AustLII Home | Databases | WorldLII | Search | Feedback

Melbourne Journal of International Law

Melbourne Journal of International Law (MJIL)
You are here:  AustLII >> Databases >> Melbourne Journal of International Law >> 2011 >> [2011] MelbJlIntLaw 4

Database Search | Name Search | Recent Articles | Noteup | LawCite | Author Info | Download | Help

Panizzon, Marion --- "Migration and Trade: Prospects for Bilateralism in the Face of Skill-Selective Mobility Laws" [2011] MelbJlIntLaw 4; (2011) 12(1) Melbourne Journal of International Law 95



International migration management lacks a formal global institution, but in its place a multi-level system of norms is taking shape. This article finds that level of skill has become a more important selection criterion than nationality in post-9/11 immigration law reforms in Europe. The result is a skill divide, which is being further entrenched by the high-skill bias of the multilateral commitments in the so-called ‘Mode 4’ of the General Agreement on Trade in Services (‘GATS’) of the World Trade Organization. It finds that when it comes to the migration of high-skill labour — a scarce resource for which there is stiff global competition — countries increasingly opt for venues outside the regulatory constraints of traditional immigration law and prefer the mobility chapters of free trade agreements. This article combines international migration studies with trade regulation and human rights law to obtain a better understanding of how normative responses to economic migration differ, overlap and interact. Through an exposition and analysis of bilateral migration agreements, I describe their re-emergence in response to skill-selective immigration laws and trade agreements. I identify an ‘agreement dualism’, consisting of trade agreements attracting the highly skilled, corrected by bilateral migration agreements reintroducing nationality-based selection criteria at the risk of infringing the most-favoured-nation clause of GATS. The point of convergence in both regimes is their absence of migrant workers’ rights protection.


Migration is a ‘universal experience’ which affects the lives and wellbeing of far more persons than those actually crossing borders in search of better opportunities.[1] As a cosmopolitan phenomenon, migration touches upon many other opportunities and problems of humankind, including the welfare of states, the productivity of markets, poverty, development, trade, human rights, and the environment.[2] Unlike trade liberalisation and its related ‘linkage’ issues,[3] however, migration lacks a normative and institutional response that might address the universality of migration with sufficient unity. The ‘missing regime’ for international migration, in particular, the absence of an international architecture by which to enforce the substantial body of norms,[4] has been repeatedly deplored.[5] With the exception of a ‘thin layer of multilateralism’ addressing highly specialised issue areas, including asylum, refugees, travel, trafficking and the temporary movement of service providers,[6] no single international organisation has been able to bring under a single umbrella the various strands of this ‘global phenomenon’.[7] Even within the United Nations system, responsibility and capacity for managing migration is ‘diffused’ among different institutions.[8] There is no one treaty that comprehensively covers each step of the migratory process, spanning across departure, transit, admission, stay and return. Key areas of migration policymaking fall outside the mandates of the International Labour Organization (‘ILO’), the International Organization for Migration (‘IOM’) and the World Trade Organization, even if these organisations ‘could be understood to include significantly broader activities in this field’.[9]

Migration has often been viewed as the ‘last bastion of sovereignty’.[10] It therefore comes as no surprise that selecting the appropriate forum and level of rule-making primes the desire to mainstream the entire spectrum of migration norms (whether based on political, economic, environmental or family motivations) towards a single international normative framework. Nascent international soft law, including the common understandings of the Global Commission on International Migration, the guidelines of the International Agenda on Migration Management, the resolutions of the UN High-Level Dialogue on Migration and Development and the output of the Global Forum on Migration and Development, identify efficient practices which account for a first and inductive step towards bringing about normative, even if not institutional, coherence.[11] In contrast, international legal scholars suggest a deductive, top-down approach to unify the fragmented state of migration law by proposing to replicate, for the field of migration, the WTO experience of multilateralising trade liberalisation.[12] Thomas Straubhaar suggests a General Agreement on Movements of People,[13] Bimal Ghosh a New International Regime for Orderly Movements of People,[14] Jagdish Bhagwati a World Migration Organisation,[15] and Joel Trachtman has sketched a first draft of a General Agreement on Labour Migration.[16]

Scholars in international relations such as Alexander Betts and James Hollifield have drawn on global public goods theory to offer an explanation for this fragmented state of migration law.[17] They have found that multilateralising migration management finds its limits in the fact that migration, unlike trade, is not considered a global public good, although aspects of migration, such as refugee protection and the protection against smuggling and trafficking, come close to displaying the qualities of public goods.[18]

But, as Betts has well described, the public goods analogy does not hold for labour migration.[19] Not all labour migration is identical in terms of supply and demand.[20] Whereas the transboundary movement of highly skilled workers qualifies as a private good, the regulation of low-skilled migration amounts to a club good, the ‘governance benefits’ of which are, according to Betts and Lucie Cerna, excludable, but not rival.[21] Depending on whether the particular migrant labour in question is an abundant or scarce resource, states will opt for either restrictive regulation or facilitative liberalisation. In the case of low-skilled labour, of which there is abundant supply, states may wish to keep the benefits of regulated supply to themselves, whereas in the case of high-skilled labour, which is in scarce supply, states will exclude others from the benefits of having secured and facilitated access to this scarce resource. Consequently, in both cases the normative solution chosen will aim to prevent other states from enjoying benefits. The benefits, however, are not the same: for low-skilled labour the benefit is regulation, for high-skilled labour, the benefit is the resource itself. Overlaps in terms of the type of governance chosen exist to the extent that, for the management of both high and low-skilled labour, states opt for a governance framework that is unilateral, or at best bilateral, not multilateral.

The main difference between the governance of high or low-skilled labour relates to the type of market intervention chosen. In the case of high-skilled labour, the chosen norm liberalises and facilitates access to the supply; in the case of low-skilled labour, the norm restricts the benefits of regulated supply to the migrant destination country. To successfully compete for high-skilled labour against other migrant destination countries, many destination countries install skill-selective immigration laws or conclude bilateral migration agreements. Either legal instrument will be designed to enhance, rather than to restrict, the mobility of the highly skilled. The normative goal to be achieved by the bilateral governance of migration determines the legal design of the bilateral agreement: if used for tapping into the scarce resource of high-skilled migrants, bilateral agreements (‘bilaterals’) circumvent the constraints of domestic immigration law, by offering more attractive conditions of entry and stay. To keep the offer of low-skilled labour regulated and in check, bilaterals will provide for skill upgrades, pre-employment training and other capacity building incentives which serve to narrow down and qualify the type of low-skilled labour on offer.

Unlike for refugee protection, where free-rider issues arise, there is no collective action failure to be overcome in the field of labour migration.[22] There is therefore no incentive for the international community of states to conclude a multilateral agreement on labour migration. What has instead emerged is a ‘tapestry of bilateral, regional and inter-regional structures’.[23] This article will examine the interaction between these various layers and venues. It will focus particularly on the overlaps and parallels between trade and migration agreements. By introducing the term of ‘agreement dualism’, it seeks to explain and analyse the functions of these different venues that liberalise diverging segments of labour migrants, apart from graduate trainees and young professionals.

This article begins with the hypothesis that, from the perspective of global public goods theory, it is no coincidence that trade and labour migration are governed by two different types of agreements. Free trade agreements (‘FTAs’), including the WTO/General Agreement on Trade in Services (‘GATS’),[24] liberalise the high-skilled segment of labour migrants, but fail to regulate their entry, admission and return. If bilateral migration agreements open markets to lower-skilled migrant workers, they do so under heavily regulated conditions of admission, integration and return.

Having introduced the term ‘agreement dualism’ to describe the dual track of trade and migration agreements, this article goes on to examine the implications this may have on labour market structures. Against the background of labour market segmentation theory, I find that a key pull-factor for migration is labour market demand.[25] With trade liberalisation having lowered barriers to trade in goods and services, global demand for skills and talents has increased. The result is that a high-skilled migrant worker today has more in common with a native worker than with a lower-skilled migrant worker. In this ‘new’ labour market segmentation, lower-skilled migrants are subject to highly restrictive mobility schemes, while highly skilled migrants are exempt from similar regulatory constraints.[26] Thus, an increasing discrimination among migrants, rather than between natives and migrants, is identified.

The article then discusses the shift away from nationality as a selection criterion and towards discrimination along skill-levels, age, gender, education and professional experience. While there are upsides to skill-selectivity, there lies a downside in the fact that a lower-skilled migrant worker is exposed to a greater degree of regulatory intensity than their high-skilled counterpart, for the sole reason that their movement is unjustifiably associated with irregular migration.

In light of the prevailing view that the WTO/GATS system is structurally ill-suited to emerge as the new international organisation for migration, this article discusses how bilateral migration agreements have seen a renaissance in the past few years. It situates these ‘new’ bilaterals within a post-9/11 shift towards skill-selective national immigration laws in most countries in Europe, focusing on the ‘second-generation’ bilateral migration agreements which emerged post-9/11. ‘Second-generation’ bilaterals have replaced the tailor-made, nationality-specific preferences of their precursors by applying a one-size-fits-all solution.[27] This mainstreaming of preferences allows states to put all migrant source countries on a level playing field. In addition, the new bilaterals have a ‘prospective’ function in light of harmonised EU external migration policy.[28]

Studies of bilateral migration agreements are few and have often failed to situate these agreements in the wider context of global migration governance.[29] This article places bilateral migration agreements within the context of multilateral GATS Mode 4 obligations and domestic immigration law. It concludes by weighing the advantages and disadvantages of multilateralising labour migration management and inquires into the desirability and alternatives to ‘agreement dualism’. It finds that ‘agreement duplicity’ — the absence of human rights guarantees in trade and non-trade migration agreements — weighs more heavily on migrant worker discrimination than the skill divide which ‘agreement dualism’ has perpetuated.


Globalisation has changed the speed and scale at which migratory movements occur. The number of international migrants rose from 191 million in 2005 to 214 million in 2009, and this figure could double by 2050, reaching 405 million.[30] According to the ILO, migrant workers and their families account for 90% of these 214 million international migrants.[31] Economically active persons made up a total of about 105 million persons in 2010.[32] In OECD countries alone, the foreign-born population has grown by about 18% since 2000, accounted for by the rapid rise of North–South migration flows. The largest share of migrants from developing countries, an estimated 74 million (47%), reside in other developing countries. International migration is often portrayed as a ‘direct consequence’ of globalisation,[33] even if there is still less movement than at the end of the 19th century.[34]

Lower communication and transportation costs and the removal of barriers to services supply has accelerated the speed at which individuals cross borders today in search of better prospects abroad.[35] Trade liberalisation plays an important role in reducing the social, as well as the financial, distance between receiving and host societies.[36] In its wake, labour markets integrate at higher speeds. The WTO/GATS facilitates labour market integration by liberalising the temporary movement of foreign workers, as the fourth mode of cross-border services supply (‘GATS Mode 4’).

Beyond global public goods theory, two diverging theories explain the rise of migratory flows in the age of trade liberalisation. According to the Stolper-Samuelson model, it is the unequal distribution of production factors which accounts for migration.[37] Trade in goods and services can substitute for the unequal distribution of capital and labour among nations. The limits to the Stolper-Samuelson model are that production factors themselves have become more mobile, with trade flows bringing down the costs of migration to the effect that trade will no longer substitute for migration.

On the other hand, to proponents of the labour market segmentation theory, the main driver of migration is structural demand for foreign labour on the global labour market. Rather than the unequal distribution of production factors, chronic labour market demand for foreign workers creates a pull factor for migration.[38] The classic notion of the theory describes labour markets as segmented into a primary labour market occupied by native workers and a secondary labour market relegated to migrant workers or low-skilled natives. In between lays the narrow mid-level of migrant merchant middlemen.[39] Whereas natives have prospects of upward mobility, foreigners are found in secondary labour markets where jobs are dirty, dangerous and demeaning (commonly termed ‘3D jobs’).[40]

This paper expands on this classic notion of labour market segmentation theory. I argue that globalisation has created a strong global demand for skills and talents in labour markets.[41] Under this expanded notion of labour market segmentation, the high-skilled migrant occupies the same place as the native worker and competes against them on the primary labour market, while low-skilled migrants continue to occupy the secondary labour market.[42] This results in an emerging new divide, one splitting migrants along skill, education and occupational levels, rather than nationality.

The regulatory response to this labour market divide has been to treat qualified migrants more favourably in terms of entry and stay than their less qualified countrymen. This response is a by-product of post-9/11 immigration law reforms, which shift away from restrictive laws and towards (skill-) selective immigration laws.


In the previous section, I suggested that the unequal allocation of skills on the global labour market is the result of a structural demand for high-skilled labour. Trade liberalisation and the resulting integration of services and labour markets contribute to this unequal distribution of skills, causing peripheral countries to suffer from an undersupply of public services, such as health, education or utilities like water. Multinational companies affect such demand, because they use advances in technology and new communication channels to hunt globally for talent. The result is a new type of discrimination, one based on qualifications, age and gender, rather than on nationality, among the different categories of migrant workers.[43] This section focuses on demographic changes as one specific impact factor which underlies global market demand for high-skilled labour. It then explores how immigration laws have adapted to such labour market transformation. Particular attention is put on the paradigm shift away from restrictive, towards more skill-selective, normative frameworks.

Ageing baby boomers threaten the long-term productivity of the welfare state. Since the 1990s, the prospect of diminished productivity has increased competition for resourceful immigrants among destination countries. It has led countries to loosen rather than tighten their immigration laws, albeit only for select categories of foreign nationals. Positive incentive schemes were installed, to attract ‘productive’ categories of immigrants, including entrepreneurial, highly skilled or wealthy foreigners.[44] Throughout the Asia-Pacific, Europe and the Gulf, countries have moved away from restrictive immigration schemes towards more open but selective ones.[45]

In so doing, these regions started competing against traditional destination countries, such as Australia, Canada, New Zealand and the US, for the ‘best and the brightest’.[46] In Europe, points-based systems, which gave preference to the highly skilled and talented, were installed.[47] It is not surprising that a country like France, which traditionally has been open to family reunification migration, has introduced a ‘skills and talents’ admission card[48] and attempts to attract foreign students as a potential source of highly skilled migrant workers.[49]

Restrictive immigration policies of the 1990s emphasised the ‘adaptive’ qualities of the migrant and had assessed their integration propensities within a national society. The principle of ‘utility’ encourages selection based on nationality.[50] Skill-selective migration policies instead focus on the ‘competencies’ of immigrants, defined as their language, professional and life skills. Politico-societal rationales are no longer determinative. Rather, the determinative factors are now ‘market-oriented’ values, such as skills and age.[51] A paradigm shift towards market-based migration policy can be observed. Globalisation has loosened migration, like many other issues, from the grip of the nation-state. With the notion of ‘nation’ thus dissolving in favour of increasingly diverse societies, states have to look for an alternative to inform their migration policy. Like in trade or climate change, countries have found a replacement in associating more closely with the ‘market’. This same paradigm shift informs post-9/11 immigration law-making. It does not imply that the nation-state has pulled away from managing people flows. Instead, it means, as Ryo Kuboyama has put it, that the ‘modus of state interventionism’ has altered.[52] Through skill-selective policies, destination countries cooperate more closely with (private) market actors. Migration rule-making is overcoming the dichotomy that Rainer Bauböck described as migrants’ ‘real destinations being markets or cities rather than nation-states’.[53] As will be discussed below, the private sector is increasingly required to shoulder certain costs of migration management, such as skills-testing of workers or their (re)integration. This paradigm shift entails not less, but rather more, sovereign power for the state.

Even if such skill-selective immigration schemes are more desirable in economic and social terms than more restrictive models, the selectivity endorses a new divide — one among lower-skilled and higher-skilled migrants. As Antoine Pécoud and Paul de Guchteneire recall, ‘[m]obility is a privilege that is unevenly distributed among human beings’, but that in today’s globalised economy, a ‘different kind of inequality regards qualification. Today, trained workers are looked for by states and enjoy a much greater level of mobility than their unskilled compatriots.’[54] Adding to the skill divide are age-selective policies, which favour younger foreign workers over older ones, as the former integrate more smoothly into host societies. As will be discussed below, foreign students, graduate trainees and young professionals have seen their admissions fast-tracked over those of other categories of foreign workers.[55]

The downside of the high-skill bias of these new schemes is that they lead to a larger underground economy for low-skilled labour. Public services, such as entertainment, healthcare and domestic work, are jobs which national workers often refuse to take on. Employers operating in countries subscribing to skill-selective policies, like Switzerland, the UK and France, find it increasingly difficult to recruit lower-skilled migrants on the legal labour market. As employers’ demands for lower-skilled immigrants are not being catered to by official policy, they sub-contract work to irregular migrants. In turn, the enforceability of immigration law and national sovereignty over borders and territories is threatened.

