Does immigration have a Matthew Effect? A cross-national analysis of international migration and international income inequality, 1960–2005

Social Science Research 42 (2013) 683–697

Contents lists available at SciVerse ScienceDirect

Social Science Research journal homepage: www.elsevier.com/locate/ssresearch

Does immigration have a Matthew Effect? A cross-national analysis of international migration and international income inequality, 1960–2005 Matthew R. Sanderson Kansas State University, Department of Sociology, 204 Waters Hall, Manhattan, KS 66506, USA

a r t i c l e

i n f o

Article history: Received 6 October 2011 Revised 18 October 2012 Accepted 9 December 2012 Available online 21 December 2012 Keywords: Migration Development Globalization Inequality

a b s t r a c t This paper empirically assesses how immigration affects international inequality by testing the relationship between immigration and national economic development across countries in different world income groups. A series of cross-national, longitudinal analyses demonstrate that, on average, immigration has a rather small, but positive long-term effect on development levels. However, the findings also indicate that immigration has a Matthew Effect (Merton, 1968) in the world-economy: immigration disproportionately benefits higher-income countries. Moreover, the wealthiest countries reap the largest gains from immigration. Thus, from the perspective of destination countries, immigration does not appear to be a panacea for international inequality. Instead, the results indicate that immigration actually reproduces, and even exacerbates, international inequality. Ó 2012 Elsevier Inc. All rights reserved.

1. Introduction Persistent, and in many cases expanding, income gaps between countries are perhaps the most pressing issue confronting world society (Milanovic, 2007). International migration is emerging as a means of ameliorating these gaps as key international organizations place migration squarely at the front of the international development agenda. The United Nations has spearheaded the movement, beginning in 1994 with the Population and Development Conference in Cairo, which issued a Program of Action calling for, ‘‘. . .orderly international migration that can have positive impacts on both the communities of origin and the communities of destination’’ (UN, 1994). Since Cairo, the United Nations has led the formation of the Global Commission on International Migration, and initiated the High-Level Dialogue on International Migration and Development, the Global Migration Group, and the Global Forum on Migration and Development, each of which has been tasked, in different respects, with addressing the relationship between migration and development. As a result of these efforts, the migration-development nexus is now prominent in international policy circles. Over the last few years, for example, migration and development has been the topic of the United Nations’ Human Development Report (2009a) and the World Bank’s Global Economic Prospects Report (2006). Although migration has only recently emerged as a development tool, its importance for development has long been recognized. Indeed, limitations on emigration were a central tenet of mercantilist policies in Europe from the 16th to 18th centuries. Absolutist states believed that political power was directly tied to the size of the population. Large populations meant a larger supply of labor, a larger internal market, and a larger military. From this perspective, migration was seen as a zerosum proposition at the international level. The loss of population through emigration was tantamount to the loss of power and states attempted to control, to the extent possible, emigration of their populations out of a fear of being disempowered vis-à-vis rival states. E-mail address: [email protected] 0049-089X/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ssresearch.2012.12.004

684

M.R. Sanderson / Social Science Research 42 (2013) 683–697

Mercantilist perspectives on population and development lost sway to the more laissez faire approaches of the physiocrats, and subsequently, to classical and neoclassical economics, which today promotes the liberalization of markets for production factors, including labor. Migration is argued to make economies more efficient by moving labor resources from areas of lower productivity to areas of higher productivity. Thus, liberalizing international migration is a means of generating and maximizing wealth for both sending and receiving countries. From this perspective, then, migration is seen as a positive-sum proposition at the international level; it is a win–win for the world. Receiving countries gain laborers and the attendant benefits of population growth identified by the Mercantilists, but sending countries also gain from the financial remittances sent back by migrants abroad and from tighter labor markets at home, which entail higher average wages. Disagreements over the consequences of migration for national economic development, and thus, for international inequality, remain prominent. Neoclassical economic prescriptions to liberalize migration flows are becoming ascendant as the search continues for alternative means of addressing seemingly intractable levels international inequality: ‘‘In spite of a myriad of development projects, there is still a great chasm between rich and poor countries. However, one tool has as yet been left untried in attempts to bridge the divide: labor mobility’’ (Clemens, 2010). However, Mercantilist concerns about the loss of population and disempowerment of sending states vis-à-vis receiving states are the basis for contemporary concerns over ‘‘brain drain,’’ ‘‘brain gain,’’ the skill-based selective migration policies of many countries in the Global North, and other concerns about whether migration really is a positive-sum proposition at the international level: ‘‘International migration today is not an instrument for reducing inequality’’ (Castles, 2007, p. 4). How does international migration affect international inequality? Despite its emergence on the international development agenda, our understanding the relationship between international migration and international inequality remains limited by the lack of cross-national, empirical analysis. Migration is still most commonly treated as a national-level, or at best a bilateral, issue despite the fact that it is now very much an international issue. The sources and the destinations of migration are much more geographically diverse, now including the majority of all countries in the world (Castles and Miller, 2003), and the structural motivations and consequences are now situated, in part, at the cross-national level of analysis (Sanderson and Kentor, 2008, 2009; Sanderson, 2010). Moreover, the movement to elevate migration as a development strategy has gained momentum rather quickly, outpacing empirical scrutiny. Indeed, the lack of evidence supporting migration as a development strategy, and more broadly, as a means of ameliorating cross-national inequalities, is remarkable given the salience of the topic in current international discourse. This paper addresses these shortcomings by empirically analyzing the long-term relationship between immigration and national economic development. Cross-country, or international, income differences are a function, in large part, of different rates of growth within countries. Thus, the question of whether international migration affects international inequality can be addressed by empirically assessing the effect of migration on economic development levels across countries. This is the approach taken here. I begin by incorporating theory and prior research from neoclassical economics and international political economy theory, two disparate strands of research that have not engaged in much cross-fertilization, into an integrated theoretical framework. I then use this framework to develop two hypotheses: (1) immigration will have a positive, long-term effect on development levels in receiving countries; (2) but this positive effect will be larger in magnitude in countries with higher average incomes.

2. Immigration, development, and international inequality Historical evidence strongly suggests that national economic development under capitalism depends, in large part, upon the existence of a sufficient supply of cheap, flexible labor (Castles and Miller, 2003). Capitalist economies rest upon a fundamental distinction between capital and labor (Marx, 1967). Capital is a fixed factor of production; it cannot be idled without imposing direct costs on its owners. Labor, however, is a variable factor of production; it can be idled as necessary to maintain profitability. The availability of a sufficient supply of cheap, flexible labor is thus crucial for capitalist economic growth because it smooths out fluctuations in the business cycle and restrains wage growth, both of which further the process of capital accumulation. Marx (1967, p. 423) makes this clear: ‘‘. . .the characteristic course of modern industry, its decennial cycles of periods of average activity, production at high pressure, crisis and stagnation, depends on the continuous formation, the greater or less absorption, and the reconstitution of the industrial reserve army, composed of the surplus population. . . the expansions are impossible unless there is available human material, unless there has been an increase in the number of available workers irrespective of the absolute growth in the population. . .’’ Initially, capitalist production draws on internal supplies of labor, but eventually accumulation rests on the ability to procure supplies of foreign, immigrant labor. Lewis (1954) describes the role of a cheap, flexible supply of labor in his classic dual sector model of economic development. The development of national capitalist economies is predicated on the existence of dual sectors within the country: a capitalist sector and a non-capitalist, subsistence sector. The capitalist sector develops, or expands, only by incorporating labor from the subsistence sector. Early on, the capitalist sector can expand without increasing wages because there is essentially an unlimited supply of labor in the subsistence sector. By keeping wages low, this unlimited supply of labor allows higher returns to capital, which promotes a self-sustaining cycle of growth through accumulation. Capitalists reinvest some portion of the surplus value in production in order to further accumulation, which stimulates employment demand and draws in more labor from the subsistence sector. Eventually, however, growth in

