‘Subsidiarity’ is sometimes hailed as a promising ‘structuring principle’ for international law. Subsidiarity has emerged as a prominent concept in legal and political theory, not least due to its inclusion in the 1991 Maastricht Treaty on European Union and several recent US Supreme Court decisions – including on the constitutionality of the health care bill.Footnote 1 A ‘principle of subsidiarity’ regulates how to allocate or use authority within a political or legal order, typically in those orders that disperse authority between a centre and various member units. The principle holds that the burden of argument lies with attempts to centralize authority. Thus the conception of subsidiarity as laid out by the Lisbon Treaty holds that in those issue areas where the states and the EU share authority, the member states should decide – unless central action will ensure higher comparative efficiency or effectiveness in achieving the specified objectives.
Scholars appeal to subsidiarity not only to negotiate centralization and diversity in EU law,Footnote 2 but to determine the limits of sovereignty;Footnote 3 and possibly for international law more generally.Footnote 4 Some claim that a principle of subsidiarity should apply to human rights law in particular,Footnote 5 for instance, to reform the European Court of Human Rights (ECtHR).Footnote 6
Such claims go beyond the empirical point that states are central in explanations of how public international law is created. A principle of subsidiarity may serve at least two further helpful functions with regard to international law. The second role is as a way to structure legal and political debates that engages both normative and empirical premises on such issues as the allocation of authority between national and international institutions.Footnote 7 A principle of subsidiarity does indeed seem to not only explain but also provide some intellectual order to features that express the centrality of sovereign states in international law. This centrality includes: an understanding of the main social function of public international law as supplementing domestic law; the centrality of state consent in creating such legal obligations; the requirement that national remedies must be exhausted before turning to international courts and treaty bodies; the ECtHR practice of granting states a ‘margin of appreciation’ when assessing whether they are in compliance with their obligations; and the often weak treaty sanctions.
A third function of subsidiarity goes beyond the order it provides, and concerns why we have reason to value such state centrality, serving as a normative ‘constitutional principle’. As a source of normative legitimacy external to the treaties it can help justify international law – in particular these features of state centrality. Some may reasonably wonder whether subsidiarity can really support such forms of sovereignty, since subsidiarity seems to imply a challenge to such centralized powers. Yet many commentators – politicians and academics – have maintained for example in discussions about European integration that subsidiarity stands as the great limiting principle that will defend national sovereignty against incursion by the ever-expanding Brussels bureaucracy.Footnote 8
Such presumptions that subsidiarity favours the centrality of states occur in debates about reforms of the European Court of Human Rights. The centrality of states is also prominent in several contributions to international political theory, ranging from Michael Walzer’s ‘communitarian’ ‘The Moral Standing of States’Footnote 9 to John Rawls’ much critiqued ‘liberal’ Law of Peoples. Footnote 10 They both aspire to defend the normative primacy of the nation state that may undergird state-centric versions of subsidiarity. This third, justificatory function of subsidiarity is also crucial to its value as a structuring principle: such a value depends in part on whether the resulting structure merits support and deference. A normative constitutional principle of this kind may for instance help assess whether the genesis of much public international law in state consent renders it more – or less – worthy of deference. The present article addresses and challenges this normative role.
I shall argue that while a state-centric conception of subsidiarity may indeed support these features of state centrality in public international law, more defensible conceptions of subsidiarity challenge precisely these features. The upshot is that as a normatively defensible constitutional principle for international law, subsidiarity questions the centrality of states, since the most plausible conceptions of subsidiarity fail to justify the present centrality of states in international law. To defend these claims we must first discern and assess the state-centric versions of subsidiarity. The first section aids in this by presenting a backdrop of different historical and normative traditions, with significant implications for issues of sovereignty and the allocation of authority. The section explores five such theories, with particular attention to four aspects: a) whether they stress immunity for smaller units, or the obligations of larger units to assist; b) the fundamental units of normative concern: states – or individuals; c) who should have authority to specify the objectives and interests to be protected and promoted, and to decide whether more centralized action is required: the member units, or central authorities; and d) whether this version of subsidiarity is amenable to be applied not only to relations between individuals and their state, but also to relations between different levels of legal authority. These five versions are drawn from the history of political ideas. While possibly not exhaustive of the taxonomy of principles of subsidiarity, they suffice to indicate these crucial variations.
