Introduction: Magnitude and Complexity of Migration in the Middle East

The world's highest ratio of migrants to national population is to be found in the Middle East,Footnote ¹ and the region is one of the most fascinating arenas in which to observe international labor flows. Economic migration and forced displacement have led to the formation of a highly integrated regional labor market. Labor migration is one of the most dynamic economic factors in the Middle East, and remittances sent home by migrant workers in the region exceed the value of regional trade in goodsFootnote ² as well as official capital flows.Footnote ³ The growth of migrant labor in the Middle East was both rapid and massive and went hand in hand with the development of the oil economy. The stock of migrants went from 800,000 to 1.8 million between 1970 and 1975.Footnote ⁴ In the 1980s, the Middle East became the largest market for migrant labor the world has ever known, and, according to Sharon Stanton Russell and Michael Teitelbaum, just before the 1991 Gulf War the oil-rich states of the Arab Gulf taken together numbered more than seven million migrants, five million of whom were workers.Footnote ⁵ Migration to the oil-rich countries accounts for an overwhelming part of migratory trends in the region. In the Gulf Cooperation Council (GCC) countries, some 37.1 percent of the population is foreign, with the United Arab Emirates (UAE) hosting up to eighty-one percent of migrants, and foreign workers constituting at least seventy-five percent of the workforce in Saudi Arabia, eighty-two percent in Kuwait, and almost ninety percent in Qatar and the UAE at the beginning of the 2000s.Footnote ⁶

Since the mid-1950s, the region has also played host to one of the largest refugee populations in the world. Numbering approximately 900,000 in 1950, 1.3 million in the 1960s, 1.6 million in the 1970s, and 2 million in the 1980s, the Palestinian refugee population has grown steadily since 1947 to 4.5 million today.Footnote ⁷ Not including Sudanese internally displaced (IDP) and international refugees, the extended region itself holds more than six million refugees, who include some 4.5 million Palestinian refugees (exiles from 1947, 1967, and 1973 and their descendants) and approximately 1.5 million Iraqis in Syria and Jordan since 2006.Footnote ⁸ In smaller but significant numbers, the region has also been harboring refugees from the horn of Africa: Eritreans in Sudan, Yemen, and Saudi Arabia since the 1960s and the war of independence against Ethiopia, as well as Somali populations from Ethiopia and Somalia in Yemen since the 1980s and increasingly since 1991. Palestinians as well as other refugees (Eritrean, Somalis, Sudanese, etc.) from first to third generation are generally included in foreign labor statistics, notably in GCC states.

Considering the magnitude of the migratory phenomenon in such a highly strategic region, one may find it hard to understand the relative paucity of research on the politics of migration. Early work lamented the absence of migration-related research and data on the Middle East. Part of the gap in knowledge has been filled, but mainly for sending countries and mostly at the macroregional level. An extensive body of literature deals with the macroeconomic effects of labor migration in sending countries. The impact of remittances on local economies is still a question open to debate, and its evaluation in terms of development efficiency and sustainability remains controversial.Footnote ⁹ However, migration to the Gulf states is considered to have accelerated the development of Egypt and Yemen, of certain regions of India, Pakistan, Bangladesh, the Philippines, Indonesia, Sri Lanka, and Malaysia, and to have sustained the direct livelihood of many in the Palestinian-occupied territories as well as the survival of the economy itself. More detailed analyses on the sociology of labor migration are still to be done, especially in receiving countries. After a surge in interest in intraregional population movements in the 1970s and 1980s in the wake of the oil boom, migration research in the Middle East remained confined to applied economic research and international organization reports. One of the reasons for neglecting the impact of migration on Middle Eastern politics can be attributed to the fact that migration trends were expected to be temporary and meant to decline in the 1990s.Footnote ¹⁰ Since the 2000s, demographers, historians, and political scientists have taken a second look at one of the key factors of social and economic change in the region both in receiving and sending countries.Footnote ¹¹

The complexity of the underlying patterns of migration from and within the Middle East has long been emphasized,Footnote ¹² and different migration systems are centered on subregional zones (the Mashrek, the Arabian Peninsula, the Afro-Arab zone) with different emigration and immigration drivers (population, economics, security, international relations). The dominant features of mobility in the region nonetheless lie with two major factors: the oil economy and conflict. The labor-intensive development of the oil economy in GCC states and Libya since the 1970s accounts for massive levels of economic migration taking the form of contract labor, often deemed temporary, while the Arab-Israeli conflicts (1947, 1967, 1973), the first Gulf War (the Iran-Iraq conflict, 1980–1988), and the two subsequent ones (the Gulf wars that started in 1991 and 2003) account for persistent waves of refugees. Conflict is generally presented as a key determinant for population movements in the Middle East.Footnote ¹³ Be it the Palestinian exiles in the Gulf, the migrants’ exodus from Kuwait in the wake of the 1991 invasion by Iraq, or the Iraqi presence in Jordan and Syria after 2003, the main factor behind these short-lived and seemingly short-term waves of forced migration is war. More general structural and political determinants of labor-force mobility have rarely been explored.

