Méthodologie

Méthodologie

Transparency Register

Données de LobbyFacts : On a utilisé un jeu de données de LobbyFacts pour chaque batch (2019, 2021, 2022, 2024). - On récupère les budgets

The EU Transparency Register is a joint initiative by the European Commission and European Parliament (the EU Council belatedly joined in 2021) which offers registrants the opportunity to publish data on their EU lobby activities. The Register is voluntary, although lobbyists are incentivised to join. Information on the official EU Register is provided by registrants themselves, making it the sole responsibility of those organisations. It is recognised that some entries in the official register are inaccurate, incomplete or otherwise misleading. LobbyFacts presents some data differently from the original EU Register. Since September 2021, lobby consultancies are only required to declare their annual revenue from lobbying in ridiculously large bracketed thresholds, which includes the bracket 1 million euros+. In order to provide a more precise figure, LobbyFacts calculates a total for such registrants based on a tally of the individual sums declared per client per year. As these figures are provided in brackets ie. 100,000-199,999 we take the lower figure. This sum is then provided as the field ‘Lobbying costs for closed financial year’. It is a more precise figure than the total on the official register. But as we use the lower of the bracketed figures, likely the LobbyFacts figure is a major under-estimate.

Furthermore since September 2021, non-commercial organisations now need to provide a total annual budget while other organisations have to provide an annual lobby spend. For non-commercial organisations, only their historical lobby spend will be available, and they will not appear in rankings according to lobby spend after September 2021. Data taken from the official EU Register also include several hundred ‘duplications’, where organisations have joined the register and then re-registered under a different identification number. It is ongoing work in progress for LobbyFacts to link multiple entries from the same registrant together. Campaigners consistently demand more resources for the secretariat of the EU Register to monitor entries and enforce the rules to boost the quality of the data. Data on the official Register is in constant flux as new organisations join and registrants update their entries; the data on LobbyFacts reflect these changes. Our website is refreshed once a day, following publication of the official daily xml file update. This synchronisation process might explain any short-term differences between data on LobbyFacts and data on the official Register. LobbyFacts has been collecting these daily xmls since February 2012. However, we are aware of some gaps in our data archive. Most importantly, between 20 September 2021 and 19 March 2022, all registrants were required to update their EU Register entries according to a new format. During this period, and in fact until 13 May 2022, the daily xmls did not include all financial data from all registrants. Until the Register secretariat remedies this and provides the missing data, LobbyFacts has decided to only show data until 19 September 2021, and then again from 13 May 2022 onwards. There are other, smaller gaps in our data collection, including from 2013.

https://transparency-register.europa.eu/guidance/guidelines_en#ref-4-how-to-register

  • Règle de l’UE sur l’enregistrement auprès du transparency register

Donc on se retrouve avec des grosses différences pour les ONG entre avant 2021 et après. Ex = BEUC : déclaré par LobbyFacts (2020) : 3 millions. Déclaration en 2024 : 6.680.000 GreenPeace : 600.000 selon LobbyFacts (2020). Déclaration en 2024 : 1.842.157

Pour les traitements et les analyses (2022, 2024), nous avons utilisé la variable de LobbyFacts (lobbying.costs).

  • Et les RDV avec les top-management de la CE

Since 1 December 2014, the European Commission has published online lists of all meetings that Commissioners, their Cabinet members, and the Directors-General have had with organisations in the register. The data on LobbyFacts is extracted from Commission datasets (and LobbyFacts then matches reported lobbyists with data from the official EU Register). LobbyFacts includes all published meetings since 1 December 2014. The meetings data is available under Search, while individual datacards for each registrant also provide a full list of its meetings held. Since 1 January 2025 the Commission has additionally published information about the lobby meetings held by its officials with “management functions”. This information is also now published on LobbyFacts, as part of the high-level meeting data. There are a few important things to note about all of this meetings data: a registrant may well have had other lobby meetings with lower-level officials in the Commission, but the only published data covers elite officials only.

the data that the Commission publishes may not be perfect. It could include duplications, omissions, and delays. For example, a meeting attended by two separate officials can sometimes be reported twice, even though only one meeting was actually held, which can distort the figures. Nonetheless, in the view of LobbyFacts, whether or not a registrant has met with the elite of the Commission, is one good indicator of lobby influence in Brussels. (LobbyFacts website)

  • Sur les rencontres avec les députés européens.

Integrity Watch propose de regarder les rencontres avec des lobbies, mais fait une récolte quotidienne de cette (page)[https://www.europarl.europa.eu/meps/en/full-list], car le jeu de données n’est pas compilé par l’UE. Il faudrait demander à Integrity Watch de nous donner des jeux de données, mais dans tous les cas on n’a pas d’info avant juin 2019.

Partie de Cécile

  • Sur les groupes experts

Est-ce qu’on a la bonne information sur les groupes d’expert ? Et si oui, quelle est sa granularité ? (Annuelle, instantanée au moment de la création du jeu de données) Est-ce que les membres des sous-groupes sont présents ? Ou les membres intermittents des groupes ? On peut essayer de vérifier cette information pour les données que l’on a. -> Observers / members…

TODO

  • Le “Gros tableau” pour chaque année (avec les variables ci-dessous / catégories d’organisation).

  • Pour 2019 et 2021, ajouter ep.passes

  • Faire une études des “ressources” : ressource (Lobbying.costs + ETP + Bruxelles) / Institutionnelle (OPC, ROADMAPS, In_EG, Meetings, InterGroup, Forum, EP Passes).

  • Certains tableaux sont pas lisibles

  • Corrélation et régression au sein des catégories

  • Faire l’ensemble des statistiques descriptives en fonction des catégories d’organisation

  • Quels sont les variables qui concourent le plus à la participation à un grand nombre de groupe d’expert ?

  • Heatmap pour 2022, 2021, 2019

  • Le “Gros tableau” pour chaque année (avec les variables ci-dessous / catégories d’organisation)

  • Une partie description, par de lien avec les groupes d’expert

  • Une partie corrélation groupe d’expert et les variables ressources, sur la population et dans les catégories

  • Une partie pour voir comment toutes les variables concourent à la participation des organisations au groupe d’expert (multivariée : ACM / modélisation)

  • Une partie sur les full-players

  • Faire les traitements de 2022 en 2024 et inversement

  • + Utiliser la variable catégories d’organisation pour chaque corrélation que l’on veut tester

  • Est-ce que les analyses sont compréhensibles

  • Tableaux à corriger