Some of these conceptions are more convincing than others, when assessed from normative perspectives committed to cosmopolitan normative individualism. Such normative traditions hold firstly, that individual human beings are the only relevant units of intrinsic, ultimate normative concern. This is of course not to deny that societies, norms and other social features matter normatively, but rather that the chief value of culture is the value of cultural membership, both intrinsic and instrumental, for individuals.Footnote 11 Secondly, this normative perspective is cosmopolitan in that every person is accorded equal status as ultimate units of moral concernFootnote 12 – which is distinct from institutional or legal cosmopolitanism that is committed to a global, centralized legal order. One reason for basing the discussion on such normative commitments is that they, as vaguely stated as here, are shared across several of the most generally discussed normative theories of legitimacy, such as Rawls-inspired contractualism,Footnote 13 republicanism,Footnote 14 and some but not all ‘communitarian’ normative theories.Footnote 15 Other communitarian traditions may be inclined to grant communities’ ‘ways of life’ and established traditions greater normative weight.Footnote 16
The second section presents the version of subsidiarity found in the EU’s Lisbon Treaty as one central case, with attention to the assumptions and role as a legal and multi-level principle. This ‘Lisbon subsidiarity’ has several distinct features, strengths and weaknesses that are often ignored – which may explain why so many endorsed the principle when first presented in the Maastricht Treaty. The weaknesses are more severe for the more state-centric version of subsidiarity allegedly found in international law, or so I argue in the third section. The implication is that a principle of subsidiarity worth endorsing cannot help justify the state centrality of international law, its defenders’ high hopes notwithstanding.
Traditions of subsidiarity
There are several versions of subsidiarity, with very different implications for the allocation of authority. They differ as to the objectives of the member units and the central authorities, the domain and roles of member units such as states, how they allocate the authority to apply the principle of subsidiarity itself, and how they conceive of the relationship between different levels of political authority. To mention some central variations: a) does the conception of subsidiarity stress the immunity of member units, or instead insist on the obligations of larger units to assist as they – or the member units – see fit? b) what are the fundamental units of normative concern? States – or individuals? c) who should have authority to specify the objectives and interests to be protected and promoted – and to determine whether more centralized action is required? The upshot is that the choice of conception of subsidiarity has drastic implications for the preferred institutional configuration, including the appropriate authority of international institutions vis-à-vis domestic authorities.
Lisbon subsidiarity shares some – but not all – weaknesses of what we may call ‘state-centric subsidiarity’, where states and their interests as traditionally conceived dominate other bodies and concerns. State-centric subsidiarity has several normative weaknesses when assessed by cosmopolitan normative individualism as sketched above. These weaknesses have implications for the alleged justification state-centric subsidiarity provides for present public international law. To get a better grasp of the strengths and weaknesses of state-centric subsidiarity, consider five alternative theories of subsidiarity.
The five accounts draw on insights from Althusius, the American confederalists, economic federalism, Catholic personalism, and liberal contractualism, respectively. These are presented in an order of descending authority for states.
Liberty: Althusius
Althusius (1563–1638), ‘the father of federalism’ developed an embryonic theory of subsidiarity drawing on Orthodox Calvinism. Communities and associations are instrumentally and intrinsically important for supporting (‘subsidia’) the needs of the holy lives of individuals. Political authority arises on the basis of covenants among such associations; it would seem that his theory can also acknowledge further levels of such covenants, e.g., among regions of states. The role of the state is to co-ordinate and secure symbiosis among associations – on a consensual basis. The notion of symbiosis may be interpreted as requiring deliberated consent about the common ends shared by member units, since it involves ‘explicit or tacit agreement, to mutual communication of whatever is useful and necessary for the harmonious exercise of social life’.Footnote 17 Althusius recognized that deliberation will not always yield agreement, particularly not in matters of faith. In such cases, he counselled religious toleration – that is: immunity for the local units from interference by more central authorities. A further feature of particular relevance for assessing such ‘state-centric subsidiarity’ is that this conception appears to take the legitimacy of existing subunits for granted. One might agree with Althusius’ claim for immunity in cases of well-functioning democracies, or other defensible forms of collective self-determination. But his approach fails to identify such standards for legitimate associations, be it regarding their treatment of members, their proper scope of activity or their legal powers. From the point of view of normative individualism, this lacuna is regrettable. Further elaborations of Althusian accounts of subsidiarity might generate some scope conditions on the nature of member units, and on standards for power allocation among them. But such restrictions are not readily apparent. This concern is perhaps most vividly underscored by the fact that the South African practice of apartheid and separation into ‘homelands’ was long regarded as justified precisely by this tradition of subsidiarity, of ‘sovereignty in one’s social circle’.Footnote 18 Some guidance might follow from the value of freedom as absence of constraints by public authorities, which forms the basis of Althusius’ argument for associations’ immunity. But the grounds and scope of this paramount interest in non-intervention remain to be identified. This lack of principles for assessing candidate member units is a weakness shared with some more recent political theories in the communitarian tradition. Thus Michael Walzer’s presumption for respect for sovereign borders even when some brutality occurs within has been roundly criticized.Footnote 19
Note that on Althusius’ conception of subsidiarity the common good to be pursued by a central political unit is limited to those undertakings deemed by every subunit to be of their interest, compared to their status quo. While this account allows for negotiation among subunit representatives based on existing preferences and resources, full agreement on ends is not expected – which is why subsidiarity is required in the first place. This account of the common good does not seem to allow any accommodation of differential organizational resources or bargaining positions; and any coercive redistributive arrangements among individuals or associations are deemed illegitimate. The upshot is that this account suffers severe problems when some subunits – associations or states – lack normative legitimacy, or when inter-state inequities raise fears of domination, or concerns for distributive justice.
