Methodologies, tools and techniques in small area estimation: an overview

Small area estimation has received numerous attention in recent decade due to increasing demand for reliable small area statistics from both public and private sectors. Traditional direct estimation requires the domain-specific sufficiently large sample, and in reality domain-specific sample data are not large enough for all small areas (even zero for some small areas) to provide adequate statistical precision of their estimates. This makes it necessary to 'borrow strength' from data on related multiple characteristics and/or auxiliary variables from other neighboring areas through appropriate models, leading to indirect estimates. Typically, there are two types of methodologies that have been developed for indirect small area estimation - which are the statistical model-based and economic microsimulation approaches. The first methods are based on explicit area level and unit level models that have been studied through various statistical tools and techniques including simple ratio estimates, (empirical-) best linear unbiased prediction (E-BLUP), empirical Bayes (EB) and hierarchical Bayes (HB). In contrast, the spatial microsimulation based approaches are using different reweighting techniques such as GREGWT and combinatorial optimization (CO). This presentation will provide an overview of those methodologies, tools and techniques in small area estimation, along with a critical comparison between different reweighting techniques.