To Stay or Return? Explaining Return Intentions of Central and Eastern European Labour Migrants

Introduction

The EU enlargements of 2004 and 2007 resulted in significant new migration flows in Europe.¹ Fassmann et al. (2014) calculated that by 2011, almost 5 million citizens from Central and Eastern European (CEE) countries resided in the ‘old EU’ (cf. Okólski, Salt 2015). This study focuses on CEE labour migrants in the Netherlands. According to official statistics, the Netherlands hosted almost 180 000 CEE migrants in 2012 (Statistics Netherlands 2012). However, as these figures are based on Dutch population registers (GBA) and many CEE labour migrants do not register, the actual number of CEE migrants in the Netherlands is probably much larger. Recent research, using advanced statistical estimation techniques, estimated that in 2010 about 340 000 CEE nationals were present in the Netherlands, either temporarily or permanently, and either registered or not (van der Heijden, Cruyff, van Gils 2013).

A crucial question, in addition to the size, is the nature of the new East–West migration in Europe, particularly regarding the politically sensitive issue of return migration. Are CEE labour migrants only temporary residents or are they here to stay, like many ‘guest workers’ in the 1960s and 1970s? Looking back at the history of guest worker migration to Europe, several authors argue that the permanent settlement of guest workers and their families only followed after the recruitment of guest workers ended and the borders were de facto closed for them (Sassen 1997; Engbersen 2012). If this is correct, we may expect current CEE labour migrants to settle to a lesser extent than the former guest workers did. After all, the borders in the EU are and will remain open. Polish research shows that the massive migration of the post-accession period was indeed accompanied by significant reverse flows of return migrants (Kaczmarczyk 2013).

This paper further explores this issue by using survey data from 654 Polish, Romanian and Bulgarian labour migrants in the Netherlands, whose return intentions are examined by testing hypotheses based on four migration theories: the neoclassical economy; New Economics of Labour Migration; cultural assimilation theory, and transnationalism theory. The research question is: to what extent can the return intentions of CEE labour migrants in the Netherlands be explained by their economic success or failure, their transnational activities and/or their socio-cultural integration? The remainder of this article first describes some peculiarities of CEE migration to Western Europe, in particular the question of return migration. The next section elucidates the four theories, discusses recent findings on the topic and presents three hypotheses. The survey and the methodology used in the analyses are described before the main research findings are presented, and these are followed by the conclusion and discussion.

CEE labour migration and the question of return

Various authors argue that CEE labour migration to Western Europe is different from previous migration flows because many current CEE labour migrants do not intend to settle permanently. As Okólski (2001, 2012) describes, Poland has a tradition of temporary labour migration as workers from rural areas in Poland worked in urban industries, or later abroad, while still residing in their villages. This resulted in ‘incomplete migration’: a sort of circulation of individual household members, often repeated, and characterised by short-term employment abroad and a very high proportion of earnings remitted or repatriated to the migrant’s home country, where the costs of living was substantially lower (Okólski 2012: 35; cf. Okólski 2001). This migration is ‘incomplete’ in the sense that it does not result in settlement. Moreover, individual trips of workers are usually of very short duration, although migrants may be abroad for a large part of the year as a result of repeated journeys (Okólski 2001: 107). Wallace (2002) also points out the tradition of circular and temporary labour migration in Central Europe: rather than permanent one-way migration (the dominant pattern until recently) there has been a predominance of short-term, circulatory movements backwards and forwards across borders. This would be better termed mobility than migration (Wallace 2002: 604). Note that both authors refer to patterns of temporary and circular migration that existed prior to the EU enlargement.

Favell (2008: 703), writing about post-accession migration, also emphasises the temporary nature of these migratory movements. Rather than immigrants, CEE labour migrants are ‘regional free movers’: more likely to engage in temporary circular and transnational mobility, governed by the ebb and flow of economic demand, than in long-term, permanent migration. Other authors label the specific transient character of CEE labour migration with concepts like ‘liquid migration’ (Engbersen, Snel, de Boom 2010; Engbersen, Snel 2013), ‘intentional unpredictability’ (Eade, Drinkwater, Garapich 2007; Drinkwater, Garapich 2015) or ‘deliberate indeterminacy’ (Moriarty, Wickham, Salomońska, Krings, Bobek 2010). All authors agree that the specific context of East–West migration in Europe enables this new, transient character of migration: the history of circular labour migration in some CEE countries; the relatively short distances combined with the differences in wages and prosperity between sending and receiving countries; but above all, open borders in Europe that enable frequent movement. Moreover, many CEE labour migrants do not have fixed ideas about the duration of migration and tend to keep their options open. For instance, as Isański, Mleczko and Seredyńska-Abou Eid (2014: 5) found for Polish labour migrants: many of the participants travelled to more than one country, their stays either abroad or in Poland were usually temporary, and they frequently returned to Poland for short periods of time. This new pattern of multiple returns replaces the former trend of settlement in the destination country. Finally, also Fihel and Grabowska-Lusinska (2014: 30) also found ‘repeating migration patterns’ typical for Polish labour migrants: Polish nationals still tend to engage in temporary, back-and-forth mobility for employment, also because of the significant costs of moving their family to the destination country (cf. Anacka, Matejko, Nestorowicz 2013; Fihel, Górny 2013; Kaczmarczyk 2013; Krings, Bobek, Moriarty, Salamońska, Wickham 2013).

Others nuance this idea of transient East–West migration in Europe and point to indications of a long-term or even permanent settlement of CEE nationals in Western Europe. For instance, as Holland, Fic, Rincon-Aznar, Stokes and Paluchowski (2011) showed, the current economic recession in the main receiving countries combined with moderate economic growth in sending countries like Poland has not yet resulted in the expected massive return migration. Previously, Eade et al. (2007: 33–34) showed, in a study on Polish migrants in the UK, that Polish migrants have various migration strategies including single, short-term migration to earn money to be spent upon return (‘hamsters’), circular migration alternating between work abroad and at home (‘storks’), open-ended plans for the future (‘searchers’), but also settlement (‘stayers’). Engbersen, Leerkes, Grabowska-Lusińska, Snel and Burgers (2013), in a later article, also point out the variety of migration patterns of CEE labour migrants: in addition to temporary, circular and transnational migration, they also found long-term settlement. In a study about Polish migrant workers in Norway, Friberg (2012: 1601–1602), shows that many Polish labour migrants settle in Norway. Friberg describes three ‘stages’ in Polish migration to Norway. Initially, many migrants go abroad to earn money for a better life for themselves and their families in Poland, without planning permanent relocation. However, not all migrants return. Some enter the ‘transnational commuter stage’: going to Norway regularly to earn money, but with their families and friends still in Poland. The third stage is that of settlement, when the primary household moves from Poland to Norway. Although Friberg observes all three ‘stages,’ this contradicts previous assumptions of CEE labour migration as typically a temporary or ‘liquid’ phenomenon. As White (2011, 2014) found in her study of Polish families in the UK after EU accession, particularly Polish families do not return. As Kaczmarczyk (2013: 112) observes, for many Polish migrants ‘return’ can often mean only a short break between periods spent abroad. This, again, highlights the liquid or fluid character of CEE labour migration (Anacka et al. 2013; Fihel, Górny 2013; White 2014).

