Theoretical and empirical research in economics suggests that bilateral migration triggers bilateral trade through a number of channels. This paper assesses the functional form of the impact of migration on trade flows in a quasi‐experimental setting. We provide evidence that the relationship is not log‐linear. In particular, at small levels of immigration (stocks) the elasticity of trade to migration is quite high, and it declines to zero at about 4,000 immigrants. If immigration stocks exceed such a level, the evidence suggests that trade will not increase anymore. This suggests that for cross‐country network and other effects flowing from immigration to materialise at a significant level for trade, a high‐enough level of immigrant stocks is necessary. But there appears to be satiation as immigrant numbers increase.

No abstract

No abstract