Previous studies of migration in the Southern Balkans have been fragmented not only along national but also along disciplinary lines. Moreover, the fields of study of ‘forced’ and ‘voluntary’ migration have been separate, with few interconnections. Our collection intends to bring these different traditions together, and in addition aims to promote mutual understanding between the citizens of the countries of the region, who are now increasingly linked through transnational and globalized networks. In order to achieve this goal, the volume adopts a dual approach to the study of migration in the area. On the one hand, it focuses on migration flows and intercultural exchanges exclusively within the region; on the other hand, it looks at contemporary migration against the background of historical developments during the last two centuries, while at the same time proposing alternative readings of this history. From a methodological point of view, this study reserves a special place for aspects of cultural history often lacking in the literature of migration—most notably, religion, personal name strategies, gender, family strategies, and ‘memory’. It takes an interdisciplinary approach to migration—with insights from history, anthropology, sociology, economic geography, political economy, and oral history—and utilizes empirical studies of the societies and polities under examination rather than more theoretical elaborations. Our aim is to go beyond the narrow focus of national narratives that have dominated historiography in the region, by identifying transnational or trans-regional bonds from Ottoman times until the present day and also by bringing in comparative perspectives, where appropriate. We consider that migration patterns, coping strategies of individual migrants, and public discourses on migration in the present cannot be explained solely by contemporary developments, since they are deeply influenced by cultural traditions from the past.

The remainder of this chapter—like the contributions that follow—is organized historically. We start with a short discussion of migration in the Ottoman Empire, also considering some aspects of the late Ottoman Empire that are relevant to issues of migration in subsequent historical periods (such as heterogeneity versus homogeneity and the issue of nationalism). This is followed by a brief discussion of the implications of World War II and its aftermath for migration in the region. Next, we look at migration during the Cold War period and developments since the collapse of the communist regimes in Eastern Europe. We conclude by looking briefly at some consequences for intra-regional migration of the economic crisis that started in 2008.

1.1 Migration in Ottoman Times

Immigration, emigration, and internal migration all played a crucial role in the history of the Ottoman Empire. Immigration was a characteristic of the hey-day of the Ottoman Empire when many people found a safe haven and better economic prospects there. A typical case was that of the Spanish Jews who fled persecution in Spain and arrived in large numbers in the late fifteenth and early sixteenth centuries, settling in urban centres in the Balkans—especially in Salonika and Constantinople (Mazower 2004, pp. 47–52). In the late nineteenth and early twentieth centuries, when the Empire was in decline and disintegrating, many Ottoman subjects (mainly Christians and Jews) left for the USA (Ipek and Çagayan 2008). From the very beginning the Ottoman Empire was a ‘moveable empire’, as Kasaba (2009) phrases it, and the diverse populations in the Empire were frequently on the move—either voluntarily or forced by the authorities. The remarkable population transfers (sürgün) organized by the Ottoman authorities in the early period of the Ottoman Empire involved transfers of Muslim, Christian, and Jewish populations within the Empire (see, e.g., Hoerder 2002, pp. 111, 113, 117), while the devșirme—the levy of young Christian boys for service in the Ottoman army—also implied migration over large distances within the Empire, in this case of individuals (Lucassen 2009, pp. 21–22).

Besides these population movements between different parts of the Empire, there was also considerable movement within the (Southern) Balkans itself. Seasonal movements of agricultural labourers or of shepherds between winter and summer pastures, as well as the migration of itinerant artisans and traders, were all part of everyday experience. Part of this itinerant population subsequently settled in new locations. In Chap. 2, Petko Hristov describes these traditional forms of migrant labour, focusing on the Slavic-speaking part of the Southern Balkans and more specifically on Šopluk and Mijak, two mountainous border regions with a long tradition of migrant labour. There was also an almost continuous movement from the mountains to the plains, or in the opposite direction, depending on the security and health conditions prevailing in the Empire. The late Ottoman period saw new migration movements. Following the growth of trade in the late seventeenth century, thousands of Greek merchants established themselves in the cities in the northern part of our region as well as in the Northern Balkans, Central Europe, and southern Russia (Vermeulen 1984, p. 230). During the nineteenth century, cities in the Slavic-speaking regions—previously inhabited mainly by Turks, Greeks and Jews—saw their religious and ethnic composition transformed owing to the increasing migration of Slavic and Vlach speakers (Sahara 2011).

While the creation of nation states erected barriers to the free movement of shepherds, itinerant merchants, and all those who wanted to try their luck across state borders, it is difficult to get a clear picture of the volume and state regulation of cross-border migration within the Southern Balkans in the early phases of state formation. Some sources suggest that cross-border migration greatly diminished. Fatsea (2011), for example, found no evidence of the presence of foreign labour in the Athenian labour market during the founding phase of modern Athens (1830–1850).Footnote 1 A clear case of cross-border movement is the migration from the small and poor Greek state to Asia Minor in the years immediately after independence (1832). This migration—a continuation of migration flows since the eighteenth century—contributed significantly to the growth of the Greek population there (Adanır 2009, p. 64; Kitromilides 2008, p. 287).

1.2 Heterogeneity and Homogeneity in the Southern Balkans During the Last Period of Ottoman Rule

The frequent migrations during the Ottoman period contributed to the heterogeneity of the Southern Balkans in terms of language, religion and the origin of its inhabitants.Footnote 2 The ‘ethnic’ diversity of the Southern Balkans in Ottoman times is often perceived as being due to two factors. The first factor is the large number of ‘ethnic’ groups living there—a complex situation that many authors describe using the same ethno-national categories as we use today, such as Greeks, Turks, Bulgarians, Serbians, Montenegrins, Albanians, Gypsies and Jews. Second, these groups did not live side by side in neatly separated territories, but were to a large degree intermingled, especially in Macedonia and Thrace (see Sax map in the Appendix). Though such a picture gives some idea of the ‘ethnic’ complexity and cultural diversity of the region, it is potentially misleading, because it applies present-day categories to the past. People at the time did not see themselves in terms of such categories, although they slowly learned—and were taught and obliged—to do so.

How did the peasants—the overwhelming majority of the population—see themselves before the nationalist world-view reached them? Let us look briefly at the Orthodox population of the Balkans.Footnote 3 Simple peasants at the time would have seen themselves first of all as members of a family and village. Next, they would have identified themselves as members of a religious community—in our case, the Christian Orthodox one. The Ottoman Empire was organized into religious communities, the so-called millets. The Orthodox millet was known as the Rum millet and those who belonged to it called themselves Christians or—less frequently—Rum (in Turkish), Romios (in Greek), or some equivalent in another Balkan language. Though a religious label, the term also had a Greek connotation. The Orthodox millet was dominated by a Greek-speaking elite and the Greek language, called Romeika in Greek at the time, was the language of the church or millet and of the schools run by it. Besides the Orthodox millet, there were Jewish, Armenian and Muslim ones: each consisted of people speaking different languages. However, peasants were not only members of families, villages and religious communities. There were many named groups at an intermediate level, between village and religious community, such as the Hashiots (very poor, Greek-speaking Orthodox Christians), Dönme (Jewish Muslims), Vallahades (Greek-speaking Macedonian Muslims), Gagauzes (Turkish-speaking Orthodox Christians), and Pomaks or Torbesh (Slav-speaking Muslims).

Before the rise of nationalism the term ‘Bulgar’ (or some variant of it) did exist, but usually had the connotation of ‘poor Slav-speaking peasant’. Upwardly mobile Slav-speakers became Hellenized and no longer considered themselves Bulgars. Greek-speakers usually referred to themselves as Romios, but a Bulgar as a member of the Greek-dominated Orthodox Church or Rum millet could do so as well (see, e.g., Detrez 2013). The notion of ‘Turk’ also had a different meaning than it has today—being used for Muslims, even when their native language was not Turkish but Albanian, Greek or Slav. It also had connotations of status. For the Ottoman elite, the Osmanlı, the term ‘Turk’ had the connotation of country bumpkin (see, e.g., Lewis 1971, pp. 19–20; Mazower 2001, p. 51).

