When Color Hurts


The above (interactive) image comes from WalletHub’s “States Most & Least Dependent on the Federal Government” article and, as you may notice, is unspeakably ugly. And as I have previously mentioned, I’m color blind, so for me, the ugly image is also almost completely unreadable. Thanks, folks! Ever hear of the most common type of color vision deficiency? Sheesh.

So I did y’all a favor, WalletHubbers, and recreated the map with a monochromatic, easier-to-read color scheme—admittedly without all fifty gradations in color, but that was overkill anyway. And mine is interactive, too:

BTW, I should say thanks to WalletHub for including the data! That allowed me to crib the numbers from their website and create a CSV to import into the spiffy free GeoCommons interface, which includes all kinds of tools for mapping geographic data. So thanks to GeoCommons, too!

Now, something odd about the data… If you mouse over each state, you get a lone integer, namely WalletHub’s ranking from 1 to 50 of their calculation of that state’s dependency on the federal government. To calculate the ranking WalletHub used a weighted average of three factors: 1) Return on Taxes Paid to the Federal Government, 2) Federal Funding as a Percentage of State Revenue, and 3) Number of Federal Employees Per Capita. (BTW, I have to give credit to the designers of the ugly map for including a key in the lower left corner; I couldn’t figure out how to do that with the GeoCommons interface, so I have to admit falling short of my own expectations in that regard.)

I realized that mapping the first criterion would give me a chance to demonstrate one of my favorite ideas in depicting data with color. If one is using a color palette that transitions from a warm color to a cool color, it must through what I think of as an “inflection point” in between. There’s an obvious inflection point for the “Return on Taxes Paid to the Federal Government” data: the break-even point at which a states get a dollar back for every dollar paid in taxes.

So I made another map. If you click on the box in the upper right corner of my map, you’ll get a “Layers Palette” menu. (This might all be easier if you view the map on GeoCommons instead of embedded on this page. Unselect the top data set (“by rank”) and select the lower one (“by return on taxpayer inves…”). Now you should see states colored red, white, and blue.

If a state pays more taxes than it receives back in federal support, it appears blue. States that receive back a bit more than they get in return (less than $1.25 for every dollar in taxes) appear white. And if a state receives significantly more money than it contributes in taxes, it appears pink or red. Again, you can mouse over to see the actual values to discover that Illinois gets back only $0.56 for every dollar in taxes whereas South Carolina receives a whopping $7.87 from the feds for eery dollar in taxes.

Coincidentally, many of the blue states are, well, “blue states,” while many of the red states are “red.”

Even aside from this “coincidence,” I think the color scheme works because it makes clear the givers versus the takers at the opposite ends of the data spectrum. I could find science-y data to do this with, too, but these data landed in my lap, so why not use them to make my point?

(The Huffington Post reported on this story, BTW, which is how it came to my attention. So a final thanks to all my crazy liberal friends on Facebook who post these kinds of things…)