With exact totals, SSA tables allow opposite changes in annual name turnover between 1980–1989 and 2008–2017

We tested whether public name counts determine the direction of change in annual name turnover. We used babynames 1.0.1, exact applicant totals, and 1,060,844 published rows for 1980–2017. The model assumes that every absent positive count is below five and puts no extra constraint on hidden names across years. Turnover is total variation between adjacent birth-year distributions. We compared nine annual transitions in 1980–1989 with nine in 2008–2017. Two complete records preserved every published count, every total, and the same complete count spectrum in each year. They differed only in which hidden singleton names recurred. One record gave a fall of 6.54 percentage points for female records and 2.47 points for male records. The other gave rises of 9.08 and 7.84 points. Thus the stated public information did not determine the direction of change. These values are feasible examples, not sharp contrast endpoints or estimates of actual hidden names. We also give an exact annual upper bound and a conservative lower bound. Exact fractions checked all 74 transitions. Counts agreed with exports from two prior SCI studies. A repeat matched all six main analysis outputs byte for byte. The package contains sources, provenance, code, results, proofs, figures, and offline instructions. The result requires exact matching totals and does not identify a cause of name choice.

baby names · cultural change · total variation · count suppression · partial identification · data audit