UK general election 2017 results presented as a hex bin map
The [first] general election of 2017 was held 3 months ago by the time I am writing this article. Why the wait? Well, I was fairly absorbed by preparations for this year’s FOSS4G conference in Boston.
At FOSS4G Boston I had the chance to see an excellent talk on #FakeMaps by Steven Feldman. The complete slides for this standing space only presentation are available here. This talk was inspiration enough to finally put my ideas into making a better election results map. Besides, there is also the small matter of the German election later this year and I want to be able to present these results as soon as they become available.
What’s the problem with traditional election result maps? There is a great blog article from the aftermath of the 2015 general election by Kenneth Field who also happens to be one of my former lecturers at Kingston University. Ken provided some of the maps for the #FakeMaps slides and I don’t want to reiterate here the valid points which have already been made about the traditional approach.
The next two maps are from this article in London’s Metro ‘newspaper’. First up, the traditional choropleth map. I bet a quick look on this map makes Theresa feel much better about her hung parliament.
However, this map does not represent the popular vote. Using hexagons for mapping has been a recent trend which has been continued for the reporting on the 2017 election results. Usually the results are presented as a hex cartogram as seen in this example from the same source.
Yuck! I am not a huge fan of cartograms and you don’t want to get me started on heatmaps. IMHO, these abominations of map making do not give much better insight than a tradionational choropleth map. The geographic shape is a mess and individual constituencies may have between 15,000 and 75,000 votes each. This isn’t ideal for a visualisation of the popular vote.
My objectives for a better election results map were to present the popular vote where people live. For this reason I distributed election results according to the distribution of people in the Global Human Settlements population grid. I also want to show the margin of victory and use shaded party colours to display the percentage of the majority vote in each grid cell.
This map was completely drawn with the Open Source desktop QGIS. Please refer to the map for the data attribution.
Keen cartographers will notice that I use the Web Mercator projection. This is not ideal but I am a mostly a web developer these days and make maps for presentation in web mapping frameworks such as Leaflet or Openlayers.
The projection is not a big problem since I distribute and scale symbols according to the vote density.
You can download a higher resolution pdf version of this map here.
Why medium? My current blog is hosted on the free Openshift 2 service. This service will be suspended next month and I haven’t decided yet whether I want to go through the upgrade process to Openshift 3 or just use medium as my preferred platform.
Please visit GEOLYTIX website if you are interested in bespoke map visualizations and other aspects of location intelligence. Perhaps we can help you to turbo charge your spatial data.
