Sensitivity for private mechanisms for one YEAR, one PUMA

The max_records_per_individual for the whole big dataset is 7, so one person can appear seven times (in 7 years I guess) in the database. Can we assume that the sensitivity for any sub table associated to one particular year and PUMA is 1 ?


Yep, on this data-set that’s a correct assumption… in fact, it’s a little more generous than that: Each individual is associated with at most one record per year (across all PUMAs).