Analyzing corrected votes in the EU Parliament

Contains Python notebooks with the data research for the following stories:

• Här är de Svenska ledamöterna som oftast trycker fel i EU-parlamentet: ”Det är för mycket”
• Svenska ledamöter är näst sämst i EU på att rösta – trycker på fel knapp och rättar i efterhand

Contact [email protected] for questions.

1. Collect data.ipynb

Run this notebook to scrape MEP + voting data from VoteWatch. Data will be stored locally as csv files (under data).

2. Analyze.ibynb

Does some very basic aggreagtion of the data, focusing on how often members of parliament correct their votes, a proxy for how often they "vote wrong".

The notebook exports aggregated data to csv files (output) containing the following columns:

• Number of votes in total
• Number of votings when MEP was present. Excludes all votings when the MEP has been absent.
• Presence (%): Share of votings when MEP was not absent.
• Particiation (%): Excludes not only absence, but also votings when MEP has been registred to not have voted.
• Number of 'wrong votes': "Wrong votes" are defined as votings where the MEP has been present and corrected a vote after the session.
• Share of votes that were 'wrong' (of the votings where the MEP was present)
• Number of corrected votes in total: This column also include corrected votes made after absence.
• Share of votes that were corrected (of the votings where the MEP was present)