The Diversification of Intra-European Movement

Abstract

The literature on free movement among the 28 member states of the European Union (EU) divides along two main themes. The first set of studies analyses the nature and type of intra-European movement. Here, the central focus tends to be one of the main migration flows within Europe – namely, the movement of Central and Eastern Europeans (CEE) who are mostly citizens of the EU (Black et al. 2010). In migration studies, this freedom of movement within the EU is increasingly articulated in terms of ‘East’ to ‘West’ migration. Many studies show how this migration pattern within Western Europe is expanding beyond the seasonal and circular forms of labour migration that initially characterized CEE migration, to one characterized by more permanence (e.g., Recchi 2008; Favell 2008; Eade et al. 2006; Düvell and Vogel 2006; Engbersen et al. 2011, 2013; Castro-Martin and Cortina 2015). To illustrate, Engbersen et al. (2011, 2013) contend that along the two axes of attachment to country of origin and destination, there are now four categories of intra-European movement: (1) bi-national; (2) circular; (3) settlement and; (4) “footloose”. This typology captures the fact that some migrants preserve transnational ties that attach them to their region of origin as well as their host countries while others remain permanently in receiving societies, later reuniting with family members or establishing new families in the receiving country (ibid.). Others still may continue their expedition to other parts of Europe or may rather end up “footloose”, experiencing problems accessing the labour market in the receiving country as their ties with the home country fade (ibid.). Studies also show that different types of migration are related to different stages of migration, moving from an initial stage of temporary work abroad, through transnational commuting to permanent settlement, where these phases largely depend on the labour market opportunities in the destination regions (Friberg 2012).

1 Introduction

The literature on free movement among the 28 member states of the European Union (EU) divides along two main themes. The first set of studies analyses the nature and type of intra-European movement. Here, the central focus tends to be one of the main migration flows within Europe – namely, the movement of Central and Eastern Europeans (CEE) who are mostly citizens of the EU (Black et al. 2010). In migration studies, this freedom of movement within the EU is increasingly articulated in terms of ‘East’ to ‘West’ migration. Many studies show how this migration pattern within Western Europe is expanding beyond the seasonal and circular forms of labour migration that initially characterized CEE migration, to one characterized by more permanence (e.g., Recchi 2008; Favell 2008; Eade et al. 2006; Düvell and Vogel 2006; Engbersen et al. 2011, 2013; Castro-Martin and Cortina 2015). To illustrate, Engbersen et al. (2011; Engbersen et al. 2013) contend that along the two axes of attachment to country of origin and destination, there are now four categories of intra-European movement: (1) bi-national; (2) circular; (3) settlement and; (4) “footloose”. This typology captures the fact that some migrants preserve transnational ties that attach them to their region of origin as well as their host countries while others remain permanently in receiving societies, later reuniting with family members or establishing new families in the receiving country (ibid.). Others still may continue their expedition to other parts of Europe or may rather end up “footloose”, experiencing problems accessing the labour market in the receiving country as their ties with the home country fade (ibid.). Studies also show that different types of migration are related to different stages of migration, moving from an initial stage of temporary work abroad, through transnational commuting to permanent settlement, where these phases largely depend on the labour market opportunities in the destination regions (Friberg 2012).

A second set of studies considers this free movement as a form of socio-economic participation on the European labour market, where migrants are seen as a key driver of economic integration. This stream in the literature focuses more on the subsequent effects of free movement, largely taking it for granted that this movement is essentially about ‘labour migration’ (Pascouau 2013; Martín and Venturini 2015; Barslund et al. 2015). While some studies underline the phenomenon of “social dumping” produced by migrants’ readiness to work for low wages in bad working conditions (Amelina and Vasilache 2014), others question if this is a form of welfare – rather than labour – migration. Still others argue that liberalization of movement in the EU is producing a workforce that is more aware of the European dimension of the labour market (Andrijasevic and Sacchetto 2015).

Based on a research conducted in Austria, Sweden, the Netherlands and Turkey in 2013–2016, this chapter largely contributes to the first line of literature on intra-European movement with one key argument: diversification of CEE migration. While drawing the main migration corridors from CEE countries to the less-focused cases of Austria, Sweden, the Netherlands, and Turkey, we argue that migration patterns within these corridors are diversified on two levels: time (i.e., temporary versus permanent migration), and; socio-economic status (i.e., high-skilled workers versus non-workers). This typology is also presented in Chap. 1, as Fig. 2.1. Consequently, in that typology we choose eight categories or types of migrants (TOMs) that cover most of the surface of the typology. These are: 1) knowledge workers; (2) entrepreneurs; (3) manual workers; (4) persons working in private households; (5) sex workers and trafficked persons; (6) students; (7) non-working spouses and children, and; (8) beggars and homeless people. This categorization on time and socio-economic status is an important contribution to the literature as it avoids the typical singular reliance on migrants’ ethnicity/national citizenship (which holds limited analytical relevance) to focus on their labour-market position instead.

