Following the UK's 2017 general elections, I was curious as to how our own electoral system compares to others used in democracies around the world. A quick wiki search got me interested in the D'Hondt method of seat allocation.
It seemed like a fun little code exercise to implement it and see how it's results would compare against the results of the UK's 'first past the post' system.
The data from the latest election wasn't yet available, so I took 2015 data from the Electoral Commission's Website.
A version of this data can be seen in data/2015-05_GeneralElection.json, which was converted to JSON for easier processing.
At the moment it's just looking at:
- First past the post: What we're doing now. There's a bunch of constituencies which can each elect one person to represent them at parliament
- Party list proportional representation (National level): What if all votes across the country were tallied up, then seats assigned based the D'Hondt seat distribution.
- Party list proportional representation (County level): What if each county could send a certain number of people. For this distribution I have taken all constituencies within the same county and summed them to get a seat count for that county.
These were the ones I was interested in seeing at the time, but the framework is there to implement other electoral systems as well.
probably not. It was a fun bit of code-kata, but it's not my job or anything.
Cider-ware - Use, Distribute and modify to your hearts content. If you like it or find it useful (and we ever meet) feel free to buy me a cider.