A Guide to Mapping at the Urban Institute
A Guide to Mapping at the Urban Institute
PolicyViz Newsletter #57
A couple of weeks ago, some colleagues at the Urban Institute asked me to pull together a document that would outline some of the major considerations around creating maps, different tools people could use, and common questions they might ask. They summarized the question they wanted to answer as follows: “Say somebody comes to your office and says, ‘I need to make a state map, what should I use, and what do I do next?’ Instead of explaining the answer in person, it’s written down.”
Well, I’m not cartographer, but that sounds like something that’s right up my alley. I first consulted two resources: my Better Data Visualizations book (#humblebrag) and Kenneth Field’s book, Cartography. My goal wasn’t to go super deep into maps and map theory, but I wanted to hone in on the most important issues.
In the end, with the help of my Urban colleagues Will Curran-Groome and Aleszu Bajak, I settled on 5 sections:
Step 1: Should you use a map to visualize your data? Here, I wanted to make sure people know that just because you have geographic data, doesn’t mean you must make a map.
Step 2: What kind of map should you create? I showed a few different map projections (we use the Albers projection at Urban), examples of tile grid maps and others, along with a discussion of the underlying tension when creating data-driven maps.
Step 3: What tool should you use to create your map? We use three primary data visualization tools for our published work at Urban: Datawrapper, Excel, and R, so I created a table laying out the pros and cons of each.
How to create maps in these different tools? This is the part that took the longest—specific tutorials in how to create several map types in each of those three tools.
Additional Resources, Books, and Blogs. Can’t have enough resources, right?
Obviously, a lot of the content here is specific to Urban, especially around map colors and fonts, but I think the rest is valuable to anyone interested in making high-quality data-drven maps. You can check out the entire document on the Code@Urban site. I hope you’ll find it useful!
Thanks again,
Jon
Podcast: Summer Listening
I am unofficially on break from the PolicyViz Podcast. Actually, that’s not entirely true—I’m getting myself ready for a new season. I’m planning on doing more “themes” this season and by that I mean, I’m going to do several episodes dedicated to qualitative data and data visualization and several episodes dedicated to data visualization tools. Obviously, I’ve had guests talk about both topics in the past, but I’m going to zoom in on some of the specific considerations around both.
Anything on your mind you want to learn about? Anything on the show you’d like to see improve, changed, or highlighted? Message me in the Substack app, Twitter, LinkedIn, or just through my site and let me know what’s on your mind!
Things I’m Reading & Watching
Books
The Fire Next Time, James Baldwin
Holding It Together: How Women Became America's Safety Net, Jessica Calarco
Qualitative Literacy: A Guide to Evaluating Ethnographic and Interview Research, Mario Luis Small and Jessica Calarco
The Oxford Handbook of the Social Science of Poverty, edited by David Brady and Linda Burton
Articles
TV, Movies, Music, and Miscellaneous
Veep, HBO Max
The Bear, Hulu
Welcome to Wrexham, Hulu
Note: As an Amazon Associate I earn from qualifying purchases.
Support the PolicyViz Newsletter
Please help support this newsletter by rating and reviewing the PolicyViz Podcast on iTunes or any of my books on Amazon. More ⭐️ ⭐️⭐️⭐️⭐️ reviews go a long way towards attracting more listeners and more readers.
Hi Jon.
Thanks for this great article about dataviz and mapping. My works focus on mapping and dataviz at the North American level so your article is right up my alley!
About the projections:
- I would indicate to never use the Mercator projection. Although it is everywhere and many online tools only use this one (mostly because of its historical predominance), it is the worse projection someone can use in the 21st century. It distorts the world (as you pointed out with Greenland who in reality is supposed to fit 14 times in Africa and not be bigger) and many would say that it has a very colonialist point of view (as it makes the Northern Hemisphere appear much bigger than the Southern Hemisphere). If someone can choose a projection, I would choose anything but Mercator.
About Choropleth:
- I would recommend to stay away from Choropleth as they create a lot of "perception distortion" by highlighting bigger areas even though the value that is represented by the color is rarely based on surface area. Big states in the United States like Texas or big provinces in Canada like Quebec will always stand out more in a Choropleth even though the represented data might not even be about the surface area of these regions. The Hex Grid you mentioned is much better to reduce this "perceptive distortion".
Cheers!
Dominique