Liberty: Confederalists
Similar conclusions emerge from confederal arguments for subsidiarity based on the fear of tyranny. This view starts from plausible assumptions that individuals should not be subjected to the arbitrary will of others in matters where no others are harmed, and that smaller groups are more likely to share preferences. Contributors to this conception include the American confederalists; Montesquieu held that common interest is easier to see in a smaller setting, and Fritz Scharpf who makes similar arguments for subsidiarity in the European Union.Footnote 20 This may be thought to be best secured by decentralized government with powerful member units, as on the Althusian account. Montesquieu suggested that homogeneity within member units secured unanimity, and this should be combined with a limited central agenda. Thus subunits may veto decisions, or decide by qualified majoritarian votes. For instance, an agreement on the common end of defence should move on to a discussion of the best means of defence, rather than a more expansive discussion of other common ends.Footnote 21 In effect this line of thought supports a proscriptive, state-centric version of subsidiarity.
Three weaknesses of this view merit mention, some shared with the Althusian conception. Firstly, the exclusive focus on tyranny as the sole ill to be avoided is questionable. Such a minimalist conception of relevant interests may be regarded as responses to the bewildering pluralism of world-views and conceptions of the good life. To be sure, this version of subsidiarity may reduce the need for agreement across member units. By limiting the relevant set of interests, confederalism might be thought to avoid such contestation. But it is not correct that agreement is always easier to reach in small democracies with homogeneity of socio-economic circumstances and closed borders, where politicians are less likely to pursue own advantage, and where demands are stable over time. And these conditions are unlikely – as the American federalist debate made clear.Footnote 22 So if this conception of subsidiarity is used to identify legitimate subunits, few candidates would be accepted.
The disagreement between libertarian, republican or ‘communitarian’ concern for non-domination versus liberal normative perspectives is not the only contested issue. A second normative concern is that perfect homogeneity is never achieved – even in small communities. As Madison pointed out, since it is unlikely that smaller units are completely homogeneous, the plight of minorities in local communities is uncertain. Tyranny and other forms of domination of concern for republicans or libertarians may emerge more easily in small groups while it may be easier for minorities to muster courage in larger settings.Footnote 23 In the context of the European Union, similarly, abuse of centralized powers is not the only risk: ‘local abuse’ of minorities by their domestic government is often a significant threat. So the risk of tyranny and other forms of domination does not support the strong immunity that this conception of subsidiarity accords member units. Some human rights protections by central authorities might thus be well grounded. Thirdly, the best justification of this conception appears limited with regard to its conception of the common good: only as mutual advantage among the units. This means that subsidiarity will not allow net transfers to units that require more support in order to protect and promote the interests of their individual members, or to compensate for inequities in the bargain power among member units.
The upshot is that such state-centric subsidiarity entails high risks for individuals, because member states enjoy immunity and close to veto power. There are clear tensions between individuals’ interest in avoiding domination by their member unit, and the conception of subsidiarity which permits states to not so bind themselves. These problems should be kept in mind when considering both the implications of Lisbon subsidiarity, and for other state-centric conceptions of subsidiarity as a structuring principle for international law as a whole.
Efficiency: economic federalism
The third conception of subsidiarity holds that powers and burdens of public goods should be placed with the populations that benefit from them. Decentralized government is to be preferred insofar as a) local decisions prevent overload, or b) targeted provision of public goods – i.e., ‘club goods’ – is more efficient in economic terms. This conception of subsidiarity seems to match some of the reasons why ‘Euro-sceptics’ are wary of some forms of European co-operation that adds little value.
