Reframing School Size as Institutional Infrastructure: Moderating Socioeconomic Disadvantage in Rural Schools in South Korea

This study investigates how school size influences the relationship between students’ socioeconomic background and academic achievement, with a particular focus on whether community-linked program participation mediates this effect. Using data from the Korea Educational Longitudinal Study (KELS), we analyze a sample of 5,291 elementary students through a two-way fixed effects model to control for unobserved individual and school characteristics. To address endogeneity, the study adopts school site area (log-transformed) as an instrumental variable and further applies a generalized structural equation model (GSEM) for causal mediation analysis in rural small schools. Results show no significant school size effect in metropolitan and mid-sized cities, but a protective effect of small schools for socioeconomically disadvantaged students in rural areas. Moreover, participation in community-linked programs significantly mediates this relationship. These findings highlight the need to evaluate rural small schools not solely by cost-efficiency but through their structural connection to local communities and educational equity.