ARENA Working Papers
WP 01/8


Federal Inequality Among Equals

A Contractualist Defence


Andreas F�llesdal




Federal political orders often exhibit a conflict between the ideals of equality and political autonomy, since individuals in different sub-units often enjoy systematically different standards of living conditions. While federal arrangements may be theoretically attractive to avoid despotism, such federal inequality would appear to conflict with the principles of egalitarian cosmopolitans. The paper argues that individuals' interest in equal shares of income and wealth may be legitimately weighed against their interest in political control enjoyed by their sub-unit, as long as the inequalities do not engender misery, domination or unfair procedures. The reasons for sub-unit autonomy include reducing the risk of domination, increasing the responsiveness to local preferences, and reducing the burdens of decision-making. These arguments also suggest that states may not always be the appropriate sub-units in legitimate federal orders.

1 Introductory remarks

Federal political orders often exhibit a conflict between the ideals of equality and political autonomy. Individuals in different sub-units often enjoy systematically different standards of living conditions, partly as result of the political powers enjoyed by these sub-units. This paper concerns the legitimacy of such federal inequality.

Federal arrangements may seem an attractive possibility for cosmopolitan political theorists criticized for requiring a world state. A typical response is that moral or normative cosmopolitanism -- equal respect owed every affected person -- does not entail institutional or legal cosmopolitanism in the form of a risky and unstable unitary world state ( Beitz 1994 ; Pogge 1992a a). Indeed, at least since Immanuel Kant federal arrangements have been favored as less prone to despotism, yet compatible with cosmopolitan principles of justice ( {Kant 1970 #11690} ). Still, much recent political philosophy has focussed on principles for unitary states with a central sovereign (symptomatically, cf. {Rawls 1993 #9780} , xxii).

The federations of concern here are non-unitary political orders. The central and multiple regional loci of government enjoy final legislative or executive authority with regards to some functions, often by way of constitutionally enumerated powers ( Riker 1993 , 509).

While federal arrangements may be appealed to in theory, egalitarian cosmopolitans may not permit them in practice. Strictly egalitarian distributive requirements, for instance in the form of a global difference principle, would seem to require highly centralized legislative and executive powers to regulate all interaction with distributive implications. Indeed, one of James Madison's arguments for a federation of states was that it would prevent "an equal division of property, or for any improper or wicked project" (Madison 1961a). Hence we must ask whether cosmopolitan political theories can allow political autonomy of sub-units to such an extent as to allow substantial economic and social inequality among citizens of different sub-units. The question of concern to us is not whether some inegalitarian federal arrangements might be a second-best improvement within reach. For instance, some political autonomy over certain policy sectors is often accorded as part of political bargains among sovereign states in forming a federation. The issue is instead whether federal arrangements with substantial inequality may be normatively legitimate.

This question is of practical relevance for work on global justice, given the recent normative defenses of global egalitarianism in the form of a global difference principle (Beitz 1979, 152; Barry 1989, 187-9; Scanlon 1974, 202-203). Such standards may appear so counterintuitive to the voting populations of rich states that all appeals to global justice are dismissed as a slippery slope best avoided. The fear of what ideal justice requires may prevent smaller steps to alleviate the abysmal conditions of our non-ideal world, be it in the form of transfers of goods and services or support for domestic and international institutional re-design. If principles of justice for federations impose less stringent demands, this reaction may be avoided, and powerful rich states may legitimately opt for federal responses [1] .

Similarly, fears of such egalitarian conclusions may keep richer states from institutionalized cooperation with poor states. Thus the club of rich states in the European Union is currently committed to regional funds and agricultural subsidies, thereby "demonstrating consistency and solidarity ... between the Member States and between their peoples" (Treaty on European Union 1997, art. 1). These criteria and policies are subject to intense reformulation in preparation of the envisioned ascension of poor applicant states. An unmodified commitment to equalize living standards would entail politically unacceptable costs, since the applicant states GNP/capita is only 20-11 % of the EU average. Admission of the applicant states and their citizens as equals would therefore create tensions unless economic inequality can be defended within the European political order.

This paper explores how the claims individuals may have against institutions may legitimately depend on whether these are unitary or federal institutions. I shall defend the view that federal arrangements can legitimately engender somewhat unequal shares of benefits and burdens among citizens of different sub-units. Individuals' interest in equal shares of income and wealth may be weighed against their interest in political control enjoyed by their sub-unit.

The remaining part of this introduction provides a brief sketch of some elements of liberal contractualism. Section 2 considers some answers to the question "Why Equality?" Section 3 explores, only to dismiss, a defense for inequality within federations on the basis that these arguments for equality fail to apply. Section 4 considers liberal contractualist reasons for sub-unit autonomy. Section 5 identifies three reasons why even the sub-unit poor may benefit from such sub-unit autonomy even at some economic cost.I argue that such reasons permit some deviations from the egalitarianism defended in section 2. The discussion also highlights the precarious role of states as the sub-units of normatively defensible federal political orders.

First, some elements of liberal contractualism used as the normative bases for this exploration of federal inequality.

Many normative political theories have a normative egalitarian premise that all affected parties are worthy of equal concern and respect. This commitment to equal respect is cosmopolitan, in the sense of being universal: those on the inside of state and sub-unit borders have fundamentally the same moral standing as outsiders.