This theory has broader ambitions than Althusian subsidiarity, in that it not only is concerned with competence allocation, but also aspires to provide standards for subunit identification. Member units do not enjoy veto powers, since free-riding member units may be overruled to ensure efficient co-ordination and production of club or public goods, especially those that are non-excludable and inexhaustible. This approach will allow multi-level and overlapping legal orders that each secures desired co-operation.
One weakness of this view is that it is limited in scope to such goods that are to mutual advantage – leaving to the side issues of how to allocate responsibilities for transfers and social justice among individuals and across units. A further challenge concerns conflicts among objectives and clubs: economic efficiency may counsel broader joint action such as removal of tariffs, while the maintenance of existing industries or culture among a smaller group may count in favour of subsidies or protectionist measures. Who should decide between these two options? Economic federalism also suffers from standard weaknesses if such elements of economic theory are regarded as a theory of normative legitimacy. Several issues are then not addressed, such as preference formation, and the reliance on Pareto improvements from given utility levels ignores what some critics describe as the unjust impact of unfair starting positions. Also note that arguments of economic federalism may recommend to remove issues from democratic and political control, and instead place them with market mechanisms or other non-political yet accountable arrangements within or among subunits. Such arguments have been used in defence of the ‘democratic deficit’ of the EU.Footnote 24 Note finally, that this tradition may question the presumption in favour of member states as the appropriate subunits. Indeed, this conception may support placing powers not only with bodies above the state, but also with substate regions: subsidiarity should go ‘all the way down’.
Justice: Catholic personalism
The Catholic tradition of subsidiarity is first expressed in Pope Leo XIII’s 1891 encyclical Rerum Novarum, which argued that the state should support lower social units, but not subsume them.Footnote 25 The doctrine was further developed in Pope Pius XI’s 1931 encyclical Quadragesimo Anno against fascism,Footnote 26 and political implications were pursued by the personalist Jacques Maritain.Footnote 27 The Catholic Church sought protection against socialism, yet protested capitalist exploitation of the poor; in effect often agreeing with economic federalism. The conception is committed to normative individualism: the human good is to develop and realize one’s potential. Theological versions will explicate this account of human potential as made in the image of God, but other philosophical positions would develop and draw from ‘thick’ albeit more agnostic conceptions of human nature. A hierarchy of associations allows persons to develop skills and talents, and assist those in need. The state and other political bodies must serve the common interests thus conceived, and intervene to further individuals’ autonomy and flourishing. On this account, subsidiarity should regulate both the allocation of legal authority and its exercise. It allows both territorial and functional applications of the principle, placing several issues outside the scope of politics – e.g., regarding religious affairs. Subunits do not enjoy veto rights. Indeed, even application of the principle of subsidiarity is sometimes best entrusted to the centre unit. Non-intervention into smaller units may often be appropriate, both to protect individuals’ autonomy, as required for their proper development, and to economize on the scarce resources of the state or other larger unit. Conversely, state intervention is legitimate and required when the public good is threatened, such as when a particular class suffers. This account of subsidiarity does allow taxation and other means of transfers across member units, as required for the ‘common good’ as defined by the authorities. Thus influential ‘social market’ economists such as Wilhelm Roepke were in broad agreement.Footnote 28
A weakness of this view concerns its premise on a contested conception of human nature as determined by a comprehensive controversial normative theory – theological or otherwise. It is thus easily subject to criticism when applied among parties who fail to share these values. One implication is that this account cannot settle beyond reasonable disagreement which subunits and cleavages should be embedded – e.g., regarding families, or labour unions – nor the role of the state or international bodies. This does make this theory more flexible and adaptable to various forms of social organization and societal development. However, problems do arise when coupled to the strong yet contested views about what the specific responsibilities should be, e.g., whether extended families, the state, the EU or the international community should be the guarantor of last resort to secure basic needs.
