The Distorted States of America: What the 2016 Election Really Looked Like

Population is distributed unevenly across the United States. Consequently, traditional maps of election results can be highly misleading. One approach to addressing this distortion is through cartograms¹. Cartograms replace the area of a spatial unit on a map, such as US state or county, and replace it with a thematic variable such as population. Here, the area of each US county is proportional to the number of people who voted in the 2016 Presidential election. The color of each county is proportional to the Trump-Clinton vote. The result is the Distorted States of America.
In this cartogram, there is slightly-but-visibly more blue due to a combination of (a) Clinton won the popular vote and (b) counties that voted for Trump tended to have a higher vote percentage for their candidate compared with counties that voted for Hillary. That is, the county are is proportional to the total number of people who voted, not the number of people who voted for the winning candidate.
The majority of urban counties voted for Clinton, though significant exceptions included the counties of Maricopa (Phoenix), Tarrant (Fort Worth), and Suffolk (Long Island). The Trump vote tends to rise with increasing distance away from major urban areas.
Data and software: Election results were painstakingly collected from each states’ Secretary of State. Surprisingly, these federal election data are not freely available. The county data is from the US Census² and the country boundaries from ESRI³. The map was developed using ESRI’s ArcMap³ and the cartograms were created using ScapeToad⁴.

1. https://en.wikipedia.org/wiki/Cartogram
2. https://www.census.gov/geo/maps-data/data/cbf/cbf_counties.html
3. https://www.esri.com/en-us/home
4. https://scapetoad.choros.ch/