M.R. Sanderson / Social Science Research 42 (2013) 683–697

685

the capitalist sector effectively exhausts the supply of surplus labor from the internal subsistence sector, causing wages to increase (i.e., the Lewis turning point) (Ranis and Fei, 1961). In response to upward pressure on wages, capitalists search for new sources of labor abroad, stimulating inflows of foreign labor and effectively creating an international labor supply system (Portes and Walton, 1981; Sassen, 1988). The availability of an essentially unlimited international supply of labor was a key impetus to capitalist economic development in the Global North. Thomas’s (1973) seminal analysis of the 19th and early-20th century Atlantic economy demonstrates the profound role of immigration in US economic development. Thomas (1973) demonstrates that major inflows of immigrant labor in the US preceded expansions in fixed capital investments and building activity. The very availability of such large supplies of immigrant labor stimulated both the large-scale investments in fixed capital and the innovations in production methods that furthered capital accumulation and economic growth: ‘‘The introduction of automatic machines eliminating human skill over a wide field was stimulated by the incursion of such a considerable volume of cheap labor; it was profitable to invest in capital-saving equipment. . .thanks to immigration of cheap labor, the US was able to take full advantage of those innovations by ‘widening’ her capital structure with enormous benefit to her physical productivity and economic power’’ (Thomas, 1973, p. 165). More recent studies lend further support to earlier historical studies while more precisely estimating the economic effects of immigration. Much of this research focuses on how immigration affects the wages of residents in host countries. Borjas (1995) describes the gains from immigration in terms of an immigration surplus: the increase in national income that accrues to native residents as a result of immigration. Most studies estimate the immigration surplus to be positive, but relatively small, compared to the size of the host economy. For example, following several assumptions about the elasticity of wages and labor supply compositions, Borjas (1995) estimates the immigration surplus to be approximately 0.1% of GDP in the US, which in a $14 trillion economy equates to roughly $14 billion per year or $47 per person. Understanding how immigration affects wages is a step toward understanding the broader effect of immigration on aggregate incomes (Dustmann et al., 2008), and a series of econometric studies have taken this next step. This research generally shows that immigration raises per capita incomes, usually by 1–2% of GDP, in a wide array of host countries. Most of this research, however, is cross-sectional, which cannot shed light on the long-term implications of immigration, and is limited to developed countries, including Norway, Sweden, and Finland (Feridun, 2004, 2005, 2007), the US (Borjas, 1995, 2003; Ottaviano and Peri, 2008), Canada and Mexico (Aydemir and Borjas, 2007), Spain (de Arce and Mahia, forthcoming), the UK and Ireland (Barrell et al., 2010), and Greece (Cholezas and Tsakloglou, 2009; Lianos et al., 1996). An exception to the exclusive focus on developed countries is a recent study of Thailand (Pholphirul and Rukumnuaykit, 2009). There are a few econometric studies that further expand the scope of the analysis from one country to multiple countries. The World Bank’s World Economic Prospects report (2006) included a global simulation that estimated incomes in developed countries would increase by 0.4% given a 3% increase in immigration from less-developed countries. Two background papers for the UN’s Human Development Report (2009a) provide similar findings. One study found that developed countries would capture approximately one-fifth of the increase in income resulting from a 5% increase in immigration, resulting in gain of approximately $190 billion (van der Mensbrugghe and Roland-Holst, 2009). Another study analyzed annual immigration flows in 14 OECD countries over the period 1980–2005 and found that a 1% increase in immigration raised per capita incomes by 1% (Ortega and Peri, 2009). Again, however, these studies are restricted largely to developed countries, limiting the cross-national applicability of the findings. In perhaps the only previous cross-national study to include less-developed countries, Felbermayr et al. (2010) estimate a model on 163 countries in 2000 and found that immigration has a positive, but relatively small, effect on per capita incomes, increasing incomes by 0.22%. That is, on average, a 10% increase in the stock of immigrants would be associated with an increase in per capita incomes of 2.2%. There is a broad consensus that the relationship between immigration and economic development becomes evident only over a long-term time horizon (Goldin et al., 2011). Thomas (1973) found that the time lag between inflows of immigrants and economic expansion in the US was quite long, often requiring years to manifest, because immigration induces structural transformations in the basic foundations of the economy, including capital-labor relations and production techniques. Recent research more precisely estimates that gains in productivity, output, and income from immigration require approximately 10 years to manifest, as firms restructure production operations to incorporate immigrant labor, and workers, both immigrant and native, transition toward new employment opportunities requiring different skill levels (Peri, 2010; Zavodny, 2011). Immigration is thus a key, but relatively under-investigated, explanation of national economic development. Prior research demonstrates that countries that have been able to draw on an international supply of flexible, cheap labor have been able to expand their economies, and that immigration has a positive, but relatively small, effect on average incomes in destination countries. Previous research, however, has neglected the world-economic context in which international migration occurs and therefore overlooked the possibility that immigration may not be uniformly beneficial for all countries. Indeed, there is reason to believe that wealthier countries gain disproportionately from immigration. High-income, or core, countries maintain their advantaged position vis-à-vis other countries coercively, through political-military force, but also by monopolizing new technologies that drive leading sectors (Modelski and Thompson, 1996) and keep them at the more profitable end of the product cycle (Vernon, 1966). Leading sectors are the ‘‘basic stimulants’’ of national economies and leading, core economies are the ‘‘spark plugs of the world economy’’ (Modelski and Thompson, 1996, p. 71). Control over new forms of technology and innovation produces monopoly profits in industries employing these technologies but also generates huge spillover effects throughout the economy. The economic surplus generated from these leading sectors then finances the

