Immigration and economic growth: do origin and destination matter?

ABSTRACT

This article assesses the heterogeneous effects of immigration on economic growth depending on both the origin and the destination countries. Following the development of a growth model augmented by human capital of immigrants, we estimate it in a dynamic panel setup using the system-GMM estimator. We find that the growth-enhancing effect of immigration is significantly larger when immigration flows from developed to developing economies than when it does to those that include both developed and developing economies. We interpret these results as evidence of immigrants from developed countries bringing with them their advanced knowledge into the developing countries.

KEYWORDS:

• Immigration
• economic growth
• human capital
• institutions

JEL CLASSIFICATION:

• F22
• O15
• O41

Acknowledgements

An earlier version of this paper was written when B.-Y. Kim was a visiting fellow at the Institute of Economic Research, Hitotsubashi University (Kang and Kim, 2012). The authors would like to give thanks to Takashi Kurosaki and anonymous referees for their valuable comments and suggestions.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ The article defines the major industrialized countries as those joined OECD before 1970 and being ranked above 18th place with respect to per capita real GDP, 1960. Appendix A shows the list of the major industrialized countries.

² There are two main definitions of migration, that is, being born in and being a citizen of a foreign country. According to Ozden et al. (2011), the place of birth definition is superior for determining the physical movement of people.

³ As shown in Borjas (1999, 2013), the short-run effect of immigration raises the rate of return to capital and lowers the wage because the ratio of return to capital and the wage are determined by the ratio of capital to labour. Over time, however, the higher rate of return to capital will induce an increase in the size of the capital stock and so the effect of immigration disappears in the long run. Hence, the more immigrants bring in their own physical capital, the smaller the short-run effect of immigration is expected. Note that capital is somewhat immobile across countries, as found in Feldstein and Horioka (1980). That is, an immigrant is not able to carry much physical capital such as machines and buildings as suggested in the theoretical studies (Barro and Sala-i-Martin 1995; Dolado, Goria, and Ichino 1994; Boubtane, Dumont, and Rault, 2016). Along the same line with these studies, we assume that immigrants do not bring a substantial amount of physical capital. This assumption also implies that the amount of physical capital carried by immigrants does not differ depending upon where they come from.

⁴ Conceptually there are the following four distinct types of immigration according to the characteristics of host and origin nations: (1) immigration from MICs to MICs, (2) that from MICs to non-MICs, (3) that from non-MICs to MICs and (4) that from non-MICs to non-MICs. Given that this article focuses on the effects of advanced knowledge and institutions carried by immigrants on growth of host nations, we look at immigration from MICs but exclude immigration from MICs to MICs. The small sample size found in immigration from MICs to MICs was another consideration.

⁵ One may prefer using the data on immigration flow instead of immigration stock in accordance with theory presented in the previous section. However, the data on immigration flow are not available from Global Bilateral Migration Database and thus one should construct the flow data by differencing between the immigration stock in t period from that in (t + 1) period. We justify the use of immigration stock instead of immigration flow in more detail in the text.

⁶ The major industrial countries include a total of 18 nations, namely Australia, Austria, Belgium, Canada, Germany, Denmark, Finland, France, UK, Ireland, Italy, Netherlands, Norway, New Zealand, Sweden, US, Switzerland and Luxembourg.

⁷ Freund and Bolaky (2008) and Chang et al. (2009) show that the growth effect of trade openness is significantly positive only if certain complementary domestic reforms are undertaken, including deregulations of business, financial developments, better education or rule of law, labour market flexibility, and so on.

⁸ Given that the average of $immMI{C}_{i,t}$ to non-MICs is 0.51% in Table 2 and the value of $immMI{C}_{i,t}$ increases two times, the growth rate of $immMI{C}_{i,t}$ amounts to 100% because [(1.02–0.51)/(0.51)]*100 = 100%. This is translated into an increase in real GDP per capita by 11.4% because the coefficient on $immMI{C}_{i,t}$ in column 3 in Table 4 is 0.114.