Auto-regressive Distributed Lag Model on Climate Change Impact on Outmigration in Nepal

Abstract

Climate change exacerbates Nepal's exposure to environmental hazards like landslides, floods, and droughts, which degrade agricultural productivity and drive forced displacement, particularly from rural areas to cities and overseas. This study investigates the interplay of climatic variables (maximum/minimum temperatures, rainfall) and economic indicators (paddy yield, GDP per capita) on human outmigration from 1993 to 2021. Employing time-series analyses, including augmented Dickey-Fuller tests for stationarity, correlation matrices, autoregressive distributed lag (ARDL) models, and error correction mechanisms, we disentangle long- and short-term dynamics. Long-term findings reveal significant relationships (5 percent level) between outmigration and both current and lagged GDP per capita, lagged maximum temperature, lagged rainfall, and lagged migration rates, highlighting delayed climatic impacts and persistent economic influences. Short-term analyses identify GDP per capita (5 percent significance) and prior migration trends (10 percent significance) as immediate drivers, emphasizing path dependency in migration decisions. Granger causality tests demonstrate unidirectional links: minimum temperature Granger-causes outmigration, GDP per capita influences maximum temperature, and paddy yield affects both minimum temperature and GDP per capita (all 5 percent significance). These results underscore the dual pressures of economic precarity and climatic stressors, where declining agricultural productivity and erratic weather patterns synergistically exacerbate migration. The study advocates for integrated policy.