Bayesian Spatio-Temporal Small Area Estimation Using INLA: An Application to Household Income in Spain

Tutor: Jonatan Andrey Gonzalez Monsalve

This thesis will investigate Bayesian spatio-temporal models for small area estimation using R-INLA. Real data from the Spanish Household Panel (2016–2023), provided by the Institute for Fiscal Studies, will be used. Direct estimators and hierarchical models incorporating spatial and temporal dependence will be compared in terms of estimation accuracy, predictive performance, and uncertainty quantification.