Gellner (1997, p. 20) argues that cultural diversity or differentiation as well as its maintenance over time was central to agrarian societies like the Ottoman Empire:

Agrarian society encourages cultural differentiation within itself. Such differentiation greatly helps in its daily functioning. Agrarian society depends on the maintenance of a complex system of ranks, and it is important that these be both visible and felt, that they be externalised and internalised. If they are clearly seen in all external aspects of conduct, in dress, commensality, accent, body posture, limits of permissible consumption and so forth, this eliminates ambiguity and thus diminishes friction.

Notwithstanding the enormous cultural differentiation and complexity, the student of the Southern Balkans can hardly miss the cultural similarities that also existed across linguistic and religious boundaries (Detrez 2013), and in this case migration played a role as well. Weakley (1993, p. 130) provides us with a good example. Writing about the houses constructed by Greek itinerant builders over a vast area in the Southern Balkans, he comments on people speaking different languages and professing different religions:

The differences are seen in slight variations in plan and in certain detail elements…. The key design element was a single housing form for a pluralistic and diverse culture (ibid., italics original).Footnote 4

This unity of Balkan culture can be observed not only in architecture, but also in music, cuisine, kinship, religious beliefs and practices, and in other cultural domains.Footnote 5 Friedman (2011), the Balkans constitutes a linguistic area with common characteristics: Balkan languages, despite their diverse origins, show many similarities in their grammar, called Balkanisms.Footnote 6 Perhaps more important for our introduction is the domain of religion: there is abundant evidence of syncretic religious beliefs and practices in regions where Muslims, Christians and Jews lived in close proximity. Syncretism existed in many places in the Balkans and the rest of the Ottoman Empire.Footnote 7 Remnants of such syncretism still exist in the region and are visible, for example, in the sharing of sacred places (Albera and Couroucli 2012; Duijzings 2000). In Chap. 7, on Albanian immigrants in Thessaloniki, Ifigeneia Kokkali considers syncretism in Albania and Albanian history. In an attempt to explain the remarkable way in which Albanians integrate into Greek society—and more specifically the role that religion plays in this process—she looks at the context of reception as well as Albanian history and the ‘cultural baggage’ that Albanians bring with them. A certain ‘syncretic attitude’ towards religion is one element of her analysis.

1.3 Creating National Identities

Before the advent of nationalism, the major identity split in the Balkans was between Muslims and Christians. Most Muslims were Turkish-speaking, but those who spoke other languages were often also called Turks and could identify themselves as such. There was a similar correspondence between Christian and Rum. In the early phases of Balkan nationalism when the Rum millet was still intact, Orthodox Albanian and Slav speakers tended—as Rum—to participate in the Greek struggle for independence, though some of them subsequently became fervent nationalists of their own linguistic nation.Footnote 8 Whatever assimilation existed in the Ottoman Empire occurred mainly within the millets. Among native Muslim groups this tendency continues to exist in some places to this day. Smaller Muslim groups tend to identify with, and to assimilate to, the dominant Muslim minority within a particular country or region rather than the dominant Christian majority.Footnote 9 This illustrates that ‘in many ways the legacy of the Ottoman millet system has endured as religion continues to be an important differentiating factor among people’ (Poulton 1997, pp. 20–21).

The unity of the Orthodox community started to break down when the Greek national Church was founded in 1833. The real split came, however, with the foundation of the Exarchist Church in 1878. Although the original goal was ‘only’ to introduce the Bulgarian language in the liturgy, the church soon became an instrument in the Bulgarian national struggle. From that moment, Bulgarian-speaking peasants were confronted with a difficult choice: to remain in the Greek-dominated Patriarchist Church and ‘become Greek’ or to go over to the Exarchist Church and become Bulgarian. Whole villages, particularly in areas to the south bordering on the territory of Greek-speakers, became divided into Greek and Bulgarian parties. Similar processes took place in Albanian-speaking and Aromanian-speaking Orthodox villages (Van der Plank 2004, p. 90; Vermeulen 1984, pp. 242–243). Schools played a major role in inculcating a national identity: during the nationalist campaigns in Macedonia, at the end of the nineteenth and beginning of the twentieth century, the number of Greek, Bulgarian and Serbian schools increased dramatically.

Greek-speaking Orthodox individuals, belonging to the culturally dominant group within the Rum millet, were confronted with such difficult choices to a much lesser extent, but even they had to learn to be Greek. As Kitromilides (1989, p. 169) writes of the Greek communities in Asia Minor, the identification with the Greek nation ‘had to be instilled and cultivated, or “awakened”, as older nationalist historiography might say, through a crusade of national education’. Dragostinova (2008, 2011) writes in a similar way about the Greeks who lived on the Bulgarian Black Sea coast. Though most of these Greeks supported the Greek national idea, they had strong ties to their places of birth, were reluctant to emigrate and some opted for Bulgarian nationality. Moreover, the Greek government organized there ‘a massive enterprise of national persuasion by dispatching activists’ (Dragostinova 2008, p. 167; 2011, pp. 24–31).

1.4 Unmixing Populations and ‘Cleaning’ the National Territory: Forced and ‘Voluntary’ Migration in the Age of Nation State Formation

The rise of nationalism in the late eighteenth and nineteenth centuries had great consequences for the Southern Balkans. In the early nineteenth century, the entire region was still part of the Ottoman Empire; by 1923 the Southern Balkans were divided between five nation states—Greece, Bulgaria, Serbia, Albania and Turkey. Greece was the first to break away with its War of Independence of 1821–1828, ultimately achieving recognition of a small territorial state (consisting of Attica, the Peloponnese and the Cycladic islands) in 1830–1832. Further territorial expansions included the Ionian Islands (1864) and Thessaly (1881).Footnote 10 Nevertheless, by 1912 over five million Greeks (mostly Ottoman citizens) remained outside the territory in Macedonia, Epirus, Thrace, Asia Minor, Cyprus, the Aegean Islands, Crete and southern Russia (Petsalis-Diomidis 1978, p. 15). In 1912 the Balkan League, consisting of Bulgaria, Serbia, Greece and Montenegro, attacked Ottoman troops in an attempt to expel the Ottomans from the peninsula. This was succeeded by the Second Balkan War, started by Bulgaria (see below). As a result of these two Balkan wars, Greece’s territory expanded massively to include Crete, much of Macedonia, Epirus and the Aegean Islands. Sovereignty of the Aegean Islands was finally confirmed by the Treaty of Sèvres (1920), although the disastrous Greco-Turkish war that followed resulted in considerable loss of territoryFootnote 11 as determined by the 1923 Treaty of Lausanne.

Following the Greek pattern, Bulgarian revolutionaries engaged in a series of uprisings against the Οttomans, but were unable to dislodge Ottoman power. It was not until Russia declared war on the Ottoman Empire that the peace of the San Stefano Treaty (1876) created the first Bulgarian state. This was bitterly opposed by the Austro-Hungarian and British empires, and in 1878 the Treaty of Berlin reduced the new Bulgarian state to a ‘principality’ that was 37.5 % of its original size, returned Macedonia to Ottoman rule and created a new autonomous state of Eastern Rumelia—also Ottoman territory (Crampton 2005, pp. 81–83). Just seven years later, Eastern Rumelia’s military coup of 1885 reunited it with Bulgaria; this was formalized in Bulgaria’s declaration of full independence in 1908. As a result of the subsequent First Balkan War, Bulgaria greatly expanded. Dissatisfied with the division of the spoils in Macedonia, however, Bulgaria then attacked Greek and Serbian positions, thus starting the Second Balkan War. It lost this war, as well as most of the territorial gains obtained as a result of the first war. The net result of both wars was not insignificant, though: Bulgaria had obtained Eastern Thrace and its access to the Aegean Sea. Ten years later, at Lausanne, it would lose this as well.

