Misused EU funds
Organisation: Direkt36 (Hungary)
Publication Date: 04/08/2016
Size of team/newsroom:small
DescriptionThe story is a classic case of "follow the money" investigations. We exposed how substantial amount of European Union funds went to companies that later (sometime not much time later) went bankrupt. EU subsidies play a very important role in the Hungarian economy. We wanted to show how this money is spent and we were also looking for potential flaws in the system. The story was produced for a general audience but of course we knew that it can draw special attention from EU circles. We were informed that after its publication, one Hungarian MEP turned to OLAF (EU's Anti-Fraud Office) because he thought that our story should prompt an investigation. OLAF is very secretive so we don't know whether they indeed started an investigation.
What makes this project innovative? What was its impact?In Hungary, data journalism and data-driven reporting has still not taken roots widely. The foundation of our story was, however, a massive data analysis. We scraped and cleaned the whole EU grant database then linked it with the company registry to identify the grant recipients who went bankrupt. This analysis was done on hundreds of thousands of records. As part of the in-depth analysis we were also looking for cases in which the company went bankrupt not long after receiving the EU money. These were especially interesting because they raised the suspicion of fraudulent activities. As we reported, one of these companies is owned by a businessman who is now the CEO of a major state company. Even though he is a public official now, his own company swallowed EU funds that most likely will never be recovered.
Technologies used for this project:For the scraping, we relied on the work of Daniel Sparing, a data analyst who was helping us as a volunteer. He used Linux for collecting data on the pages of each grant then transformed each site with Linux to a csv file. Then he used R for producing one big csv file with all the data inside. In order to link this dataset with the company registry, we also needed extra help. The EU grant database contains only company names but no company IDs like a tax number. This was a huge obstacle because without an ID that we have both in the EU grant system and in the company registry it would have been very difficult to link the two separate databases. We could have tried to link the two by using the company names but that would not have been very reliable (as we know, text is always the dirtiest part of data). Fortunately, a research group called Microdata at the Central European University helped us out. They had a method to identify and add tax IDs for companies in the EU grant database. Once we had this, we could link that database with the company registry. For cleaning the data, we used Openrefine. For linking and analysing the datasets, we simply used Excel (vlookup and other functions).
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