[PDF] Immigration and Economic Growth in the OECD Countries 1986-2006

View - Download Immigration and Economic Growth in the OECD Countries 1986-2006

Previous PDF

Next PDF

DISCUSSION PAPER SERIES

Forschungsinstitut

zur Zukunft der Arbeit

Institute for the Study

of Labor

Immigration and Economic Growth in the

OECD Countries, 1986-2006

IZA DP No. 8681

November 2014

Ekrame Boubtane

Jean-Christophe Dumont

Christophe Rault

Immigration and Economic Growth in the

OECD Countries, 1986-2006

Ekrame Boubtane

CERDI, University of Auvergne

and CES,

University of Paris 1

Jean-Christophe Dumont

OECD

Christophe Rault

LEO, University of Orléans (CNRS UMR 7322)

and IZA

Discussion Paper No. 8681

November 2014

IZA

P.O. Box 7240

53072 Bonn

Germany

Phone: +49

-228-3894-0

Fax: +49-228-3894-180

E-mail: iza@iza.org

Any opinions expressed here are those of the author(s) and not those of IZA. Research published in

this series may include views on policy, but the institute itself takes no institutional policy positions.

The IZA research network is committed to the IZA Guiding Principles of Research Integrity.

The Institute for the Study of Labor (IZA) in Bonn is a local and virtual international research center

and a place of communication between science, politics and business. IZA is an independent nonprofit organization supported by Deutsche Post Foundation. The center is associated with the University of Bonn and offers a stimulating research environment through its international network, workshops and conferences, data service, project support, research visits and doctoral program. IZA engages in (i)

original and internationally competitive research in all fields of labor economics, (ii) development of

policy concepts, and (iii) dissemination of research results and concepts to the interested public. IZA Discussion Papers often represent preliminary work and are circulated to encourage discussion. Citation of such a paper should account for its provisional character. A revised version may be available directly from the author.

IZA Discussion Paper No. 8681

November 2014

ABSTRACT

Immigration and Economic Growth in the OECD Countries, 1986
-2006 This paper offers a reappraisal of the impact of migration on economic growth for 22 OECD countries between 1986 -2006 and relies on a unique data set we compiled that allows us to distinguish net migration of the native -born and foreign-born by skill level. Specifically, after introducing migration in an augmented Solow-Swan model, we estimate a dynamic panel model using a system of generalized method of moments (SYS-GMM) to deal with the risk of an endogeneity bias of the migration variables. Two important findings emerge from our analysis. First, there exists a positive impact of migrants' human capital on economic growth. And second, the contribution of immigrants to human capital accumulation tends to dominate the mechanical dilution effect while the net effect is fairly small. This conclusion holds even in countries with highly selective migration policies.

JEL Classification: C23, F22, J24, J61, O41, O47

Keywords: immigration, growth, human capital, generalized methods of moments

Correspond

ing author:

Christophe Rault

LEO, Université d'Orléans, CNRS UMR 7322

Faculté de Droit, d'Economie et de Gestion

Rue de Blois - B.P. 26739

45067 Orléans Cedex 2

France

E-mail: chrault@hotmail.com

We are very grateful to two anonymous referees for their useful comments and suggestions. We would like to thank Fréderic Docquier, Raquel Carrasco, John Martin, Stefano Scarpetta, Gilles Spielvogel, and Bertrand Wigniolle for discussion and helpful comments on earlier drafts. We would also like to thank participants at the European Economic Association (EEA) 2010 Congress in Glasgow, European Society for Population Economics (ESPE) 2011 Conference in Hangzhou, and the International Workshop on Immigration and Economic

Growth 2011 in Amsterdam for helpful

comments. E. Boubtane acknowledges financial support from the Agence Nationale de la Recherche of the French government through the program Investissements d'avenir ANR-10-LABX-14-01. The usual disclaimer applies.