In this article, we examine migration patterns in the oil-rich countries using the dual lens of war and political/economic factors and analyze the impact these patterns have had on regional integration. Considering the scarcity of sources on migration policy, asylum policy, and refugee law in a context where most macrohistorical migration trends can be accounted for by informal practices, one must complement administrative and political documentation with oral history and interviews. We focus on the empirical case study of the migration politics surrounding the Eritrean refugees during the war of independence against Ethiopia (1962–1991) using administrative archives to highlight the political dimension of labor migration.

Labor Migration in the Middle East: A Critical Narrative

The narrative of labor migration in the Middle East is well known and thoroughly documented.Footnote ¹⁴ The literature has largely emphasized the economic and demographic determinants of labor import,Footnote ¹⁵ and there have been many contributions to the analysis of the political economy and political demography of Arab intra-regional migrations in the 1970s and the 1980s.Footnote ¹⁶ Oil-rich states have long argued that their migration policies were in fact depoliticized along the line of classical economics, arguments in favor of laissez-faire policy.Footnote ¹⁷ Their deeds belie this argument: What they have in fact engineered is a political management of migration flows. Contrary to Nazli Choucri, who considers the initial phases of Arab labor migration in the GCC countries “individual,” “private,” and nonpolitical compared to a “state”-managed Asian labor import in the 1980s,Footnote ¹⁸ and in support of Sharon Staton Russell, we will argue that migration to the region has always been politicized along lines that vary across time and countries.

After the Second World War, the development of oil production in the sparsely populated Arabian Peninsula led to a massive increase in labor demand and an urgent need for foreign workers in Saudi Arabia, the United Arab Emirates, Kuwait, Qatar, and Bahrain. The oil-producing countries’ demand for labor was mainly met by regional inflows from highly populated neighboring Arab countries like Yemen and Egypt and, to a lesser extent, Jordan, Syria, and Iraq, or displaced populations like Palestinian refugees after 1947. Almost all categories of workers were targeted by public and private recruiters in the oil-producing states, from domestic laborers and construction workers to blue- and white-collar workers, in the private and public sector.

The shift in price-setting power for crude oil from Western-owned private or semi-private oil companies (ARAMCO or British Petroleum) to the Arab states of the Peninsula led to a massive increase in oil prices, further enhanced by the regional political context. During tafra (the period of high oil income from 1973 to 1986–1987), growth, state building, and massive inflows of mainly Arab immigrants shaped the countries political development. The steady increase in oil demand and the oil embargo of 1973–1974 generated enormous income for oil-producing countries' and put them on a rapid but “extensive” development path heavily reliant on labor import. Within a decade, oil revenues tripled (from two hundred billion dollars for the 1971–1975 period to six hundred billion dollars for the 1976–1980 period), and oil-financed socioeconomic development programs targeting infrastructures, education, industry, services, agriculture, etc., triggered a massive flow of contract labor into economies lacking a sufficient and adequately skilled or trained workforce. Most oil-producing countries therefore hinged heavily on foreign labor to achieve economic development. As a result, in the 1970s, seventy-two percent of the labor force in the GCC countries was foreign. The most cited studies on migration during the oil boom come from the numerous data surveys, policy-oriented reports,Footnote ¹⁹ and articlesFootnote ²⁰ conducted by the International Labor Organization's team of economists.

Interestingly enough, economic and demographic factors of migration fail to account completely for the counterintuitive variations in the volume of migration during the late-1980s economic recession. Analysts agree that the collapse of oil revenue caused neither a large-scale reexport of foreign labor nor a drastic fall in regional migration levels.Footnote ²¹ The lack of correlation between economic factors and migration is mainly due to political pressure on migration policies. Recession nevertheless saw a change in migratory trends. With migration patterns extending further and further “eastward,” immigration from Southeast Asia started increasing while Arab labor was either left unrenewed or simply replaced.

