The drivers of choosing a migration destination

Several studies (De Jong and Gardner Reference De Jong, Gardner, De Jong and Gardner1981; Faist Reference Faist2000; Hagen-Zanker Reference Hagen-Zanker2008; Hemmerechts et al. Reference Hemmerechts, De Clerck, Willems and Timmerman2014) have categorized the drivers of migration decision as follows: (1) societal-level (macro) studies explain aggregate migration trends as a part of the social, political, and economic development of a country; (2) household-level (meso) studies link migration cause or perpetuation to household or community based on social ties and networks; and (3) individual-level (micro) studies examine individual values, desires, and expectations focusing on the cost-benefit calculation of potential immigrants. Since this article focuses on the macro level, it adopts the word “drivers” commonly used for the macro level analysis (Carling and Collins Reference Carling and Collins2017). Based on these works (Carling and Collins Reference Carling and Collins2017), we adopt Dudu’s (Reference Dudu2018) re-categorization which divides the characteristics of migration destinations into four groups: security-based, labor market, social, and geographic drivers (see Figure 1).

Security-based drivers

Security-based drivers have links with migration caused by obligations (human insecurity) due to the public or private actions (like war conditions, slavery, and human trafficking) and environmental disasters (like climate change or destructive earthquakes) (Dudu Reference Dudu2018; De Jong and Gardner Reference De Jong, Gardner, De Jong and Gardner1981; George Reference George and Clifford1970). A valid legal residence status is a security measure (Eurostat 2011) that eliminates the fear of deportation (Fasani Reference Fasani2014). Gaining a legal residence document is easier in a destination country with an open border immigration policy (Bartram Reference Bartram2010; Velasco Reference Velasco2016), therefore, potential immigrants may consider a country’s immigrant-friendly policies when choosing a destination (Carling Reference Carling2002; Ozcurumez and Yetkin Aker Reference Ozcurumez and Yetkin Aker2016). An immigrant with a work permit can enjoy the opportunity to work, unemployment benefits, and bargaining power over wages and working conditions (Bailey Reference Bailey1987; Fasani Reference Fasani2014; Kossoudji and Cobb-Clark Reference Kossoudji and Cobb-Clark2000; Rivera-Batiz Reference Rivera-Batiz1999).

Despite the visa restrictions, “a well-founded fear of persecution in their home country based on their race, religion, nationality, political opinion, or membership in a particular social group” (UNHCR 1951) is the reason for seeking asylum. However, these people can also apply for long-term residence permits abroad if they meet the requirements.

Running away from social and/or political pressure is a strong security-based migration motivation. For example, LGBTQ+ people may want to leave their home countries due to social pressure and choose a country with broader democratic rights. Similarly, highly skilled women suffering from gender inequalities may decide to migrate abroad (Elveren and Toksöz Reference Elveren and Toksöz2019). Some minority groups might face social exclusion such as alienation and lack of freedom (Bhalla and Lapeyre Reference Bhalla and Lapeyre1997) and social coercion, that is, the constraint experienced as a result of people’s attitudes and opinions (Aseh Reference Aseh1955; De Crespigny Reference De Crespigny1964), which hinder their participation in daily social and economic activities (Chakravarty and D’Ambrosio Reference Chakravarty and D’Ambrosio2006). People may wish to emigrate to a country where they can avoid social exclusion (Abrahamson Reference Abrahamson1995), even if this means becoming a minority multiple times (e.g. being an immigrant LGBTQ+).

People may desire to immigrate when their home country does not provide democratic conditions—the right to be protected, the right to participate in the political process, and the right to equal status (Cedefop 1998; Lawson Reference Lawson1991; Sirkeci and Eroğlu Utku Reference Sirkeci and Eroğlu Utku2020). Escape through emigration from social coercion and political oppression is related to identity (Benson and O’Reilly Reference Benson and O’Reilly2016) because of the emigrants’ desire to express themselves freely.

The EUMAGINE project conducted a survey and in-depth semi-structured interviews in Morocco, Turkey, Ukraine, and Senegal between 2010 and 2013. The results demonstrated the role of perceptions on human rights and democracy over migration aspirations (Carling and Schewel Reference Carling and Schewel2017; Hemmerechts et al. Reference Hemmerechts, De Clerck, Willems and Timmerman2014). It showed that the EU has a positive image in terms of democracy and human rights in the migration-impacted districts of Turkey (Timmerman et al. Reference Timmerman, Hemmerechts and De Clerck2014).

