Resumen
Welfare comparisons may be sensitive to the assumptions made about economies of scale within households. This paper uses recent advances in sequential stochastic dominance techniques to show how to test for the robustness of poverty and housing quality comparisons to assumptions about economies of scale. The method is applied to a comparison of migrant and non-migrant households in Honduras. While simple comparisons based on per capita income and the estimated rental value of the dwelling suggest that migrants do better than non-migrants for poverty, but worse for housing quality, none of the two groups is found to dominate the other for reasonable welfare thresholds when assumptions regarding economies of scale are relaxed.