Albanian nationalism had no equivalent of the Philhellenism of the European elites, or even the pan-Slavist movement in Russia (Vickers and Pettifer 1997, pp. 1–2); with no support from the ‘Great Powers’ (i.e., the international community), Albanian independence was to be both delayed and fractured.Footnote 12 From the very outset, Albanian aspirations conflicted with Greek ones, and initially this strengthened Albanian support for the Ottomans. After the 1908 rise to power of the ‘Young Turks’, by 1909 the Albanian Kosovars were in revolt, and by 1911 Kemal had agreed in principle to Albanian demands for autonomy—in line with those conceded to other Balkan nationalities (Kondis 1976, p. 54). However, an Albanian provisional government was achieved only as a consequence of the two Balkan wars and the 1913 Treaty of Bucharest. After a period of anarchy, the borders of Albania were finally settled with the Corfu Protocol of 1914.Footnote 13 Before this could be applied, World War I broke out, leaving a poorly-equipped new Albanian state to cope with it (Kondis 1976, pp. 132–133, 137).

What is now the Republic of MacedoniaFootnote 14 was for a very short time part of ‘San Stefano Bulgaria’ until it was returned to the Ottoman Empire by the Congress of Berlin in 1878. After the Balkan Wars, with the Treaty of Bucharest (1913), Macedonia was divided between Greece, Serbia and Bulgaria. When the Kingdom of Serbs, Croats and Slovenes was formed after World War I, it became part of the federal state that during World War II was transformed into the Socialist Federal Republic of Yugoslavia. During both world wars the Yugoslav Republic of Macedonia—as well as Greek Eastern Macedonia and Thrace—was temporarily occupied by Bulgaria, which still claimed it as its own. Macedonia gained its independence only in 1991, after the break-up of Yugoslavia.

The modern Republic of Turkey is, of course, what remains of the former Ottoman Empire. As a member of the Axis Powers, the Ottoman Empire belonged to the camp of the losers of World War I. In 1919, Kemal Ataturk started a revolt against both the Ottoman government and the occupying allied forces: the Turkish War of Independence (1919–1923) resulted in the end of the Ottoman Empire and the creation of modern Turkey.

Carving out more or less homogeneous nation states was not an easy task, especially in the Via Egnatia region with its mosaic of cultures characterized by hybridity and syncretism. No wonder it became an arduous and cruel process, notwithstanding the emphasis on heroism and national glory in the historiography of the ­individual nations. Creating homogeneous nation states and national cultures required a lot of work in many domains: a national history had to be written and the national language, religion and landscape had to be purified—freed from ‘foreign’ elements. Religion and national identity became strongly linked, especially in Turkey and Greece, with Albania constituting the only real exception. There, the continuing co-existence of Islam and Christianity in its Orthodox and Catholic variants prevented a close link between national identity and religion while leaving some space for the continuation of syncretic attitudes and practices.

Purifying the nation also implied demographic engineering, the unmixing of ethno-religious populations, or the ethnic cleansing of the new national territories’ ‘foreign populations’ (Sigalas and Toumarkine 2008; Zürcher 2008). Landscapes were nationalized by giving places with ‘foreign’ names new, ‘national’ names. Cultural monuments that evoked the presence of others were often used for new purposes or neglected, if not destroyed. Forced migration and other, more obnoxious, ways of ethnic cleansing such as massacres and the burning of villages were recurrent phenomena.

In the Southern Balkans, the Greek War of Independence marked the beginning of this process, leading to the deaths not only of soldiers but also of many civilians. The Greeks, perceiving they were on the winning side, seized their chance and cruelly murdered many Turkish civilians—for example, at the massacre of Tripolitsa. The Turks took revenge in Constantinople, Smyrna and on the island of Chios (see, e.g., Rodogno 2012, p. 66), with this last massacre, especially, gaining much attention in Europe at the time. Woodhouse (1977, p. 136) concludes, ‘On both sides atrocities were appalling.’ Those Muslims who had been able to save their lives fled the small independent Greek state and migrated northwards to Ottoman-held territory (McCarthy 1995, pp. 10–13). In the Russian-Turkish War (1877–1878) the victorious Russian troops, supported by Bulgarian irregulars, took revenge on the Turks (McCarthy 1995, pp. 65–81). Large numbers were killed and many fled to Ottoman territories. Cities like Thessaloniki and Istanbul were flooded by masses of refugees (see Akyalçin-Kaya 2011 for Thessaloniki). According to Mazower (2001, p. 11), over the period 1878–1913 some 1.7–2 million Muslims migrated voluntarily or involuntarily from the Balkans to what later would become the Republic of Turkey. By the 1880s, the Ottoman administration had already sent experts to sensitive border areas such as Thrace to see if refugees could be settled there to make these areas more Turkish (Adanır 2006, p. 177).

The two Balkan Wars (1912–1913) mentioned above were a combination of fighting, burning villages, massacring people and putting people to flight. As the International Commission to Inquire into the Causes and Conduct of the Balkan Wars wrote, ‘The Turks are fleeing before the Christians, the Bulgarians before the Greeks and the Turks, the Greeks and Turks before the Bulgarians, the Albanians before the Servians’ (Carnegie 1914, p. 154).

World War I (1914–1918) brought new population movements, mainly as a result of changing state borders, and was followed by the Greco-Turkish War which was at least as violent as the Balkan Wars. Until the Balkan Wars, the main (forced) migration movement had been that of Muslims from the Balkans to Turkey. As a result, by 1923 over 20 % of the population of Turkey consisted of people with muhacir (that is, refugee) background (Zürcher 2003, p. 6). The wars ‘caused a staggering refugee problem that was estimated by independent observers to have involved up to three million people’ (Kitromilides 2008, p. 256).

Although a cynic or fervent nationalist would perhaps say that the formation of nations and national identities had made a big leap forward after (and as a result of) all these wars, violence, ethnic cleansing and mutual hatred had increased as well. Those who did not belong to the dominant nation became minorities, and usually enemies ‘within the walls’ as well. Moreover, in several cases these minorities lived in border areas, claimed by the neighbour. Under the circumstances, the best solution politicians could think of was the exchange of populations—a method considered legitimate at the time (see, e.g., Brunnbauer and Esch 2006, p. 11). And ‘despite the great human hardship engendered by population exchanges, the improvement in regional stability cannot be ignored’ (Barutciski 2003; see also Clark 2006, pp. 223–246). The largest and most well-known of these exchanges is that between Greece and Turkey (1923) and it is also the one that is best studied, though mainly from the Greek perspective. In Turkey there used to be little interest in the exchange, although interest in the topic is increasing (Kitromilides 2008, p. 269). The Greek-Turkish population exchange was laid down in the ‘Convention concerning the Exchange of Greek and Turkish Populations’ signed in Lausanne on 30 January 1923. It came into force when it was included in the peace treaty of Lausanne (24 July 1923). Greeks were defined as ‘Turkish nationals of the Greek Orthodox religion established in Turkish territory’ and Turks as ‘Greek nationals of the Moslem religion established in Greek territory’ (Hirschon 2003, p. 8). Most Greeks of Istanbul and Turks from Western Thrace—about 100,000 in each case—were excluded from the exchange. The convention covered all those who had become refugees since 18 October 1912. This included about 1.2 million Greek Orthodox persons and slightly fewer than 400,000 Muslims. Since many Greek Orthodox people had already left Asia Minor over the period 1912–1923, the number actually falling under the conditions of the exchange (that is, those leaving after 1923) was considerably lower—around 350,000. The corresponding number for the Muslims was about 190,000.Footnote 15

As the Greek-Turkish population exchange is better documented, and also as we are more interested in the region where the five states involved in our book border each other—the area we call the Via Egnatia region, we consider here mainly the other two cases of population exchange, between Bulgaria and Turkey and between Bulgaria and Greece, respectively, agreed in September 1913 and in November 1919. Here we will not go into detail on these exchanges since both cases are well-treated in Chap. 3 by Raymond Detrez, who shows how the expulsions and exchanges of populations relate to nation-building and irredentist policies. Vukov’s contribution (Chap. 4) also relates to the population exchanges in which Bulgaria was involved, although his interest is less in the population exchanges themselves and more in what happened to Bulgarian refugees from Thrace and their offspring long after they settled in Bulgaria. He demonstrates that many people of Thracian origin are still involved in Thracian organizations and in claims for compensation for lost property.