This change in the composition of the migrant labor force took place progressively. Asian workers had been present in the Gulf since the 1940s, notably in former British residency states. The circulation of indentured labor and the activity of the British East India Company in the Indian Ocean saw a first inflow of Commonwealth workers from Pakistan and India to the trade, service, and administrative sectors. The 1820 Treaty, which put Gulf Trucial states or sheikhdoms under British supervision for commercial and strategic matters, further accelerated this process. In the 1930s, the British imported large numbers of Indian workers in order to secure control of the oil sector in the Gulf.Footnote ²² In 1975 there were around two million foreign workers in the oil-producing states, sixty-eight percent of them Arab and the rest mainly from Asia, with a small proportion of skilled occidental workers. In 1983 the number of foreign workers increased to five million, of whom fifty-five percent were Arabs. The proportion of Arab workers kept declining throughout the 1980s and the 1990s as Arab nonnational populations continued to grow in absolute numbers in the GCC countries. The change was more brutal in some countries than others: Between 1975 and 1985, the relative share of the Arab foreign workforce in Saudi Arabia went down from ninety to thirty-two percent.Footnote ²³ In the 1980s, 44.3 percent of the three million Egyptian expatriates returned from Libya and the oil-producing countries, particularly in 1986 and 1989. Often called the “third and fourth phases” of migration,Footnote ²⁴ Asian migration became more complex in the 1980s and onward, including an increasing variety of nationalities (South Korea, Taiwan, Indonesia, Philippines, Thailand); this trend afforded the GCC countries more autonomy vis-à-vis Arab-sending countries on the regional and international political scenes.

As early as the 1960s, Arab immigrants started to be considered a source of political activism and a potential threat to the states and regimes of the GCC. Arab diasporas constituted a transnational network through which issues could travel and political action could be organized outside the purview of the host states, if not in direct opposition to them. The risk of “foreign agitation” was obviously put forward by GCC states in moments of crisis, as Robert Vitalis has shown for the 1950s labor unrest in Saudi Arabia's Eastern province. But the GCC states also came to see them as a more substantial or existential threat to national identity. Naturalization, which the oil-rich countries had opened to “ethnic” Arab migrants in the 1960s and the beginning of the 1970s, had become virtually impossible by the mid-1970s. Strict nationality and citizenship laws based on lineage were enforced, preventing non-nationals from gaining access to socioeconomic and political rights and cutting them off from the benefits of patrimonial welfare politics. The change in the composition of immigration and the selective preference given to Asian migrants in the 1980s may also be seen as an illustration of these “anti-integration” policies aimed at controlling the boundaries of the nation and excluding non-nationals from both the welfare system and the polity.Footnote ²⁵ Disenfranchised and likely to be “passive observers of political processes rather than potential activists or claimants on social services and other benefits of citizenship,”Footnote ²⁶ Asian workers were not meant to gain access to indigenous resources and political participation.

Beyond its economic rationale and the response to immediate market incentives, the selection of foreign workers illustrated a regional political strategy. Oil-producing states have justified their labor import policies on the basis of cost effectiveness. But along with the mechanic effects of push-pull factors and labor shortages, the dynamics of labor migration also require a thorough analysis of the policies and motivations that help determine the nature and volume of migrant labor flows. And in the case of the Arab region, politics as much as economic rationale seem to have shaped the trends of labor circulation.

Several “migration crises”Footnote ²⁷ occurring at different levels point to the political nature of migration. Most interestingly, their study lends credit to the idea, initially posited by Nazli Choucri about the 1980s phases of Asian migration, that the origins and composition of migrant worker populations should be interpreted as the joint result of public policies of both sending and receiving countries.

The case of Thailand is exemplary at the binational level. Yearly streams of labor from Thailand to the GCC rose from a few individuals to 105,016 between 1973 and 1982,Footnote ²⁸ with a large number of Thai workers going to Saudi Arabia, and continued until the Saudi government ruled out Thai immigration after a diplomatic clash in 1990. After a heist and the murder of three Saudi diplomats and a businessman in Bangkok, the Saudi government gave a mot d'ordre of not renewing 250,000 Thai workers’ visas and work permits in June 1990, signaling the exclusion of Thailand from a blooming migration circuit in Southeast Asia.Footnote ²⁹

The Gulf War of 1991 provides a massive illustration of the important role played by migration politics. In the wake of the invasion of Kuwait by Saddam Hussein, the Palestinian Liberation Organization (PLO) and the governments of Jordan and Yemen expressed more or less explicit support for Iraq. It not only proved to be a clumsy international positioning but also a dangerous political move for their citizens in the GCC countries. The 1991 “migration crisis” led to the expulsion of 800,000 Egyptians from Iraq, Jordan, and Kuwait;Footnote ³⁰ 731,800 Yemeni from Saudi Arabia (according to the Central Statistical Bureau in Sana'a); and close to a million returnees from other GCC countries,Footnote ³¹ as well as some 350,000 Palestinians from Kuwait and 100,000 from Saudi Arabia.Footnote ³² In total, an estimated two million Arab workers and dependents left the Gulf. This movement had massive repercussions on the political economy of labor migration in the Middle East.Footnote ³³ First and foremost, it accelerated the replacement of Arabs by Asian and Southeast Asian workers and clearly highlighted the political nature of labor imports.

