Based on a welfare-maximization model of skilled migration where education generates a positive externality, this paper examines whether the early view regarding brain drain's (BD) negative impact on source countries and the Bhagwati tax (BT) associated with it, is compatible with the recent more optimistic BD-induced brain gain view. I derive BD's impact on education, welfare, optimal education subsidy (s), and a combination of s and BT, when residents' (emigrants') weight in the government's objective function is 1 (1 − β), with β ε [0,1]. I find that: i) education, welfare and s are higher (lower) under an open than under a closed economy for 1 − β larger (smaller) than the ratio of source-country to host-country income; ii) s and BT are 'policy complements,' i.e., they are positively related; and iii) BT increases with β and reaches a maximum at β = 1. Two implications and a proposal are: a) The early literature focused on resident – rather than on migrant – welfare (the β = 1 case), which is precisely where the optimal BT is largest; b) A second policy instrument should be useful, especially if there are constraints on making changes in the other one. Thus, as opening up the economy implies a lower s, raising BT should be beneficial if, say, parents' and teachers' organizations make it politically difficult if not impossible to reduce s; c) A proposal for collecting the tax is presented.