To Settle or to Leave Again? Patterns of Return Migration to Poland During the Transition Period

  • Published in:
    Central and Eastern European Migration Review, Vol. 2, No. 1, 2013, pp. 55-76
    Views: 18450

The article discusses the notion of return migration with regard to its permanency and temporariness. In reference to selective patterns of return migration, factors conducive to permanent returns and to re-emigration, i.e. subsequent migration after the return, are examined with the use of a logistic regression model. Analyses demonstrated in the article are devoted to return migration to Poland in 1989-2002 and based on the 2002 Polish census data. The obtained results confirm earlier findings on the major role of the level of human capital and family attachments in shaping the nature of the return waves. It was revealed that return migrants who decided on a longer stay in Poland were more often living in Polish urban areas, and had higher human capital and stronger family attachments to Poland, when compared to re-emigrants. It was also observed that return migrants possessing dual nationality were the most likely to engage in re-emigration, while descendants of Polish emigrants tended to settle in Poland on a more permanent basis.

Keywords: return migration, Poland, re-emigration, temporary migration

Introduction

It can be argued that two distinct waves of return migration can already be identified since the fall of the communism in Poland: return migration of the transition period, encompassing Polish emigrants and their descendants who returned to Poland mainly in the 1990s; and a return wave that started after the accession of Poland to the European Union in 2004, as a consequence of the economic crisis of recent years. These two waves of returns to Poland differ in several aspects from previous waves of returns to Poland that took place in the XX century. Among other differences, they have been of a less permanent nature (Anacka, Fihel 2012a) than earlier return waves, which is in line with what has been observed in other countries and contexts (Cassarino 2004).

The goal of our article is to contribute to the discussion on permanency and temporariness of return migration. The research question we aim to answer in this article is what factors increase the likelihood of permanent return migration to Poland and re-emigration, i.e. subsequent emigration after return to Poland. We address this question in the framework of selectivity of migration by identifying individual characteristics of return migrants that make them prone to remain in Poland and to re-emigrate after their return. Basing on Polish and international literature on the topic, we take into consideration the following socio-demographic characteristics of return migrants: gender, age, education, family status, citizenship, place of birth, place of residence in Poland, duration of stay in Poland and year of return.

In pursuing our goal we examine the case of return migration of the transition period to Poland in the 1990s and at the beginning of 2000s, with the use of the 2002 census data referring to migrants who returned to Poland for permanent residence in the intercensal period (precisely, from 1989 to mid-2002). We believe that our results can contribute to the scientific discussion on the topic of permanency and temporariness of return migration even though we are examining an already historical wave of returns to Poland. At the same time, however, it should be acknowledged that the degree of temporariness has presumably been higher in post-accession return flows, but, as of today, lack of adequate quantitative data makes it impossible to conduct analyses of that kind for most recent returns. Moreover, we are convinced that our contribution fills an important gap in the research on return migration to Poland during the transition period, since the degree of permanency in this return flow has not been explored with the use of quantitative indicators.

Theoretical background

Return migration and temporary migration – perspectives of the destination and home countries

Several main topics can be distinguished in the literature on return migration: selectivity of return migration and motives of returnees, as well as consequences of the return for both returnees and countries to which they return. Alongside the growing importance of temporary and fluid forms of mobility, a discussion on the interrelations between return migration and repetitive mobility forms (e.g. circular migration) has also been evolving (cf. Cassarino 2004).

It can be argued that, in the 1970s and 1980s, the return migration literature was largely devoted to consequences of return migration for migrants themselves and for countries to which they were returning, with the main attention paid to the role that returnees might play in the development of these countries (e.g. Cerase 1974). Earlier works on return migration thus more frequently took the perspective of the home country (country of emigrants’ origin, to which they returned) and not the destination country (country from which migrants returned). The latter perspective has been more popular in recent works on return migration, which implies some shift of interest in return migration studies towards an examination of determinants of returns and attempts to draw a line between definitive returns and repetitive mobility (Dustman, Weiss 2007).

The ‘traditional’ definition of return migration states that it is the movement of emigrants back to their homelands to resettle. Migrants returning for vacation or an extended visit without intention of remaining at home are generally not defined as return migrants (Gmelch 1980: 136). Such an understanding of return migration implies permanency of a return move, which is not that realistic in the present era of intensive mobility. Nevertheless, even nowadays, it is often assumed that we can talk about return migration only when migrants have some longer-term plans regarding stay in the home country after their return. According to the OECD glossary, returning migrants are persons returning to their country of citizenship after having been international migrants (whether short-term or long-term) in another country and who are intending to stay in their own country for at least a year.1

While the aforementioned vision of return migration is usually applied in most studies on return migration employing the perspective of the home country, the destination country perspective tends to alter this ‘traditional’ approach. Because researchers in the destination countries cannot easily follow migrants after their departure from the studied country, identifying a direction and assessing permanency of the given departure from the destination country is frequently problematic. In particular, the assumption, often made in this stream of return migration literature, that leaving migrants return to their home countries (cf. Constant, Massey 2002; Dustman, Weiss 2007) can be questioned in some contexts.

At the same time, notwithstanding the perspective taken by researchers – be it the perspective of the home or destination country – some specific aspects in return flows are acknowledged in the migration literature. In general, it can be argued that migrants have some special attitudes toward their home country that shape their propensity to return migration. Gosh (2000) introduced this aspect in the form of the category of ‘homesickness’. Dustman and Weiss (2007) posited a preference of migrants for consumption in their home country, over consumption in the destination country. Other authors frequently discussed the ‘myth of return’ being preserved in various migrant communities in the destination countries (cf. Bolognani 2007; Ganga 2006).

An important line of research on return migration investigates the selectivity mechanisms among migrants in general, and return migrants in particular. Among the latter, the most frequently addressed socio-demographic characteristics affecting selectivity include: gender, age, human capital, family situation, earnings and various attachments of migrants to the home and destination countries, such as possession of a house, citizenship and declared identity (cf. Dustman, Kirchkamp 2002; Constant, Massey 2002). In general, the evidence for patterns of selectivity among return migrants is mixed, especially when we differentiate between studies conducted in the countries of origin and destination. In our view, this is related, among other things, to crucial differences in definitions of return migration employed in these two groups of studies. Therefore, in the sections to follow, while demonstrating results of studies pertaining to selectivity of return migration, we comment also on the definitions applied by their authors.

For the home country perspective, results of studies basing on LFS data, where return migrants are individuals who had taken up international emigration and subsequently returned to their households in the home countries, are worth mentioning. For example, Martin and Radu (2012) examining return migration to five home countries of Central and Eastern Europe – Czech Republic, Hungary, Lithuania, Poland and Romania – in the years 2002-2007,2 demonstrated that return migrants were on average younger than the general group of migrants. In comparison to natives, returnees were also more likely to be single, live in one-person households, and possess an intermediate or higher education level. In contrast, in interviews with post-accession Polish return migrants registered in the LFS data, Anacka and Fihel (2012b) found out that return migrants tended to be older, possess either a higher or a primary level of education, and originate from urban areas in Poland, when compared to migrants who remained abroad.

