Not so long ago, we said that Jon Schwabish seems to be “all over the place”, making a reference to his restless stream of in-depth articles, insightful presentations and constant social media engagement – especially on Twitter and, in another level, with HelpMeViz, one of the remarkable new initiatives of 2013, highly praised by the data visualization community.
For the past nine years, Jon worked as an economist for the U.S. federal government, more precisely in the Congressional Budget Office (CBO). There, he was responsible for opening up the way for data visualization not only as an analysis tool, but more importantly, as an effective way to convey information. He has also conducted research on inequality, immigration, retirement security, data measurement, food stamps, and other aspects of public policy in the United States. Now, he has joined Urban Institute, where he will be continuing his economic research agenda and working with the communications team.
In parallel, Jon runs public workshops about data visualization and presentation techniques, teaches “Statistics and Information Visualization” at MICA, and is becoming a regular presence in conferences as a keynote speaker – like this one at Tapestry 2014.
In the midst of all of this, Jon still managed to find time to answer some questions about data visualization illiteracy, major trends in the field, his new work, and much more.
Visual Loop (VL) – Jon, when you took that Tufte one-day course, did you imagine that one day you’d be on the stage of an event such as Tapestry, talking data visualization? What drove you to get so involved in this field, so quickly?
Jon Schwabish (JS) – It’s really amazing when I think about it. A few years ago, I was working on my economic research agenda and now, in addition to that research, I am in a role where I help others improve the way they analyze their data and communicate their findings to a wider audience. I am extremely fortunate to be surrounded by friends, colleagues, and managers who support and encourage my work. Tufte’s course—though in retrospect did not provide much in the way of practical tools—opened my eyes to different and more strategic ways of presenting data. And having worked with data in my own research for many years, I felt it a natural step to improve how I visually communicate my work.
My involvement with data visualization was driven by a few factors. First, I wanted to explore ways to get more people to read my research and to have it make a difference. The research I’ve undertaken over the past decade — at the Congressional Budget Office and now at the Urban Institute — has addressed public policy challenges in such areas as food stamps, retirement security, and income inequality. To make informed decisions about those programs, policymakers need objective information and facts. Today, as the modalities by which people receive information change (through, for example, more mobile devices), data visualization offers a means by which research and analysis can be disseminated in more useful ways.
Second, people in the data visualization field are incredibly willing to share their knowledge and time to help others improve the way they present and communicate information. The first few practitioners and researchers I met, including Stephen Few, Andy Kirk, Robert Kosara, and Kim Rees, really helped and encouraged me to create visualizations and write about the challenges I faced. I’ve relied on these folks—and many, many others—for ideas, editing, and feedback on a wide range of projects.
VL – And besides Tufte, what other authors and practitioners have become a reference to you?
JS – I rely on the entire data visualization, journalism, technology, and presentation communities to help me think about the best ways to disseminate research. I can only give a short list here (and apologies to the friends I left out!):
- Excel: Jon Peltier and Jorge Camões;
- Journalism: Alberto Cairo, Amanda Cox, Hannah Fairfield, and Derek Willis;
- Teaching: Enrico Bertini, Stephanie Evergreen, Andy Kirk, Jen Lowe, and Cole Nussbaumer;
- Technology, color, storytelling, and just all-around good things: Bryan Connor, Lynn Cherny, Francis Gagnon, Robert Simmon, and Santiago Ortiz.
I also spend a lot of time in the presentation technique and design field, relying on authors like Nancy Duarte, Carmine Gallo, Garr Reynolds, and Gavin McMahon. Interestingly though, the presentation field feels less like a community than data visualization. There doesn’t seem to be the same discussion, critique, and debate as in data visualization even though there’s plenty to argue about. But the work of these authors (and, of course, many others) is equally as important, especially in the sciences and social sciences, where good, basic presentation design and techniques could really improve how knowledge is disseminated.
VL – You must have faced tremendous challenges during the time at CBO, promoting better ways to visualize and present data. How serious is the visual/graphic illiteracy problem in the U.S. Congress, and in politics in general? And how does that affect the average citizen?
JS – My colleagues at CBO were tremendous at recognizing and embracing the importance of data visualization in how the agency communicates its work. Very early on, I had great support, which right away led to infographics, CBO Snapshots, and improvements in the agency’s regular reports.
