2017 年 26 巻 1 号 p. 67-83
This study aims to measure the impact of conflict on education outcome in the long term and to find out school factors that can effectively improve academic performance of students who have been exposed to conflict in the past. In order to examine the school factors which mitigate the conflict's negative impact on the education outcome, the research principally employs a multi-level regression model structured with cross-level interaction terms between school level characteristics and individual level conflict experience. The study focuses on the case of Timor-Leste which experienced intensive conflict for its independence in 1999, and used the 2015 EMIS individual data, results of 2015 national examination of 9th grade students completing the basic education cycle as well as the 1999 conflict data.
The study found, first, that students who were born during or before the conflict period (1999-2002) in any provinces of Timor-Leste that were exposed to severe conflict tend to receive lower national examination score in the end of the 9th grade. However, the study also found that the negative impact of a student's conflict experience in Timor-Leste is mitigated if the proportion of permanent teachers (shown as a proxy of teacher quality) in the school increases while the negative impact of conflict, on the contrary, is aggravated if the number of students in a school increases. The study further found that the existence of a school fence (shown as a proxy of school safety) raises average examination score of the school by 0.55 points compared to schools without fences, and this result can be possibly explained by the context of post-conflict situation.
This research contributes not only to making more effective education policy and system in post-conflict countries, but also serves as an additional research example in the field of measuring the impact of conflict on education outcome where not many researches have been conducted up to date.