The Impact of Flooding on Education of Children and Adolescents: Evidence from Pakistan

Abstract

This study traces short- to long-term adverse effects of the colossal flood 2010 on educational outcomes of children and adolescents (age 5–16 years) in the flooded districts of Pakistan. Taking advantage of the flood — a type of quasi-natural experimental research design we utilized a difference-in-differences (DID) approach with inverse probability of treatment weights (IPTWs) to estimate the impact of the flood on educational outcomes by using a household surveys’ dataset (six waves). We compare educational outcomes out-of-school or dropout from school of — children and adolescents in the flooded households with the educational outcomes of individuals of same age groups in the non-flooded households before, during and after the flood. Our findings reveal that, on an average, 39 out of 1000 children and adolescents in the flooded districts, compared with their counterparts in the non-flooded districts, were not admitted in any educational institutions and 16 of them dropped out from schools during the flood. The effect of flood on education of children and adolescents, then, disappeared after 2–4 years after the flood. The education outcomes of children and adolescents in flooded households in rural areas compared with their peers in non-flooded districts were severely affected by the flood. Mirroring the impact of flood on education sector to the current heavy flood 2022 in Pakistan or pandemic COVID-19 is similarly compelling nations around the world for closure of their schools and educational institutions. The findings of this study may have some policy implications in terms of identifying the most vulnerable children and adolescents to mitigate the adverse impact of the natural disasters such as flood or pandemic on education outcomes and particularly significant to pinpoint shocks of disasters that have large and long-run impacts on human capital accumulation.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 19 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.