In general, the graphic illiteracy problem in government and public policy is serious. Just look at the charts William Gray shows in his @FloorCharts Twitter feed. This graphic literacy challenge is important: If policymakers and those trying to affect policy can’t effectively communicate their ideas—and visualization is a key tool in those efforts—it will be hard to convert those ideas into improved policy.
If policymakers and those trying to affect policy can’t effectively communicate their ideas—and visualization is a key tool in those efforts—it will be hard to convert those ideas into improved policy.
VL – But there have also been several improvements in recent years, right? Can you mention some of the positive developments you witnessed/participated in during the time you stayed at the CBO – even from other U.S. Federal agencies?
JS – Yes, there have indeed been great strides made in the past few years. In addition to the visualization efforts I worked on at CBO, I’ve seen changes at lots of other agencies: the Census Bureau, for example, has its own visualization library and has opened up its datasets in different ways (for example, a new API was launched last year).
But in some ways, government agencies have been slow to embrace new technologies and to understand changes in the ways people receive their information. Agencies still publish PDF report after PDF report with what appears to be little thought to how the research will reach the target audience, or even a wider audience. Agencies could use data visualization as a building block to a larger communications strategy that might include additional tools and techniques to disseminate findings (such as blog posts, infographics, slide decks, podcasts, and interactive visualizations). Improvements might also simply take the form of better layouts and graphics in existing publication types. These challenges are driven by the skill set of existing staff, revising existing agency mandates, and the overall motivation of managers and staff.
One of the other important developments I’ve seen is that some agencies are taking a more strategic look at whether and how they can release their data. Just moving from releasing tables in PDF format to Excel files is a major step, but I’ve also seen some agencies explore how to create APIs or flexible data extraction tools. As governments open their data—and not just open them up, but open them up in ways that are easy for people to use—more people can lend their knowledge and insight trying to solve our policy challenges.
Agencies could use data visualization as a building block to a larger communications strategy that might include additional tools and techniques to disseminate findings
VL – And now a new challenge, at the Urban Institute! Congratulations! What can you tell us about your role there and the people you’ll be working with?
JS – Yes, thanks. It was not an easy decision to leave CBO—the work I did there over the past 9 years was extremely rewarding and I was proud to have led the agency’s data visualization efforts. At the Urban Institute, I’ll be splitting my time between continuing my economic research agenda and working with the communications team. In both roles, I’ll have heavy responsibilities in Urban’s data visualization efforts.
My hope is that I can act as a strong link between researchers and the communications team, as well as improving upon and helping to create new types of visualizations. And that’s not just on the production side—it will be increasingly important, I think, to feature the communications strategy in project applications to help ensure that the research reaches the target audience. Urban already has a great team — Tim Meko and Daniel Wolfe, for example—have already created some terrific projects (such as Our Changing City), so I’m looking to add my voice and to act as a bridge between the research and the communications teams. One of my goals on the data visualization side is to help create visualizations that will make Urban the place where policymakers and the public will go not just for great research, but also for tools and resources that will help them bring their own perspectives to these issues.
One of my goals on the data visualization side is to help create visualizations that will make Urban the place where policymakers and the public will go not just for great research, but also for tools and resources that will help them bring their own perspectives to these issues.
VL – In a certain sense, as an economist you’ve been dealing with ‘big data’ way before it became…well… so ‘big’! How do you see the ongoing discussion about the merits and disadvantages of something that has been around for much longer than people normally think?
JS – It is sort of funny because when the “big data” term started growing, many economists I knew shrugged—we’ve been using large datasets for years from Census data (hundreds of thousands of observations) to Social Security Administration data (millions to tens of millions of observations and, if you’re lucky, hundreds of millions of observations). But the new big data—billions if not trillions of observations—not only introduces significant technological challenges of storing and processing, but also begs the question of how are we as researchers and visualizers going to comprehend and analyze such enormous amounts of information?
I’ve also found that the big data discussion is often centered on the technological challenges of getting enough memory, processing speeds, and visualization tools. Aside from national defense, I haven’t seen a lot of big data applications to social and economic policy challenges, but I’m excited to see some recent developments in that area and to see how different groups might be able to work together.
VL – Speaking of ‘buzzwords’ , which other significant trends excite you the most, at this time?
JS – Five trends stick out to me as the most exciting: big data (which I mentioned above), open data, teaching, storytelling, and tools. Open data, at least as it relates to governments, is interesting because it requires a change in the way governments go about their business and how they view and change their existing mandates. For example, agencies whose main goals are to produce research reports but now want to release their underlying data will have considerable technical and internal process considerations. I’m eager to see how agencies will change and adapt to these different demands.
