Does geographic mobility contribute to educational outcomes: is migration timing the missing link in internal migration

Abstract The rapid economic growth and urbanization have significantly increased geographic mobility among individuals and families. While international migration has long been a topic of debate, internal migrations are gaining increasing attention. Using data from China Migration Dynamic Survey, this paper pays special attention to the role of internal migration in educational attainment of the next generation. Specifically, by examining the migration-induced (non-)disruption in schooling, this study aims to control certain confounding factors and explore the role of migration timing as a potential missing link. The findings reveal that non-disruptive migration (before children aged 5) results in increased children’s schooling years for both left-behind children (LBC) and migrant children (MC). Conversely, disruptive migration tends to have no significant effects on children’s education in most cases. To bolster the robustness of these findings, an alternative sample of younger children aged 6–20 is constructed to assess potential dropout rates. Additionally, to address concerns regarding endogenous issues, we instrument the (non)disruption with the average (non-)disruption rate and the employment rate of the migration year in the hometown and destination respectively for MC and LBC. Meanwhile, propensity score matching (PSM) technique is employed to gauge the average treatment effect by comparing the disrupted and non-disrupted. The results consistently reaffirm the main findings, highlighting the importance of interventions to mitigate disadvantages associated with migration timing. In particular, policies should focus on children who experience migration during their schooling years and ensure supportive educational environments to minimize potential disruptions.