Small area estimation : estimating selected economic statistics for provinces of the Philippines

Abstract

Different methods of small area estimation are discussed in detail together with a summary of the applications for each method. A case study consisting of an application to small area estimation of selected Philippine economic data (employment, salary and value of output) are described. Seven different estimators are applied to this problem of estimating intercensal small area characteristics; 3 synthetic estimators, a SPREE estimator, a RATIO-SPREE estimator and 2 regression-based estimators. In addition, estimates of the variance of each estimator are derived. The various estimators are evaluated on the basis of their performance in replicating the 1988 Census of Establishments (CE) preliminary estimates. The absolute relative error percentage and the root mean square error are used in assessing the performance of each estimator. The SPREE is a best choice based on percentage absolute relative error while the RATIO-SPREE estimator is the most accurate in terms of root mean square error.