Regional Integration and Labor Migration

Regionalism and regional integration have long presented difficulties for scholars struggling to capture the “evolving architecture of regionalization.”Footnote ³⁴ Most research on regionalism has gone beyond the analysis of geographical or politically integrated units in an effort to describe complex sets of interactions within groups of actors or zones. For that matter, Louise Fawcett contrasts the soft regionalism of the Arab world and the hard regionalism of the European Union, insisting on a spectrum that goes from one to the other, from a sense of awareness or community to the consolidation of pan- or subregional groups and networks through common institutions and organizations.Footnote ³⁵ Social-constructivist approaches applied to political identity in the Arab worldFootnote ³⁶ have helped define the political integration processes beyond the crude measurement of hard factors such as state power, intergovernmental cooperation, supranational delegation of sovereignty, and formal institutionalization of interdependence. In the Middle Eastern case, in order to evaluate the efficiency of economic integration processes, one needs to measure the degree of integration between labor markets provided through both formal and informal kinds of regulation. And this task requires, among other things, that labor migration be brought under scrutiny. Our interest lies in the overlapping of these two aspects of regional integration, and in this respect it is critical to take into account the politics of labor migration in the Middle East.

On the one hand, the literature on Arab regional integration generally laments the weakness of intraregional exchanges and the inefficiency of regional institutions. It tends to paint a “gloomy picture” of Arab integration, depicting instead the conditions of Arab “disintegration.”Footnote ³⁷ Regional integration is deemed to have failed, according to economic indicators and institutional criteria posited by the theoretical literature on regionalism.Footnote ³⁸ United Nations and World Bank reports confirmed in the late 1990sFootnote ³⁹ that indicators of liberalization of trade and economic interdependence were still low despite institutional proliferation, including the 1998 Greater Arab Free Trade Agreement. Between 1985 and 1990, intraregional trade accounted for only six and seven percent of total export from and import to the Middle East. The integration of labor markets since the 1960s, epitomized by the case of the oil-rich countries, is nonetheless generally understated as an economic factor of integration. One could argue that it has not only compensated for the failure of commodity market integration and weak corporate capital flows, but also that it has constituted a pioneering experiment in global labor-market integration that resulted from successive phases of globalization.

On the other hand, the most salient characteristic of Arab regional integration is the sense of community based on a shared language,Footnote ⁴⁰ a set of shared values, and the loose sense of collective identity linked to Arabness. Beyond essentialism, the politics of identity in the Middle East has been a key issue and a strategic factor in the social construction of the “Arab world.” The cultural dimension of Arab integration was massively represented in the literature in and on the region during the Cold War, especially as it served and illustrated the endeavor of pan-Arab politics. The politics of shared identity led Arab thinkers like Zaki al Arsuzi to evoke the possibility of an Arab regional citizenship that would unite all Arab national identities within an “exemplary republic.”Footnote ⁴¹

Discourses on Arab integration in fact dwell more on the ambition of a political project that aimed to unify and homogenize the region rather than on actual institutional integration.Footnote ⁴² Through the creation of the League of Arab States in 1945 and subsequently of the United Arab Republic (UAR) in 1958, Arab political leaders and transnational activists sought unityFootnote ⁴³ rather than integration. Arabism was presented as a challenge to a postcolonial nation-state-centered regional order based on ethnocultural grounds by Egyptian and Syrian presidents Gamal Abdel Nasser and Shukri al-Quwati, who signed the treaty founding the UAR. In turn, the UAR served as an instrument of hegemony for both “Arab republics” and for the Ba'ath party, enabling them to monopolize the legitimate discourse on Arabness and Arab integration for at least two decades. Historians of Arabism and analysts generally overlook the role of labor market integration within the wider process and focus on the intellectual and institutional history of regional integration. Three factors can account for this lack of attention. First, the main actors of migration management, the oil-rich states, came from countries that were not portrayed as key actors of Arabism. Second, migration politics are “low” politics compared to high politics and diplomatic endeavors. Third, most of the regulation and institutions relevant to labor migration were, and still are, semiformal or informal, bilateral at best, and failed to give birth to standing regional organizations. Oral history and the history of administrative practices regarding asylum and entry and exit policies open up an additional dimension of regional integration significant enough to reappraise the process as a whole.

