Most of us pay te tax we are supposed to to secure our welfare. But there are a group of people who avoid taxes or pay a lot less then they should by transfering their assets to tax havens. Paradise papers gave a unique insight into a world that is usually closed for insight and review – a world where some of the world’s most wealthy companies and individuals withhold large amounts of tax through advanced tax planning.With solid and massiv research and the cross linking of several data sources we build the so called "Svensklistan", a database of the Swedes we found in paradise papers. By finding and adding data such as personal numbers, adresses, income and tax information we were able to paint a pretty good picture of who this people are. In the visualization, wich is what we are submitting here, you can follow each person as a dot by scrolling through the different variables. We show how old they are, what they earn, how much they pay in taxes, where they live, and much more – in a beautiful and smooth visualization.
What makes this project innovative?
The Paradise papers consists of 13,4 million documents from two law firms and company index in 19 of the worlds tax havens . The documents were reviewed by 90 media companies from 67 countries. It contains data about 180 countries over a time period of 70 years. Through extensive searches we were able to find 2000 Swedes in the leak. We then created a database of the swedes by cross linking the list against different databases. Our purpose with the visualisation was to show every individual, but as the bigger picture. How do you tell the story of 1200 people, without revealing their identities? Our idea with a dot for each person gives you a overview of the swedes and scrollytelling is how we let our users take it all in. In step two we placed them all out on a map (at the centerpoint of their zipcode) to show where in Sweden they live.
What was the impact of your project? How did you measure it?
The publishing of our visualization had a big impact online. Hundreds of thousands of people were intrested in knowing more about the Swedes in the leak. Since this was an international leak we decided to release a version in english as well, which grew very popular when we put it on the subreddit Data is beautiful. More then 200.000 people abroad read it. Several news stories were made from the material, which got many consequences. Leif Östling, the highest representative for Confederation of Swedish Enterprise, had to resign and a big debate about moral and welfare took place. In Sweden the government has suggested several changesfor the Tax Agency with reference to Paradise papers.
Source and methodology
Paradise leak first came to the German newspaper Süddeutsche Zeitung. The documents were then shared with the International Consortium of Investigative Journalists, ICIJI. Three Swedish editors were invited to participate in the cooperation: Mission review (Uppdrag granskning), SVT News and the news agency TT. Already at an early stage, we decided to chart which Swedes were in the leak and why. We couldn't dig deep into every single case, but we wanted to tell as much as possible about all Swedes in the leak and present the results in an understanatble way to the audience.The original source is Paradise leak where we through extensive searches found Swedish names in different types of documents. To find out more about who these people were, we had to generate social security numbers for all of them, which was done with manual searches in SPAR. We managed to identify 1,200 of the swedish person with a Swedish social security number and then proceeded to obtain further information such as address, tax information, income, company assignments, etc. We did this by combining the list with a variety of registries, such as SPAR (Statens personadressregister, The states person-adress register), the company register in Sweden (Bolagsverket) and Acta Publica. We used a database with coordinates for the center point of all Sweden's zip codes (used to place the people on the map in the middle of the zip code area in which they live)Income capital, tax information from the Swedish Tax AgencyPopulation register
Few tools were involved. For analysis we used Excel and Acces and Postgresql. The map was made in Cartodb and the main visualisation in D3.js with some data preparation with Pandas.
The following people had a crucial role in making the visualization:Ola HjalmarssonLinnea Heppling (me)Linda Larsson KakuliFredrik EdgrenLot's of more people worked with the leak, but the names above are the ones involved with the visualization.