Fig. 2.1

CEE migration before and after the 2004 and 2007 enlargements

Country (Before/After): Austria (2002/2013), Netherlands (2003/2013), Sweden (2000/2012), Turkey (2003/2012)

Source: Sert (2014), updated by the author

Migration corridors are utilized here as frames of observation. The concept of corridors enables us to observe different forms of asymmetry in movement (Carling 2015). The destination countries chosen for this research represent different stages in relation to CEE migration. To illustrate, while Austria and the Netherlands issued transitional arrangements concerning intra-European ‘labour migration’ during the 2004 and 2007 enlargements, Sweden opted out of such a policy choice. Turkey is included in the analysis as a case study with a very different immigration regime, since the EU regulations on CEE migration are not applicable. Moreover, the corridors are not evaluated solely at the country level. Different urban regions were also included in the analysis adding leverage to the analysis of the diversity of CEE migration. Two urban regions were included for each country: Vienna and Linz in Austria, The Hague and Rotterdam (the Netherlands), Stockholm and Gothenburg (Sweden), and Edirne and Istanbul (Turkey). The concentration on urban regions uncovered further diversification of CEE migration, since we observed more transnational patterns in bordering urban regions, such as Linz and Edirne. Still, considering a multilevel perspective, the country level is more determinant than the specific urban level to identify responses, policies and even consequences of CEE migration, where cities located in the same country often seem to display similar responses and register the same consequences.

The chapter proceeds as follows. The first section presents the methodology, underlining the issues of inaccessibility and incomparability of available data. The second part provides a brief outline of the historical background of East-West migration before the 2004 and 2007 enlargements. Subsequently, the effects of the eastern enlargement of the EU are analysed with a specific focus on the transitional arrangements where the four countries chosen for the research represent comparable cases. The fourth section defines and details the main migration corridors. The fifth part depicts the types of migrants involved in each country and urban region, with a focus on the feminization of migration. The final section presents some concluding remarks.

2 Methodology and Issues

As outlined in the methodological considerations of Chap 1, research teams in Austria, Sweden, the Netherlands and Turkey collected available data on CEE migration pertaining to their respective countries, and tried to synchronize these figures to produce systematic comparisons. Each research team outlined and analysed existing data on the number of CEE migrants across four data clusters: (1) country of origin, gender, educational levels, stocks and flows; (2) migration motives (i.e., labour, family, and/or study); (3) duration of migration (i.e., temporary versus permanent), and; (4) labour market participation (i.e., whether employment was formal or informal). All the teams followed a dual approach on methodology.

First, as mentioned in Chap. 1, they utilized secondary sources, reviewing the literature of previous research as well as available official statistics. This provided with a valuable source on stocks and flows of migrants including basic demographic features such as age and gender, but lacks important information on education, professional background or legal status. As such, although CEE migration in Europe is a highly debated issue, finding comparable data proved difficult. To demonstrate, the Central Population Register (CPR) in Austria works with a system of continuous reporting on the changes of main residences at the municipal level covering all persons who have registered a main residence in Austria for more than 90 consecutive days. This provides a valuable source for an overview on stocks and flows of migrants including basic demographic features such as age and gender, but lacks important information on education, professional background or legal status. In the Netherlands, population statistics provide reliable information about the registered ‘migrants’ in the country. However, only those migrants who intend to stay in the Netherlands more than 4 months need to register, making statistics incomplete and socially selective. There are analyses of CEE migration derived from representative samples from these population registers, but the problem remains that many CEE migrants are not registered. There are also survey studies based on non-representative sampling of CEE migration, which provide better insight into the more volatile category of temporary (and often less integrated) CEE migrants in the Netherlands, but the unemployed and other non-working migrants tend to be absent in these. In Sweden, data on CEE migration were derived from official statistics (i.e., from Statistics Sweden, the Swedish Work Environment Authority, the National Board of Health and Welfare, and the Swedish Higher Education Authority). Like the Netherlands, there is a time dimension in the population registers where only those migrants who declare their intention to stay for at least one year in Sweden are included in the population statistics. Moreover, like in Austria the official data hardly provide any information on the educational levels of CEE migrants. In Turkey, the problem of availability of data is more critical than other cases. International migration in Turkey became a policy concern only in the late ‘90s, and collection and distribution of data have been considered as statistically important only after the establishment of the Directorate General of Migration Management in 2014, as envisioned by Law 6458 on Foreigners and International Protection (YUKK) adopted in 2013.

Secondly and additionally – especially where no official data were available – teams conducted qualitative interviews to estimate the scale of CEE migration and to uncover the different types of migration. Within this second approach, semi-structured interviews with different stakeholders were conducted. These included representatives of CEE migrants in the different migrant organizations, officials from local government, relevant private outfits (e.g., labour recruitment agencies), and NGOs involved in migration-related areas (e.g., housing and education). In that regard the upcoming paragraphs are the outcome of the exploring, classifying and summarising quantitative data gathering steps in this project. But overall, although CEE migration in Europe is a highly debated issue, finding data – especially comparable data – proved difficult.