There is a clear legal difference between the three population exchanges: the first two were voluntary, the last one obligatory. Many of the people involved in the Bulgarian-Greek and Bulgarian-Turkish exchanges were nevertheless forced to leave, either before, during or after the exchange. In practice, the difference is one of degree. The main criterion of selection in the Bulgarian-Turkish and the Greek-Turkish exchanges was religion. In the language of the day and also in many subsequent publications, Turks are said to be exchanged for Greeks; however, the selection criterion was not language as West Europeans would expect, but religion. The Karamanlides (Turkish-speaking Christians in Anatolia), for example, were forced to migrate to Greece; and the Vallahades (Greek-speaking Muslims from what is now Greek Macedonia) had to leave Greece for Turkey.

There are some characteristics that these population exchanges have in common, and these are points which also recur in the contributions by Detrez and Vukov. All three states involved in the exchanges were very keen on ‘cleaning’ border regions from populations considered to be a fifth column or ‘foreign element’. The border regions were, moreover, considered ideal locations for settling incoming refugees, returning kindred peoples from afar. If the original population was removed, these refugees could be settled in those houses; and if native but ‘alien’ local groups were still living there, they could be induced or forced to leave. It was sometimes considered important to keep a claim on territories on the other side of the border. In these cases, ethnic kin living just across the border had to remain there, rather than ‘return to the fatherland’ as, for example, in Venizelos’ policy regarding the Greeks in Western Thrace when this region came under Bulgarian control (Adanır 2006, p. 182). Another recurring theme is that people often had to migrate more than once. This happened, for example, to the Greeks on the Turkish coast, to the Bulgarians in Thrace, and to the Turks after the Russian-Turkish war. A final point to note is the radicalization among refugees—for example, the muhacir, who played an important role in the expulsion of Greeks from the Aegean coast before the Greek-Turkish war (see, e.g., Van der Plank 2004, p. 73).

Though the exchanges of populations contributed to the security of the region, the human price paid for it was tremendous (see, e.g., Clark 2006, pp. 223–246 on the Greek-Turkish exchange). Kitromilides (2008, p. 266) sees yet another drawback to the exchange of populations as a policy of ethnic cleansing:

By reducing ethnic pluralism through such radical means, the exchange prevented the modern national societies that emerged from it from learning the skills and internalizing the values necessary for the practice of toleration, mutual respect of social groups and recognition of otherness.

Ultimately, this had consequences for the way that immigrants—not only those from other Balkan countries—were received in the 1990s.

1.5 World War II and its Aftermath

The Axis Occupation, resistance movements and the civil wars of the 1940s had a deep impact on the societies of the Southern Balkans. At the end of the decade, the establishment of communist regimes in Albania, Yugoslavia and Bulgaria led to the onset of the Cold War which would seal off Greece from its northern neighbours for half a century. Economic hardship, violence and demographic politics led to mass migration, both within and across the borders of nation states. Although the Occupation created a new situation, some of the issues that had divided Balkan nations in the past continued to play a crucial role—territorial claims, irredentism, ethnic minorities, and the consequences of the interwar exchange of populations. Some of the events of the 1940s are the source of continued tensions between states in the Southern Balkans up to the present day, most notably the Macedonian Issue and the expulsion of the Albanian-speaking Muslim minority of Greece (Chams) to Albania.

The war created a major split between the states: while Albania, Greece, and Yugoslavia were occupied by the Axis Powers and developed coordinated resistance movements, Bulgaria actually joined the Axis and facilitated the invasion of Greece and Yugoslavia by German troops. Through this move, Bulgarian troops were allowed to occupy Serbian Macedonia, the eastern part of Greek Macedonia, and Western Thrace. This temporary annexation was seen by Bulgarian nationalists as a reunification of the Bulgarian nation. Italian troops occupied Albania and most Greek territory, while the Germans occupied Central Macedonia, Athens, Thessaloniki and the island of Crete. After the surrender of Italy in September 1943, Italian-held territory came under German control, but most of Italy’s weaponry ended up in the hands of the partisans. In September 1944, following a change of regime in Bulgaria, Bulgarian troops evacuated Greek and Yugoslav territory and took part in the final battles of the war against Nazi Germany. Turkey remained neutral until the last months of the war, when it joined the Allies.

Many of the migration movements during World War II were of a temporary nature.Footnote 16 Between 1941 and 1943 nearly 23,000 Greek citizens fled to Turkey. Among them were soldiers of the defeated Greek army and partisans who had received an order to join the Greek forces under the command of the Greek government-in-exile in Cairo. The rest were refugees from islands of the Northern Aegean who fled in panic as the Germans approached or were driven by hunger. They remained in Turkey for the rest of the war.Footnote 17 Similar temporary movements of refugees also occurred in the border areas between Greece and its northern neighbours, Albania, Yugoslavia and Bulgaria. Towards the end of the war, the newly created People’s Republic of Macedonia received increasing numbers of refugees from the Macedonian minority in Greece. Among them were many families from the border area, but also Macedonian resistance fighters who had served within the ranks of the Greek Partisan Army ELAS, as they had been promised equal rights after the war. In May 1944, however, they came into conflict with the ELAS leadership and joined the Yugoslav Partisan Army. In 1945 there were 8,500 Macedonian refugees from Greece in Yugoslav Macedonia. The next year, with the outbreak of violence by right-wing irregulars, which often specifically targeted Macedonian villages, their numbers swelled to 20,000 (Michailidis 2004, p. 46–47).

By far the most important migration waves during the war years occurred as a result of the Bulgarian occupation of Eastern Macedonia and Western Thrace. The Bulgarian occupation forces adopted everywhere a policy of forced Bulgarization. In contrast to what happened in Yugoslav Macedonia, in Greece these Bulgarization policies were particularly harsh, as the majority of the inhabitants, after the interwar exchange of populations, were Greek-speaking. People were forced to acquire a Bulgarian identity card, an early uprising in September 1941 was bloodily suppressed and Greeks were forced or encouraged to sign applications for ‘voluntary migration’ (Aarbakke 2004, p. 38, 2005, p. 167; Kotzayeoryi-Zymari 2002, p. 139). Others, especially army officers, academics, and civil servants, were forcibly expelled (Kotzayeoryi-Zymari 2002, p. 140). As a result of these pressures, about 150,000 Greeks left their homes and moved west to German-held territory. In 1941 another 12,000 Muslims from Western Thrace fled to Turkey (Kotzayeoryi-Zymari 2002, p. 152–155).

Bulgarian authorities further pursued their Bulgarization policies by bringing in about 100,000 Bulgarian settlers, many of whom were former inhabitants of the region. Associations of Thracian Bulgarians in Bulgaria played an active role in these policies. In a proclamation of May 1941 their representatives condemned the way in which they had been ‘expelled’ from the area under the Neuilly Treaty and claimed the right to return to their homes in the now ‘liberated’ area (Aarbakke 2004, p. 384). In September 1944 all the settlers returned to Bulgaria, together with about 9,000 Bulgarian-speaking residents of the area who became refugees for the first time. The stream of refugees from Eastern Macedonia and Thrace continued over the following months, and at the end of 1944 there were some 100,000 refugees in Bulgaria (Aarbakke 2004, p. 386).

