Map Critique: Data in a Map Bucket

I wanted to look at Here’s How America Uses Its Land, by Dave Merrill and Lauren Leatherby, from July 31, 2018 on The two data reporters framed the app as a way to look more closely at how wealth is created (which is a whole problematic Bloomberg framing for a different critique). Merrill and Leatherby preface the piece by saying how well the economy is doing, but that they want to ‘decipher’ how the ‘1.9 billion-acre jigsaw puzzle of cities, farms, forests and pastures that Americans use to feed themselves, power their economy and extract value for business and pleasure.’

Merrill and Leatherby use the six land types designated by the U.S. Department of Agriculture, and display land use by the quarter million acre.

This is both a map and not a map – they use the implied rearrangement of all of those quarter million acre squares, from their original geographical arrangement into bands by use type. Merrill and Leatherby then step through each of the six buckets, showing first where that land use is located around the country and then how much area it would occupy if squished all together. They break down the larger designation, such as ‘special use areas,’ and how much of that is a national park, an airport or a golf course.

The map is a presentation of the data but is presented as a way to understand land use. The framing, helping us understand where the last quarter’s growth came from, is somewhat lost in the post itself.

Some choices that Merrill and Leatherby make are to exclude Hawaii and Alaska, because the ‘[s]pecial-use land and forestland make up the biggest land types in those states.’ This may keep the percentages from being weighed by Alaska’s area, but Hawaii and Alaska are part of the U.S., and if we have vast areas that are forestland, that seems important to consider. The one outside source that is referenced is The Land Report magazine. I wanted similar facts to be included about more of the land types, or to have information about how much of the economy is created by each sector, as is done for the urban environment.

In Against Infographics, Daniel Rosenberg contrasts the art maps of Gert Jan Kocken with the example Edward Tufte has made sure we are all aware of, Charles Minard’s figurative maps. Rosenberg says of the map, ‘[n]ot only does it combine all the data, it rigorously excludes everything else.’ This pinpointing is missing in the Bloomberg map. Many of the most interesting details are in among several other things, though the data given was not consistent.

What I think works best about this treatment is that many of us have a sense of how big Connecticut is in a way that we might not know what two million acres is. However, this is geared for a U.S. audience, or at least a U.S. centric one. The assumption of familiarity with the areas used is only a helpful tool for those of us for whom the size of Florida is a helpful explanation.

While my main critique was that Merrill and Leatherby used the data they had rather than collating from various sources, I enthusiastically did the same. Similar to the portal that Sarah showed us a few weeks ago, Washington, D.C. has an open data portal from which I could find most of the vectors I wanted. I wanted to show both how the land use was divided up in DC, and also how a single source of data missed a good part of what was going on. I used the Future Land Use, from 2011, as well as other data available from the city data portal, to help give the land use some context and a double check. For example, the classification for ‘Institutions’ in land use was largely places that are universities, but when comparing with the Universities and College Campuses, there was some but not complete overlap.  

One of the things that I was most curious about when I first saw the Bloomberg map this summer was the amount of land, in any of the categories, that was not usable for its named type of activity because it was a road or a parking lot. The DC data included a map of roads, which to my great appreciation was the shape of the road rather than a line of the center. Looking at the roads data, about 36% of area that is classified as residential in land use is a road. So while the area shown for residential use is true, zoning or tax wise, it is only two thirds true in terms of where a house could be built. This looking at the data and also at what isn’t in the data is part of what I wanted to do because I left the Bloomberg map with a lot of questions that I wanted to be able to answer.

I know that any project has to be built within constraints, and I appreciate the map Lauren Leatherby and Dave Merrill created for what it did show.


Leave a Reply

Your email address will not be published. Required fields are marked *