Discrimination in Migration Regulation

Discrimination in Migration Regulation

Fordham International Law Journal Volume 36, Issue 3 2013 Article 2 A Minimal Liberal Defense of (Some) Discrimination in Migration Regulation Tome...

2MB Sizes 0 Downloads 4 Views

Fordham International Law Journal Volume 36, Issue 3

2013

Article 2

A Minimal Liberal Defense of (Some) Discrimination in Migration Regulation Tomer Broude∗



Hebrew Univeristy of Jerusalem

c Copyright 2013 by the authors. Fordham International Law Journal is produced by The Berkeley Electronic Press (bepress). http://ir.lawnet.fordham.edu/ilj

A MINIMAL LIBERAL DEFENSE OF (SOME) DISCRIMINATION IN MIGRATION REGULATION Tomer Broude' INTRODUCTION .................................... I. THE RELEVANCE OF THE MINIMAL LIBERAL FRAMEWORK TO MIGRATION REGULATION ........ II. THE GENERAL CASE OF DISCRIMINATION ..... III. ABUSIVE DISCRIMINATION IV. MIGRATION-ENCOURAGING AND MIGRATIONDETERRING DISCRIMINATION ................. CONCLUSION ............................. .......

575 579 583 5....... 87 593 598

INTRODUCTION The regulation of migration-any regulation of migration-is inherently discriminatory against prospective migrants.' Immigration policy determines who gets in, often consigning those who do not, to less advantageous living conditions in their country of origin. After all, most migrants are simply interested in improving their quality of life.2 Quantitative restrictions and qualitative selection criteria imposed by national immigration policies constitute "at-the-border" discrimination. This is augmented by "behind-the-border" measures that treat immigrants less favorably than citizens after admission (such as lower levels of social protection). Neither "at-the-border" nor

Sylvan M. Cohen Chair in Law, Hebrew University of jerusalem; Visiting Professor, University of Virginia Law School. 1. "Discriminatory" here is not a value-statement, it is merely referring to objectively differential treatment: " [T]he term, 'discrimination,' has value implications that can never be completely cradicated, though they can be Stcrilizcd for specific empirical and descriptive purposes." See Kenneth Arrow, The Theory ofDiscimination, itn DISCRIMINATION IN LABOR MARKETs 3, 3 (Orley Ashentelter & Albert Rees cds., 1973). 2. Among migrants, the "vast majority . . . move in search of better economic opportunities." Only ten percent are considered refugees whose migration stems from a fear of persecution. See INT'L ORG. FOR MIGRATION, WORLD MIGRATION REPORT 2005: COSTS AND BLNEITS OF INTLRNATIONAL MIGRATION 379-81 (2005).

575

576

FORDHAM INTERNATIONAL LAWJOURNAL

[Vol. 36:575

"behind-the-border" distinctions sit well with egalitarian ideals, but their reduction or removal is often perceived as a threat to the cohesion and welfare of political communities in migrationreceiving states, including those who consider equality to be a core social value. A tension therefore emerges between egalitarian mores and communitarian concerns.4 The debate on immigration policy in liberal, democratic, immigration-receiving states can thus be formulated as a debate over boundaries of unequal treatment of immigrants. To the extent that "open borders"-advocated by some cosmopolitans-are not politically viable (and would clash with communitarian perspectives), it becomes necessary to ask what are the morally permissible boundaries of "behind-the-border" discrimination? This communitarian-egalitarian tension is well-reflected in the ambivalence of international law. It is generally accepted that immigration regulation is the state's prerogative." With few exceptions, states are not obligated to introduce immigrants to their territory, determining conditions for admittance (and citizenship) as they see fit. However, various sources of 3. Cominunitarians have argued that state-based distinctions, including restrictions on immigration, are consistent wvith egalitarian principles. See, e.g., Michael Blake, Distibutiejustice, State Coercion and Autonomy, 30 PHIL. & PUB. AFF. 257 (2001); Thomas Nagel, The Problem of Globalfustice 33 PHIL. & PUB.AFF. 113. 129 (2005). 4. I use the term communitarian here broadly, referring to any approach that attaches m1oral significance to historical and/or cultural communjities, from which one

can derive the "respect that foreigners owe to a historic comiunity and its internal life." Michael Walzer, The Woral Standing oj States: A Response to Four Critics, 9 PHIL. & PUB. AFF. 209, 212 (1980). 5. Compare Audrey Macklin, Freeing Migration from the State: Michael Trebilcock on Aigration Policy, 60 U. TORONTO I.]. 315, 320 (2010), with Audrey Macklin, A Theory About Theories of Migration, and a Comment About Michael Trebilcock's Theoly (2009). available at http://www.law.utoronto.ca/documents/conferences2/Trebilcock09 Macklin. pdf (" [T]he concern animating theoretical debates about migration is less about liberty than about somie version of equality."). 6. See generaly Joseph Carens, Aliens and Citizens: The Casefor Open Borders, 49 RLV. Pot. 251 (1987). 7. References here to "cosmopolitanism" and "coLmunitarianisn" deliberately condense a wide spctrum of approaches. For a imore nuanced survey of different approaches to immigration see Michael j. Trebilcock, The Law and Economics of humigrationPoli,. 5 AM. L. & ECON. RLV. 271. 294-96 (2003). 8. T. Alexander Alcinikoff, InternationalLegal Norms and Migration: A Report. in MIGRATION & INTERNATIONAL LEGAL NORMS (T. Alexander Aleinikoff & Vincent Chetail eds., 2003). 9. The chief exceptions being asylun-seckers under the 1951 United Nations Convention Relating to the Status of Refugees, and the 1967 United Nations Protocol

20131

DISCRIMINATION IN MIGRATION REGULATION

577

international law including labor law, 10 human rights law," and economic law,"1 express a generally negative position towards discrimination against immigrants (though for divergent reasons). The aspiration of equal treatment of immigrants challenges the discrimination that is inherent in the general right to exclude, yet also cohabits with it. This inconsistency is even more intricate because equal treatment does not necessarily walk hand in hand with inclusiveness either. Quite the contrary: rigid applications of "behind-the-border" non-discriminatory treatment of migrants can constitute effective barriers to immigration. Indeed, egalitarian ideals and rules of nondiscrimination may have the counterintuitive effect of restricting immigration-greater protection of rights leading to a lower number of migrant beneficiaries.' 3 If disparity between the quality of life drives migration, then inflexible equal treatment may raise the beam for migration higher, deterring it and perpetuating global inequality. Drawing the moral bounds of justifiable discrimination therefore has real implications for the design of national and international immigration law and policy. This is the case with respect to temporary migration, particularly in areas of low-skilled labor, where the economic demand for