Even more dramatic was the fate of the Albanian-speaking Muslim minority (Chams) of Epirus during the war. In 1940 about 20,000 Chams lived in this area, which had become part of the Greek nation state in 1913. Most of them inhabited the fertile lowlands, and frictions over land subsequently became a source of conflict with their Christian neighbours. Although in 1926 the Chams were exempted from the exchange of population foreseen in the Lausanne Treaty, large portions of Cham lands were expropriated and distributed to Greek refugees from Asia Minor or to local landless peasants. Others sold their properties, being led to believe they would be included in the exchange and forced to leave for Turkey. As a result, large numbers of the now landless Chams fled to Albania, where they formed a restless community. When war broke out in 1940 between Italy and Greece, the Italian authorities exploited the feelings of bitterness among the Cham communities both in Greece and in Albania and set up special Albanian-speaking military units to help them fight first the Greek National Army and later the Resistance. After 1943 the German army followed their example. Some Chams, however, fought together with Greek partisans in the left-wing Resistance units of ELAS. During 1943–1944 violence escalated and atrocities were committed by both sides. Between June and September 1944 the non-communist resistance organization EDES launched a major attack on Cham communities, in which about 2,000 Chams were massacred, women raped, and mosques and houses burnt to the ground. The survivors of these massacres, about 18,00–20,000 individuals, fled to Albania. The Greek census of 1951 registered only 77 Albanian-speaking Muslims in Greece. After the end of communist rule in Albania, the Cham Issue resurfaced, when Cham communities in Albania and the diaspora launched an international campaign to reclaim their properties.Footnote 18

The second half of the 1940s produced new waves of mass migration within the Via Egnatia area. The establishment of communist rule in Albania, Yugoslavia and Bulgaria led many Muslims of these countries to seek refuge in Turkey, mainly because of the restrictions on religious practices (Icdygu and Sert, this volume). The largest streams of refugees, however, were linked to the violence of the Greek Civil War. By the end of 1947 there were already 25,000 refugees in Albania and 18,000 in Yugoslavia (Danforth and Van Boeschoten 2011, p. 46). During the same period, 327 refugees, mostly women and children, arrived in Bulgaria. In 1948 the left-wing Democratic Army forcefully evacuated villages from border areas under its control to Albania and Bulgaria (Danforth and Van Boeschoten 2011, pp. 53–54). In the same year, the partisans evacuated about 20,000 children from northern Greece to Albania, Yugoslavia and Bulgaria (Danforth and Van Boeschoten 2011). After the defeat of the Democratic Army in August 1949 the number of political refugees who had left the country reached a total of 140,000. However, as the neighbouring countries were unable to cope with the enormous flows of destitute and hungry people, the majority of the refugees were moved to other countries of Eastern Europe, where most of them would remain for over 30 years. About 4,500 refugees stayed in Bulgaria (Aarbakke 2004, p. 392) and 30,000 in Yugoslavia (Kofos 1964, p. 168). Most of the refugees in Yugoslavia were Macedonians who settled in the People’s Republic of Macedonia.

In the aftermath of the Greek Civil War, many political refugees were stripped of their nationality and their properties were confiscated. Deserted villages of the border areas—especially those formerly inhabited by Slavic-speakers—were resettled with people from other areas of Greece, deemed to be of ‘healthy national beliefs’. Until 1974 the majority of political refugees settled in Eastern Europe were not allowed to return to Greece. Therefore, in the 1950s and 1960s tens of thousands moved to Bulgaria and Yugoslav Macedonia, in order to be closer to the villages of their birth and to enjoy the milder climate. Mass repatriation became possible only after 1981 when the Socialist Party PASOK was voted into power. However, a decree of 1982 restricted the right of repatriation to ‘Greeks by birth’. Since then, many Macedonian refugees have been refused permission to enter Greece even for short visits (Danforth and Van Boeschoten 2011, p. 36; Monova 2008). This painful legacy of the past is at the core of the Macedonia Issue which has poisoned relations between Greece and its northern neighbour.

1.6 The Cold War Period (1947–1989)

The political architecture of Europe that resulted from World War II was absolutely central for Balkan countries and their management of borders. The Yalta Conference had placed Greece unambiguously in the western sphere. Under the Truman Doctrine, both Greece and Turkey, despite the latter’s ambiguous position prior to and during the war, were increasingly seen as strategically important to the West; moreover, the Marshall Plan of 1947 nudged Turkey into a more pro-Western position. In contrast, the Balkan countries to the north had all initially fallen under the Soviet sphere of influence in 1945: by 1949, Tito had distanced Yugoslavia from Soviet influence and received substantial western aid and loans, followed in 1953 by a friendship treaty with Greece and Turkey (Jelavich 1983, p. 328). Albania broke relations with Yugoslavia after 1948 and became the most doctrinaire and isolated country in the region, closely allied with Moscow until 1960, when it shifted its allegiance to China (Jelavich 1983, pp. 331–333). Bulgaria was a close ally and adherent to Stalinist policies, remaining the region’s closest ally of Moscow right until the final collapse of the communist bloc.

Albania became the first regional target of US and British security services, which recruited some 300 Albanian dissidents from Greece, Italy and Egypt and sent them in as guerrillas to organize anti-communist uprisings over the period 1950–1952 (Jelavich 1983, pp. 378–379). All were killed, and Albania effectively sealed its borders and isolated itself from the entire region until 1991. Thus, there was no migration of any sort for the duration of the Cold War.

Greece, despite the largest US investment in the world from the Marshall Plan and military aid (Mazower 2001, p. 119), suffered from a very weak economy, even with massive capital inflows. This led to extraordinary levels of emigration—initially to the USA and after 1960 to northern Europe, primarily Germany (Vermeulen 2008). Migration flows within the region followed the general Balkan pattern of ethnic rationalization (i.e., creating a mono-ethnic state), with removal of Greek citizenship from minority groups, the emigration of minorities to Bulgaria, Yugoslavia and Turkey (see Içduygu and Sert, this volume), and the immigration of ethnic Greeks. In particular, as political relations between Turkey and Greece soured over Cyprus, the Greeks of Istanbul became vulnerable. Just under a third of them did not possess Turkish citizenship, and in March 1964 Turkey denounced the 1930 Convention on Establishment, Commerce and Navigation and began deportation of Greeks on the grounds of national security (Alexandris 1992, pp. 280–281). By September 1965, over 6,000 Greeks had been deported from Turkey; many were wealthy businessmen who arrived penniless in Greece (Alexandris 1992, pp. 284–285). The fear of political persecution also led Greeks in possession of Turkish citizenship to join the ranks of refugees, and an additional 30,000 were reported by the Turkish press as having left Istanbul. Naturally, this had a massive impact on the size of the Greek community in Istanbul and contributed to the ongoing ethnic rationalization common to both Turkey and Greece.

Yugoslavia was not only a federal state, but a multi-ethnic one in which historic ethnic tensions were suppressed through the overarching ideology of the Communist Party of Yugoslavia. There were tensions between developed and less developed regions, but traditional ethnic rivalries were the major concern. Many Muslim Yugoslavs—especially Albanians—took advantage of the improved relations between Turkey and Yugoslavia after the signing of the Balkan Pact and various treaties, and just under 200,000 migrated to Turkey during the Cold War period. They consisted of Albanians, Slav-speaking Muslims from Macedonia (Torbesh) and Bosnians as well as Turks, but all claimed Turkish descent (see Içduygu and Sert, this volume). With the creation of the People’s Republic of Macedonia in 1943, Yugoslavia addressed the long-standing issue of regional identity (although the major part of the historical territory lay in Greece). In 1944 the government in Skopje declared Macedonian to be the republic’s language and a commission chose a Macedonian local dialect that was the furthest removed from Bulgarian and Serbian (Jelavich 1983, p. 399). An official history and literature were contrived, along with an autocephalous Macedonia Orthodox church; ultimately, these resulted in serious tensions with Bulgaria, in particular. From the early 1950s, Yugoslavia tended towards economic liberalization along with greater freedom for its citizenry compared with the rest of the communist bloc. Citizens were free to travel, and tourism was actively welcomed with few visa controls. Labour emigration was also freely permitted. By 1975, persistently high levels of unemployment had resulted in over a million Yugoslavs and their families working abroad—representing 20 % of the actual labour force still in the country (Jelavich 1983, p. 392). Most were working in Germany and Austria, but there was a small presence in almost every European country. Yugoslavia was the second largest source country for labour migration across Europe, after Turkey (Kupiszewski 2009, pp. 427–429).