686

M.R. Sanderson / Social Science Research 42 (2013) 683–697

expansion of coercive, military-political power, which further supports the advantaged position of core countries. As leading sector technologies and innovations diffuse outward from core countries to more peripheral countries, competition increases, profitability levels decline, and growth rates slow (Modelski and Thompson, 1996). Immigration therefore likely has positive, but uneven, effects across broad income groups in the world-economy. On the one hand, immigration into core countries may stimulate relatively higher levels of innovation and investment because these countries are at the leading edge of the product cycle. Economic production in wealthier core countries is driven by industries in leading sectors that generate relatively higher levels of surplus value. These sectors have more highly-developed technological infrastructures and more highly-educated labor forces that augment the investment and innovation-inducing effects of immigration on development levels. On the other hand, immigration into more peripheral countries may also promote development, but the investment and innovation-inducing effects of immigration are likely weaker, or are dampened, by the lower surplus value-generating forms of economic production that predominate in these economies. This paper tests these propositions in a series of cross-national, longitudinal analyses. 3. Data and method 3.1. Key variables and hypotheses The dependent variable is GDP per capita in constant 2000 US dollars. It is taken from the World Development Indicators dataset (World Bank, 2009). GDP per capita is the most widely used measure of economic development, or aggregate living standards, in the cross-national sociological literature. The key independent variable is the stock of international migrants per capita. This variable represents cumulative, long-term immigration into the host country and it is taken from the United Nations Trends in International Migration Stock (2009b). It is logarithmically transformed (ln) to correct for a skewed distribution. The United Nations defines an international migrant as a person who is living in a country outside of their country of birth. These data are compiled from national population censuses and registers that identify migrants by place of birth (i.e., the foreign-born population) or citizenship (i.e., the foreign population). Approximately 78% of the countries with data on the international migrant stock identify migrants based on the place of birth (179 of 230 countries) and the others identify migrants based upon citizenship (42 of 230 countries). International migrant stocks is considered the most ‘‘global and comparable’’ measure of international migration currently available (United Nations, 2008). Moreover, international migrant stocks is particularly well-suited to this study because the time series spans 40 years, allowing an analysis of long-term effects, and it is only available at 5-year intervals, reducing its sensitivity to short-term shocks (United Nations, 2009b). Based upon prior research, I expect that, on average, international migration will have a relatively small, but positive, association with per capita incomes over a decennial time period: Hypothesis 1. Higher levels of international migration will be associated with higher levels of economic development 10 years later. To test whether immigration has differential effects across countries in the world-economy, I place countries into one of three world income groups: low-income countries with incomes below the 25th percentile; middle-income countries with incomes between the 25th and 75th percentiles; and high-income countries with incomes above the 75th percentile. Countries are assigned to world income groupings based upon their income in a particular year (1960, 1970, etc.) in order to allow for the possibility of upward or downward movement in the world income hierarchy. I expect that immigration will be more beneficial for long-term economic development in higher-income countries. That is, I expect immigration will have a Matthew Effect at the international level: Hypothesis 2. International migration should have a larger positive long-term effect in higher-income countries compared to middle-income and lower-income countries. 3.2. Estimation techniques and sensitivity tests This study uses time series of cross-sections data to estimate autoregressive random effects and fixed effects models with a 10-year lag between the independent variables and the dependent variable. This estimation strategy addresses two important methodological issues associated with studying the migration-development relationship over time and across countries. First, a primary purpose of the paper is to estimate the long-term effect of immigration on development. To do this, it is necessary to specify a time lag between the independent and dependent variables. However, lagging the dependent variable is also important methodologically, as it mitigates against endogeneity bias, or reverse causality. I hypothesize that immigration levels will have a positive effect on economic development levels, but it is possible that higher economic development levels have a positive effect on immigration levels. Specifying a contemporaneous relationship between immigration and development leaves open the question of endogeneity bias. For example, it would be difficult to determine the direction of causality if the models related immigration levels in the year 2000 to development levels in the year 2000. This is a problem for cross-sectional research. However, by incorporating time into the model, longitudinal models can reduce the

M.R. Sanderson / Social Science Research 42 (2013) 683–697

687

problem of endogeneity bias by satisfying the time order criterion of causality. For example, it is a non sequitur to state that development levels in the year 2000 influence immigration levels in 1990. Thus, the models include measures of the dependent variable 10 years after the measures of the independent variables. Specifying a 10-year lag is not arbitrary. Prior research has shown that the effects of immigration require approximately 10 years to become evident (Peri, 2010; Thomas, 1973). In addition, a decennial time horizon compromises between a desire to examine the relationship over as long of a time period as possible while also maximizing the number of observations available for the models. Because international migration data are only available in 5-year intervals, the shortest time-frame possible for these models is 5 years, which is not sufficient to examine long-term effects. However, 15-year time lags significantly reduce the number of observations available for each model. Thus, a 10-year time period maximizes the number of observations while allowing any long-term relationship to manifest. Second, in addition to the problem of endogeneity bias, the models address the problem of unobserved heterogeneity. By pooling multiple cross-sections over time, the data incorporate variation over space (i.e., countries) and time (i.e. years), allowing the models to incorporate, or time-invariant, country-specific factors not explicitly included in the model (Stimson, 1985). Unobserved effects are central to the problem of causal inference, and TSCS data provide an advantage over cross-sectional data in addressing the issue of unobservable influences (Halaby, 2004). To correct for heterogeneity bias, fixed and random effects models allow each country to have a unique intercept. The two approaches differ, however, in how each treats the country-specific effects. Fixed effects models transform the estimation equation by using the within transformation to eliminate the time-constant, unit-specific effect. This is equivalent to introducing a series of dummy variables to allow each country to have a unique effect (Wooldridge, 2002, pp. 267 and 274). Random effects models treat the country-specific intercepts as part of the error term and consider them as ‘random’ draws from a larger population (Wooldridge, 2006). A sequence of preliminary specification tests was conducted to determine the appropriate analytic model. The Breusch– Godfrey Lagrange multiplier test and the Wooldridge test for autocorrelation in panel data both indicated that the residuals were autocorrelated. To address serial correlation in the residuals, the residuals are modeled as a first-order autoregressive (AR[1]) process. The Breusch-Pagan Lagrange multiplier test indicated that country-specific heterogeneity is present in the data, suggesting the need to use a random or fixed effects technique. To select between random and fixed effects models, I conducted a Hausman specification test of the assumption that the random effects estimator is uncorrelated with the independent variables. The Hausman test indicated that the random effects estimator is consistent. As a result, I present random effects models with an AR(1) correction. However, in order to test the sensitivity of the results to modeling technique, I also present first-order autogregressive fixed effects models. Fixed effects models do not allow time-constant terms, precluding the possibility of including indicators for world income groups in the models. Thus, I present separate fixed effects models for high-income, middle-income, and low-income countries to test whether the effects of international migration vary across world-income groups (Hypothesis 2). The results presented in the following tables are robust to additional sensitivity tests. Although the analyses incorporate data on the largest possible sample given data availability, it is possible that the estimates might be sensitive to sample composition. Thus, two different resampling techniques were used to estimate the standard errors for the coefficients: bootstrapping and jackknifing. Both techniques are data-dependent, nonparametric approaches to estimating standard errors from the observed distribution of the sample. Bootstrapping constructs samples of the standard errors by taking random draws of N observations from a N-observation dataset. For this analysis, the bootstrapped standard errors were calculated based upon 1000 repetitions of sample size N for each model. Jackknifing repeatedly calculates the statistic by omitting one randomly selected observation from each sample. As a result, jackknifing has been used to check the robustness of the estimates to influential observations in addition to its utility as a resampling technique. Because the data are clustered, one country-year (i.e., observation) was omitted from each sample when calculating the jackknife standard errors for the estimates. The estimates are presented below with standard errors estimated using random and fixed effects models. These estimates were robust to both bootstrap and the jackknife resampling techniques (results from these supplemental analyses are available upon request). 3.3. Countries included in the analyses The analyses include data on all countries with populations greater than one million that have complete data on the variables in the models. The dissolution of the Soviet Union artificially inflated the population of international migrants at the global level, as large proportions of populations living in the former Soviet Union suddenly became ‘international migrants’ in the successor states without necessarily moving into these states. To avoid problems associated with these discontinuities in the data, successor states from the former Soviet Union are excluded from the analyses. Because cross-sections of data are pooled over time, the unit of analysis, or observation, is the country-year. Countryyears are selected on the basis of data availability. That is, country-years are included in the analysis if they have information on the independent variables at time t and the dependent variable at time t + 10. On the contrary, a country-year would not be included if it is missing data on the dependent variable at time t or any of the independent variables at time t 10. The overall time span for the study covers the period 1960–2005 because the first wave of data is available in 1960 and the last wave of data is available in 2005. In order to maximize the sample size for each model, the sample size varies across models, a common approach in cross-national sociological research. The set of countries included in the analyses varies across models in order to maximize the amount of information available for each model. Appendix A lists the countries included in the most inclusive model (Models 1 and 2). Together, the set