Relating to the Status of Reiugees. See Convention Relating to the Status of Refugees, July 28, 1951, 189 U.N.T.S. 150; Protocol Relating to the Status of Refugees, Jan. 31, 1967, 19 U.S.T. 6223, 606 U.N.T.S. 268. Refugees are not the subject of this Article's analysis. 10. See Convention Concerning Migration for Employment, July 1. 1949. I.L.O. No. 97, art. 6, 1616 U.N.T.S. 120 [hereinafter ILO 97]; see also INTI LABOUR OFFICE, MU TILATERAL FRAMEWORK ON LABOUR MIGRATION: NON-BINDING PRINCIPIES AN) GUIDELINES FOR A RIGHTS-BASLD APPROACH To LABOUR MIGRATION §8.4.4 (2006). 11. Stemming from the general rule on equality in accordance wvith the International Covenant on Economic, Social and Cultural Rights, art 3, Dec. 16, 1966, 993 U.N.T.S. 3 [hereinafter 1((PR]. For discussion, see infra Part IV. 12. See General Agreement on Trade in Services, Apr. 15, 1994, Marrakesh Agreement Establishing the 'World Trade Organization, art. I(2) (d), Legal Instunents-Results of the Uruguay Round, 33 I.L.M. 1125 [hereinafter GATS] (precluding, generally, discrimination between sources of temporary service providers who are essentially labor migrants). For a discussion on the relationship between economic nondiscrimination and human rights-based equal protection see also Toiner Broude, The Wost-Favoured Nation Principle, Equal Protection, and Migration Policy, 24 GLO. IMMIGR. LJ. 553 (2010).

13. See Martin Ruhs & Philip Martin, Numbers vs. Rights: Trade-Offs and Guest Worker Programs, 41 INT'. MIGRATION REV. 249, 254 (2008).

578

FORDHAM INTERNATIONAL LAWJOURNAL

[Vol. 36:575

foreigners in migration-receiving countries is often steep, and the temptation to provide them with fewer rights is high.14 Without prejudice to the fundamental debate over "open borders" and the general liberalization of migration, in this article I develop a plausible case in defense of (some) "behindthe-border" discrimination in migration regulation. The overall claim is that within identifiable limits, it is morally acceptable, and at times even preferable, to treat immigrants less favorably than residents. The framework of analysis I employ is a specific application of the "Minimal Liberal" rule of social choice developed by Amartya Sen in his initial exposition of the Liberal Paradox, as explained below.1 Unlike the weak Paretian decision rule with which it will be contrasted, in the Minimal Liberal framework the collective efficiency of decisional outcomes is not considered. Instead, each participating actor can determine the outcome of at least one choice between alternatives, as an expression of individual liberty. When relevant actors represent host- and sender-societies, the results are illuminating. In Part I, I will provide justification for the employment of a Minimal Liberal framework in the context of migration regulation. In Part II, I will set out the Minimal Liberal defense of discrimination in migration. In Part III, I will apply the same framework to discriminatory treatment that simultaneously constitutes a violation of migrants' substantive human rights, arguing that this latter type of discrimination cannot be justified. In Part IV, I apply the framework to two different forms of "behind-the-border" discrimination in migration regulationimmigration-encouraging discrimination, and immigrationdeterring discrimination-showing that while both of these are defensible, migration-encouraging discrimination is preferred. In the conclusion, I summarize the defense of discrimination in migration regulation and its implications for public policy.

14. On the morality of diflerential conditions for guest workers, see Robert Mayer, Guestworkers and Exploitation, 67 RLV. POL. 311, 322 (2005). 15. See generally Amartya Sen, The Impossibility of a ParetianLiberal, 78.1. POL. ECON. 152 (1970).

20131

DISCRIMINATION IN MIGRATION REGULATION 1.

579

THE RELE VANCE OF THE MINIMAL LIBERAL FRAMEWORK TO MIGR4 TION REGUL4TION

The qualified defense of discrimination in migration regulation is neither a general justification nor a reproach of migration restrictions, nor of exclusionary communitarian approaches to immigration; and by no means is it a rejection or an embrace of egalitarianism or cosmopolitan views of migration. Rather, it is a recognition of the communitarianegalitarian tension and the need to reconcile it through a moral analytical framework that is not substantially tilted towards either end of the spectrum, but rather engages in a process of public reasoning between them. Moreover, the aim of this reasoning is not to present the types of discrimination defended here as mere second-best compromises or least-worst policies, but to justify them as morally preferable not only to unbridled discrimination but also, in some constellations, to purist nondiscrimination. This said, the analysis is not a paternalist justification of discrimination against immigrants merely because in some way it might be good for them. A crude argument in favor of discrimination might rely on migrants' supposed objective interests as perceived by immigration-receiving states, pointing out that some migrants who suffer discrimination in their hostcountry may nevertheless be better off than they would have been with no migration at all.16 However, this argument is deficient. It rests mainly on the potential migrants' objectively inferior initial condition, inferring from it a preference for migration with discrimination-any discrimination. Taking into account these migrants' presumed objective welfare, this argument ignores their first overall subjective preference, which would surely be immigration with equality. Also, it assumes that incumbents in immigration-receiving states prefer discriminatory treatment of immigrants. However, from their communitarian perspective they might actually prefer full exclusion of immigrants, and from egalitarian viewpoints they might prefer nondiscriminatory immigration. Indeed, 16. For a more sophisticatcd yet related argument, see Mayer, supra note 14 at 319 (suggesting Lha[ moderately exploitativc guest worker programs are acceptable if the

most likely non-exploitative alternative worsens the plight of the disadvantaged).