From among works on return migration applying the destination country perspective, the work by Dustman and Kirchkamp (2002) is worth mentioning because it employed a rather ‘orthodox’ approach to the definition of return migration. The studied group of Turkish migrants encompassed persons who had returned from Germany to Turkey in 1984 and were interviewed in Turkey afterwards, in 1986 and 1988. Selectivity of return migration was examined via determinants responsible for shortening migrants’ stay in Germany. According to the obtained results, highly educated Turkish migrants and those possessing family ties established in Turkey prior to migration were more likely to return to Turkey sooner.

Another interesting contribution by Dustman and Weiss (2007), also addressing selectivity of return migration by examining determinants of shortening migration duration, can be treated as an approach employing a rather ‘unorthodox’ definition of return migration. The authors, with the goal of proposing a theory of return migration, understood return migration as the type of migration one usually has in mind when referring to a migration as being temporary. Return migration describes a situation where migrants return to their country of origin [for at least one year] by their own choice, often after a significant period abroad (Dustman, Weiss 2007: 238). In other words, in the quoted study, return migration was treated as an inherent element of temporary movements.

Taking LFS data for the UK as a starting point, Dustman and Weiss (2007) proposed an interesting theoretical economic model in which they developed a relatively nuanced perspective. According to this model, three main categories of determinants of return migration can be identified: 1) a migrant’s greater preference for consumption in the home country when compared to the country of immigration; 2) higher purchasing power of the host country currency; 3) accumulation of human capital in the host country, which, when transferred to the home country, increases the productivity of the migrant back home. It deserves attention that, according to this approach, not only the level of human capital accumulated by migrants abroad, but also the degree of its transferability to the home country played a role in decisions about return migration.

Another contribution to research on selectivity patterns in return migration was delivered by Constant and Massey (2002), who, focusing on migrants from Italy, Greece, Turkey and the former Yugoslavia, examined a powerful dataset – the German Socio-Economic Panel for the years 1984-1999. They defined return migration as any absence from the panel [i.e., from Germany] for three consecutive years (Constant, Massey 2002: 23), silently assuming that, when leaving Germany, migrants go to their home countries. In the proposed definition, they thus assumed some permanency in return migration, since only longer stays outside of Germany were considered as return migration.

Results obtained by Constant and Massey (2002) did not provide any evidence for positive selection of return migrants with regard to their human capital and socio-economic performance in Germany or for selection with regard to gender, age and legal status in Germany. According to the quoted study, return migration was, first of all, associated with family attachments to the home country (partner and children in the home country) and the fact of sending remittances. Conversely, migrants possessing family and other attachments to Germany, such as German citizenship and those who declared German identity, were less likely to return to their home countries.

When demonstrating the complexity of mechanisms pertaining to return migration, it is also worth mentioning the conclusions from another study by Constant and Zimmerman (2011), based on the same dataset and examining the same migrant groups as in the study by Constant and Massey (2002). Differences in the results obtained in these two studies speak for themselves. Constant and Zimmerman (2011) examined selectivity of circular migration, operationalised as relatively numerous departures from Germany (the studied destination country) associated with a reasonable period of time spent outside Germany. They demonstrated that dual citizens, migrants not owning a dwelling in Germany, the youngest and the oldest age-groups, as well as migrants possessing families in the home countries were more likely to engage in circular migration. This was not the case for migrants with higher education and those who were more attached to the German labour market, who were less interested in that type of mobility.

In our opinion, the presented diversity of definitions and of mixed results from studies on return migration demonstrates that an investigation of the permanency and temporariness of returns is worth undertaking, in order to deepen our understanding of selectivity mechanisms of return moves. This is in line with the argument of Constant and Massey (2002), who claim that the nature of return migration differs with regard to permanent and temporary mobility logics. Though the task is complex, in this article, we make an attempt to conduct such an investigation by analysing the already historical return flow to Poland during the transition period.

Return migration to Poland – concepts and main areas of research

Several phases of return migration to Poland in the XX century have been identified in the literature on the topic (cf. Anacka, Fihel 2012a; Bade 2000; Gawryszewski 2005; Kołodziej 1998; Slany, Małek 2002). After the political and economic transition that took place in Poland in the late 1980s, we witnessed two distinct waves of returns. Within the first one, not only Polish emigrants but also their descendants returned to Poland in the 1990s, attracted by new opportunities that arose in their home country after the abolition of the communist system (Iglicka 2002; Klagge, Klein-Hitpaß, Fihel, Kindler, Matejko, Okólski 2007). The most recent returns to Poland, within the second wave, were linked to the mass emigration of Poles to countries of the European Union (primarily to the United Kingdom) after its enlargement in 2004 (for more see, Bieńkowska, Ulasiński, Szymańska 2010a, b; Szymańska, Ulasiński, Bieńkowska 2012). The economic crisis spreading across Europe since 2008 accelerated and – probably – stimulated the reverse movement, from Western Europe to Poland.

Of these two return waves, the post-accession return migration was of an incomparably larger scale. Estimates say about over half a million Poles returned to Poland in the post-accession period (cf. Anacka, Fihel 2012a), whereas return migration numbers after the transition period do not exceed 100 thousand (Iglicka 2002; Fihel, Górny, Matejko 2006). Given differences in the scale and political context of these two return moves to Poland, topics addressed in studies devoted to them also differ.

Studies devoted to return migration of the transition period, apart from the study by Fihel et al. (2006) based on the 2002 census data, were usually of a qualitative nature. A mixed, though largely quantitative, approach was taken by Jończy (2003) in his complex studies on emigration processes – with their causes and consequences – from the Opole region, which is inhabited by a large number of dual Polish-German citizens. Even though Jończy (2003) does not use the term return migration, his works are devoted, among other issues, to identifying mechanisms of intensive circular migration between Poland and (mainly) Germany. Among other things, he found out that, during the transition period, men and individuals possessing children were more likely to work permanently abroad (visiting Poland only occasionally), whereas women and individuals without children were more eager to engage in circular movements between Poland and Germany.

In other studies devoted to return migration to Poland in the transition period, the most frequently addressed topics concerned motives of returns and adaptation of return migrants to the transforming Polish economy and society. Results of the conducted research revealed that, following the typology of Cerase (1974), all types of returns were observed in Poland at that time: returns of failure, returns of conservatism, returns of retirements and returns of innovation (cf. Iglicka 2002). Returning migrants were considered as agents of innovation, bringing valuable human capital to Poland (Fihel et al. 2006; Górny, Kolankiewicz 2002). Only return migrants from Germany coming back to the region of Upper Silesia were more likely than other returnees to pursue a ‘return of failure’ (Koryś 2002).