3 CEE Migration Before the Enlargements

Migration from CEE countries to Austria, Sweden, the Netherlands, and Turkey occurred long before the EU enlargements. In Austria, two groups of CEE migrants have been of particular importance. The first are refugees from the communist countries (who continued to arrive until the late 1980s) and the second are labour migrants and their families. Due to its neutrality during Cold War, Austria received three major waves of refugees: from Hungary in 1956, Czechoslovakia (1968) and Poland (1981). For many, Austria was only a transit country on their way to other destinations (Enengel et al. 2014). At the beginning of the 1960s, Austria was confronted with a growing need of additional labour and started to recruit workers from countries such as Turkey and Yugoslavia (ibid.). Despite the original model – a rotation of temporary ‘guest workers’ – many of these migrants stayed and brought their families to Austria. To this day, Turkish- and Balkan-origin workers comprise the most important and growing foreign-born groups in the country (ibid.).

Like Austria, the Netherlands also received refugees from Hungary, Czechoslovakia, and Poland during the same periods (1956, 1968 and 1981, respectively) (Bonjour 1980: 48; also cited in Ostaijen et al. 2014). Unlike other countries, the Netherlands also faced a unique type of movement of those Polish soldiers who had fought with the Allied forces to liberate the Netherlands from German occupation during World War II. These soldiers often stayed and married Dutch women. There was also marriage migration of women from CEE countries, particularly from Poland, who married Dutch men, the so-called ‘Polish brides’ (Dagevos 2011; also cited in Ostaijen et al. 2014). Again, like Austria the Netherlands has also been an important destination for labour migration. Even before and shortly after World War II several thousand CEE nationals, mainly from Poland and Slovenia, arrived to work in the Dutch mines (Brassé and Van Schelven 1980; also cited in Ostaijen et al. 2014). 50,000 Poles were estimated to work as seasonal workers in Dutch horticulture in the late 1980s (Dagevos 2011: 31; also cited in Ostaijen et al. 2014).

Like Austria and the Netherlands, Sweden was (and still is) a destination for humanitarian immigration, and in increasing numbers (Boguslaw 2012; also cited in Zelano et al. 2014). Unlike the two other cases, Sweden followed a more cautious immigration policy in terms of labour migration. In the wake of the economic downturn at the beginning of the 1970s, Sweden allowed labour migration only if demand could not be met by the existing domestic workforce. These obstacles for free movement were removed with Sweden’s accession into the EU in 1995 (Zelano et al. 2014).

Immigration from CEE followed a different trajectory in Turkey. In the Early Republican Period, usually categorized under Balkan migrations, there were large influxes from Bulgaria, Romania, and Yugoslavia. After the war of independence, as part of its nation and state building efforts, the country needed human capital to build a homogenous nation state and welcomed migrants of Sunni-Islam origin from countries such as Bulgaria, Romania and Yugoslavia. In the period of 1940 and 1945, more than 20,000 people migrated to Turkey with Bulgarians, Romanians and Yugoslavians representing 73%, 19% and 8% of this migrant community, respectively (Korfalı et al. 2014). During the Cold War, a similar trend was observed. Approximately 800,000 migrants entered Turkey, those from Bulgaria forming the largest group followed by those coming from Yugoslavia (ibid.). Large-scale migration from Bulgaria during this time was due to Bulgaria’s negative policies against its minorities, including those of Turkish ethnicity. Overall, the history of CEE migration to Turkey has been dominated by migration of people of Muslim origin from Bulgaria, Yugoslavia and Romania, usually looking for a safe haven. This pattern thus somehow resembles the humanitarian migrations in other countries under research. Lately, with the neo-liberal economic policies of governments in Turkey after 1990, different types of migration have been observed.

4 Enlargements, CEE Migration, and Transitional Arrangements

The EU enlargements bringing in the mostly Eastern European countries of the A10 in 2004 and Bulgaria and Romania in 2007 (the EU-2), had significant effects on both the mobility of people within the EU and near abroad (i.e., Turkey). Figure 2.1 depicts the volume of CEE migration in the four countries under research before and after the 2004 and 2007 enlargements, showing a substantial increase in the number of CEE migrants in Austria, Sweden, and the Netherlands, but a decrease in Turkey.

Accordingly, as of 2013 there were a total of 330,000 officially registered persons in Austria who were born in the CEE enlargement countries. This figure was one third higher compared to 2002. The number of CEE migrants has also risen in the Netherlands. Just before the EU enlargement of 2004, in 2003, there were about 62,000 CEE residents in the country. This number was only 50,000 in the 1990s. As of 2013, the number of CEE migrants in the Netherlands increased to almost 180,000, three times more than the same figure in 2003. Sweden also faced an increasing number of CEE migration. From 2000 to 2012, the number of individuals born in one of the CEE EU member states increased by almost 80%, from 80,000 to 150,000.

In Turkey, there was a reverse pattern: the proportion of CEE migrants of the total number of regular migrants who received residence permits in Turkey declined from 40% in 2003 and to 22% in 2011. This can be explained by two factors. First, Turkey has begun to receive higher numbers of migrants from non-CEE countries in general, likely causing the ratio of migrants from the CEE countries to fall. Second, the number of CEE migrants has declined as they have shifted their migration destination towards the EU following the home country’s EU accession. The finding reiterates the position of Turkey as a reference case in the IMAGINATION research.