580

FORDHAM INTERNATIONAL LAWJOURNAL

[Vol. 36:575

aggregating the conflicting moral sentiments of incumbents leads to the conclusion that to them, immigration accompanied by discrimination might be a third-best, first-worst alternative-a necessary evil. Arguably, the question of whether discrimination against immigrants is morally justified should rather take into account the views and preferences of both incumbents and of immigrants who might suffer discriminatory treatment, because the interests of both would be affected. A plausible way of doing this would be to model a global society that includes both communities, and then to apply some rule of social choice to their preferences. The initial problem with this approach, of course, is that both strict communitarians and cosmopolitans would object to the application of a decision-making rule that takes into account the preferences of both groups. In a communitarian's ideal world, while acknowledging that migration-receiving and migration-sending states are part of the same international society, 17 there would be no need or justification for the receiving community to consult the preferences of potential immigrants. Incumbents would be free to design a policy according to their community's interests and values. By contrast, in a cosmopolitan's ideal world, there would be no need or even justification to consult the invariably unjust preferences of the incumbents. Reality is more complicated than both theories, which, for all their intellectual merit, describe and justify ideal-type world orders of justice. In fact, migrants cannot ignore the preferences of immigration-receiving states who hold legal power to regulate entry; incumbents cannot ignore immigrants' preferences, because their control over migration is de facto limited, as evidenced by the high numbers of undocumented immigrants, 8

17. International society is not necessarily a community, but may be understood as a society composed of distinct communities, with varying degrees of political pluralism.

See generally Michael Walzer, International Society: What is the Best that We Can Do? (June 2000) (unpublished manuscript), available at hIIp://,w.sss.ias.edu/files/ papers/papereight.pdf. 18. See GORDON H. HANSON, COUNCIL ON FORLIGN RELATIONS SPLCIAL REPORT: THL ECONOMIC LOGIC OF ILLEGAL IMIilGRATION No. 26, at 3 (2007) ("In the past Len years, the US population of illegal

immigrants has risen from five million to nearly

20131

DISCRIMINATION IN MIGRATION REGULATION

581

and also because their governments invariably allow some immigrants to enter their territory, de jure. The question, therefore, warrants scrutiny through realistic theory, taking into account the preferences of both migrants and incumbents, extracting concessions and compromises from otherwise 'transcendental' or ideal-type paradigms. As Sen has written, instead of clinging to unrealistic theories in the pursuit of social justice, "what is needed ... is an agreement, based on public reasoning, on rankings of alternatives that can be realized"; ' furthermore, "in seeking resolution by public reasoning, there is clearly a strong case for not leaving out the perspectives and reasoning presented by anyone whose assessments are relevant, either because their interests are involved, or because their ways of thinking about these issues throw light on particular judgments."20 These words ring especially true with respect to migration, a field that lies well within the realm of "non-ideal" theory.21 Recourse to public reasoning involving both incumbents and migrants is generally justified by the exigencies of the real world in which people live and die and migrate. In this real world, constituencies are not faced with a binary decision (to allow or disallow immigration), but rather with much more textured policy choices that translate into a ternary decision: no immigration, immigration with "behind-the-border" discrimination, and/or immigration with equality. Cosmopolitans who would ideally choose open, egalitarian borders are forced to consider their priorities in the face of restrictive immigration policies. And communitarians who would ideally choose to isolate or protect their community must nevertheless acknowledge their own propensity to introduce immigrants and hence, the need to choose between discriminatory and egalitarian terms. twvelve million, prompting angry charges that the country has lost control over its borders."). 19. AMARTYA SEN, THE IDEA OFJUSTICE 17 (Allen Lane ed., 2009).

20. Id. at 44. 21. Although Sen pays tribute to Rawls by critiquing his transcendentalism, Rawls

was cognizant of the strengths of "non-ideal" theory, in which one seeks "policies and courses of action that are morally permissible and politically possible as well as likely to be effective." JOH1N RAWLS, THE L.AW OF PEOPLES: NVITH "THE IDEA OF PUBLIC RLASON REVISITED" 89 (1999).

582

FORDHAM INTERNATIONAL LAWJOURNAL

[Vol. 36:575

Crucially, turning to such a process does not ex ante prejudice the substantive outcome towards either cosmopolitan or communitarian ends of the spectrum. A cosmopolitan approach would have already excluded the forms of discrimination contemplated here, through its advocacy of both open borders and equal treatment. A communitarian approach would have led to unilateral migration policy reflective of a general disinclination towards immigration, perhaps banning it entirely. In contrast, a process of consultation with the preferences of both incumbents and immigrants can be outcome-neutral, yet reasoned, leaving the question of "behindthe-border" discrimination open to debate. However, we have still to determine the rule of social choice to be applied to these preferences. The method of aggregating preferences must also be ex ante outcome-neutral. One obviously applicable decision-making rule of social choice would be Pareto optimality. A weak Pareto rule would seek to identify policy on immigration and discrimination that cannot better reflect the preferences of incumbents, without worsening the position of migrants, in terms of its preferences, assuming that this reflects aggregate welfare. Weak Pareto efficiency would, however, be a flawed social choice paradigm, because it would be prejudicially cosmopolitan. Paretian logic is ab defintio a method of considering the preferences of relevant agents, for the purpose of optimizing the efficiency of the combined/ aggregate set of agents. Here, a Pareto analysis would identify the optimally efficient rule for the combined (global) population of migrants and incumbents. This goes beyond a decision rule that merely takes into account the preferences of both communities, because it treats them as a single community -it is the aggregated welfare of all groups that is to be promoted/ maximized, not the relative welfare of each group. This background cosmopolitanism of Paretian analysis is borne out by the results of the analysis below, in which it consistently pushes towards immigration rules that are both inclusive and non-discriminatory-a cosmopolitan ideal. Furthermore, this suggests that in the immigration context, Pareto efficiency has a blind spot with respect to possible distinctions between different forms of discrimination (i.e., all discrimination is Pareto inefficient, period).

20131

DISCRIMINATION IN MIGRATION REGULATION

583

An alternative to the Paretian paradigm, engaging preferences of both incumbents and immigrants without considering them as components of a single group, would be the Minimal Liberal approach. It is suitable here because it deliberately disrupts Paretian logic. Under the Liberal Paradox, Minimal Liberal choices will never be Pareto optimal." The Minimal Liberal framework has no prerequisite of broad egalitarianism. Yet it takes into account the preferences of different groups under equal terms: each group gets to determine at least one social choice. Thus, if there are three alternatives for the final decision-x, y, and z-there are three choices to be made: x/y, x/z, and y/z. Each group may determine the outcome of at least one of these binary decisions. In the present case, these choices would include, to begin with, three alternatives: immigration with equal treatment, immigration with discriminatory treatment, and finally, no immigration. The Minimal Liberal framework provides a rule of social choice that is neither prejudiced towards open migration nor to the question of "behind-the-border" discrimination. The Minimal Liberal framework is not necessarily to be preferred in analyzing migration policy, but it has the distinct advantage of being relatively neutral with respect to the communitarian-egalitarian/cosmopolitan debate and nonprejudicial towards a particular result, focused on public reasoning in a particular way. Let us see where it takes us. II. THE GENERAL CASE OFDISCRIAHA4TION Consider two representative individuals with distinct preferences in migration policy: Alice is an egalitariancommunitarian

in

a

high-income,

immigration-receiving

state,

whose socio-political community attaches great importance to egalitarian values; Roberto is a prospective low-skilled immigrant from a low-income state, whose chief concern is to improve his quality of his life. Let Roberto and Alice constitute international society, each representative of their social community. What

22. Sen has shown (though not without controversy) that the outcomes of a Parctian efficiency approach and a liberal approach are not the samne. See generally Scn, supra note l5.

