Penerapan Regresi Logistik dan Regresi Spasial pada Penetapan Kebijakan Anggaran Pendidikan (Studi Kasus: Alokasi Anggaran Pendidikan Kabupaten/Kota untuk Wilayah Indonesia Bagian Timur)

Abstract

Constitutional Court Decision No. 13 / PUU-VI I in 2008, the government must provide education budget at least 20% from national budget and regional budget. However, in reality there are some areas that have not been allocated a budget of education at least 20%. Variables education budget is divided into two. First, education budget which have surpass 20% and education budget which below 20% of regiomal budget. Second, the percentage of education budget allocation from total regional budget. Logistic regression is used to observe the effect of the regional characteristics to fulfillment of the education budget policy in accordance with the decision of the Constitutional Court. Some of the characteristics examined include: population based of education, revenue-sharing, special allocation fund, the general allocation fund, poverty, and enrollment. Unlike the Logistic Regression, spatial regression is used to see the effect of the spatial variables used. From the analysis of logistic regression, there are three variables that affect the determination of the education budget policy. The third variable is the general allocation fund, a special allocation fund, and area. While the spatial regression analysis, obtained five variables that affect the percentage of the education budget allocation. These five variables include: revenue-sharing, general allocation fund, a special allocation fund, the number of people who graduated from college, and enrollment at age 19-24.