Estimating Mortality at Extreme Ages. Does Data Structure Matter? Evidence from Italian Women Aged 105+
Using classic parametric mortality models, including the Gompertz and exponential models, we evaluate how the absence of survivor data affects parameter estimates, uncertainty, and model-based conclusions. By systematically contrasting truncated and non-truncated settings, this study provides a robust empirical foundation for contexts where only death data are available and contributes new evidence to the debate on whether human mortality reaches a plateau among the oldest-old.
Authors
- Giorgia Capacci
- Graziella Caselli
- Carlo-Giovanni Camarda
- Silvia Capuano
- Marco Battaglini
Institutions
- Institut national d'études démographiques (FR)
- Sapienza University of Rome (IT)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-04-22
- DOI
- https://doi.org/10.5281/zenodo.19697405
- Primary Topic
- Insurance, Mortality, Demography, Risk Management
- Type
- article
- Field-Weighted Citation Impact
- 0.00