The VAT hustlers
Organisation: Danish Broadcasting Corporation (Data Team) (Denmark)
Publication Date: 04/06/2016
Size of team/newsroom:small
DescriptionPeople and companies pays taxes and VAT to Denmark in order to support common goods such as public health care, schools, roads, pensions etc. But when companies pay the VAT it rests on a fair amount of trust from the tax authorities towards the companies rather than control. The risk of getting caught if you cheat the authorities is very low. As a result, the tax-system is victim of comprehensive fraud with VAT every year. The Data Team – and the documentary-team at DR – decided to investigate how, where, and who are behind the fraud. This investigation was so comprehensive, thorough, and deep that we were able to reveal a huge network of scam-businesses with links to international fraud too. We revealed who is behind the fraud, how it works and who benefits. We wrote stories revealing companies that not only cheat the tax authorities, but also the costumers when they sell old, filthy candy to discount price. And not the least we wrote about the links between the scam-businesses in Denmark and terrorists in Spain. A link, which according to VAT-expert Richard Ainsworth from USA, never has been as clearly documented. This investigation is a masterpiece of data analysis based on the unfolding of a network in a dark, secret world of crime where no information is handed easily. The output was more than 50 articles published at www.dr.dk, 3 interactive visualizations and a joint-venture with the documentary team at DR which led to 3 documentaries for TV primetime and news at the main evening news program. Credit: Bo Elkjær, Nis Kielgast, Benjamin Hughes Dalsgaard, Mads Rafte Hein, Mathias Sommer, Troels Kingo Larsen and Mads Nilsson
What makes this project innovative? What was its impact?Most significantly the project revealed how a network of suspected terrorists and Syrian foreign fighters took part in extensive fraud with public funds in Denmark. The network of suspects included among others a Danish citizen on the US terrorism watch list, several Spanish terror suspects arrested in a large crackdown in 2015, a Spanish fugitive former soldier who since has been reported killed fighting for ISIS in Syria and a Danish citizen killed in a US drone attack in Syria last December. Common for all these – and several other similar suspects – were, that they operated as directors and strawmen in several companies suspect of or investigated for different kinds of fraud, including extensive VAT fraud. The story was investigated in several iterations, initially taking off with groundbreaking use of social network analysis software, which made it possible to sift through and analyze large datasets and different public databases to identify suspicious companies and directors that needed closer examination. The result of the data analysis was combined with classic investigative reporting, revealing how terror suspects and foreign fighters in Syria were set in as directors and strawmen in different companies – some of which, it turned up, had company addresses in empty mailboxes, vacant lots and even, it turned out, a publicly owned meter cabinet. It has previously been widely reported that Intelligence services and Law Enforcement Services worldwide suspected that fraud and VAT fraud was used to fund terrorism and the war in Syria. But for the first time it has been possible to report independently on direct connections between identified terror suspects and Syrian foreign fighters on one side and fraudulent companies and networks of organized criminals on the other side. The investigation spanned several countries as Denmark, Sweden, Norway, Poland and Spain and utilized data from public registries and criminal records from several other countries.
Technologies used for this project:The investigation was initiated by collecting different datasets from public records, including company registries, bankruptcy records, directories and records from the Danish Food and Drug Administration (Fødevarestyrelsen). Some datasets were scraped, mostly using basic web scraping techniques in Google Drive, using ImportXML and Xpath queries. Datasets were cleaned up using OpenRefine, which can clean up, align and combine different datasets. Notably, the ability to merge datasets with the “cell.cross” command was highly useful. After cleaning up the data it was fed into the social network analysis program Sentinel Visualizer. This made it possible to identify several nodes of importance, including companies, directors, vehicles, addresses, telephone numbers and email addresses. Most importantly, we took to the streets, testing the data. This involved visiting company addresses and the home addresses of the involved directors. This “verification by hand” of the collected data was extremely important. In recent years a lot of communications in Denmark between companies and public authorities have been digitalized. This has its benefits, but also open up a floodgate for fraud. The investigation shows that public authorities only verify claimed information too little and too late.
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