Apart from specific studies by Jończy (2003), the permanency of returns in the 1990s has not been heavily discussed in the literature on Polish return migration of the transition period. Górny and Osipoviĉ (2006), in their qualitative study examining this aspect, arrived at the conclusion that the ‘myth of return’ was strong among Poles returning from Great Britain, but the decision to stay in Poland for good was dependent on the family situation of returnees and their opportunities on the Polish labour market. These outcomes pertain, however, to a specific group of returning migrants – second-generation British Poles – thus they cannot be easily extrapolated to other groups of returnees.

It can be argued that, in comparison to research on return migration to Poland during the transition period, studies on post-accession return migration are more complex and employ quantitative indicators to a greater extent. It is beyond the scope of this article to demonstrate all topics addressed in this line of research; it is, however, worth presenting selected results pertaining to the selectivity of return migration and its permanency and temporariness. 

Selectivity patterns of post-accession return migration to Poland were best addressed in a unique, large-scale, though regional, study conducted in 2010-2012, which was devoted to post-accession return migration to three southern Polish voivodeships (cf. Bieńkowska et al. 2010a, b; Szymańska et al. 2012). Its results clearly demonstrate that selectivity of return migration differs for men and women. In particular, highly educated women were the least eager to return to Poland. Evidence for the role of age in selection of return migrants was mixed for different regions, though a somewhat higher tendency to return was observed among the youngest migrants, aged below 30.

The same study, focusing on Małopolska region, also provides an interesting typology of returning migrants, dividing them into those who intended to remain in Poland (70 per cent of all returning migrants) and those who planned subsequent migration. Individuals who had a job in Poland constituted only one third of the latter group, whereas half of it was comprised of circular migrants who treated work abroad as their main income-generating strategy. Among circular migrants, men and individuals with vocational training prevailed – 60 per cent of circular migrants and one third, respectively. Permanent migrants were relatively diversified, but the authors of the study distinguished four main types – tourists, specialists, investors and actors of change (cf. Bieńkowska et al. 2010a).

In another study by Iglicka (2010), conducted on unrepresentative sample of 200 return post-accession migrants, over one fifth of respondents declared that they would definitely re-emigrate. Matejko (2010), conducting a qualitative analysis of post-accession return migrants, discussed intentionally accomplished returns and intentionally unaccomplished returns. It is thus clear that, in contrast to studies on return migration to Poland during the transition period, the topic of permanency and temporariness of returns takes a prominent place on the research agenda in studies devoted to post-accession return moves.

With regard to the definition of return migration, the literature on returns of the transition period is not fully consistent, but it can be argued that the most frequently applied definitional dimensions relate to the nature of people’s emigration history before they arrive in Poland, and to their attachments to the Polish culture and land. For example, Heffner and Sołdra-Gwiżdż (1997), focusing on return migration to Upper Silesia, limited their analyses to people who lived on the territory of Upper Silesia and had close and direct family ties with Poland, or were descendants of inhabitants of the Upper Silesia and/or came to Poland for retirement. Thus, connections to the region where heavily stressed in this approach. 

A more general and formal definition of return migration was proposed by Iglicka (2002), who studied returnees from the US, the United Kingdom and Germany. According to her approach, a return migrant was defined as a person who was born in Poland, left Poland and settled abroad (notwithstanding whether the decision about settlement was made by him/her or his/her parents) and came back to Poland after 1989 (Iglicka 2002: 23-24). Thus, this approach excluded descendants of Polish emigrants from the group of returnees, even if they possessed Polish citizenship.

The latter group was included in the definition of return migrants employed by Górny and Osipoviĉ (2006), who studied second-generation Poles coming to Poland from the United Kingdom. In this research, a return migrant was defined as a person who had at least one Polish parent (i.e. held Polish citizenship) and was raised in a Polish environment in the United Kingdom (for more explanations, compare Górny, Osipoviĉ 2006). Although such a definition can be considered quite specific, it encapsulates an important group of individuals with Polish origins who arrived in Poland in the 1990s.

On the one hand, it can be argued that none of the definitions of the transition period return migrants encompassed all types of returnees coming to Poland in the 1990s and 2000s. On the other hand, it should be stressed that the substantial variety of migrant types returning to Poland after 1989 makes it particularly difficult to propose one universal operationalisation of return migration.

Data and method of analysis

Data and measurement

Analyses presented in this article are based on data from the 2002 census pertaining to return migrants who came to Poland for a permanent stay in 1989-2001 or during the first half of 2002 (until 20th May 2002 – the moment of the census).3 The total number of Polish return migrants recorded in the census equalled 69 704 individuals. From this group, we excluded from our analyses persons aged less than 18 at the moment of their return to Poland, since we did not intend to examine those who most likely had not taken part in the decision-making process regarding a return to Poland. After also excluding persons for whom information necessary to conduct our analyses was incomplete, we analysed 45 813 persons in total.

We define a return migrant as a person having Polish citizenship that registered for a permanent stay in Poland in 1989-20024 after having spent some time abroad notwithstanding his/her place of birth (cf. Fihel et al. 2006). In this way, we include in our analysis not only Polish emigrants, but also their descendants who were most likely raised in Polish families and inherited Polish citizenship from their parents. An inflow of such people was a characteristic feature of the transition period return wave. With regard to the place of birth, we should also acknowledge that people born on the pre-war Polish territory incorporated into the former Soviet Union in 1945, after the Second World War, were considered foreign-born persons in the 2002 census. For the studied population of return migrants (at least 18 years old at the moment of return), such a group could encompass at most 1 029 persons (born on the present territory of Belarus, Lithuania, Russia or Ukraine before 1946).

It should be also noted that a serious drawback of the definition of the studied population of return migrants is that the moment of return migration, as registered in the 2002 census, was not necessarily directly linked to a trip to Poland but to a permanent registration in Poland.  Consequently, on the one hand, some groups of return migrants were not captured in the 2002 census. This applies to two main categories: 1) migrants who came to Poland, but did not register for a permanent stay, and 2) migrants who left Poland before or after 1989 and returned in the period 1989-2002, but never cancelled their permanent registration in Poland. On the other hand, Polish emigrants who only registered for a permanent stay in Poland, but never returned to Poland were recorded as return migrants in the 2002 census (though their absence was recorded). Such a situation can take place when an emigrant inherited, for example, an apartment in Poland, and it was beneficial for him/her to have a permanent registration in Poland. Taking into account that such individuals are not ‘real’ return migrants, we excluded from our analyses all return migrants for whom the date of re-emigration was recorded as earlier than the date of return – 216 adults at the moment of return.

In the group of return migrants, we differentiate between re-emigrants and return migrants staying in Poland until mid-2002 – the moment of the census. By ‘re-emigrant’ we mean a return migrant who had emigrated again from Poland after his/her return in 1989-2002 and was absent at the moment of the census. They were recorded in the census as persons possessing permanent registration in Poland but being absent at the moment of the census. It should be acknowledged that return migrants who took up international trips and returned to Poland by 2002 are missing from the studied group of re-emigrants, even though they should be considered as re-emigrants. However, information about intercensal mobility is not available in the census data (though it was partly collected).

