In a competitive constant-returns small open economy world prices are the appropriate shadow prices for traded goods in public sector cost-benefit analysis. If there are at least as many factors of production as there are goods produced and traded, it is straightforward to derive 'foreign-exchange-equivalent' factor shadow prices. Bertrand (American Economic Review, 1979) has argued that the case of more goods being produced and traded than there are factors (referred to as 'diversification' below) is of greater empirical relevance, and that in this case it is impossible to define factor shadow prices. In this paper, I argue that this view is incorrect, in the sense that if the economic forces which bring about the diversification of production are consistently and explicitly modelled, just enough information is available to define the factor shadow prices. The cases considered are: (i) a small open economy in which the government's objective is to influence the distribution of income, in which case diversification could only result from an irrational use of the policy tools available to the government; (ii) a small open economy in which there are political constraints on production levels in some activities, in which case world prices are not the correct shadow prices for the constrained activities; and (iii) a closed economy (which can be thought of as a model in which world prices are endogenous) in which the shadow prices of goods are not world prices.