Following the two enlargements, the number of CEE migrants in Austria, Sweden and the Netherlands did not increase overnight. Some member states opted for transitional restrictions, which allowed member states to temporarily restrict the right of workers from the new member states to move freely to another member state to work. Such transitional arrangements had been applied in most of the previous enlargements of the EU. Among the four countries under research here, it was Austria and the Netherlands that have issued such transitional arrangements in relation to intra-European movement (See Table 2.1).

Table 2.1 Transitional arrangements (Source: Sert (2014), updated by the author)

Austria put in place transitional rules during both the 2004 and 2007 enlargement processes. Concerning the labour market access of CEE migrants, Austria declared that workers from the new member states needed a work permit sponsored by the employer. After one year of legal employment, workers were free to move within the labour market. After 18 months, those family members residing with the workers also received this right. The restrictions were lifted for the EU-10 in May 2011 and for Romania and Bulgaria by the end of 2013. In fact, transitional rules had not restricted freedom of settlement in Austria where the citizens of the new EU-member states could arrive as students, retirees or to join family members. They also had the right to found businesses, where they could be self-employed.

A year after lifting the transitional rules (2012), net migration from the CEE to Austria reached a zenith of 20,000. Because there is not municipal-level, register-based census that provides figures on employment of individuals during this time, the effects of the end of the transitional rules in May 2011 can only be made based on employment data and extensive in-depth analysis provided by the Federal Ministry of Labour, Social Affairs and Consumer Protection (Enengel et al. 2014). Hence, the impact of liberalization one year after the end of the transitional provisions amounted to an increase of almost 27,000 persons or in relative terms about a one-third increase compared to April 2011 (ibid.). Based on the expert interviews conducted in Austria, some of these workers were not new immigrants, but had already been in Austria and had only legalized their existing employment. Interestingly, almost half of these new workers were commuters who did not move to Austria (ibid.).

Similarly, the Netherlands also issued transitional restrictions, which required that workers from the new member states obtain a work permit for the first two years of their employment. After 2007, when the restrictions for Poles and residents from the other new member states of 2004 were revoked, the figure of migrants from these countries rose. Studies conducted in the Netherlands during this time illustrate the implications of the transitional restrictions on CEE immigration and give a sense of comparison between the cases of 2004 and 2007, as transitional restrictions were still in effect in the latter case (e.g., Weltevrede et al. 2009, Engbersen et al. 2011, Engbersen et al. 2013, Snel et al. 2014, also cited in Ostaijen et al. 2014). Focusing both on registered and non-registered CEE labour migrants, these studies show that because of the transitional regulations that were still effective towards the citizens of EU-2 enlargement countries of 2007, many Romanian and Bulgarian respondents were often self-employed rather than holding formal jobs (ibid.).

Contrasting the Austrian and Dutch cases, Sweden did not employ any transitional restrictions during either enlargement. In 2004, Sweden was the only EU member state that refrained from implementing transitional rules, where the UK and Ireland implemented minor ones to moderate the impact of the prospective immigration flows. Similarly, in 2007 Sweden again elected not to apply restrictions. Additionally, in 2008 Sweden liberalized its labour migration policy. With the change, instead of having state institutions evaluate the demand for immigrant labour via labour market tests, authorization to conduct the assessment was delegated to employers (Zelano et al. 2014). Having a work permit in Sweden now only depends on having a job offer with a wage that is in accordance with a collective agreement or at the same level as collective agreements in the industry concerned (Wadensjö and Gerdes 2013). Despite cautionary calls that insisted on transitional restrictions, studies by Doyle et al. (2006), Ruist (2013) and Wadensjö and Gerdes (2013) have proved that the fears of welfare abuse were exaggerated; immigrants from the expanded EU show no dramatic deviation in earnings, work conditions, educational levels or social welfare provisions (Zelano et al. 2014).

Overall, in Sweden – where no transitional restrictions were applied – the reservations on CEE immigration proved entirely unfounded. In Austria and the Netherlands, where transitional restrictions were employed, the restraints only affected formal employment. Immigrants from the CEE member states still arrived in both countries, working on a self-employed basis.

5 Main Migration Corridors

Migration corridors are tracks that form under specific conditions and enable movement of people within a geographical setting, in our case between the CEE countries and Austria, Sweden, the Netherlands, and Turkey. Inspecting the four country cases under research, the emergence of different migration corridors facilitating the movement of migrants can be explained by the classic push–pull factors theory, where migration is determined by the presence of attracting pull factors at destination, and repelling push factors at origin (Lee 1966). Demographic growth, low living standards, lack of economic opportunities, and political repression are typically cited push factors in the place of origin. Demand for labour, availability of land, good economic opportunities, and political freedoms that attract migrants to certain receiving countries are common examples of pull factors (Castles and Miller 1998; Sert 2010). Regarding CEE migration towards our four country cases, geographical proximity, historical ties, and political conditions such as endorsement of transitional periods are also among the facilitating conditions of movement. Table 2.2 identifies the main migration corridors between the CEE countries and Austria, Sweden, the Netherlands and Turkey based on the volume of immigrants from these countries.