This is taken to mean that every affected individual's interests, suitably delineated, must be secured and furthered by the social institutions as a whole ( Dworkin 1978). Contractualist theories hone this vague commitment by invoking the notion of hypothetical consent (O'Neill 1989, Habermas 1991 , 235). The principles of legitimacy we should hold institutions to are those that the affected persons would unanimously consent to under conditions that secure and recognize their status as appropriately free and equal, thus manifesting "our respect for the reasonableness of others" (Macedo 1990). Such principles allow each of us to "justify one's actions to others on grounds they could not reasonably reject" (Scanlon 1982, 116; cf. Barry 1989, 8). The specific conditions, and the significance of consent, vary among these theories in ways that need not concern us here.

Among the interests of individuals that can command general agreement for purposes of such arguments about legitimate social orders are the satisfaction of basic needs and all-purpose means for pursuing one's conception of the good life.Furthermore, individuals must be acknowledged to have an interest in procedural control over the social institutions that shape values, goals, options and expectations.

The topic of concern here is principles for assessing the rules of institutions that have a pervasive impact on us. Institutions are rule-governed practices established and maintained through the threat of force. They provide the backdrop for the distribution of important goods, powers, burdens and obligations necessary for a variety of interests. Institutions also shape our expectations and values, and changes in institutions challenge our ability to maintain coherence and continuity in our lives. These considerations give us good reasons to claim a share in the political authority to make and change these rules.

The contractualist approach leads us to search for principles for institutions, against which no reasonable objections can be made. Three features are relevant for the following arguments. Such principles of justice, or particular institutions, are not generated by the process of checking whether equal respect is secured. Instead the procedure of hypothetical consent provides checks drawn from interests at stake, but does not aim for a deduction of principles or institutions. Secondly, the process allows for several alternative principles. Thus the set of principles may be underdetermined, in the sense that alternative principles may all be unobjectionable. Thirdly, a set of principles may allow a variety of sets of institutional arrangements, each of which satisfies the distributive requirements of liberal contractualism.

This tradition might thus allow that several different institutional arrangements are just. The unity provided by such a theory is hence not one of deduction - blueprints of institutional design are not on the agenda.Instead, the normative assessment of institutions show that they are consistent with, and can be regarded as an expression of, a view of individuals as enjoying equal respect. For instance, the role of the state is underdetermined. It remains an open question whether universal basic income at some level is appropriate, or whether the state should be less directly involved in the form of transfers [2] . This is partly a matter of the long-term effects of moral hazard and incentive effects on the recipients, but also legitimately a question of historical fit and decision by democratic fiat. Similarly, consider the variety of institutional arrangements in European states concerning such important topics as old-age pensions and health insurance. Careful assessment of their consequences may show that reasonable objections can be made against some such arrangements, but it is by no means clear that only one of them is best by standards of distributive justice. One important task in the EU is thus to reform roughly just domestic institutions as required by increased interdependence, while respecting individuals legitimate expectations.

2 Why Equality?

Some liberal theories appear to take for granted that equal respect for all entails equal shares -- be it of goods, opportunities, resources, or initially un-owned things (Cohen 1989;Dworkin 1981a a, Dworkin 1981bb,Dworkin 1987; Sen 1980; Steiner 1994). Others, such as Rawls' theory Justice as Fairness, give a similar impression, since the principles famously require equal shares of economic and social goods, except insofar as inequalities benefit all Rawls 1971.

I here explore the room contractualism allows for inequalities -- i.e. when substantive inequality survive reasonable objections. Whether substantive inequality constitutes a violation of the commitment to normative equality, and hence gives rise to justifiable feelings of inferiority and second-rate citizenship, can best be determined by considering normative arguments for equal treatment. Such analysis of arguments for equality is particularly important when our intuitions about equality appear to conflict with other intuitions, and where an intuitive "weighing" of these intuitions is controversial or unsatisfactory. The task in this section is precisely to determine what reasons can be offered for equal shares of benefits. Four grounds for lamenting inequalities can be identified on the basis of the interests at stake. [3]

a) Prevent Misery

If we seek to avoid reasonable objections to normative principles, surely acceptable institutions must engender and distribute benefits so as to meet the basic vital needs of all, to secure their survival. Human rights can be interpreted and defended as such conditions on domestic and international regimes so as to secure the satisfaction of such needs (Follesdal 1991). The current world order fails dismally on this point: those with nothing to sell in markets cannot buy food, and large economic inequalities can even prevent wage earners from acquiring sufficient food. Differences in relative political power perpetuate abysmal prenatal health care for the poor, and international regimes fail to include obligations of international support as final resorts, when domestic resources run out or when government powers are grossly abused

But this argument from basic needs and human rights does not require equality of condition (Raz 1986; Miller 1995, 191). Rather, this consideration only prohibits drastic inequality regarding certain specific goods, insofar as these inequalities engender misery.

b) Prevent Domination

A social order is objectionable if some individuals can drastically restrict the attractive options of others, prevent deliberation or otherwise leave them at the mercy of the powerful. The reason is that individuals have an interest in maintaining control over the social factors that shape their own lives -- in particular if the alternative is that others wield such control. One important strand of recently resuscitated republicanism has focussed on this interest in avoiding subjection to the arbitrary will of others (Pettit 1997, Skinner 1998). Large inequalities of wealth or income opportunities can prevent the less privileged from exercising control over their lives and subject them to the arbitrary bargaining power of the powerful in various spheres of life.