Bulgaria had agreed at the end of the war that those parts of Pirin Macedonia within Bulgarian territory should be ceded to Yugoslav Macedonia when the Balkan federation was established. It even recruited teachers skilled in the new Macedonian language that had been selected by the Yugoslav republic (Crampton 2005, p. 190). This unpopular policy was immediately dropped when Yugoslavia was expelled from Cominform in 1949. Expression of a Macedonian identity was criminalized in the 1960s, and the 1965 census was the last one to recognize the Macedonian language (Crampton 2005, p. 199). At the end of the war, pressure to leave was put on ethnic Turks in Southern Dobrudja (acquired in 1940), as Bulgaria wanted the land for collectivization. Some 156,000 Turks left in the early 1950s (Marushiavoka and Popov n.d., p. 46), although there were still about 750,000 remaining in the country in 1965 (Jelavich 1983, p. 368). After Bulgaria and Turkey reached agreements in 1968 on family reunification, an uncertain number (114,000 according to Marushiavoka and Popov) left for Turkey (see Içduygu and Sert, this volume). The largest and most dramatic flow of ethnic Turks to Turkey occurred in 1989, as the final stage of an aggressive assimilation campaign that had started in 1984, with the criminalization of Turkish dress, language and personal names. Some 350,000 Turks (including Bulgarian-speaking Muslims or Pomaks) fled to Turkey between June and August of 1989 (Içduygu and Sert, this volume). Emigration of political dissidents occurred primarily in the immediate post-war period—that is, until the early 1950s. According to one source of data, some 8,000 Bulgarian political refugees had been resettled in Europe by the mid-1950s, the largest communities in Yugoslavia, Greece and Turkey (IOM 2003, p. 13). Despite its strict adherence to the Soviet line, Bulgaria did encourage some temporary migrations. In particular, starting in the 1960s, Africans were admitted as university students. From the late 1960s, links with various African countries facilitated large temporary skilled migration flows out of Bulgaria for doctors, teachers, engineers and others (Crampton 2005, p. 195). By 1981, there were over 2,000 Bulgarian doctors working in Libya, for example, providing much-needed foreign currency and remittances. However, there was no regional Balkan labour migration either from or into Bulgaria prior to 1989.

Turkey during the post-war period became a major emigration country—in particular as Gastarbeiter in West Germany. In the 1970s, after the oil shocks, Turkish emigration started to be directed towards the Middle East (Adaman and Kaya 2012, p. 5). Turkey also shifted its immigration policy towards discouragement of immigration on the grounds that there was no demographic need. Some ethnic flows did occur, primarily from Greece and Bulgaria (see Içduygu and Sert, this volume).

1.7 The Post-Communist Political Upheavals

1.7.1 The Early Migration Phase

Unsurprisingly, the 1989 revolutions in Eastern Europe had major ramifications for the Balkans. With the collapse of the Iron Curtain, the region entered a new phase of population movements, encompassing ethnic, refugee and economic migrations. Perhaps less obviously, the first country to experience emigration pressures was Albania, which had been the most isolated for the entire post-war period. In June and July of 1990, some 5,000 Albanians sought refuge in western embassies in Tiranë (Vullnetari 2007, p. 31). Subsequently, thousands crossed the land and sea borders to enter Italy and Greece in an irregular fashion. The total number of those who ­illegally entered Greece is unknown, but is estimated at over 200,000 for 1991 alone (Fakiolas and King 1996, p. 176).

Bulgaria’s transition from communist rule followed the general pattern across Eastern Europe, although the fall from power of Zhivkov in November 1989 was more the result of a ‘palace coup’ than a popular revolution (Crampton 2005, p. 212). One of the first acts of the new regime in December was to revoke the ban on Turkish names. This led to the return of a third of the 360,000 who had left for Turkey earlier in the year. However, as a result of the sharp economic decline—especially affecting the ethnically mixed areas of Bulgaria—an additional 150,000 ethnic Turks left for Turkey over the period 1990–1991 (IOM 2003, p. 14). The removal of restrictions on travel also permitted ethnic Bulgarians to move in search of work—especially in Western Europe, the USA, Canada, and Australia. However, strict immigration requirements and Schengen visa controls meant that only the highly skilled could migrate legally. The majority crossed borders illegally (especially into Greece) and also were recruited for seasonal agricultural work in Greece (see Hatziprokopiou and Markova, this volume). This pattern of irregular and seasonal labour migration extended subsequently to Italy and Spain.

The collapse of the Socialist Federal Republic of Yugoslavia, and the complex and brutal civil wars that accompanied it, resulted in massive flows of refugees and IDPs (internally displaced persons) from and within the region. By the end of 1995, 350,000 Croatian Serbs had left Croatia; from Bosnia, more than 2.6 million were displaced and 1.2 million found refuge abroad. In 1998, 350,000 fled their homes in Kosovo, and the following year 350,000 ethnic Albanians fled to Albania, 250,000 to Macedonia and 70,000 to Montenegro. Conflict in Macedonia in 2001 led to 150,000 ethnic Albanians fleeing, mainly to Kosovo (Baldwin-Edwards 2005, p. 33). Other than the involvement of Albania and Macedonia in these flows, the impact on neighbouring Southern Balkan countries was not large. Relatively small numbers of refugees arrived in Greece and Turkey.

Greece had acquired an immigrant population over the 1980s—a complex mix of legal, semi-legal, and illegal migrants—from various countries including those in Eastern Europe and the Middle East. By 1990, it was of the order of 2–3 % of total population (Baldwin-Edwards and Apostolatou 2009, p. 235)—reflecting the country’s increasing political and economic stability. The Greek political reaction to irregular border crossings by Albanians en masse in December 1990 was highly negative, and reinforced by near-hysterical reports in the mass media which constructed a stereotype of the ‘dangerous Albanian’ (Baldwin-Edwards 2004a). Thus, a new immigration law was rapidly drafted to replace the previous one of 1929; its primary rationale was the allegation of criminality and the need to protect Greece from aliens. The new law made no practical provision for legal immigration, but implemented several new mechanisms of expulsion and deportation, as well as implementing major parts of the Schengen Agreement (Baldwin-Edwards and Apostolatou 2009, p. 235). Immediately, the Greek police began mounting regular operations known as skoupa (broom) to round up undocumented immigrants and expel them, generally to Albania. In 1992, 277,000 Albanians were summarily expelled without legal process, and 221,000 in 1993. From 1992 to 1995, 250,000–282,000 immigrants (predominantly Albanians) were expelled annually, although there were multiple expulsions of the same individuals. Reyneri (2001) has argued that this procedure was actually a form of circular migration, since Albanians who wished to get a free ride home would ensure that they were detected in the skoupa roundups. Small numbers of other nationalities were also expelled, primarily Iraqis, Romanians and Pakistanis (Baldwin-Edwards and Fakiolas 1998, Table 7). Despite these repressive measures, the stock of unregistered immigrants in Greece (primarily Albanians) climbed rapidly and by 1995 had reached an estimated 600,000 of which fewer than 100,000 had legal residence (Baldwin-Edwards 2004b, Fig. 1).

