2012_3-4_Melegh

Attila Melegh

Net Migration and Historical Development in Southeastern Europe since 1950*

 

* The study is a background paper for the following project: SEEMIG Managing Migration and its Effects in South-East Europe – Transnational Actions Towards Evidence Based Strategies. The project is funded under the third call for proposals of the South-East Europe Programme. The information published here reflects the author’s views and the Managing Authority is not liable for any use that may be made of the information concerned. The writing of this study was supported by the Institute of Advanced Studies at CEU. Special thanks to Szabrina Csánó, who helped in the construction of the database. Also thanks to József Böröcz and Arland Thornton for valuable comments. And special thanks to Márta Kardulesz and Ágnes Anek for their help in editing the paper.

 

This essay formulates some basic developmental patterns in Southeastern Europe (focusing on the area between Italy and the Caspian See) on the basis of some longer term macro statistics on net migration and other macro statistical time series. It demonstrates that in furthering an understanding of longer term developmental patterns, the world system approach is helpful in a modified form. In the case of state socialist economies the direct intervention of world capitalism had a long lasting impact on the migratory links of the countries within the region. Countries that were unable to counterbalance the collapse of local industry became sending countries and were partially re-ruralized and partially pushed into large scale emigration. The analysis lends credence to the neoclassic macro-economic theory of migration, but its validity with regard to per capita GDP differentials is strengthened if it is linked to positions in global hierarchies. The key point is that it is not simply GDP differentials that matter, but rather positions within the global economy, which themselves are in part the results of historical processes and linkages. In addition, people seem to have clear ideas of developmental scales which correspond quite accurately to actual per capita GDP figures. Thus people may well be aware of global inequalities and may even have clear ideas of complex sequences that might orientate them in their decisions regarding migration.

keywords: net migration, macro statistics, state socialist economies, global economy, patterns of migration, global history

 

Introduction

Migratory links and channels form a web around the world. As in the case of the exchange of ideas, images, capital, goods and services, countries and regions are integrated into a global flow of people.1 Concerning spatial units (countries and/or regions) it is important to consider various modes of integration, since as with the global flow of capital, there are various patterns according to which regions and countries are integrated into the global flows and systems of flows. None of the countries is completely isolated, and there is no country or region to or from which migrants simply embark or depart, as most regions and countries produce both emigration and immigration at the same time. It is also widely claimed that in areas and countries in which emigration has dominated there is a gradual move towards a pattern of receiving more immigrants than losing emigrants.2 Also more and more countries are moving into a middling or transitional position, including the North African and many Latin American countries. Little research or theoretical work has been done on the question of how these complex modes of integration develop historically. Migration flows are related to other social processes, which makes analysis difficult, but, more importantly, the analytical focus has been too narrow to further any subtle grasp of how the patterns of interrelated processes have changed in human history according to positions in a global system. There is a need to re-contextualize historically and regionally all of the major theories of migration that emerged over the course of the last three decades.3

Theoretical Problems

Classical and neoclassical macro and micro theory seeks to discern mechanisms based on wage differentials and labor market processes without a historical understanding of the developmental perspective. Structural-historical and world system theories have arrived at the clear premise according to which transition from rural to non-rural economies and the intrusion of world capitalism create a scenario for massive emigration. From the theoretical perspective of intervention and the break-up of “traditional” systems, scholars of this intellectual approach also argue that colonial or historically established links matter, but they have no systematic idea of how longer term changes happen beyond the specific periods leading to massive social transformation or establishing specific links.4 This is exemplified by the following summary by Douglas Massey:

International migration originates in the social economic, cultural and political transformations that accompany the penetration of capitalist markets into non-market and pre-market societies (as hypothesized under world system theory). In the context of a globalizing economy, the entry of markets and capital-intensive production technologies into peripheral regions disrupts existing social and economic arrangements and brings about a displacement of people from customary livelihoods, creating a mobile population of workers who actively search for new ways of achieving economic sustenance.5

Network theory and cumulative causation are also relevant to an understanding of historical change, as they help explain why and how established migration flows continue. Nonetheless, they are not adequate as explanations of why such flows might dry out or become less intensive, nor for that matter do they shed much light on how these flows can become cyclical. Furthermore, these theories offer little insight into the ways in which transitional or intermediary countries are integrated into the global flow and how this integration might change.

Concerning longer term and more empirical approaches to the question of how migratory integration of countries and regions varies over time, we have only a few hypotheses. One is the idea of migration transition, which was developed by the geographer Zelinsky, who modeled the idea of demographic transition as established in the 1930s in the United States and Europe.6 Zelinsky argues that gradually, following an increase in emigration, because of socio-historical processes countries of large-scale emigration become countries of net immigration within the framework of a fairly linear development.

This theory is related to migration hump or migration curve theory, according to which over time and with increasing income levels countries may move from increasing to decreasing flows of emigration and then to immigration.7 In other words, upon reaching a certain level of economic wealth, countries produce more migrants as the migrants or potential migrants are actually able to finance and organize a move to better-off countries, while an increase of wealth actually reduces the incentive for massive emigration. This is a non-linear idea of progress and may serve as an interesting starting point, but this theory also focuses on one transition and lacks a complex approach to migratory integration that would combine not only wealth differentials, but also related historical processes of economic integration into the world economy. Moreover, this premise regarding the gradual move toward immigrant status is actually not true with regard to many countries, as there can also be reverse processes, as we will see below.

Debates on migration and development focus on the analysis of a complex interrelationship between migration and developmental processes, but generally the temporal perspective is rather limited and/or the discussion remains on a rather superficial level, listing several factors and mechanisms without actually measuring and systematically demonstrating the mechanisms and the importance of various factors.8 This is undoubtedly a consequence of the lack of appropriate and comparable statistics and actual data, but a more systematic historical analysis is still missing.9

It is also worth mentioning that there are some descriptive analyses on the history of migration in the last century, but while they may be very informative and sometimes brilliant, they are either very specific in time and analysis or actually rather broad and fail to give a systematic analysis of how countries are integrated into a global flow of people and global processes of development.10 In addition, in the history of migration most analysts stress the importance of political events, but fail to consider the role of other relevant social processes. This is especially true when countries representing varying political systems are included in an analysis of long term change.

