Temporary international migration, shocks and informal insurance : analysis using panel data

We use panel data for rural Kyrgyzstan to examine households' international migration response when faced with shocks. Using a household fixed effects regression model, we find that while a drought shock increases the likelihood of migration, winter and earthquake shocks reduce the likelihood of migration. We use a simple theoretical framework to illustrate the trade-off between two effects of a shock for a household: loss of income and increase in the need of labor services. We show that migration increases when the former effect of a shock dominates, it reduces when the latter effect dominates. We explore these mechanisms by examining how the migration-response to shocks changes in the presence of alternate coping mechanisms and by evaluating the effect of shocks on a household's decision to send and recall a migrant member. We find that when households have easier access to informal finance the migration-response is muted only for shocks for which the adverse income effect dominates. Our findings also suggest that while shocks for which the loss of income effect dominates have a greater effect on the decision to send a migrant, shocks for which the need of labor services effect dominates only affect the decision to recall a migrant. These findings provide evidence in favor of the proposed mechanisms through which shocks affect temporary migration.