Again, this argument does not support equal distribution tout court. The prevention of domination prohibits only those inequalities that impact objectionably on the distribution of control over individuals' lives.

c) Ensure fair procedures

A further ground for equality also stems from our interest in controlling the social factors that shape our lives. Many social procedures and mechanisms require for their fairness a roughly equal distribution of procedural input levers. Some adversarial procedures illustrate this: If legal trials are to regularly identify the guilty, competent counsel must represent both parties. Democratic arrangements likewise require a broad dispersal not only of formal political power, but also of education and income and wealth, since relative shares of these levers often matter for the real value of formal political power (Dahl 1985, 55). Similarly, markets provide an efficient allocation of goods relative to a base line -- but only under conditions that include information about alternative buyers and sellers, the likely consequences, inability to create oligarchies, etc. Certain forms of inequalities in information or organizational resources may therefore prevent efficiency.

Note that such arguments primarily apply when we have independent standards for determining what outputs the procedure should generate. Moreover, such considerations may sometimes favor unequal distribution of formal levers such as voting powers. Consider, for example, that citizens of differently populated member states of the EU enjoy different representation in the political bodies. Small populations enjoy more formal power than the principle "one person one vote" would allow -- even after the changes decided in Nice December 2000. Germany, with more than 80 million citizens, has ninety-nine members in the European Parliament and ten votes in the Council of the European Union. So each Member of European Parliament (MEP) from Germany represents more than 820 000 Germans, and each German vote in Council represents more than 8 million Germans.In contrast, each MEP from Ireland represents 240 000 Irish citizens (out of a total of 3.4 million), and each Irish vote in Council represents 1.2 million Irish. The Luxembourgians, in all 400 000, enjoy even more formal political power: each of their 6 MEPs represents 70 000 people, and each of their two votes in Council represents 200 000.

Such outcomes are typical of the bargain between small and large states joining in federal arrangements. Yet this formal inequality has not been perceived as disrespectful, presumably because citizens have assumed that there are good reasons for the skewed power. For instance, arguments can be made that smaller political units need such over-representation to reduce the risk of being permanently outvoted in political decisions (Follesdal 1998a).

d) Equal Shares of products of co-operation.

Individuals may claim equal shares of certain goods when they have contributed equally to the production of the benefits, and when no one can be said to have prior claims to the benefits -- for instance when there is no prior agreement regarding distribution and each party's contribution cannot be determined. [4] When several individuals jointly labor to produce goods, they have equal claims regarding these goods. [5]

We may think of social institutions as the social practices that are maintained by the use of legal powers. Sanctions enforce the public rules and provide public assurance of general compliance, and authoritative interpretations apply the rules to new or difficult cases.

Legal rights in a broad sense are aspects of such social institutions. Within a state, Hohfeldian legal claims, powers, privileges and immunities are constituted by the rules of rights of practices (Hohfeld 1964). These rights include political power, property rights and even income. Take money as an example: it exists only as part of a social practice regulated by rules defining legal tender, where it is common knowledge that all accept the currency in return for goods and services (cf. Coleman 1990 : 119). And an individual who owns something has acquired it according to public rules regulating entitlements. Insofar as she has complied with these rules, the object is clearly hers, and not anybody else's. But her claim of ownership is only true -- and can only be made sense of -- because these rules of ownership are publicly known and generally complied with by those participating in that practice. While her entitlements are hers, the entitlements are entitlements only because others regulate their actions according to public rules.

Legal rights are thus goods that are products of co-operation. These rights are aspects of social institutions, and they only exist insofar as these practices are maintained, which all law-abiding citizens do. Someone enjoys these claimrights and immunities only when the participants in the practice generally recognize and act according to the rules specifying these rights. The general compliance with these rules constitutes these legal rights.

Those who participate in this sense are not those who are producing material goods with the expectation that their expectations of reward will be honored. Rather, those who produce the legal rights are all those participate who regulate their actions according to the rules of the practice, for instance by refraining from taking the material goods identified as the property of others. They thereby cooperate in maintaining the practices defining property.

The argument for equal shares of products applies to this account of the nature of legal rights. Citizens have equal moral claims on how social institutions should regulate the legal distribution of political power, income and other legal rights, where there are no prior claims on such goods. This is not an argument for a policy of providing each person with an equal amount of money per year, but rather a condition on the institutions when operating dynamically, creating and honoring expectations among those choosing employment, and selecting places of work. This is precisely the issue concerning how institutions should affect the distribution of these goods, e.g. through rules of acquisition and transfer that shape individuals preferences and aspirations, and the incentives and expectations of desert. Since these legal rights exist only through the cooperation of all, all participants in social institutions have a prima facie equal moral claim to the legal rights that arise within such institutions [6] .

I submit that this account provides an argument in favor of Rawls' egalitarian principles for Social Primary Goods -- that is, political and civil rights, and equality of opportunity and income and wealth (Rawls 1999).These goods are legal rights, rights-clusters specified by rules governing the practices maintained by citizens.

To summarize, the arguments surveyed only apply to a limited subject, namely the distributive impact of institutions and policies. Moreover, the arguments regarding equality are limited in scope: they do not support equality of quality of life generally. Rather, the arguments are addressed to institutions, and require that they on the one hand prevent certain inequalities, namely those that are instrumental in maintaining misery, domination or skewed procedures. And the institutions should engender equal shares of certain goods, namely income, wealth, educational and employment opportunities for talents, and political and civil rights - what Rawls calls social primary goods. These arguments may leave some room for inequality in the distribution of other goods.