Labor market drivers

The labor market drivers are more employment opportunities, higher wages, lower cost of living, and better working conditions (like fewer working hours) (Tabor et al. Reference Tabor, Milfont and Ward2015). According to Jennissen (Reference Jennissen2003), Keynesian economic theory attributes international migration to employment rate differences (Piore Reference Piore1979), while neo-classical economic theory (Sjaastad Reference Sjaastad1962) explains it as the consequence of wage differences, which lead to differences in income (Harris and Todaro Reference Harris and Todaro1975).

Social drivers

Social drivers include cultural similarities. For example, interviews with people who migrated from the UK to New Zealand revealed that cultural similarities affected the migration destination. They believed that living in New Zealand is similar to living in the UK due to the cultural overlap, but ultimately life there is better than in the UK (Tabor et al. Reference Tabor, Milfont and Ward2015).

Another advantage for easier adaptation is speaking the language of the destination country or having linguistic proximity between the immigrant’s language and the language of the destination country so as not to face a language barrier (Eurostat 2009; Tabor et al. Reference Tabor, Milfont and Ward2015). If newcomers cannot speak the primary language in the destination country, then they generally work in the enterprises of earlier immigrants from the same region. These migration networks can help the newcomer integrate more quickly (Dudu Reference Dudu2018).

Migration networks are created through kinship or cultural ties (Guilmoto and Sandron Reference Guilmoto and Sandron2001; Hagen-Zanker Reference Hagen-Zanker2008). Such networks help newcomers to adapt by providing support regarding jobs, housing, education, and cultural issues (Beine et al. Reference Beine, Docquier and Özden2010). Another feature of migration networks is remittances from the destination to the origin country (Day and İçduygu Reference Day and İçduygu1999). Migration networks also provide information to newcomers before they arrive (Hagen-Zanker Reference Hagen-Zanker2008; Roseman Reference Roseman1983).

Geographical drivers

A moderate climate (De Jong and Fawcett Reference De Jong, Fawcett, De Jong and Gardner1981; European Commission 2006; Lee Reference Lee1965; Tabor et al. Reference Tabor, Milfont and Ward2015; Thompson Reference Thompson2017) and environmental quality, including landscape (Berger and Blomquist Reference Berger and Blomquist1992; Tabor et al. Reference Tabor, Milfont and Ward2015), cause retirement migration from powerful economies to countries with moderate climates to reduce the risks of living on a low retirement income (Karakaya and Turan Reference Karakaya and Turan2006; Özerim Reference Özerim2012; Südaş Reference Südaş2009; Williams et al. Reference Williams, King and Warnes1997). It also includes lifestyle migration, which refers to migration to achieve a more fulfilling lifestyle (Torkington Reference Torkington2010). For example, UK citizens living in Portugal’s Algarve region have migrated for lifestyle reasons, including a moderate climate, slower pace of life, healthier diet, more sociable culture, and more leisure opportunities (Torkington Reference Torkington2010). Interestingly, Turkish migration stock in the Mediterranean-European countries like Italy, Spain, and Portugal—presumably attractive due to their moderate climate—has been increasing slowly but steadily in recent years (UNDESA 2017b).

The destination choice is also affected by the distance between the destination and the country of origin (Dedeoğlu and Genç Reference Dedeoğlu and Deniz Genç2017). Distance increases travel costs and time, and limits the available means of transportation. Thus, a nearby destination country may be attractive for feeling close to the home country, while a potential immigrant is more likely have options to travel by road, rail, or sea.

Figure 1.

Drivers of migration destination.

Source: Dudu (Reference Dudu2018).

Modeling the drivers of migration destinations

Linear regression is a commonly used technique for modeling migration destination (Rogers Reference Rogers2006). Some studies (Nica Reference Nica2015; Pânzaru Reference Pânzaru2013; Winter Reference Winter2019) showed the importance of economic drivers such as the cost of living, income opportunities, GDP, and the unemployment rate. Others (Geis et al. Reference Geis, Uebelmesser and Werding2015; Mayda Reference Mayda2010; Tabor et al. Reference Tabor, Milfont and Ward2015) demonstrated the significance of non-economic factors such as quality of life, safety, environment, and cultural similarity, as well as a migrant-friendly perception, good education and health systems, and the existence of migration networks.

