Hierarchical Bayesian Approach for Addressing Multiple Objectives in Poverty Research for Small Areas.

Nowadays the information extracted from data should be the key to good policy, therefore, analysts must make the best possible use of all available information. However, data availability often is limited by cost or for other reasons. Consequently, there is the need to use data from different sources. Our goals are to develop hierarchical models and to demonstrate their ability to improve inferences about quantities for which there are meager data. When a hierarchical model can be found to rep- resent the situation properly, analysis of that model often can be used to extract most or all of the relevant information and so provide the best possible estimates. The application considered will include small area estimation in the context of the EU Statistics on Income and Living Con- ditions. In developing the hierarchical model, we use together survey data and population registers. As for the implementation of the hierarchical model, we propose to use Bayesian methodology assisted by Monte Carlo Markov Chain.