Development Accounting in a Heckscher-Ohlin World

This paper tries to contribute to the strand of literature that investigates the question to what extend differences in per capita income between countries are due to differences in factor endowments like human- and physical capital on the one hand and due to differences in technology on the other hand. In particular, I am trying to assess to what extend structural transformation, ie the ability of a country to specialize in the production of goods that intensively use the factors with which it is abundantly endowed, has an important role in determining cross country income differences. I find that when productivities are country specific, for realistic parameter values structural transformation plays little role and productivity differences between countries remain large. However, when I allow for factor augmenting technology differences and factors are complementary in sectoral production, there seem to be large differences in the productivity of physical capital that are strongly correlated with per capita income, while human capital seems to have an inverse hump shape. This result is ad odds with Caselli (2005), who finds that poor countries use capital more efficiently than rich countries, while having a lower productivity of human capital. Finally, I use trade data and the Heckscher-Ohlin-Vanek equations to assess the plausibility of my calibrations and find a good fit for the model with factor specific productivities and complementary factors.