The Elusive Triple Win: Can temporary labour migration dilemmas be settled through fair representation?

Temporary Labour Migration Programmes (TLMPs) are controversial because they are caught in a dilemma between global and domestic justice. From a global justice perspective, TLMPs expand opportunities for workers in poor countries to access labour markets of rich countries and they improve the situation of origin countries through remittances. From a domestic justice perspective, TLMPs violate principles of domestic equality because they always afford migrant workers more restricted rights than those enjoyed by citizens and long-term residents of the host country. Although this dilemma cannot be fully resolved, we argue that TLMPs can be morally justified and recommended if they are characterized by ‘fair representation’ in policy design and implementation and also meet certain democratic legitimacy conditions. TLMPs can be justified if they provide triple benefits for destination and origin countries as well as for migrants themselves, yet the relevant benefits can only be achieved cooperatively through transnational governance in which each of the three groups of actors is fairly represented. This conclusion is supported by our interpretation of the democratic principle of including all affected interests. Under such conditions, some rights of temporary migrant workers in host countries can be regarded as a legitimate outcome of negotiations, while others need to be fixed in advance under a democratic principle of equal protection of all subjected to the laws. Democratic legitimacy also requires that migrants enjoy protection and participation rights as citizens of their countries of origin as well as local citizenship in their countries of residence.


Introduction *
Europe and the United States are, once again, debating the expansion of temporary labour migration programmes (henceforth TLMPs), especially for admitting lower-skilled migrant workers (see .e.g.Bier 2021 1; European Commission 2020 2; The Economist 2020 3 ).We define TLMPs as policies that grant migrant workers temporary residence and work permits upon arrival, employ migrants under a restricted set of rights, and do not guarantee an 'upgrade' to permanent residence after some time.This paper makes a novel contribution to long-standing debates about the ethical desirability of TLMPs.We propose a new normative framework and ask whether 'fair representation' in the design and implementation of TLMPs can address the underlying moral dilemma of how to balance global and domestic justice concerns, and thus also justify variations in the specific content of such policies.
TLMPs have a long and global history (e.g.Hahamovitch 2003) and they are today the dominant form of labour immigration policy-making in high-income countries (e.g.Ruhs 2013).Research and policy debates about TLMPs have evolved over time, as has the terminology used to describe and discuss them.Most research on 'guest worker' policies in the United States (most notably the Bracero programme in 1942-64) and Europe (the Gastarbeiter programmes from the 1950s up to the mid-1970s) concluded that these programmes largely failed because they did not achieve their stated policy aims and instead generated a range of adverse impacts, including the emergence of labour market distortions and segmentations (e.g.Piore 1979) along with the growth of a structural dependence on the continued employment of migrant workers in the host country, and permanent settlement of many supposedly temporary migrant workers.The slogan "there is nothing more permanent than temporary foreign workers" was a popular summary statement of the perceived failure of these policies in democratic high-income countries the late 20 th century (see, e.g.Castles 1986;Martin and Teitelbaum 2001).At the same time, the large-scale guest worker policies of the Gulf States, which were more successful in ensuring the temporariness and return of migrant workers, were widely considered as unacceptable in democratic countries.
The early 2000s saw the emergence of new research and policy debates that moved beyond "what went wrong" and the "inevitable failure" of TLMPs (e.g.Castles 2004) to questions about whether and how new TLMPs can be designed to avoid the adverse consequences of past guest worker policies (e.g.Martin 2003;Ruhs 2006).Much of this debate was driven by international organisations and initiatives concerned with liberalizing labour migration around the world, especially for lower-skilled workers.For example, the Global Commission on International Migration, a body encouraged by then UN Secretary General Kofi Annan in 2003, concluded that "states and the private sector should consider the option of introducing carefully designed temporary migration programmes as a means of addressing the economic needs of both countries of origin and destination" (GCIM 2005, p. 16).Similarly, the World Bank and United Nations Development Programme (UNDP) called for the expansion of TLMPs, especially for lower skilled workers, because of the large benefits of migration for poverty reduction and human development (World Bank 2006;UNDP 2009).
In more recent years, research and policy debates about TLMPs have been expanded to include 'circular migration programmes' which facilitate temporary but repeated stays of migrant workers (e.g.European Migration Network 2011) and international 'skills partnerships' that combine temporary labour migration with skills training and development in origin countries (e.g.Clemens 2015).The recent Global Compact on Safe, Orderly and Regular Migration, adopted by the UN General Assembly in December in 2018, calls on governments to implement labour mobility schemes for "temporary, seasonal, circular, and fast-track programmes in areas of labour shortages" (GCM 2018).The need for cooperation between countries of migrants' destination and origin in the design and implementation of TLMPs has become an increasingly important new feature in policy debates.For example, the EU's "New Pact on Migration and Asylum" calls for expanded cooperation between the EU and non-EU countries on labour migration (European Commission 20204 ).
