just write about it?
PB: Well, there’s a couple things at play, the first is that I’ve been at the Post since 2014, and so, there are, you know, and I write about politics, so this stuff is naturally cyclical. And so, I’ll look at how does presidential approval relate to midterms, what’s the effect of the State of the Union on presidential approval ratings. So there are certain things that you can go back to over and over again, because they have your data, your point about COVID and partisanship, I’ve written about, got it, quite right about it, biweekly at this point, just because it is. I keep expecting the trend to break, but instead the trend keeps exacerbating, and part of this is related to the decreased interest in vaccination among Republicans, so on and so forth. But, I mean, that is something that I keep revisiting in part because, you know, and honestly, because I keep getting challenged, and being like, well, just yes, thought it was bad last summer, but wait till the winter, and so it happens in northeast, so we wait till the winter. What happens? Cases still bad in [inaudible 00:23:04] counties. I mean, it’s just, you know, so you want to keep revisiting that, and so, now we’re like, okay, so what happens if there’s another wave that happens this summer, what does that look like, and where is that disproportionately affected. Just this morning, I wrote about the effects of vaccination, and the positive benefits of them. It’s something that, particularly, because we’re talking about something that has an obvious and immediate public health benefit, it’s worth elevating when you can.
JS: Right. Philip, thanks so much for coming on the show. You got the midterms coming up in a few months. Well, let me just ask one last question, I guess, then, because I was going to about to sign off, but then I was like, one last question. So we’re in March, so we’ve got, I think at least three big things going on. In the US, we’ve got the midterms coming up in November, we’ve obviously got the pandemic continuing, we’ve got Ukraine and Russia, and I’m sure many other breaking news, things will happen between now and then. But when you look forward to the next six months, do you see your time spent primarily on COVID, and then the midterms, or are you just sort of like, I’m just going to do what the news is, and the interesting data stories that I can tell?
PB: Yeah, man, I don’t know what I’m writing about on Friday. I mean, like, I remember going – I went to Indiana University in the fall of 2019, and I remember talking to them, and they were all just curious, you know, what are you guys going to do for the 2020 election. I was like, okay, here’s what our plan is, we’re going to do X, Y, and Z, and voter file stuff, and yada, yada, yada. And then March 2020 happens, and it just, then June 2020 happens, it’s just like, there’s just no way, and I’ve just given up on it. Am I going to write about the midterms? Yes, absolutely, 100%. Is that what I anticipate writing about in November? Yeah, sure. October-November? Yeah, absolutely. Is that what I may end up writing about? Well, actually, I have to write about the emergence of Godzilla from the Atlantic Ocean, like, who the… So we will see.
JS: By the way things are going, we might have to see that article.
PB: You never know.
JS: All right, man, thanks so much for coming on the show, I really appreciate it, and glad to know we’ve got a Western New York connection.
PB: You got it.
JS: All right, thanks so much.
PB: Thanks, man.
Thanks everyone for listening to this week’s episode of the show. I hope you enjoyed that, I hope you’ll check out Philip’s column at the Post, and you should also sign up for the newsletter; I’ll put links to both in the show notes, so you can check it out. If you’d like to support the show, I’ve got a number of new ways that you can interact with me and you can support the show financially. You can go over to my Patreon page where we now have just a single $5 tier, so for five bucks a month, you get a sneak peek at who the guests are coming up, and you can even send me some questions. That’s one way you can support the show. Another way you can support the show is through my paid newsletter, where you can get some more behind the scenes work, you can even get on a Zoom call with me, I’ll be setting these up every month for paid subscribers. You can get on Zoom call with me, and we can chat about whatever it is you want to chat about, primarily data visualization, maybe some specific challenges you’re having, maybe some ways that you’re thinking about solving a data visualization, whatever it is, you can check that out. And the other way you can support the show is heading over to the Winno app. The Winno app is a new tool in which I am sending out two or three text messages per week with just a little clip of the visualization I think did a good job in just one kind of tiny aspect, maybe it’s a good label, it’s a good color, or it’s a good technique that I think I might use in my work. So check those three options out. Plus, if you want to just share the show with your friends, your family, your networks, please do so, please subscribe to the show and check out all the great content coming your way on the PolicyViz blog YouTube channel and elsewhere. So until next time, this has been the PolicyViz podcast. Thanks so much for listening.
A number of people help bring you the PolicyViz podcast. Music is provided by the NRIs. Audio editing is provided by Ken Skaggs. Design and promotion is created with assistance from Sharon Stotsky Ramirez, and each episode is transcribed by Jenny Transcription Services. If you’d like to help support the podcast, please share it and review it on iTunes, Stitcher, Spotify, YouTube, or wherever you get your podcasts. The PolicyViz podcast is ad free and supported by listeners. If you’d to help support the show financially, please visit our PayPal page or our Patreon page at patreon.com/policyviz.