584

FORDHAM INTERNATIONAL LAWJOURNAL

[Vol. 36:575

social rule should apply to the treatment of migrants like Roberto in Alice's country? Alice and Roberto both have three alternatives: ximmigration with full equality (i.e., upon admittance) Roberto will enjoy the full range of rights Alice does; y-immigration without equality (i.e., Roberto can physically migrate and work in Alice's country but he will not be included in Alice's community and will be subject to fewer rights than those accorded to Alice, as a temporary guest worker for example); and finally z-no immigration at all (i.e., Roberto is refused at the border and his migration is precluded entirely). Alice's preferences (in decreasing order) would be z, x, y. As a communitarian (and out of uncertainty and concern about the effects of immigration on her own personal welfare) she might rather not welcome immigration under any terms, but if immigration is nevertheless allowed, her egalitarian values mandate inclusion and equal treatment. Physical immigration subject to exclusion from the community and discriminatory treatment is to her the least attractive result. In contrast, Roberto's preference ordering is x, y z. His first choice would be inclusive and non-discriminatory admittance, as an immigrant with full rights. Moreover, he believes that immigration with social exclusion and discrimination will still improve his quality of life. The alternative that leaves no chance of migration is least desirable to him. What is the proper social choice among these alternatives, taking into account both Alice's and Roberto's preferences? A weak Paretian rule would lead to x-open and egalitarian migration (it is the only alternative among the three for which Alice and Roberto share a preference, over discriminatory immigration). Let us apply, instead, for the reasons already explained, a Minimal Liberal rule of social choice, that is liberal in that it grants each individual (or represented group) the ability to determine at least one social preference. The binary choice between x (inclusive, egalitarian immigration) and y (immigration with discrimination) cannot be one of the choices allocated to either Alice or Roberto, because x is a preference that is common to both of them, and would therefore not fulfill the terms of Minimal Liberal analysis. This leaves us with two choices to be made, x/z, and y/z.

20131

DISCRIMINATION IN MIGRATION REGULATION

585

Now, if the specific choice were between x (inclusive immigration with equality) and z (no immigration), a Minimal Liberal perspective valuing individual choice would arguably follow Alice's preference and take z as superior to x (if Alice does not wish to include Roberto in her community, society should support this, regardless of Roberto's desire).2 This demonstrates that Minimal Liberal analysis is not prejudicially cosmopolitan-leaving the fundamental choice regarding inclusion of immigrants in the incumbent community to Alice. The second choice is between y (immigration with exclusion and discrimination) and z (no immigration). A Minimal Liberal framework would prefer y, following Roberto (if Roberto is so intent on migrating that he is willing to accept discrimination, he should be allowed to do so; even though Alice would prefer life without Roberto, under y she is not forced to include him in her community with equal rights-her original preference to avoid including Roberto in her community is respected) who must in any case be left with at least one meaningful choice. The outcome is therefore y-immigration permitting "behind-the-border" discrimination. Notably, this result is not path-dependent. Even if Roberto's y/z choice is sequenced before Alice's x/z choice, the result is still y. Hence, in a Minimal Liberal approach, a ban on migration (z) might be specifically socially preferred to migration with full rights (x) (as a binary choice made by Alice), but-counterintuitively-immigration with exclusion and discrimination (y) would be the overall socially preferred alternative. This exemplifies Sen's Liberal Paradox because in their individual ordering of preferences, both Alice and Roberto prefer inclusive immigration with equality over non-inclusive, discriminatory immigration (the Paretian outcome), and yet the Minimal Liberal outcome is of immigration with discrimination.24 This analysis presents a coherent defense of discrimination in migration regulation. People in immigration-destination 23. One might wonder why Alice's preference to exclude Roberto takes precedence over Roberto's preference for migration. This is because x would require Alice to accept Roberto in her community, and so impinge on Alice's liberty, just as Sen's "prude" would be required to read LadT Chatterly's Lover against his will.

24. See Sen, supra note 15, at 155.

586

FORDHAM INTERNATIONAL LAWJOURNAL

[Vol. 36:575

states might prefer no migration to egalitarian migration, but migrants prefer discriminatory migration over no migration. When immigration occurs-that is, when receiving states allow it to happen, explicitly or implicitly-discriminatory migration conditions prevail: This result appears similar to Howard F. Chang's exposition of an "immigration paradox."25 According to Chang, from a selfinterested welfare perspective, citizens in migration-receiving countries prefer temporary immigration (non-inclusive and likely discriminatory immigration policy), but this violates their liberal-egalitarian ideals. Moreover, the costs of egalitarian inclusion would be so high that ultimately, restrictive immigration policies are adopted to the detriment of would-be migrants. Chang's analysis is, in fact, quite different from the Minimal Liberal analysis suggested above, most obviously because it focuses only on incumbent citizens' preferences and does not factor immigrants' preferences into the analysis. Chang's analysis is helpful in understanding the _formation of Alice's preferences-how the communitarian-egalitarian tension leads to a first preference of no immigration-but it does not analyze

the general social choice that would refer to both Alice and Roberto's preferences, as a form of public reasoning. In this sense, the present analysis begins where Chang's ends. The analysis also differs in its conclusions. Chang's proposed solution to the communitarian-egalitarians' predicament is to urge them to adopt cosmopolitanism, allowing open and egalitarian immigration. This is of course a difficult, if not question-begging, solution, simply invalidating the communitarian end of the spectrum; but as Chang acknowledges, it is in any case politically unlikely. Chang's conclusion is that guest worker programs are the best that cosmopolitans, communitarian-egalitarians, and indeed prospective immigrants, can hope for. By contrast, in the present analysis, discriminatory terms of migration are not normatively second best, even if they might be second best from the individual perspectives of both incumbents 25. Howard F. Chang, The Inunigration Paradox: PovertY, Distributive justice, and LiberalEgalitarnsm,52 DEPAL L. RFV. 759, 764 (2003).