This shift in research and policy debates over the past 20 years towards greater engagement with the question of how policy design might improve the outcomes of TLMPs, and towards a greater emphasis on the need for cooperation between destination and origin countries, has led to a growing body of normative analyses of such programmes (e.g.Chang 2002;Mayer 2005;Bell 2006;Carens 2008 and2013;Lenard and Straehle 2010;Lenard 2012;Ruhs 2013;Miller 2008 and2016;etc.).This paper contributes to this growing literature.Our starting point is that, from a normative perspective, TLMPs raise a fundamental dilemma between global and domestic justice.
From a global justice perspective, TLMPs expand opportunities for workers in poor countries to access labour markets of rich countries, thus improving their and their families' household incomes and human development.At the same time, from a domestic justice perspective, TLMPs violate principles of domestic equality because, regardless of their design, they always afford migrant workers more restricted rights than those enjoyed by citizens and long-term residents of the host country.As we discuss in section 2, most TLMPs, especially those for admitting lower-skilled migrant workers, restrict migrants' right to free choice of employment, access to the welfare state, opportunities for family reunification, andmost obviouslysecurity of residence in the host country.
The guest worker programmes of the 20 th century were designed to maximize the benefits for host states and paid less regard to the interests of migrants and source countries even if these were claimed as beneficial side effects.New advocacy of TLMPs in the 21 st century invokes instead the 'triple win' for all three categories of actors involved.However, host country interests have remained dominant in the creation and operation of TLMPs.If the justification of TLMPs requires benefits for migrants and source countries, it is not likely that these can be secured in the prevailing pattern of bilateral treaties.
Most existing normative analyses of TLMPs are focused on issues related to policy design, especially on the question which rights of temporary migrant workers can be justifiably restricted and for how long.For example, Carens (2008Carens ( , 2013) ) suggests that it is justifiable to restrict some specific rights of temporary migrant workers including their access to noncontributory welfare benefits (such as social housing) for some time.Martin Ruhs (2013) and Gillian Brock (2020) similarly discuss a list of rights and their permissible restrictions that could justify TLMPs.We take a different approach in this paper.We argue that TLMPs that realize 'triple benefits' through procedures of fair representation are not merely ethically acceptable but can also be normatively recommended provided they meet some conditions that we specify in this paper.We take the concept of 'triple benefits' seriously and ask: What are the substantive and procedural conditions for TLMPs to be justified on these grounds?
In our response to the dilemma of balancing domestic and global justice concerns we aim to overcome the opposition between sovereign immigration control and global utilitarianism.While the former view prioritizes rights of destination states to choose whether to admit labour migrants, how many and under what conditions, the latter prioritizes benefits for source countries and prospective migrants over standards of social justice in host countries.As an alternative we propose that destination states have global justice duties which they can partly deliver through well-designed TLMPs that are compatible with their domestic obligations of social justice.The triple benefit should not be interpreted only in utilitarian terms as a strategy for the overall maximization of welfare, but as the condition under which it is possible to reconcile domestic with global justice concerns.
A second element of our response is that the normative acceptability of TLMPs is not merely a matter of social justice but also of democratic legitimacy.We reject views that consider democratic inclusion only in relation to the host state (e.g.Walzer 1983).We argue instead that TLMPs should satisfy three principles of democratic inclusion.First, all relevant affected interests must be included through fair representation of destination countries, source countries and temporary migrants in negotiating TLMPs.Second, in destination states temporary migrants must be included not as future citizens but as subjects to the laws.This requires that they are granted equal protection by the laws and opportunities to contest these laws.The third principle of democratic inclusion is that temporary migrants are citizenship stakeholders in their countries of origin.These countries have duties to help them realize their life projects and to involve them in shaping the future of these societies.The principles of including all subjected to the law and all citizenship stakeholders apply to destination and origin states respectively and determine substantive conditions for justifiability of TLMPs, independently of the outcome of fair negotiations shaping the specific content of such programmes.
We develop our argument in four steps.We begin, in section 2, with a discussion of the 'anatomy' and sources of the dilemma at the heart of TLMPs, drawing on relevant empirical research.Next, section 3 reviews existing normative responses to the dilemma raised by TLMPs and discusses difficulties with assessing whether TLMPs generate 'triple-benefits' for migrants as well as their countries of origin and destination.We then develop our normative argument on domestic and global justice conditions (section 4) and democratic legitimacy conditions (section 5) for TLMPs.The conclusion identifies and discusses briefly some of the practical challenges with designing and implementing TLMPs that meet the justice and democratic legitimacy conditions we identify in this paper.We conclude that our proposed new approach to TLMPs can help to address, but never fully resolve the inescapable ethical dilemma that such programmes raise under real world conditions.

Anatomy of the dilemma: Expanded migration under restricted rights
Most TLMPs are unilateral programmes that are designed and implemented by destination countries.Where bilateral programmes exist (and their number has grown in recent years, see e.g.Peters 2019;Chilton and Posner 2018), most of the bargaining power has remained with host rather than origin countries.As a consequence, the vast majority of today's TLMPs are largely 'made' in host countries.They are typically consequentialist policies designed to benefit primarily the interests of the host country's population alone.A key objective of all TLMPs is to promote economic growth, partly through helping to fill labour and skill shortages, in the host country.A second objective, the practical significance of which varies across countries and over the business cycle, is distributional, i.e. to make sure that TLMPs do not harm (certain groups of) domestic workers.