Dedeoğlu and Genç (Reference Dedeoğlu and Deniz Genç2017) examined Turkey as a home country, focusing on migration drivers for 31 European destination countries. They drew on the gravity model, which emphasizes distance over migration destinations. Their main findings were as follows: better economic conditions strengthen Turkish nationals’ migration to that country; populations in the home and destination countries increase migration stock; Turkish migration stock in Europe significantly influences the migration choices Turkish nationals make; and “volume of immigration decreases with distance and conversely increases with contiguity” (Dedeoğlu and Genç Reference Dedeoğlu and Deniz Genç2017, 14).

Data

This article uses panel data blended from the databases of the OECD, Eurostat, and the World Bank. The central hypothesis is that the labor market, which was a key factor during the twentieth century, is no longer the only significant driver for choosing a migration destination in the twenty-first century. This hypothesis is tested with data for four drivers: security-based, labor market, social, and geographic drivers.

The article focuses on documented (i.e. legal) immigrants. The dependent variable is the number of first residence permits issued (i.e. for more than 12 months; the abbreviation is “FirstPermit”) to Turkish nationals (Eurostat 2019b). Since people prefer to immigrate to countries where they can more easily obtain a residence permit, the security-based drivers are effective over the documented immigrants. A person who fears persecution in Turkey can apply for and hold a long-term residence permit in the EU because the line between the concepts “voluntary” and “forced” migration is blurred (De Haas Reference De Haas2021) (see Table 1).

Table 1.

First permits issued for Turkish nationals by the member states which signed labor force agreements with TurkeyFootnote ⁴

Source: Eurostat (2019b).

In 2018 almost half of the first residence permits received by Turkish nationals were issued for family reasons, which is a social driver resulting from the connection with migration networks. In contrast, although labor migration used to be an important reason for Turkish nationals entering Europe, only around 16 percent of the permits were issued for labor reasons (Eurostat 2019b) (see Table 2). Other studies (Kirişçi Reference Kirişçi2007; Kulu-Glasgow and Leerkes Reference Kulu-Glasgow and Leerkes2013; Timmerman et al. Reference Timmerman, Lodewyckx and Wets2009) also report that currently many Turkish immigrants continue to move to the EU for family reasons.

Table 2.

Reasons for first permits issued for Turkish nationals in 2018 by EU member states that signed a labor force agreement with Turkey

Source: Eurostat (2019b).

The following are the independent variables:

Methodology

The panel data included 11 years (from 2008 to 2018) and 28 EU countries (including the UK), yielding 308 observations. After transforming the model into log-log form, all the independent variables for Turkey (EmpT, WageT, WorkHourT, LivingCostT, FreedomsT, and ColdDaysT) were subtracted from all the independent variables for the member states (EmpMS, WageMS, WorkHourMS, LivingCostMS, FreedomsMS, and ColdDaysMS). Since using data at the macro level does not allow the development of a more complex approach to the interaction of macro and micro levels, we assume that the differences between the home country and destination are highly correlated with migration motivation (Dudu Reference Dudu2018; Sirkeci Reference Sirkeci and Uğuzman2018). For example, if the difference in employment rates between the two countries is high, the probability of migrating for employment reasons to the country with a high employment rate from the country with a low employment rate is likely to be high. Thus, while the dependent variable remained the same, new independent variables were created considering these differences:

$\rm{In(EmpDif)=ln(EmpMS)-ln(EmpT)}$

$\rm{ln(WageDif)=ln(WageMS)-ln(WageT)}$

$\rm{ln(WorkHourDif)=ln(WorkHourMS)-ln(WorkHourT)}$

$\rm{ln(LivingCostDif)=ln(LivingCostMS)-ln(LivingCostT)}$

$\rm{ln(FreedomsDif)=ln(FreedomsMS)-ln(FreedomsT)}$

$\rm{ln(ColdDaysDif)=ln(ColdDaysMS)-ln(ColdDaysT)}$

The estimation was made using the ordinary least square (OLS) model. The following model was used for the estimation:

