Indirect Fertility Estimates in Historical Italy: A Bayesian Approach

Michail Raftakis , University of Bologna
Riccardo Omenti, University of Bologna
Nicola Barban, University of Bologna

This paper examines Italy’s comparatively late demographic transition, marked by substantial subnational disparities, by developing the first long-run, province-level series of overall fertility that explicitly incorporates male fertility. Using the rich but under-digitized historical censuses of ISTAT, we employ large language models to transcribe and harmonize tabulations and apply a Bayesian hierarchical framework to generate indirect fertility indicators for men and women across 1861–1961. These estimates allow us to chart the evolution and spatial dispersion of fertility trajectories and to quantify systematic sex differences in timing and intensity, yielding a more granular portrait of Italy’s transition and a replicable approach to subnational demographic reconstruction. Preliminary results show that subnational fertility trajectories were far from uniform: the more urbanized and industrial North-West entered decline first, while in the broader North and Centre the First World War and the 1918–19 influenza pandemic marked an inflection after which fertility either continued its pre-war slide or stabilized at lower levels. By contrast, several Southern provinces had reverted to pre-war fertility by 1931 as wartime disruptions receded and traditional family formation resumed. Over time, these divergent paths widened dispersion and produced a pronounced North–South gradient, yielding an increasingly polarized fertility map. Throughout, male fertility levels exceeded female levels, consistent with older paternal ages, remarriage, and a wider male reproductive span.

See extended abstract

 Presented in Session 70. Flash Session New Data, Methods and Comparative Perspectives in Historical Demography