While there are considerable variations in the policy design of TLMPs around the word (see e.g.Ruhs 2018), all such programmes are characterized by a fundamental trade-off: they provide opportunities for labour migration while at the same time restricting at least some of the admitted workers' rights (compared to the rights of citizens and log-term residents in the host country).By definition, migrants' right to reside in the host country is time-limited, at least initially.Many countries allow some migrants admitted through TLMPs to transfer to permanent residence status after some time (often after 5 years), but only if they meet certain criteria.In democratic countries, strictly temporary programmes, where there are no opportunities at all for migrants to upgrade to permanent residence status, are generally limited to lower-skilled (including seasonal) admission programmes.In non-democratic countries, such as the oil-rich Gulf States, strictly temporary programmes are also used for medium and higher skilled migrant workers.
Most TLMPs, except those for admitting highly skilled migrant workers5, also restrict a range of other rights of migrant workers.These almost always include the right to free choice of employment in the host country's labour market.The vast majority of TLMPs issue work permits that limit the employment of the admitted migrant to the employer specified on the permit.Changing employers may be possible after some time, but it usually requires a new work permit application.From the host country's perspective, a fundamental rationale of TLMPs is to help reduce labour and skills shortages in specific occupations and/or sectors.If the admitted migrants were free to take up employment in any occupation or sector, TLMPs would not be able to meet one of their fundamental objectives.Some countries (such as Ireland) have introduced policies that allow temporary migrant workers to switch employers freely within certain sectors or occupations after some time (e.g. one year after admission under the TLMP).However, even when the initial tie between worker and specific employer is lifted, the restriction on employment to the occupations or sectors perceived to be in 'shortage' typically remains.
Most TLMPs also restrict migrants' access to welfare benefits, especially to targeted and non-contributory benefits such as social housing and social assistance.Again, the extent to which access to welfare is constrained varies across countries but there are at least some restrictions in most countries.Two rationales are usually given by policy-makers for these restrictions (see the discussion in Ruhs 2013).The first relates to material effects, specifically to the fiscal costs of providing temporary migrants with access to welfare benefits.The common policy objective of "maximizing the net-benefits from labour immigration" typically includes "minimizing the fiscal costs"and in many countries this objective is pursed partly through restrictions of migrants' access to certain welfare benefits, especially for lower-skilled workers (although typically not to basic public services such as primary health care).
A second reason for why many countries restrict, at least temporarily, some welfare benefits for migrants admitted under TLMPs relates to normative ideas, including perceptions of fairness and justice among the host country's population.Welfare states have been designed as inherently national projects that are based on a social contract between citizens and the state and that redistribute from rich to poor among citizens (and often also including long-term residents).While the preferred principles for redistribution among existing residents (e.g. based on 'need', 'reciprocity', or 'universal access') vary across countries, when it comes to regulating newly arrived migrants' access to welfare benefits, there is a popular and widespread view among the populations of many net immigration countries that prior contribution and reciprocity should be the guiding principles (see e.g.Reeskens and Oorshot 2012).In other words, new migrants are widely seen as having to "earn their rights" to welfare benefits in the host country.Restricting the social rights of new migrants admitted under TLMPs, especially their access to needs-based welfare benefits, is therefore not only perceived as an issue of costs and benefits but also one that raises deep-seated questions about fairness and domestic justice (also see Ruhs and Palme 2018).
Many but not all TLMPs also restrict migrants' rights to family reunion, although this varies considerably across programmes targeting high-and lower-skilled workers (for the latter family rights are typically more restricted).In practice, a key policy tool to restrict family reunion is to require a minimum income threshold that migrants must meet to be able to bring their dependent spouses and children.How high this threshold should be has been subject of considerable public and policy debates in many countries (e.g. for the UK, see Sumption and Vargas-Silva 2016).The centrality of the minimum income threshold indicates that a key concern many countries have about family reunion relates to fiscal costs (rather than to issues of fairness or justice that are important additional considerations in the case of welfare benefits).

Normative responses and the elusive "Triple Win"
Most normative critiques of TLMPs are based on arguments relating to human rights, equal membership, and/or exploitation.From a human rights perspective, critics have argued that TLMPs violate the fundamental principles of universality, indivisibility and inalienability of human rights, and that the rights restrictions under TLMPs are often incompatible with the UN Convention on the Rights of Migrant Workers and Their Family Members.6 Universalistic rightsbased normative frameworks naturally oppose the idea of different categories of residents with different sets of rights (see e.g.Weissbrodt 2008; Taran 2000).