1 Introduction

International migration to OECD countries, notably labour migration, has increased signicantly over the past decades. Between 1997 and 2007, in most southern Euro- pean countries, the United Kingdom, the United States, and several Nordic countries, immigrants contributed to more than 40% of net job creation. In 2007, the share of immigrants in employment reached 12% on average in OECD countries (OECD, 2009). In many developed countries the rst eects of population aging can already be felt in the working age population as baby boomers begin to retire in large numbers while younger cohorts are too small to replace them. In this context, labour migration will continue to play a signicant role in the medium and long term. Specically, interna- tional migration is expected to account for all labour force growth between 2005 and

2020 in the OECD area as a whole.

At the same time, many countries have recently adapted their migration system to make it more selective vis-a-vis skills and education. Traditional settlement countries (Australia, Canada, New Zealand, and the United States) have implemented skills-based migration programmes for a long time which now serve as models to other countries. The United Kingdom, Denmark, and the Netherlands have recently reformed their mi- gration system to give more priority to highly educated migrants through a point-based migration system. Furthermore, the European Union has adopted a new directive, the European Blue Card, to attract highly qualied migrants to the European labour mar- ket. This Directive does not prevent EU Member States from having their own system of national residence permits for highly skilled migrants, but such national permits can- not grant the right of residence in other EU Member States that is guaranteed under the Blue Card Directive. Accordingly, most European countries have also implemented specic migration programmes to attract highly skilled foreign workers. For instance, Austria adopted a point-based immigration scheme-Red-White-Red Card in July 2011. This system aims to attract highly qualied persons and skilled workers in shortage occupations who wish to settle, with their families, permanently in Austria. Also, the United Kingdom changed its point system in 2011 towards greater selectivity. The Highly skilled migrant programme has been replaced by an `Exceptional Talent' visa for applicants who are `internationally recognised leader or emerging leader' in their eld. This trend is most likely to continue, and could even be reinforced, in the future. These changes in migration trends and policies prompted us to reconsider the eco- nomic impact of migration. Empirical economic analyses have been ourishing in recent years in two key areas likely to in uence public opinion on migration, namely the labour market impact of immigrants (Borjas, 2003, 2009; Angrist and Kugler, 2003; Lubotsky,

2007; Ottaviano and Peri, 2008)

1and the scal impact of immigration (Auerbach and

Oreopoulos, 1999; Storesletten, 2000, 2003; Hansen and Lofstrom, 2003).

2However, the

debate is relatively quiet on a third major area of interest: the impact of migration on economic growth. This is precisely the question addressed by this paper.1 See, for instance, Longhiet al.(2005, 2008) for recent meta-analyses.

2See, for instance, Rowthorn (2008) or Leibfritzet al.(2003) for a review.