Migration theory and return migration

Although migration research has traditionally focused on immigration to Western countries, there is a growing literature about return migration (Constant, Massey 2002; Cassarino 2004; de Haas, Fokkema 2011; Carling, Pettersen 2014; van Meeteren, Engbersen, Snel, Faber 2014; de Haas, Fokkema, Fihri, 2015). Some of these studies focus on actual return migration, others examine return intentions of previous migrants. Return migration may be defined as the movement of emigrants back to their homelands to resettle (thus not for vacation or extended visits without the intention of remaining at home) (Gmelch 1980: 136). Gmelch’s definition does not mention the possibility of temporary return as a break between two periods of working abroad. However, as we shall see, only a small minority of the respondents in the present research intend to move to another EU country after leaving the Netherlands. The research seeks to exclude the possibility of short-term temporary return or back-and-forth movements to the home country by asking respondents how long they intend to stay in the Netherlands. The focus of this article is the return intentions of CEE labour migrants. Obviously, return intentions are often at variance with actual return behaviour – as the ‘myth of return’ of many former guest workers clearly shows. However, various authors claim that return intentions are significant in their own right. Return intentions show the return motivations of migrants, even if social, economic and political constraints prevent actual return. Moreover, return intentions sum up the respondent’s attitude towards the migration experience, and can affect behaviour other than actual return – for instance, investments in social contacts, skills and assets (de Haas, Fokkema 2011; Carling, Pettersen 2014).

Constant and Massey (2002) and later de Haas and Fokkema (2011) showed that mainstream economic migration theories offer radically opposed interpretations of return migration. Here, we will supplement two economic migration theories – neoclassical migration theory and New Economics of Labour Migration – with two quite common sociological perspectives: socio-cultural integration theory and transnationalism-theory. As we shall see, each perspective offers specific assumptions about the determinants of return intentions of migrants, which will be tested in this study.

Economic migration theories

Dominant economic migration theories contain divergent assumptions about return migration and the determinants of return decisions (Constant, Massey 2002; Cassarino 2004; de Haas, Fokkema 2011; Hołda, Saczuk, Strzelecki, Wyszyński 2011). Neoclassical economics (NE) explains migration as the result of individual cost–benefit trade-offs. Migrants decide to migrate or to return in order to maximise expected net lifetime earnings (Borjas 1989). If the expected lifetime benefits of migration (the wage gap between the origin and destination country) are higher than the costs (transport, first settlement in the destination country), migration is a rational choice. From this perspective, it is expected that economically successful migrants will settle permanently in the destination country and eventually have their spouses and children come over from the origin country. Return migration, on the other hand, is a sign of ‘failed migration.’ Return migrants are the ‘losers’: migration did not bring them the expected lifetime earnings or the costs were too high (Constant, Massey 2002). As de Haas and Fokkema (2011: 757) put it boldly, while ‘winners’ settle, ‘losers’ return. Regarding return intentions, we can hypothesise that the greater the migrants’ economic success, the less they intend to return (NE hypothesis).           

The New Economics of Labour Migration theory (NELM) contains opposite ideas about return migration. NELM interprets migration as a strategy of households and extended families to spread income risks and to overcome market constraints in the sending country. Migration provides an additional income gained elsewhere (Stark 1991). Migrants generally leave for limited periods of time to meet certain income targets. From this perspective, migrants are seen as ‘target earners’ and are expected to stay in the destination country as long as they are unable to meet the income target, for instance because of unemployment or high costs of living. Migrants who have saved and remitted enough financial or human capital to realise their investment plans are expected to return. NELM views return migration not as ‘failure’ (as NE does), but rather as an indicator of economic success (de Haas, Fokkema 2011: 759). The NELM hypothesis is thus: the greater the migrants’ economic success, the sooner they intend to return.

Previous research on return migration and return intentions has produced mixed outcomes about both economic hypotheses. The findings of Constant and Massey (2002: 27) are more consistent with NE rather than NELM: migrants with stable, full-time employment are more likely to stay; unemployment or marginal employment strongly predicts return migration. However, holding this factor constant, a higher socio-economic position in terms of earnings or occupational prestige does not contribute to stronger return intentions. This outcome is not in line with NE. Also, their finding that sending remittances is strongly associated with return migration is – in their opinion – related to NELM. Constant and Massey therefore argue that the two economic arguments are neither mutually exclusive nor contradictory. Some migrants are ‘target earners’ who return after having reached their earnings target, others are ‘income maximizers’ who will stay when economically successful (idem). De Haas and Fokkema (2011: 771) also find mixed outcomes with regard to the economic determinants of return intentions and conclude that a ‘one-size-fits-all’ theory is unlikely. Return migration and return intentions are likely to depend on initial migration motives, opportunities in both sending and receiving countries, and specific features of immigrant groups. In a recent study of Moroccan migrants in several European countries, de Haas et al. (2015) find that return intentions are more closely related to socio-cultural integration (or assimilation) than to structural or economic integration. Carling and Pettersen (2014) report similar findings: after incorporating factors like cultural assimilation and transnational ties in the analyses, the socio-economic indicators of employment status and financial stress hardly contribute to the explanation of return intentions.

Despite these previous research outcomes, the present research anticipates that in the specific case of CEE labour migrants in Western Europe, economic success in the host country increases the propensity to stay and thus decreases their return intentions. The reason is that the relatively short distances and open borders in Europe make it easy for successful CEE migrants to stay in the host country and simultaneously have frequent and intense contacts with the country of origin. The following hypothesis will therefore be examined:

The greater the economic success achieved by CEE labour migrants from Poland, Romania and Bulgaria in the Netherlands, the less they intend to return soon (H1).