A different but equally influential critique of TLMPs is based on the idea that a democratic community must provide all its residents with equal terms of membership and access to citizenship rights.Michael Walzer (1983) famously argues that if democratic communities are not prepared to admit migrants as equal members, they should not admit them at all.David Miller (2008) makes a similar argument that also acknowledges that the requirement of equal membership and citizenship rights is likely to result in fewer migrant workers admitted.
A third line of normative arguments against TLMPs is based on the idea that such programmes are inevitably exploitative.Exploitation can and has been defined in different ways.Daniel Attas (2000), for example, suggests that guest worker programmes are exploitative not because of a lack of political rights (which, Attas argues, are not required) but because they deny migrants full economic rights which gives rise to 'unequal exchange'.Attas suggests that giving migrants the same economic rights as local workers, including the right to free choice of employment, would remove the exploitative element of TLMPs.
Most normative justifications and defenses of TLMPs have typically been based on a 'realistic approach' to the ethics of migration coupled with a focus on migrant agency and global utilitarianism.For example Howard Chang (2002), writing about TLMPs in the United States, and Daniel Bell (2006), writing about TLMPs for domestic workers in East Asia, start with the observation that equal citizenship rights for migrant workers are politically unfeasible, and that the most likely alternative to TLMPs would be exclusion of migrants.Emphasizing the agency of migrants and the benefits of migration for individuals, both Chang and Bell support TLMPs that restrict some of the rights of migrant workers, under certain conditions (see also Ruhs 2013).Arguments for TLMPs based on global utilitarianism, i.e. based on the idea that TLMPs can and do generate large economic benefits for migrants and their familiesare typically motivated by concerns with poverty reduction in poor countries (e.g.Pritchett 2006) and/or reducing global inequalities (e.g.Milanovic 2018).
A key idea that is common to most existing normative justifications of TLMPsand that also plays a central role in our own normative argument developed in the next two sectionsis that TLMPs can generate 'triple wins' for host countries, migrants and their countries of origin.We prefer the terminology of 'triple benefits' in order to signal that we count among benefits that TLMPs help states to comply with duties of justice and build cooperative relations with other states, which would not normally be understood as 'wins'associated with purely selfinterested goals.Our position differs from a utilitarian one in that we do not advocate a policy design that maximizes overall benefits but one that guarantees fair conditions under which all actors can mutually endorse the benefits achieved by each of them.
Defining the meaning of 'benefits' for host countries, migrants and origin countries, and establishing that they exist under TLMPs, is by no means easy.This is because the consequences of migration for individuals, communities, and countries as a whole are multidimensional and they canand typically doinvolve competing impacts (or trade-offs) for each of the three groups (host countries, migrants and countries of origin).
For example, from the host country perspective, the employment of migrant workers admitted through TLMPs does not only generate a range of economic and social impacts (e.g. on labour markets, public finance, economic growth as well housing, community cohesion etc.) but has also distributional effects (e.g.different costs and benefits for different groups of people such as employers and workers).Moreover, we know that the impacts of labour immigration vary between the short-and longer-run.For example, while immigration may lower the wages of some competing domestic workers in the short run, it may lead to an increase in these workers' wages in the longer run as labour demand adjusts to the immigration-induced increase in labour supply (e.g.Dustmann et al 2008).Some of these effects (e.g.labour market impacts) are much easier to define and measure than others (e.g.social effects).This multitude of impacts makes it very difficult to speak about an 'overall impact' (or benefit) for the host country.To assess the overall effect, the various types of impacts need to be considered, their relative importance assessed (i.e., weighted), and trade-offs need to be managedwhich is an inherently normative exercise as it involves decisions on what type of impacts, and for what groups of people in the host country, should be prioritized.
There are similar considerations and challenges with establishing the overall benefit of TLMPs for origin countries.Labour emigration generates remittances (which are thought to have mostly positive impacts on people left behind in origin countries (see e.g.Clemens and McKenzie 2018;Acosta et al 2006), and it can also be associated with a transfer of skills when migrants return.At the same time, the outflow of skilled workers can have adverse brain drain effects, at least for certain countries (e.g.Docquier and Rapoport 2012).Most origin countries perceive some responsibility for protecting the rights and welfare of their workers abroad.This can lead to trade-offs in perceived impacts and policy objectives vis-à-vis TLMPs: origin countries benefit from remittances generated by TLMPs but this may come at the price of restricted rights for their workers abroad.As a consequence, establishing whether a particular TLMP generates an overall benefit for a specific origin country is not as simple as suggested by some advocates of TLMPs.
Finally, from the perspective of migrants, participation in TLMPs also involves a range of economic and social consequences for themselves and their families, and some of these impacts may be conflicting, which complicates an overall assessment of their benefits.Perhaps the most obvious example is the potential trade-off between economic gains from employment abroad and the sometimes severe loss of rights and freedomsas is the case with temporary migrant workers in the Gulf States.A second trade-off concerns family life.Migrant workers with families can experience family separation as a heavy burden.Often they face a dilemma between performing their duties of care (especially for minor children) or providing their families with better income through temporary migration.Here again, overall benefits are not easy to determine, even when assuming that temporary migrants make the choices that are best for them.Some of these choices may be made under conditions of insufficient information or selfdeception and migrants can come to rationally regret them in retrospect.However, it would be unacceptably paternalistic to assume that it is in migrants' best interest not to participate in TLMPs.Instead, the goal should be to enable them to do so under conditions that will mitigate the trade-offs, e.g. by securing their fundamental rights in the destination country or providing them with multiple re-entry visa that allow them to visit their family members at home.