3 Domestic Equality, Global Inequality?

A defense of federal inequality may argue that these arguments for equality fail to hold within federal arrangements, regarded as cooperation among sovereign states. I shall suggest that this strategy is not satisfactory.

On the standard view of the states system, these reasons for equality fail to hold across state borders. Sovereign states enjoy broad formal powers of external sovereignty, in the sense that there are few if any decision-making body above them, and states enjoy immunity from forcible intervention in determining the social institutions which shape citizens' lives.

Several of the arguments sketched above apply to international inequalities of income and wealth. Natural facts and the actions of other powerful agents often restrict the range of alternatives. Market competition, oligarchic trans-national corporations, structural adjustment policies and international human rights norms affect a state's ability to determine its international and domestic policies. However, in a system of sovereign states, international inequalities in income do not fully determine the domestic distribution of control over non-material goods. A claim to international equality of income and wealth is not obvious on the basis of the concern to avoid misery or domination, or to secure fair processes, as long as states enjoy some de facto external sovereignty. Government policies can buffer the impact of international economic inequalities, and they may seek to insulate the domestic population from the impact of international economic inequality.

In addition, foreigners in poor states cannot obviously claim that they participate in shared institutions with individuals in rich states. The 'constitutive' argument for equal shares of products of co-operation thus does not apply across state borders as traditionally conceived.

This sketch indicates how Liberal Contractualism might in principle allow inequality across state borders within a system of sovereign states, -- though only within limits concerning misery, domination and fair procedures. However, the premises do not match our present world order. Citizens of different states participate in a wide range of shared practices and regimes across state borders. The global economic interdependence wrought by trade and financial institutions makes individuals more vulnerable to cross-border effects of misery, domination and unfair procedures - and make it less plausible to regard domestic economies as constituted solely by co-operation among compatriots [7] .

The level of interdependence is clearly very high in several federations. In the European Union, member states have pooled sovereignty to such an extent that it seems plausible that similar distributive principles must apply as within a state [8] . Shared institutions, including freedom of movement for capital, workers, goods and services; and for some states monetary union with a common currency, render citizens vulnerable to shocks otherwise buffered by government interaction. The political units of the region are very close intertwined. It seems implausible to claim that citizens of one European state maintain domestic social institutions that are separable from those of neighboring states and of the union.

The upshot of this is that it seems difficult to defend economic inequality within federal arrangements, since the arguments for equal shares, and against inequality, apply among individuals who share social institutions -, as is the case in federal political orders. Yet there are other arguments for permitting inequality, namely for the sake of sub-unit autonomy in federal arrangements.

4 Why Sub-unit Autonomy?

Sub-unit Autonomy: Split legislative and executive competence

Federal arrangements have been presented as solutions to a wide range of perceived problems suffered by unitary governments, in order to secure peace, institutional innovation, efficiency, liberty and the like. Many forms of local autonomy will allow inequality across sub-units, in apparent conflict with normative egalitarianism.

According to the account presented above, legal autonomy is a cluster of legal powers in Hohfeld's sense. It is a product of cooperation that should be shared equally among citizens. The question of concern here is how to allocate such powers between the central and the sub-unit level. We do that by comparing the likely effects - benefits and burdens, risks and opportunities - of alternate allocations of authority. For our purposes, the interesting versions of local autonomy are those where some final legislative or executive authority is permanently allocated to the sub-units of a political order, rather than placed with a central unit. Delegation of administrative authority to sub-units is not at issue.

The legitimacy of such split authority can be assessed by a hypothetical contract between representatives of joining nations or states, deciding on the terms of their federation without knowing which nation or people they represent (Norman 1994). These approaches can easily be suspected of reifying social groups and being committed to normative communitarianism instead of maintaining normative individualism, unless firmly based on arguments appealing exclusively to the interests of individuals. Several reasons may be offered for why individuals can reasonably seek protection and furtherance of some of their interests within such a non-unitary political order.

Larger units protect against domination.

One of the historical arguments for non-unitary political orders is that they protect against unjust domination. In a federal system "the parts are so distant and remote that it is very difficult, either by intrigue, prejudice, or passion, to hurry them into any measures against the public interest." ( Hume 1793, 514-15; cf. Madison1961a, b; Beer 1993, 266; Sunstein 1994, 323). Two arguments can be discerned. The federal level of deliberation and legislation provides a second chance to protect against abuse by local majorities. Secondly, larger units provide better checks on tyranny since majority coalitions in larger polities are less likely to actively harm a local minority.

Protection against abusive power is surely valuable. However, the added deliberation provided by two-stage decisions is not a feature of the split authority we address. This benefit stems from the political authority enjoyed jointly by sub-units, in what is sometimes called interlocking federalism ({Scharpf 1985 #18940}). It primarily holds where enhanced ability to block decisions is important. This argument is weaker under other circumstances, for instance when the status quo is drastically unjust, or when other interests than avoiding domination are more important -- such as the need for joint action to secure basic needs and fair distributions. The second argument begs a central question of the political orders we consider. If larger units indeed provide better checks against tyranny, it remains unclear why sub-units should have powers at all.

