The (de)globalization of migration: has the polycrisis period changed the patterns of global migration? (OeNB Bulletin Q3/24)

Jonathan Fitter, Anna Katharina Raggl, Paul Ramskogler ¹

JEL classification: F22, F60, J11

Keywords: international migration, bilateral migration, (de)globalization, diversification of migrants

The second half of the 20th century was characterized by increasing global economic integration, which gained even more speed after the fall of the Iron Curtain. The global economic crisis in 2008–10 seemed to have put a first halt to this process: the ensuing recovery was uneven and shallow in many world regions. At the same time, many advanced economies started to question the benefits of globalization: The United Kingdom leaving the EU and the trade war between the USA and China can be seen as manifestations of this skepticism (Thompson, 2022). More recently, the COVID-19 pandemic and the Russian invasion of Ukraine highlighted the fragility of supply chains, imposing further challenges on globalization. Whether this leads to a permanent trend reversal or is merely a transitory development remains to be seen (see, for example, Goldberg and Reed, 2023, and a recent IMF staff discussion note, Aiyar et al., 2023, for an assessment of the risks associated with increasing policy-driven geo-economic fragmentation).

While global trade and capital flows are more closely monitored (see, for example, Abelianski et al., 2024), another dimension of bilateral flows receives less attention in the context of (de)globalization: the bilateral flow of people. However, the composition of migration has important repercussions for trade channels, productivity, innovation and thus, ultimately, monetary policy. Conventional wisdom is that migration has been becoming increasingly global, more frequent and more diverse. Whether this is the case and/or whether migration flows are joining deglobalization trends often attributed to trade and capital flows has not been empirically assessed recently.

This study focuses on the question whether bilateral migration flows show signs of deglobalization, building on a concept developed by Czaika and de Haas (2015), who suggest measuring the degree of globalization along three dimensions: the intensity of migration, the spread/diversification of migration with respect to source and destination countries, and the average distance traveled by migrants. In this study, we apply the approach to estimates of global bilateral migration flows from Abel and Cohen (2019) in its most recent version (October 2022) and provided by Abel (2019) to address the following questions: First, looking at global migration flows , has migration continued to become more globalized since 1990? Has it become less globalized in the most recent past, in line with deglobalization tendencies often attributed to other cross-border flows (capital, trade)? Second, how do world regions differ with respect to the diversity of source countries of emigrants and destination countries of immigrants and their degree of migration globalization? Specifically, how does Europe differ from the rest of world? And how has this been changing over time?

These questions are not only intriguing in their own right but also integral to a comprehensive assessment of (de)globalization tendencies in global flows of capital, goods, services and people. Migration, in particular, has important implications for economic outcomes in both receiving and sending countries. In the euro area, the first and most direct impact is on the size of the working-age population. Especially against the background of shrinking working-age populations, immigration can counteract population decline at least in the short to medium run. Note, for instance, that the recent labor market boom in the USA can be largely explained by a surprise surge in immigration (Edelberg and Watson, 2024). In both the EU and the euro area, for example, the labor force would have shrunk between 2012 and 2022 without workers from foreign countries ² . During the COVID-19 pandemic, on the other hand, weak net migration may have contributed to subdued labor force developments (ECB, 2022). Second, immigration does not only alter the size but also the age structure of a population. As immigrants tend to be relatively young, migration can counteract population aging in aging societies. It can contribute to lower old-age dependency ratios, which have been increasing strongly in most advanced economies (see, for example, Peri, 2020). The fact that immigrants tend to be young is also one of the reasons why several studies also find a positive fiscal impact of immigration (see, for example, OECD, 2021). Further, immigrants can contribute to a slowdown of population aging in advanced economies through, on average, higher fertility rates. Third, migration has been shown to foster productivity, thereby fueling economic growth. This effect is usually attributed to the complementarity of immigrants’ skills to those of natives. With immigrants entering the labor market, natives move to different occupations that often require the performance of more complex tasks, linguistic and communication proficiency, etc. In other words, natives often upgrade their jobs as immigrants enter the labor market (see Foged and Peri, 2016.) Fourth, migration is related to other cross-country flows, such as trade and capital flows. Specifically, trade between home and host countries has been shown to increase as a result of migration (see, for example, Iranzo and Peri, 2009; Egger et al., 2012; Felbermayr et al., 2015; Bahar and Rappoport, 2018; OECD, 2022), just like FDI has been shown to increase with higher migration between two countries (see, for example, Buch et al., 2006; Javorcik et al., 2011; Kugler and Rapoport, 2007).

Thus, migration has an impact on economic variables, and in terms of growth and productivity, the impact is positive. But how does this relate to the globalization of migration? Does the globalization of migration have an effect on economic variables to an extent that it becomes relevant from a monetary policy perspective? In fact, there is accumulating evidence that higher cultural heterogeneity of migrants has positive impacts on productivity growth and innovation (see literature section below). As far as this is the case, migration affects the natural rate of interest. This is when it becomes relevant for monetary policy.