Table 2.2 Main migration corridors (Ranking) (Source: Sert (2014), updated by the author)

In view of that, Austria’s main migration corridors are with Romania, Poland, Hungary, and the Czech Republic. An analysis of the Statistics Austria Population Register shows that immigrants from the new EU member states – the EU-10 of 2004 and EU-2 of 2007 – comprised almost 300,000 persons in 2013, with growth in 2002–13 being 26% for the EU-10, and 95% for the EU-2 (Enengel et al. 2014). The increase of migrants from Romania and Bulgaria was much more noticeable compared to immigrants coming from bordering countries: the Czech Republic, Hungary, Slovakia as well as Poland, who had already been arriving in Austria in larger numbers during the 1990s (ibid.). While geographical proximity and historical and transnational ties seem to play a more important role in creation of migration corridors with Hungary, Poland, and the Czech Republic, political conditions (i.e., accession to the EU) appear to be also relevant for movement originating in Romania.

Like in Austria, Poland, Hungary, and Romania constitute important migration corridors to the Netherlands. Bulgaria is an additional important corridor. In the Netherlands, the number of immigrants from Poland almost tripled in 2004–13, from approximately 60,000 to 110,000, while the number of Hungarians grew one-third to almost 20,000 persons (Ostaijen et al. 2014). In 2013, almost 21,000 Bulgarians arrived in the Netherlands, a five-fold increase since the 2007 enlargement (ibid.). Romanian migrants, comprising almost 18,000 people the same year, nearly doubled after the enlargement (ibid). These figures are derived from the Dutch Population Registers. As many labour migrants do not register in the Netherlands, it is assumed that the actual number of CEE migrants is probably much higher (ibid.). Recent studies, operating advanced statistical estimation techniques, projected that in 2010 about 340,000 CEE nationals (both registered and not) were present in the Netherlands, either temporarily or permanently (Van der Heijden et al. 2013, also cited in Ostaijen et al. 2014).

Like in Austria and the Netherlands, Poland, Romania and Hungary form the major migration corridors to Sweden, followed by Estonia and Lithuania. Based on figures provided by Statistics Sweden, there are almost 75,000 Polish, 22,000 Romanians, 16,000 Hungarians, 10,000 Estonians and 9000 Lithuanians in Sweden (Zelano et al. 2014). Like in Austria, geographical and political factors are important in the formation of these corridors.

Like in other countries, Romania is a main migration corridor also towards Turkey, along with Bulgaria, Ukraine and Moldova (Korfalı et al. 2014). Partially because of its location and partially because of its status, Turkey has rather different migration corridors than other countries under research here (i.e., countries like Ukraine and Moldova). While the link with Moldova can be explained by the historical and linguistic closeness of the Gagauz people living there (they speak Turkish), the corridor with Ukraine is mostly the result of economic factors.

6 Types of Migration (TOMs) and the Feminization of Migration

Based on earlier studies, comparative insight and inductive reasoning, we can categorize migration patterns within these corridors on two levels: duration (i.e., temporary versus permanent migration), and; socio-economic status (i.e., high- versus low-skilled migration). Consequently, a typology of CEE migration is fashioned depicting eight different categories or types of migrants (TOMs). These are: (1) knowledge workers; (2) entrepreneurs; (3) manual workers; (4) persons working in private households; (5) sex workers and trafficked persons; (6) students; (7) non-working spouses and children and; (8) beggars and homeless people. The typology is structured following an account of abductive reasoning (Yanow 2012), which is formed along formal–informal non-employment levels, taking into consideration different characteristics such as migrants’ skills, the sectors in which they are employed, their vulnerability as well as the duration of their migration (i.e. temporary versus permanent). The typology shows that CEE migrants should not be considered a homogenous category, but rather a highly differentiated one. As previous studies such as Engbersen et al. (2013) have proposed, the typology here also includes high- and low-skilled migrants as well as temporary versus permanent ones. Unlike previous studies, though, the effort of creating a typology of CEE migration here is realized not within a single country but in four different countries (and eight different urban regions). Regarding the typology, there are some major findings relevant for all four countries and urban regions (See Table 2.3). Table 2.3 provides a comparative synthesis of the typology with certain drawbacks on internal coherence. There is an asymmetry of information, for instance, for the same category of migrants, the main activity sectors for some countries and urban regions, and about chronological evolution.

Table 2.3 TOMs within countries and urban regions (Source: Sert (2014))

To begin, there is a small presence of knowledge workers in all countries, but there are more differences than similarities. In Austria, high-skill positions in the labour market are usually reserved for Austrian citizens, with a certain level of improvement in this situation (Enengel et al. 2014). With regard to the urban regions of Linz and Vienna, we have less region-specific information. In the Netherlands, despite the conventional belief that CEE migrants hold low-skill jobs, there is an observed heterogeneity in the labour market positions of the migrants from the CEE, where migrants also have high-skill employment (Ostaijen et al. 2014). While there is less region-specific information with regard to Rotterdam, we observe Romanians and Hungarians taking on some positions in international organizations in The Hague. Contrary to Austria and the Netherlands, the highest shares within the skill-level distributions in occupations of migrants in Sweden are detected in high- and medium-skill jobs (Zelano et al. 2014). While high-skill jobs are accessible for migrants in Sweden, we observe a decreasing pattern in Stockholm compared to Gothenburg. Although migrants taking high-skill positions are exceptional in Turkey, Istanbul is considered to attract highly skilled migrants. The conventional wisdom especially in Edirne holds that migrants from the CEE have higher skills than the local population.

