• Growing Migration Inequality: What do the Global Data Actually Show?

World Migration Report 2024: Chapter 4

chapter-color-orange
Chapter 4
Growing Migration Inequality: What do the Global Data Actually Show?

Who migrates internationally and where do they go? International migration globally between 1995 to 2020

Chapter Item

In seeking to answer this question, it is important to acknowledge that the ability to offer a perspective at the global level – as part of this World Migration Report – is challenging. As widely acknowledged over many years, statistics to support our collective understanding of international migration patterns and trends are not as well developed as those available in other domains. However, there has been renewed interest in and action on migration statistics, with several major initiatives launched or under way in recent years.47

While migration flow statistics are limited to specific, narrow geographies (see Chapter 2 for discussion),48 a global picture on international migration patterns and trends can be drawn from international “foreign-born” migrant population data.49 Analysis of long-term migrant stock trends allows for insights into where people migrate to, and which countries they emigrate from.50 The UN DESA statistical estimates are widely acknowledged as the main data source on international migrants globally, with separate databases compiled on various categories of migrants (such as migrant workers, missing migrants, internally displaced persons, refugees and asylum seekers).51

Since this chapter re-examines international migration from the perspective of opportunity (or lack thereof), the circumstances of forced displacement are set to one side, in recognition of the lack of choice and the related losses associated with being forcibly displaced. Data on international displacement (refugees and asylum seekers) have, therefore, been subtracted from the international migrant statistics collected by UN DESA in order to produce an estimated total of international migrant stock minus forcibly displaced.52 For a full description of the methods, see Appendix C.

For this analysis, we have also used HDI, which allows for a complementary perspective to that provided by macroeconomic analysis based on country income data. Such macroeconomic contributions to our understanding of global migration have analysed migration-related data against economic indicators, such as gross domestic product or the average income of a household. The outcome of this research has been fruitful, but there is a substantial body of literature suggesting that migration is motivated by income considerations as well as a range of other factors.53 Just as development is more than economic, opportunity to improve well being beyond economic aspects affects migration trends worldwide. Our analysis, therefore, draws upon the broad set of indicators represented in the HDI (see discussion of the HDI in Appendix A). More specifically, our analysis utilizes HDI and migrant stock data from 1995 to 2020. Beginning the analysis in 1995 allows for the inclusion of more countries that did not have reportable data when the HDI was first published; it also allows for geopolitical changes in Eastern Europe following the dissolution of the former Soviet Union. At the time of writing, the most current migrant stock data available are from 2020. However, the effects of COVID-19 on migrants and migration are likely to be significant and may have important impacts on migration patterns well into the future (see Chapter 5 for further discussion).
 

Who has migrated?

As noted above, while the global number of international migrants has increased substantially over the past 25 years, rising from approximately 161 million migrants in 1995 to 281 migrants in 2020, the proportion of international migrants has only slightly increased, rising from 2.8 to 3.6 per cent of the global population over the intervening years. Table 2 shows the difference between 1995 and 2020, disaggregated by United Nations region.54 While absolute numbers of immigrants have increased by tens of millions across all regions, the share of international migrants as a proportion of each region’s population has only marginally increased in Africa, Asia, and Latin America and the Caribbean, while Europe, Northern America and Oceania have seen the proportion of international migrants rise by around 4 percentage points or more in each.

 

Table 2 . Immigrants by United Nations region, 1995 and 2020
Chapter Item

 

Source: UN DESA, 2021.

Table 3 shows both emigrants (origin) and immigrants (destination) further disaggregated at the country level, with the top 20 countries for each category listed in descending order. Countries in Europe and Asia feature as both origin and destination countries for tens of millions of migrants.

 

Table 3. Top 20 countries of origin and destination, by number (millions) and proportion of total population
Chapter Item

 

Sources: UNDP, 2020; UN DESA, 2021.
Note: Uzbekistan did not receive an HDI score until 2000. At that time, the HDI classified Uzbekistan as a medium HDI country. As per UN DESA definitions, emigrants are “foreign born” such that major political changes (e.g. 1947 Partition, dissolution of the Soviet Union) can be reflected in data (further discussion of definitions can be found in Chapter 2). Some categories of international migrant are not included (see methods in Appendix C).