Turkey was a recipient of Muslim refugees from the Yugoslav war—primarily from Bosnia and Herzegovina and Kosovo (see Içduygu and Sert, this volume). While some ethnic migrations within the region continued (e.g., Muslims from Greece and Bulgaria), the predominant character of immigration into Turkey began to change into that of (unauthorized) labour migrations. In particular, small migration flows from Albania to Turkey started with the Albanian ‘embassies crisis’ of 1990, with later family reunifications and other flows. Starting from this period, the phenomenon of ‘trader-tourism’ also emerged as an important survival strategy: traditionally practised in East European counties, and involving Turkish goods, Bulgarians were particularly well positioned to engage in this economic activity (Konstantinov 1996).

1.7.2 The Consolidation Phase (1997–2008): Limited Peace, Regularizations and Border Controls

With the Dayton Accord of 1995, it looked as if the Yugoslav war period was over. However, in 1998 civil war broke out in Kosovo between Serb and Albanian Kosovars. Some 350,000 people, mostly Albanian Kosovars, fled as IDPs or refugees. The following year, 450,000 fled to Albania, 250,000 to the Republic of Macedonia and 70,000 to Montenegro. According to Kirişci (2001) about 18,000 fled to Turkey. With peace in 1999, some 600,000 returned to their homes in Kosovo alongside a reverse exodus of Serbs and Roma who fled to Serbia and Montenegro. Two years later, conflict in the Republic of Macedonia resulted in 150,000 ethnic Albanians fleeing—mostly to Kosovo (Baldwin-Edwards 2005, p. 33). EU governments have insisted on refugee returns to the region, although this simplistic policy choice has not been feasible for a variety of reasons. Around one million persons had returned by 2005, but this did not undo any of the ethnic cleansing created by war.Footnote 19

Over this same period, different migration issues emerged in the other countries of the region. For Greece, the issue was the ballooning of an irregular immigrant population, of which the majority were Albanian. For both Turkey and Greece, there was the problem of irregular transit migration. In both cases, the Northern European countries put political pressure on the countries concerned to deal with ­irregularity of all sorts. In late 1999, in the context of improved Greek-Turkish relations, an agreement was signed for cooperation in various matters of combating crime. This was supplemented by a protocol signed in 2001 detailing readmission procedures for irregular migrants. By this point, Turkey’s application for EU membership had stalled and Greek-Turkish relations were less comfortable. As a result, Turkey was reluctant to implement the protocol and accepted merely 3–8 % of the requested readmissions over the period 2004–2006 (Baldwin-Edwards 2006a, p. 120). Improved coastguard patrols from 1999 onward deflected irregular crossings from Turkey towards the Greek land border and the River Evros, but the numbers were still small in comparison with irregular crossings from Albania. By 2007, detected crossings from Turkey were beginning to approach those from Albania at 34,000 compared with 43,000 (Maroukis 2008, Table 16), and this rising trend continued. As the numbers of detected irregular migrants climbed from 2006, there was no change in the proportion re-admitted by Turkey; in fact, the highest readmission number was in the first year of operation, in 2002, at 645 returns (Içduygu 2011, Table 2).

The other major issue was that of visa requirements. Greece had joined Schengen in 1999 (fully implemented from 2000) and was required to enforce the Schengen rules with regard to visas. The other countries in the region followed different policies on visas and border controls. All of the countries of the former Yugoslavia retained visa-free travel with each other; Albania and the Republic of Macedonia granted visas at the border for each other’s nationals; Bulgaria granted visa-free travel to citizens of Macedonia (Baldwin-Edwards 2006b); and Turkey allowed visas to be bought at the border for citizens of almost all countries in the region, with the exception of Bulgaria. The visa requirement for Bulgarian nationals was waived in 2001, allowing a three-month stay for tourism (see Parla, this volume). This was occasioned by Turkey’s partial adaptation to Schengen; Greece also removed the visa requirement for Bulgarians in 2001 (see Hatziprokopiou and Markova, this volume).

Greece by 1997 had acquired an estimated immigrant stock of some 700,000 of which only 60,000 were with legal status (Baldwin-Edwards 2004b, Fig. 1). There was considerable political pressure on the government to regularize immigrants—since the mass deportations of 200,000 a year had failed to prevent rising numbers (Baldwin-Edwards and Fakiolas 1998). In 1997 Greece started a two-stage regularization programme, which yielded the first hard information about irregular immigrants in Greece. Out of a total of 371,641 applicants, 241,561 (65 %) were Albanians, 25,168 Bulgarians and 16,954 Romanians. The total number of persons covered (including family members) by the White Card awards was 462,067 with an estimated 150,000 who did not apply. Regularization programmes were held subsequently in 2001 (with a new immigration law), 2005 and 2007. Immigrants started to acquire a more secure presence in Greece. By 2006, the second generation of immigrants had reached an estimated total of 220,000 with Albanians at around 110,000–representing 30 % of Albanian residence permits (Baldwin-Edwards 2008, p. 38).

Albanian irregular emigration continued apace, especially to Greece and Italy, with another large exodus in 1996–1997 after the economic collapse associated with fraudulent investment schemes, the so-called pyramid schemes (Jarvis 1999). Whereas the original border crossings into Greece had been over the mountains, this was in fact a very dangerous route: mostly men arrived in Greece this way. Women and children started arriving in the mid-1990s, mainly through other routes—with underground visas, smuggling arrangements and even trafficking. By 2001, the predominant mechanism was the purchase of underground visas from the Greek consulate in Tiranë (Baldwin-Edwards 2004a, pp. 52–53). By 2010, the Albanian population had increased to an estimated 700,000 (Vullnetari, this volume). Barjaba (2000) formulated an ‘Albanian model of migration’, consisting of survival emigration, abnormally high emigration rates, high but decreasing irregularity, volatility, back-and-forth migration and strong instability in migrant behaviour. This model actually holds for most poor Eastern European former socialist states, including Bulgaria (Kupiszewski 2009, p. 446).

Turkey’s geographical position in the region has always made the country—and Istanbul especially—a major conduit for migration flows from the Middle East into Europe. In the 1990s the ‘Southern Balkan route’ through Turkey, then Greece and onwards to other EU countries, came under political scrutiny (Içduygu 2004, p. 309) with respect to irregular border crossings. The situation of Turks of Bulgarian origin is quite fascinating as a case study of both identity and adaptation between two countries, in particular, the shift in their status from that of potential citizens to systemic irregularity (Parla, this volume). Moreover, with changes in the Bulgarian citizenship law in 1998 and 2001, many returned to Bulgaria and applied for return of their Bulgarian citizenship. The resulting dual citizenship provided considerable advantages—in particular, involving border-crossing and trading of goods. Reports suggest that locals on both sides of the border resent what they see as the unfair privileges of dual nationals (Özgür-Baklacioglu 2006, p. 324–325).

1.7.3 Post Economic Crisis

Since 2008, the near-collapse of the Greek economy alongside economic depression throughout the region (with the exception of Turkey) has altered the character of migration flows. A large number of Albanians residing in Greece returned to Albania over the period 2007–2012. Greek statistics are incapable of revealing the extent of this, but reports from Albania suggest a figure of 180,000 returns (Kathimerini, 15 Jan. 2013), with potential impacts on the weak Albanian welfare system. Albanian border guard reports indicate that 64,060 Albanians returned from Greece in 2007, along with negligible numbers from other countries (Gedeshi and Jorgoni 2012, Table 1). Limited evidence suggests that the degree of rootedness in Greece has been the primary determinant of remaining. Thus, it is mainly single migrants (male workers) who return to Albania. Border apprehensions indicate that circular (irregular) labour migration to Greece has continued, alongside circular migration of those with valid Greek permits or citizenship, as well as authorized seasonal employment in agriculture (see Vullnetari, this volume). In the case of Bulgarian migration to Greece, Bulgaria’s EU accession in 2007 has made cross-border flows, and circular migration patterns, far easier. Thus, the labour markets of Bulgaria and Greece have become more interconnected, and flows of capital as well as people have increased correspondingly (Hatziprokopiou and Markova, this volume).