Peter Schuck has described immigration law as a case in point for the discrepancy that Roscoe Pound identified as the one between ‘law-on-the-books’ and ‘law-in-action’.[56] Whereas ‘law-on-the-books’ reflects a policymaker’s decision about the direction of immigration law, ‘law-in-action’ describes the deviation from ‘law-on-the-books’ created by market demand or other policy constraints. In the field of migration, ‘law-in-action’ has often been related to the non-enforcement of immigration law.[57] In this article, we relate the term of ‘law-in-action’ even more specifically to the non-enforcement of readmission procedures, which is often the result of a laissez faire attitude displayed by countries of origin towards irregular stays of their citizens abroad. Schuck introduces the additional category of ‘law-in-their-minds’, which represents the visionary solution that the ‘losing’ actor — the one whose interests were ignored in the law-making process — would advance.[58] In our case, ‘law-on-the-books’ is well represented by skill-selective immigration law, whereas ‘law-in-action’ describes how the absence of loopholes for lawful entry of low-skilled labour in many skill selective schemes has led employers to recruit migrants in unauthorised stays.[59] In our scenario, the ‘law-in-their-minds’ represents the criticism that migrants’ countries of origin have voiced against skill-selectivity. Their concerns with such schemes are twofold: first, brain drain may be exacerbated, which is detrimental to their development interests and secondly, they have no outlet for their large pools of surplus low-skilled labour. Rather than sending only their ‘best and brightest’ to Europe or the US, countries in West and North Africa advocate expanding access to destination countries for their lower-skilled and often unemployed surplus workers.[60] Migrant source countries have thus sought ‘correctives’ to the high-skill bias of selective immigration law.

In response to such criticisms, countries like France or Spain have ex post facto corrected the high-skill bias of their national immigration laws by designing bilateral migration agreements, which partially reverse the skill-selective (point-based) system of their migration law in favour of nationality-based recruitment quotas.[61] In sum, the skill-selectivity of current migration law has required adjustments due to political pressure from source countries and labour market realities in destination countries. As a result, certain countries in Europe are considering watering down their skill-selective migration law in favour of providing more openings for low-skilled labour. An alternative to immigration law reform, however, is entry into trade agreements. These display certain advantages over national immigration law, which render such agreements a more attractive option for liberalising the mobility of highly skilled workers, in particular, service suppliers. The highly skilled migrant worker increasingly falls outside the formal channels of immigration law. Thus, there appears an increasing normative complexity composed of domestic migration law, trade agreements and bilateral migration agreements. Within this framework, ‘agreement dualism’ is emerging in parallel to, or outside the constraints of, domestic migration law. For occupations in high demand, market access will be liberalised through channels featuring the fewest possible barriers, such as trade agreements. For lower-skilled workers, bilaterals are used to restrict mobility.

What I identify as an ‘agreement dualism’ reflects the increasing bifurcation among immigrant labour, which is employed at either the high-skill or low-skill end of the job spectrum, with few occupations in the middle income range.[62] Moreover, the ‘agreement dualism’ also reflects what economists describe as the ‘polarisation’ between declining manufacturing employment and the concomitant growth of low-skilled services occupations,[63] the latter divide being stifled by immigration and female labour force entry.[64] It is not surprising that the legal responses tend to reflect both phenomena. FTAs and GATS liberalise the temporary movement of highly skilled services professionals, while opening markets to lower-skilled service suppliers and to some extent to workers in agriculture, manufacturing, and mining remains the reserved domain of bilateral, non-trade migration agreements.[65] While the international mobility of the lower-skilled worker is often over-regulated, usually in bilateral migration agreements, there are fewer restrictions if the highly skilled are targeted. Moreover, the liberalisation of movement of high-skill service suppliers occurs through FTAs, which contain little in terms of regulatory constraints on admission, post-admission and return/reintegration, as opposed to bilateral migration agreements which are biased towards restricting entry and encouraging return. ‘Agreement dualism’ creates inequalities in terms of access and post-admission rights amongst different migrant worker categories.

So far these new inequalities have received little scholarly attention, at least in law. What should states do to redress these inequalities? As Pécoud suggests, source countries would have to be compensated for the lack of labour market access for their lower-skilled workforces.[66] Destination countries could offer broader access for agricultural products from migrant source countries to make up for borders being closed to lower-skilled workers. The fact that the WTO offers the opportunity for such cross-sectoral trade-offs facilitates such compensatory adjustments.

The next section introduces the link between trade and migration and discusses what type of labour migration WTO members liberalise in the context of the so-called Mode 4 of GATS. By adding this level of analysis, I explain the extent to which it is desirable for the migrant worker and for the labour sending country to maintain ‘agreement dualism’, defined above as the divide between trade and migration agreements. I also ask to what extent trade and migration regimes ought to be brought into uniformity so as to overcome the discriminatory effects of the skill divide. In the final analysis, trade and migration agreements diverge less than at first sight. What matters more are their mutual deficiencies in the protection of migrant workers’ rights.


At the multilateral level, the WTO/GATS is the only international regime to treat the liberalisation of product (services, goods) and factor (capital and labour) movement under one single umbrella. Whereas historically, the stillborn International Trade Organization should have dealt with labour, its successor, the provisionally applied General Agreement on Tariffs and Trade (‘GATT’) of 1947 was more limited in its scope.[67] When the GATT was merged into the WTO in 1994, services trade was put onto the WTO agenda through GATS. In the context of services trade, the WTO was given the mandate to liberalise the temporary movement of natural persons. However, industrialised countries resisted attempts to grant the WTO/GATS jurisdiction over labour issues, such as direct employment of non-nationals by local firms, working conditions, immigration, residence and stay.[68] With its regulatory mandate for labour mobility having been short-circuited, the WTO is prevented from taking on a bolder position in the field of migration.

Alongside ILO core labour standards and the non-refoulement principle, the temporary movement of service providers is one of the few international obligations limiting national sovereignty over migration.[69] The temporary movement of natural persons which is to be ‘progressively liberalised’ under GATS art XIX, is defined in GATS art I:2(d) as the fourth mode of service supply, thus the commonly used abbreviation of ‘Mode 4’ of GATS. Mode 4 services are those that are delivered ‘through the presence of natural persons of a Member in the territory of any other Member’.[70] Despite GATS’ multilateral architecture, WTO members liberalise the temporary movement of service supplying persons on the basis of bilaterally exchanged commitments, which are then multilateralised in a second step. The next section will identify GATS’ main weaknesses, which have kept it from developing into a global migration regime.

A Architectural and Jurisdictional Limitations of GATS as a Global Migration Institution

The temporary movement of natural persons is only liberalised by GATS to the extent to which a WTO member has scheduled a horizontal or sector-specific commitment into its Schedule of Commitments. To date there are 153 such GATS Schedules of Commitments, one for each WTO member with the exception of the EU countries, which share a single EU-27 schedule.[71] Multilateral liberalisation under GATS functions through a bottom-up approach and is thus quite different from the top-down approach of other WTO agreements. Unlike the GATT, which liberalises trade in goods from the top down, there is, as Trachtman notes, ‘no market access for service supplying persons, unless a Member has committed to liberalising the temporary movement of services-related migrant labour’.[72] GATS does not require WTO members to open their markets to foreign labour, yet once a binding commitment is made, the foundational legal principles of GATS, such as most-favoured-nation (‘MFN’) treatment or transparency, attach. What makes GATS somewhat inadequate as a regime for managing migration is the fact that it is less concerned with WTO members actually removing barriers to the movement of persons than with ensuring that such removals occur reciprocally and if barriers persist, that they are applied in a non-discriminatory manner.[73]

Inversely, the fact that the WTO includes temporary migration within its ambit, albeit limited to services occupations, is a reference to the aborted International Trade Organization which should have included labour issues.[74] Mode 4 also stands as a concession towards developing countries asked to sign onto the Agreement on Trade-Related Aspects of Intellectual Property Rights (‘TRIPS Agreement’).[75]

Mode 4 of GATS has a narrow scope regarding persons whose movements are liberalised under GATS. First, persons seeking employment, as well as those in manufacturing or agricultural work, are excluded. Secondly, only the temporary, and not the permanent, movement of service suppliers is liberalised, even if GATS does not stipulate the duration of stay.[76] The Scheduling Guidelines of 2001 specify however, that each member should identify in its schedule of commitments the permitted length of stay for the categories of persons included therein.[77] Thirdly, market access (art XVI) and national treatment (art XVII) are conditional in GATS, which means that WTO members are free to insert qualifications, conditions and other limitations.[78] This flexibility to condition national treatment offers labour-receiving WTO members the opportunity to protect their domestic workforce from wage-downward competition by foreign service suppliers. The flexibility in the WTO/GATS to condition, limit and qualify national treatment and its attendant consequence of discriminating against a non-national service supplier in terms of wage levels, access to social welfare benefits, educational grants, subsidies, and other market conditions would not be possible under either international human rights law or international labour standards, in particular the 1998 ILO Declaration on Fundamental Principles and Rights at Work,[79] if these were properly applied in UN and ILO member states. Nonetheless, such in-built flexibility in the national treatment obligation of GATS is politically necessary to continue to motivate WTO members to progressively liberalise services trade in GATS more generally.

Despite the flexibility to protect domestic workers from foreign competition, WTO members have nonetheless entered commitments which focus on the highly skilled segment. This is the result of yet another type of flexibility. As the GATS Annex on Movement of Persons Supplying Services under the Agreement (‘Annex TMNP’) underlines: ‘Members may negotiate specific commitments applying to the movement of all categories of natural persons supplying services under the Agreement’.[80] The high-skill bias results from GATS only encouraging, but not prescribing, commitments, which span across the entire skills range.[81]

Evidence for the skill-bias in GATS is found in the categories of persons whose movement is liberalised: intra-corporate transferees account for the largest share (43%), followed closely by business visitors (24%) and the category of executives, managers, and specialists setting up a commercial presence (25%). The rest is divided up between contractual service suppliers (4%) and the category most WTO members describe as ‘other’ (4%).[82] The ‘other’ category is comprised of service providers who are not formally trained, who can be skilled, such as fashion models, sportspersons and artists, but who can also be occupied in the low-skilled, low-paid jobs with few opportunities for upward mobility, commonly referred to as ‘McJobs’.[83]

Another factor which accounts for the skill bias is the Special and Differential Treatment obligation of GATS art XIX:2. It allows developing WTO members an ‘appropriate flexibility … for opening fewer sectors, [and] liberalizing fewer types of transactions … in line with their development situation’.[84] The Doha Development Round’s well-meaning ‘Recommendations of the Special Session of the Council for Trade in Services’, which forms part of the Doha Work Programme and is otherwise known as the ‘July 2004 Package’,[85] exacerbated this asymmetric liberalisation strategy. Only industrialised country members were required to offer more market access in Mode 4: ‘In the services negotiations, Members shall implement the Less Developed Country (‘LDCs’) modalities and give priority to the sectors and modes of supply of export interest to LDCs, particularly with regard to movement of service providers under Mode 4.’[86] This strategy backfired precisely because it left industrialised WTO members without any expectation of reciprocity. All it did was further reduce the willingness of developed countries to offer commitments of interest to developing countries, such as for persons in lower-skilled or non-formally trained services occupations.[87]

Further exacerbating the skill-bias of GATS Mode 4 are proposals by emerging economies, such as India’s ‘Service Provider Visa’. Commonly known as the ‘GATS Mode 4 visa’,[88] it envisages fast-tracking visa procedures to facilitate the admission of contractual service suppliers, who are service providers dispatched abroad by a firm of the person’s country of origin. India’s proposal relies on ‘sending country’ firms establishing transnational business networks with foreign firms abroad.[89] LDCs, however, often fail to utilise globally active firms which could dispatch workers. Instead, LDCs advocate for workers being directly employed by destination country firms.[90] Domestic employment remains an unresolved issue in GATS, as is apparent from the ‘artificial’ distinction which art I:2(d) and the Annex TMNP make between a foreign national’s ‘foreign’ and ‘domestic’ employment. The prevailing opinion, which is also held by the WTO Secretariat, is that GATS was not intended to liberalise the direct recruitment of foreign nationals by national service suppliers.[91] According to some, however, the Annex TMNP seems to include foreign employees of a natural or juridical person of the host country (thus a domestic or local employer). Such a broad reading would serve developing country interests.[92] However, for the time being, WTO members’ commitments in Mode 4 of GATS have remained not ‘meaningful’ to developing country demands for more openings in lower-skilled segments of service suppliers.[93]

GATS Mode 4’s high-skill bias is another example of Schuck’s dichotomy of ‘law-on-the-books’, and ‘law-in-action’ in the field of migration. ‘In-action’, GATS is clearly a tool for countries to select the best and the brightest among migrants, despite the fact that ‘on-the-books’, GATS encourages temporary mobility of natural persons across the entire skill spectrum.

I also ascribe the skill-asymmetry of GATS to its architectural limitations, particularly the binding nature of commitments, the scheduling structure, the MFN clause and the immigration law carve-out. These features, which the following section describes in more detail, fail to offer the necessary incentives for WTO members to open their markets to lower-skilled service providers.[94] They also reduce the prospects for WTO/GATS to become a global migration steering tool.[95]

First, Uruguay Round negotiators took care to remove any resemblance between the temporary movement of natural persons and ‘immigration’, and did so by terminologically reducing the definition of the temporary movement of natural persons to a ‘mode’ of service supply (GATS art I:2(d)).[96] Secondly, ‘GATS Mode 4 remains embryonic’,[97] because the MFN clause of GATS obliges WTO members to multilateralise a market opening to all 153 WTO members.[98] The binding nature of the MFN discourages WTO members from offering market access commitments in Mode 4 in the first place,[99] because in no other area of trade are interests as emotive, culturally biased and politically-laden as with regard to the temporary movement of persons, where, as Antonia Carzaniga notes, ‘there might be greater willingness to be more open for some nationalities than for others’.[100] Thirdly, the dichotomy between the cyclicality of labour markets and the binding nature of GATS commitments[101] leaves WTO members without the ‘flexibility in and external sovereign control’ over services and labour markets, which are often exposed to economic downturns.[102] Fourthly, the ‘immigration law carve-out’ of the Annex TMNP excludes measures relating to residency, permanent migration, citizenship, border security, and visa policy from the scope of GATS.[103] This absence of jurisdiction over key areas of migration policy-making further reduces the utility of GATS as a global migration steering tool.

Selected FTAs have made advances over GATS in precisely these regulatory areas by imposing a return obligation on migrant workers, as in the 2006 Japan–Philippines Economic Partnership Agreement,[104] or by providing for pro-mobility visas to fast-track entry procedures for certain categories of workers. For example, the North American Free Trade Agreement[105] has a one-year, renewable ‘Trade–NAFTA’ visa for professionals (uncapped in 1994 for Canadians and in 2004 for Mexicans).[106] In addition, the Asia-Pacific Economic Cooperation operates a Business Travel Card for temporary business visitors, and in its FTAs with Singapore and Chile, the US has fast-tracked entry of professionals occupied in listed specialty jobs by offering the tailor-made H1-B1 visa, capped at 5400 professionals for Singapore and at 1800 professionals for Chile, even though the agreement liberalises market access for a much wider range of persons.[107] These FTAs thus come closer to fulfilling the far-reaching mandate to regulate migration with which some Uruguay Round negotiators had wanted to equip GATS. In so doing, these FTAs bridge the gap between the liberalisation-biased GATS Mode 4 and the security-oriented bilateral migration agreements.[108]

Unlike with bilateral migration agreements, GATS lacks the regulatory mandate to address the difficult issues associated with labour migration, such as overstays, brain drain and migrant worker exploitation. This also explains why WTO members have ‘lacked the comfort’ to liberalise trade in the services provided by low-skilled workers.[109] In lieu of GATS, destination countries will rely on other venues and formats, namely bilateral migration agreements, for managing lower-skilled mobility. Consequently, ‘agreement dualism’ is emerging as the global forum and governance format in the field of labour migration, reflecting the skill divide on the global labour market.

B The Added-Value of Trade Agreements: Bypassing Immigration Law Constraints

Trade liberalisation generates distributional inequalities, including the widening of wage disparities, which lead to the rise of migratory flows, at least in the short run.[110] With tariffs and non-tariff barriers disappearing, states can no longer rely on tariffs, technical and other non-regulatory hurdles to constrain the ease and speed at which migrants can move internationally. While restrictions on trade have gradually decreased, a parallel increase of the selectivity under which immigration laws admit foreign nationals is observed. The skill divide in migrant labour, defined as the diverging admission practices depending on the individual migrants’ skill levels, is one consequence of this broader phenomenon.

Departing from the hypothesis that the skill divide is a direct consequence of market demand reacting to trade liberalisation, this section analyses the extent to which the emergence of an ‘agreement duality’ between trade and non-trade agreements is a consequence or a catalyst of the selectivity in terms of types and categories of migrant labour admitted into a market. The notion of ‘agreement duality’ is thereby introduced to describe the parallelism and substitution effects between trade and non-trade agreements. The concept draws on a study by Maurice Schiff in 2006, who finds that migrant labour supply could be more effectively tailored to match labour market demand, and immigration more optimally regulated, by shaping the sequence in which trade and bilateral migration agreements are concluded.[111] This section queries the extent to which this duality is itself a policy tool that allows immigration services to respond to the multilateral reduction of barriers for trade in goods and services in the WTO, or whether it is rather an unintended consequence of trade and migration economics. It asks whether trade ministries have appropriated one increasingly important category of migration, the temporary mobility of service suppliers, because the multilateral trading system of the WTO in GATS considers the temporary movement of service suppliers as falling within the scope of trade; or whether the economic proximity of trade and migration generates market effects, which empower governments to respond with skill- rather than nationality-biased discrimination. The bifurcation of agreements available for steering foreign labour demand, with trade agreements targeting the high-skilled and bilaterals restricting the mobility of the lower-skilled under this second scenario, constitutes a policy response which decision-makers could not have more optimally tailored to trade the effects of market realities.

In the wake of trade liberalisation, destination countries were left with very few types of border barriers to control immigration. Since the 1970s they have thus made a stricter and more widespread use of remaining border measures, such as visas — one of the few barriers left untouched by efforts to multilateralise trade in the framework of the GATT 1947 and later on in the WTO. Specifically, Annex TMNP carves out immigration law from the jurisdictional scope of GATS. This carve-out in principle works to legitimise the discriminatory use of visa or other border barriers.

