The Bangalore reclassification happened in 2011 when it split into multiple administrative zones, and our data switches from city-level to aggregated-zones starting that year. The aggregates are supposed to be equivalent, but the actual numbers jumped 8%, which might reflect either a genuine change in what's being measured or residual data-quality differences. Should we add a note about this reclassification and its potential impact on trend analysis?
The total doesn't add up to the sum of the parts. Are overlapping categories included?
Good catch.