20131

DISCRIMINATION IN MIGRATION REGULATION

and migrants. Neither would discrimination regulation be a compromise between different simply the proper outcome of a social choice individual choices. Vithin this framework, migration is therefore morally vindicated.

587

in migration positions. It is rule based on discriminatory

III. AB USIVE DISCRIMINATION Does this mean that all discrimination in migration regulation is justified in the Minimal Liberal framework? After all, a key element of the analysis is that Roberto prefers being treated unequally as an immigrant over remaining in his home country. Perhaps conditions in Roberto's country are so bad that even abusive discrimination in Alice's country is acceptable to him. One can imagine that impoverished immigrants are often willing to suffer abusive treatment in immigration-receiving countries, and even embark to the new country with the foreknowledge that they will be degraded. Does the analysis therefore justify abusive discrimination against immigrants? One could argue that if Roberto is willing to forfeit his rightsincluding not only freedom from discrimination, but also freedom from humiliation-he should be allowed to do so. An analysis of the "optimal contract" between immigrationreceiving states and immigrants could lead to a similar conclusion. Cox and Posner justify discrimination between types of migrants on a contractual basis, arguing that immigrants who make lower "country-specific investments" (e.g., temporary immigrants, in comparison with permanent immigrants) value their immigrant rights less.26 States are therefore justified in minimizing the costs of endowing such immigrants with rights. However, under this logic, temporary migrants who see no value in any investment in their host society value their rights least, willing to forfeit even basic rights (such as the freedom from abusive discrimination). To avoid this outcome, Cox and Posner simply assume that immigrants will not migrate to countries in which their basic rights are abused.27 Sadly, this is not the case,

26. Adam B. Cox & Eric A. Posner, The Rights of Migrants: An Optimal Contract Framework, 84 N.Y.U. L. REV. 1403, 1418-22 (2009). 27. Id. at 1407-08.

588

FORDHAM INTERNATIONAL LAWJOURNAL

[Vol. 36:575

as the multitude of cases of the abuse of immigrants' basic rights attests. In contrast, the present Minimal Liberal framework of analysis takes account of immigrants' potential willingness to be subjected to abusive discrimination, as well as of host communities' interests in avoiding it. In other words, both Roberto's desperation and Alice's egalitarian preferences count. The Minimal Liberal framework can justify discrimination in migration regulation, but not discrimination that concurrently violates migrants' basic substantive human rights (including, for example, core labor rights). Let me now clarify which forms of discrimination in migration regulation I am defending, and which forms of discrimination I am not. Discrimination against immigrants, whether through public or private actions or both, too often takes the forms of oppression, harassment, victimization, hostility, and exploitation that violate basic human rights.29 By no means do I purport to defend these types of discrimination. Having said that, it is possible to distinguish between types and degrees of discrimination, and one should not derive from the immorality of some abhorrent practices a general rule on all discrimination in migration regulation. This limited approach is defensible on a few interrelated grounds. First, egregious problems such as these are morally problematic in themselves, regardless of the role discrimination plays in them. Wage theft and sexual abuse are surely rights violations that cannot be tolerated whether they are discriminatory against migrants, or not. This is true even of racial discrimination, which persists "regardless of immigration

28. See INT'L LABOUR ORG., Declaration on Fundamental Principles and Rights at Work, June 18, 1998, 37 I.L.M. 1233 (1998) (recognizing that the elimination of discrimination in the workplace is itself a core labor right). However, this would refer to discrimination not derived fron an immigrants status, but fromn her race. colour, sex, etc. See id. 29. For one survey, see generally S. POVLRTY L. CTR., UNDLR SILGE: LIFL FOR LowINCOML L.ATINOS IN THL SOUTH (2009), IhIp://www.spicectr.org/sitcs/dcfaul/files/

downloads/UnderSiege.pdf.

20131

DISCRIMINATION IN MIGRATION REGULATION

589

status."" Abusive discrimination against immigrants is usually contingent upon the existence of other substantive wrongs. Second, there is also legal ground to distinguish between different forms of discrimination. The international law of economic, social, and cultural rights recognizes that nondiscrimination, as an obligation, is contingent on the existence of substantive rights and should not be understood as "an autonomous right to be free from discrimination.", 1 In civil and political rights, nondiscrimination is similarly guaranteed with respect to substantive protected rights,2 as well as in the more specific principle of equality "before the law," requiring only "in this respect" "equal and effective protection" against discrimination.a These might be difficult distinctions to draw, but they permit us to distinguish between discrimination in migration regulation that relates to self-standing violations of rights and discrimination that does not. Third, discrimination can exist even when there is no violation of substantive rights, within the policy space that lies above autonomous human rights. Such discrimination should be assessed on its own merits-such as aims, effects and 30. Id. at 32. But s James A. Goldston, Holes in the Rights Framework. Racial Disc rmnatin, Citizenship, and the Rights of Non-Citizens, 20 ETHICS & INT'l XFF. 321, 322-35 (2006) (demonstrating how citizenship and immigration laws are used for the abuse of racial and ethnic minorities). 31. See M. MAGDALENA

SEPTLVEDA, THE NATURE OF OBLIGATIONS UNDER THE

COVLNANT ON ECONOIc, SOCIAL AND CULTURAL RiGHTS 380 (2003). The United Nations Committee on Economic, Social and Cultural Rights has opined, that "[t]he Covenant rights apply to everyone including non-nationals. such as refugees, asylumseekers, stateless persons, migrant workers and victims of international trafficking, regardless of legal status and documentation." Comm. on Econ., Social and Cultural Rights, General Comment No. 20: Non-Discrimination in Economic, Social and Cultural Rights, 42nd Sess., 1 30, U.N. Doc. E/C.12/GC/20 (July 2, 2009) [hereinafter General Comment No. 20]. It did not go so far, however, to say that these rights must be applied equally to Citizens and non-citizens. Id. 32. See ICCPR, supra note 11; see also UN Human Rights Comm'n. General Comment No. 18: Non-discrimination (1989), in Compilation of General Comments and General Recommendations Adopted by Human Rights Treaty bodies, U.N. Doc. HRI/GEN/1/Rev.7 (May 12, 2004) [hereinafter General Comment 18] (noting that Article 2 of the International Covenant on Civil and Political Rights ("ICCPR") limits the scope of the rights to be protected against discrimination to those provided for in the covenant). 33. See ICCPR, supra note 11, art. 26. The Human Rights Committee ("HRC") has opined that the "enjoyment of rights and freedoms on an equal footing, however, does not mean identical treaLntc[ in every instance." General Comment No. 18, supra note

32, [ 8.