Methodological Remarks

In this essay I identify some basic developmental patterns in Southeastern Europe on the basis of some longer term macro statistics provided by the United Nations (UN) World Population Prospects (WPP) website.11 I focus on net migration as estimated by the UN as a residual of population growth minus natural growth. This is a problematic source, as it incorporates the problems of population enumeration as well, but there are no other comparable sources available for the period in question.

It is worth citing various authors who have published findings in the recent Prominstat project reviewing various data systems, including migration flows. They have arrived at conclusions such as the following:

In the study, we have presented a detailed analysis of the availability, reliability and comparability of data on international migration flows in 27 European countries (all EU Members States except Bulgaria and Romania, plus Norway and Switzerland). Our conclusion is that internationally comparative research on migration flows in Europe is currently generally not possible. The main problem is the comparability of data, in particular the differences in definitions and sources used in various countries and in the coverage of the statistics. These differences imply that comparing migration flows in various countries would be often like comparing pears and apples.12

Furthermore net migration rates hide whether countries in which similar levels and the same overall direction (positive or negative) of net migration prevail actually have the same levels of outflow and inflow. Thus a country with a net migration rate of negative five people per 1,000 inhabitants could be a country with zero immigration and rate of five in outmigration, but it could also be a country into which there is large-scale immigration, but this rate of immigration is surpassed by the emigration rate by five people per 1,000 inhabitants. This remains hidden, and this lack of information is a significant problem that needs to be addressed through the collection of more information on the actual rates of emigration and immigration.

Nonetheless, the rate of net migration can be a very useful measurement if one looks at the data systematically. With reference to possible methodological problems, it can be understood as an overall sum of “personal” levels of integration into global flows of people, and this actually avoids some of the pitfalls of migration statistics in terms of definitions and the actual underestimation of immigrants and more importantly of emigrants.13 Altogether, change will be assumed when the figure for a country in which there is a negative, positive or zero rate of net migration shifts in terms of scale or direction.

In the analysis additional longer term statistics on GDP and other economic and labor market indicators will also be used coming from various sources, such as the World Bank, International Labor Organization (ILO) or local statistics. Regarding per capita GDP figures, I follow Böröcz14 when looking at changes such as percentages of world average and evaluating historical development of various regions and countries accordingly. Here I do not use his ideas concerning global weight, regardless of the fact that in the case of migration population and economic size matters.

In this essay I focus on the area between Italy and the Caspian Sea. I identify subregions in an inductive manner on the basis of changes in net migration. Nonetheless, I capitalize on the insights of historians like Wallerstein and Berend, according to which Southern and Eastern Europe have something in common if longer term historical processes are analyzed. This approach is based on the premise that these countries were integrated into global-colonial capitalism in a rather similar manner, especially during the nineteenth and the early twentieth centuries.15

Generally this regional linkage is forgotten when state socialism as a rivaling form of modernity appeared in the late 1940s, and there is an overdue emphasis on political changes and factors. The period of state socialism is either ignored or it is seen as a “frozen” moment as far as longer term regional patterns of migration are concerned. In my view we need to go back to proper historical comparative social and economic analysis without inbuilt teleological assumptions, as such an analysis provides a better perspective from which to understand migratory changes in the region in question. This is true for the period between the 1950s and 1960s and the so-called transitional period between 1988 and 1995.

As mentioned above, the idea of the region below is an inductive one. This is true in the sense that at the moment I disregard ideas of historical regions such as the “Balkans,” the “Eastern Block,” or “Mediterranean” territories. I do this not because I find these ideas useless or lacking validity from the perspective of many aspects of historical change, but rather because one needs to be more open in dividing and linking these regions when social processes such as migration are analyzed.

Net Migration and Historical Development in Southeastern Europe between 1950 and 2010

Global Changes in Net Migration and Europe

If one looks at larger regions, one notes that the larger areas of the world are sending regions, while North America and Europe are, overall, the ones that receive migrants on a cross-continental level.

In the beginning of the period in question, Europe was a sending region, and it became a net immigrant area in the early and mid-1960s. Other regions, such as Asia and Latin America, moved from a zero rate of net migration to negative levels, then to a level of less than one person per one thousand. Africa has been always on the negative side, while North America has always been positive.

Figure 1. Net migration rate by larger regions, 1950–2010

Source: World Population Prospects (WPP) 2010 revision

The shift that took place in Europe, from a continent in which emigration exceeded immigration to a continent in which immigration began to prevail, in all likelihood was due to several factors. One of these factors was reduced transatlantic migration, which never returned to its pre-1920 levels.16 The other was the dramatic transformation and the decline of large-scale rural systems in Europe, especially in areas like Southern and Eastern Europe, where various efforts were made to solve an agrarian crisis and the problems emerging due to large landed estates and to strengthen the competitiveness of agrarian economies. The key point from the perspective of migration history was that these rural societies lost people on a dramatic scale, and actually the 1950s and 1960s was the turning point when rural production and rural producers became a minority in Europe and in many other areas of the world.17 This meant large-scale migration to cities and, as a related process, intra-European and intercontinental geographic mobility. Another factor was the final collapse of the European colonial system, because of which until the 1950s colonized areas had been major recipients of emigrant populations coming from Europe. It is also important to note that while the colonial system existed, the arrival of various local groups from the colonies was seen as negative, preferably obviated by the arrival of immigrants from other “European” populations, even when there was a dire need for laborers.18 Thus until the collapse of the colonial systems there was no real counter flow of migrants, and in the 1950s and early 1960s only colonies that had liberated themselves from colonizers sent larger groups of migrants to Europe.19 The other major factor was related to the fact that many of the European countries had industries that were in need of migrant workers and, in addition to the desperate search for much promoted “European” sources of labor, programs were started in the 1960s to attract immigrants from Algeria, Morocco and Turkey. This has been widely demonstrated and widely theorized.20

After 1980, Europe surpassed the plus 0.1 percent level of net migration, and by the first decade following the turn of the millennium net immigration rates of more than plus 0.2 percent can be observed, in relation to the relevant population figures. Thus the shift that took place in the early 1980s was for Europe and North America an intensifying immigration pattern, while other regions primarily figured as sources of emigrants. There is a clear link here to the new cycle of globalization after 1980, a new cycle of openness that increased the relative loss or gain of the population on behalf of the major regions.21

Thus altogether a pattern came to prevail in Europe as the continent evolved from the status of a source of emigrants to a new home for immigrants. It never reached the levels of North America, but a relatively small proportion of emigrants from Asia, Africa, or Latin America came to work or settle in Europe. I now turn to Southeastern Europe, a region of the continent which before the Second World War was a major source of migrants in migratory links beyond and within Europe.