 

Between 1995 and 2020, only a few countries changed from being among the top 20 migrant origin countries (with Portugal, Belarus, the Republic of Korea and Afghanistan included among the top 20 in 1995, but replaced by 2020 by the Bolivarian Republic of Venezuela, Romania, Egypt and Viet Nam). We can see, however, that there are far fewer medium HDI countries of origin by 2020 and no low HDI countries; however, this relates in part to the development progress by countries and their recategorization (discussed further below). The prevalence of high and very high HDI countries as origin countries is quite stark by 2020, accounting for 16 of the 20 top origin countries.

In terms of destination countries as at 1995 and 2020, compared with the top 20 origin countries, there was greater change evident, with five countries dropping out of the list (Pakistan, Côte d’Ivoire, Argentina, Israel and Uzbekistan), being replaced by Spain, Thailand, Malaysia, Kuwait and Japan. With the exception of the Russian Federation, Kazakhstan, India, Jordan and Ukraine, all of the destination countries in both the 1995 and 2020 top 20 lists experienced increases in numbers and proportions of immigrants over this period. Further, Table 3 shows the substantial increase in numbers of immigrants experienced in many destination countries, most notably in the United States of America, Saudi Arabia, Germany, the United Kingdom and the United Arab Emirates. This highlights that while it may be useful to discuss international migrants at the global and regional levels, there are distinct long term country-to-country corridors that account for large proportions of international migration, potentially masking the extent to which migration remains highly uneven globally.55

 

Migration trends through the prism of human development

Current data indicate that most international migrants (79.6% or 190 million) reside in very high HDI countries. We can see, for example, that all of the top 10 countries of destination in Table 3 are very high HDI countries, and the majority of the remaining top destination countries in Table 3 are also very high HDI (with the rest being high HDI countries). This is consistent with long-term trends and existing knowledge that shows that international migration has developed over time as a means for households, families and communities to realize opportunities, including substantial increases in household income via international remittances.56

The current data also highlight that most of the top 20 origin countries are very high (8) or high (8) HDI countries. By 2020, the remaining four origin countries were medium HDI countries.

This is also shown in Figure 5 below, which clearly highlights that international migrants are concentrated in very high and high HDI countries, being most pronounced for immigrants, but also showing significant prevalence among emigrants. In other words, there is a lot more migration occurring in the more developed countries in the world, with lower numbers and proportions in medium and low HDI categories. Interestingly, and contrary to the mobility transitions analysis discussed above (see Figure 3), the very high HDI countries combined have produced a high proportion of emigrants relative to the aggregate population (4.6%), which is higher than high, medium and low HDI categories. Further, in numerical terms, very high HDI countries produced 76 million migrants, second only to high HDI countries (86 million).

 

 

Figure 5. Immigrants and emigrants by Human Development Index country category, 2020
Chapter Item

 

Sources: UN DESA, 2021; UNDP, 2020.
Note: Some categories of international migrant are not included (see methods in Appendix C).

This snapshot in Figure 5 shows that many more emigrants were born in wealthier countries and seem to have moved to other wealthier countries. Other earlier analysis, however, seems to show very different patterns to Figure 6 below, in which 2005 HDI data are used.57

 

Figure 6. Association between Human Development Index scores and immigrant/emigrant stocks, 2005
Chapter Item

 

Source: de Haas, 2010:4, reproduced in de Haas, 2020.
Note: Categorization by author (not UNDP’s HDI 4 categories).

 

In Figure 6, the association between HDI and international migrants is represented, although an author-created fifth category of “very low HDI” based on HDI scores is used (not among UNDP’s four categories), and “average migration values” are applied rather than aggregated migrant stock and population data by category.58 Figure 6 shows that emigrants as a percentage of population are lower from high and very high HDI categories compared with medium HDI, which appears consistent with the “mobility transitions analysis (Figure 3), but different to the current empirical evidence in Figure 6 above.