688

M.R. Sanderson / Social Science Research 42 (2013) 683–697

of countries included in the analyses are not, strictly speaking, a random draw of from population of countries, but it is generally representative of the world population in terms of both regional distribution and income distribution. Thus, the set of countries included in the analyses represent the most comprehensive test of the hypotheses that is possible with extant data. 3.4. Control variables The analyses also include controls for an array of economic, political, and demographic variables that measure the external relations and internal structures of countries. Previous cross-national studies have identified international trade and foreign direct investment as important external, or cross-national, explanations of development outcomes (Kentor, 1998, 2001; Kentor and Boswell, 2003; Kentor and Jorgenson, 2010; Chase-Dunn, 1975; Timberlake and Kentor, 1983; Shandra et al., 2004, 2005; Jorgenson et al., 2007; Jorgenson, 2009; London, 1988; London and Smith, 1988). In order to distinguish the effect of international migration from these other international factors, the analyses include controls for exports per GDP to measure the degree to which a country is integrated into the world-economy and FDI stocks per GDP to indicate the extent of foreign economic influence in the country. FDI stocks are logarithmically transformed (ln) to correct for a skewed distribution. The measure of international trade comes from the World Development Indicators dataset (World Bank, 2009) and the measure of FDI comes from the World Investment Report (UNCTAD, 2005). Several factors internal to countries are likely to affect development levels. Democratic political structures are more likely to respond to public opinion and special interest groups concerned with issues related to development than more repressive political structures. The analyses control for this effect by including a measure of democratic development, or domestic political structure, in the country. Values on this variable range from 10 to +10, with lower scores indicating more authoritarian political structures and higher scores indicating more democratic political structures. These data are taken from the Polity IV dataset (Marshall et al., 2006). Similarly, the state plays a crucial role in economic development (Evans, 1995), as they shape the policy context in which national markets operate and serve as an intermediary between broader, international factors and national populations. Thus, the analyses include a variable to control for the influence of the state on development. A state’s capacity to influence development is largely a function of its size, which is often measured as government consumption per GDP. This measure is taken from the World Development Indicators dataset (World Bank, 2009) and it includes all government expenditures for the purchase of goods and services. The level of investment from domestic sources likely influences economic development levels. Countries that have higher levels of investment from firms situated within their boundaries are likely to have more robust and vigorous economies because higher proportions of profits are retained within these economies. Similarly, previous studies have found that domestic investment and foreign investment have different effects on development levels over time (Dixon and Boswell, 1996; Firebaugh, 1992). Thus, the analyses control for levels of gross domestic investment per GDP. This measure is taken from the World Development Indicators dataset (World Bank, 2009). Because the dependent variable is the average income in an aggregate economy, it is important to control for the distribution of aggregate income in the country. The analyses control for this effect by including a measure of the degree of income inequality in a country with the Gini coefficient, a standard measure of inequality. These data come from Deininger and Squire (1996). To control for the degree of technical variation in these data when this variable is included in the models, the analyses also include controls for whether the data are: (a) derived from income or expenditure data; (b) measured at the household or individual-level; and (c) measured as gross or net of taxes and transfers. Because they are technical variables for a control variable, the models presented only report the results for the income inequality variable, but results for these technical variables are available upon request. Finally, in order to capture time-specific effects that may affect the level of economic development but are not explicitly controlled for in the analysis (e.g. military conflicts, famines, refugee crises), the analysis includes a linear time trend, a common approach in longitudinal modeling (Wooldridge, 2006). 4. Results Tables 1 and 2 provide bivariate correlations and descriptive statistics for the variables included in the analyses. Table 1 includes statistics for all countries with information on all the variables included in the analysis. International migrants represent 3% of the population in the typical country and the average per capita income level is $7179. As hypothesized, international migration is positively correlated with the level of economic development 10 years later, and notably, the correlation is strong (r = 0.60). Table 2 presents correlations and descriptive statistics for the two key variables, international migration per capita and GDP per capita, across world income groups. As hypothesized, the relationship between international migration and longterm economic development varies in magnitude across countries in world income groups, ranging from a moderately strong positive correlation in high-income countries (r = 0.36) to a positive but slightly weaker correlation in middle-income countries (r = 0.36) to a negligible relationship in low-income countries (r = 0.02). The data also illustrate substantial levels of cross-national inequalities in both stocks of international migrants and average incomes. The average proportion of international migrants in high-income countries (6%) is three times larger than the average proportion of international migrants in

689

M.R. Sanderson / Social Science Research 42 (2013) 683–697 Table 1 Zero-order correlations and descriptive statistics (N = 118). GDPpc

(a)

(b)

(c)

(d)

(e)

(f)

(a) Int’l Migration (ln) (b) FDI Stock (ln) (c) Exports (d) Democracy (e) State Strength (f) Inequality (g) Domestic Investment

0.60 0.04 0.21 0.52 0.76 0.43 0.09

0.15 0.23 0.51 0.59 0.33 0.08

0.34 0.10 0.19 0.07 0.06

0.17 0.41 0.07 0.25

0.54 0.25 0.09

0.42 0.05

0.10

Mean Std. Dev.

$7179 $9016

0.03 0.03

0.11 0.14

0.26 0.15

4.53 6.43

0.14 0.06

0.40 0.06

Note: International migration per capita and FDI stock per GDP are not log-transformed in descriptive portion of table.

Table 2 Zero-order correlation and descriptive statistics for key variables by world income groups. High GDPpc

Middle GDPpc

Low GDPpc

Int’l Migration (ln) GDPpc Mean GDPpc S.D. Int’l Mig. Mean Int’l Mig. S.D.

0.36 $20,357 $7283 0.06 0.04

0.31 $2796 $1848 0.02 0.02

0.02 $334 $107 0.02 0.02

N (countries)

38

63

17

Note: International migration per capita not log-transformed in descriptive portion of table.

middle-income (2%) and low-income (2%) countries. Cross-national inequalities are even greater in terms of average incomes. Average per capita incomes in high-income countries ($20,357) are approximately seven times larger than average incomes in middle-income countries ($2796) and average incomes in middle-income countries are approximately 8 times greater than average incomes in low-income countries. These relationships are investigated further in the multivariate analyses below. Table 3 provides results from GLS random effects models with correction for a first-order autoregressive process (AR1). Model 1 tests Hypothesis 1: that cumulative immigration has a positive effect on the level of economic development 10 years later. The results support the hypothesis. Higher levels of immigration are associated with higher long-term levels of economic development across 122 countries. The findings also indicate that economic development levels have increased over time in constant dollars, as indicated by the positive term for the time trend in constant dollars, but the positive effect of immigration is robust to this trend. Because the international migration variable is log-transformed, the coefficient should be divided by 100 and then interpreted as the change in per capita incomes associated with a 1% increase in immigration. Thus, the effect of immigration is relatively small in magnitude, as a 1% increase in the stock of international migrants per capita raises per capita incomes 10 years later by approximately $6.60 on average. Model 2 presents results from an initial test of Hypothesis 2: that immigration has a larger positive long-term effect in higher-income countries compared to middle-income and low-income countries. To test this hypothesis, the model includes dummy variables indicating position in the world-income hierarchy (low-income is the reference category) and multiplicative terms indicating an interactive effect between international migration stocks and world-income groups. The findings support the hypothesis. International migration is disproportionately beneficial for long-term economic development in countries with higher average per capita incomes and high-income countries gain the most from immigration. The product term allowing for an interaction between international migration and high-income country indicates that a 1% increase in the immigrant stock in high-income countries raises average per capita incomes 10 years later by $14.32. The effect of immigration is nearly three times greater in high-income countries than in middle-income countries, where a 1% increase in the immigrant stock is associated with only a $5.02 increase in average per capita incomes 10 years later. The main effect for immigration indicates that immigration also has a positive effect in low-income countries, raising average per capita incomes by $3.20, but this effect is approximately 1.5 times smaller in low-income countries compared to middle-income countries and nearly five times smaller than in high-income countries. Thus, the results provide evidence that immigration has a Matthew Effect at the international level, exacerbating average income gaps between high and middle-income countries and between middle and low-income countries. Models 3 and 4 assess whether the relationship between international migration and economic development is robust to controls for international and domestic attributes of countries. Model 3 includes controls for foreign direct investment, domestic investment, and international trade. The results indicate that international migration has a positive, long-term effect on average per capita income in all countries, but the effect is only significant in high-income countries. Every 1%

