Analisis Regresi Kuantil B-Splines Monoton Naik pada Hubungan Rata-Rata Lama Sekolah dan Pengeluaran Rumah Tangga Per Kapita di Provinsi Yogyakarta

Abstract

The multidimensional problem in various countries that is always become the government's attention is the problem of poverty, Indonesia is no exception. Poverty is often associated with the education sector due to the function of education as a driving force of the transformation of society to break the chain of poverty. The pattern of relations between poverty and the education sector can be seen from the relationship between the level of education (mean years of schooling) and poverty level (per capita household expenditure). DI Yogyakarta is still the province with the largest percentage of poverty on the Java island despite showing a downward trend since 2007. This study aims to look at the relationship between the level of education (mean years of schooling) and poverty level (per capita household expenditure) in DI Yogyakarta Province 2016. The model that is suitable for determining household characteristics is quantile regression with the Increased monotone B-Splines method that links the mean years of schooling and per capita household expenditure. Estimation results based on the quantile regression model with Increased monotone B-Splines method found that households with the lowest education level are said to be very poor households if monthly per capita expenditure is less than 322,205 rupiah and is said to be a poor household if monthly per capita expenditure is between 322,205 rupiah to 426,666 rupiah. Meanwhile, households with the highest level of education are said to be very poor households if monthly per capita expenditure is less than 3,410,965 rupiahs and is said to be a poor household if monthly per capita expenditure is between 3,410,965 rupiahs up to 4,676,718 rupiahs