Socio-cultural integration theory

While economic theories explain return migration and return intentions primarily with reference to economic factors, the third theoretical perspective focuses on the socio-cultural dimension of immigrant integration. Socio-cultural integration theory (SIT) postulates a negative relation between the socio-cultural integration of migrants and return intentions. ‘Socio-cultural integration’ can be understood in terms of identification with the host country, social contacts with native citizens, participation in social institutions of the host country and speaking its language (Snel, Engbersen, Leerkes 2006; de Haas, Fokkema 2011). SIT also assumes a negative association between the length of stay in the destination country and return intentions. The longer migrants stay, the more they become integrated in the receiving country, the more difficult return becomes in practice (in terms of both financial and emotional costs), and the more they are inclined to settle permanently. Therefore, the SIT hypothesis proposes: The stronger the socio-cultural integration of migrants, the less they intend to return.

Several studies confirm the assumption on socio-cultural integration and return migration. Constant and Massey (2002: 20) find negative relations between various indicators of attachment to Germany (like having a spouse in Germany or ‘feeling German’) and the likelihood of return migration. De Haas and Fokkema (2011: 771, 773) also find a negative relation between the level of socio-cultural integration (measured in terms of friendship with natives, participation in native organisations, language fluency, and having ‘modern values’) and return intentions. In their study about Moroccans in various European countries, de Haas et al. (2015) found a clear negative association between socio-cultural integration (measured by indicators such as watching Moroccan television, attitude towards mixed marriage, contacts with non-migrants and feelings of belonging) and return intentions. However, as they argue, the causality is not clear. Migrants who intend to return soon may feel less motivated to integrate into the host society. Most studies find the expected negative relation between duration of stay and return migration (Borjas 1989; Dustmann 1993; Waldorf 1995; Constant, Massey 2002: 19; Jensen, Pedersen 2007). Carling and Pettersen (2014) combine socio-cultural integration and the transnational ties of migrants. They find that migrants with strong socio-cultural integration and weak transnational ties are less inclined to return. Return intentions are most prevalent among migrants with strong transnational ties and weak socio-cultural integration. However, the weakness of this analysis is that it says nothing about the independent effects of socio-cultural integration and transnational ties on return intentions.

The socio-cultural integration of CEE labour migrants can be expected to increase their inclination to stay. Migrants who have frequent contacts with Dutch natives in the host country, who speak the Dutch language and who are interested in Dutch news and Dutch developments, will be less inclined to return. The present research will therefore test the following hypothesis:

The greater the socio-cultural integration of CEE labour migrants from Poland, Romania and Bulgaria in the Netherlands, the less they intend to return soon (H2).

Transnationalism theory

The fourth theoretical perspective is transnationalism theory (TT). Transnationalism refers to the process by which immigrants forge and sustain multi-stranded social relations that link together their societies of origin and settlement (Basch, Glick Schiller, Szanton Blanc 1994: 7). Transnational ties and activities as such are not new, as the classic study The Polish Peasants in Europe and America by Thomas and Znaniecki (1918–1920) shows. However, modern transport and communication technologies enable cheaper and far more frequent communication and travel between sending and receiving countries than previously. As a result, contemporary migrants can be integrated in the receiving country and also have strong attachments to the sending country. The point here is that, according to TT, transnational ties and participating in transnational networks also affect migration intentions and migration decisions (Samers 2010: 95–97). For instance, involvement in migrant networks makes migration cheaper and thus more likely (Massey 1990). With regard to return migration, TT postulates a positive relation between the transnational activities of migrants (sending remittances, frequent home visits, intensive communication with the ‘folks back home;’ cf. Snel et al. 2006) and return migration, independently of economic success or failure in the destination country. More generally: According to transnationalists, returnees prepare their reintegration at home through periodic and regular visits to their home countries. They retain strong links with their home countries and periodically send remittances to their households (Cassarino 2004). The central TT hypothesis regarding return intention proposes: the more migrants are engaged in transnational activities, the more they intend to return. The TT hypothesis thus mirrors the SIT hypothesis. Whereas a strong attachment with the host country results in a weaker inclination of migrants to return, strong attachments with the origin country contribute to stronger return intentions.

Various studies observe a positive relation between migrants’ transnational activities and return migration. Duval (2004: 52) finds a positive link between home visits and return migration which he sees as the completion of the migration cycle. Others nuance this view by saying that return migration is not necessarily the end of a ‘migration cycle’ (King 2000; Ley, Kobayashi 2005; Sinatti 2011; Oeppen 2013). Besides permanent return, there are various forms of temporary return. All the same, all these authors report strong positive relations between transnational ties or involvement and temporary or permanent returns. Constant and Massey (2002: 20) also found that having a spouse and children in the origin country and sending remittances to friends and family back home contribute to the likelihood of return. Itzigsohn and Saucedo (2002: 738) and later Bilgili (2014) find positive relations between various forms of economic and socio-cultural transnational activities and return intentions. However, using similar indicators, de Haas and Fokkema (2011: 772) come to different conclusions. In their study, transnational ties or transnational economic activities like sending money or goods to the home country are not related to return intentions. They only find a strong positive relation between investments in the origin country and return intentions. Again, one can ask what the causality is. Do migrants who invest in their origin country have stronger return intentions or is it the other way around? In their recent study, de Haas et al. (2015) again find positive relations between investments and, to a lesser extent, social ties to people in Morocco and return migration intentions.

Also in the case of CEE labour migrants, a positive relationship between their various transnational activities and their return intentions is to be expected, although this relationship may have different backgrounds for CEE migrants in Western Europe than for other (non-EU) migrant groups. For CEE migrants, transnational activities are not just a means to maintain contact with or to support the wider family, as applies to many other migrant categories. Some CEE labour migrants have their spouse and children still at home. For them, home visits and financial transfers to the home country are part and parcel of their own family life. In general, we assume that CEE labour migrants generally do not send remittances or invest in the home country only to support extended family members or a local community but also for their own future plans in the country of origin – after having returned. This is in line with the pattern of ‘incomplete migration’ in the EU pre-accession period (Okólski 2001). Moreover, White (2014) found that Polish families who settle in the UK scale down transnational practices, such as return visits to Poland and keeping well-informed about Polish current events. The following hypothesis will therefore be tested:

The greater the transnational activity of CEE labour migrants from Poland, Romania and Bulgaria in the Netherlands, the more they intend to return soon (H3).

