Cartograms of election results
By Dejan Vinković: Cartograms are maps with deformed geometries that substitute land area with some other thematic mapping variable, such as election results. Conveying information in this way often produces more informative visualization than ordinary maps, especially when used for online interactive data access.
Election results are particularly interesting for cartograms because the land area can be in a large disparity with the population density. A normal mapping can give a false impression that a political option that preformed well in highly populated areas does not have such a good results because they cover a small land area. A cartogram, on the other hand, will clearly show the true strength of that political option. A nice demonstration of this effect can be seen in the example of US presidential elections, where the “red” states cover a larger land area than the “blue” states and this may give a wrong impression that they are dominating the overall election results.
Even though cartograms have a long history, they are rarely used in the media coverage of elections. For example, this type of visualization of election results has never been used in Croatia. Hence, for the first time ever, we created such a map for the daily newspaper Jutarnji list, where the mapping polygons are municipalities and cities (556 of them in total).
Here we provide the data for cartogram’s borders and a small javascript code in d3.js for utilizing this dataset in an interactive visualization. The cartogram was created with the “diffusion-based method for producing density-equalizing maps”, invented by M. T. Gastner and M. E. J. Newman and published in 2004 in PNAS. The authors also provide the software for transforming an ordinary map into a cartogram. Our starting point was a map of Croatia with the adjusted number density of voters per each municipality and city. The density had to be adjusted in order to give each voter equal representative value in elections, since Croatia has 10 election districts electing 14 parliamentary seats per district, but with significant differences in the number of voters in each district.
Here is an example of how things differ between an ordinary map and a cartogram. Below is a map of Croatia with municipalities and cities colored by the percentage of votes for MOST – a new party founded only a few months before the elections which emerged as a complete election surprise and took a total of 19 seats out of 140. From this map it is hard to see how they managed to take so many seats.
However, if we look at the cartogram below, we immediately notice a large area covered with very good election results. MOST was very successful in Zagreb, the capital city, and in a few other cities, mostly in Dalmatia.
The cartogram is accessible in json format, under Creative Commons License:
Cartogram of Croatia by Dejan Vinkovic is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Based on a work at http://oraclum.eu/cartograms-of-election-results/.
We also provide an example how to draw this map with d3.js and use it for interactive visualization. The json contains coordinates for drawing polygons in svg format. Each polygon has a unique ID number (the official Municipality/City Code Number) that will connect this polygon with the data. HERE is a csv file with some example data – the key columns are “kod_lokalne_jedinice” (IDs of polygons) and “glasova_za_postotak” (percentages of votes).
The visualization code that connects all these together using d3.js is HERE.
Recent Comments