If it is already hard to figure out the overall benefits for each of the three actors involved, it is even more difficult to weigh them up and aggregate them into an overall formula, as would be required by a global utilitarian calculus.We argue that this is also not necessary.As assessing different types of impacts of TLMPs (and migration more generally) is an inherently controversial and subjective process, we argue that the existence and magnitude of triple benefits of TLMPs cannot be determined without the involvement, or at least the representation of the interests, of all the affected actors (host country, origin country and migrants themselves) in policy design and implementation.
Why, one may respond, is 'revealed preference' not enough to demonstrate the existence of triple benefits of TLMPs?Doesn't the fact that migrants participate in TLMPs, and that origin countries seek to (and sometimes do) cooperate in TLMPs, show that such programmes generate benefits for all sides?The first reason why we reject this argument is the well-known existence of large power asymmetries between high-income host countries on the one hand, and lower-income sending countries and individual migrants on the other.All too often, destination states simply assume benefits for migrants and source countries in order to justify the pursuit of their own interests in TLMPs.
Secondly, revealed preferences are not sufficient for achieving possible benefits for all sides if collective action problems arise.In the absence of TLMPs, labour migration is still likely to occur but in circumstances that lead to far less beneficial outcomes than could be achieved under effective and fair TLMPs.Migrants are likely to choose irregular routes, which makes them much worse off in terms of migration costs and protection of their rights, and undermines the rule of law in the host state; destination countries can damage the interests of source countries by poaching high skilled migrants without sharing in the costs for their human capital development; source countries can blackmail destination states by threatening to facilitate irregular migration and refusing to readmit their undocumented nationals when these are deported.
When collective action dilemmas arise within states, they can be overcome by governments imposing cooperative solutions or creating incentives for all actors to prefer these solutions over defection.TLMPs, however, involve independent states and thus require coordination between actors that are not subjected to the same laws and do not routinely cooperate on multiple issues.In such contexts, what counts as a benefit for each and all of the actors must be established through deliberation rather than the revealed preferences formed independently in non-cooperate settings.We argue therefore that fair representation of interests in negotiating TLMPs and in their ongoing governance is crucial for determining what counts as triple benefits and for achieving these.
In a normative account we cannot, however, leave completely open what kinds of interests ought to be accepted by the other actors as morally justified and democratically legitimate.We address the question of justification in the next section and that of legitimacy in section 5.

Domestic and global justice conditions for TLMPs
A first premise in any liberal justification of TLMPs must be that the migration involved is voluntary.In a forthcoming book, Valeria Ottonelli and Tiziana Torresi (forthcoming) criticize normative migration theorists for assuming that most migrants intend to stay permanently in their destination countries and ultimately want to become citizens of those countries (see also Ottonelli and Torresi 2010).They provide empirical evidence that this is generally not the case and argue that liberal host states ought to accommodate temporary migrants' life projects, which include an intention to return after reaching specific savings targets. 7Ottonelli and Torresi also argue that persons can be considered as having voluntarily chosen temporary migration even if such choices are made under conditions of a lack of opportunities in their countries of origin.Regarding as involuntary all migration that happens in contexts of large opportunity disparities would ignore migrants' agency and disrespect their life projects.However, the emphasis on migrants' voluntary choices also leads these authors to reject enforced temporariness as a feature of TLMPs: "The obvious cases in which temporary migration is non-voluntary are all those programs by which migrants are forced to return to their country of origin after a fixed period of residence abroad as guest workers" (forthcoming).In our view, this stance is self-defeating, as it would imply that temporary migrants could only realize their goals if they were accepted into permanent immigration programmes and could freely choose whether and how long to stay.Instead of considering only what conditions in the host society would allow temporary migrants to pursue their life plans and remain free in changing them, we believe that TLMPs should aim to achieve a triple benefit and must thus factor in the interests of host and origin countries alongside those of the migrants.If they fail to do so, states will not be motivated to provide opportunities for temporary labour migration in the first place.
An alternative view could consider TLMPs as second-best options under non-ideal conditions where state borders are not fully open (e.g.Chang 2002).Enforced temporariness would then only be acceptable conditionally because rich states fail to comply with their moral duties to open their borders for immigration from poor countries whose citizens enjoy fewer opportunities because of their morally arbitrary circumstances of birth (Carens 2013).If the goal is open borders, then one might consider free movement arrangements between states (such as those in place in the European Union) in the current world as approximating this ideal.By contrast, TLMPs can hardly be justified as a step towards open borders since they are premised on destination states' immigration control and impose limits on migrants' right to stay.