Because of the legal conditions that determine the entry to the formal labour market, the category of entrepreneurs is observed as an important path for formal employment in all four countries. In Linz and Vienna, there are many one-person companies especially in the construction and care sectors. Similarly, in Rotterdam and The Hague, self-employment in the construction sector is common. Self-employment as a means of formal employment is also observed in Sweden, where Estonians have a higher share. Interestingly, in Sweden the number of self-employed female migrants is increasing both in Stockholm and Gothenburg, pointing to a feminization of CEE migration in this category. Self-employment as a means of formal status in the labour market is also detected in Edirne and Istanbul, mostly in the form of small enterprises. Thus, there is a clear relationship between the presence of legal restrictions on employment in the formal labour market in the host country and the use of self-employment as strategy of migrants to achieve formal status. To illustrate, as the transitional rules were lifted in Austria and the Netherlands, there was – as expected – a decrease in the trend of self-employment. Similarly, as more regulations were applied in the berry-picking market in Sweden, the number of self-employed migrants in this sector rose (Zelano et al. 2014).

Manual workers are commonly observed in the cases under research except for Turkey. In Austria, the tourism, construction and agricultural sectors employ large segments of CEE migrants, both in the formal and informal employment schemes. In Vienna, there are mostly Slovakians and Hungarians taking on jobs in the gastronomy sector and viticulture. In Linz, there are more agricultural workers in the rural areas. In the Netherlands, there are seasonal workers in horticulture and agriculture, mainly composed of Polish migrants. Regarding Rotterdam and The Hague, we have less region-specific information due to the lack of data. In Sweden, there are low-skilled workers in both the formal and informal labour market in construction, cleaning and welding. Both in Gothenburg and Stockholm, we observe that the number of young, male migrants has been increasing since 2004. While there are a few cases of young Moldovan males in the construction sector in Istanbul, generally in Turkey internal migration of the local Kurdish population – and lately the increasing number of asylum seekers from Syria – supply the demand for manual workers.

Persons working in private households as a category are identified in all four countries. In Austria, there is a trend of circular migration of females especially in the care sector for the elderly as well as informal employment of migrants as cleaners. The category is observed both in Vienna and Linz, where migrants from Poland, Slovakia (and increasingly Romania) work in private households. Similarly, in the Netherlands there are also migrants who usually work in informal arrangements in the domestic care and cleaning sectors. Both in Rotterdam and The Hague, the category is mostly composed of Bulgarians, which is expected to change with the end of transitional arrangements. With regard to the care sector in Sweden, while official numbers show a small immigrant overrepresentation in the health sector, information about unregistered migrants is rather patchy at best. In Stockholm 7% and in Gothenburg 9% of registered migrants work in the nursing and cleaning sectors. In Istanbul, irregular domestic labour – both in temporary and permanent arrangements – is common and is largely dominated by female Bulgarian and Moldovan migrants. In all four cases, a dominant issue in relation to migrants working in private households as care workers, nannies, cleaning persons, and gardeners is the problem of registration, where the distinction between regular versus irregular migrants becomes difficult to distinguish.

Another category where the distinction between formal and informal arrangements is hard to differentiate is sex workers and trafficked persons. In Austria, there are both registered and unregistered sex workers, and forced prostitution can be observed in both cases. While most registered sex workers in Vienna are Hungarian and Romanian, there is insufficient information about the urban region of Linz. In the Netherlands, prostitution is legalized under certain conditions and CEE migrants are known to be victims of trafficking, but there are no official statistics. In The Hague, 25% of registered sex workers are known to be from CEE countries. In Sweden, there are young women staying for short periods of time. In Stockholm, there are few women from Romania, Hungary, Bulgaria, Moldova, the Baltic countries and Poland working in the sector. In Gothenburg, the Roma from Bulgaria and Romania are more visible. In Turkey, human trafficking, especially from Moldova, is a known phenomenon and for both urban regions the category is relevant, but less is known. Overall, on the one hand some sex workers are not registered and working informally. On the other hand, there are many registered sex workers that may represent a voluntary choice of profession in a formal setting. Within this latter group there are also many that are forced into prostitution. Thus, sex workers and trafficked persons are usually mixed in this category.

Students as a category of CEE migration are rising in volume in all four countries. In Austria, the number of CEE students is increasing where Bulgarians constitute a large category. In the Netherlands and Turkey, there are also many CEE students. In Sweden, the number of CEE students is also increasing with gender differences. Student mobility is an expanding phenomenon not only among the member states of the EU, but also in Turkey (Findlay et al. 2012; Van Mol and Timmerman 2013). Students may utilize education as a means of access to permanent residency within a country (Gribble 2008), but further research is needed to understand the motivations and future plans of CEE students in the four countries.