This study contributes to the literature in several ways. First, we address a topic that has received a lot of attention recently in the context of capital and trade flows and apply it to international migration dynamics. Second, we use a conceptual framework for assessing the globalization of migration developed by Czaika and de Haas (2015) and apply it to more recent data on migrant stocks. This allows us an assessment of (de)globalization trends of international migration up to the year 2020, an update compared to the existing results for 1960 to 2000. Third, we use estimates of global bilateral migration flows (Abel and Cohen, 2019). This way, the analysis is not diluted by previous migrants that show up in the stock figures although the actual year of migration might have been years or even decades ago. Migration flow estimates show current migration behavior and thus allow us to better study current trends. Fourth, we focus on the EU and assess how the intensity and diversification of migration has changed over time. This will feed into a follow-up study that will be dedicated to the assessment of possible implications for the size and age structure of the euro area population.

The remainder of this paper is structured as follows. In the next section, we will discuss the concept of measuring the degree of globalization of migration. After describing the data in section 2, we will present the results in chapter 3. Section 4 concludes.

1 The impact of migrant diversity on economic outcomes in the literature

The only paper that explicitly addresses the globalization of migration is that of Czaika and de Haas (2015). However, there is a body of literature that is concerned with the economic relevance of migration diversity. Migrants influence the economy in various ways: they impact demography, labor markets and productivity; migration has a fiscal dimension, and it also relates to trade and capital flows (see, for example, OECD, 2014; Koczan et al., 2019). An expansion of the workforce almost inevitably leads to an increase in a country’s GDP. But there is also extensive research on the impact of migration on GDP per capita, and the majority of empirical studies find a growth-enhancing effect of immigration (see, for example, Brunow et al., 2015; Engler et al., 2023).

In the context of this study, however, the obvious question is whether the spread of migration, i.e. the heterogeneity of (im)migrants with respect to their origin, influences economic outcomes, and, if so, how. There is a limited but growing body of literature that addresses this question, and most analyses conclude that the diversity of immigrants with respect to their place of birth boosts economic performance in the recipient countries, regions, subregions or cities. The mechanism behind this positive impact is usually described as the complementarity in knowledge and skill sets that come with migrants that were raised and trained in different countries. Bahar et al. (2022) argue that this increase in the skill sets that countries can draw upon also enables a country to become active in a broader set of fields and to become more economically complex. They empirically explore this relationship and find cross-country evidence that countries with a birthplace-diverse population indeed exhibit higher economic complexity. This finding holds in particular for diversity among highly educated migrants and for countries with intermediate levels of economic complexity. They further provide evidence that the underlying mechanism of birthplace diversity boosting economic complexity is the increasing diversity in host countries’ export baskets. Bove and Elia (2017) investigate whether cultural diversity brought about by immigration has an impact on economic growth and, if so, to what extent. They find a robust and positive relationship over long time periods. Similarly, Alesina et al. (2016) find that the diversity of immigrants relates positively to measures of economic prosperity. The results suggest that skill complementarities between immigrant and native workers are driving the effect. Trax et al. (2015) show at the firm level and for Germany that cultural heterogeneity in firms increases productivity levels (while the mere share of foreign workers in a firm does not). In addition, the cultural heterogeneity in the region where a firm is located, matters for plant-level productivity. Discussing evidence on the impact of migrant heterogeneity with respect to origin but also to skills, motives, culture, etc., Brunow et al. (2015) conclude that its impact is, on balance, positive on innovation and economic growth. Ortega and Peri (2014) explore the interrelationship between trade, migration and income per capita. They find that openness to immigration has a positive and robust long-run impact on income per capita, with the main impact of immigration operating through total factor productivity. The degree of diversity in migration flows has an additional positive impact on income: the set of skills in the labor force is better differentiated, and some evidence suggests that innovation activity is higher with more diversity among migrants.

These recent examples of studies investigating the impact of the diversity of migrants – instead of their numbers – on economic outcomes provide evidence of a positive impact of birthplace diversity. Underlying mechanisms are predominately related to (skill) complementarities between native and migrant workers and associated productivity improvements and innovation.

2 Measuring the degree of (de)globalization of migration

Along with the increase of trade and cross-border investment flows in the course of globalization, a common narrative is also that migration flows keep growing and are getting increasingly complex. At the same time, it is argued, albeit less commonly, that the share of people living outside their country of birth is actually rather small and did not change significantly over several decades, i.e. between the 1960s and 2000s, and after that only increased to a still moderate 3.6% in 2020 (UNDP, 2020). Czaika and de Haas (2015) challenged the common notion that global migration patterns have become increasingly complex over the past decades, with a formerly more clear-cut distinction between typical immigration and emigration countries becoming diluted over time and the initially few main corridors – often following colonial ties – becoming increasingly broad. They do so by investigating – in a structured way – the changes in the magnitude and the diversity of global migration during 1960 and 2000, using data on migrant stocks that measure the number of foreign-born individuals living in the reference countries at a given point in time.