While we endorse the widening of free movement opportunities, we do not believe that this is an alternative to, or substitute for TLMPs.Free movement areas are typically created between countries with comparable levels of average incomes. 8In contrast, TLMPs open up legal migration channels between states set apart by great disparities with regard to their levels of economic opportunities and social citizenship.In democratic states, governments have a mandate to promote the common good of their citizens and residents and they cannot be expected to open up borders for free movement from states with vastly lower levels of opportunities and welfare.They must be able to show that admitting economic migrants benefits not only these migrants and their countries of origin but also their own populations.Where disparities are small, free movement can be defended on precisely these grounds; where they are large, only controlled immigration can potentially meet the triple benefit condition.
Furthermore, immigration control is not merely justified on grounds of a priority of domestic social justice, but also because controlled migration makes it possible to address the interests of sending countries and migrants.It allows to do this through governance mechanisms instead of expecting them to be taken care of by the invisible hand of uncoordinated individual choices of migrants, their families and host country employers.Free movement is not always optimal to realize benefits for source countries as shown by the example of the EU where EU enlargement in the 2000s has significantly enhanced East-West labour mobility leading to severe depopulation and loss of skilled workers in some new member states (Krastev and Holmes 2019).Under non-ideal world conditions, immigration control is thus not a regrettable fact resulting from wealthy states' failure to comply with their moral duties to open their borders, as Carens would see it; it is instead a fundamental premise of the duties that such states have not only towards their own citizens but also towards people in poorer countries.
If the migrants participating in TLMPs do not have a normative claim to free movement, what kind of admission claims do they have?Unlike asylum seekers or family members of settled refugees and immigrants, temporary labour migrants do not have individual admission rights.Instead, their admission requires (different types of) consent by all three actors.Migrants themselves can apply and need to be accepted by destination states, or they can be invited through active recruitment and need then to consent.By contrast, because of the fundamental human right to leave any country, the source country must not control, and thus does not have to consent to, the participation of individual migrants in a TLMP.Instead, it is essential that source countries have a say in the general conditions under which their citizens live and work abroad and can act to protect them when their rights are violated.This applies particularly to the case of low skilled labour migrants.Programmes for the temporary migration of highly skilled migrants or international students are less likely to be associated with exploitative conditions for migrants (although some degree of exploitation can also occur in these programmes, see e.g.Costa 2017).Here, interests of sending countries that ought to be taken into account concern mainly the so-called brain drain that can occur under certain circumstances and a fair return on their investment into the human capital of these migrants.
Although TLMPs require these forms of mutual consent and temporary migrants do not have individual admission rights, it does not follow that destination states are under no obligation to admit any temporary labour migrants at all.Even 'weakly cosmopolitan theories' agree that wealthy countries of the Global North share responsibilities for securing fundamental human rights and the satisfaction of basic human needs in the Global South (Miller 2016).Development economists have convincingly demonstrated that temporary migrants' remittances contribute more to poverty reduction and human development in lowincome countries than Official Development Assistance (e.g.World Bank 2017).Under these circumstances, opening up temporary labour migration programs can become an important way how rich states can alleviate poverty and improve the situation in other parts of the world.This argument runs counter to intuitions that destination states cannot be duty-bound to do something that is anyway in their self-interest.Yet duty and self-interest are two rationales for action that do not exclude each other.Partners in a cooperative scheme participate because they expect to benefit but also have duties of fairness to ensure that the other participants benefit as well.In the case of North-South relations, there are additional asymmetric duties that rich countries in the North have towards poor ones in the South.If they can meet these at least partially through well-designed TLMPs without thereby damaging their self-interests, this removes an important cost and feasibility excuse for shunning their global justice duties and strengthens the case for temporary labour migration programs.9 Let's take stock of our normative conclusions so far.We have argued that migrants' participation in TLMPs must be voluntary and can be regarded so if these programmes match their own temporary migration projects.Migrants do not have individual claims to be admitted under such programmes but wealthy destination states may still have special and global justice duties to provide ample opportunities for temporary migration.The triple win justification ensures that destination states can legitimately give priority to the promotion of opportunities and social rights of their own citizens and residents, but it obliges them also to take into account the interests of migrants and source countries in the design of such programmes.These conclusions deviate significantly from most of the existing normative literature.However, they are not yet very specific and leave the content of TLMPs and the conditions under which temporary labour migration can be regarded as satisfying a triple benefit justification wide open.

Democratic legitimacy conditions for TLMPs
We explore therefore a second normative perspective that complements principles of justice with those of democratic legitimacy.This perspective brings several problems into focus.First, decisions on TLMPs adopted unilaterally by a destination state have large spill-over effects on other countries and their citizens.Second, once inside the country, temporary migrants are fully subjected to the host country's laws and coercive state power but they are not adequately represented in the making of these laws and, as pointed out by Ottonelli and Torresi (2010), may actually not be interested in host country politics because of their plans to stay only temporarily.Third, temporary migrants are citizens of their countries of origin and may need to rely on these countries' governments for protection of their interests, but often find that their absentee status diminishes their political clout or that home country governments use them only instrumentally for their own economic or political purposes.