Types of Development in Southeastern Europe

Southeastern Europe has shown increasing diversification of net migration rates over the course of the past sixty years. In the 1950s it was more or less homogenously a net emigrant region (with the exception of countries in the south west of the Soviet Union). After changes that took place between the 1960s and 1990s, it lost this homogeneity and some parts became immigrant areas, while others became or remained emigrant areas.

 

Figure 2. Countries analyzed according to developmental types

Type one on this map is the region that includes Southern European countries such as Italy and Greece, but also countries of Central Europe. Type two is comprised of the countries of the so-called Balkans, while type three contains areas that once were the south western edge of the former Soviet Union around the Black Sea. Type four covers major areas of the former Yugoslavia, but as will become apparent this type merits further analysis and can be included in the region of the Balkans.

One can identify four types of developmental patterns that are related to relative wealth and processes in the economic and employment structures. These patterns reveal distinct trajectories of development based on macro figures. The four types can be summarized as countries:

• that were emigrant countries in the 1950s and the 1960s and then gradually became immigrant countries (type one),
• that remained emigrant countries throughout the period (type two),
• that were immigrant countries and then became emigrant countries (type three),
• that oscillated between emigrant and immigrant status (type four).

 

Type One: from Emigrant to Immigrant Status

The first type is comprised of countries that had a negative migration rate in the 1950s, but where migration rates became positive parallel to the process observed when taking the entire continent into consideration. Type one contains Southern and Central European countries outside the Balkans and the post-

 

Figure 3. Type one: net migration in selected countries that became immigrant countries, 1950–2010

Source: WPP 2010 revision

Soviet countries: Italy, Greece, Slovenia, Austria, the Czech Republic, Poland,

Slovakia, and Hungary. This pattern could be easily expanded to include other major Southern European countries, such as Spain and Portugal.

This type is a linear developmental pattern that shows a strong linear regression when time and net migration are related, very much in line with migration transition theory.

 

Figure 4. Net migration over time in type one, 1950–2010

(All data points in type one, five-year intervals marked by midpoints)

Source: WPP 2010 revision

 

This type represents the overall European pattern of development. Whether state socialist or capitalist, the countries were basically sending countries in the 1950s and 1960s. Some of them were extremely open for relative large-scale outmigration, such as Greece, which experienced the outmigration of hundreds of thousands after the Greek civil war, mainly from rural areas. Some had a clear negative rate of net migration in the 1950s, such as Italy, Hungary, Slovenia and Austria (and other Southern European countries, such as Spain and Portugal). Countries like Bulgaria in type two also produced large-scale emigration in the 1950s. Beyond longer term rural crises and transformation and post-war resettlement processes,22 this dominance of the emigration pattern may show that, for instance, the well-known Hungarian exodus in 1956 was not due solely to political reasons, as has often been argued.23 Most of the people who left were young (less than twenty-five years of age), primarily skilled male workers (two thirds of them) living either in Budapest or regions of the country that traditionally had been sources of migrants leaving for Austria and/or the West.24 Many of these people would have looked for jobs in areas demanding industrial labor even without political motives, as was the case in Italy and Portugal, for example. This emphasis on social processes, however, should not be misunderstood as a dismissal of the clear relevance of political factors, such as the opening of the border.

The negative net migration rates began to approach zero in the 1970s and in some cases even became positive. Rates in Austria became positive between 1960 and 1965. Italy, Greece, and Slovenia followed in the 1970s. Other countries crossed the zero line during the 1980s, and in the case of Slovakia even as late as after the collapse of state socialism. It is also important to note that these countries actually never got out of the negative 0.5 and positive 0.5 percent range of net migration.

There is a peculiar feature of this linear migration transition in terms of net migration. Namely, concerning income gaps, many of these originally emigrant countries did not change their positions in comparison with the major target areas. For instance, the income gap between Hungary and Germany can hardly be said to have closed over the course of the last four decades of the twentieth century, nonetheless between 1954 and 1999 Hungary followed a cyclical pattern of migration flows toward Germany. These net flows (the sum of Hungarian citizens moving between Germany and Hungary) follow the change in the income gap, thus offering support for macro-economic arguments. Nevertheless, Hungary also became an immigrant country while at the same time maintained its emigrant character toward some of its main historical target areas.

 

Figure 5. Net migration flow and per capita GDP ratios between Germany and Hungary, 1954–1999

Source: Maddison databank and Statistisches Bundesamt

 

Thus we have to look for the combination of internal change in the transformations of employment structures and additional macroeconomic changes in order to explain the change of net migration in these countries on a macro level.

As theorized by the world system approach, one very important factor in this transition could be that agriculture, which was once an important element in the economic performance of these countries, declined to a very low level, a decrease which of course was followed by changes in the employment structures.25 And as a parallel process, the service sector overtook the other sectors, and in all the countries of this type this sector grew to comprise more than sixty percent of the share of the labor force. State socialist countries experienced a greater decline not only in agriculture, but importantly in industry as well. But it is important to note that in comparison with countries belonging to the other types, each of these countries was able to stabilize a larger industrial share above 30 percent

Figure 6. Some countries if type one that became immigrant countries. Share of agriculture (percent of GDP), 1960–2010

Source: World Bank, Development Indicators

of the GDP and could maintain substantial employment levels in this sector, at least for men. According to World Bank Data this share is between 40 and 50 percent, with the exception of Greece. Overall, after the collapse of state socialism, state socialist countries basically smoothed into the developmental patterns of capitalist countries within this type and region, and they experienced a one-time great loss of productive sectors beyond the slow gradual decline during the state socialist period.