The Informal Politics of Labor Migration in the Middle East: Migration as Diplomacy

The process of regional integration in the Middle East is atypical with regard to other models such as Europe or even the Americas: It has mainly been fueled by formally or informally regulated labor-force transfers between countries in the region.Footnote ⁴⁴ Therefore, we argue that migration policy should be analyzed as an indirect form of foreign policy that uses the selection of migrants and quasi-asylum policies as diplomacy.

At the regional level, patterns of worker mobility organized along the lines of regional and international politics have been a key to understanding what could be called the Arab “migration diplomacy.” This diplomacy is nonetheless to be understood as including formal and informal, public and private diplomacy. Bypassing the channels of formal institutions and agreements, the politics of workers’ circulation has been shaped by the decisions of both public and private actors, state administration, and embassies on a bilateral and temporary basis, and by firms and recruitment companies on a contractual business basis. Most sending countries have created public institutions to manage and organize labor emigration while at the same time retaining a public-private partnership in the management of migration flows: Overseas employment ministries, agencies, and offices flourished in Asia in the 1970s (the Overseas Employment Corporation established in 1976 in Pakistan and 1979 in Bangladesh, the Office of Overseas Employment Service Administration in Thailand, the Korea Overseas Development Corporation in South Korea). But similar institutions also started to appear in the Arab sending countries in the 1970s. The 1971 Egyptian constitution mentions the competence of the state in managing migration flows, and the 1981 presidential decree (n°574) defines the competence of the Ministry of Emigration Affairs and Egyptians Abroad until the Ministry of Manpower and Emigration was created in 1996 (presidential decree n°31) and the Supreme Committee of Emigration was created in 1997.

In the receiving countries, the politics of labor import from the 1960s to the 1980s was characterized by a clear preference granted to Arab immigrants in oil-rich countries that was not formalized by bilateral agreements or migration quotas. Nevertheless, the patterns of regional migration politics followed those of regional integration politics.Footnote ⁴⁵ In the context of Arab regional integration politics, fostered by pan-Arab ideologies, migration was the only field of actual integration between states and economies. Labor migration, its regulation, and the social, cultural, and financial flows that go with it are the very domain in which Arab integration has taken concrete form.Footnote ⁴⁶ Arab migrant worker flows and the remittances of capital they generated “have been the most important feature of regional integration”Footnote ⁴⁷ in the Middle East.

The sequence of Arab labor circulation and regional integration of labor markets lags behind the institutional history of Arabism. Five major historical phases are generally identified in the saga of Arabism: the intellectual premises (nineteenth and early twentieth century), the rise (from the Palestinian uprising of the 1930s to the Egyptian revolution of 1952), the consolidation (from the Egyptian to the Iraqi revolution of 1958), the decline (1958-1967), and the demise after 1967.Footnote ⁴⁸ The first wave of regional integration through labor migration came as a result of the war of 1947-1948 and the creation of the state of Israel, which caused the exodus of around 700,000 Palestinian Arabs, who relocated mainly in the then-Jordanian West Bank or Cisjordan area but also fled to Jordan, Lebanon, and Syria, and in smaller numbers to the Arab peninsula. The 1967 war generated subsequent waves of Palestinian refugees who either left the Palestinian-occupied territory or the first country of asylum to reach the oil-rich states and the better work opportunities they offered. Between 1947 and 1960, the population of Kuwait doubled, and the proportion of non-Kuwaiti in the population rose to fifty-three percent in 1965 with around thirty percent of all non-Kuwaitis being Palestinian.Footnote ⁴⁹ In 1975, twenty-five percent of the Saudi Arabian workforce was Yemeni (from North Yemen). Around sixty percent of the Jordanian labor force was working in the Gulf during the 1970s, including both Jordanians and Palestinians with Jordanian passports. The share of foreign Arab population in the oil-producing countries rose dramatically and started raising problems of national integration for the receiving states, which were later reformulated as issues of national security. The governments in Kuwait, Saudi Arabia, and the UAE started development policies in the late 1980s, raising requirements for participation by nationals in the workforce.