TLMPs raise issues of democratic inclusion and representation and the interests that the three actors bring to the negotiations must therefore be vetted for their democratic legitimacy.There has been a vigorous debate among political theorists about competing principles of inclusion.We adopt a pluralist view and propose that democratic legitimacy requires combining three distinct principles, each of which has different scopes and domains of application (Bauböck 2018).
The first of these principles is the inclusion of all relevant affected interests.This principle responds to the question of whose interests need to be taken into account in a collectively binding decision.It applies specifically to extraterritorial spillover effects of decisions taken by one particular polity.Interpreted in the right way, the principle does not call for including those whose interests are affected as citizens and voters in the polity taking the decision.Instead, it requires an adequate representation of interests in the deliberation before the decision or in the decision itself.On many policies that have negative extraterritorial effects, consulting the governments of other affected states and NGOs representing the interests of affected populations may be enough to generate legitimacy for the decision.Some authors have also argued that the affected interests principle applies only to negative effects, as outsiders can be assumed to agree to receiving a benefit. 10If a country does not comply with carbon emission targets and thereby harms the future of humankind it can be legitimately pushed by outsiders to change its policy, but if it adopts a policy that reduces its emissions below an agreed and fair target, then outsiders should not have a say in deciding how it does so.However, the triple benefit of TLMPs is not of this kind.As we have argued above, benefits for migrants and sources countries do not come about as side-effects of TLMPs unilaterally adopted by destination states and pursuing only their economic selfinterest.Instead, triple benefits can only be secured through cooperation.We propose therefore that the democratic principle of including affected interests calls for international decision-making procedures on TLMPs in which the interests of destination states, origin states and migrants are fairly represented.
The second principle, which hasin our view wronglybeen considered as a rival to that of including affected interests is that of including all persons subject to the law and/or coercive political power.This principle provides legitimacy to restrictions of individual autonomy by political authorities through demanding that they at the same time secure this autonomy through constitutionalized liberties and by enabling individuals to contest these authorities.This principle has generally a narrower scope than that of including affected interests, as it applies primarily within the territorial jurisdiction of a particular polity, but its domain of application is broader as it does not merely concern the legitimacy of particular decisions but of the whole system of government institutions and their coercive powers.
Applied to temporary labour migrants, the principle identifies them as subjected to the law and authorities of the host society in a comprehensive way that gives rise to corresponding claims to inclusion.In contrast to most authors, however, we do not think that subjection is a sufficient condition for a claim to full citizenship status.The case of temporary labour migrants illustrates this point.On the one hand, they are not transients like tourists, travelling business people or border commuters who are also subjected to the law but whose presence in the territory is less essential for their life projects.On the other hand, they are also not like settled immigrants and native-born residents who locate the centre of their future lives in the territory.For that reason, they do not have a claim to citizenship and voting rights in the host country.Yet they do have claims to equal protection by the laws to which they are subjected as well as to contestation of these laws (e.g. through staging protests).The 'all-subjected principle' is not compatible with the view that temporary migrants' rights can be entirely derived from agreements fleshed out in negotiations representing all affected interests.Nor can the content of such rights be fully determined by considering how their life plans ought to be accommodated by host states, as Ottonelli and Torresi propose.Both of these considerations will play an important role in justifying deviations from a standard of equal rights for workers and residents in the host society.But before justifying deviations, we first need to assert a default standard of equality.The principle of equal protection by the law for all subjected to the law provides that standard.
The third and final principle is that of including all citizenship stakeholders.It identifies those individuals whose lives are linked to a particular polity in such a way that they have a claim to full citizenship status.We suggest thatunlike long-term settled immigrantstemporary migrants generally do not have a claim to citizenship in their host countries.They may, however, have a claim to citizenship and voting rights at the local level, since local citizenship is structurally open for all who take up residence in the municipalityas illustrated by local voting rights for mobile EU citizens but also for third country nationals in many European and South American states (Pedroza 2019).We think that offering local citizenship and voting rights to temporary migrants is important as it provides them with additional protectionsymbolically through a status of temporary membership and practically through the attention that candidates have to pay to the interests of potential voterseven if these do not turn out in large numbers, as temporary migrants are unlikely to do.More important is, however, the claim of temporary migrants to effective citizenship in their countries of origin.As temporary absentees they should not lose their voting rights and as a particularly vulnerable 'diaspora' group they need effective consular and diplomatic protection.Because they are and remain citizens, countries of origin also have special duties to assist them in realizing their life plans through facilitating remittances, return migration, and reintegration after return.
Principles of democratic inclusion can in this way fill some of the gaps that a normative account based on principles of justice leaves open.They explain why the content of TLMPs should be determined by negotiations in which the affected interests of source countries and migrants themselves are fairly represented, why and in which respect temporary migrants can claim equal treatment with other residents in the host society and why they have claims to transnational citizenship, which include membership at the local level in the host country and citizenship rights in the sending state.

