Existing work examining the trade effect of commonwealth membership does not account for sample selection, unobserved heterogeneity and multilateral resistance in estimation, leading to biased estimates. Our analyses improve on all these fronts. Unlike earlier work, we also consider services...