The immigration law carve-out of GATS generates two consequences. First, it allows countries to tighten or re-install immigration barriers which can keep ‘undesirable’ — economically inactive migrants — out.[112] Secondly, WTO members are encouraged to make use of GATS Mode 4 and FTAs as an alternative to domestic immigration law. Paradoxically, this carve-out of GATS, which was designed to strengthen domestic immigration law, in fact works to weaken it, at least when it comes to highly skilled workers.

Certain countries have been inspired by the GATS immigration carve-out to experiment with bifurcating the admission procedures so as to facilitate access for highly skilled service providers within, rather than outside, domestic immigration law. For example, Australia’s ‘e-visa’ facilitates the temporary entry of skilled key personnel (executives, managers, and specialists) outside the regular immigration system.[113] Sponsorship by a domestic employer acts as an in-built safeguard against disrespect of Australian immigration and employment laws.[114]

As shown, the absence of jurisdiction in GATS over immigration barriers, including visa policy, stay, and permanent residence, has its benefits. Trade agreements can be used to attract much needed human skills and resources, precisely because they are allowed to bypass immigration law constraints which would otherwise deter entry and restrict mobility. Another advantage of using GATS Mode 4 to liberalise the temporary movement of natural persons is the broader bargaining space of the WTO. Trade agreements, at least in their second-generation post-Uruguay Round (1994) templates, cover a wider range of issues (from market access for agricultural products over intellectual property rights to government procurement) than non-trade, migration agreements.[115]

This section has found that WTO members’ commitments in Mode 4 of GATS have exacerbated, rather than diminished, the skill divide and illuminated its adverse consequences for human rights. The ‘agreement dualism’ in the field of migration is explained by an economic and legal trigger: the economic effects of trade liberalisation on immigration policy formulation (skill selectivity) and the inclusion, under the umbrella of the WTO, of the temporary movement of persons in GATS. The immigration law carve-out of Annex TMNP (and the jurisdictional and more implicit architectural limitation of the scheduling structure of GATS commitments) further entrench the agreement duality, because these two legal limitations inadequately equip GATS with the tools to restrict human mobility where there is no demand (and often a surplus) on the global labour market. The consequence is a rise in bilateral migration agreements.


The most visible expressions of current global migration governance are bilateral migration agreements. Compared to ‘formal multilateralism’, which has shaped the refugee, labour standards and passport regimes of the interwar years, Betts finds that ‘bilateral or trans-regional structures … have evolved more rapidly’.[116] In the field of economic migration, ‘[t]he principal type of international arrangement, formal or informal, between states’ is, according to Trachtman, the ‘bilateral labour agreement’.[117] In light of the emerging network of bilateral migration agreements, one can no longer speak of a ‘total absence of international regulation [as one of] the most intriguing aspects of migration’.[118] Whereas their recent rise consolidates the international law of migration, scholarship has remained inconclusive as to their specific functionality.

Out of the 176 bilateral labour agreements that OECD countries had concluded by 2004,[119] approximately 25% remain unimplemented.[120] As Kathleen Newland notes, ‘formal, set, legally binding negotiated agreements are not the modus operandi of government networks’.[121] However, these findings are either limited to the narrower category of bilateral labour agreements, or alternatively to the more comprehensive bilateral migration agreements, which countries in Europe have been concluding since the late 1990s with their counterparts in Africa and Latin America. My hypothesis is that such broader bilateral migration agreements were installed to address the dual pressure of migrant source country critique over the detrimental effects of skill-selective immigration law for development of, and increasing market demand for, lower-skilled labour.

In response, bilateral migration agreements reintroduced nationality as a selection criterion so as to water down selection based on skills, education and professional experience. In so doing, the new bilateral agreements have corrected skill-selective immigration norms. Drawing on the example of France, this section assesses the extent of this ‘corrective’ role in light of the multilateral commitments in Mode 4 of GATS and post-9/11 immigration law reforms.

Bilateral migration agreements evolved out of friendship, commerce and navigation acts in the late 18th and 19th centuries.[122] The first wave of bilateral migration agreements emerged between 1950 and 1970. Some such ‘first-generation’ agreements compensated a former colony for the loss of free movement towards its former motherland and thus showcased an early corrective function. Others, namely the old ‘guest worker’ agreements, were aimed at recruiting the supply of migrant workers necessary to reconstruct post-World War II Europe or to stimulate the US economy (such as the Bracero programme with Mexico).[123] In the case of the Canada–Caribbean–Mexico Seasonal Agricultural Worker Program, the function was to delay technological innovation in sunset industries.[124] Since the late 1990s, a new set of agreements has been pioneered, mostly by countries in Europe bordering the Mediterranean, namely France, Spain and Italy. Such ‘second-generation’ agreements have incorporated pre-existing labour market access quotas that had been scattered among the various precursor agreements — such as those covering the exchange of graduate trainees, young professionals, or seasonal agricultural or fishery workers — into a single but comprehensive framework.[125]

In the field of labour migration specifically, bilateral agreements can have quite different, if not contrary purposes. For the highly skilled migrant, the purpose of a bilateral migration agreement is to eliminate as many barriers to entry as possible. Bilateral migration agreements for attracting the high-skilled basically relax the stricter requirements of domestic immigration law by easing visa requirements and eliminating economic needs tests. Low-skilled migration is characterised by a surplus offer of workers, which receiving countries aim to contain. In this case, bilateral migration agreements will seek out partnerships with the labour sending country. In contrast to the rather lean templates applicable to highly skilled, the typical template designed for managing the mobility of the lower-skilled is more comprehensive and multidimensional. Such bilateral migration agreements will offer cooperation on education, trade facilitation and development, in return for obtaining a concession by the sending country on border security, document security and overall cooperation with efforts aimed at restricting human mobility.

Policy-oriented compilations have systematically recorded the numbers and types of bilateral migration agreements across countries. Other research has identified bilateral migration agreements as substitutes or complements to multilateral or regional trade agreements.[126] Few of these studies, however, differentiate between pre-9/11 and post-9/11 templates, despite marked differences between the two. A normative evaluation of bilateral migration agreements, spanning across the spectrum of multilateral, regional, and unilateral venues of migration law-making, remains outstanding.

France has pioneered ‘second-generation’ bilateral migration agreements for Europe. It designed these in the wake of its immigration law reform of 2006/07.[127] The following section primarily draws on the French example to extract more general lessons on post-9/11, ‘second-generation’ bilateral migration agreements. A particular focus is placed on evaluating their interaction with the skill-selectivity of national immigration law and the Mode 4 commitments of GATS. A corrective, sovereignty-enhancing function, and a prospective, sovereignty-delegating function, are thus identified.

A The French 2006/07 Bilateral Migration Agreements

The French immigration law reforms of 2006/07 were guided by the aim of ‘selective migration’ (‘immigration choisie’).[128] They aimed to lower the number of family reunification migrants in proportion to the entry of ‘professional’ migrants. Thus, ‘French immigration policy [was] redirected towards encouraging economic migration and matching it better to the needs of the French economy’.[129] To this end, France actively sought, under new admission categories, the skilled and talented, intra-corporate transferees, and students.[130]

In the context of its 2006/07 immigration law reform under the guidance of Nicolas Sarkozy, then Minister of the Interior, France in 2006 began designing its second-generation bilateral migration agreements.[131] Between 2006 and 2010 France concluded 15 such agreements. Out of these 15, nine qualify as ‘classic’ versions and were concluded with Benin, Burkina Faso, Cameroon, Cape Verde, Congo, Gabon, Senegal, Tunisia and Lebanon. There are six ‘lighter’ templates without a readmission obligation — with Brazil, Russia, Serbia, Macedonia, Mauritius and Montenegro — because these countries already have in place a readmission agreement with either France or the EU. France is continuing to aggressively pursue a policy of multiplying the number of such agreements, so that more will soon be concluded.[132] In a similar vein to France, albeit earlier, Spain also started concluding bilateral migration agreements: between 2001 and 2010, agreements were signed with Algeria, Colombia, Cape Verde, Ecuador, Gambia, Guinea-Bissau, Mali, Mauritania, Morocco and Niger. Switzerland joined this competition for new migration agreements and concluded migration partnerships with Bosnia, Serbia and Kosovo in 2009, and with Nigeria in 2010.

Debate over the objectives pursued by bilateral migration agreements remains somewhat inconclusive. Some believe that they were designed to dissipate tensions with former colonies in West Africa, which had been disproportionately affected by the high-skill orientation of France’s 2006/07 immigration law reform. In that function, the bilaterals would ‘actively solicit’ low-qualified labour and complement the common interests France and source countries shared with respect to the management of migratory flows.[133] Others have found that France’s bilaterals further entrench the skill-selectivity of its new immigration law.[134] On a third view, bilateral migration agreements can be seen to qualify as ‘migration partnerships’ so as to more effectively manage the risks of migration, which destination countries cannot adequately address on their own. This partnership logic engages the country of origin’s responsibility for migration. The bilateral migration agreement thus sensitises the source country to the negative effects entailed by irregular migration, while focusing attention on the benefits that migration, if well managed, can have for development (such as remittances, skill and technology transfers, tourism and foreign direct investment).

Unlike the first generation agreements, which were one-dimensional in direction and content, these post-9/11 migration agreements are multifaceted policy tools. They must offer a broad enough platform for negotiating the types of trade-offs a balanced partnership requires. In order to entice a country of origin to share responsibility in migration, the bilateral agreement must address a broad panoply of issues, ranging from combating irregular entry and stay, to development and labour migration.

B Corrective ‘Migration Partnerships’

Unlike readmission agreements, which overwhelmingly cater to the destination country’s concerns about irregular migration, the goal of migration partnerships is to strike a balance of interests on a reciprocal and symmetrical basis.[135] For example, many countries of origin are interested in gaining market access for their surplus, low-skilled, non-formally trained and frequently unemployed workforce. To the extent that France’s new migration agreements offer lawful entry to those categories of migrants in which the source country has an ‘export’ interest, these agreements provide a certain degree of ‘migration partnership’. In the final analysis, the partner country’s bargaining power will determine the extent to which the agreement with France will correct the skill-selectivity of France’s immigration law.

French bilateral migration agreements correct the skill-selectivity of French domestic law in three ways. First, they add the admission category of young professionals to those admission categories which French domestic law opens towards third countries. The young professionals category is exclusively reserved for those citizens who originate from countries that have signed on to a bilateral migration agreement or an agreement on the exchange of young professionals with France. To say that France’s new bilaterals typify a partnership approach simply because they have added this admission category would, however, be overrating their ‘partnership’ value,[136] because young professionals hardly qualify as low-skilled migrants.

Secondly, skill-selectivity is corrected by the fact that a partner country is offered the possibility to add professions to the French occupational shortage list, which eliminates the economic needs test (‘ENT’) for third country nationals. This test, which acts as a market access barrier, requires a French employer to screen for a French or EU national before engaging a third country national for the job. Under French migration law, the ENT is eliminated if an occupation qualifies as a shortage occupation.[137] By concluding a bilateral migration agreement, the source country can add professions not already on the list to France’s list of 30 shortage occupations. Whereas the list of 30 should originally have contained around 60 professions, the number was intentionally lowered, so as to give countries partnering with France in a bilateral migration agreement the possibility to bargain for additional jobs to be put on to the list. In concrete numbers, the pact with Senegal added 108 professions, the one with Benin and Congo 15, the one with Gabon 9, the one with Tunisia 78, the one with Burkina Faso 64 and 88 are projected in the draft agreement with Mali.[138] A wider spectrum of jobs is thus open to citizens originating from a country that has signed a migration agreement with France than to citizens whose country is not linked to France through such an agreement. The agreements thus fulfil a developmental objective and more broadly comply with demands made by developing WTO members that fewer occupations be regulated by economic needs tests.[139] In this way, France’s bilateral agreements contribute to the relaxation of an important barrier to market access as promoted by GATS art XVI.[140]

As a WTO member, however, France is bound by the MFN provision of GATS art II. Under the MFN clause, France is required to treat service suppliers from all 153 WTO members equally in terms of market access, and must refrain from offering preferential access conditions (such as quotas, adding new categories of persons including young professionals, or relaxing or eliminating economic needs tests) to service supplying persons from select WTO members — for example, only to those which happen to cooperate in readmitting nationals, third country citizens and stateless persons in the context of a bilateral agreement. However, because WTO scholarship is inconclusive as to whether young professionals fall under the scope of GATS for the time being, France can continue to keep the category of ‘young professionals’ exclusively open to those countries which have signed onto a bilateral migration or a young professionals agreement with France. However, more at risk of infringing the MFN obligation of GATS is the fact that France’s bilateral migration agreements grant a relaxed ENT for additional occupations if a country cooperates with France in the framework of a bilateral migration agreement. By offering these countries the possibility of a relaxed ENT for more than the 30 professions that French domestic law has opened on an MFN basis, France’s bilaterals may be unduly discriminating against third countries. Because the bilateral migration agreement treats a certain third country more favourably in terms of market access than those which have not signed onto such bilateral deals, France may be in violation of MFN treatment under GATS art II.[141] However, France[142] — like Switzerland[143] and the US[144] — listed an MFN exemption in 1994, for Francophone countries, which would justify discrimination against third countries. In principle, such ENT exemptions ran out in 2004. However, to the extent that this preference still operates, a closer look reveals that it is potentially less in violation of the GATS MFN than at first sight. First, the lists eliminate the ENT, but only relax, rather than eliminate, the requirement to obtain a work authorisation. Secondly, the bilateral preference, defined as the opportunity to add occupations to France’s list of 30 occupations in the context of a bilateral migration agreement, eliminates the French and EU preference, but does not, in principle, allow these additions to deviate from the list of 150 occupations. A national from a country signing onto a bilateral with France will be exempted from competing against a French or EU national/resident, but continues to compete against a national/resident from a Newly Acceding EU Member State (‘NMS’) or from a third country, since the jobs opened on a bilateral basis must either figure already on the list of 30 occupations, or must be chosen from the list of 150 occupations. When selecting jobs to add to the list of 30 shortage occupations, France’s partner country does not have an entirely free hand. The bilateral preference is thus watered down by the EU preference. Consequently, the French agreements eliminate the domestic component of the ENT, but maintain the EU–NMS preference. A national from a country having signed onto a bilateral migration agreement with France will thus have to compete with a national from potential EU accession countries and NMS.

For example, out of the nine occupations added to the list of 30 occupations in the context of a bilateral migration agreement, four may already be part of the 150 occupations open for preferential access for potential EU accession countries and NMS. For instance, the pact with Gabon lists nine additional occupations which supposedly seem to be exclusively accessible to Gabonese citizens. A closer look, however, reveals that four of these already figure on the list of 30 professions open on a multilateral basis to all third countries. Yet none of these jobs are of ‘export’ interest to Gabon, since the Gabonese economy does not ‘produce’ many IT specialists, insurance experts, legal counsels in insurance matters, commercial banking specialists, or technical maintenance officers.[145]

In the final analysis, whether a French bilateral migration agreement qualifies as a ‘migration partnership’, and thus also caters to the interests of the source country, can be measured by the number of low-skilled occupations which figure on the list of shortage occupations. More precisely, if the jobs listed on the occupations list attached to a specific bilateral migration agreement showcase a low number of matches to the occupations which France has on its list of 30 occupations open to all third country citizens or 150 occupations open to NMS, chances are high that the sending country has had some say in the bilateral negotiations over the skill-levels required of, and jobs open to, its citizens in France. Inversely, if the jobs listed in the bilateral migration agreements closely match the professions and occupations in France’s generic list, we can safely conclude that the sending country did not have much bargaining power in negotiating labour market openings for its citizens in the bilateral agreement concluded with France.

A third way to correct the skill-selectivity of French migration law is by re-introducing annual quotas, which the immigration law reform of 2006 had abolished. While these apply to the different admission categories foreseen under French law, the quota ‘secures’ admission for citizens originating from a country that has signed on to a bilateral migration agreement with France. The prioritisation of labour market access for nationals from the partner country marks yet another return to nationality as a selection criterion. However, as will be seen below, these quotas are only indicative and thus, non-binding. There is no overall third-country quota for each admission category, out of which the ‘bilateral’ quota is given priority. Whereas this fact waters down their actual normative value for the source country, it diminishes the risk of their incompatibility with superior law, in particular the MFN obligation of GATS art II.

Finally, France treats nationals of countries with which it has signed bilateral migration agreements more favourably than other third country nationals in terms of extensions and renewals of admission cards, which can be harsher (such as in the case of skills and talents) or more lenient than under France’s migration law. However, like any other EU member state, France no longer holds the rule-making power to introduce short-term visas beyond the Schengen visa. Thus, France’s bilateral migration agreements cannot deviate from EU visa rules.[146] Annex 1 of the Common Consular Instructions[147] lists Benin, Burkina Faso, Cameroon, Cape Verde, Congo-Brazzaville and Congo-DRC, Gambia, Gabon, Guinea, Guinea-Bissau, Egypt, Haiti, Morocco, Mauritius, Mauritania, Moldova, the Philippines, Serbia and Montenegro, Senegal, and Tunisia as countries whose citizens need a Schengen visa for entry into the Schengen territory. Consequently, the French agreements on migration with Benin, Congo-Brazzaville, Senegal, Tunisia, and Mauritius cannot waive this requirement. Inversely, if the EU has concluded joint visa relaxation/readmission agreements with certain countries, which either lower the price of a Schengen visa from €60 to €35 or waive the Schengen visa for certain categories of persons, France cannot reintroduce stricter rules.