590

FORDHAM INTERNATIONAL LAWJOURNAL

[Vol. 36:575

proportionality. 4 For example, minimum wage rules establish a substantive standard of adequate remuneration but do not guarantee equality, and different occupations and qualities command different incomes. Gender-based pay discrimination can be pursued in salary tiers well above minimum wage, but in this case "equal pay for equal work" laws intervene, establishing nondiscrimination as a principle that detaches itself from a substantive norm. In short, not all discrimination was created equal. Fourth, focusing on the worst treatment of migrants (which is usually also discriminatory, at least in orderly legal environments that otherwise prevent such objectionable conduct against incumbents) hardens and "absolutizes" the normative approach to discrimination, obscuring the relative nature of discriminatory wrongs. Being harassed by police or denied access to justice simply because one is an immigrant is surely different from receiving lower wages, paying a higher tax, or enjoying less comprehensive healthcare insurance. In order to assess discrimination in immigration regulation, we must treat its worse manifestations-what I have referred to as "abusive" discrimination-separately from other forms of discrimination, not because they are somehow unimportant, but rather because they are the easy cases. Let's return to Alice and Roberto, who now have four alternatives to choose from. As before, egalitarian immigration (x) or no immigration (z) can be chosen, but now also two distinct types of potential discrimination after immigrant admission: abusive discrimination that concurrently violates Roberto's substantive human rights (e.g., he will be denied effective legal protection, or will be employed under conditions tantamount to slavery)-marked as y;; and discrimination that does not (e.g., Roberto's employment will be restricted to certain jobs, or he will be paid less than Alice for the same work, or receive fewer social benefits)-marked as y2. As before, Alice initially prefers no immigration, but if immigrants are admitted she would ideally prefer an egalitarian standing; between the alternative scenarios of discrimination, she strongly prefers that Roberto's human rights not be violated. 34. See General Comment No. 20, supra note 31, [ 13.

20131

DISCRIMINATION IN MIGRATION REGULATION

591

Alice's preferences are ordered (best-first to worst-last) z, x, Y2, YI. Roberto, however, is so intent on migration that he is willing to forfeit some of his autonomous human rights and face abuse. Non-abusive discrimination would be far better; ideally, he would wish to be treated equally and enter Alice's community. Hence, Roberto's preferences, in diminishing order are x, y2vy. z. Importantly, this formulation explicitly takes into account Roberto's (reluctant) propensity to enter into a situation in which he will be discriminated against in ways that violate his substantive human rights, by ordering y, over z. Notably, both Alice and Roberto prefer egalitarian immigration to discriminatory migration policy, and that to abusive discrimination. They part ways on the question of whether no immigration is the better alternative or not. Let us again apply the Minimal Liberal framework. As before, if the only choice is between z and x (no immigration and egalitarian immigration, respectively), Alice's choice should prevail, and on the same grounds. If the only choice is between z and y 1 -no immigration and abusive discrimination-again, Alice's choice should prevail, on the grounds that she should not be forced to tolerate the committal of human rights abuses in her own country because human rights are a central tenet of her community values. This leaves the third and final possible choice between z and y2, 3 which should be left to Roberto (he must still determine at least one choice), who clearly prefers the latter. Consequently, under a Minimal Liberal rule of social choice, no immigration is preferred to both egalitarian and abusively discriminatory migration, but non-abusive discrimination in migration policy is superior to no immigration. The outcome is y2, but not y;: discrimination is tolerated, but not any discrimination. Discrimination that concurrently violates substantive human rights is unjustified. The Liberal Paradox and the Paretian bias towards cosmopolitan outcomes holds: on a weak Paretian basis, the outcome would have been x (egalitarian immigration), because both Alice and Roberto prefer it to T; and T2. Furthermore, this outcome is not path-dependent. Had the choice z/y2 been left to 35. Choices xl,

x/v2, and yl/y2 are excluded: in each of them Alice and

Roberto have shared preferences, and would not satisfy the conditions of the Minimal

Liberal franework.

592

FORDHAM INTERNATIONAL LAWJOURNAL

[Vol. 36:575

Alice, leaving the choice z/y; to Roberto, the social choice outcome

would

have

been

yi,

resulting

in

an

apparent

justification of abusive and discriminatory treatment of migrants. This is false, because under Minimal Liberal conditions, leaving the choice between z and 2 (no immigration vs. non-abusive discrimination) to Alice would not be justified in the same way that allocating the choice between z and T, to her would be. While Alice has a direct interest in preventing members of her community from abusing human rights, she has no direct moral interest in preventing non-abusive discrimination against immigrants that are not admitted to her community. Positing otherwise would require Alice to extend her preference for nondiscrimination not only to immigrants to her country but to people anywhere-and that would be a cosmopolitan view that has not been ascribed to her. Two variants of this analysis are worth considering. One suggests that I have been too careless in formulating Roberto's preferences; another suggests that I have been too charitable with Alice. In the first variant, Roberto is not keen enough on migration so as to accept abuse. His preference ordering then changes from x, y, y, z to (the more sensible, perhaps) x, y Z, ', while Alice's preferences stay the same. The outcome in this case does not change, and is justified a fortiori: both Alice and Roberto now prefer z to Ti. Alice prefers z to x, but Roberto prefers y2 to z. This is essentially the same as the general case of a three-way choice between equal treatment, discrimination and no immigration, discussed in Part III above. The second variant exposes Alice as a merely hypocritical egalitarian. Vhile she prioritizes the alternative of no immigration (z), for selfish reasons her second-best choice is not egalitarian immigration (x), but non-abusive discriminatory migration regulation (y2), which is less costly and threatening to her community (indeed, she might even prefer abusive discrimination to egalitarian immigration, because it might deter immigration altogether). Her preferences would then be Z, Y2 , y1 (or in the worst case, z, y2, yl, x). However, taking both Alice's and Roberto's preferences into account, the social preference remains-discrimination in migration regulation, without abuses of substantive human rights.