It is not straightforward how to measure the degree of globalization of migration, and while there is literature that relates international migration to globalization, usually understood as openness to trade and foreign capital, research on (the measurement of) the globalization of migration is very scarce. Czaika and de Haas (2015) suggested a framework to quantitatively assess this question by considering three different dimensions: the intensity of global migration, the spread of migrants across origin and destination countries, and the average distance covered by global migrants. ³ The intensity of migration is measured by the share of migrants in the population. The spread of migration, i.e. the origin-country diversity of immigrants and the destination-country diversity of emigrants, is measured as unity minus the Herfindahl-Hirschman Index of concentration: the higher this spread measure, the more diverse the migrants. And the third dimension, the distance of migration, is measured as the average distance between migrants’ origin and source country in a given year or period. Following Czaika and de Haas (2015), we combine these three dimensions to obtain an index of emigrant globalization and an index of immigrant globalization. The former summarizes how globalized emigrants are: how many emigrants are there, how diverse are their destinations, and how far away do they move. The latter index informs about the degree of globalization of immigrants: how many immigrants does a country have, how diverse are they, and from how far away are they? Finally, we combine these two subindices to obtain an index of migration globalization that measures the integration of a country into global migration processes, considering both emigrants from and immigrants to a country. Under this concept, migration becomes more globalized if the intensity of migration increases, migration is more diverse (with respect to origin and destination countries) and the distance becomes longer. Please refer to section A1 in the annex for a comprehensive explanation of the indicators constructed.

We use this approach to address the question of (de)globalization of migration, using recent data on global migration flow estimates in addition to migration stock data. We also use the most recent data on migration stocks, currently available for 1990 to 2020, to directly compare and update the results by Czaika and de Haas (2015). In all dimensions, it is crucial to look beyond the global average that might hide important regional heterogeneities. Geographically, our particular emphasis is on the EU.

3 Data

The core data we use in this analysis are estimates of international migration stocks and estimates of international migration flows. For a comprehensive overview of currently available data on international migration, see Buettner (2022).

3.1 Global bilateral migrant stocks

The UN’s International Migrant Stock 2020 dataset (UNDP, 2020) is currently the most complete global database for bilateral stocks of migrants. It contains estimates of the total number of international migrants by age, sex and origin for the mid-point (July 1) of every fifth year between 1990 and 2020. The data are available for 232 countries and areas ⁴ of the world and are based on official statistics of the foreign-born population. Most of the estimates rely on population censuses but also use information from population registers and nationally representative surveys. Generally, international migrants are equated with the foreign-born population, which is possible for most countries. Whenever the necessary information on the country of birth is not available, and that is the case for approximately 20% of the countries/areas, the country of citizenship is used instead. See section A6 in the annex for further information.

3.2 Global bilateral migrant flows

Migrant stocks also reflect to a large extent past migration patterns. Because we are mostly interested in current trends and developments, estimates of international migration flows are more interesting for us: they are not “diluted” by previous migration behavior and reflect concurrent migration dynamics. Migration flow data have additional important advantages over stock data. First, they are more appropriate for policy analysis, as flows can show direct reactions to policy changes and do not include migration movements that happened years or even decades ago independent of current policies. Second, migration flow estimates take into account return migration : a movement from A to B and back to A would be reflected and counted as two movements while it would not show up in migrant stock data once the return migration is completed. Third, flow data also better reflect onward migration : a movement from A to B to C would be counted as two movements while in migrant stock data, it would be interpreted as one movement from A to C once the onward migration is completed.

However, a global dataset on migration flows is not available, as in many countries migration flow data are not collected. In addition, even if the data are available for certain countries, it is difficult to compare the data for different countries, as underlying definitions often differ. We thus rely on estimates in our analysis. In particular, we make use of the bilateral international migration estimates provided by Abel (2019) and Abel and Cohen (2019), who – motivated by the lack of migration flow data – estimate global bilateral migration flows drawing on UN international migrant stock data. For this analysis, we use the most recent update of the database (Abel, 2019, version 6, October 28, 2022). This update has been prepared to incorporate current UN migrant stock estimates (UNDP, 2020) and current UN population statistics updates (UNDP, 2022). Following the periodicity of the UN migration data, the dataset provides estimates of migration flows between country pairs over five-year periods from 1990–1995 to 2015–2020 (in each case from July 1 onward). The estimates cover 225 countries for the periods 1990–1995 to 2000–2005, 226 countries for 2005–2010, and 229 from the period 2010–2015 onward. Abel and Cohen (2019) provide six different estimates of global bilateral migration flows and show validation exercises that help to choose the estimates according to the intended use. We use migration flows estimated by the closed demographic accounting method (pseudo-bayesian), as these estimates exhibit the best performance in validation exercises, in particular when the bilateral dimension of the data is of importance, as in this application.