\begin{align}{\rm{ln}}\left( {{\rm{FirstPermit}}{_{ijt}}} \right) &= {{{\rm{\beta}} _0}} + {{\rm{\beta}} _1}{\rm{ln}}\left( {{\rm{EmpDif}}{_{jt}}} \right) + {{\rm{\beta}} _2}{\rm{ln}}\left( {{\rm{WageDif}}{_{jt}}} \right) + {{\rm{\beta}} _3}{\rm{ln}}\left( {{\rm{WorkHourDif}}{_{jt}}} \right)\\ &\quad + {{\rm{\beta}} _4}{\rm{ln}}\left( {\rm{LivingCostDif}}{_{jt}} \right) + {{\rm{\beta}} _5}\left( {{\rm{FreedomsDif}}{_{jt}}} \right)\\ &\quad + {{\rm{\beta}} _6}{\rm{ln}}\left( {{\rm{ColdDaysDif}}{_{jt}}} \right) + {{\rm{\beta}} _7}\left( {{\rm{ExOttoman}}{_{ij}}} \right)+ {{\rm{\beta}} _8}{\rm{ln}}\left( {{\rm{TMigSt2005}}{_{ij}}} \right)\\ &\quad + {{\rm{\varepsilon}} _{ijt}}\end{align}

A high employment rate indicates a high probability of finding employment (Ortega and Peri Reference Ortega and Peri2009). Similarly, higher wages (European Commission 2006) attract immigrants, so here the expected relationship is also positive. Better working conditions in a country, like more leisure time, is a positive motive (De Jong and Fawcett Reference De Jong, Fawcett, De Jong and Gardner1981). Accordingly, countries with lower-than-average normal weekly hours in the main job are more attractive, so a positive relationship is expected. On the other hand, a higher cost of living (Berger and Blomquist Reference Berger and Blomquist1992; Cedefop 1998) is negatively related to applying for a work permit.

The perception that Europe supports democracy and human rights increases the probability of migrating there (Timmerman et al. Reference Timmerman, Verschragen, Hemmerechts, Timmerman, Clycq, Levrau, Van Praag and Vanheule2018). Likewise, since Turkey has greater cultural/historical proximity with countries ruled by the Ottoman Empire and Turkish nationals still have relatives in these countries, a positive relationship is expected. Conversely, since a moderate climate increases immigration, a negative relationship is expected.

The descriptive statistics show a wide range of residence permits issued yearly for Turkish people by each member state. For example, one country issued only three long-term residence permits per year whereas another issued more than 17,000 (see Table 3). The total number of residence permits issued follows a similar trend to the number of residence permits issued for Turkish nationals from the countries with labor force agreements with Turkey (see Graph 2).

Graph 2.

Number of first permits issued for Turkish nationals from EU member states.

Source: Eurostat (2019b).

Table 3.

Descriptive statistics

Turkey’s employment rate, wages, working hours, and democracy level are lower than those in EU member states, whereas its climate is more moderate. Turkey’s employment rate has gradually increased since 2008, reaching a peak of 51.98 percent in 2018 (OECD 2018), although this was still lower than the EU-28 average. Wages are also lower in Turkey. Average weekly working hours range from 30 to 42 hours in the EU, while they had decreased from 51.7 to 47 hours in Turkey by 2018 (OECD 2019). Turkey’s democracy level, which was already lower than any EU country, has fallen sharply from 46.15 to 25.12 since 2008 (World Bank 2019). When comparing the average heating degree days with countries in the EU, like Greece, southern Italy, and Spain, Turkey has a moderate Mediterranean climate (Spinoni et al. Reference Spinoni, Vogt, Barbosa, Dosio, McCormick, Bigano and Füssel2017) (see Table 3).

Table 4 shows us that some variables are highly correlated. For example, FirstPermit and TMigSt2005 are highly positively correlated at 0.85. Signing a labor force agreement with Turkey significantly strengthens the migration network between two countries, encouraging a significant number of Turkish citizens to live in these member states. Another high correlation is between lnWageDif and lnLivingCost at 0.92. The reason for this high correlation is that people can adopt different product search strategies to buy a similar product with a different brandmark (Committee on Finance US Senate 1995) thus paying less when they cannot afford that product anymore.

Table 4.

Correlations