Schooling and destiny : behind the inequality in education performance in Indonesia

Abstract

In recent decades, important policies and programs-including nine-year compulsory education, pro-poor programs and the improvement of teacher quality-have been implemented to improve the availability and quality of schooling in Indonesia. However, to date, comprehensive research on the equity of the outcomes of these policies across the many schools and districts in Indonesia has never been conducted. This study aims to carry out a comprehensive analysis of education outcomes in Indonesia, covering the changes and patterns of education resources, the trends, patterns and determinants of school progression, and the patterns and determinants of student performance as measured by national exam scores. Survival and multilevel analyses are applied on a novel dataset constructed from a number of secondary datasets gathered from various sources. These secondary datasets include administrative data on schools, teachers and national exams, as collated by the Ministry of Education and Culture (MoEC), the National Socio-Economic Survey (SUSENAS), the National Labour Force Surveys (SAKERNAS), the Village Census (PODES) and the administrative data on education finance from the Ministry of Finance. This study finds a continuous improvement over time in education progression, both in administrative data and household survey data. The equality of education survival across communities, for example, between urban and rural areas, between males and females, and across socio-economic statuses of the community, has been improving. However, the results also reveal that Indonesia still faces serious problems in school discontinuation. The decision to continue schooling is found to be affected not only by individual and household characteristics but also the characteristics of the districts in which the children reside. Gender is an important factor in school continuation, but the effect of gender varies across districts. Socio-economic characteristics of households measured by parents' education and economic status play an important role in determining school continuation. Against expectation, this study finds a trivial association between student continuation and education resources when measured by budget allocation and the availability of schools at the district level. Analyses of student performance suggest that student performance measured by national exam scores at junior secondary school in four tested subjects (Indonesian language, mathematics, science and language) varies across schools and across districts. Student performance is affected by the gender of the students and the effect varies across schools and districts. Students from higher socio-economic families perform significantly better than those from lower socio-economic families. Students perform even better if the schools have a larger proportion of students from higher socio-economic families. Unexpectedly, controlling for parental characteristics, school input has only a small influence on student performance. There is also a contradictory result that the budget for education allocated by the district government is negatively associated with all performance, except Indonesian language. Levels of gender development at the district level have a significant influence only on maths performance, but it is the performance of males, not females, in Indonesian language and English that is more influenced by the level of gender development in the labour market.