3.3 Further data sources

In addition, we use data on the total population from the UN (UNDP, 2022) and the distance between countries, measured by the geographical distance between their largest cities from the CEPII Gravity Database (Conte et al., 2022).

4 Empirical findings

4.1 Intensity of global migration flows

The number of international migrants continuously increased in the past decades. In 1960, an estimated 77 million people lived outside the country they had been born in (United Nations, 2009). By the year 2000, this number had grown to 173 million and increased further to 281 million by 2020. ⁵ Considering the simultaneous growth in the global population, relative numbers provide a more insightful perspective and show a rather different picture: The global share of migrants, i.e. the number of people living outside their birth country relative to the total global population, remained fairly constant at or below 3% from the 1960s to 2005, before it moderately increased in the past 15 years, to reach 3.5% in 2020 (see chart 1 ⁶ and table 1 in section 4.4, which summarizes trend changes in all the indicators discussed). Currently, the stock of migrants – in absolute and relative terms – is at an all-time high. Political instability, geopolitics, inequality, and climate change are part of the story behind this recent increase (see, for example, IOM, 2024). Nevertheless, migrants still constitute a small share of the global population.

Chart 1

When looking at the estimates of global migration flows, i.e. the number of people that move from one country to another in a five-year period, we see that in relative terms, they remained fairly constant: In the period 1990–95, around 1.4% of the global population migrated, and in the period 2015–20, this share was approximately the same. This implies that the absolute values grew at rates comparable to overall population growth⁷ and migration did not become more frequent between 1990 and 2020. Positive relative migration flows –even constant ones – mean that the relative stock of migrants can still increase, and so it did.

At this point, it is important to keep in mind that UN migration stock data are based on the country-of-birth-concept, whereas the flow estimates aim at counting each act of migration. So the flow estimates do not merely reflect changes in the stocks of migrants. They aim at counting each “regular” outward migration, where a person leaves their country of birth, but also each return and transit migration (i.e. onward migration). Each outward migration increases the global stock of migrants, whereas each return migration reduces it. Transit migration does not change the overall global migrant stock. The relative importance of these three types of migration in the overall flow estimates thus determines how stocks react to flows. Azose and Raftery (2019) show that return and transit migration constitute considerable parts of global migration. They provide estimates of migration flows by the type of movement between 1990 and 2015 and find that approximately 10% of global migration flows can be attributed to transit migration and around 30% to return migration, with the remaining 60% being typical emigration.⁸

What we thus observe in terms of global migration intensity is that while migration did not become more frequent in relative terms between 1990 and 2020 (i.e. constant relative migration flows), the presence of migrants did increase (i.e. rising relative migrant stocks).

Czaika and de Haas (2015) find considerable heterogeneities in migration intensities across world regions. Their analysis concentrates on the period from 1960 to 2000 and finds that emigrant intensity, measured by migrant stocks, decreased somewhat for Europe, which had the highest emigration rate, and also for Africa. Emigration rates from other regions, namely the Americas and Oceania, increased between 1960 and 2000. Immigration rates, on the other hand, increased for Europe (and the Americas), with Europe changing from a region predominantly sending migrants to a receiving region.

Looking at more recent years and using migration flow estimates (see charts A1 and A2⁹ in the annex), we find notable variations in migration intensities and their trends across regions. Globally, however, this yields fairly constant migration rates between 1990 and 2020, but an increasing stock of migrants relative to the population. This first dimension does not contribute to an increase in the globalization of migration (flows) between 1990 and 2000, at least not globally. Regionally, we see mixed evidence: While in the EU, both emigration and immigration intensities increased (which would result in a higher globalization index), in other world regions, we observe little changes in both emigration and immigration rates (Asia) or significant changes in only one of the two (non-EU Europe, Latin America, Oceania). North America saw increases in emigration rates, with immigration rates dropping. So migration intensities will feed rather differently into regional migration globalization indices (see table 1 for an overview).