A related key element may be that during the period under discussion these countries were always able to maintain a global position above the world average, and most of them actually were able to improve this positive gap relative to the global average.

Figure 7. Type one: countries that became immigrant countries. Per capita GDP as related to world averages 1950–2008

Source: Maddison databank

All state socialist countries suffered a quick and dramatic decline toward the average in the early 1990s, but they soon got back to levels above the average. This decline in income in the early 1990s, together with the relevant political changes, produced additional emigration, as noted above in the case of Hungarian emigrants departing for Germany. But overall, former state socialist countries within this type maintained a global position that calmed this wave of immigration, and macro structures allowed the move to an overall positive net migration rate. Even more importantly, with the reentry into a relatively open capitalist system (and being in the upper layer of these countries), they began to receive greater number of immigrants even within the region. Slovenia became “attractive” as a goal for immigrants from the territories of the former Soviet Union, the Czech Republic for immigrants from Vietnam and the Ukraine, and Hungary for immigrants from Romania, China, and the Ukraine. The more prosperous successor states of the previously federative countries (the Czech Republic and Slovenia) also received larger numbers of migrants from states previously within the same federative formation.26

It is worth taking a closer look at how these changes in the place of a country in a global hierarchy on the basis of per capita GDP were related to changes in net migration. One could consider the example of Greece.

 

Figure 8. Net migration rate and per capita GDP difference from world averages in Greece, 1950–2010

Source: WPP and Maddision databank

In the case of Greece the link between the two processes is very clear, and actually the change in net migration is well correlated with changes in the difference between Greek per capita GDP and the world average. Changes in income levels were soon followed by a shift in net migration. By the end of the 1970s larger groups of Greek emigrants returned home, as they found the country more stable and prosperous. It is also important that in the 1990s citizens of Albania, Bulgaria and Romania, countries that had experienced a large-scale collapse of local industries, found it more and more attractive to go to Greece. Also as of the 1980s the whole upper Mediterranean region became a target area for migrants coming from and through North Africa. In addition, in the case of Greece the border with Turkey became a central point of entry of undocumented immigrants coming from Asia.

The above processes lend considerable credence to the arguments above according to which positions in such global hierarchies do matter. But the relationship needs further investigation, as there are cases in which it is not that clear or other mechanisms can be identified. Hungary constitutes one such example.

Figure 9. Net migration rate and per capita GDP difference from the world average in Hungary, 1950–2010

Source: WPP and Maddison databank

Between the 1950s and the mid-1980s the processes that were underway in Hungary resembled those in Greece, although the country did not become an “immigrant” country as early as Greece. Even more importantly, the situation in Hungary began to differ substantially at the end of 1980s and early 1990s. At that time there was an increase in the outflow of migrants to Austria and Germany, as noted above. Thus a decline in the overall global position led to “expected” changes. But most probably due to some underestimation of outmigration and its relatively prosperity in comparison with neighboring countries with significant Hungarian speaking minorities (Romania and the Ukraine), Hungary was itself an attractive goal for immigrants, and the inflow from Romania, for instance, as a sending country of type two was larger than the increase in the outflow of citizens of Hungary. This linkage can be well demonstrated for the late 1990s and early 2000s, especially with regards to the category of labor permits, by the following graph:

 

Figure 10. Immigration from Romania to Hungary, 1995–2005

Source: Hungarian Central Statistical Office (HCSO) migration, Maddison databank

As this graph illustrates, one cannot simply look at individual countries, but must consider larger systems containing various types and dynamics of development. Surely historical links and other mechanisms of cumulative causation on a behavioral level also matter and shape processes indicated by macro positions and structures.

 

Type Two: Countries that Remained Sending Countries throughout the Period in Question

Type two countries started out like type one countries, but they have not completed any kind of transition toward net immigration. Thus we can see that Southeastern Europe has been increasingly diverse with regard to an overall mode of migratory integration and its historical trajectories.

From the perspective of overall trends, migration rates in these countries were with very few exceptions consistently negative, but within this there was a cyclical move with some extreme values in the early 1990s.

 

Figure 11. Type two: net migration in countries that remained emigrant countries, 1950–2010

Source: WPP 2010 revision

 

Figure 12. Type two: countries that remained emigrant countries. Net migration over time, 1950–2010 (All data-points in the “sending” type, five-year intervals marked by midpoints.)

Source: WPP 2010 revision

In the 1950s and early 1960s these countries were rather similar to type one countries. In other words one notes the beginnings of a transition toward immigration. However, already in the early stages some of the countries had relatively large-scale negative net migration around and beyond negative 0.5 percent. Later, during the late 1960s and early 1970s, the state socialist countries like Romania and Bulgaria seemed to follow the transition seen in type one, but this shift remained short-lived. In the same period, Turkey, the only capitalist country in type two, was experiencing intensified outmigration due to the guest worker programs promoted by Germany and Austria, countries in which by this time migration rates were positive.

A dramatic outmigration scenario emerged in the form of massive outflow during the early 1990s in the former state socialist countries, which in the case of Albania was so intense that it reached 3 percent of the total population. This extreme figure is actually the one that reduces the consistency of this type. As the only non-socialist country, Turkey represented a different pattern, and actually it began to approach zero, as had Bulgaria and Romania in the 1970s. Turkey was also able to maintain its more balanced integration into the global flow of people. Surely, in combination with other factors size also plays a role in this process, as smaller countries, especially when they are undergoing unsettling transitions or changes, can produce massive outflows.

 

Figure 13. Type two: countries that remained emigrant countries. Per capita GDP, 1950–2008

Source: Maddison databank

As the GDP figures (as related to global averages) reveal, as opposed to type one, most of the countries in this category remained consistently below the global average. The sole exception was Bulgaria, which actually did surpass the global average in the 1970s. Turkey began to approach the average in the 1960s, and ever since then it has been moving in parallel with the global average. Thus development patterns can be related to global positions and changes in these positions if one measures them according to per capita GDP.