Survey and measurements

The data used in this article are derived from a survey among labour migrants from Poland, Bulgaria and Romania in the Netherlands (N = 654). The fieldwork explicitly focused on recently arrived labour migrants and not on CEE nationals in general. Labour migrants are migrants who came to the Netherlands with the intention of working. The survey examined the labour market position and incorporation of CEE labour migrants (Engbersen et al. 2013; Snel, Faber, Engbersen 2014). The interviews were conducted between October 2009 and February 2011 in various Dutch municipalities, including Rotterdam and The Hague, two medium-sized cities (Breda, Dordrecht) and some rural towns (in agricultural areas). The field work took a year and a half because CEE labour migrants were interviewed in ten municipalities one by one. The face-to-face interviews were held in the mother tongue of the respondents. Native-speaker interviewers recruited respondents at places frequented by CEE labour migrants, such as Polish shops or churches, or on internet fora. Respondents were also approached in the street by the interviewers upon hearing their mother tongue or by identifying their car’s nationality plates. The respondents were deliberately selected from diverse sources; for example, care was taken not to overuse certain locations. Respondents were also recruited by snowball sampling. At the end of each interview, respondents were asked whether they knew of fellow nationals who could be interviewed. A structured questionnaire was used containing 213 questions about issues such as the migration history of respondents, their labour market and housing conditions, social integration in Dutch society, their transnational activities, and future plans.

Return intentions

A crucial survey question for this analysis concerns the expected duration of stay in the Netherlands. The answer categories for this question ranged from only three months to ‘more than 10 years’ and ‘forever.’ Here, responses are grouped into four categories: (0) ‘up to 2 years;’ (1) ‘2 to 10 years;’ (2) ‘more than 10 years/permanently;’ and (3) ‘don’t know.’ The latter category will be treated not just as ‘missing’ but as a meaningful category. Migrants who don’t know how long they will stay embody the ‘intentional unpredictability’ which is assumed to be typical for post-accession CEE migrants (Eade et al. 2007).

Economic success

Two indicators were used to measure migrants’ economic success: (1) labour market participation; and (2) occupational status. First, respondents were asked about their current labour market situation. Possible answers were: employed on a permanent contract; employed on a temporary contract (including employment through an employment agency or another intermediary); self-employed (‘having own business’); working on an informal (‘verbal’) contract; work in one’s own household; searching for a job; being a student or a pupil; or something else. Respondents in the latter category could specify what they were doing at the moment. A new variable was constructed containing five categories: (0) employed on a permanent contract; (1) employed on a temporary contract; (2) (registered) self-employed;² (3) working on a verbal/informal contract; and (4) non-working.

Occupational status

Second, respondents were asked about their occupation in an open question. The answers were categorised according to the Erikson Goldthorpe and Portocarero (EGP) occupational class scheme (Erikson, Goldthorpe, Portocarero 1979; 1983). The EGP scheme consists of eight occupational categories: (I) higher-grade professionals; (II) lower-grade professionals; (III) routine clerical non-manual workers; (IV) small employers and independent workers; (V) manual supervisors and foremen; (VI) skilled manual workers; (VIIa) semi-skilled and unskilled manual workers (not in agriculture); and (VIIb) (unskilled) agricultural workers. For this study a final category was added to the EGP scheme: ‘survival strategies,’ which refers to informal income-generating activities such as being a street musician or selling newspapers on the street (probably street newspapers). The official EGP scheme does not take these informal work strategies into account. For interpretation reasons, the variable occupational status was reduced to four categories: (0) high (occupational categories I to III); (1) medium (IV to VI); (2) low (VIIa, VIIb and ‘survival strategy’); and (3) non-working. Respondents who had no job at the time of the interview were categorised as persons with no occupational status. This number of inactive respondents is the same as the number of inactive respondents for the variable ‘labour market situation.’

Socio-cultural integration

The survey contained questions about three aspects of socio-cultural integration: (1) frequency of contact with Dutch natives (at work, in the neighbourhood and during free time); (2) speaking the Dutch language at work (with the boss/manager or with colleagues) and with friends/acquaintances during free time; and (3) following Dutch news and developments in the Netherlands through various channels (colleagues, friends, TV, internet, radio, newspapers). The response categories for the first two sets of questions ranged from (0) ‘never’ to (4) ‘very often;’ for the last set of questions the range was (0) ‘daily’ to (4) ‘never.’ An extra category ‘not speaking the Dutch language’ was added to the questions about Dutch language proficiency, for people with no command of Dutch at all. The three aspects were combined in a scale to measure the socio-cultural integration of respondents in the Netherlands. The answers to the questions about following Dutch news and Dutch developments are reversed in such a way that they point in the same direction as the other two sets of questions. A principal component analysis (PCA) was conducted on the 12 items of the three sets of questions; the results are presented in Table A1 (in the Annex 1).³ Only one item (contact frequency with native Dutch at work) has a small factor loading (0.340). After deleting this item, the remaining items load high enough on the first component, representing socio-cultural integration. A scale taking the mean of the 11 items was constructed, including only respondents with valid answers on at least nine items (Cronbach’s alpha: 0.812). The lowest score on the scale is 0; the highest score on the scale is 4.27. The higher the score on the scale, the stronger the migrant’s socio-cultural integration in Dutch society.

Transnational activities

The survey contained several questions about the transnational ties and activities of CEE migrants in the Netherlands. The focus was on transnational activities defined as cross-border activities of an economic, political or socio-cultural nature (Snel et al. 2006). Three indicators of transnational activities often found in transnationalism studies were used:

• frequency of contact with friends and family in the origin country: in two questions respondents were asked how often they are in contact with (1) family and (2) friends and acquaintances back home;
• frequency of home country visits: number of times per year a migrant visits the home country;
• sending remittances: respondents were asked what proportion of their monthly income they spend to support their family, friends or others. If this question was not answered, the proportion was calculated by dividing the amount of money they remit by their income.

An index based on the four above-mentioned questions was constructed by assigning scores (0, 1 or 2) to the different variable categories (note that the separate effects of the three types of transnational activities were not analysed). Table A2 (in the Annex 1) shows the distribution of these scores, which were summed creating an index with a minimum of 0 (no transnational activity) and a maximum of 8 (high transnational activity). As some respondents did not answer all four questions, the sum was divided by the maximum score that a respondent could reach (for example, if a respondent only answered three questions, his or her maximum score was 6). This created a more reliable index, ranging from 0 (no transnational activity) to 1 (high transnational activity). The higher the index score, the more CEE migrants are transnationally active.

Duration of stay

The duration of stay in the Netherlands is measured by the total time elapsed since the respondent’s first visit to the Netherlands for work (thus including periods possibly spent in the home country or elsewhere outside the Netherlands). The duration of stay was measured in months because this is a more exact measurement than years, and because some respondents first came to the Netherlands to work less than a year before the interview. The average duration of respondents’ stay is 27.5 months (slightly over 2 years).

