Zoning Schools by Proximity, Radius, or Network: Spatial Method Performance and its Equity Implications in Langsa City

Abstract

The school zoning system used in Indonesia's New Student Admission (PPDB) is meant to equalise access to education, yet in Langsa City it is still drawn from administrative boundaries that say little about whether students can actually reach a school. This study compares three ways of delineating zones for public junior and senior high schools, judged by how well each covers residential areas: the Thiessen polygon, the buffer, and network analysis (service area). All three were built in a Geographic Information System, with a Model Builder workflow used to automate the mapping. Coverage varied markedly across methods. At the junior high level the Thiessen polygon reached the highest residential coverage at 98.21%, ahead of the service area at 79.63% and the buffer at 75.37%. At the senior high level the ranking changed: the buffer led with 88.50%, followed by the service area at 79.21% and the Thiessen polygon at 71.62%. The reversal shows that no single method is best everywhere; methods that account for the road network and travel distance describe accessibility more faithfully, while purely geometric partitions can mislead where schools are few and clustered. Automating the workflow in Model Builder made the analysis reproducible and transferable to other regions. Together the results offer a defensible, accessibility-based footing for designing fairer educational zones.