2. Literature Review and Theory
In the broad context of the study of economic inequality within societies, one of the most prominent topics is the relationship between inequality and economic prosperity of a country, which is studied by many authors such as Kenworthy (2003) or Wren (2013). With this study, I want to contribute to this research by focusing on one possible mechanism within the inequality-economic prosperity relationship, namely migration. Attracting and retaining highly qualified people is crucial to the development of a knowledge-based economy in which ideas and innovation drive technological and social progress. Moreover, the perception and acceptance of immigration within a society are largely shaped by the skill level of immigrants. Understanding how high-skilled people and in contrast, how low-skilled people choose their country of destination when migrating is, therefore, crucial – for sending and receiving countries alike. Migration and especially cross-border migration is a highly complex topic and the individual decision to migrate is influenced by a multitude of drivers. However, when looking at the differences between high- and low-skilled migrations, economic drivers are especially important, because more than for other drivers such as political or social drivers they affect high- and low-skilled people differently. Moreover, Parey et al. (2018) point out, that studying high-skilled migrants is theoretically interesting because this group usually faces the lowest migration barriers and thus for this group the individual internal motivation should be most relevant for their migration decisions.
Assuming that migration is at least partly determined by the desire to realize economic opportunities abroad and assuming that high-skilled migrants should benefit from the opportunities in less equal destination countries, while low-skilled migrants should benefit from a compressed wage distribution in more equal destination countries, a stream of literature, building on Borjas (1987), studies the relationship between inequality in sending and receiving countries and the consequences for high- and low-skilled migration. Within this thinking and with having in mind the effect of high- and low-skilled migration for economic prosperity, this study empirically tests with immigration data from 20 countries from the Organisation for Economic Co-operation and Development (OECD) from 1985 to 2010 the effect of inequality in a receiving country on high- and low-skilled migration to this country, so that the research question that this study seeks to answer is: Does economic inequality influence the immigration of high- and low-skilled individuals?
2. Literature Review and Theory
This paper builds on the literature on, firstly, the economic impact of high-skilled immigrants, secondly, reasons for the migration of people to a country, and thirdly, the study of inequality in a country in order to examine the influence of inequality in a receiving country on the skill level of immigrants to this country.
The idea of this study and the reason why this study might be interesting is based on the assumption that high-skilled immigration positively affects economic prosperity. No matter how different the studies on this question are, they all agree that high-skilled migration is beneficial for the economic prosperity of a country. See for an overview of contemporary research Hanson et al. (2018), Pekkala Kerr et al. (2017), and Boeri et al. (2012). Furthermore, the study does not address the question of why people emigrate, but it does address the question of where people decide to migrate. Therefore, it builds on the idea of the push-pull model of migration, developed by Lee (1966), assuming that conditions in countries of destination assert a pulling influence on prospective migrants. The theories of migration capabilities and migration aspirations pioneered by Sen (1985) and followed by de Haas (2010) and Carling & Talleraas (2016) expand this concept. De Haas defines migration capabilities as the material and social capital a person can deploy for the migration initiation (de Haas, 2010, p. 16). Migration aspirations, on the contrary, develop according to de Haas from people’s awareness of conditions in another place which makes them want to seek a better life and a local surrounding that cannot offer these opportunities (de Haas, 2010, p. 17). Migration aspiration is, thus, the conceptual idea that will be used in this study to link conditions in a country with the shape of immigration to this country. Geis et al. (2013) for example find with micro‐data for France, Germany, the UK, and the USA that socio‐economic characteristics in destination countries shape the aspirations and consequently the decisions of migrants.
The specific literature on the relationship between inequality and skill-level of immigrants has so far been largely based on the so-called Roy-Borjas model of migrant selection. Building on the so-called Roy model in economics, designed by Roy (1951), that considers the influence of skill-level on the self-selection of workers within a country when deciding between working in different sectors of the economy, Borjas (1987) expands this reasoning to migration and the self-selection of migrants when deciding to migrate to one or another country. Because high-skilled migrants should benefit from the opportunities in less equal destination countries, while less skilled migrants should benefit from more equal conditions in more equal destination countries, the theory suggests that high-skilled migrants are drawn to migrate to less equal countries, such as the United States, while less skilled migrants should favor migrating to more equal countries, such as the Scandinavian countries (Borjas, 1987; Parey et al., 2018). Thus, country-specific characteristics of the income distribution are seen to determine the average skill level of immigrants to this country. Building on this theory, the relationship between the inequality of a country of destination and the self-selection of high-skilled and low-skilled migrants has been studied empirically with mixed results.
Some studies focus on historic migration patterns over a long period. Stolz and Baten (2012) analyze data from 52 sending and five receiving countries for the years from 1820 to 1909, while Abramitzky et al. (2012) focus on Norway-to-US migration between 1850 and 1913. The historic results from both studies from the time of mass migration suggest confirming the Roy-Borjas model of self-selection.
Many of today’s studies on international migrant selection focus on migration from Mexico to the US. Some studies find evidence for negative selection that is consistent with the Roy-Borjas model that the less skilled are those most likely to migrate from the country with high earnings inequality (Mexico) to the country with lower inequality (US). Ibarraran & Lubotsky (2007), Moraga (2011), and Kaestner & Malamud (2014) find that Mexican migrants to the US are negatively selected on earnings and skill level, which is explained by different returns to labor market skills between the US and Mexico. Other studies by Chiquiar & Hanson (2005) and Orrenius & Zavodny (2005) find that migrants from the US to Mexico are in the middle of Mexico’s education, wage and skill distribution, which is inconsistent with the negative‐selection hypothesis from the Roy-Borjas model.
