Yesterday the New York Times published an interactive map of the 2016 presidential election which succeeded in annoying a lot of cartographers. The NYT's Extremely Detailed Map of the 2016 Election allows you to explore the 2016 presidential election at the voting precinct level.
The map is a great tool for exploring how many votes were cast for each presidential candidate at precinct level. It also allows you to see at a glance which precincts overwhelmingly voted for either candidate. It does this by shading each precinct by the percentage of votes cast for the winning candidate. The darker a precinct is colored red on the map then the higher the percentage of votes cast for Donald Trump. The darker a shade of blue then the higher the percentage of votes cast for Hillary Clinton.
It is this choice to shade precincts by the percentage of votes cast for a candidate that has upset a lot of people. The reason why many people are arguing that the NYT map is misleading is because it places too much visual weight on the large rural precincts won by Donald Trump and distorts the overall result of the election. For example Jon Schleuss of the LA Times posted this direct comparison of the NYT election map with the LA Times Election map -
Both are maps of the same precinct level data. However the LA Times map shades the precincts by the number of people who live there rather than by the percentage of votes cast for the winning candidate.Therefore in the LA Times precinct election map more visual weight is given to precincts with the most voters rather than to the most partisan precincts. The result is a much more accurate map of the number of votes cast for each candidate in California.
If you want a detailed explanation of the problems with the NYT election map then you should check out Kenneth Field's Cartographic Hyperbole post of the map. Kenneth Filed's considered thoughts on election maps also feature prominently in the Wired's Is the US Leaning Red or Blue? It All Depends on Your Map. The Wired article looks closely at how the different cartographic and data visualization choices you make can greatly influence the story your maps tell. The article is illustrated with a number of different maps of the 2016 presidential election visualizing the data in a number of different ways.
The upshot of all this criticism is not that the NYT election map is wrong. It is just that the visualization choices made have resulted in a map which could easily mislead users about the level of support for the winning candidate in the 2016 presidential election.
No comments:
Post a Comment