When trying to decide what to map I unsurprisingly thought of the 2016 presidential election. I wanted to discuss everything from demographics to voter suppression, through the use of maps. Due to time constraints, I decided on a foundation to base our discussion on and create two maps using Tableau showing the popular votes for Clinton (fig. 1) and Trump. (fig. 2)
One thing that comes to mind when I view these maps is Richard Jean So’s article “All Models Are Wrong”. There seems to never have been a truer statement when faced with these maps.
The first and most obvious piece of missing information is the electoral college vote, which neither of these maps represent and was one of the clear reasons Trump won the 2016 presidential election. If any alien visiting earth for the first time were reading these maps, they might say “Clinton won the popular vote, therefore she won the election.” But this is false and the maps fail to show that. There were many more elements affecting the outcome of the 2016 election that are not shown here; lobbying, number of visits to states by candidates, and Russian interference to name a few, all of which could be their own map.
Using Tableau to create these maps was a bit of a learning curve. I downloaded a database from the Federal Election Commission of the United States of America and cleaned the data myself for the first time. I chose Tableau because it gave me the freedom to choose my own large set of data without manually inputting, and frankly because I cannot seem to get access to Carto since the workshop a few weeks ago.
In conclusion, I’m interested in what these maps do not tell us, and also what they can. If we could create maps and accompanying diagrams to show us the popular vote, voter suppression, Russian collusion, candidate visits, monetary donations, racism, sexism, and electoral votes, could we predict the next election? Could we find out how to change it?