Less is discerned about the qualitative features of the migration of non-working spouses and children, but quantitatively this category is visible in numbers in all four countries, while receiving less attention both in policy and academic circles. In Austria, children constitute a small portion of CEE migration. In Vienna, there are 7931 children registered, where cross-border school attendance is an emerging issue. In Linz, the category has minor relevance, mostly dominated by Romanians. In the Netherlands, family-related migration is increasing and mostly led by females, i.e., women migrating first and husbands joining later, which is very different than the historical migration of single males. In Rotterdam, a large share of CEE migration is related to families. In The Hague, the number of CEE children going to school has doubled since 2010. In Sweden, the second largest reason for registering for a residence permit is intention of family members to accompany their spouses and parents. Similarly, in Turkey the highest proportion of residence permits granted upon education is at the primary level (i.e., mostly children). For the latter two cases, region-specific information is not available regarding this category.

In contrast to the category of non-working spouses and children – which receives little policy and academic attention – the category of beggars and homeless people emerges as a sensitive issue that attracts public debate, although the numbers of this type of migration are rather low compared to the other categories. In Austria, begging is a sensitive public issue, but there are no reliable figures. In Vienna, beggars are often commuting from the neighbouring countries, especially from Slovakia and Hungary. In Linz, a growing number of beggars of Roma descent are observed. In the Netherlands, the number of beggars and homeless people are low and decreasing. In Rotterdam, there are non-registered Bulgarians selling journals, or street musicians. In The Hague, there is a public debate about homeless CEE nationals, mainly Polish. In Sweden, there are 379 homeless EU citizens; the majority are males with an average age of 38 years old from Romania as well as Poland, Bulgaria, Slovakia and Spain (Zelano et al. 2014). In Stockholm, NGO Crossroads receives on average 120 visits every day. In Gothenburg, the numbers are increasing with the majority being female migrants. In Turkey, beggars and homeless from CEE is a rare phenomenon.

Overall, within many categories there is an observed feminization of migration, which is described here as an increasing number of female migrants in CEE migration. In Austria, the gender composition of CEE migration has changed a lot since the end of the Cold War. Female migrants from CEE (being 15 years and older) have an increasing presence and importance, clearly outnumbering males. In 1991, the proportion of male migrants among all CEE-migrants ranged at around 70% with only little variation between the sending countries and with around one third not having reached the age of 30 years Today, young and middle-aged female migrants have gained importance and visibly exceed the number of males. Women from CEE are more visible in the service sector, in child care as well as caring for old-aged people in private households. These occupations have strong demand in Austria, thus providing a considerable number of jobs, but generally pay modest incomes and are often organized on an informal basis. The urban regions of Vienna and Linz do not show a variance from this trend on the national level. While the figures of residents that were born in the EU-8 do not show big differences in gender patterns in the whole of Vienna urban region, the study by Lechner et al. (2010) referring primarily to the (daily) commuters in the Austro-Hungarian border region (southeastern parts of the region of Vienna) show clear gender- and age- specific patterns. Accordingly, about 75% of the Hungarian workers are male. Many of these migrants choose to commute to Austria daily, as they can stay and live in western Hungary (with lower living and residential costs) and work in the border regions of Austria for higher wages than they would earn at home. Thus, geographical proximity is a determining factor of this type of migration, where males show a stronger presence than females.

In the Netherlands, there is also a tendency towards feminization of migration, where the majority (albeit a small portion in some cases) of the officially registered CEE migrants are females (51% of Polish and Bulgarian, 54% of Hungarian, and 60% of Romanians). While the proportion of females among the labour migrants is low (Nicolaas 2011, Jennissen 2011), females are overrepresented in migration for family motives (Gijsberts and Lubbers 2013). Looking at the ratios at the level of urban regions, in Rotterdam the total percentage of females is more than the males in all three municipalities. The difference is profound among Romanian migrants, where females are overrepresented in all municipalities (63% in Rotterdam; 62% in Schiedam and 72% in Lansingerland). Similarly, the male-female ratio in The Hague region is also in favour of the female migrants, where the total percentage of females is more than the males in all three municipalities. In the municipalities of The Hague and Westland, women are overrepresented among both Romanians (59–60% female) and Hungarians (63–4% female). In the municipality of Delft, a similar overrepresentation is true for the Polish (62% female).

In Sweden, gender distribution shows a variance between registered and unregistered migrants. While the flows of regular and registered CEE migrants are characterized by a rather even gender distribution, the group of CEE immigrants overlooked by the official records is dominantly male. In 2012, of the 148,998 individuals living in Sweden born in one of the EU-10 countries the gender distribution was 45% male and 55% female. For the CEE as a whole, the gender distribution has been more or less the same throughout the period 2000–12, with a somewhat larger share of women than men (an average of 42% men, 58% women). The same is true for the five largest immigrant populations, with the exception for the Polish and Lithuanian ones, where the share of males have increased slightly during the 10 years period between 2000–2012 (Statistics Sweden 2013b). Of the 2732 citizens from the CEE region who obtained Swedish citizenship in 2012, 1627 were women and 1105 were men (Statistics Sweden 2013a). Polish women consist the largest category representing 36% (975) of all CEE naturalizations in Sweden in 2012. In the urban region of Stockholm, women had a larger share than men in both the municipality of Haninge and Stockholm (51 and 56% respectively). In Södertälje, the ratio was 48% women, 52% men. In the region of Gothenburg, women were the larger category in all three municipalities ranging from 53 to 60%.