690

M.R. Sanderson / Social Science Research 42 (2013) 683–697

Table 3 Autoregressive (AR1) GLS random effects regressions of GDP per capita on international migrants per capita. Model 1

Model 2

Model 3

Model 4

660.53*** (5.02)

320.02+ (1.82)

534.91 (1.30)

58.81 (0.14)

High Income

9767.1*** (8.66)

28296.1*** (9.17)

21813.7*** (7.27)

Middle Income

3200.7*** (3.88)

2407.1 (1.24)

2866.8 (1.45)

Int’l Mig. (ln)  High Income

1432.8*** (4.59)

5488.5*** (6.87)

2949.4*** (3.87)

Int’l Mig. (ln)  Middle Income

502.9** (2.97)

294.1 (0.76)

199.5 (0.49)

Int’l Mig. (ln)

FDI Stock (ln)

1299.6 ( 1.05)

Exports

12.08 ( 0.79)

Domestic Investment

13.25 (0.55)

Democracy

64.63+ (1.80)

State Strength

82.28+ (1.68)

Income Inequality

66.68* ( 2.16)

Year

107.39*** (13.62)

91.71*** (10.72)

88.76*** (3.97)

76.14** (3.29)

N (obs) N (countries) R2 (within) R2 (between) R2 (overall)

857 122 0.28 0.11 0.1

857 122 0.37 0.67 0.61

230 63 0.5 0.8 0.75

252 83 0.31 0.78 0.71

Notes: Coefficients with z-scores in parentheses. + p < .10. * p < .05. ** p < .01. *** p < .001 (two-tailed tests).

increase in the stock of international migrants increases average per capita income by $54.88 in high-income countries. Although the product term for middle-income countries and the main effect term for immigration (indicating the effect of immigration in low-income countries) are both non-significant, it is informative to compare the magnitude of the coefficients across world income groups. In this respect, Model 3 indicates that immigration is up to 18 times more beneficial for per capita incomes in high-income countries ($54.88 per person) compared to middle ($2.94 per person) and low-income countries ($5.34 per person) when controlling for theoretically-important explanations of national development. Notably, the results also show that neither foreign investment stocks nor international trade affect economic development levels. This is an interesting finding in light of the previous cross-national literature on economic development, much of which demonstrates that FDI in particular has a negative effect on economic development in the long-term (Kentor, 1998; Chase-Dunn, 1975). It could be that previous studies report positive effects for foreign investment and trade integration because international migration was excluded from the models. That is, instead of reflecting the effects of foreign investment and international trade, these coefficients may have instead captured the effect of international migration. However, this study cannot be definitive on the issue because it differs from these previous studies in important ways. Specifically, it includes developed countries (many previous papers focus exclusively on less-developed countries) and it models relationships over a shorter time horizon (some previous papers specify time lags as long as 40 years). Regarding the latter, it may be that the time horizon modeled in this analysis (10 years) is simply not long enough to capture the long-term negative effect of foreign investment on economic development. These are worthwhile issues for future cross-national research. Model 4 includes controls for important internal attributes of countries: level of democracy, state strength, and income inequality. Again, the results demonstrate that immigration is only beneficial for high-income countries. Each 1% increase in immigration raises average per capita incomes by $29.49 on average 10 years later. Model 4 also suggests disproportional effect from immigration, with income gains of $29.49 per person in high-income countries, but only $1.99 per person in middle-income countries and just $0.58 per person in low-income countries. Again, however, these interpretations are only heuristic because the coefficients for middle and low-income countries are not statistically significant. The control variables all

691

M.R. Sanderson / Social Science Research 42 (2013) 683–697 Table 4 Autoregressive (AR1) fixed effects regressions of GDP per capita on international migrants per capita. High-Income

Int’l Mig. (ln)

Middle-Income

Low-Income

Model 5a

Model 5b

Model 6a

Model 6b

Model 7a

Model 7b

2699.4** (3.01)

6883.4** (3.18)

84.44 (1.03)

604.9* (2.41)

5.37 ( 0.63)

554.5** ( 4.21)

Democracy

994.5+ (1.96)

14.87 (0.08)

0.308 (0.07)

State Strength

62.2 ( 0.38)

5.615 ( 0.22)

3.196 ( 0.59)

Inequality

70.5 (0.63)

17.32 (1.01)

3.702 ( 1.98)

Year

24.37*** (1.27)

10.46 (1.27)

6.26*** (7.21)

2.621* (2.59)

4.29*** (4.49)

2.816* ( 2.60)

N (obs) N (countries) R2 (within) R2 (between) R2 (overall)

160 29 0.16 0.06 0.07

65 17 0.23 0.06 0.02

332 65 0.16 0.06 0.07

78 32 0.21 0.05 0.06

219 42 0.11 0.01 0.01

18 8 0.85 0.31 0.11

Notes: Coefficients with z-scores in parentheses. + p < .10. * p < .05. ** p < .01. *** p < .001 (two-tailed tests).

have statistically significant effects on economic development and the relationships are in the direction specified by prior research. Higher levels of democracy and stronger states are associated with higher levels of long-term economic development, but higher levels of income inequality are associated with lower levels of long-term economic development. Table 4 presents results from first-order autoregressive fixed effects models in order to assess the robustness of the findings to modeling technique. Separate models are estimated for countries in each of the world income groups because fixed effects modeling does not allow time-constant variables. Two sets of models for each income group are provided. The first model for each group (the ‘‘a’’ model) attempts to replicate Model 2 by testing the effect of international migration only against a linear time trend in development levels. In addition, this strategy maximizes the number of countries included in the analyses. The second model for each group (the ‘‘b’’ model) adds the significant control variables from Models 3 and 4 in order to assess the effect of immigration net of these important explanations of economic development. The results provide additional support for Hypothesis 2. Again, immigration is disproportionately beneficial for higherincome countries and the effect is largest in high-income countries. Across the models, the results show that immigration is up to 10 times more beneficial for high-income countries ($68.83 per person in Model 5b) compared to middle-income countries ($6.04 per person in Model 6b) and that immigration may actually be detrimental for average per capita incomes in low-income countries (-$5.54 per person in Model 7b). Both Models 5a and Model 5b indicate that international migration has a positive, long-term effect on economic development levels in high-income countries: each 1% increase in the stock of international migrants in high-income countries raises per capita incomes between $26.99 and $68.83 on average 10 years later. The results also provide some evidence that immigration has a positive, long-term effect on development in middleincome countries. The effect of international migration is markedly smaller in magnitude and is positive in both Models 6a and 6b, but it is only significant in Model 6b, where it shows that a 1% increase in immigration increases per capita incomes by $6.04 on average over 10 years (Model 6b). Finally, the results demonstrate that immigration is slightly detrimental to long-term economic development in low-income countries. The coefficient for immigration is negative, but not significant, in Model 7a but when controls are included in Model 7b, the coefficient indicates that a 1% increase in the stock of international migrants decreases per capita incomes by $5.54 on average over 10 years in low-income countries. The results of Model 7 are intriguing. On the one hand, they are somewhat contrary to Hypothesis 1, which posited that immigration would have a positive effect on long-term economic development levels across countries in all world income groups. On the other hand, they are consistent with Hypothesis 2, which posited that immigration would have stronger positive effects in higher-income countries. It is important to note that the sample sizes for the low-income country models are relatively small, especially in Model 7b, where only 8 countries are included. Thus, some caution is warranted in interpreting this finding. However, the finding suggests the importance of analyzing the effects of immigration across world income groups. 5. Discussion The world-economy is characterized by large, and in many cases expanding, income gaps between countries. The International Monetary Fund (2010) estimates that the citizens of Qatar hold the highest incomes in the world, earning an aver-