Immunity from larger unit

Consider a political order where individuals' interests and preferences vary according to parameters such as geography and resources, or tastes and values. At least four reasons can be discerned for placing some decisions with sub-units.

a) Avoid Domination

Historically, one important argument stems from the "republican" concern to avoid subjection to others, in this case the population of the larger political order. Security against interference can be of great value when the population of the sub-unit has different interests than those of the majority in the federation. Such concerns were indeed central to Althusius, "the father of federalism" ( Althusius 1995, Hueglin 1999).

b) Allow Institutional Experimentation

Another classical argument for sub-unit autonomy is to allow room for institutional experimentation and innovation, "compass and room enough to refine the democracy," (Hume 1793, 514). "Experiments in living" (Mill 1962) allow citizens to learn from the experiences of other sub-units [9] .

c) Secure fit to local circumstances

Thirdly, local powers should be kept because they must address the "local circumstances and lesser interests" which center unit may neglect or overrule because of the "great and national objects" ( Madison 1961a, cf. Beer 1993, 124). Central government and common legislation ignores local variations in preferences, while sub-units can act on those preferences and hence provide a better fit of policies to local circumstances. Decentralized decision-making allows for the creation of "club goods" or "internalities" for the subsets of individuals who prefer them and are prepared to pay for them ( Musgrave 1959 , 179-80; Olson 1969 ; Oates 1972 , cf. Beer 1993, 182). However, as the federalists noted Madison , there is a risk of majority tyranny both centrally and in sub-units. So local government must also be restrained to curb "what pretends to be local self-government, but is, too often, selfish mismanagement of local interests by a jobbing and born� local oligarchy." ( Mill 1969 , 116). Of course, liberal contractualism will not allow considerations of economic efficiency to be the sole criterion for allocation of powers. Distributive standards must also apply - hence we need to know more about why such autonomy may violate equality.

An important application of this argument concerns federal orders that are created on the bases of pre-existing political units. Sub-unit autonomy allows preexisting political units to maintain some control over institutional change in those areas where common policies are not required. Individuals' interest in maintaining legitimate expectations thus can support sub-unit autonomy.

This argument does not require that institutions or cultures remain unchanged, but, rather, that the affected individuals control the speed and form of change.So this is an argument that will hold during transition, entering federal arrangements, but that may lose force over time. Moreover, the required local influence over decisions may often be secured by including them in central decision-making bodies through interlocking arrangements, rather than by granting them some powers.

This argument does not support the maintenance of unjust institutions or expectations in ways that prevent eradication of injustice.The concern to prevent subjection and to honor expectations may be overruled when other important interests are at stake.So it remains unclear why such autonomy should override egalitarian principles of justice.

d) Reduce burdens of decision-making

A fourth reason for allocating some powers with sub-units is to reduce the burdens of decision-making. Good decisions will require much information when there are large local variations. If the only individuals affected by the decisions are in one sub-unit, local decision-makers are likely to have a better grasp of affected preferences and alternatives ( Smith 1776 , 680). No clear benefit arises if outsiders are required to gather and reflect on such information, hence their efforts and resources may be better employed elsewhere.

To conclude, there are some identifiable benefits for citizens under sub-unit autonomy, where they wield more political influence over the sub-unit agenda than they would have enjoyed under a unitary political order. But by assumption, the political decisions taken centrally - in a unitary political order, for short - ensure a more egalitarian distribution of benefits and burdens, as required by the arguments laid out above. It seems clear that this gain in political influence can sometimes be advantageous -- even for those who are left economically worse off in such federal arrangements than they would be under a unitary political order. Let us call them the sub-unit poor. Their economic loss may be outweighed by the benefits provided them by more political power at the local level, due to some sub-unit autonomy.

5 Why sub-unit poor may be better off with sub-unit autonomy

The arguments presented above suggest at least three main reasons why the sub-unit poor may prefer sub-unit autonomy to increased economic benefits.

a) Reduces risk of domination

The division of political agendas reduces the vulnerability of sub-unit citizens to the views of others. The immunity accorded by split agenda allows sub-unit some protection against domination or intervention from other sub-unit. Such immunity may not only protect against ill will in the larger population, but also against incompetence or insufficient attention.

This argument holds not only for geographical units, but may also apply to minority cultural or religious groups living intermingled among others -- thus supporting consociational arrangements.

b) Provides increased responsiveness to own interests

Secondly, sub-unit autonomy over certain issues allows increased responsiveness to each affected person's interests. The sub-unit poor will have more voting power on these issues, and this can be of value. This is important in policy areas where the sub-unit population has special interests or circumstances that make it important to shape institutions and policies accordingly. The fit between policies and circumstances can be at stake by central decisions that ignore local circumstances. Sub-units with shared geography, resources, social institutions, culture or other features make for similar interests and policy choices among members of the sub-units. Added immunity and political power over such local issues for the sub-unit poor can be more important for their life plans and expectations than having somewhat more economic resources available under a unitary political order. The authority to shape institutions may be more important than a marginal increase in the resources to use within institutions that presumably fit the local circumstances less well. Even the economically worse off (within strict limits) may have reason to prefer democratically chosen policies & institutions within one sub-unit, such as a concern to maintain population in a region, desire to maintain traditional industries, etc.The alternative is to subject such decisions to centralized authority where a majority in the larger polity may overrule a local consensus. Indeed, the larger polity should have such powers insofar as the local costly choices can trigger redistributive obligations from the surrounding political order. An illustration can be seen in the German Federal Republic, in the redistributive obligations among the sub-unit L�nder. The constitution warrants central legislation when required to maintain "uniformity of living conditions beyond the territory of any one Land". Recent debates concern why rich L�nder should subsidize poorer L�nder that refuse to reform their industries. A central issue is therefore what range of outcomes and policies the sub-unit population should be responsible for in, in the sense that they should bear the full economic burden of their collective choices.