Conclusions
This paper has argued that temporary labour migration programmes are caught in a dilemma between requirements of domestic justice in migrant receiving states that call for maintaining equal rights and standards of social justice for all subjected to the laws of the country and global justice duties of wealthy states to provide more opportunities for (temporary) migration from poorer countries.Where cross-country disparities of economic development and social rights are very large, free movement would undermine the capacity of states to deliver domestic social justice.Regulated temporary migration can, however, be justified as a response to the dilemma, if it provides benefits for receiving states, sending states and the migrants themselves.
However, what counts as relevant benefits is difficult to determine ex ante.The impacts of TLMPs on the three actors involve possible gains and losses for each of them.Moreover, achieving triple benefits also depends on cooperation between the actors and cannot be expected if each aims to maximize their own interests without coordinating with the other actors involved.We have therefore proposed that triple benefits can only be achieved if TLMPs are negotiated and governed in such a way that all affected interests are fairly represented.This conclusion is supported by our interpretation of the democratic principle of including all affected interests.Under such conditions, some rights of temporary migrant workers in host countries can be regarded as a legitimate outcome of negotiations, while others need to be fixed in advance under a democratic principle of equal protection of all subjected to the laws.It is not our ambition in this paper to list those rights that are conditions for fair TLMPs and those that can be flexibly negotiated.Our aim is instead to show that the domestic vs. global justice dilemma can be addressed in this way even though it cannot be fully overcome.
Our paper provides a new normative framework for discussing the desirability of TLMPs, and for debating their design and implementation in practice.As such, it raises many issues that require further research and analysis.To conclude, we highlight briefly three important practical issues and challenges that arise when trying to fairly negotiate and govern a TLMP along the lines we have suggested.Two of these challenges emerge from the problem of structural asymmetries of power between the three categories of actors that must be addressed to ensure procedural fairness.The first asymmetry is that between wealthy receiving countries and poor sending states, the second is that between states that are by their very nature organisations representing collective interests vis-à-vis other states and migrants that lack representation of their special interests.
One way of overcoming the asymmetric dominance of destination states in negotiations about TLMPs would be to involve international organisations (such as the International Labour Organisation or the International Organization for Migration, two UN agencies) as convenors of the discussions, which could thus be conducted as multilateral negotiations where destination and origin states have formally equal powers.In such a setup, the international organisations' role would be to facilitate deliberations among what are meant to be equal parties in the negotiation, without actively supporting one side or another.While it would be naïve to think that the involvement of an international organisation as a convenor would eliminate power asymmetries between rich destination countries and poorer sending states, it may go some way toward reducing them.This might be especially the case if TLMPs are negotiated between groups of destination and origin countries and/or if TLMPs are negotiated as part of wider migration policy agreements that include areas where origin countries have more leverage (e.g. as part of attempts to reduce irregular migration through cooperative policies between destination, origin and transit countries).
The second challenge, which is how to best represent the interests of temporary labour migrants in multilateral negotiations about TLMPs, is equally important and difficult to address.The problem is that temporary labour migrants do not have their own organizations to represent them and are less motivated to spend time and resources for creating them than other categories of workers.Therefore, they need vicarious representation through, for example, NGOs.11Involving NGOs in decision-making processes is often criticized by arguing that they lack democratic legitimacy.In our view, this critique is less convincing where direct representation cannot be achieved and when the issue at stake needs to be addressed through transnational governance rather than domestic government (Macdonald 2008).Identifying an NGO that is suitable for this purpose is, however, not a straightforward issue, as many of the existing civil society organisations dealing with labour migration are focused on representing specific groups of migrant workers or specific aspects of labour migration and protection.
In the negotiation of TLMPs, migrants ought to be represented not only in discussions but also in the decision-making, possibly through some kind of veto power.Multilateral TLMPs involving several host and source countries offer not only better chances for evening out power asymmetries between states, but also for giving representatives of temporary labour migrants an effective voice in the negotiations.NGOs representing migrants and IOs acting as conveners of the discussions should also have a strong role at the implementation stage where they could be in charge of independent monitoring of these agreements.
A third important issue relates to non-democratic countries.Our approach may be described as semi-(non)-ideal theory in the sense that we presuppose a non-ideal context of global distributive injustice with large discrepancies of opportunity between countries in the global South and global North.At the same time, we have assumed that governments in destination and origin countries will adequately represent the best interests of their citizens and that they accept fundamental principles of democratic inclusion.Real-world liberal democracies often fail to meet these assumptions, but they are at least normatively committed to them.By contrast, non-democratic states involved in TLMPs may reject the conceptions of domestic justice and democratic legitimacy that inform our normative argument.On the one hand, we accept that our normative argument has limited purchase for TLMPs concluded between governments lacking democratic legitimacy.On the other hand, we remain optimistic that "model TLMPs" worked out between democratic states and international organisations might provide new standards that could eventually also be accepted by non-democratic destination or origin states, which might adopt them to bolster their domestic legitimacy and international reputation.