In the sending countries, exit visa procedures were lightened in the 1960s and 1970s. Egypt's political and cultural capital within the region directly benefited its diaspora.Footnote ⁵⁰ Egypt suspended previous restrictive emigration policies in 1973 while launching the infitah (open door) politics, and in 1974, exit visa requirements were suppressed for emigrants.Footnote ⁵¹ Egyptian migrants within the Arab world in the 1960s were predominantly skilled workers, mainly teachers and administrators. At the beginning of the 1970s, they joined with flows of Palestinian, Jordanian/Palestinian, and Syrian but also Iraqi and Lebanese workers, giving rise to a massive wave of both skilled and unskilled migration to the Gulf. In turn, the sending countries were also attracting other categories of immigrants. Egypt and Lebanon served as an educational hub in the region with prestigious universities for students and intellectuals from the rest of the region and Africa; Lebanon attracted unskilled workers from neighboring countries; and Egypt attracted Sudanese unskilled workers. The idea of an integrated Arab labor market was pervasive in the 1970s, and some policy makers considered treating the Arab region as a single territory, with mobility being one of the many forms of interaction possible between state units. In 1957 the League of Arab States commissioned its Economic Council to promote the principle of freedom of movement for workers, and in 1965 the first conference of Arab Labor Ministers endorsed the principle. In 1967 an agreement was signed by Jordan, Syria, and Egypt to call for free circulation of workers in the region.

None of the main receiving countries signed the agreement, and the oil-producing countries steadily refused to sign any binding convention concerning the access to their labor markets for foreign nationals. Even though oil-rich states participated in the Arab integration process in a subdued but extremely efficient fashion, they failed to reap the political benefits of this participation. The Arabian Peninsula was indeed considered a minor eulogist of Arabist politics compared to the vocal advocacy of Nasser's Egypt, Ba'ath-led countries like Iraq and Syria, or the socialist People's Democratic Republic of Yemen since 1970.Footnote ⁵² In the case of the oil-producing countries, pan-Arabism was, as Michael Barnett puts it, an “informal institution” rather than a set of organizations.Footnote ⁵³

The change in the selection of migrants engineered in the 1980s was also done with a view to securing more independence for oil-rich states from their traditional sources of labor. The political momentum of the oil-rich states can be measured through their ability to control migration patterns from the 1980s onward. The informal coalition was institutionalized in 1981 with the creation of the GCC states. Oil-rich countries proved able to implicitly shape the dynamics of regional politics. In 1983 fifty-five percent of the five million foreign workers in the GCC states were Arabs. As previously explained, the proportion of Arab labor slowly declined throughout the 1980s, and by 1995 Asian workers outnumbered Arab nationals by one million in the GCC countries.Footnote ⁵⁴ Putting cost-of-labor factors to the fore, which heavily favored the import of Asian labor, the oil-producing countries diversified away from the integrated Arab labor market and staunched the quasi dependence of certain sectors of their economies on Egyptian labor.

Migration policy is indeed an indirect instrument of foreign policy, and the opening and closing of borders on the receiving or sending side have been used as tools of political leverage, as shown by Nazli Choucri in the case of the Libya-Egypt migration couple.Footnote ⁵⁵ Beyond regional logics, bilateral meetings and negotiations between state officials, diplomats, businessmen, and recruitment agencies contributed to engineer the patterns of labor migration between “migration couples.” These processes hardly ever gave way to formalized and perennial agreements. Much more than the short-term political weapon they have often been used as, migration flows have proven in the long run to be a key factor shaping regional politics.

Migration Policy as Asylum Policy by Proxy: The Palestinian and the Eritrean Case

Today, more than eighty-eight percent of registered Palestinian refugees (4.7 million) are located in Arab near-eastern countries,Footnote ⁵⁶ and more than half a million of them were still in the GulfFootnote ⁵⁷ after the expulsion of approximately 450,000 from the region after the 1991 Gulf War due to the position adopted by the Palestinian Authority. Neither Saudi Arabia nor the Gulf states are party to the 1951 Geneva Convention, the 1965 Casablanca Protocol of the Arab League of States for Palestinian refugees, or the 1967 Organization of African Unity Protocol that guarantees protection and assistance to international refugees. However, the link between population dynamics and international violence has been discussed mainly within the context of the Palestinian refugee case, and the refugee flows in the Middle East clearly highlight the interconnection between migration and displacement.Footnote ⁵⁸ Indeed, the treatment of Palestinian refugees has become a central issue to the extent that it can be taken as symbolic of Arabism and pan-Arab politics as a whole. The equation between Arabism and the Palestinian cause started with the first exodus of 1948, as shown by Constantin Zurayq, the Syrian historian who first coined the word nakba. Footnote ⁵⁹

Even though Saudi Arabia was not a United Nations Relief and Works Agency (UNRWA) asylum country and did not recognize the Casablanca Protocol status for Palestinian refugees, the kingdom hosted Palestinian refugees as early as 1948. The state was simply applying the Protocols’ provisions de facto through a liberal entry and residence regime for Palestinian exiles of 1948 and 1967. As a token gesture of Arab solidarity and an act of informal diplomacy, the Arab managers of the ARAMCO recruitment sections opened an office in Beirut in 1948 in order to reach the Palestinian population. After 1967, large numbers of Palestinian refugees were recruited and sponsored by the ARAMCO, then a public-private Saudi-US company.Footnote ⁶⁰