Lower levels of emigration from low HDI countries is apparent in both figures; however, the two sets of bivariate analyses highlight different rates of emigration from wealthier countries. To explore the difference between the emigration data for high HDI categories represented in Figures 5 and 6, we first looked at changes since 1995. Overall, there appear to be two important but distinct change processes occurring:

  • Significant changes in HDI classification; and
  • Intensifying migration to, as well as from, highly developed countries.

These are now discussed in turn.

Human development index changes since 1995: the up and up

The HDI was developed by economist Mahbub ul Haq and first used by UNDP in 1990 as the centrepiece of its 1990 Human Development Report in an effort to better encompass human aspects in the analysis of development, previously dominated by economic indicators.59 Initially, the HDI covered 130 countries, increasing to 163 in 1995 and progressively reaching a total of 189 countries (see Table 4). All countries that have been reclassified over time have moved into a higher classification in accordance with HDI methods, with the exception of the Syrian Arab Republic (dropping from medium to low in 2015).60 By 2019, 66 countries (or 34%) were classified as very high HDI, and a further 53 (or 27%) were high HDI.61

 

 

Table 4. Number of countries in HDI classifications, 1995 to 2019
Chapter Item

 

Source: UNDP, 2020.

 

So, in part, we can see that reclassification of countries helps explain different migration patterns at different points in time. However, when keeping the 1995 HDI classifications constant (i.e. not adjusting outputs for reclassifications over time), we can also see that there are specific underlying migration dynamics occurring beyond reclassification issues.

Figure 7 below shows the “stepladder” phenomenon over time, even when 2019 classifications are applied across all years (represented by the black dotted lines), so that:

  • There is a marked increase in “migration to” by HDI category (graphs on the left of the series), so that very few people migrate to a low HDI country, more migrate to a medium HDI country, more again to a high HDI and the largest number to a very high HDI country (even when applying 2019 categories).
  • There is a distinct pattern across Figure 7, which shows that “migration from” one HDI classified country to another category (graphs on the right) also follows the “stepladder” principle of moving up. However, reclassifications have clearly impacted on this pattern over time, resulting in a more pronounced emphasis on the very high HDI category.
  • Of particular interest is the “migration within” data (middle graphs), which show significant differences by HDI classification: higher levels of migration to a country with the same HDI classification occur for low to low HDI countries and very high to very high HD countries. We can also see the impact of reclassification, most pointedly for very high HDI countries. Nevertheless, emigration both to and from very high HDI countries is a distinct and clear feature in current migration trends.

 

 

Figure 7. Migrants to, between and from each of the four HDI categories (low, medium, high and very high), 1995–2020
Chapter Item

 

Sources: UNDP, 2020; UN DESA, 2021.
Notes: “Migration to” plots refer to migration to that HDI category from the other HDI category countries; “Migration from” plots refer to migration from that HDI category to the other HDI categories. The data points at the five-year intervals in the colour bands reflect the HDI categorization at that time; the black dotted lines use 2020 HDI classifications across all data points (i.e. 1995 through to 2020). Some categories of international migrant are not included (see methods in Appendix C).

 

Two important conclusions can be drawn from these data:

  1. It is clear that migration from high and very high human development countries to other high and very high countries is pronounced and has increased significantly since 1995 (even accounting for recategorization of countries).
  2.  A question arises as to whether the degree of shift relevant to the migration “hump” model is as relevant today as it previously has been – the bivariate data analysis shows correlations that would benefit from deeper examination.

Of particular relevance is the important factor of policy, and how countries’ visa and mobility policies have evolved over time. As highlighted in the discussion above (and modelled in Figure 2), such policies can enable migration options to be transformed from “impossible dreams” into concrete options, and recent research has pointed to growing mobility inequality.62 To explore this further we examine mobility agreements at the regional level (e.g. the Schengen agreement and the ECOWAS free movement protocol).