C Sovereignty-Delegating Public–Private Partnerships

The extent to which a bilateral migration agreement will correct the high-skill bias of French migration law will be determined on a case-by-case basis depending on the partner country’s negotiating power and, in the final analysis, on the sending country’s access to EU regional bloc formation. For example, if the migrant-sending country has access to EU (neighbourhood) policies it will succeed in obtaining a relaxed ENT for lower-skilled occupations. With regard to countries, like Senegal, Mauritius and Tunisia, France’s migration agreements do realise a measure of partnership because the skill-bias of French migration law is being corrected.

In contrast to their corrective function, which applies on a case-by-case basis, the prospective function of a bilateral migration agreement is not determined by negotiating power, but simply by the presence or absence of an EU-wide readmission agreement. The prospective nature of bilaterals relates to the difficulty of getting certain migrant source countries to sign on to collective, EU-wide readmission agreements. France’s bilaterals, like those of Spain, Italy and other EU member states, temporarily step in for the absence of an EU-wide readmission agreement. By agreeing to take back persons who are in irregular stays in France, a source country can at least temporarily avoid signing onto an EU-wide readmission agreement which entails taking back persons whose stay is not only irregular in just one, but in all of the 27 EU member states. The exception to this rule is the French bilateral migration agreement with Senegal (2006), which only stipulates a customary international legal obligation towards Senegalese nationals in unlawful stays in France (art 3), without requiring the readmission of third country nationals.[148] Other French bilaterals, such as the one with Cameroon (May 2009), require the foreign government authorities to take back their own citizens (art 3) and third country nationals (art 4).[149] Identical versions are contained in the agreements with Burkina Faso of 2009,[150] with Benin of 2008[151] and with Congo of 2007.[152] The absence of an EU readmission agreement often counts as a key criterion for France selecting potential target countries for its new bilateral agreements.

Source countries have often resisted deepening or widening their duty towards all 27 EU member states to take back categories of persons who are in irregular stays in an EU member state. These countries may find it easier to accept an EU-wide readmission agreement if, as a first step, they can conclude a bilateral migration agreement with an individual EU member state. In this first step, the source country will often deepen its readmission obligation in terms of persons covered, meaning that in addition to taking back its own nationals, it will agree to also repatriate third country nationals and stateless persons who are in irregular stays in the partner country.[153] At the conclusion of an EU-wide readmission agreement, the source country takes the second step, which is to widen this readmission obligation in terms of geographical scope to all 27 EU member states.[154] Although irregular migration, including readmissions, is one of the few areas of migration policy which has been communitarised according to art 79 of the Lisbon Treaty,[155] EU member states continue to have at their disposal some competencies in the field of readmissions. France, like all other EU member states, retains the power to conclude bilateral migration agreements with migrant source countries. This shared competency allows for precisely the type of sequencing readmission obligations which have proven effective for preparing the grounds for an EU-wide solution.

France’s new bilaterals do not only prospectively delegate sovereignty upwards towards the EU organs and other international organisations;[156] they also delegate cost-intensive migration management tasks downwards to private actors. Delegation of rule-making power to employer unions, industry associations and to private security agencies is a recurrent feature of post-9/11 migration agreements.[157] Aside from patrolling a migrant source country’s borders, tasks delegated to the private sector include preventing the falsification of documents, skills-testing, pre-departure training and accompanying the repatriation of unauthorised migrations. The delegation norm is usually contained in annexes, which form an integral part of the agreements. Based on a delegation norm, the government ministry will conclude private–public partnerships. Examples include partnerships with commercial banks to reduce the costs of remittance transfers or to grant tax breaks on migrants’ savings and partnerships with the diaspora to co-fund projects financed partially by the diaspora or partnering with industry associations to establish pre-employment training centres. Compared to their Spanish counterparts, the French agreements are particularly efficient at operationalising what Bhagwati labels the ‘diaspora model’, whereby the destination country shifts some responsibility for the root causes of migration and for irregular migration to the diaspora.[158] For instance, France conditioned its co-financing of diaspora investments upon the diaspora’s willingness to denounce those of their fellow countrymen and women who were in irregular stays in France.[159] This particular condition has been removed following intense criticism, and French migration policy instead engages, via tax benefits, the individual migrant workers to invest savings into projects in their country of origin. If a country signs onto a bilateral migration agreement with France, its tax break scheme will be extended to a wider category of persons than the one applicable under French law.

France’s bilateral migration agreements allow states to shift the costs of migration management to non-state actors. Consequently, bilateral migration agreements are not only multidimensional in terms of focus, but also in terms of actors. This institutional multiplicity sets the French pacts apart from their counterparts in trade and their precursors in migration. FTAs remain one-dimensionally focused on market admission and explicitly exclude the ‘collateral issues of management related to migration, such as recruitment, remittances and return’.[160] In this sense, bilateral migration agreements are more comprehensive migration ‘institutions’ than the ‘embedded governance’ model of the ILO, WTO or World Bank which deal with selected and fringe aspects of migration, such as cross-border supply of services or remittances flows.[161] The upside to delegating rule-making downwards to the private sector is the sharing of costs. The downside is the lack of parliamentary oversight and the consequential absence of transparency, including a destabilisation of human rights guarantees.[162]

D Multifunctional, but Asymmetric, ‘Migration Partnerships’

French migration agreements are multifunctional agreements, in terms of actors involved and substance covered.[163] This multifunctionality is a result of another emerging migration paradigm known as the whole-of-government approach. The whole-of-government approach requires that destination and source countries mainstream policies related to migration into a single understanding, so as to increase the coherence of migration policy formulation. In the EU context, the whole-of-government approach materialises under the so-called ‘Global Approach to Migration’, which has a similar purpose.[164] This policy mantra aims to avoid the recurrence of a situation like the one off the coast of Senegal, where EU fishery subsidies had put artisanal Senegalese fishermen out of their jobs. They turned to earning their living by smuggling individuals over the sea onto the Spanish coast, whom are then often intercepted by EU FRONTEX guards.[165] Bilateral migration agreements of the post-9/11 era tend to codify into binding treaty structures the whole-of-government approach.[166] The motivation to incorporate all policies necessary for effective migration management is driven by the desire to broaden the bargaining space for the destination and source country to arrive at a balanced, binding solution.[167] The question is whether, in practice, such bilateral migration agreements do in fact establish a balance whereby the destination and the source countries’ interests and concerns are equally represented.

Despite their comprehensiveness and the partnership rhetoric, France’s new agreements continue to be asymmetrically tilted towards France’s interests as a migrant destination country. Nowhere is this more obvious than in the fact that, for France, ensuring effective migration control outweighs their desire to secure workers for occupations where a shortage exists in the French labour market. This is apparent first from the fact that, currently, the largest number of labour migrants originate from Mali and Morocco[168] — precisely the two countries that are refusing to sign on to a migration pact with France. Secondly, the labour market access quotas granted in the agreements are only maximum benchmarks.[169] As such, they are indicative preferences, which encourage but do not bind French immigration authorities to attain a certain threshold, but not to go beyond it.[170] These maximum benchmarks are also rather low in terms of the number of persons when compared to the number of persons who have moved to France in the past. Despite correcting, in some cases, the skill-selectivity of French law, France’s bilateral migration agreements reinforce French national security interests above all other migration-related policies and thus strengthen French sovereignty over admissions.

E Preliminary Conclusion: Prospects for Bilateralism

I have traced the historical origins of the binary treaty law on economic migration. These origins are reflected in the high-skill bias of national immigration law reform, which at the turn of the millennium came to replace the family reunification schemes and quota systems of the post-WWII period. At the turn of the millennium, countries in Southern Europe, such as France, Spain and Italy, moved to curtail family reunification migration through wide-scale immigration law reforms. With demographically induced labour shortages on the rise in Europe and elsewhere, demand for lower-skilled migrant labour had to be considered alongside the policy goal of combating irregular migration. A policy mix had to be found which married both objectives in a mutually supportive way, that could then be sold to migrants’ countries of origin.

The policy outlets identified are bilateral migration agreements. Their advantage over domestic immigration law is that they operate on the basis of reciprocity. This allows for the opening of labour markets in exchange for cooperation in combating irregular migration. Unlike the one-dimensional focus of precursor agreements (guest worker schemes), ‘second-generation’ agreements correct skill-selective immigration schemes while preparing the grounds for EU-wide readmission/visa relaxation agreements and EU mobility partnerships. Yet, undermining their ‘partnership’ rhetoric and the label of ‘shared responsibility’ is the fact that these package deals betray an asymmetric, rather than balanced, exchange of interests, with the rationale of migration control and security dominating the market access interests of migrant source countries.

Nowhere is this more evident than when attributing interests to each of the three functions that France’s new migration agreements are said to fulfil. First, the value added by the bilateral migration agreements in terms of market access opportunities is quite low. While these have the potential to correct the skill-selectivity of French migration law, the level of additional market access is quite insignificant when compared to domestic migration law. For example, bilaterally agreed quotas are only indicative, rather than binding. Secondly, nationals from the partner country can be admitted without competing against another third country national in additional professions, but their admission procedure will not otherwise be fast-tracked, because France cannot autonomously deviate from Schengen visa rules. Thirdly, the agreements do not offer admission channels for categories of persons which domestic law does not provide, with the one exception of young professionals. Yet, where a source country’s negotiating power is sufficiently high, the bilateral agreement has the potential to correct the skill-selectivity of French migration law.

France’s new bilaterals prepare the ground for concluding EU-wide regional readmission agreements by expanding the categories of persons subject to repatriation obligations. In so doing, the balance of interests again tilts in favour of the migrant destination country, and of wider Europe. To the extent that France’s bilateral migration agreements prospectively delegate tasks to the private sector and upwardly to the EU, they realise a measure of ‘law-in-action’ of migration governance, reflecting the reality that the nation-state no longer controls the full extent of migratory movement. France’s new agreements balance the ‘law-in-their-minds’ of the source country against France’s common ‘law-on-the-books’.[171] In cases where this balance succeeds, France’s bilateral migration agreements contribute to unifying the fragmented laws of migration.


There are certain issue areas in the field of migration which cut across bilateral migration agreements and FTAs. Two significant duplicities are: the liberalisation of market access for graduate trainees; and the absence of protection for migrant workers’ human rights. Whereas the first such duplicity can be qualified as a ‘positive’ conflict of competency, the second one amounts to a ‘negative’ conflict of competency. Earlier, I introduced the notion of ‘agreement dualism’ to describe the split between different types of agreements (trade versus non-trade) over the level of skills, education or occupational function of foreign persons whose movement is liberalised. I also referred to ‘agreement dualism’ to describe the vertical level of regulation (unilateral and bilateral versus multilateral) at which a certain issue area in the field of international migration law is being managed. I found that in the absence of a free-rider problem in the field of labour migration (unlike in refugee law), destination countries regulate low-skill mobility — which is both highly on offer, and highly at risk of irregularity — through bilateral agreements. Inversely, high skill migration, which produces benefits over which nations compete, remains the domain of unilateralism with the exception of certain bilateral agreements and a ‘thin layer of multilateralism’ in GATS Mode 4.[172] While ‘agreement dualism’ remains the prevailing form of interaction in international migration management, there are instances of ‘agreement duplicities’, such as the absence of protection, in both FTAs and bilateral migration agreements, of migrant workers’ human rights.

Whereas the notion of ‘agreement dualism’ refers to parallels among international agreements, the concept of ‘agreement duplicity’ has been used to describe overlaps in the external dimensions of EU migration policy, in particular among migration agreements concluded by EU member states and those where the EU has acquired competencies in the field (like EU readmission and visa relaxation agreements or EU mobility partnership agreements).[173] So far, the notion of ‘agreement duplicity’ has been confined to EU studies where it relates to situations of concurrent or shared competencies between the EU and its member states in the field of migration, such as over labour market access for third country nationals and development cooperation.[174] In this article, I propose to expand the notion of ‘agreement duplicity’ to describe situations where there are positive and negative conflicts of rule-making authority.

The following section describes how within an ‘agreement dualism’ that splits along a skill divide into trade and non-trade templates, there are some overlaps in terms of coverage and absence of such coverage, and thus moments of ‘agreement duplicities’.

A Overlapping Competencies: Promotion of Education-Specific Migration

A relatively new phenomenon in the field of economic migration is the mobility of students and graduate trainees, liberalised by both FTAs and the bilateral migration agreements. The fact that both trade and migration regimes are converging over these two categories of migrants, demonstrates the importance of these categories for labour markets and more generally for productivity and welfare. Thus, an overview of these categories of workers is warranted.

The new French migration agreements liberalise three types of education-related admission channels into French territory: access to employment for foreign students, internships, and — within the category of ‘migration of professionals’ — the exchange of young professionals.[175] Young professionals are only admitted on the basis of reciprocity with the partner country. They constitute a category of admission conditional on the conclusion of a bilateral migration agreement or an agreement on the exchange of young professionals with France. In contrast, eligibility for the ‘access to employment for foreign students’ extends to all third country citizens who have completed their studies with a diploma equivalent to a Master’s degree and have sufficient means to support themselves in France, regardless of whether a migration agreement has been signed.[176] To be eligible, the diploma must have been issued by an institution linked to the French education system. This affinity to the French education system gives de facto preference to migrant source countries which happen to be former colonies of France. Finally, the intern or graduate trainee is the only one among the three types of admission channels discussed that was created exclusively by a French bilateral migration agreement. This arrangement is similar to the admission card granted to ‘intra-corporate transferees’, which, alongside the skills and talents visa/admission card, flags the shift towards a skill-selective immigration system in France.[177]

France’s bilateral migration agreement with Mauritius of 2009 establishes a sui generis version of the intra-corporate transferee admission card. Compared to the one established by French domestic law, this sui generis version allows more emphasis to be placed on the educational benefit of intra-corporate mobility, specifically when that movement takes place from a French firm’s subsidiary in Mauritius to that firm’s headquarters in France.[178] It is offered to Mauritian students desiring to move to France to pursue an internship at a firm in France, but also — and here is the difference from French migration law — to Mauritian employees of French firms established in Mauritius or of a Mauritian enterprise linked by a partnership to a French firm, who move to France to pursue an internship, if this internship includes both theoretical and practical education. The precondition for such internships is the conclusion of a tripartite agreement between either: the educational establishment in Mauritius, the French firm in France, and the student; or the sending firm in Mauritius, the French firm in France, and the employee. In both cases, these tripartite conventions stipulate the duration of stay, the content of the training module, and the conditions of stay, accommodation and social security in France.[179] Compared to the intra-corporate transferee visa established by the French immigration law of 2006, the Mauritian internship visa is broader in terms of eligible persons, but shorter in duration, as it is reduced from three years to one year.[180]

With a similar aim of promoting development through education, FTAs have added ‘graduate trainees’ to the workers they cover. This category has infrequently appeared in WTO members’ GATS schedules, but offers submitted in the course of the Doha Round of WTO negotiations signal that there is more to come. One reason for WTO members’ reluctance to open this category in Mode 4 of GATS may be that a graduate trainee is only partially motivated to move abroad to supply services. The trainee’s other rationale to move to another WTO member country is to be a consumer of the services, such as training and education, provided there. Graduate trainees thus combine the elements of a Mode 4-type movement with a Mode 2-type service supply.[181] Under the narrow definition of Mode 4 in art I:2(d) GATS, graduate trainees fall outside the category of a Mode 4 service supplier, because they often work only part-time and receive on-the-job training courses or schooling during the rest of their time. Instead, they constitute a category of quasi-service supplying natural persons. Pursuant to the EC-27 Schedule of Commitments in Mode 4 of GATS,[182] graduate trainees are only allowed to move from headquarters to a subsidiary or between subsidiaries.

The EC–CARIFORUM EPA of 2008 liberalises the mobility of graduate trainees in a similar way to France’s agreement with Mauritius — namely from a subsidiary in a Caribbean country to the firm’s headquarters in Europe.[183] It is the first trade agreement to establish this new type of movement and thus to expand on the developmental potential of the admission category of ‘graduate trainees’. This additional type of movement is the only one offering skill-upgrading possibilities to graduate trainees from Caribbean Forum (‘CARIFORUM’) countries and thus delivers an important development dividend for the CARIFORUM economies. The definition implies that a domestic employer domiciled in the EU will recruit graduate trainees from the CARIFORUM countries, as opposed to a foreign firm, and is likely to fall outside the scope of GATS. Consequently, the EC is under no obligation to generalise this opening to all WTO members on the basis of the MFN clause of GATS.[184]

Graduate trainees, students, and participants in professional exchanges or stagiaire programmes were a traditional domain of the ‘first-generation’ bilateral migration agreements. Dating back to the 1950s when colonies gained independence from their motherlands, ‘first-generation’ migration agreements compensated a former colony for the loss of free movement with the motherland. Such compensatory adjustment included liberalising mobility for graduate trainees, but also for scientists, technical personnel and politicians. If graduate trainees were to fall under the scope of GATS, a very significant number of bilateral agreements would be inconsistent with the MFN obligation and would need to be revised.

A threefold policy agenda is pursued by liberalising the movement of graduate trainees, students and young professionals: first to tap a resource for potentially highly skilled foreign labour without the host country engaging directly in brain drain; secondly to recruit persons whose potential for integration is higher than that of older migrants; and thirdly, to generate revenue for local destination country education services. The category of graduate trainees is simply a further means to promote skill-selectivity in migration, with a time lag. As mainly younger persons will qualify as graduate trainees, this category not only discriminates in terms of age, but reinforces the skill-selectivity of migration under the cover of promoting development through education.

