Rainfall Shocks and Risky Migration: New Evidence from Nepal Anti-Trafficking Data

Speaker: A/Prof David Smerdon

Affiliation: The University of Queensland

Location: Level 6 Boardroom (629), Colin Clark Building (#39), St Lucia Campus

Teams Meeting: https://teams.microsoft.com/meet/42479477765668?p=Mo75WMOeZIBHDobgN9

Password: VR7WR2Ly

Abstract: 

Nepal's open border with India sustains one of the world's most notorious corridors for cross-border sex trafficking, yet very little is known about who crosses and why.  For example, estimates of the number of Nepali girls trafficked each year run from five thousand to thirty-five thousand. That climate shocks feed this flow is among the most repeated and least tested claims in the surrounding policy field. We study the question using the administrative archive of an anti-trafficking organisation

whose monitors watch Nepal's border crossings: roughly 11,500 recorded interceptions over 2012–2025, with 10,100 linked interviews of the people (overwhelmingly young women and girls) its staff stopped en route. We link interview counts by home district and month to satellite rainfall anomalies across all 77 districts in a fixed-effects Poisson design with distributed lags. Trafficking interceptions rise by 27–32% in total across the seven months following a single month that is unusually wet for that district and time of year. The estimate is stable across thirty-one specification rebuilds, nine

sample definitions and seventy-three covariance estimators, as well as a battery of robustness checks including falsification and permutation placebos. Using what limited data exists in this context in Nepal, we then explore candidate channels for the response: agricultural income, recorded disasters, and rainfall's effect on the organisation’s own monitoring and recording practices. To our knowledge, this is the first study to provide causal evidence of a significant push-factor for human trafficking in Nepal.