2 Though there are few doubts about the impact of a labour shock due to migration on aggregate GDP growth, the eect is not so obvious with regard to per capita GDP growth. Indeed, in the standard augmented neoclassical growth model developed by Mankiwet al.(1992), an increase in permanent migration ows has a negative impact on long-term economic growth because of capital dilution, which might be compensated by a positive contribution of new migrants to human capital accumulation (Doladoet al., 1994 ; Barro and Sala-i-Martin, 1995). Consequently, in this framework, whether or not migration positively eects per capita GDP growth crucially depends on the scope of migration and its demographic and educational structures. Due to a lack of harmonized international data on migration, few empirical studies have tried to estimate the impact of permanent immigrant ows on economic growth while accounting for educational attainment. The closest related paper is Doladoet al.(1994). They estimate|as we do herein|a structural model including immigrants' human capital. However, they do not observe migrants' education level and use educa- tional attainment of the population in the country of origin as a proxy. Moreover, their analysis covers the period from 1960{1985, which was characterized (until the second oil shock at the end of the 1970s) by low-skilled migration concentrated in the manu- facturing sector. In the past two decades the characteristics of international migration has evolved considerably and its impact therefore needs to be reconsidered. This is the purpose of this paper. This paper is also related to the recent studies that analyse the eects of economic openness and diversity on GDP per capita. Felbermayret al.(2010) estimate the ef- fect of the stock of migrants on per capita income using cross-sectional country data. Andersen and Dalgaard (2011) consider temporary cross-border ows of people as a measure of global integration and evaluate the eect of the intensity of travel on GDP per capita. Ortega and Peri (2014) estimate the eect of economic openness, jointly considering migration and trade on income per person. In line with studies on birth- place diversity and economic development, they take into account diversity by country of origin within the stock of immigrants. Alesinaet al.(2013) estimate the eect of diversity of migrant birthplaces on growth. They build diversity indicators from data on immigration population by country of birth and education. This paper departs from existing studies by considering the eect of permanent ows of immigrants by country of birth and skill level on productivity growth. We focus on permanent migration|movements that the receiving country considers are for the long term. We exclude temporary visitors (i.e. tourists and businessmen) because we are mainly interested in the economic consequences of long-term immigrants. Moreover, we focus on newly arrived immigrants rather than the immigrant population as a whole. Indeed, the socioeconomic characteristics of immigrants has changed over time and ows better re ect these changes than the stock of immigrants. We also independently iden- tify the eect of net migration of the foreign- and native-born by skill level. 3 Specically, we contribute to the existing literature in two ways. First, we compile a unique data set on net migration that includes data on country of birth and skill level from various data sources for 22 OECD countries between 1986{2006. Moreover, spe- cic attention is devoted to producing robust measures of the educational attainment of recent immigrants as well as native-born expatriates who return to their home country. Second, our estimations are based on the SYS-GMM for dynamic panel data models developed by Arellano and Bover (1995) and Blundell and Bond (1998) that permit one to deal with the (potential) endogeneity of migration variables. However, the con- sistency of the SYS-GMM crucially depends on the validity of the instruments used. Therefore, in contrast to previous studies, we carefully follow some of the recommenda- tions of Roodman (2009) and Bazzi and Clement (2013), introducing both internal and external instruments in our estimation. We think that this way of proceeding is a useful complement to standard specication tests for getting valid instruments and obtaining robust econometric results. Our econometric investigation provides evidence showing that, over the period con- sidered, the impact of migration on productivity growth via human capital accumulation and capital dilution is signicant with the expected signs (i.e. respectively positive and negative). Furthermore, in almost all OECD countries, the former dominates the latter.

Therefore migration

ows tend to have a positive (though small) impact on economic growth, even in countries which have highly selective migration policies. Moreover, simulations based on these results indicate that, all else equal, a one percentage-point increase in foreign-born net migration would have increased productivity growth by three-tenths of a percentage-point per year on average for the 22 OECD countries con- sidered. The remainder of the paper is organised as follows. Section 2 provides a short review of the literature. Section 3 outlines the theoretical model. Section 4 describes the econometric strategy and the data, and presents the empirical results. Section

5 discusses the implications of the results. Finally, Section 6 oers some concluding

remarks. Technical details of the theoretical model and data sources are contained in the appendix.

2 Direct and indirect eects of migration on economic

growth: an overview of the literature International migration potentially has direct and indirect eects on economic growth. Firstly, migration can be viewed as a demographic shock. Indeed, in the textbook Solow- Swan growth model, an increase in migration has a negative impact on the transitional path to the long-term steady state where all per capita variables are nonetheless stable. Even in this framework however, migration aects the age structure of the population of the destination country because migrants tend to be more concentrated in active age groups compared to natives. Consequently migration reduces dependency ratios and 4 potentially has a positive impact on aggregate savings,

3which nally could result in

higher total factor productivity (TFP) growth.

4Yet, this transmission channel has not

been directly considered in the literature. Secondly, migrants arrive with their skills and abilities, which supplement the stock of human capital in the host country. To our knowledge, Doladoet al.(1994) were the rst to introduce migration into the Solow-Swan model augmented by human capital. In this framework the contribution of immigrants to human capital accumulation compen- sates (at least partially) the negative capital dilution eect associated with population growth. The authors estimate their model for 23 OECD countries between 1960 and 1985.
More recently, several authors have included migration in endogenous economic growth models. This literature considers the impact immigrants have on technological progress, notably their contribution to innovation.