Consider the concerns about different, incompatible conceptions of human flourishing. Koskenniemi argues against dismissing the state:
as long as there is no wide agreement on what constitutes the good life, the formality of statehood remains the best guarantee we have against the conquest of modernism’s liberal aspect by modernism’s authoritarian impulse.Footnote 29
Such disagreement should lead us to be wary of granting the authority to make decisions about flourishing to any authorities – regardless of how centralized or decentralized. But such disagreements notwithstanding, the member units might still be able to agree on certain basic needs, human rights, etc that are less open to challenge. So this concern about value pluralism is compatible with subunits pooling certain powers to secure those objectives they and others agree are in citizens’ interest. Indeed, human rights treaties that limit state sovereignty may be more robust against criticism based on value pluralism: These treaties and international efforts are not clearly aimed at promoting flourishing, but rather largely aimed at preventing abuse of state power that causes human harms. There is arguably more – but not full – agreement on such topics than on the details of human flourishing.
A liberal contractualist case for subsidiarity
Finally, consider a brief sketch of subsidiarity compatible with liberal contractualism of the kind associated with John Rawls, Tim Scanlon or Brian Barry.Footnote 30 Such a tradition may acknowledge a – limited – role for subsidiarity, for reasons tied both to the interests of members of the subunit and the interests of other members of the larger legal and political order. The set of interests is limited, to accommodate the pluralism among citizens concerning their conception of the good life. Firstly, individuals are acknowledged an interest in controlling the social institutions that in turn shape their values, goals, options and expectations. Such political influence secures and promotes four important interests. Agreeing with the republican claim of confederalists, political power helps protect our interest in ensuring that the institutions remain responsive to our best interests as we see it; and secondly helps us avoid domination by others. In modern polities this risk is arguably reduced by a broad dispersion of procedural control in the form of universal suffrage. Thirdly, such democratic control over institutional change also helps us maintain our legitimate expectations, expressed as an interest in regulating the speed and direction of institutional change. This interest is secured by ensuring our informed participation so as to reduce the risk of false expectations. When individuals share circumstances, beliefs or values, they thus have a prima facie claim to share control over institutional change to prevent subjection and breaking of legitimate expectations. Those similarly affected are more likely to comprehend the need and room for change. Insofar as this holds true of members of subunits, there is a case for subsidiarity in two ways. Central authorities should seek to support member units’ democratic and informed decision-making, and they should also respect member units’ immunity against influence – as long as the decisions respect the best interests of its members and avoid local domination. The fourth interest that supports subsidiarity concerns its role in character formation. The principle of subsidiarity can foster and structure political argument and bargaining in ways beneficial to public deliberation, and to the character formation required to sustain a just political order. Public arguments about subsidiarity may facilitate the socialization of individuals into the requisite sense of justice and concern for the common good. For this purpose the principle of subsidiarity need not provide standards for the resolution of issues, as long as it requires public arguments about the legitimate status of subunits, the proper common goal, and the likely effects of subunit and centre-unit action.
In addition to these reasons concerning the interests of those within the member unit, there are further reasons for subsidiarity stemming from the interests of those in the rest of the legal order. They have an interest to avoid political power over issues that do not affect them, and that do not threaten the important interests of others. Such authority would require them to spend resources to determine alternatives and their consequences on affected parties, without any affect on themselves. They may reasonably want to be able to spend such resources on their other interests. One of the institutional implications of such accounts is, inter alia, support for some version of granting democratic states a ‘margin of appreciation’ as the European Court of Human Rights does, in determining whether the state complies with its human rights obligations. This margin allows different laws and policies – within a range – to accommodate different natural and social circumstances and in acknowledgement that domestic courts may be better informed about likely effects and alternatives.Footnote 31 Such a margin can narrow as a result of institutional learning and scrutiny of the reasons offered for alleged national exceptional circumstances.
There are several weaknesses of this conception of subsidarity. It only provides a limited role and weight for this principle. Its defence of immunity is limited when vital interests or human rights are at stake, or when the common interests include distributive or redistributive norms, such as equalization of living conditions across the member units. Such objectives are found in several federations, e.g., as stated in the German Grundgesetz: ‘the maintenance of legal or economic unity, especially the maintenance of uniformity of living conditions beyond the territory of any one land’.Footnote 32 Such objectives cannot be ignored by most members of the larger legal order – to the contrary, it seems that most principles of subsidiarity would require the larger order to tend to these objectives. However, this account does not single out states as the only relevant subunits. A second challenge to such accounts is disagreements about the conception of the person that would justify the substantive interests that ground the various arguments – Rawlsian attempts to ‘bracket’ such disagreements notwithstanding. A further weakness of such liberal contractualist arguments is that they under-determine subsidiarity. That is: other rules for the exercise of political power could serve the same interests. And the case for subsidiarity would have to be filled out by theories of institutional design in order to suggest suitable institutional reforms. Whether subunits should enjoy veto, votes or only voice is a matter of the likely effects when it comes to protecting human rights and other vital interests, on character formation, and on the likely effects to ensure that institutions remain sufficiently responsive to the best interests of citizens.