The case of Bulgaria can be cited in clear support of our argument. Its cyclical change in net migration is paralleled with some delay by cyclical changes in per capita GDP. Overall in Bulgaria net outmigration declines when the per capita GDP approaches the global average, while outmigration rises steeply when the GDP collapses in relative terms.

 

Figure 14. Net migration and per capita GDP difference in Bulgaria, 1950–2010

Source: WPP 2010 revision and Maddison databank

 

Actually this period is related to the huge exodus of the Turkish minority.27 But it seems that again this exodus was due not simply to immediate ethno-political considerations, as was suggested in the literature and in public discussions. The target country, Turkey, rose above the world average in this period. So, very much like the case of Romanian and Hungarian migratory links, in the case of the relationship between Bulgaria and Turkey the ethnic component could be simply the behavioral link between changing macro positions and group level actions. As I will demonstrate later, there is considerable empirical evidence indicating that Bulgarians (and Albanians) see themselves as inferior in development to Turkey. So ethnic considerations, the break-up of state socialism, economic hierarchies and the collapse of per capita GDP together create scenarios in which a larger exodus may happen. Probably the same historical development took place in Romania with regard to its Hungarian and German minorities. Furthermore, it seems that as opposed to the neoclassical economic approach in migration theory, in cases of large-scale outmigration it was not the actual differential that mattered with regard to the receiving areas, but a relative position in comparison with global averages, which is not an individual level phenomenon. One notes large differentials between type one countries such as Hungary and the Czech Republic versus many other European countries, but until now these countries have not produced large-scale outmigration in these directions, while countries around and below the global average did.28 We can relate these mechanisms to world system theory, looking for explanations according to global positions.

Changes in the composition of an economy according to sectors may also offer some insight into how this sending pattern remained dominant in this group of countries. Concerning the composition of the economies since the 1980s, one notes changes similar to those that took place in type one countries, but the collapse of these economies in the state socialist and capitalist periods is sharper and had longer-term consequences. For instance, in Albania (the country that produced the greatest exodus over the course of the entire period and over the whole region) the industrial collapse was not only vertical, but actually stabilized at a very low level of around 20 percent. As opposed to type one countries, these countries hardly surpassed 30 percent of GDP with regard to industry, which shows that they were not able to “attract” enough global industrial capital even to achieve the levels of former state socialist countries in type one.

Figure 15. Type two: countries that remained emigrant countries. Industry (percent of GDP), 1960–2010

Source: World Bank Data

Actually some of these countries experienced “re-ruralization,” which was a clear sign of the collapse of the employment structure and also an indication that people were desperately looking for lower value opportunities at a time when social security had also been shattered. Even in Romania, where the share of agriculture declined from the late 1980s, male employment in agriculture increased from 25 percent to 40 percent during the 1990s in terms of total male employment. The share of agriculture in female employment was consistently high in these countries, though this rate declined during the transition period.

This shows that in the case of some former state socialist countries the intrusion of global capital led to larger scale outmigration not because of the rediscovery of a “traditional” pattern, but because it could ruin an alternative type of modern industry, somewhat defended locally as long as the state socialist framework existed.

The industrial collapse and the inability to regain the losses in the service sector of the economy that came in the wake of this collapse led to a massive and continuous exodus in countries that were not able to surpass or to remain above global average income in the region. Countries that were above world averages were able to re-strengthen industry and expand the service sector substantially, and these two sectors thus could slow down the exodus of the early 1990s. In other words, they were able to attract larger numbers of immigrants to counterbalance outmigration.

 

Figure 16. Type two: countries that remained emigrant countries. Agriculture (percent of GDP), 1960–2010

Source: World Bank Data

Type Three: Countries that Become Emigrant Countries

 

The post-Soviet countries in the south-western segments of the Soviet Union show a very different developmental pattern. Type two countries were close to type one countries, as they were all emigrant countries in the 1950s. In the 1950s and early 1960s, however, type three countries were either those countries in which there was zero net migration (Ukraine, Azerbaijan) or immigrant countries (Georgia, Moldova and Armenia) that received larger numbers of migrants from various parts of the Soviet Union, including Belarus, Russia, and the Ukraine. In the case of these countries, the scale of positive net migration was much higher than the rates ever reached on an overall European level or in “classic” immigrant countries like France. They were relatively highly developed countries in the Soviet Union which not only were the beneficiaries of investment in industry, but also had higher quality agriculture, which in the state socialist system was actually overvalued due to internal market problems.

Figure 17. Type three: net migration rates in countries that became emigrant countries, 1950–2010

Source: WPP 2010 revision

 

Figure 18. Type three: decrease. Net migration over time in countries that became emigrant countries, 1950–2010

(All data points in type three, five-year intervals marked by midpoints.)

Source: WPP 2010 revision

There is a claim widespread in history writing and especially in the historiography of ethnic groups and smaller nations according to which internal migration within the Soviet Union and in other state socialist countries was a forced process orchestrated by the political authorities. Unquestionably political authorities did have a role in geographic mobility, but it can also be demonstrated that the areas that received migrants actually enjoyed a higher level of economic prosperity in comparison with many parts of the Soviet Union.29

Russia was main loser of population during the first decades following the Second World War. The Ukraine had a positive net migration rate comparable to the negative net migration rate that prevailed in Russia. The Soviet Republic of Moldova gained large number of migrants due to the rapid growth of industrial production. In the 1960s employment grew in the Caucasian Soviet Republics. Georgia actually lost a large number of Armenian migrants to Armenia. Azerbaijan lost some of its importance in the oil industry. This outmigration was not very significant, and it remained well below negative 5 percent.

Concerning GDP hierarchies, unfortunately there is no systematic data for these countries before 1988, only sporadic figures. According to the Maddison databank, these countries were well above world averages in terms of per capita GDP in the 1970s. They were almost 50 percent higher than the global mean per capita GDP. There is some evidence according to which this position was more or less maintained until 1990. So one can assume that this trend began earlier, and also that this had been the case in the 1960s and even probably the 1950s. As Böröcz has shown, the whole USSR was above the world averages between 1950 and 1989, and the republics under consideration were seen as well-to-do.30

The change in the migration pattern and the switch to large-scale exodus was surely due to the collapse of these economies during the break-up of the Soviet Union and the transition away from state socialism. The collapse was so dramatic that most of them fell from above the global average positions to 50 percent of the world average of per capita GDP, and only Azerbaijan and Armenia got back above the average again after a certain period. These intraregional differences and the individual linkages to the global position of the relevant countries may prove the point that the relative position with regards to global averages can be an important factor in the migratory profile of a country. The case of Moldova demonstrates this very clearly.

