Are we sure the metric is being calculated the same way across all rows? Let's spot-check a few.
The Barcelona revision in 2019 was stated as a 6% adjustment, but the documentation doesn't say whether it was a correction of error, a redefinition, or a methodological improvement. The type of revision matters: error corrections suggest past figures were wrong, while redefinitions suggest they measured something different. Should we clarify which it was?
Region definition changeI'm convinced by your argument that the regional patterns are statistically significant rather than noise — that's a crucial insight for how we should interpret the trend.