To conclude: these different conceptions of subsidiarity support drastically different powers and immunities for the member units. The normative assessment drawn from cosmopolitan normative individualism has challenged those conceptions that stress the immunity of member units regardless of their internal policies, and that grant them veto rights concerning any proposed centralized policies. Such conceptions seem to assume that the fundamental and ultimate units of normative concern are member units – states or communities – rather than individuals. Another conclusion concerns the authority to specify the objectives and interests to be protected and promoted – and to determine whether more centralized action is required. While basic needs and other vital interests may be relatively uncontroversial, it seems more difficult to defend authority used to promote controversial conceptions of the good life, such as aspects of Catholic social doctrines, – or economic efficiency over other important societal objectives, for that matter.
Subsidiarity in the Lisbon Treaty – compared and assessed
To illustrate how various conceptions of subsidiarity lend support to different aspects of a legal order above the state, consider the Lisbon Treaty of the European Union, which entered into force in 2009.Footnote 33
At least five aspects of the Lisbon Treaty underscore the prima facie preference for state authority over that of the EU. All five points illustrate why a principle of subsidiarity was explicitly introduced into the EU treaties starting with the Maastricht Treaty: member states sought to defend against unwarranted centralization and domination by Union authorities.
Firstly, as all treaties go, the Lisbon Treaty requires explicit consent by every EU member state. The competences and objectives of the EU are thus said to be identified, specified, and conferred by member states. Each state has determined that it is in their interest, broadly conceived, to pool some authority in order to better pursue certain joint actions. This is, of course, compatible with a prevalent claim among many observers that the EU’s power has expanded beyond the member states’ original intent. In the case of the European Court of Justice, JHH Weiler has shown how it took it upon itself to in effect ‘constitutionalize’ the Community’s legal structure.Footnote 34
Secondly, the treaty states that where the states and the EU share authority, EU competences are to be exercised respectful of a principle of subsidiarity. These points are both stated in Article 5.1:
The limits of Union competences are governed by the principle of conferral. The use of Union competences is governed by the principles of subsidiarity and proportionality.
The principle of subsidiarity in the Lisbon Treaty – ‘Lisbon subsidiarity’ – is stated thus:
Under the principle of subsidiarity, in areas which do not fall within its exclusive competence, the Union shall act only if and in so far as the objectives of the proposed action cannot be sufficiently achieved by the Member States, either at central level or at regional and local level, but can rather, by reason of the scale or effects of the proposed action, be better achieved at Union level.Footnote 35
A third example of subsidiarity in the Lisbon Treaty is the ‘yellow card’ procedure, an innovative arrangement that allows national parliamentary review of proposed EU legislation.Footnote 36 National parliaments monitor proposed EU decisions and may appeal those thought to violate Lisbon subsidiarity.Footnote 37 This mechanism addresses the tensions and risks of undue centralization when member units disagree with each other or with the central authorities whether joint action is required and efficacious, and when they disagree about how to weigh the different objectives. The ‘yellow card’ procedure concerns precisely this competence, and increases the role of member states. Some may worry that this procedure offers insufficient protection against ‘competence creep’ toward the centre, since the procedure only applies to certain kinds of legislation, and the national parliaments can only appeal to the legislative institutions, and not even to the Court of Justice of the European Union.
Further protection of member state authority might be offered by a fourth case of subsidiarity found in the Lisbon Treaty’s inclusion of various human rights protections. A principle of subsidiarity does not guarantee against domination or abuse of power, be it from central or more local authorities. The Lisbon Treaty includes human rights norms in several ways. Regulations of central authorities follow from Lisbon Treaty Article 6, further specified in Protocol 8: the EU shall accede to the ECHR and thus become subject to its Court (ECtHR), as are all member states already. The added layer of human rights norms of the ECHR makes human rights even more visible, and may add to the protection of individuals, in accordance with several conceptions of subsidiarity. Lisbon Treaty Article 6 also makes clear that the Union recognizes the rights, freedoms and principles of the Charter on Fundamental Rights, and that the Charter shall have the same legal value as the treaties. This Charter explicitly endorses a principle of subsidiarity. Its Preamble ‘reaffirms, with due regard for the powers and tasks of the Community and the Union and the principle of subsidiarity’, and Article 51 holds that the Charter is ‘addressed to the institutions and bodies of the Union with due regard for the principle of subsidiarity’. What are the implications of such ‘due regard’ for subsidiarity? The Treaty states that the EU must act and legislate consistently with the Charter, and EU courts will rule out EU legislation which is incompatible with it. The Charter also applies to the member states, but Article 6 explicitly states that the Charter shall not extend the competences of the Union. Again, this can be seen as an expression of subsidiarity, in that EU courts will react only insofar as member states implement EU law, not regarding the states’ other legislation and policies.