Personal characteristics

The analysis takes account of personal characteristics such as sex, age, educational level and origin. Males are coded with (0), females with (1). Ages range from 18 to 64 years. Educational level is divided into four categories: (0) low (no education, only primary or lower secondary education – up to the age of 15); (1) medium (higher secondary education, including high school); (2) high (higher vocational (polytechnic) and academic education); and (3) other. The latter category contains people who are still studying and people who reported an unclear educational level. The origin of respondents refers to their home country (Poland, Romania or Bulgaria).

Findings

The respondents

Table 1 shows the personal characteristics of the respondents (N = 654). More than half of them were Polish (58 per cent); the others came from Bulgaria (25 per cent) and Romania (17 per cent). Compared to the number of registered migrants from these three countries in 2011, relatively few Polish migrants were interviewed, whereas Romanians and particularly Bulgarians were somewhat over-represented.⁴ More men than women were interviewed (respectively 57 per cent and 43 per cent). The majority of respondents are quite young; almost 53 per cent are aged 30 or below. Bulgarian respondents are on average slightly older compared to the Poles and Romanians in the sample. With regard to educational level, Table 1 clearly shows that current CEE labour migrants – in contrast to a previous generation of labour migrants (‘guest workers’) from the 1960s and 1970s – are generally well educated. Only a minority of respondents (16 per cent) have a low level of education in the sense that they completed only primary or lower secondary school (up to the age of 15). Only Bulgarian respondents more frequently have low educational levels. Almost two-thirds of the respondents (64 per cent) have a medium-level education: they have finished high school or a similar education up to the age of 18. One-fifth of the respondents (one-third of the Romanians) have completed a university or polytechnic education. Compared to the Poles, Romanians and Bulgarians have on average lived somewhat longer in the Netherlands. More than one in five Polish respondents first arrived in the Netherlands two years at most before the interview took place. Only about one in seven respondents had lived in the Netherlands for four years or more. The latter outcomes seem to be in line with previous analyses that emphasise the temporary nature of CEE migration (Wallace 2002; Okólski 2012; Engbersen et al. 2013). However, it may also be that the relatively short migration history of most CEE labour migrants does not yet allow for longer residence in the Netherlands.

Table 1. Personal characteristics of respondents

Source: Dataset Central and Eastern European Migrants, N = 654.

Return intentions

Table 1 also shows the return intentions of respondents, the central dependent variable in the analyses. Respondents were asked how long they intend to stay in the Netherlands. Most migrants in the sample intend to leave the Netherlands in two to ten years’ time (30 per cent). 18 per cent of respondents, most of them Polish nationals, wanted to leave the Netherlands earlier, within two years. When asked where they intend to go after leaving the Netherlands, the vast majority (almost 80 per cent) of the latter category said they want to return home. Only the minority intend to work in another EU country after leaving the Netherlands (14 per cent). About a quarter of all respondents want to live in the Netherlands for at least ten years or permanently. A relatively large proportion (28 per cent) doesn’t know yet if and when they will return to their origin country. Especially Bulgarian respondents (more than 40 per cent of them) often don’t know whether they will stay in the Netherlands or return to their home country, and when.

Explaining return intentions

The main research question in this study asks how the return intentions of CEE labour migrants in the Netherlands can be explained. Table 2 shows the outcome of a multinomial logistic regression analysis that estimates the return intentions of respondents with various (personal) characteristics. The return intentions are categorised in the same way as those in Table 1. Respondents who intend to return within two years (first column), within two to ten years (second column) or who don’t know if and when they want to return (third column) are compared with respondents who intend to stay in the Netherlands for at least ten years or permanently (reference category). For each category, two models are estimated. Model 1 incorporates only the personal characteristics of respondents (origin country, sex, age, education and duration of stay); Model 2 is the full model that incorporates all relevant independent variables.

Table 2. Relative risk ratios (exp(B)) from a multinomial logistic regression analysis of return intentions (reference category is ‘more than 10 years/permanently’)

* p < 0.05; ** p < 0.01; *** p < 0.001 (two-tailed)

¹ Cell is empty.

² Redundant because category has already been taken into account in labour market participation.

Source: Dataset Central and Eastern European Migrants.

We first compare the odds that respondents intend to return to their home country in the short term (within two years) with the odds that they intend to stay for ten years or longer (reference category). Looking at Model 1, various personal characteristics (country of origin, sex, age and duration of stay) seem to have significant effects. Note that the educational level of migrants has no effect on their return intentions. However, when including other relevant independent variables (labour market participation, occupational status, transnational activities and socio-cultural integration) in the analysis in Model 2, the significance levels of some of the personal characteristics change. Two of the personal characteristics (country of origin and duration of stay) lose their significance, and one of the categories of education (‘high’) gains significance. Taking all other factors into account, Polish, Romanian and Bulgarian respondents no longer differ in their return intentions. Duration of stay loses significance because it is significantly related to the level of socio-cultural integration, one of the new variables in Model 2 (Pearson’s r: 0.34; p = 0.000). In other words, it is not so much the duration of their stay as such that affects migrants’ return intentions as their level of socio-cultural integration in Dutch society. Two other personal characteristics remain significant in Model 2: age and sex. The older the respondents, the smaller the odds they intend to return to their home country within two years, compared with the odds of intending to stay for at least ten years (odds ratio: 0.91). The odds that male respondents intend to return within two years is three times higher than the odds that female respondents intend to do so (odds ratio: 3.13). Moreover, the category ‘high educational level’ gains significance: Polish, Romanian and Bulgarian labour migrants with a high educational level are more likely to intend to stay for just two years at most (compared with the odds that they intend to stay for at least ten years) than those with a low educational level.

In addition to these personal characteristics, two other factors included in Model 2 affect the return intentions of respondents: their level of socio-cultural integration and their transnational ties. The more respondents are involved in transnational activities and have transnational contacts with friends and family in their country of origin, and the less their socio-cultural integration into Dutch society, the higher the odds they intend to leave the Netherlands within two years, compared to the odds that they intend to stay for at least ten years (odds ratio 1.04 and 0.15 respectively). The two socio-economic factors included in the analysis in Model 2 play a limited role in the respondents’ return intentions. The sizes of the odds ratios of the occupational status’ categories are relativelylarge, however, the effects do not appear to be significant.

