Project description:

Le Monde had access to the flight data of some 800 aircraft transiting 50 of France’s main business airports and some 15 foreign equivalents between the start of the French lockdown on March 17 and the end of travel restrictions on June 1.

Among those flights, we found out that several CEOs used their private jet to reach their resorts and thus not respecting the law.

Impact reached:

The autorities said they will do extra-checks on the travel reasons, after people were angry finding out the travel restrictions were not applied to CEOs.

Most of the top CEOs we named have asked not to be tracked anymore by the websites like FlightRadar, etc. (which we weren’t using).

Techniques/technologies used:

Emmanuel and I used mainly Python and R, for the first part of the investigation : namely crossing the registration number of the jets that have flown during the lockdown with their registered owner (we scraped lots of national plane-owners registeries), or with other files (the Paradise Papers, Russian individuals and entities subject to UE restrictive measures, FAA Block list…).

After that, we also used Python and R to make a scoring of the flights that we should fully investigate : for instance flights that were leaving Paris on Friday and coming back on Sunday or Monday, suggesting someone was leaving for the weekend, or flights going to festive places like Ibiza or Cannes, flights that used jets (instead of big commercial aircrafts).

From the 800 flights we had at the beginning, we picked 20 flights to fully investigate (according to the scoring and the registered owner), and we investigated online (using osint, checking social medias of the people we thought were on the flight, findng witnesses who posted from the airport when an flight-of-interest was landing, using satellites images to check if the resorts were occupied by the owners, etc.) and offline (going in the field to find witnesses). This last step was necessary to check who was on the plane (because it’s not always the plane owner).

What was the hardest part of this project?

We found 800 flights to investigate. There was too many we could humanily check, so the main part of the work (and the most difficult) has been to find databases to cross the registration number of the planes in order to create a scoring that could help us find the ones that had the most potential to create a breaking news story. It worked, since, using this method, we found several top-CEOs using their jets to go get some fresh air during the lockdown, while saying they were travelling for important work reasons.

This kind of investigation would never have worked without open data (the radar data from ADSB-Exchange) and without our technical skills, since the private jet business is precisely selling its clients (top CEOs) the fact that no one will ever know the latter used their services. Human sources are very rare in this kind of environement.

What can others learn from this project?

Other journalists can replicate our investigation in their own country, applying the same method as we did : getting data from ADSB Exchange, applying the same scoring method and then investigating each flight-of-interest using osint and field work.

The subject has not aged at all : the travel restrictions are more than ever in place. France just announced it would not welcome anymore most travellers from outside the EU, for instance.

Project links:

www.lemonde.fr/economie/article/2020/10/14/comment-au-printemps-et-malgre-les-restrictions-de-deplacements-de-grands-patrons-ont-voyage-en-jets-prives_6055911_3234.html

docs.google.com/document/d/19GvyOi9pjLlqz3uam1D3T95aUUX1TngRsqwHsh5wes8/edit?usp=sharing