This case for sub-unit autonomy, even at some economic cost, holds only in some circumstances. The political autonomy of the sub-unit must concern policies and institutions where local fit matters for individuals' interests and concerns. This leads to at least two constraints. The sub-unit's opportunity set must not be unduly limited by the sub-unit's relative share of resources. For instance, the sub-unit must be able to implement acceptable educational or health care arrangements. Secondly, these decisions should not systematically disadvantage a minority in the population. Sub-unit autonomy is problematic from the point of view of individuals whose interests systematically differ from the majority's in ways that matter, and whose interests would be better served by institutions and policies secured through majoritarian decisions in a unitary political order. If members of such minorities also are the sub-unit poor, they would be better off both in terms of institutions and in economic terms with central decisions, and it becomes difficult to defend local autonomy.

Note that this argument holds among individuals in a sub-unit with shared circumstances -- be they resources, values or existing institutions. Those similarly affected are more likely to comprehend the need and room for common policies. But this argument does not single out states as the only relevant sub-units: States with multiple cultures or large natural variations may not satisfy these conditions.The arguments may also support non-state units, such as cross-state regions, cultural minorities, or other groups who share institutions or practices [10] ( Follesdal 1996 , Follesdal 2000b ).

c) Reduces burdens of responsiveness to others

A third reason for allocating powers to sub-units is to reduce each person's burdens of responsiveness. Sub-unit autonomy for decisions that solely affect individuals in that sub-unit relieves others from responsibility and the affiliated costs. Responsible decision-making requires information gathering, consultations with affected parties, analysis of likely impacts, and the like. Each person avoids such tasks under sub-unit autonomy. It seems plausible that such reduction in responsibility may be worth paying for, in the sense that the sub-unit poor can be expected to forego some economic benefit precisely in order to gain non-monetary resources otherwise spent deliberating about policy choices on behalf of others.

This argument from alleviating the burdens of responsiveness supports sub-unit autonomy only for certain issue areas, where the decisions only impact on members of the sub-unit. Several preconditions must be met. The individuals of the sub-unit must be able to make fair decisions through functioning democratic procedures against an acceptable allocation of background resources, and so forth. Furthermore, supplemental mechanisms for central intervention and support must presumably provide added safeguards.

These three arguments suggest that the same interests that support claims to equal shares, and restrictions on inequality, can support claims to sub-unit autonomy over certain decisions. Autonomy may be preferred even at the price of economic equality -- within strict limits stemming from the interest in avoiding misery and domination, and primarily concerning issues that do not affect other individuals.

6 Conclusion

By reflecting on "why equality?" and "why sub-unit autonomy?" I have sought to resolve some of the tensions between these norms within federal political orders.

Attention to the arguments for equality and autonomy indicate why various sorts of equality and sub-unit autonomy matter for individuals' interests.

The contractualist account of equal respect suggests that goods and burdens must be distributed among individuals across sub-units in ways that avoid misery and domination, and so as to secure fair procedures and equal shares of the products of cooperation. Liberal Contractualism still allows some distributive inequality among individuals in different sub-units, insofar as such variations are unavoidable features of the immunity and autonomy required to protect against domination, and to ensure well-informed shaping of institutions and policies to local circumstances. The same interests that ground claims to equality often support sub-unit autonomy rather than a unitary political order. Thus our interest in equal shares of income and wealth may legitimately be weighed against our interest in enjoying more political influence over matters controlled by our sub-unit.

These reflections nevertheless pose an important challenge to federal orders based on states as sub-units. As noted throughout, several arguments for decentralized legislative authority suggest that authority should be placed with sub-state regions or even to non-governmental actors (cf. Howse 1995 ,273-74). These implications are also argued on the basis of the "Principle of Subsidiarity" ( Follesdal 1998b ). The case for states as the appropriate sub-units remains to be made. Given pluralism of values and shared circumstances both within and across state borders, claims that states should be privileged parties must be substantiated better than as yet. In defense of states as sub-units, the main argument would appear to be that they have historically been the sites of political decision-making, whose citizens' expectations have converged around shared institutions. As primary makers and implementers of policies, governments may remain the most plausible agents within federal orders. Yet the pluralism of cultures and values within states suggest that they may often be too large and heterogeneous for shared policies, as witnessed by regional unrest and regionally based political parties. Moreover, permanent minorities within states, such as migrant workers, cultural minorities or the unemployed, have needs that are not always satisfied by states. From this point of view, a flaw of federal agreements is that they tend to perpetrate cleavages along state borders. This embedded partitioning may limit mutual respect, and reduce the interest in political participation beyond sub-units, as witnessed in consociational arrangements ( Lijphart 1979 , cf. de Beus 1997 ) .The political authority of pre-existing states in emerging federations may thus not be legitimate in the long run. Which groups and organizations should be recognized as sub-units with political power is an important question, not addressed here. The arguments presented here suggest that the problem is not that sub-units maintain unequal standards of living, for some such inequalities may be compatible with the equal respect of all. Federal arrangements may legitimately permit some economic inequalities for the sake of sub-unit autonomy, to the benefit of each person.

[11] References

Althusius, Johannes. (1995) [1603/1614].Politica Methodice Digesta. Frederick S. Carney, transl., and Daniel J. Elazar, introd. Indianapolis: Liberty Press.