The next column in Table 2 compares the odds that respondents intend to stay in the Netherlands for a medium-to-long period (leaving in between two and ten years) with the odds that they intend to stay for at least ten years or permanently (reference category). Model 1 shows that of the personal characteristics, only age has a strong effect on the return intentions of respondents: the older they are, the greater the odds that they intend to stay in the Netherlands for at least ten years or permanently. This effect remains significant in the full model (odds ratio 0.92 in Model 2). Duration of stay also seems to affect the return intentions of respondents in Model 1, but this factor loses significance in Model 2. On the other hand, level of education gains significance in Model 2: higher-educated and medium-educated respondents are more likely to intend to stay in the Netherlands for a medium-to-long period than lower-educated respondents. Model 2 also shows that both the level of socio-cultural integration of migrants and the intensity of their transnational activities and contacts significantly affect their return intentions. The more respondents are involved in transnational activities and the less their socio-cultural integration in Dutch society, the more they intend to leave the Netherlands in between two and ten years’ time, compared with the odds that they intend to stay in the Netherlands for ten years or longer (odds ratio 1.06 and 0.29 respectively). Here as well, the socio-economic position of respondents is only related to their return intentions to a limited extent. Labour market participation is not related to return intentions at all. With regard to occupational status, only people with a low occupational status have smaller chances of intending to stay for a medium-to-long period (compared with the chance of staying for at least ten years) than people with a medium occupational status.

The final comparison is between the odds that respondents say they ‘don’t know’ how long they will stay in the Netherlands – what Eade et al. (2007: 34) call ‘intentional unpredictability’ – and the odds that they intend to stay for at least ten years or permanently. As previously, duration of stay seems to affect respondents’ return intentions in Model 1, but this factor loses significance in the full model. This is not the case with the other two personal characteristics that affect return intentions: country of origin and age. With all relevant factors taken into account, Bulgarian respondents significantly more often don’t know when they will return to their home country than Poles (compared with the odds that they will stay in the Netherlands for at least ten years or permanently). The same goes, obviously, for younger respondents: the younger they are, the more respondents are inclined to leave the future open. As in the two previous comparisons, both transnational involvement and socio-cultural integration appear to affect migrants’ return intentions. Respondents with more transnational activities and contacts and those who are less socio-culturally integrated in Dutch society are more inclined not to know how long they will stay in the Netherlands, compared with the odds that they intend to stay for at least ten years or permanently (odds ratios 1.04 and 0.33 respectively). Note finally that the socio-economic position of respondents does not appear to be related to their return intentions.

Conclusion and discussion

This article explores the return intentions of CEE labour migrants in the Netherlands. As several million CEE nationals went to the ‘old’ EU to work, the central question now is whether CEE labour migrants will remain only temporarily and then return, or are here to stay like many of the former ‘guest workers’ of the 1960s and 1970s. It is migrants’ return intentions rather than actual return that are explored here, as figures on the actual return migration of CEE labour migrants are scarce. Although return intentions will deviate from actual return, some authors argue that return intentions are significant in their own right. Return intentions show the return motivations of migrants, even if social, economic and political constraints prevent actual return, and can also affect migrant behaviour other than actual return (de Haas, Fokkema 2011; Carling, Pettersen 2014).

Following general migration theories, three hypotheses about return migration and return intentions were formulated at the outset of this research (H1, H2, H3). Neoclassical economic migration theory (NE) expects that successful migrants will intend to stay longer. New Economics of Labour Migration theory (NELM) suggests the opposite: economically successful migrants will return soon (as they have met their income targets) while failed migrants need to stay on. This article argues that in the specific case of CEE labour migrants in Western Europe, NE theory will be more valuable than NELM theory. Given the relatively short distances and the open borders in Europe, there is little reason for economically successful migrants to return, which would be expected to produce a negative association between economic success and return intentions (H1). Socio-cultural integration theory (SIT) expects that not socio-economic success or failure, but socio-cultural integration will be negatively related to return intentions (H2). Transnationalism theory (TT), on the other hand, expects that transnational activities and contacts will be positively related to return intentions (H3). A first and general outcome is that socio-economic factors (the labour market position and occupational status of migrants) does not appear to be related to return intentions at all. Whether or not CEE labour migrants are economically successful in the host country does not affect their return intentions. This means that both NE and NELM fail to explain the return intentions of CEE labour migrants. This finding is in line with several previous studies that also found no or mixed effects of economic success on the return intentions of migrants (Constant, Massey 2002; de Haas, Fokkema 2011; Carling, Pettersen 2014; de Haas et al. 2015). One possible explanation is that migrants return for a variety of reasons. Some return home after a successful migration, whereas others return after their migration has failed (Cerase 1974; Carling 2004). This is in line with Bijwaard and Wahba (2014) who found that intensities of return migration are U-shaped with respect to migrants' income, implying a higher intensity in both low- and high-income groups. Another explanation could be that the indicators of ‘economic success’ ignore the dual reference frame of many labour migrants (cf. Waldinger, Lichter 2003; Wills, Datta, Evans, Herbert, May, McIlwaine 2010). CEE migrants with uncertain jobs and a low occupational status may be less successful according to the standards of the receiving society and the prevailing academic classification schemes, but may be very successful in their own eyes.

The research findings confirm the other two hypotheses. First, the expected positive relation was found between migrants’ transnational activities and their return intentions. CEE migrants who maintain frequent contact with friends and family back home, who pay frequent home visits and who spend larger sums on remittances, intend to return to the home country sooner than migrants who are less transnationally engaged. Second, a negative relation was found between the socio-cultural integration of migrants and return intentions. Labour migrants who are strongly connected with Dutch society in several ways are more inclined to stay in the Netherlands for longer periods of time or even indefinitely than less integrated migrants, even after controlling for the duration of stay in the Netherlands. A more general conclusion of the analysis is that it is socio-cultural factors (socio-cultural integration and transnational activities) that can explain variation in return intentions, rather than the socio-economic success or failure of migrants as postulated by the economic migration theories (NE, NELM).

The research also found that higher-educated CEE labour migrants intend to stay in the Netherlands for shorter periods of time (up to two years or between two and ten years rather than ten years or longer) than lower-educated labour migrants. A possible explanation is that foreign workers who work below their level of training, as many highly skilled CEE labour migrants do, have stronger intentions to return because they do not want to do this work indefinitely (Pungas, Toomet, Tammaru, Anniste 2012). Some higher-educated foreign workers may also be more transnationally oriented and ‘hop’ from one country to another. This may also explain why they intend to stay in the Netherlands for shorter periods of time. Another unexpected outcome is that many Bulgarian respondents, more often than Poles or Romanians, don’t know how long they will stay. A possible explanation could be that the Poles in the sample relatively often work for temporary employment agencies (with clear expectations and agreements about how long they will stay in the Netherlands), whereas many of the Bulgarians are working informally. Obviously, there is less clarity for them about how long they will manage to survive working ‘off the books.’