5Walz (1995), for instance, intro-

duces migration in a two country endogenous growth model based on Lucas (1998). He nds that the sign of the growth rate eect depends on the initial specialization of the two countries and that migration is selective towards high skilled individuals. Robertson (2002) also analyses the impact of migration in an Uzawa-Lucas model with unskilled labour and shows that an in ow of relatively unskilled immigrants results in lower transitional growth. Lundborg and Segerstrom (2000, 2002) include migration in a quality ladders growth model developed by Grossman and Helpman (1991). They nd that free migration would stimulate growth, especially if it responds to dierences in labour force endowments. Similarly, in an expansion-in-variety framework, Bretschger (2001) shows that skilled migration can promote growth by decreasing the costs of research and development and also by raising the market share of certain types of goods. Most of the previous studies are theoretical and there exist very few empirical as- sessments of the impact of migration on economic growth. Furthermore, when such analyses exist they are not based on structural models and are often hampered by data constraints. For instance, Ortega and Peri (2009) analyse the eects of immigration ows on total employment, physical capital accumulation and TFP in 14 OECD coun- tries between 1980 and 2005. They nd that migration increases employment and capital stocks but doesn't have a signicant eect on TFP. Since immigration shocks lead to an increase in total employment and a proportional response in production, output per capita is not eected by in ows of migrants. However, this study does not take into3 This eect may be partially oset by remittances sent by migrants to their country of origin.

4There is increasing evidence of the impact of changes in age structure of the population on produc-

tivity (Sarel, 1995; Lindh and Malmberg, 1999; Kogel, 2005; Feyrer, 2007).

5Hunt and Gauthier-Loiselle (2008) provide recent evidence on the impact of highly skilled migration

in the United States on innovation. They nd that a one percentage-point rise in the share of immigrant

college graduates in the population increases patents per capita by 6%. 5 account the human capital of migrants,

6or their diversity by country of origin. More

recently, Felbermayret al.(2010) and Ortega and Peri (2014) use bilateral migration stocks around the year 2000 to estimate a positive relationship between the immigrant share of the population and GDP per capita in the host country. Moreover, Ortega and Peri (2014) nd that this positive relationship is magnied when diversity by country of origin within the immigrant population is taken into account. These results are in line with the ndings from studies on birthplace diversity and economic development. For example, Alesinaet al.(2013) nd a positive eect of diversity of immigration pop- ulation by country of birth and education on growth. Another approach is to use time-series analysis. Morley (2006), for instance, analy- ses the causality between migration and economic growth on data for Australia, Canada, and the United States between 1930 and 2002. He nds evidence of long-run causality running from per capita GDP to immigration but not the reverse. In another such example, Boubtaneet al.(2013) nd a positive bidirectional relationship between im- migration and GDP per capita for 22 OECD countries from 1987{2009. Note that due to the lack of harmonised data on characteristics of migration ows, time-series studies do not take into account the educational attainment of immigrants. The main contribution of this paper is to provide robust estimates of the impact of net migration ows on productivity growth, controlling for the skill composition of recent immigrants, through a clear theoretical framework which is presented in the next section.