In Turkey, the number of female CEE migrants with residence permits in 2011 was approximately 59% of total regular CEE migrants. In previous years, the share of residence permits given to women was 55% in 2008, 54% in 2009, and 55% in 2010. These high percentages of female migrants among the CEE migrants were also higher than the averages of female migrants among total migrants in Turkey. For rge same years, female migrants constituted 50% (2008), 51% (2009), 52% (2010) and 54% (2011) of total migrants in Turkey, which are below the proportion of women among CEE migrants. In a more detailed analysis, it is demonstrated that there was a gender balance of CEE migration between 2003 and 2005, male predominance in 2006 and 2007, and female predominance after 2008. A similar trend of increasing female domination is also observed, when we look at the work permits. An analysis of gender distribution of CEE migrants with work permits in the year 2008 shows that female migrants constitute only 37% of the CEE community holding employment permits. However, the difference between women and men with employment permits started closing in 2009 when women constituted almost half of the CEE migrants with work permits. Moreover, in an increasing trend the proportion of women among CEE migrants with employment permits in 2010, 2011 and 2012 rose to 52%, 65% and 67%, respectively.

However, like in Sweden the gender distribution may show a variance between registered and unregistered migrants. While the flows of regular and registered CEE migrants are characterized by recent female predominance, looking at the apprehension data (a basic variable for estimation of unofficial migration) especially in Edirne, we see that it is predominantly male. However, there is also a variance among unregistered migration between irregular labour migrants and irregular transit migrants. While female CEE migrants may largely dominate the former (like in Austria, there is a strong demand for domestic female workers), the latter is largely a male-dominated phenomenon. In Istanbul, the variance is also visible between different types of migration.

7 Conclusions

Regardless of the variances in accessibility, type and disposition of data in the four countries, the exercise of mapping of migration corridors and analysis of types of migration from CEE countries to Austria, the Netherlands, Sweden, and Turkey see interesting conclusions develop. International comparisons are challenging, where disparities in data-collection practices between countries, the changing nature of political unions (in our case for example, being a member of the EU or applying transitional rules), and informal paths of migration all contribute to the problem. The lack of data is observed on the international level, as well as supranational, regional and local levels. On the supranational level, even though CEE migration is mostly about intra-European movement, unavailability of comparable data among the EU member states on the movement of EU citizens deserves policy attention. On a regional level, inaccessibility of comparable data between urban regions in a given country also demands further elaboration. There is a need to build mechanisms that enhance data collection at these different levels.

Within this context, creating a typology of CEE migration as a heuristic device for comparison produces a tool with great exploratory value for answering further research questions. To illustrate, the almost trivial presence of knowledge workers in high-skill jobs in all four countries is instructive both for understanding de-qualification of migrants from the CEE countries in the labour market and recognizing the glass ceiling created by established structures. There is room for further research about this group, such as collecting information about their patterns of integration or modes of professional advancement. In all four countries, self-employment seems to be a useful instrument for migrants from different CEE countries to overcome the limitations created by the recognized structures, such as transitional rules in the case of Austria and the Netherlands and strict labour market regulations in the case of Sweden and Turkey.

Whether there are signs of emerging ethnic economies in these countries also remains to be explored. In Austria, the Netherlands, and Sweden, manual workers are taking seasonal jobs mostly in the agricultural sector, implying that we can predict new routines of circular migration, where the difference between registered/unregistered migrants is blurred. In all four cases, persons working in private households are parts of unregistered or irregular migration patterns, leaving room for further research to understand both the scale and nature of this type of migration. A similar reflection can be made about sex workers and trafficked persons, non-working spouses and children, and beggars and homeless people. All three groups include at-risk persons, and more research needs to be done to recognize the level and character of their vulnerabilities. The number of foreign students is increasing in all four countries, also opening a rather new venue for research.

The typology created here shows that the labour market remains an important governing factor in this specific types of CEE migration, but that new and changing patterns of free movement within the migration corridors are also in evidence, leaving room for more analytical insights and interpretation about the drivers, impacts, and modes of integration of migrants. The typology confirms that ‘CEE migrants’ should not be considered as a homogenous, but rather as a highly differentiated category with high and low skilled, as well as temporary and permanent migrants. There is also a need to address the feminization of migration, which is described here as increasing number of female migrants in CEE migration, an observed phenomenon in all four countries, maybe more in some than others. Very basically, we still do not know much about the gender relations among CEE migrants or the changing profile of female CEE migrants.

The country cases selected for this research also help us to draw conclusions about migration corridors. The cases illustrate the importance of historical context. Certain migration corridors have always been active. For example, the migration corridors between Poland and the Netherlands, or Bulgaria and Turkey have seen flows that have not eroded but rather transformed over time. In many ways, institutional regimes, rather than geographical proximity, have played a larger role in these transformations. The country cases with their different applications of transitional restrictions create specifically interesting comparisons. In Sweden, where no transitional restrictions were applied, concerns about the costs of admitting CEE migrants proved groundless. In Austria and the Netherlands, where transitional restrictions were in place, the limits only had consequences for formal employment. Deploying different tactics, such as self-employment, migrants from the CEE member states still arrived and worked in both countries. Taking everything into account, today we see a diversification of intra-European movement, where historical context, institutional structures and individual schemes have all contributed to the alteration of existing migration corridors.