20131

DISCRIMINATION IN MIGRATION REGULATION

593

IV. MIGRATION-ENCOURAGING AND MIGRATIONDETERRING DISCRIMINA TION The question now arises, what is discrimination in migration policy good for? Why undertake a moral justification of discrimination? The answer is that discrimination (and to belabor the point of the previous section, I mean only non-abusive discrimination) can be an effective tool for regulating migration. It can encourage and increase labor migration beneficial to migrants, their families and their home countries. Yet, conversely, some forms of discrimination in migration regulation can deter immigration. It is easiest to clarify this difference-between migrationencouraging and migration-deterring discrimination-through concrete examples. There are many illustrations, but here I will refer to the Israeli labor migration case of Bukhris,16 and a series of cases in the European Court ofJustice ("ECJ") relating to the employment conditions of citizens of one Member State working in another Member State. I will subject the two types of discrimination to the Minimal Liberal framework, examining whether either of them should be considered as justifiable. The Bukhris petition to the Israeli Supreme Court had all the trappings of a landmark immigration case. National legislation relating to migrant workers was being challenged on the basis of international law. The general claim was discrimination. The legislation established a tax calculated as eight percent of a migrant worker's overall income (the "Tax"), that was not applied to local workers. The Tax, the petitioners argued, discriminated against migrant workers, violating international norms of labor rights, double taxation, and international trade. Yet the Court rejected all of the petitioners' claims, on technical legal bases. Ostensibly, the evil of harmful discrimination against migrants had been upheld by the Court. This might have been the case, but for one fact omitted from this narrative: the petitioners were not at all migrant workers, and they had no real stake in the petition. 7 Rather, 36. H(

2587/04 Bukhris v. Hadera Tax Assessor, unpublished [2005] (Isr.).

37. Had the Bukhris petition been sustained, employers would have been exempt

from the eight percent tax on migrant worker's overall income ("Tax") not previously

594

FORDHAM INTERNATIONAL LAWJOURNAL

[Vol. 36:575

Bukhris himself was an Israeli floriculturist who had employed five Thai migrant workers. The second, more financially significant petitioner in the case was a corporation engaged in "human resources," at the time employing 373 Chinese citizens and 238 Romanians. The petitioners were represented by a leading Israeli tax law specialist, not by migrant rights advocates. Although the petition seemed to focus on discriminatory treatment of migrants, Bukhris was really about the taxation of their employers' profits. Bukhris therefore serves as a cautionary tale: discrimination in migration regulation is not always what it seems. We see, for example, that nondiscrimination arguments can be used not for the sake of human dignity, or the enhancement of migrants' rights, or even the preservation of communitarian values, but for promoting the self-interest of non-immigrants. And we see that discrimination can stem not from indifference or maliciousness but from economic policy, acting as a tool for regulating migration flows; in its response to the Bukhris petition, the Israeli government was entirely frank in stating that such was the goal of the discriminatory taxation system." The Tax, discriminatory warts and all, was simply intended to discourage Israeli employers from seeking foreign migrant workers, by making their labor more expensive. In Bukhris we see an immigration-deterring form of discrimination, at least by design, if not in its effect. The opposite type of discrimination-immigrationencouraging discrimination-is discernible in a series of ECJ cases in which the common quandary (simplified for present purposes) was the extent to which workers from one Member State, working in another Member State, could be granted terms and conditions of employment that were lower than those guaranteed in their host state. The questions have usually arisen when a Member State or labor union insisted that equal terms of employment be granted to foreign workers, and this was considered by the employers of the foreign workers or the paid and would have been able to request reinibursement of Taxes alircady paid. Not a penny would have gone to the migrants. 38. HCJ 2587/04 Bukhris. at 3 ("The Respondents accept that the Employers' Tax creates a negative incenLtive for the employment of foreign workers, and indeed that is

its goal.") (author's translation).

20131

DISCRIMINATION IN MIGRATION REGULATION

595

European Commission to constitute a restriction on the freedom of establishment or the freedom of movement of services. The relevant European legislation, the Posted Workers Directive (the "Directive"),39 actually mandates nondiscrimination of employees, but that only as "a nucleus of mandatory rules for minimum protection to be observed in the host country by employers who post workers to perform temporary work in the territory of a Member State where the services are provided."11 Under the Directive, employees posted to another Member State are guaranteed the terms and conditions of employment in the host state, if laid out by law, regulation or administrative provision or by collective agreements declared to be "universally applicable," with respect to a list of employment terms including maximum work periods and minimum rest, minimum paid annual holidays, and minimum rates of pay. Issues have reached the ECJ in disputes related to the application of host state minimum requirements that are not explicitly covered by the Directive. Thus, in Viking4 ' and Laval, 42 industrial action taken by labor unions in the host state (Finland and Sweden, respectively), aimed at compelling employers to apply the terms of host state collective agreements to foreign "posted" workers4 3 was found to possibly interfere with the freedom of establishment and the freedom of movement of services. 44 In Riffert, a legal requirement in the German Land of 39. See generally Directive 96/71/EC of the European Parliament and of the Council Concerning the Posting of Workers in the Framework of the Provision of Services, 1997 O.J. L 18/1 [hereinafter Posted Workers Directive]. 40. Id. preamb., 1 13 (emnphasis added). 41. See Int'l Transp. Workers' Fed'n v. Viking Line ABP (Viking), Case C-438/05, [2007] E.C.R. I-10779. 42. Laval un Partneri Ltd. v. Svenska Byggnadsarbetareforbundet (Laval), Case C 341/05, [2007] E.C.R. 1-11767. For commentary, see Uladizlaui Belavusaul, The Case of Laval in the Context of the Post-EnlargementEC Law Development, 9 GER. L.J. 2279 (2008); A.C.L. Davies, One Step Forward, Two Steps Back? The Viking and Laval Cases in the ECJ 37 INDUS. L.J. 126 (2008), and Ronnie Eklund, A Swedish Perspective on Laval, 29 CoMP. LAB. I. & POL'YJ. 551 (2008). 43. In Viking, the foreign clement was introduced by a Finnish employer's intention to re-flag its ships to Estonia, where lower terms of employment apply. See Viking, [2007] E.C.R. 1-10779. In Laval, the employer posted Latvian workers in Sweden, applying Latvian terins to their cmployment. See Laval, [2007] E.C.R. 1-11767, 2. 44. See generally Norbert Reich. Free Movement v. Social Rights in an Enlarged Union-The Laval and Viking Cases Before the E, 9 GLR. L. J. 125 (2008). For a European Constitutionalist critique, see Christian joerges & Florian RdI, Informal