In contrast to re-emigrants, we consider the remaining return migrants staying in Poland in 2002 as permanent return migrants. Clearly, on the one hand, this is a simplification, since some of them might have re-emigrated after the census. On the other hand, however, the year 2002 can be considered as a date very close to the end of the transition period in Poland, at least in the migratory framework. Poland’s accession to the European Union completely changed the context of emigration, re-emigration and return migration of Poles. Nevertheless, it should be stressed that our analyses pertain only to a specific moment in time and the category of ‘permanent migrants’ should be perceived rather as an analytical than a perfect empirical category.

Method of analysis

In analyses shown in this article, we use a logistic regression model (Greene 2003) to estimate the likelihood of return migrants to stay in Poland until 2002. This methodological approach has two important advantages: first, it disentangles the combined effects of different independent variables on the probability of permanent return to Poland; and second, it demonstrates the statistical significance of these variables. Logistic regression analysis is applicable when the dependent variable can be described by two states – for instance, migration took place or did not.

The dependent binary variable identifies permanent return migrants. Therefore, the value ‘1’ of the dependent variable designates the fact that a return migrant remained in Poland in 2002 at the moment of the census, whereas the value ‘0’ refers to the fact that a person emigrated abroad after having returned to Poland in 1989-2002.

Independent variables include basic socio-demographic characteristics of return migrants such as: gender, age (four categories: 20-29, 30-39, 40-49, 50 or more),5 education (five categories: primary education or lower, vocational, secondary without a certificate of completion, secondary or secondary vocational with certificates, higher education: BA or MA degrees), family status (four categories: single, couples without children, couples with children, single parents with children) and also place of residence in Poland (rural or urban area).

We also take into account variables related to international mobility and national identity of return migrants, such as: duration of stay in Poland after return, held citizenships (three categories: only Polish, dual citizenship Polish-German, dual citizenship Polish-other), country of birth (Poland or foreign country) and period of return to Poland (1989-1992, 1993-1996, 1997-1999, 2000-2002).6

Moreover, the model includes two statistically significant interactions: 1) between the migrant’s place of residence in Poland and educational level and 2) between the migrant’s place of birth and citizenship. The reason for including these interactions stemmed from our belief that the nature of the association between level of education and inclination to re-emigrate differs between inhabitants of urban and rural areas. Also, the fact of having dual nationality might have a varied influence on probability of stay in Poland among people born in Poland and abroad.

All information regarding characteristics of migrants referred to the moment of the census, i.e. May 20th, 2002. Such a setting is satisfactory for return migrants remaining in Poland in 2002, but problematic for re-emigrants, who may have emigrated several years before the census and for whom data on the place of residence and family status referred to the Polish household where migrants remained registered for their permanent stay. Consequently, we do not have information on the situation of re-emigrants for the moment when they re-emigrated. Therefore, in our interpretations, we focus on determinants of permanent returns in the transition period and only comment on differences between return migrants and re-emigrants at the moment of the census.

Economic characteristics of return migrants, such as labour market situation and main source of income, though valuable in explaining migratory patterns, were not incorporated into the model. This was due to the fact that, for re-emigrants, relevant variables referred to the situation in the destination country. Consequently, they could not be examined in relation to selectivity patterns in re-emigration since the situation on the labour market in the destination country could be both the precondition and the outcome of re-emigration.

Return migration to Poland – trends and socio-demographic characteristics of migrants

In the intercensal period (1989 to mid-2002), 69 704 Polish citizens returned to Poland and registered for a permanent stay. For a number of them, however, their arrival to Poland was not a definitive return: at the moment of the 2002 census, as many as 19 630 persons (28 per cent of all returnees) were not present in Poland and most of them – 17 493 persons – had stayed abroad for at least one year. Interestingly, while incidence of returns to Poland was rather stable over the whole intercensal period, re-emigrations were particularly frequent in the first years of political transition, that is, in 1989-1990 (Figure 1). As many as 42 per cent of those who returned to Poland in 1989-1990 left Poland within one year. At the same time, 73 per cent of all re-emigrants did not stay in Poland for longer than one year. There can be at least two explanations of such migratory behaviours. First, these migrants’ short stay in Poland could be related to adaptation problems some of them encountered in the transforming Polish society. Second, some migrants could have not intended to return to Poland permanently, but only to register for a permanent stay in Poland due to family or bequest reasons. In general, the average duration of stay in Poland was one year and two months for re-emigrants and almost six years for the rest of the group.

Figure 1. Returns and re-emigrations to/from Poland in 1989-mid 2002, by year, in per cent

Source: own elaboration based on the 2002 census.

A high representation of dual citizens constituted a distinctive feature of return migration during the transition period. Polish citizenship based on the ius sanguinis principle was easily retained by Polish emigrants and transferred to their children (cf. Górny, Pudzianowska 2009). Over one third of return migrants possessed dual citizenship, predominantly Polish-German, which was held by two thirds of returnees with dual citizenship. Other popular foreign citizenships included: American (18 per cent of all returnees with dual citizenship), Canadian (9 per cent), French (5 per cent) and Australian (4 per cent). The percentage of dual citizens was particularly high in the group of re-emigrants – 55 versus 27 per cent in the rest of the group of return migrants (Figure 2). This indicates that dual citizenship enhanced international mobility and made settling down in Poland less probable in the transition period.

Figure 2. Return migrants by nationality and migratory status, in per cent

Source: own elaboration based on the 2002 census.

Polish returnees of the transition period were returning, first of all, from Germany (31 per cent of returnees), but also from the US (17 per cent), Italy (5 per cent), Canada (5 per cent), France (4 per cent) and from other popular destinations of Polish emigrants in the communist period and after transition. The 2002 census also revealed the existence of an inflow, though rather small, of Polish citizens from Eastern countries, such as Kazakhstan (3 per cent of all returnees), Ukraine (3 per cent), Russia (2 per cent) and Lithuania (1 per cent). For dual citizens, there was a strong correlation between the country of last residence and the country of foreign citizenship.

It should be also noted that the share of foreign-born individuals was relatively high in the population of returnees – 29 per cent. This demonstrates the importance of the group of descendants of Polish emigrants in the return wave of the transition period. Interestingly, the proportion of such individuals was even higher – 35 per cent – when we considered only returnees who stayed in Poland in 2002. Apparently, return migration of descendants of Polish migrants tended to be more permanent than returns of Polish emigrants: from among foreign-born return migrants as many as 86 per cent stayed in Poland in 2002, whereas for persons born in Poland the respective share accounted for 66 per cent. 

