Infertility in Italy: Calibrating Nonprobability Survey Data for Population-Level Estimates

Maria Francesca Morabito , University of Florence
Emilia Rocco, University of Florence
Valentina Tocchioni, University of Florence

Infertility represents an increasing concern in high-income countries and a growing focus of research. Delayed childbearing, combined with the natural decline in fecundity with age, has increased the likelihood of failed conception attempts among couples, yet prevalence remains uncertain due to heterogeneous definitions and methods. Awareness of the magnitude of infertility is essential for managing its consequences, which range from medical aspects (e.g., recourse to Assisted Reproductive Technology) to individual psychological well-being. This study aims to provide a robust estimate of infertility in Italy by integrating data from a nationally representative survey based on probabilistic sampling with data from a non-probability online survey. A doubly robust estimation approach is applied to infer population parameters from the non-probability sample, adjusting for selection bias through auxiliary variables common to both data sources, both of which refer to the same target population. This approach leverages the cost-efficiency and timeliness of the nonprobability survey while maintaining the representativeness of the probabilistic data. Our study offers two contributions: (1) robust national estimates of the couple infertility as defined by the WHO for Italy, a country with a positive desired–actual fertility gap, and (2) a framework that is transferable to other national contexts, based on readily collected data and a coherent definition of infertility.

See extended abstract

 Presented in Session 26. Flash Session Emerging Data Sources in Demography: Digital Traces, AI and Mobile Phone Data