B Negative Conflict of Competencies: Absence of Individual Rights Protection

Trade and migration agreements have one commonality, which is their failure to protect migrants’ human rights. Neither type of agreement codifies human rights guarantees, which would protect a migrant’s movement or post-admission rights.[185] The absence of human rights guarantees is more critical for agreements which codify readmission obligations, such as France’s bilateral migration agreements. Human rights protection plays a crucial role in readmission processes. Readmission obligations trigger situations which may be human rights-sensitive in terms of the non-refoulement guarantee. Non-refoulement, which has the status of customary international law, sets limits to readmissions, because it prohibits sending migrants back to their country of origin or to any third country if their life and health are threatened there.[186] Where a source country is to readmit its citizens, third-country nationals and stateless persons who remain in unlawful stays abroad, the bilateral migration agreement should expressly refer to the universal treaty or international agreement, which binds both the migrant-sending and receiving country to non-refoulement guarantees, and also grants the rights to fair process. Most bilaterals, however, simply contain a blanket reference to human rights. They fail to expressly list the human rights treaties to which the contracting parties of the bilateral agreement have agreed to abide by. Without a clear-cut reference to a specific human rights treaty, an affected migrant is unable to raise a claim against the violation of her non-refoulement guarantee or another human right.[187] Out of France’s new bilateral migration agreements, so far only the agreement with Senegal (2006) expressly lists human rights instruments. It refers, in art 4, to the 1951 Convention on the Status of Refugees, the 1967 Protocol Relating to the Status of Refugees, as well as the 1950 European Convention on Human Rights specifically for France[188] and the 1981 African Charter of Human Rights for Senegal.[189]

In terms of movement rights, France’s new bilateral migration agreements may be at risk of infringing upon the basic human right to leave any country, including one’s own, which is codified in art 13 of the Universal Declaration of Human Rights.[190] France’s bilaterals may not be explicitly violating this basic tenet, but most of them refer in their preambles to art 13 of the EU–ACP Cotonou Partnership Agreement (‘CPA’),[191] the bias of which may amount to such a violation. What appears on its face to enhance the mutual supportiveness between the CPA and a French bilateral migration agreement, risks infringing movement-related human rights. Article 13 of the CPA calls on the African, Caribbean and Pacific Group of States (‘ACP’) countries (most of which happen to be migrant sending countries) to join France in border patrol operations to secure EU borders. In essence, France’s new agreements operationalise art 13 of the CPA to the extent that the migrant-sending country is required to close its borders to its own citizens who might seek to leave for Europe. This may amount to a violation of the human right of any person to leave any country, including one’s own, according to art 13 of the Universal Declaration of Human Rights.

In conclusion, the new bilateral migration agreements concluded by certain countries in Europe fail to stipulate post-admission human rights (such as non-refoulement) with sufficient clarity and may potentially be infringing upon movement-related human rights as well.

The situation is different for FTAs which, with the exception of the CPA, have not been restricting mobility to such an extent as to potentially amount to an infringement of movement-related human rights. Most FTAs are restrained by architectural (Schedules of Commitments) or jurisdictional (immigration law carve-outs) limitations from incorporating readmission obligations. Thus, these agreements are not at risk of incompatibility with basic movement-related or post-admission human rights tenets, with one notable exception.

The flexibility inherent in national treatment under GATS art XVII allows a destination country to treat foreign service supplying persons less favourably than native workers. Inversely, human rights instruments — with the 1998 ILO Declaration on Fundamental Principles and Rights at Work[192] and the 2006 ILO Multilateral Framework on Labour Migration[193] being at the forefront — protect against the discriminatory treatment of a migrant worker in relation to exploitation, socio-economic integration, continued education, training and skills-upgrading (in the context of intra-corporate transfers), geographic mobility, the right to change employer, and access to social security. The ILO instruments thus stand in contrast to the GATS national treatment provision, which allows countries to treat foreign service suppliers less favourably than native ones, as long as the WTO member has scheduled such limitations and qualifications in its binding commitments under art XVII of GATS.

Under GATS, WTO members are thus allowed to discriminate in terms of national treatment against non-nationals, whereas such discrimination would neither be possible under international human rights law nor ILO International Labour Standards if these were properly applied in UN and ILO member states. To remedy this conflict of norms, it is suggested that trade agreements be interpreted in the light of ILO International Labour Standards.

Most FTAs with a development dimension — such as the CPA of 2008 (arts 9 and 96) and the EC–Chile FTA of 2002 (art 44),[194] but not the EC–Mexico[195] or the EC–South Africa TDCA[196] — refer to labour standards.[197] All of the US FTAs require the partner country to sign on to labour standards so as to guarantee national treatment on that front. Yet, the risk of inequality of treatment between the foreign and domestic worker remains even in those cases where labour standards have been incorporated into the treaty text. In none of the abovementioned agreements is national treatment defined with sufficient precision so as to afford a concerned individual the right of defence under non-refoulement or due process. With the national treatment clause falling short of displaying self-executing effects, the individual service-supplying person will not have the right to complain against an employer violating her minimum labour standards.

In terms of movement-related rights, FTAs usually do not contain readmission obligations or otherwise restrict human mobility. These are therefore less at risk of infringing on movement-related rights than bilateral migration agreements. The situation is different for economic partnership agreements, such as the Japan–Philippines EPA, which lists return obligations for certain categories of workers.[198] Should these return obligations have direct effect — that is, be formulated with sufficient clarity and precision so as to create an obligation to return for the individual migrant worker (in addition to the obligation for the sending country) — the destination country (Japan) should ensure that the migrants’ return procedure occurs within the tenets of fair treatment, due process, and the non-refoulement guarantee. What goes for bilateral migration agreements also applies to EPAs, which may increasingly experiment with incorporating return and readmission obligations to ensure that the movement of persons remains orderly and lawful. EPAs ought also to be particularly sensitive to guaranteeing post-admission rights with sufficient clarity. As Alexander Aleinikoff notes, despite the migration process being built around a ‘triangular relationship among a person, a sending state, and a receiving state’, on the human rights protection issue, GATS Mode 4, FTAs and bilateral migration agreements are all largely deficient.[199]


Migration remains a key distributional problem of economic globalisation, and its management has become a high-profile issue of international politics given its implications for trade, travel, development and education. Economic globalisation and demographic demand have changed the patterns and processes through which migration takes place and trade relations unfold. This article has focused on labour migration and its relationship to trade liberalisation. It has found that the traditional version of labour market segmentation theory, according to which there is a divide in the labour market between native and foreign labour, no longer holds true in the age of globalisation. Today, highly skilled migrants occupy the primary labour market alongside native workers, while lower-skilled migrant workers occupy the secondary labour market. This divide explains the new forms of regulation and discrimination.

The shift from restrictive to selective immigration law since the 1990s is also the result of changing labour market demands. This shift comes in response to stiffened global labour market demand for high-skill labour. The transformation of immigration law-making responds to declining productivity due to demographic challenges. The welfare state strengthens its ties to the market and thus relaxes restrictions on human mobility. At the same time, governments may be driven by a nation-state’s fears about globalisation’s challenges and effects on societal cohesion, to install restrictions on certain types of human mobility. The result is a divided, rather than a unified, landscape of migration rule-making.

The free trade template, which can be multilateral or regional, is used to ‘hunt globally for talent’; whereas, in principle, bilateral migration agreements target the lower-skilled service suppliers and workers in the fields of agriculture, manufacturing, and mining.[200]

As long as trade agreements remain inadequately equipped to regulate migratory risks, destination countries will continue to implement the duality between trade and non-trade agreements. The result is an emerging ‘agreement dualism’, with the horizontal or vertical layer offering the opportunity to apply one layer as a corrective to another.

Public goods theory provides an additional explanation for the skill divide. As has been shown, in neither the low- nor the high-skilled segment is labour migration considered a global public good. This result diminishes the prospects for a multilateral solution. For that reason too, it is but a logical consequence to see a divided normative response.

On the upside, it has been seen that the divide allows lawmakers to select the appropriate level (multilateral, regional, bilateral or unilateral) and venue of regulation (trade or non-trade agreement) for a given issue area of migration (labour recruitment, readmission, or skill testing). On the downside, ‘agreement dualism’ discriminates between skilled and less skilled migrant workers. It has implications for human rights and negative consequences for source country development. For example, access to a foreign labour market is usually highly regulated for lower-skilled workers, while it is facilitated for the highly skilled through spousal work permits, automatic visa renewals and extensions, free housing and other incentives.

Within agreement dualism, two areas of overlap exist which are defined as ‘agreement duplicities’. Both trade and non-trade agreements compete over a scarce resource, namely foreign graduate trainees, students, and young professionals. More worrisome is the negative conflict of competencies, which unifies trade and non-trade migration agreements over migrant workers’ rights protection. In most cases discussed, neither trade agreements nor bilateral migration agreement adequately protect migrant workers’ movement and post-admission rights.

For various reasons, we will not see a multilateral agreement on labour migration emerge in the near future. This is mostly because only the mobility of a limited segment of highly skilled migrants qualifies as a high-demand private good. The multilateral trading system of the WTO, and in particular Mode 4 of GATS, will be in even less of a position to take on a lead role in managing international labour migration. GATS’ jurisdictional and architectural constraints are too limiting for the industrialised countries among the WTO members to consider using this forum for liberalising the temporary mobility of lower-skilled service suppliers.

Instead, bilateralism in economic migration management is here to stay. There are manifold benefits of bilateralism, ranging from correcting the high-skill selectivity of the multilateral GATS Mode 4 and domestic law, to prospecting the grounds for future regional migration agreements, and to delegating certain responsibilities in migration to the private sector. Bilateralism does, however, have its downsides, including an asymmetric balance of interests tilted in favour of the destination country, a poor human rights record, and a potential incompatibility with the MFN clause of art II GATS. Future research will need to explore, in more detail, areas of overlap between bilateral migration agreements and trade agreements, as well as between different layers of migration law making, so as to identify potential areas of inefficiencies which come about with duplicated competencies. One such area can be defined in terms of negative overlaps. It relates to the absence of human rights protection in trade and non-trade migration agreements. At first glance, the trade and migration agreements seem worlds apart; yet viewed from a migrant worker’s perspective, the two regimes are less distinct, in particular when it comes to rights protection. In that sense the trade and migration regimes are closer to each other than either are to human rights.

[*] PhD (Bern); LLM (Duke); LicJur (Fribourg); Assistant Professor, World Trade Institute, University of Bern, Switzerland. I thank the two anonymous reviewers for their valuable comments and Susan Kaplan for her editorial assistance. All translations are by the author.

[1] Stephen Castles and Mark J Miller, The Age of Migration (Guilford Press, 3rd ed, 2003) 5.

[2] See Khalid Koser, ‘Introduction: International Migration and Global Governance’ (2010) 16 Global Governance 301, 304; see also Khalid Koser, International Migration: A Very Short Introduction (Oxford University Press, 2007) 1.

[3] See, eg, Symposium, ‘The Boundaries of the WTO’ (2002) 96 American Journal of International Law 1, 1–146, for a discussion of ‘trade and’ linkages; see also Andrew Lang, ‘Reflecting on Linkage: Cognitive and Institutional Change in the International Trading System’ (2007) 70 Modern Law Review 523, 523–34; see James Fawcett, ‘Networks, Linkages, and Migration Systems’ (1989) 23 International Migration Review 671, for an account of linkages in migration.

[4] See T Alexander Aleinikoff, ‘International Legal Norms on Migration: Substance without Architecture’ in Ryszard Cholewinski, Richard Perruchoud and Euan MacDonald (eds), International Migration Law: Developing Paradigms and Key Challenges (TMC Asser Press, 2007) 467, 468.

[5] James Hollifield, ‘Migration and the “New” International Order: The Missing Regime’ in Bimal Ghosh (ed), Managing Migration: Time for a New International Regime (Oxford University Press, revised ed, 2003) 75, 98–103. Starting with Hollifield’s observation of migration’s missing regime, various authors (notably in political science literature) have identified and criticised this institutional vacuum and have called for a comprehensive migration regime: see below nn 1316.

[6] Alexander Betts, ‘Introduction: Global Migration Governance’ in Alexander Betts (ed), Global Migration Governance (Oxford University Press, 2011) 1, 13.

[7] Khalid Koser, ‘Making the Case for the Global Governance of International Migration’ (2010) The Forum Discussing International Affairs and Economics 31, 32: ‘International migration has become an archetypal example of a global issue that is still largely governed at a national level’. See also Castles and Miller, above n 1, 7, who observe a trend towards the ‘globalisation’ of migration, in the sense that no country in the world today can escape the challenges of migration. A key contributing factor for the globalisation of migration is the blurring of the distinction between a receiving, sending or transit country, as borders become more porous as result of trade liberalisation. Moreover, the fact that migrant categories no longer remain clear-cut, but change over the course of the migratory cycle, thus enabling a refugee to become a migrant worker or vice-versa, testifies to the declining impact of the nation-state and in its wake, an increasing flexibility in the categorisations of states as ‘receiving’, ‘sending’ or ‘transit’ countries.

[8] Global Commission on International Migration, ‘Migration in an Interconnected World: New Directions for Action’ (Report, Global Commission on International Migration, 5 October 2005) 65 (‘GCIM Report’); see also Aleinikoff, above n 4, 467.

[9] Joel P Trachtman, ‘Coherence and the Regime Complex for International Economic Migration’ in Rahel Kunz, Sandra Lavenex and Marion Panizzon (eds), Multilayered Migration Governance: The Promise of Partnerships (Routledge, 2011) 62.

[10] Catherine Dauvergne, Making People Illegal: What Globalization Means for Migration and Law (Cambridge University Press, 2008) 62.

[11] See, eg, Anne-Grethe Nielsen, ‘Cooperation Mechanisms’ in Ryszard Cholewinski, Richard Perruchoud and Euan MacDonald (eds), International Migration Law: Developing Paradigms and Key Challenges (TMC Asser Press, 2007) 405, 405–21, for an exposition of international soft law on migration.

[12] See, eg, Joel P Trachtman, The International Law of Economic Migration: Toward the Fourth Freedom (Upjohn Institute Press, 2009) 333, 335; contra Hollifield, ‘The Missing Regime’, above n 5, 101.

[13] See Thomas Straubhaar, ‘Why Do We Need a General Agreement on Movements of People (GAMP)?’ in Bimal Ghosh (ed), Managing Migration: Time for a New International Regime (Oxford University Press, revised ed, 2003) 110, 134.

[14] See Bimal Ghosh, ‘Towards a New International Regime for Orderly Movements of People’ in Bimal Ghosh (ed), Managing Migration: Time for a New International Regime (Oxford University Press, revised ed, 2003) 6, 6.

[15] See Jagdish Bhagwati, ‘Borders beyond Control’ (2003) 82(1) Foreign Affairs 98, 104.

[16] Trachtman, Toward the Fourth Freedom, above n 12, 347.

[17] See Alexander Betts, ‘The Global Governance of Migration and the Role of Trans‑Regionalism’ in Rahel Kunz, Sandra Lavenex and Marion Panizzon (eds), Multilayered Migration Governance: The Promise of Partnerships (Routledge, 2011) 24, 31; Betts, ‘Introduction: Global Migration Governance’, above n 6, 8–9; cf James Hollifield, ‘Regulating Migration in an Era of Globalization: A Public Good Approach’ (Chapter presented at the Global Mobility Regimes Conference, Levin Institute, 27–28 April 2009).

[18] Various multilateral treaties for refugee protection and the prohibition of smuggling and trafficking attempt to overcome free-rider problems and collective action failures: see, eg, Convention Relating to the Status of Refugees, opened for signature 28 July 1951, 189 UNTS 137 (entered into force 22 April 1954); Protocol Relating to the Status of Refugees, opened for signature 31 January 1967, 606 UNTS 267 (entered into force 4 October 1967); United Nations Convention against Transnational Organized Crime, opened for signature 12 December 2000, 2225 UNTS 209 (entered into force 29 September 2003); Protocol to Prevent, Suppress and Punish Trafficking in Persons, Especially Women and Children, opened for signature 15 November 2000, 2237 UNTS 319 (entered into force 25 December 2003); Protocol against the Smuggling of Migrants by Land, Sea and Air, Supplementing the United Nations Convention against Transnational Organized Crime, opened for signature 12 December 2000, 2241 UNTS 507 (entered into force 28 January 2004); United Nations Convention against Transnational Organized Crime, opened for signature 15 November 2000, 2225 UNTS 209 (entered into force 29 September 2003).

[19] See Betts, ‘Introduction: Global Migration Governance’, above n 6, 8–9.

[20] See ibid 17; see also Rey Koslowski, ‘Global Mobility and the Quest for an International Migration Regime’ in International Organization for Migration, International Migration and Development: Continuing the Dialogue: Legal and Policy Perspectives (International Organization for Migration and Centre for Migration Studies, 1st ed, 2008) 103.

[21] See Alexander Betts and Lucie Cerna, ‘High-Skilled Labour Migration’ in Alexander Betts (ed), Global Migration Governance (Oxford University Press, 2011) 60, 61.

[22] See generally above n 18 and accompanying text.

[23] Betts, ‘The Global Governance of Migration and the Role of Trans-Regionalism’, above n 17, 29.

[24] See Marrakesh Agreement Establishing the World Trade Organization, opened for signature 15 April 1994, 1867 UNTS 3 (entered into force 1 January 1995) annex 1B (‘General Agreement on Trade in Services’).