In Poland, returnees mainly chose urban areas (78 per cent) and, especially, major Polish cities, such as Warsaw, Cracow, Poznań, the Tricity area, and Wrocław. Permanent return migrants were particularly likely to settle in urban areas. This pertained to 82 per cent of them, as compared to 69 per cent of re-emigrants. Apart from the above-listed cities, Śląskie voivodship drew an important group of returnees (12 per cent), both due to its urban character and long traditions of emigration from this region to Germany. It may be thus presumed that return migrants tended to settle down in economically attractive regions, with good work opportunities and a high demand for specific and high skills, rather than in regions of their or their families’ origin in Poland.7

Gender distribution was almost balanced in the group of returnees, while women slightly prevailed among re-emigrants with a 55 per cent share. At the same time, return migration during the transition period should be perceived as a flow of several generations and of families. This pertained especially to permanent return migrants, which is well portrayed in their age structure, dominated by middle-aged people and children (Figure 3). People below the age of 18 at the moment of the census constituted as much as 27 per cent of permanent returnees, compared to 20 per cent among re-emigrants. Also, the presence of people of retirement age (60 or more) was more visible in the group of permanent return migrants – 11 per cent versus 5 per cent among re-emigrants.

Figure 3. Age structure of return migrants

Source: own elaboration based on the 2002 census.

The majority of return migrants lived in (formalised or not formalised) unions – 61 per cent of adult individuals were in family households in Poland. In turn, Polish households of re-emigrants were more frequently one-person households or consisted of unrelated persons – 28 per cent versus 17 per cent in the rest of the group of returnees. Among re-emigrants the share of single people was also rather high, accounting for 26 per cent of the group, whereas in the rest of the returning population it was equal to 20 per cent.

Although overall return migration was not large in scale during the transition period, it was unique in terms of human capital possessed by return migrants: 30 per cent of adult returnees (aged 18 years or more) possessed university degrees and a further 36 per cent had completed secondary education (Figure 4). Permanent returnees were particularly well educated. In this group, the share of persons with university degrees reached 33 per cent, whereas for re-emigrants – ‘only’ 21 per cent. In contrast, re-emigrants were particularly likely to possess vocational or unfinished secondary education – 38 per cent of re-emigrants, in comparison to 22 per cent of the remaining group of returnees. Representation of other levels of education was close to the average for the total group of return migrants.

Figure 4. Level of education of return migrants aged 18 and over, by migratory status, in per cent

Source: own elaboration based on the 2002 census.

Most of all adult returnees – 61 per cent – were economically active (employed or unemployed). As of May 2002, 55 per cent of adult returnees were employed, and this share was similar among those who stayed in Poland till 2002 and those who left (54 versus 58 per cent). Permanent returnees seeking employment accounted for 12 per cent of the total group, whereas among re-emigrants the respective share was visibly lower – 4 per cent. Apparently, re-emigrants were more successful in acquiring a job than were other return migrants. It should be, however, recalled that they were employed abroad and it is possible that the decision of re-emigration was related to a job offer abroad.

Every third adult return migrant remained economically inactive. The main reasons for inactivity differed importantly between re-emigrants and the other returnees. For those who stayed in Poland until 2002, the share of individuals receiving old-age and/or disability pensions almost equalled the share of people supported by other household members (47 and 44 per cent, respectively). Among re-emigrants, the percentage of retirees and disabled persons receiving benefits was visibly lower, equalling 23 per cent, whereas a predominance of individuals supported by other household members was noted (68 per cent). This is in line with results of earlier studies on the transition period return wave, which demonstrated that returns for retirement constituted a visible category among returnees (cf. Iglicka 2002). Apparently, returnees undertaking re-emigration were less likely to belong to this category.

Occupations performed by return migrants included service workers (43 per cent), white-collar workers (28 per cent), qualified (16 per cent) and unqualified (12 per cent) blue-collar workers. Moreover, the largest groups of returnees were experts (31 per cent), managers and high administrative officials (16 per cent). It is worth underlining that among permanent return migrants with tertiary education the representation of individuals belonging to the above occupational categories was exceptionally high – 89 per cent of this group. In general, returnees of the transition period frequently possessed valuable skills and work experience gained in Western countries, which were in demand in the transforming Polish economy. They were teachers, managers, medical doctors, specialists in new types of services and domains in Poland: marketing, commerce, finances, IT.

This review of the main socio-demographic and economic characteristics of return migrants of the transition period demonstrates that they constituted a unique group in terms of high human capital. A high representation of dual citizens among returnees should be also emphasised and regarded as a Poland-specific phenomenon in the context of international emigration and return migration. It is also evident that some important differences between re-emigrants and return migrants who stayed in Poland until 2002 can be pointed out. They relate first of all to the level of human capital, possessed nationality, economic activity of migrants and their family situation.

Results of the logistic regression model

The estimated model proved the statistical significance of certain characteristics of migrants for the probability that they remained in Poland until mid-2002. The hypothesis that all independent variables included in the model are jointly insignificant was rejected. All independent variables were significant at the p = 0.1 level with the exception of gender and one category of education – primary or unfinished (see Table 1). We decided, however, to keep gender in the model as a basic socio-demographic characteristic, especially because its removal would not significantly improve the model fit.8 The presented interpretations concerns odds ratios (Exp(β) in Table 1). In the case of the interaction terms, odds ratios products9 describing the increase or decrease in probability of permanent return migration are demonstrated in relevant figures (Figure 5 and 6).

As demonstrated by the results of our model, the propensity to remain in Poland after return was much higher among returnees who came to Poland between 2000 and mid-2002 when compared to return migrants who arrived in the 1990. In fact, the earlier the period of return, the lower the probability of remaining in Poland until mid-2002.

Table 1. Results of logistic regression with an independent variable: reemigrant = 1 if stayed in Poland in 1989-2002, 0 if left Poland by 2002

a *** p<0.01, ** p<0.05, * p<0.1

b Country borders as of 2002.

Source: own elaboration based on the 2002 census.

The earliest returns took place when, on the one side, the situation in Poland was very unstable, both in economic and political terms, but on the other hand, the demand for Western skills and earning opportunities for people possessing such skills were enormous. Consequently, Polish emigrants and people of Polish origin who came to Poland at that time were likely to constitute a specific group. They might have been persons capable of accepting high risk in return for high profits or experts sent by their companies to set up Polish branches of their companies operating in the country of destination (cf. Górny, Osipoviĉ 2006). As suggested by the census data, some part of this group might have left Poland when they completed their tasks in Poland and when the situation became more stable and job opportunities less attractive, as compared to the early stages of the economic transition. Moreover, it can be assumed that some group of returnees simply did not adapt to the specific environment in Poland at that time. In this realm, lack of services, to which return migrants and their families had got used to in Western countries, was of significance (ibidem).

The above interpretations are in line with the fact that duration of stay in Poland constituted a powerful predictor of permanent stay in Poland after return. Each year spent in Poland increased the probability of stay until mid-2002 by more than 300 per cent. This is an intuitive outcome, since the longer a return migrant stays in Poland, the stronger the economic and social links he/she is likely to develop with Poland, while his/her links with country of the former emigration weaken.