692

M.R. Sanderson / Social Science Research 42 (2013) 683–697

age of US$ 88,559 in 2010. At the bottom, citizens in the Democratic Republic of the Congo earned an average income of US$ 328 in 2010. The gap is remarkable. The typical Qatari garners 270 times the average income of a Congolese, earning in a little more than 1 day what an average Congolese will earn in the entire year. The gap is unprecedented. Never before have human beings lived in a context in which living standards are as unequal as they are today (Milanovic, 2007). This paper empirically assessed how immigration affects international inequality by testing the relationship between immigration and national economic development across countries in different world income groups. A series of cross-national, longitudinal analyses demonstrated that, on average, immigration has a rather small, but positive long-term effect on development levels. This finding is consistent with a large body of prior economic research documenting the generally beneficial effects of immigration on national economic development. At the national level, the findings suggest that restricting immigration on the basis of concerns over potential negative effects on aggregate incomes is not warranted, especially in high-income countries. Higher levels of immigration increase aggregate incomes in the long-term, especially in high-income countries. Certainly, the issue of immigration cannot be reduced entirely to its economic implications. However, policies restricting immigration on the basis that it is detrimental for host country residents, as a whole, in the long-term, are generally contrary to the evidence presented here. The results are less sanguine on the question of how immigration affects international inequality. The results provide some support for the position of economic liberals, who argue that international migration is a positive-sum game at the international level. Immigration indeed appears to generate income gains, albeit relatively small gains, in most countries. However, immigration is not equally beneficial for all receiving countries. The analyses consistently demonstrated that immigration disproportionately benefits wealthier countries. Immigration into high-income countries produces average per capita income gains that are up to 18 times larger (Model 3) than the gains experienced in middle-income countries, and the gains from immigration in middle-income countries are at least twice as large as the gains in low-income countries (Model 2). Moreover, the wealthiest countries gain the most from immigration. Immigration increases per capita incomes in high-income countries by up to $68 on average (Model 5b) compared to a maximum gain in middle-income countries of just $6.04 (Model 6b). Thus, the results strongly suggest that immigration has a long-term Matthew Effect (Merton, 1968) in the contemporary world-economy: countries that are already structurally-advantaged in the world-economy gain more from immigration than countries in less structurally-advantaged positions. This Matthew Effect is evident between each of the world income groups. Immigration exacerbates income differences between high and middle-income countries, between middle and low-income countries, and between high and low-income countries. The results therefore provide substantial support for ongoing concerns, rooted in Mercantilist concerns about the loss of national power via population loss, that migration is more of a zero-sum game at the international level. Immigration may be generally beneficial for receiving countries, but not all countries receive benefit equally from immigration. More importantly, the benefits of immigration in high-income countries are large enough to make the relatively marginal gains in middle and low-income countries entirely insufficient for the purpose of closing income gaps with high-income countries. Even if residents in less-developed countries could capture a larger share of the relatively modest income gains from immigration, immigration at current levels would not produce aggregate income gains sufficient to ameliorate cross-national income disparities. Thus, despite increasingly being touted as a means redressing cross-national inequalities (World Bank, 2006; United Nations, 2009c), immigration does not appear to be a panacea. On the contrary, the findings suggest that immigration may actually be a mechanism that reproduces, and indeed even exacerbates, international inequalities. In this respect, the results presented here lend credence to arguments positing international migration as a form of unequal exchange, or net transfer of value, from lower-income to higher-income countries (Emmanuel, 1972). Although international migration has been relatively neglected as an important aspect of change in the world-economy, earlier research in the world-systems analysis tradition identified migration as a crucial form of unequal exchange reproducing uneven development in the world-economy: ‘‘Labor migration is a form of development aid given by poor countries to rich countries’’ (Castles and Kosack, 1973, p. 8). Made the argument more explicit: ‘‘This additional immigrant labor force constitutes also a disguised transfer of value from the periphery to the center, since the periphery has borne the cost of training this labor force’’ (Amin, 1976, p. 362). Burawoy (1976) elaborated further: ‘‘The very sale of labor power by an underdeveloped country. . . to an economically advanced nation serves only to reinforce the relations of economic subjugation and domination’’ (Burawoy, 1976, p. 1068). The findings from this study suggest it is worthwhile to revisit international political-economic theories of migration as a means of furthering our understanding of how migration, as a relational exchange, shapes development outcomes across countries. Future studies drawing on these theoretical frameworks and employing international-comparative methods could provide important new insights into the relationship between migration and international inequality. Two limitations to this study are worth noting. First, the findings represent the average effect of immigration across countries and over time. These are cross-national-level results and should not be applied to specific countries. The effect of immigration likely varies in magnitude, or even direction, in any particular country. The analyses do, however, include country-specific effects, reducing to the largest extent possible the possibility of biased results due to the exclusion of important time-invariant, country-specific factors. Moreover, the magnitude of the effect of international migration found here is quite similar to the magnitudes reported in previous studies of individual countries, which lends credence to the notion that if the effect does differ in specific countries, it is likely not considerably different. Second, the data used in this study likely underreport the actual level of immigration because they do not include undocumented migrants. This is an unavoidable problem in migration research, but it is one that deserves mention. It should be noted, however, if undocumented migrants

693

M.R. Sanderson / Social Science Research 42 (2013) 683–697

represent a large proportion of all migrants, then the estimates presented likely underestimate the true magnitude of the effect of immigration. Indeed, the relatively small aggregate gains from immigration reported in prior research are usually attributed to the fact that migrants comprise relatively small shares of the populations (Kapur and McHale, 2005). Two additional avenues for future research are especially warranted on the basis of this study. Although immigration appears to be beneficial for aggregate incomes, it is likely that it does not benefit all host country residents equally. Indeed, economic theory suggests that a relatively small gain in aggregate income from immigration disguises a relatively large redistribution of wealth. For example, Borjas (1995) estimates that while immigration produces an aggregate gain in income (i.e., the ‘‘immigration surplus’’) it also produces a net transfer of wealth from laborers to capital owners of approximately 2% of GDP. Specifically, host country laborers working in sectors in which they must compete for employment with immigrant labor experience decreases in wage levels. These wage decreases, however, represent gains to capital owners, who benefit directly from immigration, regardless of whether they directly employ immigrant labor (Borjas, 1995). Thus, future studies of the relationship between immigration and within-country inequality would extend and refine this study, and possibly even shed new light on the political-economic implications of immigration in host countries. At the international level, the empirical question of whether international migration promotes or stems international inequality ultimately depends on the perspective taken: the perspective of destination countries (i.e., immigration), of origin countries (i.e., emigration), or migrants themselves. This study is limited solely to the former; it cannot address the question from the other perspectives. There is a general consensus that international migration benefits immigrant themselves, otherwise they would not move, and migration research increasingly consists of micro-level investigations of the implications of migration. Yet as long as national boundaries exist, the fates and fortunes of peoples will be shaped, to an important extent, by social structures that exist at the nation-state level of analysis. Until people are allowed to move without constraint across national boundaries, it will be crucial to investigate the national-level causes and consequences of international migration. In this respect, the findings from this study open up an important opportunity for future research addressing the question of how international migration affects international inequality from the perspective of origin countries. There is a real need for research that examines the effects of migrant remittances on per capita incomes in origin countries because remittances may reduce, or augment, the magnitude of the effects of immigration on destination countries. For example, if remittances are found to increase per capita incomes in origin countries, then the inequality-exacerbating effects of immigration documented in this study will be lessened in magnitude. However, if remittances are found to decrease, or have no effect on, per capita incomes in origin countries, then it will become clearer that international migration cannot ameliorate cross-national inequalities. Extant research is inconclusive on the issue of remittances and development, but to my knowledge, there are no international-comparative studies that investigate this relationship. This is certainly another worthwhile area for future research.

Appendix A See Appendix A1.

Appendix A1 Countries included in analysis.