[25] See generally Allan Findlay, ‘A Migration Channels Approach to the Study of High Level Manpower Movements: A Theoretical Perspective’ (1990) 28(1) International Migration 15; Khalid Koser and John Salt, ‘The Geography of Highly Skilled International Migration’ (1997) 3 International Journal of Population Geography 285, 289–90.

[26] See Ryszard Cholewinski, ‘Formulation and Management of Foreign Employment Policies in Countries of Destination’ in International Organization for Migration (ed), World Migration 2008: Managing Labour Mobility in the Evolving Global Economy (2008) 285, 296.

[27] Daniela Bobeva and Jean-Pierre Garson, ‘Overview of Bilateral Agreements and Other Forms of Labour Recruitment’ in Organisation for Economic Co-Operation and Development (ed), Migration for Employment: Bilateral Agreements at a Crossroads (2004) 11, 16. Bobeva and Garson coin the term ‘second-generation’ migration agreements to refer to those agreements that emerged post-9/11 in many European countries. These were a reaction to ‘first-generation’ readmission agreements which were rarely enforced. Other types of first-generation migration agreements include guest worker recruitment schemes, visa relaxation agreements for citizens of former colonies, and so forth.

[28] See Sandra Lavenex and Rachel Stucky, ‘“Partnering” for Migration in EU External Relations’ in Rahel Kuns, Sandra Lavenex and Marion Panizzon (eds), Multilayered Migration Governance: The Promise of Partnerships (Routledge, 2011) 116, for an exposition of EU mobility partnerships.

[29] See Mohammad Amin and Aaditya Mattoo, ‘Does Temporary Migration Have to Be Permanent?’ (Policy Research Working Paper Series No 3583, World Bank, 1 March 2005); see also Uri Friedman and David Zafar Ahmed, ‘Ensuring Temporariness: Mechanisms to Incentivise Return Migration in the Context of GATS Mode 4 and Least Developed Country Interests’ (Global Economic Issues Publication, Quaker United Nations Office, June 2008); Rupa Chanda, ‘Low-Skilled Workers and Bilateral, Regional, and Unilateral Initiatives: Lessons for the GATS Mode 4 Negotiations and Other Agreements’ (Report, United Nations Development Programme, April 2008).

[30] United Nations Development Programme, ‘Overcoming Barriers: Human Mobility and Development’ (Human Development Report, United Nations Development Programme, 2009) 21 <> . See International Organization for Migration (ed), World Migration Report 2010: The Future of Migration: Building Capacities for Change (2010).

[31] International Labour Organization, International Labour Migration: A Rights-Based Approach (International Labour Office, 2010) 2, 18.

[32] Ibid 2.

[33] See, eg, Barbara Schmitter-Heisler, ‘The Sociology of Immigration: From Assimilation to Segmented Integration, Assimilation, from the American Experience to the Global Arena’ in Caroline B Brettell and James F Hollifield (eds), Migration Theory: Talking across Disciplines (Routledge, 2nd ed, 2007) 83.

[34] See Directorate for Employment, Labour and Social Affairs, International Migration Outlook 2008 (Organisation for Economic Co-Operation and Development, September 2008). Foreign-born persons in 2006 accounted for about 12% of the total population in OECD countries, an increase of 18% compared to 2000.

[35] See, eg, Castles and Miller, above n 1, 7–8.

[36] Caroline B Brettell, ‘Theorising Migration in Anthropology: The Social Construction of Networks, Identities, Communities, and Globalscapes’ in Caroline B Brettell and James F Hollifield (eds), Migration Theory: Talking across Disciplines (Routledge, 2nd ed, 2007) 113, 120.

[37] According to the Stolper-Samuelson model, labour-intensive jobs are found in developing countries, while capital-intensive occupations move to industrialised countries. This unequal allocation of static production factors leads to migration.

[38] See Douglas Massey et al, Worlds in Motion: Understanding International Migration at the End of the Millennium (Oxford University Press, 1998) 28–30; see also Koser and Salt, above n 25, 289–90. Migration theory is split on the relative weight to be given to the push and pull factors of migration. While labour market segmentation theorists, such as Findlay, above n 25, or John Salt, ‘Migration Processes among the Highly Skilled in Europe’ (1992) 26 International Migration Review 484, consider the labour market divide as a key pull‑factor for migration, rational choice theorists emphasise the importance of push-factors, including an individual’s decision to migrate as a result of poverty.

[39] See Michael J Piore, Birds of Passage: Migrant Labour and Industrial Societies (Cambridge University Press, 1979) 187–205; see also Tobias Müller, ‘Migration Policy in a Small Open Economy with a Dual Labour Market’ (2003) 11 Review of International Economics 130, 138. Labour market segmentation theory is based on economic dualism, which describes how, in capital-intensive markets, migrants work on the ‘secondary labour market’ and natives on the ‘primary labour market’.

[40] See, eg, Massey et al, above n 38, 28–30.

[41] See, eg, Sophie Nonnenmacher, ‘International Labour Mobility in the Evolving Global Labour Market’ in International Organization for Migration (ed), World Migration 2008: Managing Labour Mobility in the Evolving Global Economy (2008) 23, 34–40, on the question of whether there is a global labour market.

[42] This is not to say that the traditional notion of labour market segmentation, which describes a situation in which native-born workers do not compete with the immigrant labour pool for jobs, is outdated. In fact, in certain destination countries, such as Saudi Arabia, the attractive jobs remain the reserved domain of natives, while the dangerous, dirty and difficult jobs are taken on by immigrant labour: see Salah Mahdi, ‘Labour Shortages, Migration, and Segmentation: The Case of Saudi Labour Market’ (Paper presented at the Workshop on the Interdependence of Migration Politics and Demography, Berlin, 15–20 February 2007).

[43] See Tomer Broude, ‘A Minimal Liberal Defense of (Some) Discrimination in Migration Regulation’ (Research Paper No 24-09, The International Law Forum at the Hebrew University of Jerusalem, 26 November 2009). Broude distinguishes morally unjustified discrimination among migrants, where human rights are violated, from justified forms of unequal treatment, for example, where inequality is based on quantitative restrictions or qualitative selection criteria (such as linguistic ability, education, professional experience, creative skills), the latter of which he bases on a communitarian, rather than egalitarian, model of justice.

[44] See Friedman and Ahmed, above n 29, for a discussion on dividing migration policies into ‘positive’ and ‘negative’ incentive schemes.

[45] See Ryo Kuboyama, ‘The Transformation from Restrictive to Selective Immigration Policy in Emerging National Competition State: Case of Japan in Asia-Pacific Region’ (Working Paper No 61, Centre on Migration, Citizenship and Development, 2004) 61. This shift away from ‘zero-immigration policy’ in the 1990s has been well-documented for European countries, in particular France, but has also taken place outside Europe in the Asia-Pacific region, particularly in Japan.

[46] Devesh Kapur and John McHale, Give Us Your Best and Brightest: The Global Hunt for Talent and Its Impact on the Developing World (Brookings University Press, 2005) 40.

[47] For other examples of recent policies attempting to ease entry for or recruit skilled immigrants, see Jeanne Batalova and B Lindsay Lowell, ‘The Best and the Brightest: Immigrant Professionals in the US’ in Michael Peter Smith and Adrian Favell (eds), The Human Face of Global Mobility: Internationally Highly-Skilled Migration in Europe, North America and the Asia-Pacific (Transaction Publishers, 2006) 81, 81–2.

[48] France’s new ‘skills and talents’ admission card was introduced in its immigration laws 2006-911 of 24 July 2006 and 2007-1631 of 20 November 2007: Loi no 2006-911 du 24 juillet 2006 relatif à l’immigration et l’intégration [Law No 2006-911 of 24 July 2006 regarding Immigration and Integration] (France) JO, 25 July 2006, ch V; Loi no 2007-1631 du 20 novembre 2007 relative à la maîtrise de l’immigration, à l’intégration et à l’asile [Law No 2006-1631 of 20 November 2007 regarding the control of immigration, integration and asylum] (France) JO, 21 November 2007, ch V. Like no other policy tool, the admissions card epitomises the paradigm shift of French immigration policies towards ‘selective migration’, as it empowers the French government to take part in the worldwide competition for the best and the brightest. The card is available to persons who contribute in a significant and durable manner to the economic development or the intellectual, scientific, cultural, sporting or humanitarian communities of both France and the country of their nationality. The holder is free to exercise the profession of his or her choice in France, with the exception of regulated professions. Consequently, no economic needs test is applied and no integration contract needs to be signed by its holder. The recipient of the skills and talents admission card may apply for a short-term (three month) stay or a longer-term renewable stay valid for three years. However, the applicant must present a project which he or she will pursue in France and defend that project in competition with competing applicants. Critics, including developing countries, have maintained that the skills and talents card exacerbates brain drain and thus acts as a catalyst of France’s high-skilled migration strategy. To reduce the risk of the skills and talents card leading to brain drain from developing countries, the project submitted by an applicant originating from such a country must plan a development cooperation, which benefits the cardholder’s country of origin. In addition, the validity of the card for citizens from those countries is limited to a six-year stay in principle, while even stricter conditions exist for citizens from Benin, Congo, Gabon, Mali and Senegal, for whom the skills and talents admission card is not renewable at all, so that there is no possibility of extension for citizens from those countries beyond the initial three-year stay.

[49] France foresees granting access to employment for foreign (non-EU) students, who have completed their studies with a diploma equivalent to a Master’s degree and have sufficient means to support themselves in France. To be eligible, the diploma must have been issued by an institution linked to the French education system. Students are granted a temporary residence permit for up to six months in order to ‘complete their formal education by a first professional experience in France in a perspective of returning to their country of origin’. No economic needs test is required, so the student has a free choice as to which remunerated activity he or she desires to pursue in France. However, the salary of the professional activity chosen must correspond to 1.5 times the French minimum wage. The latter criterion actually sensibly limits the scope of applicants for this admission card, because it targets medium- to high-skilled students: see Marion Panizzon, ‘Bilateral Labor Agreements and Trade in Services: Defining the Linkages, an Analysis of the Migration Pacts of France with Senegal, Tunisia, Congo, Gabon, Benin, Burkina Faso, Cape Verde, Cameroon and Mauritius’ (Commissioned Report, World Bank, forthcoming), which draws on Ministry of Immigration, Integration, National Identity and Solidarity Development, ‘Satisfying Labour Demand through Migration’ (France National Report, European Migration Network, 2010) 18 <;jsessionid=A75282FEDB3F716AFFDD76822234C306?fileID=1148> (‘EMN Report for France’).

[50] Kuboyama, above n 45, 4.

[51] Alexander Betts and Kalypso Nicolaidis, ‘The Trade-Migration Linkage: GATS Mode IV’ (Memorandum presented at the Global Trade Ethics Conference, Princeton University, 19 February 2009) 3.

[52] Kuboyama, above n 45, 4.

[53] Rainer Bauböck, ‘Towards a Political Theory of Migrant Transnationalism’ (2003) 37 International Migration Review 700, 707.

[54] Antoine Pécoud and Paul de Guchteneire, ‘Introduction: The Migration without Borders Scenario’ in Antoine Pécoud and Paul de Guchteneire (eds), Migration without Borders: Essays on the Free Movement of People (UNESCO and Berghahn Books, 2007) 9.

[55] Stéphan Vincent-Lancrin, ‘Student Mobility, Internationalization of Higher Education and Skilled Migration’ in International Organization for Migration (ed), World Migration 2008: Managing Labour Mobility in the Evolving Global Economy (2008) 105.

[56] Peter H Schuck, ‘Law and the Study of Migration’ in Caroline B Brettel and James F Hollifield (eds), Migration Theory: Talking across Disciplines (Routledge, 2nd ed, 2008) 242.

[57] Ibid.

[58] Ibid 243.

[59] Ibid 242.

[60] Kapur and McHale, above n 46, 9.

[61] See Marion Panizzon, ‘Franco-African Pacts on Migration: Bilateralism Revisited in Multilayered Migration Governance’ in Rahel Kunz, Sandra Lavenex and Marion Panizzon (eds), Multilayered Migration Governance: The Promise of Partnerships (Routledge, 2011) 216–17.

[62] See Jacob Funk Kirkegaard, The Accelerating Decline in America’s High-Skilled Workforce: Implications for Immigration Policy (Peterson Institute for International Economics, 2007) 19, citing Jeffrey Passell, ‘Unauthorized Migrants in the United States: Estimates, Methods, and Characteristics’ (OECD Social, Employment and Migration Working Paper No 57, Organisation for Economic Co-Operation and Development, 5 September 2007) 16 [33], describing the US immigrant workforce as ‘hourglass-shaped’, with many high-skilled and many low-skilled individuals. See also Saskia Sassen, Globalization and Its Discontents: Essays on the New Mobility of People and Money (The New Press, 1998) 47; David Autor and David Dorn, ‘The Growth of Low Skill Service Jobs and the Polarization of the US Labor Market’ (Working Paper, MIT, August 2010) 43.

[63] Autor and Dorn, above n 62 ,43.

[64] Ibid 43.

[65] See Sherry Stephenson and Gary Hufbauer, ‘Increasing Labour Mobility: Options for Developing Countries’ in Olivier Cattaneo, Michael Engman, Sebastián Sáez and Robert M Stern (eds), International Trade in Services: New Trends and Opportunities for Developing Countries (World Bank, 2010) 29, 57–61, noting that the bilateral labour agreement signed by Greece with Egypt covers only fishery workers, while the bilaterals signed by South Africa recruit farm and mining workers from Botswana, Lesotho, Malawi, Mozambique and Swaziland.

[66] See Pécoud and de Guchteneire, above n 54, 9.

[67] See John H Jackson, William J Davey and Alan O Sykes, Legal Problems of International Economic Relations (Thomson West, 5th ed, 2008) 218–19, describing how the US in 1945 issued a resolution on behalf of the UN Economic and Social Council (‘ECOSOC’) to convene a United Nations Conference on Trade and Employment which would be tasked to draft a charter for an international trade organisation and how three preparatory conferences were held for that purpose, in Lake Success, NY, in 1948; in Geneva in 1947; and in Havana in 1948. In US constitutional law, Congress must approve employment and labour issues, whereas trade issues, such as tariff negotiations remain the sole domain of executive (presidential) power. With Congress shifting to a ‘less liberal stance on trade matters and being less internationally oriented’ by 1950, the US Executive Branch decided not to submit the ITO Charter to Congress, thus rendering it ‘dead’. This also meant that labour and employment no longer figured on the negotiating agenda of the multilateral tariff negotiations, which were being simultaneously conducted under the framework of the GATT.

[68] Jimmie Reyna, ‘Services’ in Terence Stewart (ed), The GATT Uruguay Round: A Negotiating History (1986–1994) (Kluwer Law International, 1st ed, 1999) vol IV, 820, describing how the Dunkel Draft of 1992 ‘reflected primarily the concerns of developed countries’ and how ‘[developing countries’] concerns resulted in the adoption of a decision to continue negotiations in this area’.

[69] See generally Vincent Chetail, ‘Freedom of Movement and Transnational Migrations: A Human Rights Perspective’ in Alexander T Aleinikoff and Vincent Chetail (eds), Migration and International Legal Norms (TMC Asser Press, 2003) 48. Non-refoulement is a customary international law obligation, which describes the duty to admit those whose life and health are threatened in their home country. It has been codified in several instruments, including: the Convention Relating to the Status of Refugees, opened for signature 28 July 1951, 189 UNTS 137 (entered into force 22 April 1954) art 33(1); the Protocol Relating to the Status of Refugees, opened for signature 31 January 1967, 606 UNTS 267 (entered into force 4 October 1967); the Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, opened for signature 10 December 1984, 1465 UNTS 85 (entered into force 26 June 1987) art 3(1).

[70] GATS art I:2(d).

[71] See Martin Roy, ‘Endowments, Power, and Democracy: Political Economy of Multilateral Commitments on Trade in Services’ (Staff Working Paper No ERSD-2009-07, World Trade Organization, September 2009).

[72] See Trachtman, Toward the Fourth Freedom, above n 12, 244.

[73] See Brian Opeskin, ‘The Influence of International Law on the International Movement of Persons’ (Human Development Research Paper No 2009/18, United Nations Development Programme, April 2009) 20 [97], where the author finds, in this context, that ‘international trade law is less concerned with eliminating obstacles to the free movement of labor than ensuring that any such obstacles are reciprocal and non-discriminatory’.

[74] Jackson, Davey and Sykes, above n 67, 218.

[75] Marrakesh Agreement Establishing the World Trade Organization, opened for signature 15 April 1994, 1867 UNTS 3 (entered into force 1 January 1995) annex 1C (‘Agreement on Trade-Related Aspects of Intellectual Property Rights’) (‘TRIPS Agreement’). See, eg, Jürgen Bast, ‘Annex on Movement of Natural Persons Supplying Services’ in Rüdiger Wolfrum, Peter-Tobias Stoll and Clemens Feinäugle (eds), WTO — Trade in Services (Martinus Nijhoff, 2008) 575, 579–80.

[76] See Marrakesh Agreement Establishing the World Trade Organization, opened for signature 15 April 1994, 1867 UNTS 3 (entered into force 1 January 1995) annex 1B (‘General Agreement on Trade in Services’) annex 2 (‘Annex on Movement of Persons Supplying Services under the Agreement’) para 2 (‘Annex TMNP’).

[77] See Guidelines for the Scheduling of Specific Commitments under the General Agreement on Trade in Services, WTO Doc S/L/92 (28 March 2001) (Guidelines adopted by the Council for Trade in Services) [34]; Presence of Natural Persons (Mode 4), WTO Doc S/C/W/301 (15 September 2009) (Background Note by the Secretariat) 5, 23–4 (‘Secretariat Note on Mode 4’).