Figure 19. Type three: immigrant countries that became emigrant countries. Per capita GDP, 1950–2008

Source: Maddison databank

 

Figure 20. Net migration rate and per capita GDP difference from world averages in Moldova, 1950–2010

Source: WPP 2010 revision and Maddison databank

When Moldova was an immigrant country in terms of per capita GDP it was well above the global average, and the decline of its position was directly correlated to the shift to an overall emigrant status. Moldova actually fell to a level of 40 percent of world mean per capita GDP, and this was why it had almost a world record level of remittance dependency, as shown by Böröcz in a recent study.31 This dependency type integration into both the world economy and the flow of people can lead to a situation in which tens of thousands of children are left behind by parents seeking jobs in Spain, Italy or Greece.

In the 1990s, from the perspective of the composition of the economy on the basis of sectors, these countries showed patterns similar to those that prevailed in the countries that were sources of immigrants. The industrial sector was strong in late state socialism and it collapsed during the transition. The cases of Moldova and Georgia are especially striking, as the share of industry in GDP declined from 40 percent to almost 10 percent, followed later only by partial gains. The other country producing very intensive outmigration was Armenia (with a drop in industry from 50 to 30 percent), while the others, the Ukraine and Azerbaijan, remained relatively stable.

Figure 21. Type three: countries that became emigrant countries. Industry as a percentage of GDP since 1970

Source: World Bank Data

Agriculture also played an important role even before the collapse of state socialism, but there was significant re-ruralization of these economies and labor markets, which then led to a global devaluation of these economies and to a pressure situation. Thus it seems that changes in global positions and related processes of the fall of the share of industry and re-ruralization together changed the overall integration of these countries into the global flow of people. Since the 1990s the whole region around the Black Sea has been an emigrant region serving as a repository of labor migrants from Russia and wealthier states of Southern Europe.32

 

Type Four: Cyclical Changes in the Former Republics of Yugoslavia

 

This type requires further attention, since due to the violent collapse of this federal state there were probably developments that were “incidental” in the sense that some of the flows of migration would not have taken place without the dramatic political changes and the wars themselves. There is a consensus in migration literature that the collapse of old states and the creation of new ones may produce waves of migration.33 Further analysis is necessary in order to determine whether the overall patterns found in the key states of Yugoslavia resembled one of the above types or in fact other processes were at work, processes that led to a distinct cyclical pattern.

In the case of Serbia and Croatia, during the late 1950s and early 1960s there was a move toward higher levels of outmigration due to a large extent to guest worker programs initiated by Germany and Austria. But during the early 1970s this process was not continued, and up until the early 1990s (the break-up of the federal state itself) there was a moderately positive net migration rate, which grew with the territorial fights that influenced various ethnic groups across the emerging new borders. Then a new cycle began. This contention regarding the cyclical nature of this pattern receives a boost from two additional observations.

Figure 22. Type four: oscillation. Net migration rates in the former Yugoslav Republics, 1950–2010

Source: WPP 2010 revision

If one recalculates Brunnbauer’s data,34 one sees an oscillation in emigration in Yugoslavia even between the two World Wars, first during the great economic crisis and then after 1938. The level was shifting between 10,000 and 40,000 people, with particular emphasis on European migratory links. With regard to the overseas links, a new cycle did not emerge at the end of the 1930s most probably due to political changes concerning immigration into the United States and the overall transatlantic relationships.

Secondly, if one considers changes in the overall global position of Yugoslavia and within Croatia and Serbia, the rise of Yugoslav per capita GDP above global averages correlates with the rise of net migration to positive levels up until the early 1990s, when in fact this link breaks down, most probably due to the war and the concomitant uprooting of people. Because of the lack of comparable and consistent data for the 1970s and 1980s, it is not possible to address the question of whether these changes could be related to changes in the economic structure and the labor market.

Figure 23. Type four: oscillation. GDP per capita in some former Yugoslav Republics, 1950–2008

Source: Maddison databank

Nonetheless, overall one can conclude that before 1985 Yugoslavia was following changes observed in countries in type one. Furthermore, had the break-up of the federal state been avoided, the country might have followed the pattern observed in Slovenia and other countries in type one. Or there may merely have been a historical oscillation in terms of net migration that was simply somewhat distorted by the collapse of the federal state. In order to answer this question, one would have to pursue further analysis on the basis of more reliable data.

A Possible Behavioral Link

It is widely acknowledged that in the case of arguments such as the macro type above there is no real behavioral analysis and the actual decisions of people to migrate are simply assumed through the construction of an argument according to which the overall scenario for such decisions was created due to macro-structural changes.35 In this paper I would like to raise the possibility that there may be a more direct link. There is increasing empirical evidence of the rather “accurate” knowledge of people concerning the overall standing of their country in global hierarchies of per capita GDP. Repeated surveys indicate that in many countries around the world people are clearly aware of hierarchical development and that they position their own countries rather “well” within these hierarchies.36

Data for Bulgaria and Hungary, two countries represented in the analysis of migratory developments, are important from this perspective.

 

Figure 24. Bulgaria 2009, per capita GDP for 2006 according to country rating

Source: Melegh, Attila et al.,”Perceptions of societal developmental hierarchies in Europe and beyond: A Bulgarian Perspective,” European Sociological Review (2012), accessed March 5, 2012, doi: 0.1093/esr/jcs010.

 

Figure 25. Hungary, 2010. Average country ratings (2010),
and per capita GDP (2007) for ten countries

Source: Special thanks to Ildikó Husz and Zoltán Szántó for having an extra module in the survey, “Pénzügyi kultúra Magyarországon” [Finance culture in Hungary] (INNOTARS_08-PENZKULT) Nemzeti Innovációs Hivatal, Corvinus University of Budapest.

