This blogpost is based on an article (see here for a link to the article ‘Understanding the Impact of Migration on Innovation’) I wrote for a Policy Forum in the June 2014 issue of the Australian Economic Review.
Although immigration debates in the popular press often focus on the perceived negative aspects – e.g. newcomers to a country are ‘stealing’ locals’ jobs –immigrants may actually stimulate innovation, thereby promoting job creation and enhancing productivity in their adopted homeland (in addition to filling shortages for skilled labour). There are a number of reasons why migrants – who could be permanent settlers, temporary workers (or students), or returning expatriates – might be a catalyst for innovation. For example, people movement is one of the main ways in which tacit knowledge moves between regions. Moreover, immigrants bring embodied experience and knowledge from outside cultures and economies. They also act as a circuit breaker for ‘group think’ which would otherwise limit the way societies approach problems.
In this blogpost, I reconsider this issue in light of recent academic research on the topic. In trying to understand the innovation effects of immigration policy, my focus is on economic outcomes associated rather than social outcomes. So I don’t focus on the potential benefits from allowing migrants into Australia for family or humanitarian reasons. To be sure, economic and social outcomes are potentially inter-connected: for example, the presence of an immigrant parent (or grandparent) in Australia might enable a couple with a newborn child to enter back into the labour market more quickly. However, I gloss over these issues here.
Whether or not there are positive effects of immigration on innovation is essentially an empirical issue which can only be ascertained using comprehensive longitudinal data since the effects of immigration on jobs and growth will occur with a considerable lag. Deeper understanding of these relationships could have significant policy implications: for example, it may shape permanent settler policies and settings, temporary worker policies (e.g. working holiday maker visas) and also inducements to encourage select expatriates to return.
If there is an effect, it is important to figure out what proportion can be explained by the fact that people choose to migrate to another country and this is often done on unobservable characteristics (e.g. entrepreneurial spirit) and what proportion is due to the fact that Governments typically place conditions on who can migrate to their country, and this is often done on observable characteristics (e.g. educational attainment, skills).
The mechanism by which immigration might improve innovation (and long-run changes in productivity) is not clear, but there are a few possibilities (direct and indirect). The ‘direct mechanisms’ are: the inflow of skilled migrants may increase the number of researcher workers and thereby innovation. This knowledge might be quite different to the skills and human capital employed by the native research workers, and diversity of skills may be important. The ‘indirect mechanisms’ relate to the nature and range of networks than the immigrants might have. If these are different to the native population, there might exist an opportunity to exploit knowledge spillovers that were otherwise not available.
Recent literature on immigration and innovation using U.S. data finds a positive relationship between the two. Using data from the 2003 National Survey of College Graduates (NSCG), this paper shows that immigrants account for 24% of patents (or twice their share of the population) and the advantage is almost all due to the immigrants’ disproportionately higher share of degrees in science and engineering. Moreover, information on the types of visas provides information on which immigrants are the most innovative. The performance of immigrants (based on the visa on which they first entered the United States) is compared to natives on a range of different innovation outcomes. Although this cannot disentangle the effects of the visa from the immigrant characteristics, it is clear that immigrants whose entry visa was a student/trainee visa or a temporary work visa produce more patents (and patent licenses) and publications than natives. This effect is largely explained by immigrants’ higher education qualifications and field of study.
There is also evidence from an international survey of academics which explores the effects of international mobility on scientific performance. In this study, a comprehensive stratified sample of academic journals in biology, chemistry, earth and environmental sciences, and materials science. For the selected journals, the full list of all publications in 2009 was downloaded and the lead author sent a survey. The country of their current employment was identified and each author was asked questions about their country of origin (where they were living at age 18) and whether they had spent at least 12 months in another country as a student, postdoc or other job. This produced a cross-section of academics from 16 countries and a total of 47,304 individual scientists (with an email address).
Individuals were classified as ‘native non-mobile’ (i.e. natives who have never worked overseas), ‘returnee’ (i.e. natives who have worked overseas but since returned), or ‘foreign born’ (i.e. those who are currently working in a country other than their native country). They then examined the performance of these groups in terms of total citations to their publications and impact factor. They find strong evidence that mobile scientists outperform non-mobile scientists. Moreover, ‘returnees’ outperform ‘foreign born’, so the result doesn’t appear to be a function of selection in terms of who migrates.
Another recent study controls for the fact that individuals choose to live in a specific location (e.g. city) or work in a specific institution (e.g. university). To do this, they compare immigrant and native students within a given university lab. Data were compiled by using PhD dissertation abstracts to construct lists of students. All students were then matched to their advisors and publications, which provide a massive dataset of more than 20,000 PhD students who graduated from U.S. universities between 1999 and 2008. The results indicate a strong positive advantage for Chinese students vis-à-vis all other students in terms of research productivity (24-62%). They attribute these results to a selection effect: to get into a U.S. PhD program, Chinese students predominantly come from the elite Chinese universities (Peking and Tsinghua) who are even more selective than the leading U.S. universities. However, they also note that preferences (for an academic career) and culture (work ethic) may also play roles.
The evidence from these studies suggests an emerging consensus that immigration of skilled workers does have a positive effect on innovation. However, both Government- and self-selection play an important role in explaining the observed effects and it is not always possible to disentangle the two. Of course, these results are based primarily on U.S. data and it is not clear whether these results are generalizable – to the extent that other countries have unique institutions and cultures, and a different mix of unskilled/skilled migrants, it is not clear whether these factors will mitigate the positive results seen in the U.S. Moreover, the U.S. is likely to be the most attractive destination for immigrants, so it is also quite likely that these studies represent the upper bound on the size of the effect.