Preliminary

Combining a unique panel dataset on international migration into 34 OECD countries from 223 origins from 1985 to 2010 with data on Foreign Direct Investments (FDI) we investigate how migrant diasporas attract FDI to their origin countries. Our data enables us to control for cultural similarities between countries, an identification issue which previous studies have left largely unaddressed. We find that in the aggregate, migrant diasporas indeed contribute to attracting investment, but not by much: a 1 % increase in the migration stock from country i to country j increases the FDI stock from country j to i with 0.1 % (0.3 % without controlling for common culture). This paper is currently being revised to include new results. A summary of these result (in extension to the abstract above): […] Beneath this low coefficient lies substantial variation: Instrumenting both migration and FDI, we find evidence of economically large two-way causal effects for the treated groups. To reconcile these results, we present (preliminary) evidence that FDI and migration are "information substitutes": For a country pair, migration fosters initial FDI, but once FDI reaches a critical stock, it no longer has an effect. This nonlinearity suggests that when investing in a foreign country, firms can either learn about the profitability of an investment from migrants, or they can observe how other firms from their country fare in that market