The fifth example of how the Lisbon Treaty expresses and respects subsidiarity concerns modifications to the EU’s monitoring system for suspected human rights violations within member states.Footnote 38 The changes reflect recommendations on the basis of the 2000 ‘reactions against Austria’ – reactions by several EU member states who suspected that the government of that country was xenophobic. Monitoring arrangements have been in place since the 1997 Amsterdam Treaty,Footnote 39 but the Lisbon Treaty develops them further. The Council may seek to determine whether there is a ‘clear risk of a serious breach’ by a member state of the values of Article 2, including human rights. The new procedure requires dialogue with the suspected member state – a requirement absent from the reactions against Austria. The Council may also make recommendations to that member state. Furthermore, the range of reactions is limited. In particular, there is no procedure to exclude a member state from the EU, even as a last resort. If the European Council finds that there is a serious and persistent breach, the Council may only suspend some of the rights of the member state in question, including the voting rights of that member state in the Council. Dialogue and recommendations are expressions of the same values supporting a principle of subsidiarity, in that state sovereignty is respected as much as possible.
These five features of subsidiarity in the Lisbon Treaty illustrate different ways to regulate the allocation and use of authority. They all underscore how the member states maintain their own authority and only suffer limited risks of domination by the central regional authorities.
One important point to keep in mind when assessing Lisbon subsidiarity is that it is limited in its scope of application, since it only regulates member states of the EU – who must all have ratified the European Convention on Human Rights. They are thus prepared to allow international scrutiny and criticism of alleged human rights violations. This conception of subsidiarity thus only applies among prima facie human-rights-respecting states. This reduces some of the concerns about the ‘state-centric’ aspects of Lisbon subsidiarity – aspects that also permeate principles of subsidiarity that address international institutions.
There are several reasons for acknowledging the important roles of states in the EU, consonant with the arguments presented for subsidiarity. Such member units are sometimes better able to secure shared interests, particularly if shared geography, resources, culture or other features make for similar interests and policy choices among members of the subunits. Partly for this reason, subsidiarity limits if not proscribes intervention into local affairs except when human rights are at stake, to help protect against subjection and domination by some others. Moreover, the limited number of issues on the EU agenda serves to reduce the risk of information overload, and foster joint gains among the member states. The deliberation fostered by subsidiarity can help build community, also within the EU bodies such as the European Parliament, by preference formation towards the common good. The deliberation about ends also supports an important sense of community for a minority: that these decisions are ‘ours’, and can foster a sense among the majority about majority constraints. Deliberation may thus enforce the boundary within which majoritarianism is accepted as a legitimate decision procedure.Footnote 40
The Lisbon Treaty and the conception of subsidiarity avoid some of the problems of Althusian subsidiarity concerning a lack of standards of legitimacy for member units: All member states of the EU must be committed to human rights, at least in having ratified the European Convention on Human Rights and being subject to its Court. This constraint on permitted states would seem to render Michael Walzer’s claims about the moral standing of states more plausible as assessed from the point of view of normative individualism. Such human-rights-respecting states, but not all others, may arguably merit sovereignty based on the interests of their citizens that these states respect and promote.Footnote 41
The Lisbon Treaty is also less ‘state centric’ in that it allows – as do economic federal and Catholic conceptions of subsidiarity – centre bodies to override member units for the sake of the stated objectives – such as the four freedoms. The Lisbon Treaty also includes ‘polycentric’ arrangements, e.g., to promote club goods among those who seek a common currency, another set of shared rules among a different set of member states concerning immigration, and so forth.Footnote 42 Some aspects of Catholic subsidiarity are also shared by the Lisbon Treaty – for better and worse. On the one hand, they both allow critical assessment of the legitimacy of particular states: The state must comply with natural and divine law to serve the common interestFootnote 43 – and in the EU case, all member states must comply with the European Convention on Human Rights, and must comply with the EU treaty. On the other hand, the standards and objectives of the social order are to be taken as given – be it by the natural order, by God and his church, or by the EU treaties and their guardian the Commission.Footnote 44 Disagreement about objectives or how to balance them are difficult to handle – by several conceptions of subsidiarity, including the economic, and the Catholic, as well as in the Lisbon Treaty. Still, the broad set of – fairly unobjectionable – objectives of the EU make for a very high number of conflicts regarding trade-offs. Consider that the values and objectives include inter alia human rights, democracy, the rule of law, pluralism, tolerance, justice, gender equality, solidarity and non-discrimination, social justice and protection, and subsidiarity.Footnote 45 The political challenge is obviously not to agree to such a list of values, norms, and principles but rather to specify and order them in a defensible way, and agree to common policies in their pursuit. It is by no means clear that appeals to subsidiarity will remove such disagreements. These concerns may be somewhat reduced by the ‘yellow card’ procedure which may alleviate or at least help dissipate some such disagreements to a reasonable extent.