Finally, there are some limitations to this study. Since the analysis is based on ‘return intentions,’ which is different from actual return behaviour, it is important to complement the study of return intentions with studies on migrants who have actually returned or settled permanently in the host country. Another limitation is that the causality of the relationships is not obvious. Clear associations were found between transnational activities and socio-cultural integration on the one hand and return intentions on the other. However, it is not clear what the causality is. Do strong transnational activity and weak integration within Dutch society cause strong return intentions of CEE labour migrants in the Netherlands? Or is it the other way around: do strong return intentions make CEE labour migrants more transnationally active and less integrated into Dutch society? Or is there another factor at work, such as the migration strategies of migrants, that affects all other factors? For instance, target earners may intend to stay for shorter periods of time, have fewer contacts with Dutch society and retain stronger transnational ties with the home country than other migrant types. The cross-sectional survey data used in this research does not enable these questions to be answered.

Notes

¹ In 2004, Latvia, Estonia, Lithuania, Poland, the Czech Republic, Slovakia, Hungary and Slovenia acceded to the EU, followed by Romania and Bulgaria in 2007.

² Some respondents reported to be self-employed, but appeared to be active in informal work (for instance, street musicians or domestic cleaners) because they were not officially registered as self-employed.

³ The sample is adequate for running PCA: Kaiser-Meyer-Olkin measure (KMO) = .792. The correlations were large enough for running PCA: Bartlett’s test of sphericity χ2 (55) = 2466.243, p < .001.

⁴ According to Statistics Netherlands, ample 91 000 Poles, Romanians and Bulgarians (foreign born) were registered with Dutch municipalities in 2011; 71 per cent were Poles, 16 per cent Bulgarians, and 13 per cent Romanians (source: Netherlands Statistics, Statline; own computations). Bulgarians were oversampled because the cities of Rotterdam and The Hague had special interest in this category.

References

Anacka M., Matejko E., Nestorowicz J. (2013). Ready to Move. Liquid Return to Poland, in: B. Glorius, I. Grabowska-Lusińska, A. Kuvik (eds), Mobility in Transition. Migration Patterns After EU Enlargement, pp. 277–308. Amsterdam: Amsterdam University Press.

Basch L., Glick Schiller N., Szanton Blanc C. (eds) (1994). Nations Unbound: Transnational Projects, Postcolonial Predicaments and Deterritorialized Nation-States. Amsterdam: Gordon Breach.

Bijwaard G., Wahba, J. (2014). Do High-Income or Low-Income Immigrants Leave Faster? Journal of Development Economics 108: 54–68.

Bilgili Ö. (2014). Simultaneity in Transnational Migration Research: Links Between Migrants’ Host and Home Country Orientation. PhD dissertation. Maastricht: Maastricht University.

Borjas G. J. (1989). Immigrant and Emigrant Earnings: A Longitudinal Study. Economic Inquiry 27(1): 21–37.

Carling J. (2004). Emigration, Return and Development in Cape Verde: The Impact of Closing Borders. Population Space and Place 10(2): 113–132.

Carling J., Pettersen, S. V. (2014). Return Migration Intentions in the Integration–Transnationalism Matrix. International Migration 52(6): 13–30.

Cassarino J. P. (2004). Theorising Return Migration: The Conceptual Approach to Return Migrants Revisited. International Journal on Multicultural Societies 6(2): 253–279.

Cerase F. P. (1974). Expectations and Reality: A Case Study of Return Migration from the United States to Southern Italy. International Migration Review 8(2): 245–262.

Constant A., Massey D. S. (2002). Return Migration by German Guestworkers: Neoclassical Versus New Economic Theories. International Migration 40(4): 5–38.

de Haas H., Fokkema T. (2011). The Effects of Integration and Transnational Ties on International Return Migration Intentions. Demographic Research 25(24): 755–782.

de Haas H., Fokkema T., Fihri M. F. (2015). Return Migration as Failure or Success? The Determinants of Return Intentions Among Moroccan Migrants in Europe. Journal of International Migration and Integration 16(2): 415–429.

Drinkwater S., Garapich, M. (2015). Migration Strategies of Polish Migrants: Do They Have Any at All? Journal of Ethnic and Migration Studies 41(12): 1909–1931.

Dustmann C. (1993). Return Intentions of Migrants: Theory and Evidence. Discussion paper 274. Bielefeld: Department of Economics, University of Bielefeld.

Duval D. T. (2004). Linking Return Visits and Return Migration Among Commonwealth Eastern Caribbean Migrants in Toronto. Global Networks 4(1): 51–67.

Eade J., Drinkwater S., Garapich M. (2007). Class and Ethnicity: Polish Migrant Workers in London. End of Award Research Report, ESRC. Swindon: Economic and Social Research Council.

Engbersen G. (2012). Migration Transitions in an Era of Liquid Migration. Reflections on Fassmann and Reeger, in: M. Okólski (ed.), European Immigrations. Trends, Structures and Policy Implications, pp. 91–105. Amsterdam: Amsterdam University Press.

Engbersen G., Leerkes A., Grabowska-Lusińska I., Snel E., Burgers J. (2013). On the Differential Attachments of Migrants from Central and Eastern Europe: A Typology of Labour Migration. Journal of Ethnic and Migration Studies 39(6): 959–981.

Engbersen G., Snel E. (2013). Liquid Migration. Dynamic and Fluid Patterns of Post-Accession Migration Flows, in: B. Glorius, I. Grabowska-Lusińska, A. Kuvik (eds), Mobility in Transition. Migration Patterns After EU Enlargement, pp. 21–40. Amsterdam: Amsterdam University Press.

Engbersen G., Snel E., de Boom J. (2010). A ‘Van Full of Poles’: Liquid Migration from Central and Eastern Europe, in: R. Black, G. Engbersen, M. Okólski, C. Pantiru (eds), EU Enlargement and Labour Migration from Central and Eastern Europe, pp. 115–140. Amsterdam: Amsterdam University Press.

Erikson R., Goldthorpe J. H., Portocarero L. (1979). Intergenerational Class Mobility in Three Western European Societies: England, France and Sweden. British Journal of Sociology 30(4): 415–441.

Erikson R., Goldthorpe J. H., Portocarero L. (1983). Intergenerational Class Mobility and the Convergence Thesis: England, France and Sweden. British Journal of Sociology 34(3): 303–343.

Fassmann H., Kohlbacher J., Reeger, U. (2014). The Re-eEmergence of European East–West Migration – The Austrian Example. Central and Eastern European Migration Review 3(2): 39–59.

Favell A. (2008). The New Face of East–West Migration in Europe. Journal of Ethnic and Migration Studies 34(5): 701–716.

Fihel A., Górny A. (2013). To Settle or to Leave Again? Patterns of Return Migration to Poland During the Transition Period. Central and Eastern European Migration Review 2(1): 55–76.

