In Discussion Paper No. 1400, Stefan Manzocchi and Research Affiliate Philippe Martin conduct empirical tests on data form a cross-section of 33 developing countries to analyse whether the allocation of capital across these countries has been consistent with theoretical predictions. Under the assumption that human capital cannot be used as collateral for external borrowing, Manzocchi and Martin identify determinants of capital movements in the context of an augmented Solow model, and derive a testable reduced-form equation for the amount of net foreign borrowing per capita.

Manzocchi and Martin argue that these results are consistent over the period 1960–82 with the augmented Solow model with partial capital mobility: capital has been directed towards countries with low initial income levels, high human capital and low political instability. Inclusion of the post-1982 period weakens the results significantly, however. This strongly suggests that determinants of net capital flows to developing countries have changed in the aftermath of the world debt crisis of the early 1980s. An interesting by-product of the study is the result that net borrowing was positively correlated with growth only between 1960 and 1972.

Are Capital Flows Consistent with the Neoclassical Growth Model? Evidence from a Cross-section of Developing Countries
Stefan Manzocchi and Philippe Martin