High income countries Australia Austria Belgium Canada Denmark Germany Finland France Greece Hong Kong Ireland Israel Italy Japan Korea Kuwait Netherlands New Zealand Norway Portugal

1960

1965

1970

1975

1980

1985

1990

1995

X X X X X

X X X X X

X X X X X

X X X X X

X X X X X

X X X X X

X X

X X

X X

X X

X X

X X X

X X X

X X X X X

X X X X X

X X X X X

X X X

X X X

X X X

X X X X

X X X X

X X X X X X X X X X X X X X

X X X X X X X X X X X X X X X X X X X X

X X X X X X X X X X X X X X X X X X X X

(continued on next page)

694

M.R. Sanderson / Social Science Research 42 (2013) 683–697

Appendix A1 (continued) 1960 Puerto Rico Saudi Arabia Singapore Spain Sweden Switzerland Trinidad and Tobago United Arab Emirates United Kingdom United States Middle income countries Albania Algeria Angola Argentina Bolivia Botswana Brazil Bulgaria Cameroon Chile China Colombia Congo (Republic of) Costa Rica Cote d’Ivoire Dominican Republic Ecuador Egypt El Salvador Gabon Greece Guatemala Haiti Honduras Hong Kong Hungary India Indonesia Iran Ireland Jamaica Jordan Korea Lebanon Lesotho Liberia Libya Malaysia Mauritius Mexico Morocco Namibia Nicaragua Oman Pakistan Panama Papua New Guinea Paraguay Peru Philippines Poland Portugal Puerto Rico Romania Saudi Arabia Singapore South Africa Spain

1965

1970

1975

1980

1985

1990

1995

X X X X X X

X X X X X X

X X X X X X

X X X X X X

X X X X X X X

X X X

X X X

X X X

X X X X X X X X X X X X X X X X X X X

X X X X X X X X X X X X X X X X X X X

X X X X X X X X X X X X X X X X X X X X

X

X

X

X

X

X

X X

X X

X X X X X X

X X

X X

X X

X X

X

X

X X

X X

X X X X X

X X X X X

X

X

X

X X

X X X X

X X X X

X X X X X X X X X X

X X X X X X X X

X X X X X X X X

X X X X X X X X

X X X X X X X X

X X X X X X X X

X X X X X X

X X X X X X

X X X X

X X X X

X X X X

X

X

X

X

X

X

X

X X

X X

X X

X X X X

X X X X X

X X X X

X X X

X X X

X X X X

X X

X X

X X

X

X

X X

X X X X X X X

X X X X X X X

X X X X X X X

X X X X X X X

X X X X X X X

X

X X

X X X

X X X

X X X

X X X

X

X X

X X

X X

X X

X X

X

X

X

X X

X X

X X X X X

X X X X X

X X X X X

X X X X X

X

X X X X X X X

X X

X

X

X X X X X

X

X

X

X

X X

X

X

X

X

X

X

X

X X X

X X X X X X

695

M.R. Sanderson / Social Science Research 42 (2013) 683–697 Appendix A1 (continued) 1960 Sri Lanka Syria Thailand Trinidad and Tobago Tunisia Turkey Uruguay Venezuela Vietnam Yemen, Rep. Zambia Zimbabwe Low income countries Bangladesh Benin Burkina Faso Burundi Cambodia Cameroon Central African Republic Chad China Democratic Republic of Congo Eritrea Ethiopia Gambia Ghana Guinea Guinea-Bissau Haiti India Indonesia Kenya Laos Lesotho Liberia Madagascar Malawi Mali Mauritania Mongolia Mozambique Nepal Niger Nigeria Pakistan Papua New Guinea Rwanda Senegal Sierra Leone Sri Lanka Sudan Tanzania Togo Uganda Vietnam Yemen Zambia Zimbabwe

1965

1970

X X

X X

X X

X X X X

X X X X

X X X X

1975

1980

1985

1990

1995

X X X X X X X

X X X X X X X X

X X X X X X X X

X X X X X X X X

X X X

X

X X X X X X

X X

X

X

X

X

X

X

X X X X

X X X X

X X X X

X X X X

X X X X

X X X X X

X X X X X

X X X X X

X X X X X

X X X X

X X X X

X X X X

X X X X

X X

X X

X X

X

X

X X X

X X X X X X X X

X X X X X X X X

X

X

X

X

X X

X X

X X X

X X X

X X X

X X X X X

X

X X

X X X X X

X X X X

X X X X

X X X X

X X X X

X X X X X

X X X X X X X X X X

X X X X X

X X X X X

X X X X X

X X X X X

X

X

X

X X X

X

X

X

X X X X X X X X X X X X X X X X X X

X X X X X X X X X X X X X X

X X X X X X X X X X X X X

X X

X X X

X X X

X X X

X X X X X X X

X X X X X X X

X X X X X

X X X X

X

X X

X X X X X X X X X X

References Amin, Samir, 1976. Unequal Development: An Essay on the Social Formations of Peripheral Capitalism. Monthly Review Press, New York. Aydemir, Abdurrahman, Borjas, George, 2007. Cross-country variation in the impact of international migration: Canada, Mexico, and the United States. Journal of the European Economic Association 5 (4), 663–708. Barrell, Ray, Fitzgerald, John, Riley, Rebecca, 2010. EU enlargement and migration: assessing the macroeconomic impacts. Journal of Common Market Studies 48 (2), 373–395. Borjas, George, 1995. The economic benefits from immigration. Journal of Economic Perspectives 9 (2), 3–22.

