Greek border guards have been forcing large numbers of refugees back to a dangerous sea in ‘pushback’ operations that violate international law. We proved that a European Union border agency, Frontex, was complicit in the controversial practice. Until then Frontex had denied any knowledge of pushbacks.
The investigation led to the following.
– Frontex chief Fabrice Leggeri was grilled during a hearing in front of the EU Parliament’s LIBE committee. EU Parliamentarians then launched a formal probe into the allegation (which is ongoing).
– The EU Obudsman launched its own investigation into the allegations (ongoing).
– The EU’s Home Affairs Commissioner and the President of the Commission called on Frontex to conducte extraordinary management board meetings to address the allegations.
– Frontex then pledged to reform and bolster its rights monitoring mechanisms and to investigate all incidents. The agency launched major inquiries and created a special Working Group to explore the legality of its operations to date.
– The Greek government, which had previously flatly denied any wrongdoing despite a pile of evidence, has also pledged a thorough investigation.
– The findings of the investigation were picked up by media world-wide and some outlets (from the BBC to the New York Times) conducted their own investigations in ways that amplified the impact and enriched the original story.
– The story also helped raise the awareness among EU citizens of the continuing violations of the human rights of asylum seekers and refugees, while bolstering demands for transparency and accountability at Frontex.
We combined various methods of data collection including interviews with migrants and the tracking data of boats and aircrafts (respectively called AIS and ADSB). We sought to find out whether the aircrafts and boats of Frontex were present when migrants were pushed back. We proved this in a number of cases and, in one case, we were even able to show that a Frontex boat directly blocked the path of a migrant’s dinghy to stop it.
We first collected a list of all the Frontex assets used in the Agegan using mostly open-source data (web searches, social media photo and videos etc). We then obtained AIS data that provides the GPS location of boats (purchased from www.vesselfinder.com) and ADSB data that provides the location of aircrafts (obtained from http://adsbexchange.com). We also obtained the location of pushbacks from several NGOs as well as in interviews with migrants and the metadata of any images and videos they took during their crossing of the Aegean. We plotted all these datasets on several maps, and found the cases where Frontex assets were near pushbacks (this was done with R and Python).
Then, we enriched these findings by analysing videos and photos obtained from migrants in order to understand the details of what had happened in each pushback incidents.
You can read our methodology in more details here, with an emphasis on the visual video analysis: https://www.bellingcat.com/news/2020/10/23/frontex-at-fault-european-border-force-complicit-in-illegal-pushbacks
What was the hardest part of this project?
One of the most difficult aspects of the project was to track down all of the assets owned by Frontex in order to get their tracking data. This was a herculean task which required going through hundreds of Instagram posts, such as those of dining at a restaurant facing the port where some of the Frontex vessels were moored, going through Facebook profiles of staff, etc.
The next hurdle was to obtain the tracking data for the boats and planes, searching for it on various websites and understanding its limitations. The tracking data is generated by most aircrafts and boats in order to avoid crashing into each other. There are many providers of this type of data, each with its own limitations that required a deep technical understanding to apprehend.
We also had to figure out how to combine all of this data and create customised maps in order to make sense of it. The last challenge was to use these computer-generated maps to find the cases where Frontex assets were near pusbacks. This required looking at many maps in order to find cases where Frontex’ boats had suspicious patterns. For example, some boats would go back and forth to make waves and push migrants back towards Turkey. Some planes flew in circles while others were going back and forth in straight lines. We had to understand these patterns in the data in order to grasp how they translated in real life.
What can others learn from this project?
The project was a collaboration of several outlets convened by Lighthouse Reports which included Bellingcat, Der Spiegel “Report Mainz” (a program on ARD, the German public broadcaster) and Japanese broadcaster TV Asahi. A dozen people with very different skills and backgrounds worked together, trusted each other and shared information. This project would not have succeeded without this collaboration.
Another key lesson was that it is possible to combine various types of journalism to create a stronger story. This story cannot fit neatly under one heading: data-journalism, open-source journalism, or a human interest journalism – all of those approaches enabled us to gather solid evidence and build an enticing story.