4.2 Spread of migration flows

The global spread of migrants, i.e. the dispersion of migrants across all possible (bilateral) corridors (country-pairs), see equation 4 in section A1 in the annex, has been increasing since the 1960s, but the increase slowed down considerably in the past two decades. The green line in chart 2 shows the results found by Czaika and de Haas (2015) using data on the stock of migrants: Between 1960 and 2000, global migration was becoming increasingly dispersed across all possible migration corridors: the spread of global migration increased. The chart displays index values (based on equation 4 in section A1 in the annex) and for its application, index values are very close to one. The underlying reason is that well over 40,000 bilateral corridors are involved in its estimate and a large number of already very small shares are squared, added up and subtracted from one. The index may be interpreted as the probability that two randomly drawn migrants travel different corridors (see, for example, Alesina et al., 2003). Given the high number of possible corridors, this probability – and thus the index value – is very high (and that changes in the applications below, where the index is calculated for just above 200 countries). What is interesting for us here, however, is how the index develops over time.

The red line in chart 2 shows how the global spread of migrants has developed since 2000 on the basis of data on migrant stocks. We see that it has continued to increase, but at a slower pace. In other words, the bilateral routes migrants use have continued to get broader and more diverse, but only moderately compared to the increases observed between the 1960 and the 1980s. The blue line shows the spread of global migrants on the basis of estimates of migration flows and reveal a similar pattern: between 1990 and 2020, migrants spread more widely across different migration corridors, but the pace of increase slowed down. ¹⁰ This is also confirmed by a look at the share of “filled” corridors in the migration flow estimates: While in the period 1990–1995, 63% of all possible routes (origin-destination country pairs) were “filled” (non-zero), in the period 2015–2020, 67% were filled. Corridors that opened up between 1990 and 2020 are, for example, Iran to Brazil, Nigeria to Estonia and Sri Lanka to Romania, all of which recorded positive and increasing estimated flows in 1995–2000 or later while showing zero flows in 1990–95. When counting only corridors with a flow of 50 or more, approximately 85 opened up after 1990–95.

When distinguishing between global emigrant spreads (equation 5 in section A1 in the annex) – how dispersed are (e)migrants in terms of the destination countries – and global immigrant spreads (equation 6 in section A1 in the annex) – how diversified are (im)migrants in terms of their origins – Czaika and de Haas (2015) find for the period between 1960 and 2000 that emigrant spreads declined while immigrant spreads increased (see solid and dashed green lines in chart 3). In other words, migrants from more and more diverse origin countries were increasingly concentrated in fewer and fewer different destination countries.

When applying the same method to more recent data on migrant stocks, we find that this diverging trend has been reversed: global migrants were moving to an increasingly diverse set of destinations between 2000 and 2020, and the trend of an ever-narrower set of destinations found for the decades between 1960 and 2000 reversed (see solid and dashed red lines in chart 3). We find the same trend reversal when the indicators are calculated on the basis of estimated migration flows (instead of stocks; solid and dashed blue line in chart 3). This is an important finding: the results show that while immigrants come from a (still) diverse set of origin countries, they are no longer concentrated in an increasingly narrow set of destinations but spread out more across different destinations.

This global aggregate again hides regional heterogeneities (charts A3 and A4 in the annex and table 1). ¹¹ Latin America, for instance, exhibits both the lowest emigrant spread and the lowest immigrant spread among all world regions, but also the strongest increase over time. Immigrant spreads, on the other hand, dropped significantly between 1990 and 2020, from already low levels. Immigrants to African countries are not particularly diverse in terms of their origin countries while emigrant spreads of Asian countries have been increasing since 1990, but a recent flattening is observable. Immigrants to Asian countries do not have highly diversified origins. In North America, emigrant spreads are comparably low and stagnant, while immigrant spreads –which are not particularly high in comparison either – increased over time, especially after 2010.

Chart 3

Turning to the EU, we find that both emigrant and immigrant spreads are high in a global comparison. In the most recent periods, emigrant spreads decreased significantly, the diversity of origins of EU immigrants on the other hand remained consistently high, ranking highest among world regions. In non-EU Europe, the emigrant spread has seen a moderate increase, while the immigrant spread has grown more markedly. The latter development has been particularly driven by an underlying increasing diversity of origins from Asia and Africa, as well as a balancing across existing corridors. Generally, European migration is comparatively diverse in terms of both origin and destination.

4.3 Distances covered by migrants

Chart 4

Czaika and de Haas (2015) find that the geographical distance between origin and destination countries increased considerably between 1960 and 2000. Using more recent data on migration stocks, we find an overall increase between 1990 and 2020. This is mostly due to the 1990s and early 2000s, however. As of the period 2000–05, we do not find a further increase in the average distance covered by global migrants. ¹²

Variations at the regional level are strongly related to the remoteness of countries, with migrants from Oceania having covered by far the largest average distance among regions. Immigrants to and emigrants from North America travel on average 7,000km to 8,000km. The global average migration distance is around 4,000km. The lowest emigration distances are observed for Europe and Africa. Immigration distances are again lowest for Africa, which is related to the high share of within-continent and neighboring-country migration, as well as for Asia. The dynamics over time are limited, and the mild upward trend observed globally can mostly be attributed to increasing distances between source and destination for North American emigrants and immigrants, and African and Asian immigrants.