A similar pattern of increasing circular migration can be found in Turkey, where border ‘sticker visas’ have facilitated tourism-trading as well as tourism. From a recorded total of just over one million tourist arrivals in 2000, by 2009 Turkey was receiving 27 million tourists—of which 2.8 million were from the Western Balkans (Greece, Bulgaria, Romania, Albania and former Yugoslavia), 5.6 million were from the former Soviet bloc, and 2.2 million were from the Middle East (Erder and Kaşka 2012, Table 1). Some 32 % gave their reason for visiting Turkey as ‘business’ or ‘other’, instead of tourism activities or visiting relatives (ibid., Table 2). With a relatively open border policy alongside strict controls on formal employment and departure to EU countries, Turkey is an ideal country for informal economic activities of all kinds; this includes not only tourism-trading, but illegal employment of immigrants and acting as a staging post for irregular migration into EU countries. It is for this reason that the EU has been negotiating since 2005 to conclude a readmission agreement and force Turkey into greater conformity with the Schengen provisions—despite its not being a signatory to them (Içduygu 2011). A draft agreement was initialled in June 2012, and signed in December 2013; the EU is demanding full implementation before granting a visa waiver to Turkish nationals visiting Schengen countries.

Across the entire region, migration is characterized largely by short-term survival strategies in a period of weak economic activity, alongside a small residual of ethnic migrations. Sometimes, the two are combined. The small labour immigration into Albania, for example, is predominantly from Kosovo, Macedonia and Montenegro, as well as from Turkey (MARRI 2012, p. 56). There have also been some short-term and longer-term migration movements from Albania, Bulgaria and Turkey into the Republic of Macedonia throughout the last decade (Kupiszewski et al. 2009: Annex 1). The principal result of the economic crisis has been to push both people and governments into patterns of flexibility, in order to survive the economic downturn. The old state-centric model of controlled borders and permanent residents has lost its validity, at least within the EU and Balkan regions. The exception lies in control of borders concerning ‘outsiders’ from Asia and other lessdeveloped regions. There, the EU through its agency Frontex, is persisting with the old model, and trying to impose it on the EU candidate country Turkey, as well as overseeing its implementation in Greece and Bulgaria.

1.8 Epilogue

In compiling this book we have taken a historical perspective, starting from the late Ottoman Empire. We have adopted this approach partly because studying historical developments in the field of migration is interesting in itself and this makes us realize that international migration in the Southern Balkans is not a new phenomenon (as is sometimes argued). However, our primary reason for using a historical lens is because it aids us to better understand the present. The phenomenon of nation state formation in the Balkans necessitated a brutal process of classifying and ‘reorganizing’ the ethno-national groups inhabiting the region. Several means were used to reach the goal of relatively homogeneous nation states—such as ethnic cleansing, extermination, expulsion, forced assimilation and population exchanges. The enormous changes that these processes entailed can only be grasped if we realize how complicated the ethno-cultural and religious composition was in the Southern Balkans during the last century of Ottoman rule (see Sax map in Appendix). For a long period of time, migrations between the countries involved retained the character of ‘ethnic migrations’—members of ethnic minorities leaving their birthplaces to join co-ethnics in another country that was in many ways foreign to them. During World War II and the Cold War that followed it, the volume of migration within the region was restricted; international migration was directed mainly to Western Europe and overseas destinations. Whatever migration there was between the countries of the Southern Balkans continued to be of the ethnic type or consisted of political refugees, such as the Greek partisans who crossed the border at the end of the Greek Civil War. The major change in migration between the countries of our region came with the fall of the Berlin Wall in November 1989. Although the collapse of state socialist regimes led to a resumption of ethnic migrations, especially the ethnic cleansing of the Yugoslav wars, economic migrations also emerged. Greece became the focus of intra-regional migration, particularly from Albania and Bulgaria. That was a major change since it contributed to ethnic heterogeneity rather than homogeneity.

The first three chapters that follow deal with past migrations. In Chap. 2, Petko Hristov describes and analyses a traditional system of labour migration known as gurbet or pečalbarstvo as it existed in two mountainous regions of the Southern Balkans—Šopluk and Mijak. These traditional migration systems, which also existed elsewhere in the region, have disappeared; yet, despite their differences from modern labour migrations, the word gurbet is still used and the memory of these past migrations still plays a role in relating to the present. In Chap. 3, Raymond Detrez deals with one of the major population movements in the region in the early twentieth century—namely Bulgarians from Greece, Turkey and Romania moving to Bulgaria. He discusses the role of population exchanges in the process of nation building and irredentist policies in interwar Bulgaria. The refugees were used as tools for irredentist claims, and the process of adaptation to their new homeland was thereby retarded. In the following chapter, Nikolai Vukov looks more specifically at the history of Thracian refugees and their organizations in Bulgaria. He also pays attention to the revival of activities of Thracian organizations after the fall of the Berlin Wall, the renewed claims to compensation for lost properties and the continuing and revived memories of their homelands.

Chapters 5 and 6 are illustrative of the changes over time. The first of these, written by Ahmet İçduygu and Deniz Sert, tells the history of migration from the Balkans to Turkey from the end of the nineteenth century to the present. Like ­Detrez, they relate this history to nation building, but also to economic conditions and specific Turkish concerns, such as the perceived need for immigration to compensate for a declining population at that time. They also show that after 1990, ethnic migration decreased and irregular labour migration became more important. This last aspect is dealt with in great ethnographic detail by Ayse Parla in Chap. 6. Parla shows that over the past two decades the legal status of Turkish migrants from Bulgaria changed significantly. While in the past they were received as ethnic kin and prospective citizens, today they have become dispensable labour migrants moving back and forth between Bulgaria and Turkey.

Since the early 1990s, Greece has been the major pole of attraction for prospective migrants from the Southern Balkans, mainly Albanians and Bulgarians. This is why the last four chapters are devoted to Albanian and Bulgarian migration to Greece. Ifigeneia Kokkali asks why Albanian immigrants in Greece are so inconspicuous, why they seem to change their names and even their religion more easily than most immigrants. In trying to answer these questions she looks not only at discrimination, but also at how history has shaped conceptions about national identity among both Greeks and Albanians. She examines several aspects involved, such as the strong link between national identity and religion among the Greeks and the tradition of religious diversity and syncretism among the Albanians. In Chap. 8, Julie Vullnetari takes a very different look at Albanian migration to Greece. Her interest is in temporary or circular migration—a topic also addressed by others in this volume—and more specifically, seasonal migrants in agriculture. These migrants, drawn from the poorest strata in Albania, constitute an interesting segment of the Albanian population in Greece, specifically in view of the renewed interest in seasonal labour migration and its relation to the socio-economic development of ‘sending’ regions. In Chap. 9 , Riki van Boeschoten examines the renegotiation of gender identities among Albanian and Bulgarian migrants in Greece. She focuses on two major issues that emerge from the life stories of male and female migrants. The first is the empowerment of migrant women and disempowerment of migrant men, which seems to contradict the ‘patriarchal backlash’ in their home countries. The second is the striking differences between the gender identities of Albanian and Bulgarian migrant women. Van Boeschoten locates these trends against a backdrop of gender relations in Albania and Bulgaria and also the particularities of the migration process after 1990. In the final chapter, Panos Hatziprokopiou and Eugenia Markova examine the development of labour migration from Bulgaria to Greece over the past 20 years—placing it in the context of other forms of human and capital mobility in both directions. They argue that in this way the Balkan space is regaining the unified character it used to have in the Ottoman period. Greece, in particular, is also reacquiring some of the ethno-cultural diversity that used to characterize the Balkans—notwithstanding the recent popular hostility to immigration and immigrants, not only in Greece but in Europe as a whole.