[78] Cf Martin Molinuevo, ‘Article XX GATS’ in Rüdiger Wolfrum, Peter-Tobias Stoll and Clemens Feinäugle (eds), WTO — Trade in Services (Martinus Nijhoff, 2008) 445, 454. If a WTO member decides to modify — that is, to re-introduce — a barrier to market access or to retract the commitment, it must compensate, on an MFN-basis, all other WTO members by offering to make a commitment in another sector or mode of supply GATS, unless the modification has a trade-neutral or more favourable effect on trade: GATS art XXI.

[79] International Labour Organization, ILO Declaration on Fundamental Principles and Rights at Work and Its Follow-Up (Declaration adopted by the International Labour Conference at its 86th session, Geneva, 18 June 1998).

[80] Annex TMNP para 3 (emphasis added).

[81] See Reyna, above n 68, 820, who explains that during the Uruguay Round negotiations, developing countries noted that most of the gains realised from commitments in the fourth mode of supply accrued to developed countries as they applied mostly to managers and executives and that despite the fact that

parties were encouraged to bilaterally negotiate commitments they wished to undertake with respect to the movement of labor, … concerns of the developing countries failed to generate improved offers in the area of the movement of personnel other than managers and executives.

[82] Hamid Mamdouh, ‘Mode 4 — Definition, Commitments, State of Play in the Negotiations’ (Presentation at the WTO Symposium on Mode 4 of the GATS: Taking Stock and Moving Forward, 22 September 2008) slide 17.

[83] Secretariat Note on Mode 4, WTO Doc S/C/W/301, 24.

[84] GATS art XIX:2; see also Special and Differential Treatment Provisions in WTO Agreements and Decisions, WTO Doc TN/CTD/W/33 (8 June 2010) (Note by the Secretariat) 62.

[85] Doha Work Programme, WTO Doc WT/L/579 (2 August 2004, adopted 1 August 2004) (Decision adopted by the General Council) para 1(e).

[86] Doha Work Programme, WTO Doc WT/MIN(05)/DEC (22 December 2005, adopted 18 December 2005) (Ministerial Declaration) para 47.

[87] See L Alan Winters, ‘Developing Country Proposals for the Liberalization of Movements of Natural Service Suppliers’ (Working Paper No T8, Development Research Centre on Migration, Globalisation and Poverty, January 2005).

[88] Proposed Liberalization of Mode 4 under GATS Negotiations, WTO Doc TN/S/W/14 (3 July 2003) (Communication from Argentina, Bolivia, Chile, The People’s Republic of China, Colombia, Dominican Republic, Egypt, Guatemala, India, Mexico, Pakistan, Peru, Philippines and Thailand) para 1(3).

[89] See Proposed Liberalization of Movement of Professionals under General Agreement on Trade in Services, WTO Doc S/CSS/W/12 (24 November 2000) (Communication from India); India’s proposal for a GATS visa was endorsed by other developing countries (see ibid), but was criticised by LDCs, who feared that the Indian proposal would target only medium-skilled service suppliers and leave out the lower-skilled services professions: Rupa Chanda, ‘Movement and Presence of Natural Persons and Developing Countries: Issues and Proposals for the GATS Negotiations’ (Trade-Related Agenda, Development and Equity Working Paper Series No 19, South Centre, May 2004) 24–30.

[90] Chanda, ‘Movement and Presence’, above n 89, 24–30.

[91] Secretariat Note on Mode 4, WTO Doc S/C/W/301, [4], [16]: As the WTO Secretariat maintains, it would be illogical if host country firms could bring a claim against their own government requiring GATS treatment for foreign nationals they desire to employ; see also Antonia Carzaniga, ‘A Warmer Welcome? Access for Natural Persons under PTAs’ in Juan A Marchetti and Martin Roy (eds), Opening Markets for Trade in Services: Countries and Sectors in Bilateral and WTO Negotiations (Cambridge University Press, 2008) 475, 476.

[92] See Chanda, ‘Movement and Presence’, above n 89, 26. Developing and least developed countries facing (unemployed) surplus labour contest the narrow view, while labour-receiving countries have an interest to keep the scope of GATS Mode 4 as limited as possible, so as to retain the widest possible policy space over labour migration. Only if there is a category of workers remaining outside the scope of GATS is it possible for labour-receiving countries to give preference in terms of market access quotas to those migrant-source countries willing to cooperate on border management, re-admissions and combating irregular migration. However, if a country has to generalise such an opening ‘immediately’ and ‘unconditionally’ under the most-favoured-nation (‘MFN’) obligation of art II:1 of GATS, it is no longer possible for it to conditionally link labour market openings to enlisting source-country cooperation via bilateral migration agreements.

[93] Cf Doha Work Programme, WTO Doc WT/MIN(05)/DEC (22 December 2005, adopted 18 December 2005) (Ministerial Declaration) para 47.

[94] Cf Marion Panizzon, ‘International Law of Economic Migration — A Ménage à Trois? GATS Mode 4, EPAs, and Bilateral Migration Agreements’ (2010) 44 Journal of World Trade 1207, 1217–20.

[95] See Pierre Sauvé, ‘Been There, Not Yet Done That: Lessons and Challenges in Services Trade’, in Marion Panizzon, Nicole Pohl and Pierre Sauvé (eds), GATS and the Regulation of International Trade in Services (Cambridge University Press, 2008) 599, 624, who finds that ‘seeking to pigeon-hole temporary access to labour markets in a multilateral trade policy setting, and to address those worker categories whose enhanced temporary mobility would make the greatest impact on poverty reduction, is thus far from easy’.

[96] See Steve Charnovitz, ‘Trade Law Norms on International Migration’ in Alexander T Aleinikoff and Vincent Chetail (eds), Migration and International Legal Norms (TMC Asser Press, 2003) 242, 252: ‘at present, the movement of natural persons is discussed in the WTO mainly as a services modality, rather than in the broader context of allowing workers to gain [new skills and career opportunities]’; see also Diana Zacharias, ‘Article I GATS’ in Rüdiger Wolfrum, Peter-Tobias Stoll and Clemens Feinäugle (eds), WTO — Trade in Services (Martinus Nijhoff, 2008) 52; Bast, above n 75, 580–2.

[97] Betts and Nicolaidis, above n 51, 2.

[98] See Sauvé, above n 95, 624.

[99] See Daniela Persin, ‘Free Movement of Labour: UK Responses to the Eastern Enlargement and GATS Mode 4’ (2008) 42 Journal of World Trade 837, 856–7.

[100] Carzaniga, above n 91, 484; see also Stephenson and Hufbauer, above n 65, 61, stating that there are ‘numerous advantages of BLAs or temporary worker programs, particularly for lower-skilled categories of workers’, whereby the ‘foremost advantage is that these agreements and programs can allow considerable flexibility in the management of the labour market by the countries involved’.

[101] See Sauvé, above n 95, 624.

[102] Persin, above n 99, 857.

[103] Annex TMNP para 4.

[104] Agreement between Japan and the Republic of the Philippines for an Economic Partnership, signed 9 September 2006 (entered into force 28 August 2008) < main.pdf> .

[105] North American Free Trade Agreement, opened for signature 17 December 1992, 32 ILM 289 (1993) (entered into force 1 January 1994).

[106] See Stephenson and Hufbauer, above n 65, 38.

[107] See Kotaro Matsuzawa, ‘Movement of Natural Person and the United States: Probability of the Mode 4 Offer Improvement by the United States’ (2008) 42 Journal of World Trade 653, 657–8.

[108] It was originally proposed that the GATS contain an obligation to fast-track entry for those service supplying persons falling under GATS Mode 4 as opposed to other types of labour migration: Communication from Argentina, Colombia, Cuba, Egypt, India, Mexico, Pakistan and Peru, WTO Doc MTN.GNS/W/106 (18 June 1990) annex (‘Temporary Movement of Services Personnel’) art 9, which stated that ‘the procedures for entry for temporary stay shall be accomplished expeditiously so as to avoid unduly impairing or delaying the conduct of trade in services’ and obliged the parties to ensure that ‘their embassies and immigration offices abroad, and immigration authorities at ports of entry are familiar with the visas issued pursuant to this Annex’.

[109] Tomer Broude, ‘The WTO/GATS Mode 4, International Labor Migration Regimes and Global Justice’ (Research Paper No 7-07, International Law Forum at The Hebrew University of Jerusalem, 18 May 2007) 30.

[110] See, eg, Philip Martin, ‘Migrants in the Global Labour Market’ (Working Paper, Policy Analysis and Research Programme of the Global Commission on International Migration, September 2005) 2.

[111] Maurice Schiff, ‘Migration, Trade and Investment: Complements or Substitutes?’ (Working Paper No 89, Centre for Economic and International Studies, 14 November 2006).

[112] See Schuck, above n 56, 242; see also Commission des affaires étrangères, l’Assemblée nationale [Foreign Affairs Committee, National Assembly] (France), Rapport no 1471 de M Michel Terrot [Report No 1471 by Michel Terrot] (2009) (‘Terrot 2009 Report’), concerning the adoption by the Senate of draft laws authorising the ratification of agreements on concerted migration management and co-development between France and the governments of Benin, Congo and Senegal.

[113] Administrative Procedures for Mode 4 Temporary Entry in Australia: The eVisa System WTO Doc TN/S/W/25 (26 November 2004) (Communication from Australia) [5].

[114] Ibid.

[115] See Trachtman, ‘Coherence and the Regime Complex for International Economic Migration’, above n 9, 62.

[116] See Betts, ‘Role of Trans-Regionalism’, above n 17, 25.

[117] Trachtman, Toward the Fourth Freedom, above n 12, 206.

[118] Singh Juss Satvinder, International Migration and Global Justice (Ashgate, 2007) xv.

[119] See Organization for Security and Co-Operation in Europe, International Organization for Migration and International Labour Organization, Handbook on Establishing Effective Labour Migration Policies in Countries of Origin and Destination (2006) 179, citing Daniela Bobeva and Jean-Pierre Garson, ‘Overview of Bilateral Agreements and Other Forms of Labour Recruitment’ in Organisation for Economic Co-Operation and Development (ed), Migration for Employment: Bilateral Agreements at a Crossroads (2004) 12.

[120] Ibid 182.

[121] See Kathleen Newland, ‘The Governance of International Migration: Mechanisms, Processes, and Institutions’ (2010) 16 Global Governance 331, 335.

[122] See International Organization for Migration, ‘Illustration of Multilateral, Regional and Bilateral Cooperative Arrangements in the Management of Migration’ in T Alexander Aleinikoff and Vincent Chetail (eds), Migration and International Legal Norms (Cambridge University Press, 2003) 305, 306.

[123] Philip Martin, Manolo Abella and Christiane Kuptsch (eds), Managing Labor Migration in the Twenty-First Century (Yale University Press, 2006) 94. The Bracero programme between Mexico and the US ran from 1942–64 and admitted 5 million farmworkers.

[124] See, eg, André Irving, ‘The Genesis and Persistence of the Commonwealth Caribbean Seasonal Agricultural Workers Program in Canada’ (1990) 28 Osgoode Hall Law Journal 243; see also Tanya Basok, ‘Migration of Mexican Seasonal Farm Workers to Canada and Development: Obstacles to Productive Investment’ (2000) 34 International Migration Review 79.

[125] See Panizzon, ‘International Law of Economic Migration’, above n 94, 1232–3.

[126] See Schiff, above n 111, discussing, from an institutional economics perspective, the complementarity and substitutability of regional bloc formations in trade, as opposed to in North–South and South–South migration.

[127] See Henri de Lary, ‘Bilateral Labour Agreements Concluded by France’ in Organisation for Economic Co-Operation and Development (ed), Migration for Employment: Bilateral Agreements at a Crossroads (2004) 43, 43–54 for an account of ‘first generation’ migration agreements concluded by France.

[128] See Lettre de mission de M Nicolas Sarkozy, Président de la République, addressée à M Brice Hortefeux, ministre de l’immigration, de l’integration, de l’identité nationale et du codéveloppement, sur les priorités en matière de politique de l’immigration, le 9 juillet 2007 [Letter of Engagement from Mr Nicolas Sarkozy, President of the Republic, Addressed to Mr Brice Hortefeux, Minister for Immigration, Integration, National Identity and Co-Development, on the Priorities in the Matter of Immigration Politics, on 9 July 2007], Collection des discours publics [Collection of Public Discourses] <> .

[129] See EMN Report for France, above n 49, 3.

[130] See, eg, Meng-Hsuan Chou and Nicolas Baygert, ‘The 2006 French Immigration and Integration Law: Europeanisation of Nicolas Sarkozy’s Presidential Keystone?’ (Working Paper No 45, Center on Migration Policy and Society, 2007).

[131] Ministère de l’immigration, de l’intégration, de l’identité nationale et du développement solidaire [The Ministry for Immigration, Integration, National Identity and Consolidated Development], the French ministry in charge of French migration policy, was dissolved during the Sarkozy government’s ministerial re-allocations of 14 November 2010. However, its mandate continues to exist almost unaltered and was put under the auspices of the Ministry of the Interior, headed by Brice Hortefeux: Brice Hortefeux et le Tournant Sécuritaire (15 November 2010) France Info <> .

[132] See Commission des affaires étrangères, l’Assemblée nationale [Foreign Affairs Committee, National Assembly] (France), Rapport no 2343 de M Michel Terrot [Report No 2343 by M Michel Terrot] (2010) 5 (‘Terrot 2010 Report’), concerning the adoption by the Senate of draft laws authorising the ratification of the agreement between the French Government and the Republic of Mauritius relating to the stay and circular migration of professionals.

[133] Donata Gnisci, West African Mobility and Migration Policies of OECD Countries (Organisation for Economic Co-Operation and Development, 2008) 31.

[134] See EMN Report for France, above n 49, 33.

[135] See GCIM Report, above n 8; see also International Migration and Development, GA Res 61/208, UN GAOR, 2nd Comm, 61st sess, 69th plen mtg, UN Doc A/RES/61/208 (6 March 2007).

[136] See Ministère de l’immigration, de l’intégration, de l’identité nationale et du développement solidaire [The Ministry for Immigration, Integration, National Identity and Consolidated Development], ‘L’essentiel sur l’immigration professionelle’ [Essential Information on Professional Immigration] (Explanatory Booklet, 2009) 9 <> . To allay concerns about brain drain associated with a young professional’s admission, the visa is conditional upon the professional returning to their country of origin after his or her professional experience in France: at 4. This strict return policy targets young professionals (but also students) more stringently than the other categories of workers and stands somewhat in contradiction to a global trend in migration policy, which is to draw upon young professionals as potential resource for high skills: at 7.

[137] Arrêté du 18 janvier 2008 [Decree of 18 January 2008] (France) JO, 21 January 2008, 8. France operates two distinct occupational shortage lists, one with 30 and one with 150 occupations, which both have the effect of facilitating admission for non-French nationals onto the French labour market, by either relaxing or eliminating the economic needs test (‘ENT’). The list of 150 shortage occupations eliminates the ENT altogether and on a nation-wide basis, that is, for metropolitan France, excluding overseas territories. It applies to workers, who are either nationals or residents of Newly Acceding EU Member States (‘NMS’) — defined in the decree as Estonia, Lithuania, Latvia, Czech Republic, Hungary, Poland, Slovakia, Slovenia, Bulgaria and Romania — as long as the EU free movement regime does not apply fully to these. In practical terms this means that during the transitory period, France will not prioritise a French or EU citizen/resident over one of a NMS for a job, should that job be listed as one of the 150 occupations in shortage in France. The other list comprises 30 occupations and applies to any third country national. It differs from the list of 150 occupations in one aspect: the labour market shortage is determined regionally by department (thus, the jobs on the list for Burgundy may be different for Île-de-France or Languedoc-Roussillon), rather than applicable across-the-board for the entire reach of mainland France. The regional limitation operates as an important market access barrier to the entry of a third country national, since he or she must be willing to work in a particular French department; see also Marion Panizzon, ‘Franco-African Pacts on Migration: Bilateralism Revisited in Multilayered Migration Governance’ in Rahel Kunz, Sandra Lavenex and Marion Panizzon (eds), Multilayered Migration Governance: The Promise of Partnerships (Routledge, 2011) 215; Ibrahim Awad, ‘Dimensions of Highly-Skilled Labour Migration’ in Dossier Mobility & Inclusion, Managing Labour Migration in Europe (Heinrich Böll Stiftung, April 2010) 11.

[138] Comité Inter-Mouvements Auprès des Evacués, ‘French Agreements concerning the Concerted Management of Migration Flows and Co-Development’ (Briefing Paper, Comité Inter-Mouvements Auprès des Evacués, 10 May 2009) (‘CIMADE Briefing Paper’) 4.

[139] Proposed Liberalisation of Movement of Professionals under General Agreement on Trade in Services (GATS), WTO Doc S/CSS/W/12 (24 November 2000) (Communication from India), which proposes a ‘Reference Paper on Use of ENT’; see also Doha Work Programme, WTO Doc WT/MIN(05)/DEC (13–18 December 2005, adopted 18 December 2005) (Ministerial Declaration) Annex C Services, F-1, where WTO membership calls on the ‘removal or substantial reduction of economic needs tests’ in order to ‘achieve a progressively higher level of liberalization of trade in services, with appropriate flexibility for individual developing country Members’.

[140] See, eg, CIMADE Briefing Paper, above n 138.