Barry, Brian. (1989).Theories of Justice. A Treatise on Social Justice, 1. Berkeley: University of California Press.

Beer, Samuel H. (1993).To Make a Nation: the Rediscovery of American Federalism. Cambridge, Mass.: Harvard University Press.

Beitz, Charles R. (1979).Political Theory and International Relations. Princeton: Princeton University Press.

���. (1994). "Cosmopolitanism Liberalism and the States System." In Political Restructuring in Europe: Ethical Perspectives, edited by Chris Brown, 123-36. London: Routledge.

de Beus, Jos. (1997). "The Place of National Identity in a Well-Ordered Consociational Democracy." Acta Philosophica Groningana, Groningen, 24.

Cohen, G. A. (1989). "On the Currency of Egalitarian Justice." Ethics, 99, 906-44.

Coleman, James. (1990).Foundations of Social Theory. Cambridge: Harvard University Press.

Dahl, Robert A. (1985).A Preface to Economic Democracy. Cambridge: Polity Press.

Dworkin, Ronald. (1978). "Liberalism." In Public and Private Morality, edited by Stuart Hampshire, 113-43. Cambridge: Cambridge University Press.

���. (1981a). "What Is Equality? Part 1: Equality of Welfare." Philosophy and Public Affairs, 10, 185-246.

���. (1981b). "What Is Equality? Part 2: Equality of Resources." Philosophy and Public Affairs, 10, 283-345.

���. (1987). "What Is Equality?Part 3: the Place of Liberty." Iowa Law Review 73, 1-54.

Follesdal, Andreas. (1991).The Significance of State Borders for International Distributive Justice. Ph. D. dissertation. Harvard University : UMI No. 9211679.

���. (1996). "Minority Rights: A Liberal Contractualist Case." In Do We Need Minority Rights?Conceptual Issues, edited by Juha Raikka, 59-83. The Hague: Kluwer.

���. (1998a). "Democracy and Federalism in the EU: a Liberal Contractualist Perspective." In Democracy and the European Union: Studies in Economic Ethics and Philosophy, edited by Andreas Follesdal and Peter Koslowski, 231-53. Berlin: Springer.

���. (1998b). "Subsidiarity." Journal of Political Philosophy 6, 2: 231-59.

���. (2000a). "Global Justice As Impartiality: Whither Claims to Equal Shares?" In International Justice, edited by Tony Coates, 150-166. Aldershot: Ashgate.

���. (2000b). "Subsidiarity and Democratic Deliberation." In Democracy in the European Union: Integration Through Deliberation? Edited by Erik Oddvar Eriksen and John Erik Fossum, 85-110. London: Routledge.

Habermas, J�rgen. (1991). "Justice and Solidarity: on the Discussion Concerning Stage 6." InThe Moral Domain : Essays in the Ongoing Discussion Between Philosophy and the Social Sciences, edited by Thomas E. Wren. Cambridge, Mass.: MIT Press.

Hohfeld, Wesley Newcomb. (1964) [1919].Fundamental Legal Conceptions As Applied in Judicial Reasoning. Edited by WalterWheeler Cook. New Haven: Yale University Press.

Howse, Robert. (1995). "Federalism, Democracy and Regulatory Reform: A Skeptical View of the Case for Decentralization." In Rethinking Federalism: Citizens, Markets and Governments in a Changing World, edited by Karen Knop, Sylvia Ostry, Richard Simeon and Katherine Swinton, 273-93. Vancouver: University of British Columbia Press.

Hueglin, Thomas O. (1999).Early Modern Concepts for a Late Modern World: Althusius on Community and Federalism. Waterloo, Ontario: Wilfrid Laurier University Press.

Hume, David. (1793) Essays and Treatises on Several Subjects. Edinburgh.

Jordan, William. (1998).The New Politics of Welfare. London: Sage.

Lichtenberg, Judith. (1981). "National Boundaries and Moral Boundaries." In Boundaries: National Autonomy and Its Limits, edited by Peter G. Brown and Henry Shue, 79-100. Totowa, N.J.: Rowman and Allanhead.

Lijphart, Arend. (1979). "Consociation and Federation: Conceptual and Empirical Links." Canadian Journal of Political Science, 22, 499-515.

Macedo, Stephen. (1990).Liberal Virtues: Citizenship, Virtue, and Community in Liberal Constitutionalism. Oxford: Clarendon Press.

Madison, James. (1961a). Federalist 10. In Madison, James; Alexander Hamilton, and John Jay, John. The Federalist. Edited by Benjamin Wright. Cambridge, Mass.: Harvard University Press.

Madison, James. (1961b). Federalist 51. In Madison, James; Alexander Hamilton, and John Jay, John. The Federalist. Edited by Benjamin Wright. Cambridge, Mass.: Harvard University Press.

Mill, John Stuart. (1962) [1859].On Liberty. Edited by Mary Warnock. Glasgow: Collins.

���. (1969) [1873].Autobiography. Edited by Jack Stillinger. Boston: Houghton Mifflin.

Miller, David. (1995).On Nationality. Oxford: Oxford University Press.

Musgrave, Richard. (1959).The Theory of Public Finance: a Study in Political Economy. New York: Mcgraw-Hill.

Norman, Wayne J. (1994). "Towards a Philosophy of Federalism." In Group Rights, edited by Judith Baker, 79-100. Toronto: University of Toronto Press.