4.4 An index of the globalization of migration

The index of the globalization of migration combines the three dimensions discussed above to form an overall measure (equation 14 in section A1 in the annex). Looking at global values and the stock of migrants, we see that from the 1970s, migration started to become globalized (see chart 5). This trend gained speed in the 1990s, and the globalization of migration is currently at its highest level. This increase in the index of approximately 12% results from an increase in the intensity of migration when measured by stocks, an increase in the spread of migrants as well as the distances covered by migrants.

Chart 5

When looking at migration flows, which reflect solely current but not historical migration dynamics (see blue line in chart 5, and in more detail in chart 6, as well as the overview table 1), we observe an increase in the globalization of migration between 1990 and 2010, followed by a flattening. At the time of the global financial crisis, the trend toward increasing globalization of migration came to a halt and globalization remained stagnant until the most recent period (2015–20). The underlying developments here are a stagnant intensity of migration, a shift from increasing to flattening, even decreasing, immigrant spreads, and only minimal changes in the average distance covered by migrants in this time span.

Chart 6 further shows the difference between the two components of the globalization index: emigrant dispersion – the globalization of emigration (equation 12 in section A1 in the annex) – and immigrant diversification – the globalization of immigration (equation 13 in section A1 in the annex). Over time, these two components converged, as the globalization of emigration increased faster than the globalization of immigration. Since the intensity of migration flows did not increase, and the average distance lengthened merely in the 1990s, the overall increase is mostly related to the previous finding that emigrants spread more over different destination countries and the diversity of immigrants exhibited a sideways movement in the past three decades.

How does this compare to the flows of goods or capital? Generally speaking, the observation that a steep increase in globalization starting from the mid-1990s was followed by a mere lateral movement at a higher level since the late-2000s also applies to other measures of international economic connectedness such as trade or capital flows or to other measures of globalization such as the KOF Index of Globalization (Gygli et al. 2019; see chart 7).

Chart 6

Table 1

Chart 7

At the country-level, the results are highly heterogeneous across countries (see figure 1). Many small (island) states exhibit the highest degrees of migration globalization. After excluding countries with a population of less than half a million, Australia, the United Arab Emirates, New Zealand, Canada and Kuwait are the countries with the highest degrees of globalization in 2015–20. Many of those high-ranking countries have been typical immigration countries for decades or longer or are resource-rich countries with small populations and large numbers of immigrant workers. European countries exhibit high – but not the highest – degrees of migration globalization. Table 1 shows the direction and statistical significance of changes in the globalization index and the underlying subcomponents.

Figure 1

Source: Authors’ calculations based on Abel and Cohen (2019) migration flow estimates.

Overall, we find a high degree of migration globalization especially in countries that have a long-standing migration history (and thus large diasporas and networks) and resource-rich countries with a high demand for foreign labor (like the Gulf region). ¹³ When relating the globalization of emigration to the GDP per capita in origin countries, and the globalization of immigration to the GDP per capita in destination countries, we find a positive correlation in both cases: the richer a country, the more globalized its emigrants and immigrants. The former is likely related to affordability, with emigrants from poor countries either not being able to emigrate at all or merely to a small set of neighboring (or close) countries. This reduces emigration intensities, spreads and distances and thereby the index. The latter might be generally related to the broad attractiveness of advanced economies for migrants from very different source countries, and the corresponding willingness to move far away from home.

5 Conclusions

Our research on changing trends in the globalization of migration is motivated by the observation that the discussion of (de)globalization focuses predominantly on trade and capital flows, while another important cross-border flow – the flow of people – is often neglected. However, the degree of heterogeneity of migration has important economic implications on destination countries through its effects on productivity and innovation but also through its potential to mitigate challenges associated with aging societies. These effects are likely to influence the optimal course of monetary policy. Against this backdrop, we attempt to systematically address the question of a possible slowdown of the globalization of migration in the recent period of multiple crises.

We follow the seminal work by Czaika and de Haas (2015) and use more recent data on migrant stocks as well as – for the first time – estimates of global bilateral migration flows by Abel and Cohen (2019; most recent update from October 2022) to identify changing trends in the degree of globalization of migration, often attributed to trade or capital flows. We construct an index based on three dimensions: The first dimension, the intensity of global migration, did not increase between 1990 and 2020; the relative magnitude of global flows amounted to roughly 1.4% in the period 1990–95 as well as in our last period of observation, 2015–20. The second dimension reveals that migrants continued to spread more across all possible migration corridors between 1990 and 2020, but at a considerably slower pace than pre-1990. Between 1990 and 2020, we also observe an important change in emigration and immigration spreads: while Czaika and de Haas (2015) found that migrants come from an increasingly diverse set of origin countries and settle in narrowing destination countries, we find the opposite for the periods after 1990: Immigrants still come from diverse (albeit slightly less diverse) origin countries, but the variety of destination countries has increased (again). Thus, we find no evidence that there is an increasing concentration of migrants in fewer destination countries. In fact, quite the opposite is true: after 2000, the dispersion of emigrants over destinations increased considerably. The third dimension shows that the distance between source and destination country for an average migrant did not continue its upward trend that was observed before 2000.