[141] See Reyna, above n 68, 804–5, describing how the GATT Secretariat noted that measures under certain bilateral agreements affected trade in services and were discriminatory, that they were covered under the GATS, and that those parties desiring to maintain such measures were called upon to note an MFN exemption on the list of their MFN exemptions.

[142] See European Communities and Their Member States — Final List of Article II GATS (MFN) Exemptions, WTO Doc GATS/EL/31 (15 April 1994) annex 2. France has entered such an MFN exemption towards francophone Africa, so that the additional professions listed as shortage occupations in France’s new pacts with francophone African countries, and for which no individual economic necessity tests are required, will be consistent in terms of WTO law.

[143] See Switzerland — List of Article II GATS (MFN) Exemptions, WTO Doc GATS/EL/83 (15 April 1994) annex 2: when ‘granting permits for entry, stay and work to natural persons’ towards the EU/European Free Trade Association countries and ‘traditional recruiting areas’.

[144] See The United States of America — List of Article II (MFN) Exemptions, WTO Doc GATS/EL/90 (15 April 1994): The US automatically issues a ‘treaty trader’ or ‘treaty investor immigrant’ visa for all countries with which it has a treaty of friendship, commerce and navigation, or a bilateral investment agreement. For other examples of art II MFN exemptions facilitating the preferential admission of service supplying persons, see Secretariat Note on Mode 4, WTO Doc S/C/W/301, Annex. New Zealand vis-à-vis Kiribati (capped at 20 nationals annually) and Tuvalu (capped at 80 nationals annually); Jordan, waiving annual work permit fees towards nationals of Arab countries, Jamaica towards Caribbean Community Secretariat members, waiving work permits, Italy, guaranteeing work permits to countries of Central and Southern Eastern Europe and of the Mediterranean basin; the UK waiving work permit requirements for Commonwealth member countries, if these persons had a grandparent born in the UK.

[145] See CIMADE Briefing Paper, above n 138, 4.

[146] See Common Consular Instructions (CCI) on Visas for the Diplomatic Missions and Consular Posts [2005] OJ C 326/1, which determine for all EU member states (excluding the UK and Ireland) and Schengen-associated countries (Iceland, Norway, Liechtenstein and Switzerland), towards which third countries a Schengen visa is required.

[147] Common Consular Instructions on Visas for the Diplomatic Missions and Consular Posts [2002] OJ C 313/01.

[148] Accord entre le Gouvernement de la République française et le Gouvernement de la République du Sénégal relatif à la gestion concertée des flux migratoires [Agreement between the Government of the French Republic and the Government of the Republic of Senegal with regard to the Concerted Management of Migration Flows], signed 23 September 2006, [2009] JO 8707 (entered into force 1 August 2009). France targets countries that have neither a readmission agreement with France nor an EU-wide readmission agreement in place, when selecting candidates for its bilateral migration agreements. Most sub-Saharan African countries qualify (including Benin, Burkina Faso, Cameroon, the Republic of Cape Verde, Mali and Nigeria). Congo (1993), Gabon (2002), and Senegal (2000) have readmission agreements with France, but these only create a readmission obligation for their own nationals. In light of harmonised EU migration policy, and in particular to prepare the grounds for an EU-wide readmission agreement with these three countries, France replaced its ‘old’ readmission agreements with new bilateral migration agreements that stipulate a more fully-fledged readmission obligation. The Republic of Cape Verde in 2008 signed a fully-fledged, bilateral migration agreement with France. In 2009, the Republic of Cape Verde concluded a readmission agreement with the EU and in 2010 its agreement with France entered into force. The opposite occurred with the countries from the West Balkans: Serbia, Montenegro, and Macedonia signed on to EU readmission agreements in 2006 only to sign on to a ‘light’ version of a French migration pact in 2009, which offers labour market access for young professionals, something which the EU cannot grant on an EU-wide basis. Egypt, Mauritania, Vietnam and the Philippines are countries without a readmission agreement with either France or the EU, but which are listed by the French government as potential candidates for a bilateral migration agreement (in preparation of an EU-wide readmission agreement). Georgia, Moldova and Pakistan have a readmission agreement with the EU so there is a high likelihood that if these countries sign on to a migration agreement with France, it will be one of the ‘light’ versions not requiring a readmission obligation — similar to those France signed with Mauritius, Serbia, Montenegro, and Macedonia.

[149] Accord entre le Gouvernement de la République française et le Gouvernement de la République du Cameroun relatif à la gestion concertée des flux migratoires et au développement solidaire [Agreement between the Government of the French Republic and the Government of the Republic of Cameroon with regard to the Concerted Management of Migration Flows], signed 31 May 2009.

[150] Accord entre le Gouvernement de la République française et le Gouvernement du Burkina Faso relatif à la gestion concertée des flux migratoires et au développement solidaire [Agreement between the Government of the French Republic and the Government of Burkina Faso with regard to the Concerted Management of Migration Flows and Inclusive Development], signed 10 January 2009 (entered into force 3 January 2011) arts 10 (readmission of nationals), 11 (readmission of third country citizens).

[151] Accord entre le Gouvernement de la République française et le Gouvernement de la République du Bénin relatif à la gestion concerté des flux migratoires et au codéveloppement [Agreement between the Government of the French Republic and the Government of the Republic of Benin on the Cooperative Management of Immigration Flows and Development Cooperation], signed 28 November 2007 [2009] OJ 8705 (entered into force 1 March 2010) arts 16 (concerning own nationals), 17 (relating to third county nationals).

[152] Accord entre le Gouvernement de la République française et le Gouvernement de la République du Congo relatif à la gestion concerté des flux migratoires et au codéveloppement [Agreement between the Government of the Republic of France and the Government of the Republic of Congo on the Cooperative Management of Immigration Flows and Development Cooperation], signed 25 October 2007 [2009] OJ 8705 (entered into force 1 August 2009) arts 3.1 (own nationals), 3.2 (third country nationals).

[153] Cf Gregor Noll, ‘Return of Persons to States of Origin and Third States’ in T Alexander Aleinikoff and Vincent Chetail (eds), Migration and International Legal Norms (TMC Asser Press, 2003) 67.

[154] There are eight EU-wide readmission agreements. The first one, with Albania, entered into force on 1 May 2006 (Agreement between the European Community and the Republic of Albania on the Readmission of Persons Residing without Authorisation [2005] OJ L 124/22), it was followed by the one with Ukraine, signed on 18 June 2007, and which entered into force 1 January 2008 (Agreement between the European Community and Ukraine on the Readmission of Persons [2007] OJ L 332/48). The agreement with Russia entered into force on 1 June 2007 (Agreement between the European Community and the Russian Federation on the Readmission of Persons Residing without Authorisation [2007] OJ L 129/40), and those with Macedonia (Agreement between the European Community and the Former Yugoslav Republic of Macedonia on the Readmission of Persons Residing without Authorisation [2007] OJ L 334/7), Serbia (Agreement between the European Community and the Republic of Serbia on the Readmission of Persons Residing without Authorisation [2007] OJ L 334/46), Montenegro (Agreement between the European Community and the Republic of Montenegro on the Readmission of Persons Residing without Authorisation [2007] OJ L 334/26), Bosnia and Herzegovina (Agreement between the European Community and Bosnia and Herzegovina on the Readmission of Persons Residing without Authorisation [2007] OJ L 334/66), and Moldova (Agreement between the European Community and the Republic of Moldova on the Readmission of Persons Residing without Authorisation [2007] OJ L 334/149), all entered into force on 1 January 2008.

[155] Treaty of Lisbon Amending the Treaty on European Union and the Treaty Establishing the European Community, signed 13 December 2007, [2007] OJ C 306/01 (entered into force 1 December 2009) art 79.

[156] See Organization for Security and Co-Operation in Europe, International Organization for Migration and International Labour Organization, Handbook on Establishing Effective Labor Migration Policies in Countries of Origin and Destination (2006) 183.

[157] See, eg, Jean-Pierre Cassarino, ‘Informalizing Readmission Agreements in the EU Neighborhood’ (2007) 42 The International Spectator 179; see also Aderanti Adepoju, Femke van Noorloos and Annelies Zoomers, ‘Europe’s Migration Agreements with Migrant-Sending Countries in the Global South: A Critical Review’ (2010) 48(3) International Migration 42.

[158] The diaspora model refers to a policy measure that aims to integrate ‘present and past citizens into a web of rights and obligations in the extended community defined with the home country as the center’: Bhagwati, above n 15, 101.

[159] Hein de Haas, ‘Engaging Diasporas: How Governments and Development Agencies Can Support Diaspora Involvement in the Development of Origin Countries’ (Study for Oxfam Novib, International Migration Institute, June 2006) 68. Co-development financing was granted to diaspora organisations in France who agreed to cooperate in identifying their fellow citizens on unlawful stays in France, and thus became a sort of ‘compensation’ offered by France, partly in lieu of their source country governments.

[160] Trachtman, Toward the Fourth Freedom, above n 12, 207.

[161] Betts, ‘Role of Trans-Regionalism’, above n 17, 27–9 and Betts, ‘Introduction: Global Migration Governance’, above n 6, 14–19.

[162] See Adepoju, van Noorloos and Zoomers, above n 157, 46.

[163] See Migreurop, ‘European Borders: Controls, Detention, Deportations’ (2009–10 Report, 12 November 2010) 123.

[164] See Council of the European Union, Brussels Europrean Council: 15/16 December 2005 — Presidency Conclusions (17 December 2005) Doc No 15914/05. The ‘Global Approach to Migration’ (‘GAM’) is a strategy adopted by the European Council in 2005, which was first applied as ‘priority actions focusing on Africa and the Mediterranean’, strengthened in 2006 and expanded in 2007 to Eastern and South-Eastern regions neighbouring the European Union, and to a lesser extent to the Middle East and Asia and further strengthened in 2008. The EU GAM works by encouraging EU member states to seek a more balanced, but comprehensive sharing of responsibilities with migrant source countries relating to the three dimensions of migration: legal (labour) migration, combating irregular migration, and strengthening the ties between migration and development by engaging source countries in all-inclusive and fair migration partnerships.

[165] Sub-Committee F of the European Union Committee, House of Lords, ‘FRONTEX: the EU External Borders Agency — Report with Evidence’ (9th Report of Session 2007–08, European Union Committee, 5 March 2008) 34–5.

[166] See, eg, GCIM Report, above n 8, 68, advocating the ‘whole-of-government’ approach as a means to coordinate policy making.

[167] See Communication from the Commission to the Council and the European Parliament, ‘The Global Approach to Migration One Year On: Towards a Comprehensive European Migration Policy’ [2007] OJ C 181/17.

[168] See EMN Report for France, above n 49, 31.

[169] CIMADE Briefing Paper, above n 138, 17–18.

[170] Ibid 5. To prevent brain drain, France’s pacts with Benin, Congo and Mauritius cap the offer of the skills and talents card at maximum quotas of 150 citizens each annually, while the pact with Tunisia is capped at 1500.

[171] Schuck, above n 56, 242.

[172] See Betts, ‘Introduction: Global Migration Governance’, above n 6, 13.

[173] Meng-Hsuan Chou and Marie Gibert, ‘From Cotonou to Circular Migration: the EU, Senegal and the “Agreement Duplicity”’ (Paper presented at Migration: A World in Motion, Maastricht, 18–20 February 2010) 11–13.

[174] The notion of multilayered governance was originally developed for understanding the interaction between law-making through directives by the EC Commission and the implementation of the directives by EU member states’ legislators: see Gary Marks et al (eds), Governance in the European Union (Sage Publications, 1996); Ian Bache and Matthew Flinders (eds), Multi-Level Governance (Oxford University Press, 2004). Applied to the EU context, multilayered governance illustrates the dichotomy between the centralisation of agenda-setting and rule-making, which is unified within the EC Commission, and the fragmented and pluralistic implementation process of EU policies by the EU member states: see Guy Peters and John Pierre, ‘Governance Approaches’ in Antje Wieners and Thomas Diez (eds), European Integration Theory (Oxford University Press, 2009) 91, 95–6.

[175] Panizzon, ‘Bilateral Labor Agreements and Trade in Services’, above n 49; see also EMN Report for France, above n 49.

[176] See European Migration Network, EMN Report for France, above n 49, 17–18.

[177] The French ‘intra-corporate transferees’ admission card is valid for three years and renewable. It enables a foreign employee who has been working for more than three months for a firm incorporated abroad to be dispatched to the subsidiary or headquarters of that firm in France. As noted in EMN Report for France, above n 49, however, the employee must possess qualifications and technical skills justifying a monthly gross salary equivalent to 1.5 times more than the French minimum wage. The new intra-corporate admission card is thus reserved to the rather highly skilled and the labour market test will not be applied to graduate trainees: see EMN Report for France, above n 49, 13–15.

[178] Accord entre le Gouvernement de la République française et le Gouvernement de la République de Maurice relatif au sejour et à la migration circulaire de professionels [Agreement between the Government of the Republic of Mauritius on Residence and the Circular Migration of Professionals], signed 23 September 2008 (entered into force 16 April 2010).

[179] See French Society for International Law, Bulletin Hebdomadaire Sentinelle n°214 [Sentinel Weekly Bulletin No 214], 21 February 2010 <> .

[180] See the French migration agreement with Mauritius: Décret no 2010-1114 du 22 septembre 2010 [Decree No 2010-1114 of 22 September 2010] (France) JO, 24 September 2010, 14, art 2(1)(3).

[181] See Julia Nielson and Olivier Cattaneo, ‘Current Regimes for the Temporary Movement of Service Providers: Case Studies of Australia and the United States’ in Aaditya Mattoo and Antonia Carzaniga (eds), Moving People to Deliver Services (World Bank, 2003) 113, 139–40.

[182] Communication from the European Communities and Its Member States — Draft Consolidated GATS Schedule, GATT Doc S/C/W/273 (9 October 2006) (‘Communication from the European Communities and Its Member States’).

[183] See Economic Partnership Agreement between the CARIFORUM States of the One Part and the European Community and Its Member States of the Other Part, opened for signature 15 October 2008 (provisionally entered into force 29 December 2008) art 81 < & redirect=true & treatyId=7407> .

[184] See Pierre Sauvé and Natasha Ward, ‘The EC–CARIFORUM Economic Partnership Agreement: Assessing the Outcome on Services and Investment’ (Working Paper, European Centre for Political Economy, 2009) 17–19.

[185] See, eg, Ryszard Cholewinski, ‘The Human and Labor Rights of Migrants: Visions of Equality’ (2008) 22 Georgetown Immigration Law Journal 177; see also Susan Martin and Rola Abimourched, ‘Migrant Rights: International Law and National Action’ (2009) 47(5) International Migration 115.

[186] See Chahal v United Kingdom (European Court of Human Rights, Grand Chamber, Application No 70/1995/576/662, 11 November 1996); Saadi v Italy (European Court of Human Rights, Grand Chamber, Application No 37201/06, 28 February 2008); see Beate Rudolf, ‘Chahal v United Kingdom No 70/1995/576/662’ (1998) 92 American Journal of International Law 70, 70–4 for a discussion of the case.

[187] See Nazaré Albuquerque Abell, ‘The Compatibility of Readmission Agreements with the 1951 Convention Relating to the Status of Refugees(1999) 11 International Journal of Refugee Law 60.

[188] Convention for the Protection of Human Rights and Fundamental Freedoms, opened for signature 4 November 1950, 213 UNTS 222 (entered into force 3 September 1953) as amended by Protocol No 11 to the Convention for the Protection of Human Rights and Fundamental Freedoms, opened for signature 11 May 1994, ETS No 155 (entered into force 1 November 1998).

[189] African Charter on Human and Peoples’ Rights, opened for signature 27 June 1981, 1520 UNTS 217 (entered into force 31 October 1986).

[190] Universal Declaration of Human Rights, GA Res 217A (III), UN GAOR, 3rd sess, 183rd plen mtg, UN Doc A/810 (10 December 1948).

[191] Partnership Agreement between the Member of the African, Caribbean and Pacific Group of States of the One Part, and the European Community and Its Member States, of the Other Part [2000] OJ L 317/03.

[192] Declaration on Fundamental Principles and Rights at Work, adopted on 18 June 1998, 37 ILM 1233.

[193] International Labour Organization, The Multilateral Framework on Labour Migration: Non-Binding Principles and Guidelines for a Rights-Based Approach to Labour Migration (International Labour Office, 2006).

[194] Agreement Establishing an Association between the European Community and Its Member States, of the One Part, and the Republic of Chile, of the Other Part [2002] OJ L 352/3, art 44.

[195] Agreement Establishing an Association between the European Community and Its Member States of the One Part, and the United Mexican States, of the Other Part [2000] OJ L 276/45.

[196] Agreement on Trade, Development and Cooperation between the European Community and Its States, of the One Part, and the Republic of South Africa, of the Other Part [1999] OJ L 311/3.

[197] See Roman Grynberg and Veniana Qalo, ‘Labor Standards in US and EU Preferential Trading Agreements’ (2006) 40 Journal of World Trade 619.

[198] Implementing Agreement between the Government of the Republic of the Philippines and the Government of Japan Pursuant to Article 12 of the Agreement between the Republic of the Philippines and Japan for an Economic Partnership, signed 9 September 2006 (entered into force 28 August 2008) art 11 <> .

[199] Aleinikoff, above n 4, 2.

[200] See, eg, Stephenson and Hufbauer, above n 65, 60, noting that the bilateral labour agreement signed by Greece and Egypt covers only fishery workers, while the bilaterals signed by South Africa recruit farm and mining workers from Botswana, Lesotho, Malawi, Mozambique and Swaziland.

AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback