The current project aims to Develop estimators which enhance the "standard" techniques of small area estimation (synthetic estimators, GREGs, composite estimators) four major ways:
- by borrowing strength over time
- by borrowing strength over space (e.g. taking account of spatial autocorrelation and allowing for the modifiable area unit problem)
- by developing sample design criteria that are optimal for small area estimation
- provide improved estimates of cross-classifications (using a modified version of the SPREE approach.
Assess the effectiveness of the standard estimators and the enhanced techniques using data on unemployment, income and household composition from a representative group of six European countries. Provide software implementations that fit the computing environment of European statistical offices.