3 The theoretical model

As in Doladoet al.(1994), migration is introduced in a standard augmented neoclassi- cal Solow-Swan model where aggregate output is produced from physical capital (K), human capital (H) and labour (L) using a Cobb-Douglas function with constant returns to scale:

Y=KH(A L)1+ <1 (1)

whereAis the labour-augmenting (or Harrod-neutral) technological progress. It is a productivity parameter that grows at the constant exponential rategA. The rst channel through which migration aects the economy of the host country is essentially demographic as new in ows of foreign workers fuel labour force growth. This impact can be decomposed between net migration of foreign-born workers (M) and net migration (net return) of native-born workers (E). As we shall see in Section 4.2, it is necessary to make this distinction because the dynamics and the skill composition of these two migration streams are quite dissimilar. Note that net migration is the6 Orece (2010) estimates the impact of migration on economic growth in a gravity model using OECD data on gross migration ows for 24 countries between 1998 and 2007 and uses a proxy for education based on migrant stocks in 2000. The author nds a negative impact of migration on economic growth. 6 dierence between immigration into and emigration from the country during the period. Labour force growth is therefore given by (time subscripts are omitted for convenience): _

L= ~nL+M+E

whereenis the natural population growth rate (i.e. new entries of young people into the labour force minus retirements and deaths notably). We letmbe the net migration rate of the foreign-born (m=M=L) andebe the net migration rate of the native-born (e=E=L). Then, the model follows the Solow model and assumes that the labour force increases at a constant raten=en+m+e. Immigrants and native-born returnees bring their human capital (skills and abilities) that supplements the domestic stock of human capital.

7Inversely, those who leave the

country, take with them their human capital. This is the second channel through which migration impacts production factor endowments in this basic model. We denote byhM the average quantity of human capital that each foreign-born migrant brings along,hE the average human capital of native-born migrants, and ^hthe average human capital per worker (^h=H=L). The accumulation of human capital is thus given by:

H=sHYH+M hM+E hE(2)

=sHYmM+eEH wheresHis the fraction of resources devoted to human capital accumulation,is the rate of depreciation,M=hM=^h(E=hE=^h) is the relative human capital of foreign- born (native-born) migrants compared to the average human capital per worker in the host economy. We assume that the relative human capital of immigrants,mM+eE, is constant. The dynamics of physical capital are the same as in the Solow Model. A fraction s Kof output is saved and capital depreciates at an exogenous rate:8 _

K=sKYK(3)

Using units of eective labour (i.e.yY=AL,kK=AL,hH=AL), the production function is given in intensive form by: y=kh(4)

The evolution of the economy is determined by:

k=sKy(+gA+n)k(5) h=sHy+gA+nm M+eEh(6)7 Migrants are not supposed to bring signicant amounts of physical capital to the economy of the host country.

8Following Mankiwet al.(1992), we assume that human capital depreciates at the same rate as

physical capital. 7

The economy converges to a steady state dened by:

k =sK+gA+n

11sH+gA+n(mM+eE)

1(7) h =sK+gA+n

1sH+gA+n(mM+eE)

11(8) Substituting (7) and (8) into the production function and taking logarithms, the steady state income per eective worker is: lny=1lnsK+1lnsH(9)

1ln(+gA+n)

1ln+gA+nmM+eE

Assuming that all countries are in their steady state, this equation could be used for em- pirical analysis. Rather, we suppose that countries are growing near their steady state. The rate of growth as the economy converges to the steady state can be approximated by:yy =@lny@t ' (lny(t)lny) (10) where= (1)(gA++n) (Cf. appendixA:1). This yields: lny(t)lny=et(lny(0)lny) (11) wherey(0) is income per eective worker at some initial date. Note that, assuming a constant rate of convergenceover time, Equation (11) also holds between datestand t1: lny(t)lny=e(lny(t1)lny) For estimation purposes, we need an expression in terms of income per worker,by, rather than income per eective worker,y. Sinceycan be expressed in term ofby (byY=L), lny(t) = lnby(t)lnA(0)gAtand, using Equation (9), we nally obtain the productivity growth rate: lnby(t)lnby(t1)=gA te(t1) + (1e)lnA(0) (12) (1e)lnby(t1) +(1e)1(lnsKln(+gA+n)) +(1e)1lnsH (1e)1lngA++nmM+eE 8 Equation (12) shows that for a given,,,, andgA, the rate of growth of productiv- ity is negatively related to the net migration rate because of the capital dilution eect associated with labour force growth,n. However, this eect is counterbalanced by the positive impact of the human capital content of migration owsmM+eE. The net eect of migration on productivity growth is therefore ambiguous and depends on the relative human capital contribution of foreign- and native-born migrants (MandE), on the net migration rates (mande) and on the parameters of the production function (and).