Taking into account that the 2002 census was conducted in the pre-accession period, when Polish citizens still did not enjoy the freedom of mobility and work in the European Union countries, it is not surprising that having dual nationality was strongly related to the propensity towards permanent return migration and re-emigration after return to Poland. When compared to Polish citizens born in Poland, dual citizens tended to be more likely to re-emigrate, though this tendency was not uniform for all dual citizens. It applied particularly to Polish-German citizens who were 65 per cent less likely to remain in Poland until mid-2002 after a return, notwithstanding their place of birth – Poland or abroad (odds ratios products equal to 0.35, Figure 5) – than Polish citizens born in Poland.

Figure 5. Propensity to permanent return migration by place of birth and type of citizenship (odds ratios products) – reference group: persons born in Poland with single, Polish citizenship

Source: own calculations based on the 2002 census.

In our view, such a high propensity towards re-emigration among Polish-German citizens should be linked to results of other studies devoted to this group, which demonstrated its high involvement in circular mobility between Poland and Germany (cf. Jończy 2003). In other words, we argue that, in the case of this group, we are dealing not only with a propensity to re-emigration but specifically with eagerness to shuttle between Poland and Germany. This supposition is also in line with results obtained by Constant and Zimmerman (2011), who, though not studying Polish emigrants, identified dual citizenship (German and other) as a significant predictor of increased propensity of migrants in Germany to pursue circular migration. The outcome that place of birth did not differentiate propensity of Polish-German citizens to remain in Poland and re-emigrate most likely stems from the fact that acquisition of German citizenship by Polish citizens has not always been conditioned by the necessity to be born or even to stay in Germany, but by proof of German origins.10

In the group of dual citizens possessing citizenships of other countries, the probability of permanent return was higher than for Polish-German citizens and differed for those born in Poland and abroad. From among dual citizens possessing non-German foreign citizenship, migrants born in Poland were less likely to remain in Poland until mid-2002 than those born abroad. When compared to Polish citizens born in Poland, dual citizens with non-German citizenship born in Poland were 28 per cent (odds ratio product = 0.72, Figure 5) less likely to decide on a permanent return to Poland. Those born abroad were ‘only’ 19 per cent less likely to remain for longer in Poland than the reference group of Polish citizens born in Poland (odds ratio product = 0.81, Figure 5). Apparently, among dual citizens with non-German citizenship, descendants of Polish emigrants were more likely to pursue permanent returns to Poland than were former emigrants with two nationalities.

This is in line with the observation about return migrants with single Polish citizenship, among whom migrants born abroad were over twofold more likely to remain in Poland until mid-2002 than were Poles born in Poland (odds ratio product = 2.35, Figure 5). In fact, this group was the least likely from among all analysed sub-categories to re-emigrate after return to Poland. Possibly, such people spent too short a time in their destination country to acquire a foreign citizenship and thus their links to the country of emigration tended to be weaker than links of those who were naturalised in their destination countries. Moreover, the group of Polish nationals born abroad also includes, though not in large numbers, Polish repatriates who usually came to Poland for good. 

Gender turned out to be insignificant in predicting the propensity of return migrants to remain permanently in Poland, but other socio-demographic characteristics mattered. With regard to the age at the moment of return to Poland, the oldest group of return migrants – of 50 years and older – was the most likely to remain in Poland for longer – over two times more likely than people of ages 20-29.  This observation should be linked to the phenomenon of returns for retirement that has been observed in other studies on return migration of the transition period (cf. Iglicka 2002) and studies in other countries as well (cf. Cerase 1974). The presented model supports the observation that such returns tend to be rather definitive.

Another group of return migrants relatively eager to stay in Poland for longer were people of ages 30-39. They were 58 per cent more likely to be present in Poland during the census than were younger adult returnees, whereas older return migrants, in the 40-49 age range, were 37 per cent more likely to stay than were the youngest reference category. This can be linked to the fact that return migration to Poland during the transition period was a particularly attractive option for those who possessed some valuable skills and professional experience gained in the West. Consequently, older migrants, who had managed to acquire such experience abroad, could have been particularly efficient in adapting to the Polish labour market. Our model suggests that return migrants in their thirties were particularly successful in this realm, if we assume that acquiring a satisfactory position on the Polish labour market often constituted a precondition of further stay in Poland, as claimed by some studies on return moves during the transition period (cf. Górny, Osipoviĉ 2006).

Family status and propensity to remain in Poland after return are also linked. People living in Poland in one-person households or with unrelated housemates were the most likely to engage in re-emigration after return to Poland. Conversely, families with children were the most likely to stay in Poland until mid-2002: 67 per cent more likely than the group not possessing family households. This is in line with the general pattern that family reduces mobility, as a partner’s and children’s activities and preferences must be considered in the decisions about further migrations. 

Interestingly, couples without children and people from single parent households had a very similar propensity to remain in Poland for longer. Both groups were over 40 per cent more likely to remain in Poland until mid-2002 than the reference group. This is a counter-intuitive outcome, since couples without children are usually more mobile than single parents who have to either migrate with dependants (which is usually more complicated and less profitable) or secure some care for children left at home. It is, however, possible that some households of single parents had been separated due to return migration and that some single parents could have maintained links with their former partners who remained in a foreign country. Such a situation can stimulate re-emigration or circulation even though undertaking international mobility is complicated due to family reasons.

As demonstrated in the previous section, returnees primarily focused on urban areas in Poland, where employment opportunities for Polish emigrants were usually more attractive. On the basis of the econometric model, we examined this aspect in relation to the level of returnees’ education. In each education category, returnees residing in urban areas were less likely to re-emigrate than were returnees in rural areas (Figure 6). The group least inclined to leave Poland after return were highly educated return migrants residing in cities. This confirms the earlier observations that returning migrants’ high skills were in demand in Poland during the transition period and well-educated returnees were capable of taking advantage of attractive job prospects in Polish cities, which apparently restrained them from re-emigration.

Figure 6. Propensity to permanent return migration by place of residence in Poland and level of education (odds ratios products) – reference group: persons with University degree living in urban areasa

a Cell marked with light blue indicates insignificant category.

Source: own calculations based on the 2002 census.

Interestingly, as opposed to urban dwellers, rural residents with university educations had a relatively low propensity to remain in Poland. They were 50 per cent (odds rations product = 0.5, Figure 6) less likely to remain in Poland than were highly educated migrants residing in cities. The inclination to re-emigrate was even higher for individuals with secondary (without certificate of completion) or vocational education residing in the countryside, where persons with primary or lower levels of education were less likely to leave Poland.

A similar pattern was observed in urban areas: the propensity to leave Poland was the highest for those with a vocational level of education, relatively lower for those with secondary education, and the lowest for returnees with higher education. Such outcomes should not be surprising given crucial structural changes in Polish industry in the period of post-communist transition, notably the reduction of employment and decrease in investments (Bałtowski, Miszewski 2007). Return migrants with high qualifications were in a privileged situation, as the labour demand for experts and managers was pronounced from the very beginning of the economic transition. In this context, it can be tentatively assumed that job opportunities for returnees possessing average qualifications, in particular vocational and secondary levels of education, that were available in Polish urban and rural areas in the 1990s and at the beginning of 2000s were less attractive than professional possibilities they had abroad or while engaging in circular mobility.