696

M.R. Sanderson / Social Science Research 42 (2013) 683–697

Borjas, George, 2003. The labor demand curve is downward sloping. Quarterly Journal of Economics 118 (4), 1335–1374. Burawoy, Michael, 1976. The functions and reproduction of migrant labor: comparative material from Southern Africa and the United States. American Journal of Sociology 81 (5), 1050–1087. Castles, Stephen, 2007. Can migration be an instrument for reducing inequality? In: Paper Presented at the Metropolis Conference, Melbourne, Australia. (accessed 09.08.12). Castles, Stephen, Kosack, Godula, 1973. Immigrant Workers and Class Structure in Western Europe. Oxford University Press, London. Castles, Stephen., Miller, Mark J., 2003. The Age of Migration: International Population Movements in the Modern World, fourth ed. Guilford, New York. Chase-Dunn, Christopher, 1975. The effects of international economic dependence on development and inequality: a cross-national study. American Sociological Review 40 (6), 720–738. Cholezas, Ioannis, Tsakloglou, Panos, 2009. The economic impact of immigration in Greece: taking stock of the existing evidence. Southeast European and Black Sea Studies 9 (1), 77–104. Clemens, Michael A., 2010. Immigration Policy as a Development Tool. (accessed 15.08.12). De Arce, Rafael, Mahia, Ramon, forthcoming. An estimation of the economic impact of migrant access on GDP: the case of the Madrid region. International Migration. Deininger, Klaus, Squire, Lyn, 1996. A new data set measuring income inequality. World Bank Economic Review 10 (3), 565–591. Dixon, William J., Boswell, Terry, 1996. Dependency, disarticulation, and denominator effects: another look at foreign capital penetration. American Journal of Sociology 102, 543–562. Dustmann, Christian, Glitz, Albrecht, Frattini, Tommaso, 2008. The labour market impact of immigration. Oxford Review of Economic Policy 24 (3), 477–494. Emmanuel, Arghiri, 1972. Unequal Exchange: A Study of the Imperialism of Trade. Monthly Review Press, New York. Evans, Peter, 1995. Embedded autonomy: states and industrial transformation. Princeton University Press, Princeton. Felbermayr, Gabriel J., Hiller, Sanne, Sala, Davide, 2010. Does immigration boost per capita income? Economics Letters 107, 177–179. Feridun, Mete, 2004. Does immigration have an impact on economic development and unemployment? Empirical evidence from Finland (1981–2001). International Journal of Applied Econometrics and Quantitative Studies 1 (3), 39–60. Feridun, Mete, 2005. Economic impact of immigration on the host country: the case of Norway. Prague Economic Papers 4, 350–362. Feridun, Mete, 2007. Immigration, income and unemployment: an application of the bounds testing approach to cointegration. Journal of Developing Areas 41 (1), 37–51. Firebaugh, Glenn, 1992. Growth effects of foreign and domestic investment. American Journal of Sociology 98, 105–130. Goldin, Ian, Cameron, Geoffrey, Balarajan, Meera, 2011. Exceptional People: How Migration Shaped Our World and Will Define Our Future. Princeton University Press, Princeton. Halaby, Charles N., 2004. Panel models in sociological research: theory into practice. Annual Review of Sociology 30, 507–544. International Monetary Fund, 2010. World Economic Outlook Database. (accessed 15.05.11). Jorgenson, Andrew K., 2009. The economy, military, and ecologically unequal exchange relationships in comparative perspective: a panel study of the ecological footprints of nations, 1975–2000. Social Problems 56 (4), 621–646. Jorgenson, Andrew K., Dick, Christopher, Mahutga, Matthew C., 2007. Foreign investment dependence and the environment: an ecostructural approach. Social Problems 54 (3), 371–394. Kapur, Devesh, McHale, John, 2005. Give Us Your Best and Brightest: The Global Hunt for Talent and Its Impact on the Developing World. Center for Global Development, Washington, DC. Kentor, Jeffrey, 1998. The long-term effects of foreign investment dependence on economic growth, 1940–1990. American Journal of Sociology 103 (4), 1024–1046. Kentor, Jeffrey, 2001. The long term effects of globalization on income inequality, population growth, and economic development. Social Problems 48 (4), 435–455. Kentor, Jeffrey, Boswell, Terry, 2003. Foreign capital dependence and development: a new direction. American Sociological Review 68 (2), 301–313. Kentor, Jeffrey, Jorgenson, Andrew K., 2010. Foreign investment and development. International Sociology 25 (3), 419–441. Lewis, W. Arthur, 1954. Economic development with unlimited supplies of labor. The Manchester School 22 (2), 139–191. Lianos, Theodore P., Sarris, Alexander H., Katseli, Louka T., 1996. Illegal immigration and local labour markets: the case of Northern Greece. International Migration 34, 449–484. London, Bruce, 1988. Dependence, distorted development, and fertility trends in noncore nations: a structural analysis of cross-national data. American Sociological Review 53 (4), 606–618. London, Bruce, Smith, David A., 1988. Urban bias, dependence, and economic stagnation in noncore nations. American Sociological Review 53 (3), 454–463. Marshall, Monty G., Jaggers, Keith, Gurr, Ted Robert, 2006. Polity IV Project: Political Regime Characteristics and Transitions, 1800–2004. Center for International Development and Conflict Management, University of Maryland. Marx, Karl, 1967. Capital, vol. 1. Penguin Books, London. Merton, Robert, 1968. The Matthew effect in science: the reward and communication systems of science are considered. Science 159 (3810), 56–63. Milanovic, Branko, 2007. Worlds Apart: Measuring International and Global Inequality. Princeton University Press, Princeton. Modelski, George, Thompson, William R., 1996. Leading Sectors and World Powers: The Coevolution of Global Economics and Politics. University of South Carolina Press, Columbia, SC. Ortega, Francesc, Peri, Giovanni, 2009. The Causes and Effects of International Labor Mobility: Evidence from OECD Countries 1980–2005. Human Development Research Paper No. 6. New York: United Nations Development Programme, Human Development Report Office. Ottaviano, Gianmarco I.P., Peri, Giovanni, 2008. Immigration and National Wages: Clarifying the Theory and the Empirics. NBER Working Paper 14188. Peri, Giovanni, 2010. The Impacts of Immigrants in Recession and Economic Expansion. Migration Policy Institute (June 2010). Pholphirul, Piriya, Rukumnuaykit, Pungpond, 2009. Economic contribution of migrant workers to Thailand. International Migration 48 (5), 176–202. Portes, Alejandro, Walton, John, 1981. Labor, Class, and the International System. Academic Press, New York. Ranis, Gustav, Fei, John C.H., 1961. A theory of economic development. American Economic Review 51 (4), 533–565. Sanderson, Matthew R., 2010. International migration and human development: a cross-national analysis of less-developed countries, 1970–2005. Social Indicators Research 91 (1), 59–83. Sanderson, Matthew R., Kentor, Jeffrey D., 2008. Globalization and international migration: a cross-national analysis of less-developed countries, 1985– 2000. International Sociology 23 (4), 514–539. Sanderson, Matthew R., Kentor, Jeffrey D., 2009. Globalization, development, and international migration: a cross-national analysis, 1970–2000. Social Forces 88 (1), 301–336. Sassen, Saskia, 1988. The Mobility of Labor and Capital: A Study in International Investment and Labor Flow. Cambridge University Press, Cambridge. Shandra, John M., Nobles, Jenna, London, Bruce, Williamson, John B., 2004. Dependency, democracy, and infant mortality: a quantitative, cross-national analysis of less developed countries. Social Science & Medicine 59 (2), 321–333. Shandra, John M., Nobles, Jenna E., London, Bruce, Williamson, John B., 2005. Multinational corporations, democracy and child mortality: a quantitative, cross-national analysis of developing countries. Social Indicators Research 73 (2), 267–293. Stimson, James A., 1985. Regression models in space and time: a statistical essay. American Journal of Political Science 29, 914–947. Thomas, Brinley, 1973. Migration and Economic Growth: A Study of Great Britain and the Atlantic Economy, second ed. Cambridge University Press, Cambridge.

M.R. Sanderson / Social Science Research 42 (2013) 683–697

697

Timberlake, Michael, Kentor, Jeffrey, 1983. Economic dependence, overurbanization, and economic growth: a study of less developed countries. The Sociological Quarterly 24 (4), 489–507. UNCTAD, 2005. World Investment Report, 2005: Transnational Corporations and the Internationalization of R&D. United Nations Conference on Trade and Development, Geneva. United Nations, 1994. International Conference on Population and Development: Summary of the Programme of Action. (accessed 19.07.12). United Nations, 2008. Data collection on migration and development: towards a global consensus. Presentation by Bela Hovy, Chief of Migration Section of the Population Division, United Nations, at the Seminar on Migration Data, United Nations Institute for Training and Research (December 10). (accessed 24.02.10). United Nations, 2009a. Human Development Report, 2009: Overcoming Barriers: Human Mobility and Development. United Nations Development Programme, New York. United Nations, 2009b. Trends in International Migration Stock, The 2008 Revision. United Nations Population Division, New York. United Nations, 2009c. World Population Prospects. United Nations Population Division, New York. van der Mensbrugghe, D.,Roland-Holst, D., 2009. Global Economic Prospects for Increasing Developing Country Migration into Developed Countries. Human Development Research Paper No. 50. United Nations Development Programme, Human Development Report Office, New York. Vernon, Raymond, 1966. International investment and international trade in the product cycle. Quarterly Journal of Economics 80, 190–207. Wooldridge, Jeffrey M., 2002. Econometric Analysis of Cross Section and Panel Data. MIT Press. Wooldridge, Jeffrey M., 2006. Introductory Econometrics: A Modern Approach, third ed. Thomson Southwestern, Mason, OH. World Bank, 2006. Global Economic Prospects, 2006. World Bank, Washington, DC. World Bank, 2009. World Development Indicators. World Bank, Washington, DC. Zavodny, Madeline, 2011. Immigration and American Jobs. American Enterprise Institute for Public Policy Research (December).