$th International Conference on Economic Statistics: Statistical models for the economic transition: the new challenge in a developing world

A Bayesian spatio-temporal extension of the Fay-Herriot small area estimation model is presented, integrating temporal dynamics via evolving random coefficients and spatial dependence through a distance-based structure. The model improves the estimate reliability by borrowing strength across areas and time, capturing both overall trends and yearly increments. Its effectiveness is demonstrated using agricultural economic data characterized by spatial and temporal correlations.