The implementation of empirical best linear unbiased prediction-Fay Herriot (EBLUP-FH) on the estimation of average per capita expenditure at district level in West Sumatra Province in 2019

Abstract

The problem of equitable distribution of welfare is still a sustainable development agenda that must be completed by all parties, including local governments. An indicator to describe the welfare of the population in an area is per capita income and per capita expenditure. The lack of data given by Statistics Indonesia has caused the government to not be optimal in policymaking and implementation because they require data presentation at a smaller regional level. This study aims to estimate per capita expenditure at the district level in West Sumatra Province in 2019 using the SAE EBLUP-FH method. Based on the results of the EBLUP-FH estimation, the distribution of per capita expenditure in West Sumatra Province is very varied among districts. Several districts that are geographically close to the provincial capital area tend to have a higher average per capita expenditure than other areas. Based on the comparison between the direct estimator and the EBLUP-FH estimator, it was found that the RRMSE value of the EBLUP-FH estimator was smaller than the direct estimator. It can be said that the EBLUP-FH method can provide more precise estimation results.