To a lesser extent, Saudi Arabia supported the Eritrean cause in the war of independence against Ethiopia (1962–1991) by financing the guerrillas but also by facilitating entry and residence for exiles and guerrilla fighters. Arab countries in general, and the Gulf countries and Saudi Arabia in particular, supported the Eritrean guerrillas from the 1960s to the 1990s on an ideological basis and allowed Eritrean refugees to enter and settle in the oil-rich countries using migration politics as an asylum policy by proxy. Eritrean refugees were considered as Arabs oppressed by a colonial Christian power (Ethiopia) supported by both Israel and the United States until the Marxist revolution led by Haile Mariam Mengestu in 1974. Their struggle was understood in a geopolitical context and was seen by the Arab states around the Red Sea as a key strategic move toward protecting the “Arab identity of the Red Sea,” which they consider an “Arab lake.”Footnote ⁶¹ Rather than straightforward political support through explicit asylum rights, the stance of the Saudi government was to welcome Eritrean migrants; in turn, the financial flows of remittance from the kingdom became a crucial source of support for the freedom fighters.

Even though the historiography remains controversial on the emergence of an Arab political culture of the guerrilla, the intellectual, financial, and political connections between the first political movements promoting independence and the Arab world is clearly established. Arabicized elites from Eritrea gathered within the Muslim league in 1946, the Harekat Tahrir Iritriya (Eritrean Liberation Movement) in 1958, and the Eritrean Liberation Front, and started to claim autonomy from the Ethiopian domination that was formally established over the former Italian colony after the British army withdrew in 1952.Footnote ⁶² The essence of the political culture of the guerrilla has drawn the attention of historians and political scientists up to now.Footnote ⁶³ Primary sources and historical accounts in Arabic strongly defend the idea of an Arab identity of the independence struggle based on the support granted to the guerrilla by neighboring Arab countries. The discourse produced by political leaders in the 1960s and the 1970s was strongly influenced by Arabism and by the ideological resources it provides for claiming independence.Footnote ⁶⁴ Therefore, the presentation of Eritrea as an Arab country and of Eritreans as Arabs was acknowledged as a regional political commonplace, which led to the blooming of propaganda and pseudo-historical literature on Eritrea's “Arabness.”Footnote ⁶⁵

As a consequence of this political emphasis on the Arab identity of the guerrilla and of the Eritrean people, from the 1960s to the 1980s most of this Eritrean population, although already covered by the prima facie refugee status granted by the United Nations High Commissioner for Refugees, entered Saudi Arabia either on a pilgrimage visa or without any documentation at all. Two royal decrees issued in 1974 and 1979 temporarily granted the Eritrean Liberation Front (ELF) the status of kafil (sponsor) for Eritreans who wished to come and reside in the kingdom. Sponsorship being one of the main constraints on individual immigration for persons without a job contract, the royal decree therefore served as an “asylum policy” by proxy for Eritrean exiles. At the time, the residence permit (iqama) to be obtained through the Ministry of Interior was free of charge for Eritreans, contrary to the regular procedure. The iqama is issued on presentation of an ELF letter recognizing the petitioner as Eritrean.Footnote ⁶⁶

The 1979 Royal Decree was abolished in 1981, but various public services, including the Service of Passport and Immigration at the Ministry of Interior and the regional offices of the Ministry of Labor, remained particularly tolerant in their treatment of Eritrean “immigrants” till the beginning of the 1990s. Between 1984 and 1985, the status of kafil was briefly granted again to the ELF thanks to the outcome of the Jeddah conference, gathering all Eritrean independence parties and factions in exile.

The support for Eritrean exiles is embodied by a series of semiformal or informal administrative practices, which can be traced in the documentation gathered in local Sudanese administration where former Eritrean refugees registered after leaving Eritrea before migrating to the oil-rich countries.Footnote ⁶⁷ Eritreans could enter the country without a predefined job contract or prospective employer; the procedural burden of immigration registration was alleviated; and the public administration showed tolerance vis-à-vis undocumented migrants.

