SMALL AREA ESTIMATION DENGAN METODE EBLUP FAY-HERRIOT PADA TINGKAT KEMISKINAN DI PROVINSI SUMATERA UTARA

Abstract

National socio-economic survey in March 2019 showed that the number of poor people in North Sumatra Province was 1,282 million people or 8,83 percent of the total population. The number of poor people and the percentage of poor people are obtained by direct estimation based on the survey conducted. Direct estimation in a small area has a poor accuracy because the estimator have a large variance. The method that can be used to solve this problem is indirect estimation with Small Area Estimation (SAE). This study uses SAE with the Empirical Best Linear Unbiased Prediction (EBLUP) Fay-Herriot method to estimate the parameter of the poverty level in North Sumatra Province by city. The data used is the poverty percentage of 33 cities using seven co-variables namely the human development index, gross enrollment ratio for elementary, junior, high school, and college, income per capita, and life expectancy in North Sumatra Province obtained from the survey results from the BPS. This study shows that SAE with the EBLUP Fay-Herriot method produces a better estimation value than the direct estimation