In this framework,ceteris paribus, the in

ow of foreign workers will have a posi- tive impact on productivity growth only if new migrants are more qualied than the resident population (M>1) on average. However, this is not a sucient condition as the human capital brought by migrants should also oset the capital dilution eect. Indeed, appendixB:1 shows that, provided there is not a net out ow of human cap- ital associated with total net migration (i.e.mM+eE0),M(+)=is a sucient condition for migration to have a positive impact on productivity growth. Below that threshold the impact will however depend on other parameters of the model. The fact that migration has a positive impact on productivity growth if and only if its contribution to human capital accumulation more than compensates for the eect on capital dilution is a direct consequence of the augmented Solow-Swan theoretical framework. This would not have necessarily been the case in an endogenous growth framework or in a framework in which one considers the imperfect substitution of natives and immigrants in production (Manacordaet al., 2012; Ottaviano and Peri, 2012).

4 Econometric analysis

4.1 Empirical Model Specication

Equation (12) suggests a useful specication for the model that can be used to evaluate the impact of immigration on economic growth in receiving countries. Note that: ln gA++nmM+eE= ln (gA++n)(1mM+eEg A++n) = ln(gA++n) + ln(1mM+eEg

A++n)(13)

One can expect that

mM+eEg

A++nis small.9Using the approximation ln(1x)=x

yields the following equation for the growth rate per worker:9

The relative human capital content of migration

ows is small compared to the sum of the rate

of technical progress, the depreciation rate, and the overall labour force growth rate. The data set,

presented in the next section, indicates that mM+eEg A++nhas a mean value of 0:095 and a standard error of 0:14. 9 lnbyi;tlnbyi;t1=gA t(t1)e 1et lnA(0) (14) 1e lnbyi;t1

1e1lnsKi;t

1e1lnsHi;t

1e+1ln(gA++ni;t)

1e1m i;tMi;t+ei;tEi;tg

A++ni;t

The human capital eect of net migration is captured by mi;tMi;t+ei;tEi;tg

A++ni;t.

Following standard practice in the literature,

10we assume that the convergence pa-

rameteris constant over time and across countries. The termA(0) represents all unobserved elements (e.g. the initial level of technology, resource endowments, climate, institutions, etc.). It suggests the presence of a country-specic eect, which may be correlated with the other explanatory variables considered in the model. The model used to estimate the eect of immigration on productivity growth for a given countryiis a more general form of Equation (14): +6m i;tMi;t+gA+ni;t+7e i;tEi;t+gA+ni;t+i+ t+vi;t(15) whereiand trepresent country-specic and time-specic eects and where1;...,7 are parameters to be estimated.

4.2 Data

We consider a panel of 22 OECD countries between 1986 and 2006. In order to reduce the in uence of short-run variation, we split the sample period into ve sub-periods. Since data are missing for some periods, our panel is unbalanced with between 3 and

5 data points for each country. The list of countries, periods, and data sources of the

quotesdbs_dbs22.pdfusesText_28

[PDF] migration research paper pdf

[PDF] miles flight

[PDF] military 10 codes

[PDF] military alphabet code pdf

[PDF] mille et une nuit pdf tome 1

[PDF] mille et une nuit pdf tome 3

[PDF] mille et une nuit pdf tome 4

[PDF] mille et une nuits galland pdf

[PDF] mille et une nuits mardrus pdf

[PDF] millimeter micrometer nanometer picometer

[PDF] mindset coaching topics

[PDF] mindstorms ev3 advanced programming

[PDF] mindstorms ev3 building instructions pdf

[PDF] minecraft 123 mods

[PDF] mini sumo bot