Market share evolution

Haldir Aybara Why are there gaps in the time series every July? Is the data collection paused?

Looking at the Réunion dataset, the figures jump by 18% in 2005 with no documented cause, which is large enough to question whether we're measuring the same thing before and after. The jump coincides with a reported administrative restructuring, but we lack the bridge data to verify whether it's a genuine change or a definitional shift. Should we investigate whether Réunion's methodology was revised at that time?

Is headcount per capita across the whole window?