Many countries are concerned about losing their best scientists, engineers, and other skilled workers to emigration to foreign countries and the US in particular. It is plainly the case that many skilled workers cross national borders. The evidence regarding the brain drain from Europe to the US is surveyed in Saint-Paul (2008). The foreign born represent more than a third of PhD holders in the US Science and Engineering workforce (NSF 2007). Furthermore, these migrants make a disproportionate contribution to US science and innovation (Levin and Stephan 1999 Hunt forthcoming).
The initial literature on the brain drain was pessimistic about the welfare effects of skilled migration for source countries (see e.g Bhagwati 1976). Recent contributions have instead emphasised channels through which source countries may benefit from skilled migration, such as incentives to acquire skill given uncertain migration prospects (Mountford 1997) and diasporic networks acting as knowledge banks (Kerr 2008, Agrawal et al. 2008).
The net welfare benefit of skilled migration also depends critically on the quantity and quality of migrants who eventually return to the source country. Because migrants have likely enhanced their skills during the migration episode, the benefits from return migration may be large enough to outweigh the costs of the brain drain for source countries, even if only a fraction return, as argued in a number of recent theoretical papers (Mayr and Peri 2008, Santos and Postel-Vinay 2003) or qualitative ones (Saxenian and Hsu 2001, Saxenian 2005).
Due to intrinsic difficulties in following workers as they move across countries, the available empirical evidence on return migration of skilled workers is very limited. In a recent paper (Gaule 2010), I deploy a novel approach for measuring return migration. By focusing on academic scientists, I am able to use publicly available academic records to reconstruct career histories. I rely on the availability of fine-grained biographical data collected biennially by the American Chemical Society to guide students in their choice of graduate schools. I also take advantage of the fact that the main output of academic scientists – scientific publications – can be observed.
My hand-collected data includes 1,953 individuals and covers extensively foreign faculty affiliated with a US PhD-granting chemistry, chemical engineering, or biochemistry departments between 1993 and 2010. About half of the individuals in the sample came to the US as graduate students, one third came as postdoctoral research fellows and the rest as faculty.
The odds of returning home are less than 10%
The incidence of return migration in my sample is low. Among foreign faculty who had their first US faculty appointment after 1993, 4.5% have returned to their home country by 2010. Using out-of-sample predictions, I estimate that a further 4.3% will return to their home country before the age of 65, assuming no change in trend in future years.
Distinguishing by source country, the incidence of return migration is relatively high for Australia, Canada, and European countries but very low for China and India. In fact, I observe only one return to India and three to China, despite the fact the Chinese and Indians are the largest groups in my sample.
Figure 1. Incidence of return migration by source country (normalised by the size of the risk set)
The most productive scientists are less likely to return
I investigate the relationship between ability and the propensity to return using scientific publications weighted by journal impact factors as a measure of ability. Modelling the return migration decision as a risk in a discrete hazard model with source country fixed effects, year fixed effects and controls for age, I find that that the most productive scientists are less likely to return. This result confirms earlier findings of negative self-selection into return migration (Borjas 1989, Borjas and Bratsberg 1996) but is derived using a more direct approach.
Conclusion
The location decisions of academic scientists implicitly reveal their preferences. For the majority of them, either (a) the disutility of living in the US relative to the home country is lower (in absolute value) than the professional advantages, pecuniary, reputational or otherwise, of working in the US or (b) there is no disutility of living in the US relative to the home country.
The fact that the most successful scientists are less likely to return suggests that the professional advantages of staying in the US are relatively more important for them than for less successful scientists. This result is consistent with other studies showing that the problem of the brain drain is more pronounced in the right tail of the productivity and skill distribution (Saint-Paul 2004, Commander et al. 2008).
The new evidence presented in this column is worrisome for the perspective of source countries. However, it pertains only to migrant scientists who became faculty in the US. Whether it can be generalised to other groups of skilled migrants is an open question.