Some studies examine migration between different country-pairs. The selection of migrants from Puerto Rico to the US, for example, is shown to be consistent with the Roy-Borjas model by Ramos (1992) and Borjas (2008). Gould & Moav (2016) use data from emigrants from Israel before they decide to emigrate and find that among high-skilled people, the emigration rate to a country increases with increasing return to education in this country. Gross & Schmitt (2012) study how drivers of migration differ between low- and high-skilled migrants for migration to France. In their results, various factors determine low-skill migration, while high-skilled individuals migrate only according to financial opportunities and expected returns to skill. Parey et al. (2017) find support for the Roy-Borjas model from studying the selection of high-skilled emigrants by investigating migration decisions of graduates from German universities and using predicted earnings as an operationalization of skill level. They find that graduates who emigrate to less equal countries have higher predicted earnings than non-migrants, while people moving to more equal countries have lower predicted earnings than non-migrants (Parey et al. 2017, p. 781-787).
Only a few studies focus on migrant selection between multiple countries, but without a specific inequality focus. Borjas et al. (2018) find support with cross-country data for the Roy-Borjas model’s prediction but only due to higher pre-emigration earnings of high-skilled people, while Feliciano (2005) and Grogger & Hanson (2011) reject the Roy-Borjas model of skill bases self-selection.
There is, to my best knowledge, no study that examines specifically the influence of economic inequality in different countries of destination on the selection of migrants between these countries. While the Roy-Borjas model only predicts that migrants choose their country of destination in accordance with individual income maximization, this study specifically focuses on the role of inequality as shaping migration aspirations and thus its role as a selection criterion between destinations.
Abbildung in dieser Leseprobe nicht enthalten
Figure 1: Theoretical model by the author based on de Haas (2010) and Carling & Talleraas (2016)
It enriches the framework of the Roy-Borjas model with the idea of migration aspirations to analyze if highly skilled migrants, when choosing between OECD destination countries, are more likely to migrate to more unequal countries because they can benefit from the opportunities there, while less skilled migrants are more likely to move to more equal countries because they can benefit from the insurance of a narrower wage distribution there.
Figure 1 shows a theoretical model of the influence of inequality within a country on the share of high-skilled and low-skilled immigrants through the different effect of inequality on the migration aspirations of high- and low-skilled people.
This theory leads to two separate hypotheses that are tested individually because they can be both true or false independent from the other hypothesis. It is possible for example that inequality is an important negative-selection criterion for low-skilled migrants while the high-skilled migrant country selection is not influenced by inequality at all.
Abbildung in dieser Leseprobe nicht enthalten
To test the hypotheses empirically, I use country-year data and fixed-effects panel regression models. The fixed-effects model allows eliminating all time-invariant country-specific characteristics such as the national immigration policy. Because these factors most likely largely determine the shape of immigration and the number of high- and low-skilled immigrants, they largely affect the results. With the fixed-effects model, I only look at the within-country changes and cancel out these country specifics under the strong assumption that these characteristics are not changing over time.
For the dependent variable, I will use country-year data on migration stocks in 20 OECD countries, disaggregated by country of birth and educational level in five-year steps for the time from 1980 to 2010 from the Institute for Employment Research in Nuremberg, Germany (Brücker et al. 2013). The data-set contains data on the educational level of the immigrants, which I will use to calculate the share of high- and low-skilled immigrants as a proxy for the individual skill level. The educational level is given in the data set as low, middle, and high level of education. I calculate the share of high-skilled immigrants (share_high_edu) by dividing the number of high-skilled divided by the number of total immigrants and the share of low-skilled immigrants (share_low_edu) by dividing the number of low-skilled divided by the number of total immigrants. The final data-set contains 140 observations for each dependent variable, from 20 OECD countries (Australia, Austria, Canada, Chile, Denmark, Finland, France, Germany, Greece, Ireland, Luxembourg, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom, United States) at 7 points in time (1980, 1985, 1990, 1995, 2000, 2005, 2010).
The variable share_high_edu is coded between 0 and 1, with a minimum value of about 0.05 (France in 1980) and a maximum value of about 0.68 (Canada in 2010). The mean value is 0.25 and the standard deviation is 0.11. The variable share_low_edu is also coded between 0 and 1, with a minimum value of about 0.16 (Norway in 2010) and a maximum value of about 0.90 (France in 1980). The mean value is 0.48 and the standard deviation is 0.16.
For the independent variable, I operationalize inequality within a country with the Gini coefficient, which is a measure for the inequality of the income distribution in a country. The data for the Gini index comes from the World Bank for the years from 1980 to 2010 (World Bank, 2019). Unfortunately, many observations are missing, especially for the early years, so that I end up with only 59 observations. However, other inequality measurements such as the Uneven Economic Development index (with a one-to-ten scale) from the Fragile States Index project looking at inequality in economic opportunities through pay gaps, lack of access to vocational training or discriminatory hiring practices, cover an even smaller periods of time, which is why I am sticking to using the Gini coefficient data (The Fund for Peace, 2018). The independent variable (gini) I coded from 0 to 100, whereby 0 means perfect equality and 100 maximum inequality (World Bank, 2019). The median level 34.19, the standard deviation is 6.77, the minimum value 25.2 (Denmark in 2005), and the maximum value 57.2 (Chile in 1990).
- Quote paper
- Simon Valentin (Author), 2019, Does economic inequality influence the immigration of high- and low-skilled individuals?, Munich, GRIN Verlag, https://www.grin.com/document/497772