Within the framework of Rubin's causal model, this paper estimates the effects of
international migration on the welfare of Bangladeshi migrant households. Moving from the
estimation of the average effect, the paper disaggregates the impact on the basis of
households' quartile of expenditure and length of the migration period. The no-migration
counterfactual scenario is then used to measure the effect on inequality and to build a
transition matrix showing the relationship between migration/remittances and social
mobility. The paper argues that those who benefit most from migration are the relatively
better off households and that migration and remittances are both a source of inequality and
a vehicle of social mobility. Finally, since most of the characteristics which seem to
determine the probability of migration cannot be affected by governmental policies, it is also
argued that the resources deployed for pro-migration policies cannot directly benefit the
poorer sections of the population.