Fihel A., Grabowska-Lusinska I. (2014). Labour Market Behaviours of Back-and-Forth Migrants from Poland. International Migration 52(1): 22–35.

Friberg J. H. (2012). The Stages of Migration. From Going Abroad to Settling Down: Post-Accession Polish Migrant Workers in Norway. Journal of Ethnic and Migration Studies 38(10): 1589–1605.

Gmelch G. (1980). Return Migration. Annual Review of Anthropology 9: 135–159.

Holland D., Fic T., Rincon-Aznar A., Stokes L., Paluchowski P. (2011). Labour Mobility Within the EU. The Impact of Enlargement and the Functioning of the Transitional Arrangements. Final report. London: National Institute of Economic and Social Research.

Hołda M., Saczuk K., Strzelecki P. Wyszyński R. (2011). Settlers and Guests – Determinants of the Plans of Return Migration from UK and Ireland to Poland in the Period 2007–2009. Working Paper 84. Warsaw: National Bank of Poland.

Isański J., Mleczko A., Seredyńska-Abou Eid R. (2014). Polish Contemporary Migration: From Co-Migrants to Project ME. International Migration 52(1): 4–21.

Itzigsohn J., Saucedo S. G. (2002). Immigrant Incorporation and Sociocultural Transnationalism. International Migration Review 36(3): 766–798.

Jensen P., Pedersen J. P. (2007). To Stay or Not to Stay? Out-Migration of Immigrants from Denmark. International Migration 45(5): 87–113.

Kaczmarczyk P. (2013). Matching the Skills of Return Migrants to the Labour Market in Poland, in: OECD (eds), Coping with Emigration in Baltic and East European Countries, pp. 111–126. Paris: OECD Publishing.

King R. (2000). Generalizations from the History of Return Migration, in: B. Ghosh (ed.), Return Migration: Journey of Hope or Despair?, pp. 7–55. Geneva: International Organisation for Migration.

Krings T., Bobek A., Moriarty E., Salamońska J., Wickham J. (2013). Polish Migration to Ireland: ‘Free Movers’ in the New European Mobility Space. Journal of Ethnic and Migration Studies 39(1): 87–103.

Ley D., Kobayashi A. (2005). Back to Hong Kong: Return Migration or Transnational Sojourn? Global Networks 5(2): 111–127.

Massey D. S. (1990). Social Structure, Household Strategies, and the Cumulative Causation of Migration. Population Index 56(1): 3–26.

Moriarty E., Wickham J., Salomońska J., Krings T., Bobek A. (2010). Putting Work Back in Mobilities: Migrant Careers and Aspirations. Paper delivered at the conference ‘New Migration, New Challenges,’ Dublin, Trinity College, 3 June – 3 July 2010.

Oeppen C. (2013). A Stranger at ‘Home’: Interactions Between Transnational Return Visits and Integration for Afghan–American Professionals. Global Networks 13(2): 261–278.

Okólski M. (2001). Incomplete Migration: A New Form of Mobility in Central and Eastern Europe. The Case of Polish and Ukrainian Migrants, in: C. Wallace, D. Stola (eds), Patterns of Migration in Central Europe, pp. 105–128. Houndmills: Palgrave Macmillan.

Okólski M. (2012). Transition from Emigration to Immigration. Is It the Destiny of Modern European Countries?, in: M. Okólski (ed.), European Immigrations. Trends, Structures and Policy Implications, pp. 23–44. Amsterdam: Amsterdam University Press.

Okólski M., Salt J. (2015). Polish Emigration to the UK After 2004; Why Did So Many Come? Central and Eastern European Migration Review 3(2): 11–37.

Pungas E., Toomet O., Tammaru T., Anniste K. (2012). Are Better Educated Migrants Returning? Evidence from Multi-Dimensional Education Data. Norface Migration Discussion Paper 2012-18.

Samers M. (2010). Migration. London, New York: Routledge Taylor & Francis Group.

Sassen S. (1997). Migranten, Siedler, Flüchtlinge. Von der Massenauswanderung zur Festung Europa. Frankfurt aM.: Fischer Taschenbuch Verlag.

Sinatti G. (2011). ‘Mobile Transmigrants’ or ‘Unsettled Returnees’? Myth of Return and Permanent Resettlement Among Senegalese Migrants. Population, Space and Place 17(2): 153–166.

Snel E., Engbersen G., Leerkes A. (2006). Transnational Involvement and Social Integration. Global Networks 6(3): 285–308.

Snel E., Faber M., Engbersen G. (2014). Civic Stratification and Social Positioning: CEE Labour Migrants Without a Work Permit. Population, Space and Place 21(6): 518–534.

Stark O. (1991). The Migration of Labor. Cambridge: Blackwell.

Statistics Netherlands. (2012). Return Migration Rates of Recent Immigrants Compared to Flows in the Previous Century. Paper presented at the Conference of European Statisticians, Group of Experts on Migration Statistics, Work Session on Migration Statistics, Geneva, 17–19 October 2012.

van der Heijden P., Cruyff M., van Gils G. (2013). Aantallen geregistreerde en niet-geregistreerde burgers uit MOE-landen die in Nederland verblijven. Rapportage schattingen 2009 en 2010. Utrecht: Universiteit Utrecht.

van Meeteren M., Engbersen G., Snel E., Faber M. (2014). Understanding Different Post-Return Experiences. The Role of Preparedness, Return Motives and Family Expectations for Returned Migrants in Morocco. Comparative Migration Studies 2(3): 335–360.

Waldinger R. D., Lichter M. I. (2003). How the Other Half Works. Berkeley, Los Angeles: University of California Press.

Waldorf B. (1995). Determinants of International Return Migration Intentions. Professional Geographer 47(2): 125–136.

Wallace C. D. (2002). Opening and Closing Borders: Migration and Mobility in East-Central Europe. Journal of Ethnic and Migration Studies 28(4): 603–625.

White A. (2011). Polish Families and Migration Since EU Accession. Bristol: Policy Press.

White A. (2014). Double Return Migration: Failed Returns to Poland Leading to Settlement Abroad and New Transnational Strategies. International Migration 52(6): 72–84.

Wills J., Datta K., Evans Y., Herbert J., May J., McIlwaine C. (2010). Global Cities at Work: New Migrant Divisions of Labour. New York: Pluto Press.

Annex 1

Table A1. Results of principal component analysis (PCA)

Source: Dataset Central and Eastern European Migrants.

Table A2. Index score distribution transnational activities

Source: Dataset Central and Eastern European Migrants.