The Circulation of African Exiles: Travel Documents and Arab Politics

The concept of a regional citizenship was theorized by Zaki al-ArsuziFootnote ⁶⁸ in 1965 but never gave way to concrete administrative or political measures, such as an “Arab passport.” The Arab National Convention on Nationality, adopted on April 5, 1965, fell well short of Arsuzi's hopes by asserting that whoever possessed the nationality of any of the states of the League of Arab States was Arab (article 1). The definition chosen was tautological and merely institutional. It failed to address the question of infra- or supranational identity by referring, once again, to nation-states as sole providers of identity.Footnote ⁶⁹ In spite of the mention of Arab identity in preambles to all national constitutions in the region excepting Israel and Turkey,Footnote ⁷⁰ Arab nationals still needed passports or identity documents to circulate in the region. The processes of migrant and refugee identification in the region remain partially informal nonetheless, due to weak or rather inefficient state bureaucracies. In this context, rather than a rational-legal set of screening procedures assigning each individual a permanent label, identification has become a political process highly sensitive to representational variations but with very pragmatic outcomes, including the right to access a territory.

As Jane Caplan and John Torpey have demonstrated, the categories under which foreign nationals are labeled indicate the limits and extent of their rights, duties, and freedom of movement.Footnote ⁷¹ In the case of the Arab world in the 1970s and 1980s, the relative informality of registration and identification of migrants was linked to prevailing social, political, and cultural representations of the self and of others. The administrative procedures operate in a context in which formal bilateral or regional agreements on labor migration are virtually nonexistent.

In this context, it is particularly interesting to study the case of this marginal category of migrants in the Arab world. The history of Eritrean refugees in the Arab world illustrates the power of social representations of Arabness and indirect asylum policies. Through both state and nonstate activism, Eritrean exiles negotiated conditions of access and mobility for Eritreans in the Arab world during the 1960s and 1970s. The Eritrean transnational advocacy networks were managed by guerrilla parties like the ELF and the networks of partisan politics aimed at securing freedom of movement and military and financial support for the Eritrean war of independence. For this purpose, in the 1960s and 1970s parties in exile based in Cairo, Khartoum, Sana'a, and Beirut developed and communicated a political discourse proclaiming the Arabness of the Eritrean nation and the Arab identity of the people of Eritrea.Footnote ⁷² In 1964, during a press conference in Damascus, Idris M. Adem, Eritrean guerrilla leader in exile, asserted that “there are no others in Eritrea but Arab citizens,”Footnote ⁷³ and Eritrean political figures hammer throughout their discourses and writings the Arabness of Eritrea and of Eritreans.Footnote ⁷⁴

The consequences of this activism can be seen in the quasi-asylum policy that some Arab countries developed in favor of Eritrean refugees. They could enter Arab countries that did not recognize the mandate of the United Nations High Commissioner for Refugees (UNHCR) and the international refugee status linked to the 1951 Convention, and they were granted working and residence permits. Most Eritreans were registered as refugees in Sudan or Yemen, countries that had signed the 1951 Geneva Convention and hosted UNHCR officesFootnote ⁷⁵ and were given Convention Travel DocumentsFootnote ⁷⁶ (CTDs) that allowed them to leave their first country of asylum and travel around the region. In the context of pan-Arab politics, the government of Sudan issued vast numbers of CTDs that were valid in the Arab world, contrary to the prescription of UNHCR regulations.Footnote ⁷⁷ Most CTDs observed in the Commissioner of Refugees (COR) archives in Khartoum mention “Arab countries” or “Arab countries excluding Israel and South Africa,” which reflects the positioning of the League of Arab States in regional politics. CTDs were used as passports and were recognized and accepted as legitimate travel documents for Eritrean refugees, even in the oil-rich countries that had not signed the 1951 Convention but were the preferred destination of secondary migration for Sudanese-based exiles. COR archives host numbers of expired CTDs from the 1970s and the 1980s in cardboard boxes that display visas and stamps from oil-rich countries, some of them even bearing handwritten enumerations of the number, names, and ages of dependents. Some also bear the handwritten note “to accompany employer” alongside the visa and residence stamp delivered by the Service of Passport and Immigration of the receiving country. In this context, the CTD is not just the exact equivalent of a national passport, it also serves as a job contract and a residence permit; thus, the immigration policy can be understood as an informal asylum policy.

The liberal asylum policy of the state of Sudan gave rise to difficulties and even breaches of international law in the Middle East, where the right to a national passport and freedom to travel were much more restricted than a refugee's right to a CTD. Sudanese CTDs for Eritreans started raising suspicion in Western countries in the late 1970s, but they remained valid in the Middle East until the beginning of the 1990s and sometimes even beyond the end of the war and the independence of Eritrea in 1993.Footnote ⁷⁸ The analysis of the individual documentation of labor migrants, and in this case refugees, supports the writing of a social history of mobility that links micro and informal institutions to the structures and culture of Middle Eastern political history.