 

In both cases, in representative surveys, when respondents were asked what score they would give their own countries and several other countries between zero and ten, they provided a rather consistent hierarchical map or developmental slope. More importantly, this slope correlated well with actual per capita GDP figures for a few years earlier. The overall Pearson correlation was as high as 0.91 in the case of Bulgaria, and it was also very high in the case of Hungary, 0.89. It is also important to note that there was an overall consensus among respondents and no major differences could be found among subgroups of respondents.

The mismatch in the case of some countries could be due to misunder­standings concerning the names of the countries (the country name of the Netherlands was not correctly understood in Bulgaria). Misestimates could also be due to misperceptions concerning relative prosperity in some larger countries, such as Russia and Turkey (in the case of Bulgaria) and China and Russia (in the case of Hungary).

From the perspective of migration, this might suggest that ordinary people are fairly aware of their country’s place in global economic hierarchies, and they might even follow changes in these hierarchies. This may well entail that when the relative position of their home country declines they consider it prudent to consider moving to another country.

Interestingly, one notes that Bulgarians substantially underestimate the relative prosperity of their homeland, which may be due to prevalent pessimism that developed because of a long term negative decline of their relative place in the global economic hierarchy, as described above. This overall frustration might influence the tendency to emigrate. Thus there may be a more direct link between migration rates and macro changes than generally assumed in the literature on the subject.

Conclusions

In the beginning of the period under discussion each of the regions examined here was either following European patterns of emigration or was actually serving as a migratory target (for instance in the case of Moldova). In the 1970s and 1980s (in other words well before the actual collapse of state socialism) diverging patterns began to emerge the differences between which became acute after the collapse. Some of the sub-regions (the Balkans and the region around the Black Sea within the Soviet Union) actually became destinations for migrants from countries in which in the meantime net migration rates had become negative. This is a distinctive story of the construction of inner dependency within a larger region the countries of which had a great deal in common, and this process needs to be analyzed with particular care.

Thus smaller meaningful historical, geographic regions can constructed on the basis of migratory patterns. These regions do not follow the “classic” divisions, and the state socialist and capitalist local histories are related to one another, regardless of divergences. State socialism was not isolated from global flows, and, more importantly, it partially reproduced global hierarchies and had its own effects on international migration.

In a modified form, the world system approach is helpful in furthering an understanding of longer term developmental patterns. In the case of state socialist economies, the direct intervention of world capitalism had a long-lasting impact on the migratory links between the countries within the region under discussion. Actually, most of the former state socialist countries in the region became dependent on remittances, as shown by Böröcz.37 When state socialism collapsed in the late 1980s, the economies of the countries of the region were based on a huge industrial sector. Countries that were unable to counterbalance the collapse of local industry became sending countries and were partially re-ruralized and partially pushed into large scale emigration. Thus the break-up of socialism also did not have a uniform impact on the countries in question, and the impact also depended on historical developmental hierarchies and the related ability of the various countries to regain some of the losses in the industrial sector with gains in the service sector.

The analysis offered here lends credence to the neoclassic macro-economic theory of migration, but I argue that its validity with regard to per capita GDP differentials is strengthened if it is linked to positions in global hierarchies. It thus needs to be re-contextualized into a world system approach. The key point is that it is not simply GDP differentials that matter, but rather positions within the global economy, which themselves are in part the results historical processes and linkages. In other words, one needs to go back to the theories of global relationships, which is the subject matter of global history.

In addition, I have also argued that global hierarchies and the positions of a given country in these hierarchies may well be fairly accurately perceived by the local and migrant populations. It seems that a more direct link can be found between global structures and behaviors in the perception of global hierarchies. People seem to have ideas of developmental scales that can very clearly linked to actual per capita GDP figures. Thus people might well be aware of global inequalities and may even have clear ideas of complex sequences that might orientate them in their decisions regarding migration. This hypothesis, however, merits further research, especially from the perspective of how positions in global hierarchies are perceived by people considering emigration.

 

Bibliography

 

Appadurai, Arjan. Modernity at Large: Cultural Dimensions of Globalization. Minneapolis–London: University of Minnesota Press, 1996.

Berend, T. Iván and György Ránki. The European Periphery and Industrialization, 1780–1914. Cambridge: Cambridge University Press, 1982.

Bonifazi, Corrado, Marek Okólski, Jeannette Schoorl, and Patrick Simon. International Migration in Europe. New Trends and New Methods of Analysis. Amsterdam: University Press, 2008.

Bonifazi, Corrado. “The Evolution of Regional Patterns of Migration in Europe.” In International Migration in Europe. New Trends and New Methods of Analysis, edited by Corrado Bonifazi, Marek Okólski, Jeannette Schoorl and Patrick Simon, 107–128. Amsterdam: University Press, 2008.

Böröcz, József. The European Union and Global Social Change: A Critical Geopolitical Economic Analysis. London: Routledge, 2009.

Böröcz, József. “Regimes of Remittance Dependency: Global Structures and Trajectories of the Former Soviet ‘Bloc’.” Manuscript, 2012.

Brunnbauer, Ulf, ed. Transnational Societies, Transterritorial Politics. Migrations in the (Post-) Yugoslav Region, 19th–21st Century. Munich: Oldenbourg, 2009.

Castles, Stephen and Wise Raúl Delgado, eds. Migration and Development Perspectives from the South. Geneva: IOM International Organization for Migration, 2007.

Chase-Dunn, Christopher. “Globalization: A World-Systems Perspective.” Journal of World-Systems Research 5, no. 2 (1999): 165–85. Accessed December 17, 2012, http://jwsr.ucr.edu/archive/vol5/number2/html/chase-dunn/index.shtml.

Chase-Dunn, Christopher, Yukijo Kawano and Benjamin Brewer. Economic globalization since 1795: Structures and Cycles in the Modern World-System (1999). Accessed December 17, 2012, http://wsarch.ucr.edu/archive/papers/c-d&hall/isa99b/isa99b.htm.