Conclusions

Results pertaining to return migration to Poland during the transition period provide some valuable observations as regards mechanisms underlying permanency and temporariness of return migration. Nevertheless, the specificity of this returning flow should be acknowledged, given the specific political and economic context in Poland in the 1990s and at the beginning of 2000s. As shown by the 2002 census, the return migration of the transition period was not large in size, but diversified. It included middle-aged persons who, in the communist period, had emigrated to Western countries where they gained professional experience, later in demand while Poland was undergoing economic transition. It also included children of emigrants, often born or raised abroad. Last but not least, elderly retired return migrants who decided to spend the period of their retirement in Poland were also present in this wave.

Important differences were noted between migrants who decided on permanent returns to Poland and those who had departed by the time of the 2002 census. First of all, re-emigrants more often held dual citizenship, in particular Polish-German citizenship, which facilitated international mobility in the 1990s and at the beginning of 2000s. According to the regression model, the fact of having dual nationality was one of the most important predictors of re-emigration from Poland. At the same time, descendants of Polish emigrants born abroad were particularly likely to remain in Poland for longer periods.  

While gender constitutes an important characteristic differentiating behaviours of migrants in studies on post-accession return migration (cf. Bieńkowska et al. 2010a, b; Szymańska et al. 2012), it was  insignificant in the estimated regression model for the transition period return migrants. The latter can be linked to the fact that return migration in the 1990s and 2000s was mainly a migration of families. Re-emigrants were more often singles or not related to other members of their households in Poland, whereas migrants who decided to stay in Poland more often remained in families consisting of couples with or without children.

The youngest return migrants were particularly likely to undertake re-emigration. Similarly, studies on post-accession return migrants also revealed the particular propensity of young people to undertake a subsequent migration (Bieńkowska et al. 2010a, b; Szymańska et al. 2012). What also became apparent in the econometric model was that persons aged 50 and more, regardless of family status, were less inclined to continue emigration than were younger persons, which might be partially linked to returns for retirement. Also, return migrants in their thirties at the moment of their return to Poland were inclined to remain in Poland for longer, presumably taking advantage of their human capital, acquired in the West.

Those who stayed in Poland until 2002 were particularly well educated, whereas those who re-emigrated more often had secondary or vocational levels of education. Regardless of level of education, though, the propensity towards re-emigration was higher for rural than urban dwellers. If we assume that the level of education approximates human capital, such a result indicates that migrants deciding for longer stays or settling in Poland might have been capable of making use of their professional experience earned abroad to occupy top positions in the Polish labour market. The Polish economy, which was at the time undergoing important structural changes in industry and the service sector, was rather unfavourable for people with secondary and vocational levels of education, and such migrants were more prone to leave Poland again. It is worth noting that observations pertaining to post-accession emigration and return migration form a different picture, since highly educated migrants are less likely to return to Poland (cf. Bieńkowska et al. 2010a, b; Szymańska et al. 2012).

In general, our results are in line with studies on return migration to other countries, though they provide some additional insights into the importance of human capital and family attachments in selectivity of return migrants. As regards the interrelation between family attachments and propensity to re-emigrate, our study demonstrates how home and destination country perspectives can lead to apparently contradictory results. According to Constant and Zimmerman (2011), family attachments in the home country are conducive to circular migration. According to our results, return migrants with stronger family attachments in Poland are less likely to engage in re-emigration. When combined, the results of these two studies suggest that repetitive migration is more likely to occur when some family attachments in the home country exist, but they should not be too strong or too involving. Otherwise, migrants tend to choose to stay in the home country.

As in other European studies, migrants with high human capital were particularly prone to return to Poland for longer (Dustman, Kirchkamp 2002). Moreover, in the case of the transition period wave of returns, the claim of Dustman and Weiss (2007) that not only the level but also the degree of transferability of human capital accumulated during emigration constitutes an important determinant of propensity to return is particularly relevant. Western professional experience of return migrants was not only transferable but also in high demand in Poland in the 1990s and 2000s. At the same time, the observation of Constant and Zimmerman (2011) that better educated migrants and those less attached to the labour market of the destination country are less eager to be involved in circular migration was also reflected in our results, though seen from the home country perspective. Polish re-emigrants tended to be less educated and had worse job opportunities in the home labour market than did permanent returnees. It can be thus generalised that temporary migration is an attractive option for those who are not well anchored in the labour market of some country – be it their home or destination country.

Notwithstanding the universalism of our observations, when compared to other studies, it should be remembered that the post-communist wave to Poland had a specific nature, not only for the high transferability of migrant’s human capital, accumulated abroad. In our view, this specificity is well portrayed by other two observations. First, at the beginning of the post-communist transition, 1989-1990, the return migration was accompanied by an elevated and prompt re-emigration that often took place in the same year or the year following the return. Apparently, some return migrants spent only the very first years of economic transition in Poland, when return on their human capital was particularly high. Second, dual citizens were largely overrepresented in the re-emigration of the 1990s and 2000s, mostly because they enjoyed a freedom of mobility unachievable to Polish citizens before the accession of Poland to the European Union.

Acknowledgements

The authors would like to thank the anonymous referees for their valuable comments and suggestions.

Endnotes

1 www.stats.oecd.org

2 For Poland and Romania, shorter periods are taken into consideration: 2004-2007 and 2005-2007, respectively. For Lithuania the year 2001 is also taken into consideration.

3 In the Tables and Figures sections to follow, for simplicity of discussion, we talk about return migrants who came in 1989-2002.

4 Such a period was taken into consideration in the 2002 census.

5 We decided to include a categorical variable instead of a continuous variable since we expected likelihood of permanent return to be linked to the stage in the life cycle of a person, and not to age as such. In our view, a categorical variable best captures this issue. Moreover, as it turned out, characteristics of the models with categorical and continuous variables differed very little. 

6 This aspect was operationalised as a categorical variable since we believe that the transition period was not homogenous with regard to opportunities for emigrants and return migrants, and it was justified to identify distinct periods of returns.

7 But this is only a supposition, since the 2002 census did not include information on the place of residence in Poland before emigration.

8 For the model without gender, AIC = 2193.292.

9 The odds ratios products were calculated by multiplication of odds ratios referring to variables included in the interaction and to the relevant interaction term.

10 Policy related to this procedure changed over time, becoming more restrictive in this realm, but many Polish citizens managed to acquire German citizenship only on basis of German origins (cf. Heffner, Solga 2007).

References

Anacka M., Fihel A. (2012a). Return migration to Poland in the post-accession period, in: B. Galgoczi, J. Leschke, A.Watt (eds), Migration and Labour Markets in Troubled Times: Skills Mismatch, Return Migration and Policy Responses, pp. 143-168. Aldershot: Ashgate.