Teaching is also an interesting area in the field. The number of data visualization courses at schools and universities is growing quickly, as is the range in the departments in which they are offered: in journalism schools (see, for example, Alberto Cairo at the University of Miami), policy schools (I will be teaching a data visualization course at Georgetown’s McCourt School of Public Policy this fall), and art schools (I’m currently teaching “Statistics and Information Visualization” at the Maryland Institute College of Art). (It’s also worth mentioning the growing trend of teach coding in the nation’s elementary and secondary schools). One of the challenges is to figure out what skill sets students need in those different fields and how data visualization can help them in those pursuits. Enrico Bertini recently wrote a couple of great posts about the challenges of teaching data visualization.
I’ve also seen firsthand the developments in the individual data visualization workshops and trainings, such as those offered by myself, Andy Kirk, and Cole Nussbaumer. Questions remain about the best way to teach data visualization in a one- or two-day session, what types of skill sets people might want to learn, and how to best market those workshops to potential participants.
Storytelling may be the hottest buzzword right now. There have been some terrific blog posts over the past couple of months (see here, here, here, here, and here), as well as an interesting Data Stories podcast with Alberto Cairo, Kosara, Bertini, and Moritz Stefaner. When it comes to data visualization, how should we define a story? When is it appropriate and when is it not? How do we—and how should we—weave analytics, text, and visualization together? What is especially exciting for me is to see how people from different fields share their perspectives on how to incorporate stories—or narrative or annotation—into their visualizations.
Finally, the explosion of (often free) online visualization tools and programming languages is simply amazing: d3, R, Python, Processing, Lyra, Datawrapper, Highcharts, amcharts, and RAW, to name just a few. There are also infographic-creation tools, such as Easel.ly, Infogr.am, Lemon.ly, Visualize.me, and Visual.ly, and I’m interested to see whether those tools can succeed. What’s particularly interesting about visualization tools is that there does not yet seem to be a single tool for data visualization—and perhaps one just doesn’t exist. Is there a single tool that allows the user to go from analysis to design and graphic creation to interactivity? And yes, I’m aware that dedicated fans of some of these tools are going to yell at me (“What?! You’re crazy. R/Python/Lyra/etc. does all of that. It’s awesome.”), but I just don’t believe we have found the one-size-fits-all tool just yet, if it even exists.
What is especially exciting for me is to see how people from different fields share their perspectives on how to incorporate stories—or narrative or annotation—into their visualizations.
VL – A mention, of course, to HelpMeViz, an initiative that was very well-received by the data visualization community. How is the site doing, after almost six months?
JS – Thanks! Yes, the site has received a lot of great feedback and coverage. It is doing well and I’m now receiving one or two submissions per week. I’m of course always on the lookout for more content to help populate the site and perhaps some of your readers will be able to use it.
I’m now exploring different types of things I can do with the site. As an example, in June I’ll be partnering with Bread for the World in a live data visualization Hackathon. The event will be held in Washington, DC, and I’ll provide the data on the HelpMeViz site and live-blog the entire event. That way, the room of 20–25 people working on the visualization will (hopefully) have others chiming in from all over the world. If successful, I’d like to do more of these types of events. I’m also looking for sponsors and partners to help host data visualization contests and promotions.
VL – Finally, Jon, apart from your new work, is there anything else you’re involved that you’d like to share with our readers? Do you continue to do workshops?
JS – Yes, I continue to run my public workshops and am developing a full-day workshop on “Data Visualization in Excel” and a half-day workshop on “Presentation Design and Techniques.” I hope to be able to launch both of those this fall.
As I mentioned, I’m teaching “Statistics and Information Visualization” at MICA and will be teaching a two-day workshop as part of Georgetown’s new Executive Education program in early July. I’ll also be teaching a full-semester data visualization course at the Georgetown McCourt School of Public Policy this fall.
In the short term, I’m excited about a trip to San Francisco at the end of July when I’ll spend a week at the Federal Reserve Bank of San Francisco and then head up to Seattle to spend some time with Tableau and Microsoft. While in San Francisco, I’ll conduct research on the Disability Insurance program and help the Bank with some of their data visualization needs. I’ll also be visiting folks at Apple and Facebook. It promises to be an exciting couple of weeks all around.
I also have a few other side projects going on and hope to get a couple of them out the door by the end of the summer, so stay tuned.
VL – Thank you so much, Jon!!
JS – Thank you!