Producing reliable regional life tables is essential for capturing the disparities in mortality and life expectancy at the regional scale. These disparities provide critical insight for assessing demographic and actuarial risks that influence public health systems, social security, and financial institution schemes. In Algeria, the absence of regional life tables published by the Office for National Statistics (ONS) constrains the analysis and understanding of mortality patterns and life expectancy differences across regions. To address this gap, this study estimates mortality rates across seven Algerian regions (North Central Region, Northeast Region, Northwest Region, Central Highlands, Eastern Highlands, Western Highlands, and Great South) using data from the latest Multiple Indicator Cluster Surveys (MICS4 and MICS6), combined with official national mortality data from the ONS. However, the raw mortality rates derived from MICSs are sparse and noisy and require adjustment to ensure smoothness and completeness of mortality curves using a Bayesian relational model. Our results indicate the existence of substantial disparities in mortality and life expectancy across regions. The Great South exhibits the highest mortality levels and the lowest life expectancy for both sexes, while the Western Highlands show the most favourable profiles. These findings identify regions facing elevated demographic and actuarial risk, providing insights for insurance modelling, pension planning, and public health design. By combining survey data with statistical modelling, this study provides a practical framework for measuring and managing demographic risks in data-limited contexts.
Previous Article in event
Previous Article in session
Next Article in event
Assessing Regional Mortality Risks in Algeria under Data Sparsity: A Bayesian Relational Approach
Published:
01 July 2026
by MDPI
in The 1st International Online Conference on Risks
session Actuarial Science
Abstract:
Keywords: Keywords: Regional life tables, Mortality estimation, Bayesian modelling, Relational models, Sparse data, Subnational mortality, MICS surveys, Algeria.