Anacka M., Fihel A. (2012b). Selektywność emigracji i migracji powrotnych Polaków – o procesie „wypłukiwania”. Central and Eastern European Migration Review 1(1): 57-68.

Bade K. (2000). Europa in Bewegung. Migration vom spleen 18. Jahrhundert bis zur Gegenwart. Munich: C. H. Beck Verlag.

Bałtowski M., Miszewski M. (2007). Transformacja gospodarcza w Polsce. Warsaw: Wydawnictwo Naukowe PWN.

Bieńkowska D., Ulasiński C., Szymańska J. (2010a). Kierunek Małopolska. Charakterystyka powracających i rekomendacje, jak nie zmarnować ich kapitału. Kraków: Centrum Doradztwa Strategicznego.

Bieńkowska D., Ulasiński C., Szymańska J. (2010b). Warto wracać? Strategie zachowań reemigrantów i rozwiązania służące wykorzystaniu ich potencjału. Kraków: Centrum Doradztwa Strategicznego.

Bolognani M. (2007). The Myth of Return: Dismissal, Survival or Revival? A Bradford Example of Transnationalism as a Political Instrument. Journal of Ethnic and Migration Studies 1(33): 59-76.

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

Cerase F. (1974). Nostalgia or Disenchantment: Considerations on Return Migration, in: S. M. Tomassi, M. H. Engel (eds), The Italian experience in the United States, pp. 217-239. New York: Center for Migration Studies.

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

Constant A. F., Zimmermann K. F. (2011). Circular and Repeat Migrations: Counts of Exits and Years Away from the Host Country. Population Research and Policy Review 4(30): 495-515.

Dustman Ch., Kirchkamp O. (2002). The optimal duration and activity choice after re-migration. Journal of Development Economics 67(2): 351-372.

Dustman Ch., Weiss Y. (2007). Return Migration: Theory and Empirical Evidence for the UK. British Journal of Industrial Relations 2(45): 236-256.

Fihel A., Górny A., Matejko E. (2006). Remigracja a transfer kapitału ludzkiego do Polski w okresie transformacji, in: E. Jaźwińska (ed.), Imigracja do Polski w świetle wyników Narodowego Spisu Powszechnego 2002, pp. 29-41. CMR Working Papers 13/71. Warsaw: Centre of Migration Research, University of Warsaw.

Ganga D. (2006). From Potential Returnees into Settlers: Nottingham’s Older Italians. Journal of Ethnic and Migration Studies 8(32): 1395-1413.

Gawryszewski A. (2005). Ludność Polski w XX wieku. Warsaw: Instytut Geografii i Przestrzennego Zagospodarowania PAN.

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

Gosh B. (ed.) (2000). Return Migration. Journey of Hope or Despair? Geneva: International Organisation for Migration, UN.

Górny A., Kolankiewicz G. (2002). Reemigracja z Wielkiej Brytanii – poszukiwanie ‘nowego’ czy kontynuacja ‘starego’?, in: K. Iglicka (ed.) Migracje powrotne Polaków. Powroty sukcesu czy rozczarowania, pp. 122-160. Warsaw: Institute of Public Affairs.

Górny A., Osipoviĉ D. (2006). Return migration of second-generation British Poles. CMR Working Papers 6/64. Warsaw: Centre of Migration Research, University of Warsaw.

Górny A., Pudzianowska D. (2009). Same letter, new spirit? Nationality regulations and their implementation in Poland, in: R. R. Bauböck, B. Perchinig, W. Sievers (eds), Citizenship policies in the new Europe, pp. 123-176. Amsterdam: Amsterdam Univeristy Press.

Greene H. (2003). Econometric Analysis, 6th Edition. Prentice Hall, New Jersey.

Heffner K., Solga B. (2007). Podwójne obywatelstwo na Śląsku Opolskim, in: A. Górny, P. Koryś (eds), Obywatelstwo wielokrotnego wyboru. Interdyscyplinarne ujęcie kwestii podwójnego obywatelstwa w Polsce. Warsaw: Wydawnictwo Uniwersytetu Warszawskiego.

Heffner K., Sołdra-Gwiżdż T. (1997). Migracje powrotne na Górny Śląsk z socjologicznej perspektywy. CMR Working Papers 9. Warsaw: Centre of Migration Research, University of Warsaw.

Iglicka K. (2002). Migracje powrotne w świetle najnowszych ujęć teoretycznych. Źródła informacji i podejście badawcze, in: K. Iglicka (ed.) Migracje powrotne Polaków. Powroty sukcesu czy rozczarowania, pp. 17-22. Warsaw: Institute of Public Affairs.

Iglicka K. (2010). Powroty Polaków po 2004 roku. W pętli pułapki migracji. Warsaw: Wydawnictwo Naukowe Scholar.

Jończy R. (2003). Migracje zarobkowe ludności autochtonicznej z województwa opolskiego. Studium ekonomicznych determinant i konsekwencji. Studia i Monografie 264. Opole: Uniwersytet Opolski.

Klagge B., Klein-Hitpaß K., Fihel A., Kindler M., Matejko E., Okólski M. (2007). High-skilled return migration and knowledge-based economic development in regional perspective. Conceptual  considerations and the example of Poland. CMR Working Papers 19/77. Warsaw: Centre of Migration Research, University of Warsaw.

Kołodziej E. (1998). Emigracja z ziem polskich od końca XIX wieku do czasów współczesnych i tworzenie się skupisk polonijnych, in: A. Koseski (ed.), Emigracja z ziem polskich w XX wieku. Drogi awansu emigrantów. Pułtusk: Wyższa Szkoła Humanistyczna.

Koryś I. (2002). ‘W pół drogi’ – migracje powrotne z Niemiec, in: K. Iglicka (ed.) Migracje powrotne Polaków. Powroty sukcesu czy rozczarowania, pp. 161-183. Warsaw: Institute of Public Affairs.

Martin R., Radu D. (2012). Return Migration: The Experience of Eastern Europe. International Migration 6(50): 109-128.

Matejko E. (2010). Reintegracja poakcesyjnych migrantów powrotnych na polskim rynku pracy – wyniki analizy jakościowej, in: I. Grabowska-Lusińska (ed.), Poakcesyjne powroty Polaków, pp. 34-47.  CMR Working Papers 43/101. Warsaw: Centre of Migration Research, University of Warsaw,.

Slany K., Małek A. (2002). Reemigracje z USA do Polski w okresie transformacji ustrojowej, in: K. Iglicka (ed.) Migracje powrotne Polaków. Powroty sukcesu czy rozczarowania, pp. 78-120. Warsaw: Institute of Public Affairs.

Szymańska J., Ulasiński C., Bieńkowska D. (eds) (2012). Zaraz wracam... albo i nie. Skala powrotów, motywacje i strategie życiowe reemigrantów z województwa śląskiego. Kraków: Centrum Doradztwa Strategicznego.