Skilled migration and innovation go hand in hand, right?
Perhaps not say Kirk Doran, Alexander Gelber and Adam Isen. In their paper, “The Effect of High-Skilled Immigration on Patenting and Employment: Evidence from H-1B Visa Lotteries“, they find that H-1B migrants (the 457 visa equivalent in the United States) have an ‘insignificant average effect on patenting’. (h/t Ryan Edwards)
Further, their results “generally rule out the claim that an additional approved H-1B visa has no negative effect on the employment of other workers at the same firm” (p.4). Meaning H-1B workers – skilled migrants in large organisations – are substituting other workers.
The interesting thing about this paper is the process used. The authors study firms instead of geographic locations because of their data. They match up data from the Immigration Bureau, the Tax Office and the Patent Office [think for a moment if this was possible in Australia] and include a random sorting process because in 2006-07, companies oversubscribed to the H-1B lotto. This means its possible to compare winners and losers of which companies get to hire migrants.
These results are a serious addition to the literature and should be engaged with. The authors acknowledge how their findings go against the grain of some other research where patents have been associated with migrants but their study has an excellent methodology.
This research is difficult to apply to Australia. In the U.S. about 60 per cent of H-1B visas are computer-related, such as software engineers. In Australia, the comparative figure is about 10 per cent. While there are similarities – highly skilled relative to the native population, higher than average salaries, a majority of visas being granted onshore (continuing employment) – the nature of skilled migration in the U.S. and Australia is different. Australia has a much more diverse range of migrants by occupation type than the U.S. and the link to permanent residency through employer-sponsored is much stronger.
This does not mean we should ignore this research. Indeed, Australian policy-makers should be attempting to try and figure out exactly these questions by matching processes. The ABS is currently working on a matching process using ATO and DIBP data for migrant incomes. One assumes there is little standing in the way of a similar exercise to look at companies.
Hopefully more work on these questions will emerge, providing a detailed picture on the micro-effects of migration across different labour markets.