O'Neill, Onora. (1989).Constructions of Reason. Cambridge: Cambridge University Press.

���. (1996).Towards Justice and Virtue: A Constructive Account of Practical Reasoning. Cambridge: Cambridge University Press.

Oates, Wallace. (1972).Fiscal Federalism. New York: Harcourt Brace Jovanovich.

Olson, Mancur. (1969). "Strategic Theory and Its Applications: the Principle of 'Fiscal Equivalence': the Division of Responsibility Among Different Levels of Government." American Economic Review, 59, 479-532.

Pettit, Philip. (1997).Republicanism: A Theory of Freedom and Government. Oxford: Clarendon Press.

Pogge, Thomas W. (1989).Realizing Rawls. Ithica: Cornell University Press.

���. (1992a). "Cosmopolitanism and Sovereignty." Ethics, 103, 48-75.

���. (1992b). "Loopholes in Moralities." Ethics, 89, 79-98.

���. (1994). "Cosmopolitanism and Sovereignty."In Political Restructuring in Europe: Ethical Perspectives, edited by Chris Brown, 89-122. London: Routledge.

Rawls, John. (1971).A Theory of Justice. Cambridge, Mass.: Harvard University Press.

���. (1999) [1982]."Social Unity and Primary Goods." In Collected Papers, Cambridge, Mass.: Harvard University Press, 359-387. First published 1982 in Utilitarianism and Beyond , edited by Amartya Sen and Bernard Williams, 159-185. Cambridge: Cambridge University Press.

Raz, Joseph. (1986).The Morality of Freedom. Oxford: Clarendon Press.

Riker, William H. (1993). "Federalism." A Companion to Contemporary Political Philosophy. Robert E. Goodin and Philip Pettit, editors. Oxford: Blackwell.

Scanlon, Thomas M. (1974) [1973]."Rawls' Theory of Justice." In Reading Rawls, edited by Norman Daniels, 169-204. New York: Basic Books. First printed 1973 in University of Pennsylvania Law Review, 121, 1020-1069.

���. (1982). "Contractualism and Utilitarianism." In Utilitarianism and Beyond, edited by Amartya K. Sen and Bernard Williams, 103-28. Cambridge: Cambridge University Press.

���. (1988). "Notes on Equality." Mimeo, Harvard University .

���. (1997).The Diversity of Objections to Inequality. The Lindley Lecture. Lawrence, KN: University of Kansas.

Sen, Amartya K. (1980). "Equality of What?"In Choice, Welfare and Measurement, 353-69. Cambridge, Mass.: MIT Press.

Skinner, Quintin. (1998).Liberty Before Liberalism. Cambridge: Cambridge University Press.

Smith, Adam. (1954) [1776].An Inquiry into the Nature and Causes of the Wealth of Nations. London: Dent.

Steiner, Hillel. (1994).An Essay on Rights. Cambridge: Blackwell.

Sunstein, Cass R. (1994). "Approaching Democracy: a New Legal Order for Eastern Europe -- Constitutionalism and Secession." In Political Restructuring in Europe: Ethical Perspectives, edited by Chris Brown, 11-49. London: Routledge.

Temkin, Larry S. (1993).Inequality. Oxford: Oxford University Press.

���. (1995). "Justice and Equality: Some Questions About Scope." Social Philosophy and Policy, 12, 72-104.

Treaty on European Union. (1997). Official Journal of the European Communities C 340, 10.11.1997.

van Parijs, Philippe. (1995).Real Freedom for All. Oxford: Oxford University Press.


[1] It might be argued, e.g. by extending Pogge's arguments concerning moral loopholes, that these conclusions provide illegitimate incentives for rich states to create federal solutions and maintain their unjust privileges ( Pogge 1989 , 253; Pogge 1992b ). The issue addressed here concerns whether somewhat inegalitarian federal orders are illegitimate, not whether all attempts at creating federations are legitimate.

[2] For basic income, cf. Jordan (1998 ), van Parijs (1995 ).

[3] For details, cf. Follesdal (2000a ), and Beitz in this volume. I draw on Scanlon (1997 ), and Temkin (1993 and 1995 ) .

[4] Steiner (1994 ) focuses on the latter, in the following I pursue the former premise.

[5] Scanlon notes ( 1988 , 12) that this argument is weak: "the force of appeals ... depends on a prior claim that as participants in a co-operative scheme the individuals in question have equal claim to the fruits of this co-operation. This is an appealing moral idea, but a controversial one to serve as the starting point for an argument in support of a particular conception of justice." My account in the following seeks to make the premise more acceptable.

[6] One might think that equal impact on capabilities, rather than equal shares, would be another plausible principle. Problems of inter-personal interval comparability of capabilities count against this standard once basic functionings are satisfied, or so I argue in Follesdal 1991 . I am grateful to Andrew Kuper for reminding me of this. The concern is not that capability sets cannot be ranked - which they presumably can, even beyond basic functionings, within some ranges. The comparability problems of equality are stronger, since we must compare intervals between capability sets.

[7] Several other authors on global justice make similar points -- Beitz (1979 ), Pogge (1994 ), Lichtenberg (1981 ), O'Neill (1996 ), and cf. Miller (1995 , 104-05).

[8] Follesdal 2000a .

[9] Thanks to Michael Doyle for reminder about this argument.

[10] Thanks to Andrew Kuper for insisting on this point.