Municipality merger - #kommunebygger

Municipality merger - #kommunebygger

Organisation: Bergens Tidende (Norway)

Publication Date: 04/09/2015



Data visualisation of the year (small newsroom, large newsroom). The Norwegian government wishes to reduce the number of municipalities during the coming years. A process that involves a lot of numbers, calculations and debates not always readily accessible to the general public. We wished to engage our readers in the debate. The aim was to create a viral discussion built upon traditional journalism. Using data visualisation and storytelling through the use of data, key figures, maps and elements of gamification inspired by a similar Finnish app from a couple of years ago. The #kommunebygger app allows readers to test and merge municipalities and instantly see how that will affect the population, political/election results, debt, population projections etc. Upon submitting their newly merged municipality, the readers get a unique link and name. They can share the result with friends.

Technologies used for this project:

We started of by collecting a lot of data sets on municipal level from SSB / Kostra. We then went through them and settled on a few key figures we wanted to focus the data story/application on. The geographic dataset on municipal borders were downloaded from the Mapping Authority, but we needed a data set identifying neighbouring municipalities in terms of common geographical border. This did not exist, so we had to use GIS scripting to calculate and create this dataset. To present the best possible - and understandable - figures in terms of public/municipal debt we calculated raw debt data on the fly in the browser to show how much more/less debt the merged municipalities would get if the current “merger” would go through. The #kommunebygger features on-the-fly calculation of several key figures and not least the generation of auto texts to emphasis the significance of the numbers. Every number was put in context, either through comparison or auto texts or both - to avoid “data porn” of numbers and digits. We did not want the users to merge municipality without common geographical borders, so we created clustering algorithms to make sure only “valid” municipalities were merged. We used the cloud PostGIS service CartoDB as our main datastore, both for the modified and raw data sets and user submitted data. This allowed us to do advanced GIS queries and later analysis on the user submitted data. The app was built using our in-house developed “app template” which is built upon the javascript framework Angular JS and features a fully automated app deploy pipeline. All functionality is optimized for mobile devices, but works well on all screen sizes and devices. The maps were custom made and designed by the team using Tilemill and data from the mapping authority and uploaded to Mapbox for hosting. For the drawing of selected municipalities in the browser we used a custom made Topojson map file and LeafletJS. Charts were drawn using D3. The thumbnails and Facebook share images of user generated municipalities were genereated using a script utilising PhantomJS through a cloud service.


Follow this project

Comments (0)

You have to be connected to contribute

You have to be connected to follow

Leave this project and no longer be informed about this project

By joining this project, you will be informed by email when an update or a new contribution is posted on the website.

Thank you for your active participation !

The GEN Community Team