596

FORDHAM INTERNATIONAL LAWJOURNAL

[Vol. 36:575

Niedersachsen whereby service suppliers to the government apply the terms of a local collective agreement to Polish workers, was found by the ECJ to violate the Directive.45 And in Commission v. Luxembourg the ECJ upheld the contentions against Luxembourg's legislation that effectively expanded the terms of the Directive to a broader set of labor regulations and collective agreements. 6 What is important for present purposes, however, is the recognition that arises from these cases, that requirements of non-discriminatory treatment in labor migration regulation may impede the freedom of movement. In other words, the discrimination that was being upheld was migrationencouragmg--in contradistinction to the type of discrimination upheld in Bukhris, which as we have seen, was migrationdeterring. Let us now return to the Minimal Liberal framework of analysis. Given the choice between migration-deterring and migration-encouraging discrimination in migration regulation, which would be justified by the individual preferences of Alice and Roberto? This depends on the alignment of individual preferences. As before, let x signify egalitarian migration and z signify no immigration. The option of discrimination in migration policy, y, is now split into two distinct alternatives. The

first, y*, indicates migration-encouraging discrimination; y** indicates migration-deterring discrimination. What are Alice and Roberto's (representative) preferences? Roberto is the simpler case: as before he prefers egalitarian immigration, with a second choice of discriminatory migration policy and a worst case of no immigration. In the choice between encouraging and deterring discrimination, in both types of discrimination Roberto will be treated less favorably than incumbents, but the immigration-encouraging type increases his opportunities, and so he prefers it. His decreasing order of preferences would be x, y*, y**, z.

Politics, Formalised Law and the Social Deficit of European Integration: Reflections After the and Laval, 15 EUR. L.J. 1 (2009). 45. See generally Riffert v. Land Niedersachsen, Case C-346/06. [2008] E.C.R. 11989. 46. Conm'n v. Luxembourg, Case C-319/06, [2008] E.C.R. 1-4323.

Judgments ofthe ECJin Viking

20131

DISCRIMINATION IN MIGRATION REGULATION

597

Alice's preferences are interesting in that the distinction between types of discrimination changes her framing of egalitarian migration. She still prefers no immigration, but her second choice is now (non-abusive) migration-deterring discrimination (the logic being that it is the second-best way to prevent immigration). Egalitarian immigration is therefore Alice's third-best alternative, preferred only to immigrationencouraging discrimination which would both actively increase immigration and subject it to discriminatory terms-the worst case from Alice's viewpoint. Alice therefore orders her preferences z, y**, x, y*. We see that the commonalities between Alice and Roberto's preferences have narrowed down to one: they both prefer egalitarian migration (x) to migration-encouraging discrimination (y*). This leaves five choices to be aggregated. If the only choice is between egalitarian immigration (x) and no immigration (z), the choice should still be Alice's-no immigration-for the same reasons as before. Between no immigration (z) and immigration-deterring discrimination (y**),47 however, Roberto's choice should prevail; immigrationdeterring discrimination makes few demands of admission to Alice's community and indeed works towards the goal of less immigration that Alice herself prefers. Between immigrationdeterring (y**) and egalitarian immigration (x), Alice's preference again prevails, because the more inclusive option would require her to introduce Roberto into her community. The choice between immigration-deterring discrimination (y**) and immigration-encouraging discrimination (y*) should follow Roberto: neither forms of discrimination require Alice to accept Roberto into her community, although both entail a compromise of her egalitarian ideals. Finally, in the remaining choice between no immigration (z) and immigrationencouraging discrimination (y*), Roberto's choice should be preferred, again because it will not impinge upon Alice's sense of community. The Minimal Liberal outcome would then be y*, putting immigration-encouraging discrimination as the ultimate 47. At this point, it is important to note that the outcome of this analysis is not That is, one could now first examine the hoice between no

path-dependent. inunigration outcome.

and

unnigration-encouraging

discrimination,

without

altering

the

598

FORDHAM INTERNATIONAL LAWJOURNAL

[Vol. 36:575

social choice (maintaining the Liberal Paradox, because in weak Pareto terms, egalitarian migration would still have been preferred). Immigration-encouraging discrimination is therefore fully justifiable under the Minimal Liberal framework of social choice encompassing preferences in an egalitarian-communitarian immigration-receiving state and in the state of emigration. Does this mean that immigration-deterring discrimination is unjustifiable? So long as the option of migration-encouraging discrimination exists, and as long as it is Roberto's preference, it will be superior to migration-deterrence. However, in contrast to abusive discrimination that violates human rights of migrants, immigration-deterring discrimination in migration policy can be independently justified. One only need to consider the preferences of Alice and Roberto between x, y, and z** (compared with x, y; and z) to see that this is the case. CONCLUSION This analysis shows that (some) discriminatory treatment of immigrants can be morally justified under a Minimal Liberal rule of social choice and public reasoning. The same framework shows that violations of substantive human rights of immigrants should not be tolerated, permissible discrimination notwithstanding. Finally, the Minimal Liberal analysis shows that discrimination that encourages immigration is morally superior to discrimination that deters it, although the latter is still morally justifiable in this framework. The public policy applications of this analysis are manifold, but in general, the lesson is that inflexible demands or advocacy for rules prohibiting all discriminatory treatment of immigrants, whether in the context of human rights, labor rights or economic liberalization, should be treated cautiously. There is no question that nondiscrimination rules can impede migration which is beneficial to migrants, their families, and their home countries, but this is sometimes disregarded as an instrumental claim that is trumped by deontological moral approaches. However, in this article I have attempted to demonstrate that some discrimination can be justified not only instrumentally but also morally and analytically. It cannot be taken for granted that discrimination is always wrong. In policy terms, discrimination in

20131

DISCRIINA TION IN MIGRATION REGULATION

599

immigration policy should not be prohibited en banc as a negative occurrence. Because at least some "behind-the-border" discrimination against immigration can be morally justified, different applications of discriminatory treatment should be analyzed separately. Ultimately, this conclusion significantly relaxes the moral environment in which the immigration debate exists. Much of the philosophical debate is a dialogue of the deaf, between cosmopolitans and various forms of communitarianism. This divide is reflected in the egalitarian-communitarian tension that the present analysis aims to alleviate, through the use of a rule of public reasoning that is not prejudicial towards any particular 'transcendental' moral attitude. Indeed, the approach pursued in this article suggests a moral basis for discussion that (unlike Paretian analysis) is largely unprejudiced towards either end of the moral spectrum.