Estimated North Sumatra Province Poor Population Percentage Using Penalized Spline Semiparametric Approach and Small Area Estimation

Abstract

Poverty is one of the many problems that have not been completely resolved by the government in Indonesia, one of which is poverty in the province of North Sumatra. To estimate the percentage of poor people, data is needed in each area using the Small Area Estimation method. Small Area Estimation is used to estimate the parameters of a subpopulation that has a small scope. However, to get a better estimate, you can use an indirect estimation method, one of which is the semiparametric Penalized Spline approach. This method can be used in conjunction with small area estimation because it can connect the two components in the model between the response variable and the predictor variable which is linear and the relationship between the response variable and the predictor variable is non-linear. Based on the small area estimation model with a semiparametric penalized spline approach, the best is found in model 4 with a coefficient of determination value of 0.645 where the value is close to 1, which means the results are good to use. The average poor population in North Sumatra province is estimated at 15.38%, the highest poor population is in Pakpak Bharat at 22.66% and the lowest estimated poor population is in Deli Serdang at 7.51%.