What do these dimensions tell us about a possible deglobalization of migration? We combine them to obtain an index of the degree of migration globalization and find that – on a global scale – the globalization of migration, which was increasing constantly between 1960 and 2000, slowed down from 2005–10 onward. Between 2010 and 2020, a period characterized by multiple crises, migration did not become more globalized. When distinguishing between emigrant dispersion and immigrant diversification, our finding suggests that both did not further increase during this crisis period. Of course, there might be level effects at work here (globalization cannot increase indefinitely), yet it appears that the overall development could also reflect a changing momentum of globalization. To what extent this is the case remains to be seen in future research.

As argued above (section 1), an increase in the diversity of the immigrant population as such contains the potential to generate economic benefits (e.g. via increases in productivity, trade links, etc.). However, this will only be the case if integration into the labor market succeeds. Importantly, the fact that immigrants come from a still highly diverse set of countries underlines the need for flexibility and creativity in active labor market policies. It will be increasingly necessary to implement measures aimed at minimizing potential brain waste, i.e. an underutilization of the human capital of migrants on the labor market. Further, even if the pace of globalization is diminishing, it is still high and hence underlines the necessity for international cooperation. As shown above, there is an increasing number of destination countries and thus cooperative efforts are ever more important. These efforts should be directed toward ensuring timely information about migration trends (again, informing labor market policy) but also to harmonizing migration processes across countries.

References

Abel, G. J. 2019 . Bilateral international migration flow estimates for 200 countries. figshare. Collection. https://doi.org/10.6084/m9.figshare.7731233.v6 (version 6, update 9. figshare. October 28, 2022).

Abel, G. J. and J. E. Cohen. 2019. Bilateral international migration flow estimates for 200 countries. Scientific Data 6(82).

Abel, G. J. and J. E. Cohen. 2022. Bilateral international migration flow estimates updated and refined by sex. Scientific Data 9(173).

Abelianski, A., C. Belabed and J. Mayrhuber. 2024. Enjoy the silence? (De)globalization and cross-border investment – a gravity approach. In: OeNB Bulletin Q2/24. 60–81.

Alesina, A., A. Devleeschauwer, W. Easterly, S. Kurlat and R. Wacziarg. 2003. Fractionalization. In: Journal of Economic Growth 8. 155–194.

Alesina, A., J. Harnoss and H. Rapoport. 2016. Birthplace diversity and economic prosperity. In: Journal of Economic Growth 21. 101–138.

Aiyar, S., J. Chen, C. Ebeke, R. Garcia-Saltos, T. Gudmundsson, A. Ilyina, A. Kangur, T. Kunaratskul, S. Rodriguez, M. Ruta, T. Schulze, G. Soderberg and J.P. Trevino. 2023. Geoeconomic Fragmentation and the Future of Multilateralism. IMF Staff Discussion Note SDN/2023/001.

Azose, J. J. and A. E. Raftery. 2019. Estimation of emigration, return migration, and transit migration between all pairs of countries. In: Proceedings of the National Academy of Sciences 119(1). 116–122.

Bahar, D. and H. Rapoport. 2018. Migration, Knowledge Diffusion and the Comparative Advantage of Nation. In: The Economic Journal 128(612). F273–F305.

Bahar, D., H. Rapoport and R. Turati. 2022. Birthplace diversity and economic complexity: Cross-country evidence. Research Policy 51(8).

Bove, V. and L. Elia. 2017. Migration, Diversity, and Economic Growth. In: World Development 89. 227–239.

Brunow, S., P. Nijkamp and J. Poot. 2015. The Impact of International Migration on Economic Growth in the Global Economy. In: Chiswick, B. R. and P. W. Miller (eds). Handbook of the Economics of International Migration 1B, chapter 19. 1027–1075.

Buch, C. M., J. Kleinert and F. Toubal. 2006. Where enterprises lead, people follow? Links between migration and FDI in Germany. In: European Economic Review 50. 2017–2036.

Buettner, T. 2022. Stocktaking of Migration Data. KNOMAD Paper Series 42.

Conte, M., P. Cotterlaz and T. Mayer. 2022. The CEPII Gravity database. CEPII Working Paper 2022-05. July.

Czaika, M. and H. de Haas. 2015. The Globalization of Migration: Has the World Become More Migratory? In: International Migration Review 48(2). 283–323.