Faini, Riccardo and Alessandra Venturini. “Development and Migration: Lessons from Southern Europe.” ChilD no. 10 (2008), working paper.

Fassmann, Heinz, Ursula Reeger, and Wiebke Sievers, eds. Statistics and Reality. Concepts and Measurement of Migration in Europe. IMISCOE reports. Amsterdam: Amsterdam University Press, 2009.

Garndstaff, Peter J. Interregional Migration in the U.S.S.R. Economic Aspects. Durham, North Carolina: Duke University Press, 1980.

de Haas, Hein. “North African Migration Systems: Evolution, Transformations, and Development Linkages.” In Migration and Development Perspectives from the South, edited by Stephen Castles and Raúl Delgado Wise, 143–174. Geneva: IOM International Organization for Migration, 2007.

de Haas, Hein. The Determinants of International Migration. Conceptualizing Policy, Origin and Destination Effects. DEMIG project paper no. 2. International Migration Institute (IMI), Oxford Department of International Development, Queen Elizabeth House (QEH), University of Oxford, 2011. Accessed October 12, 2012. http://www.devstud.org.uk/aqadmin/media/uploads/4dc3c9cac333a_WP32%20The%20Determinants%20of%20International%20Migration.pdf.

Joppke, Christina. Selecting by Origin: Ethnic Migration in the Liberal State. Cambridge: Harvard University Press, 2005.

Massey, Douglas S., Joaquin Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino and J. Edward Taylor. Worlds in Motion: Understanding International Migration at the End of the Millennium. New York: Oxford University Press, 1998.

Massey, Douglas S. “Why does Immigration Occur? A Theoretical Synthesis.” In The Handbook of International Migration: The American Experience, edited by Charles Hirschman, Philip Kasinitz, and Josh DeWind, 34–52. New York: Russel Sage Foundation, 1999.

Melegh, Attila. “Migráns vagy munkás. Globalizáció és migráció a nemzetközi irodalom tükrében” [Migrant or Worker. Globalization and Migration. Review Article]. Eszmélet 16, no. 62 (2004): 83–101. (Also on the internet: http://eszmelet.tripod.com/62/melegh62.html.)

Melegh, Attila. On the East/West Slope. Budapest: CEU Press, 2006.

Melegh, Attila and Éva Kovács. “In a Gendered Space. Forms and Reasons of Migration and the Integration of Female Migrants.” Demográfia (English edition) 50, no. 5 (2007): 26–59.

Melegh, Attila, Arland Thornton, Dimiter Philipov, and Linda Young-DeMarco. “Perceptions of Societal Developmental Hierarchies in Europe and Beyond: A Bulgarian Perspective.” European Sociological Review (2012). Accessed March 5, 2012. doi: 10.1093/esr/jcs010.

Ministry of Interior Document: “Az illegálisan külföldre távozott személyek főbb adatai” [Data on Illegal Emigrants]. Statisztikai Szemle 68, no. 12 (1990): 986–1003.

Molodikova, Irina. “Patterns of East to West Migration in the Context of European Migration Systems. Possibilities and Limits of Migration Control.” Demográfia (English edition) 51, no. 5 (2008): 5–35.

Okólski, Marek. “Migration Pressures on Europe.” In European Populations. Unity in Diversity, edited by Dirk van de Kaa, Henry Leridon, Giuseppe Gesano, and Marek Okólski, 141–194. Boston–London: Kluwer Academic Publishers, 1999.

Portes, Alejandro, ed. The Economic Sociology of Immigration. Essays on Networks, Ethnicity, and Entrepreneurship. New York: Russel Sage Foundation, 1995.

Portes, Alejandro and József Böröcz. “Contemporary Immigration: Theoretical Perspectives on Its Determinants and Modes of Incorporation.” International Migration Review 23, no. 3 (1989): 606–30.

Rangelova, Rossitsa and Katya Vladimirova. “Migration from Central and Eastern Europe: the Case of Bulgaria” SEER, South-East Europe Review for Labour and Social Affairs no. 3 (2004): 8. Accessed November 7, 2012. www.ceeol.com.

Sassen, Saskia. “Foreign Investment: a Neglected Variable.” In The Migration Reader. Exploring Politics and Policies, edited by Anthony M. Messina and Gallya Lahav, 596–608. London: Lynne Rienner Publishers, [1990] 2006.

Sassen, Saskia. Guests and Aliens. New York: The New Press, 1990.

Tauger, Mark B. Agriculture in World History. London–New York: Routledge, 2011.

Thornton, Arland, Georgina Binstock, Mohammad Jalal Abbasi-Shavazi, Dirgha Ghimire, Arjan Gjonca, Attila Melegh, Colter Mitchell, Mansoor Moaddel, Yu Xie, Li-shou Yang, Linda Young-DeMarco, and Kathryn Yount. “Knowledge and Beliefs about National Development and Developmental Hierarchies: The Viewpoints of Ordinary People in Thirteen Countries.” Social Science Research 41 (2012): 1053–68.

Tilly, Charles. “Migration in Modern European History.” In The Migration Reader; Exploring Politics and Policies, edited by Anthony M. Messina and Gallya Lahav, 126–146. London: Lynne Rienner Publishers, Inc., [1990] 2006.

Tóth, Pál Péter. Haza csak egy van? Menekülők, bevándorlók, új állampolgárok Magyarországon (1988–1994) [Is there only one Homeland? Refugees, Immigrants, and New Citizens in Hungary (1988–1994)]. Budapest: Püski Kiadó, 1997.

United Nations, Department of Economic and Social Affairs, Population Division (2011): World Population Prospects: The 2010 Revision, CD-ROM Edition. Accessed December 17, 2012. http://esa.un.org/unpd/wpp/index.htm.

World Bank, World Development Indicators & Global Development Finance. Accessed 17 December, 2012. http://databank.worldbank.org/ddp/home.do?Step=12&id=4&CNO=2.

Ziesemer, Thomas. “Growth with Endogenous Migration Hump and the Multiple, Dynamically Interacting Effects of Aid in Poor Developing Countries.” Working paper series. United Nations University – Maastricht Economic and Social Research and Training Centre on Innovation and Technology, 2008.