Czaika, M. and S. Orazbayev. 2018. The globalisation of scientific mobility, 1970–2014. In: Applied Geography 96. 1–10.

European Central Bank (ECB). 2022. The role of migration in weak labour force developments during the COVID-19 pandemic. In: Economic Bulletin Issue 1. 52–57.

Edelberg, W. and T. Watson. 2024. New immigration estimates help make sense of the pace of employment. The Hamilton Project – Brookings, Wahington.

Egger, P. H., M. von Ehrlich and D. R. Nelson. 2012. Migration and Trade. In: World Economy 35(2). 216–241.

Engler, P, M. MacDonald, R. Piazza and G. Sher. 2023 . The Macroeconomic Effects of Large Immigration Waves. IMF Working Paper WP/23/259.

Felbermayr, G., V. Grossmann and W. Kohler. 2015. Migration, International Trade, and Capital Formation: Cause or Effect? In: Chiswick, B. R. and P. W. Miller (eds.). Handbook of the Economics of International Migration 1B. 913–1025.

Foged, M. and G. Peri. 2016. Immigrants’ Effect on Native Workers: New Analysis on Longitudinal Data. In: American Economic Journal: Applied Economics 8(2). 1–34.

Fransen, F. and H. de Haas. 2022. Trends and Patterns of Global Refugee Migration. In: Population and Development Review 48(1). 97–128.

Goldberg, P. K. and T. Reed. 2023. Is the Global Economy Deglobalizing? And if so, why? And what is next? NBER Working Paper 31115.

Gygli, S., F. Haelg, N. Potrafke and J.-E. Sturm. 2019. The KOF Globalisation Index – Revisited. In: Review of International Organizations 14(3). 543–574.

Hosny, A. 2020. Remittance Concentration and Volatility: Evidence from 72 Developing Countries. IMF Working Paper WP/20/15.

International Organization for Migration (IOM). 2024. World Migration Report 2024. Geneva.

Iranzo, S. and G. Peri. 2009. Migration and trade: Theory with an application to the Eastern – Western European integration. Journal of International Economics 79(1). 1–19.

Javorcik, B. S., Ç. Özden, M. Spatareanu and C. Neagu. 2011 . Migrant networks and foreign direct investment. In: Journal of Development Economics 94 (2). 231–241.

Koczan, Z., G. Peri, M. Pinat and D. Rozhkov. 2021. The Impact of International Migration on Inclusive Growth. A Review. IMF Working Paper WP 21/88.

Koech, J. and M. A. Wynne. 2017 . Diversification and Specialization of U.S. States. In: Review of Regional Studies 47. 63–9.

Kugler, M. and H. Rapoport. 2007 . International labor and capital flows: Complements or substitutes? Economics Letters 94 (2). 155–162.

Kvalseth, T. O. 2022. Cautionary Note About the Herfindahl-Hirschman Index of Market (Industry) Concentration. In: Contemporary Economics 16(1). 51–60.

Organisation for Economic Co-Operation and Development (OECD). 2014 . Is migration good for the economy? Migration Policy Debates. May. Paris.

Organisation for Economic Co-Operation and Development (OECD). 2021. International Migration Outlook 2021. Paris.

Organisation for Economic Co-Operation and Development (OECD). 2022. The Contribution of Migration to Regional Development. OECD Regional Development Studies. Paris.

Organisation for Economic Co-Operation and Development (OECD). 2023. International Migration Outlook 2023. Paris.

Ortega, F. and G. Peri. 2014. Openness and income: The roles of trade and migration. In: Journal of International Economics 92(2). 231–251.

Peri, G. 2020. Immigrant Swan Song. Finance and Development 03/2020.

Thompson, H. 2022. Disorder – Hard Times in the 21st Century, Oxford University Press, Oxford.

Trax, M., S. Brunow and J. Suedekum. 2015. Cultural diversity and plant-level productivity. In: Regional Science and Urban Economics 53(C). 85–96.

United Nations. 2009. Trends in International Migrant Stock: The 2008 Revision. Department of Economic and Social Affairs, Population Division.

United Nations Development Programme (UNDP). 2020. International Migrant Stock 2020. United Nations Department of Economic and Social Affairs, Population Division. Database POP/DB/MIG/Stock/Rev.2020.

United Nations Development Programme (UNDP). 2022. World Population Prospects. The 2022 Revision. United Nations Department of Economic and Social Affairs, Population Division. Database UN DESA/POP/2022/DC/NO. 9.

UN Trade and Development (UNCTAD). 2024. Key Statistics and Trends in International Trade 2023. Recent trade patterns: slowdown, volatility and heterogeneity. UNCTAD/DITC/TAB/2024/1. United Nations.

Annex

The annex can be found in the pdf version of this study.