Lisbon subsidiarity thus seems to avoid several but not all of the weaknesses of the various theories of subsidiarity canvassed. Two main reasons are that it applies only among member states who have ratified the European Convention on Human Rights, thereby satisfying a threshold of legitimacy; and because the Lisbon Treaty overrides member states’ immunity and veto in a wide range of issues for the sake of the shared values and objectives – conflicts among these notwithstanding. The latter in turn raises important normative challenges when opinions diverge regarding the need for central action. The central upshot for our purposes is that the justifiability of Lisbon subsidiarity is enhanced in part by reducing the state-centric features laid out in several of the theories of subsidiarity canvassed above.
State-centric subsidiarity in international law assessed
We now return to assess claims that a principle of subsidiarity may help justify several ‘state-centric’ features of international law, as a normative ‘constitutional principle’. Conceptions of a principle of subsidiarity may indeed offer a refreshing and helpful view on several aspects of international law, and on the relationship between state sovereignty and international institutions in general. Still, while some conceptions of subsidiarity may be justified from the point of view of cosmopolitan normative individualism, these do not support the centrality of states in international law. This centrality includes the primacy of state consent in creating legal obligations and an understanding of the main social function of public international law as securing the interests of states. Consider some areas where ‘the’ principle of subsidiarity may well bring order, but where this centrality of states requires conceptions of subsidiarity that are normatively unconvincing.
States as masters of treaties
States remain dominant masters of treaties: They enjoy veto rights, and immunity except in areas explicitly agreed. Treaties only bind states that have agreed to do so. States may make various reservations to these treaties, sometimes quite drastic, – as long as these reservations are not contrary to the object and purpose of the treaties.Footnote 46 Furthermore, states may exploit their bargaining power and withhold consent in order to secure a lion’s share of any benefits. Due to this central role of state consent, treaties cannot be expected to restrain tyrannical states that will simply refrain from becoming party to such treaties, and thereby maintain immunity. Nor can treaties be expected to secure fair division of benefits of international co-operation, given the large differences in bargaining power among them.
Several authors thus lament the prominence and influence of states in public international law including human rights law, e.g., with regard to states’ broad discretion, and the lack of sanctions.Footnote 47 They take this as evidence of the illegitimate bargaining power of sovereign states and the lamentable priority in international law of sovereignty over human rights. As sketched above, such immunity would still find support in Althusian and confederal conceptions of subsidiarity. But these state-centric elements are subject to severe criticism as indicated: these theories lack criteria for which legal orders may legitimately be included.
The presumption for state consent as a necessary requirement for legal obligations of states is surely open to normative questioning, especially with regard to states whose normative credentials are dubitable. Consider that states are recognized as sovereign largely in virtue of satisfying certain aspects of statehood as we know it, specified by international law – i.e., by states themselves – concerning population, territory and autonomy. The normative grounds for holding these criteria to be exhaustive of legitimate members of the community of states are absent. Michael Walzer has made similar claims, duly criticized, that
... any Leviathan state that is stable, that manages successfully to control its own people, is therefore legitimate. In a sense, that is right. In international society, Leviathan states, and many other sorts of states too, enjoy the rights of territorial integrity and political sovereignty.Footnote 48
The sense of ‘right’ here is highly controversial: as a matter of legality, Walzer may well be correct. But it remains unclear why all states thus identified should enjoy such sovereign immunity. Insofar as it is states that de facto control territories and populations, effective and sustainable compliance – i.e., problem solving – may require states’ consent. But state consent seems insufficient to determine whether the authority of an international institution is legitimate – in particular for dictatorships and other normatively worrisome states. Their consent does not suffice to grant the international institution normative legitimacy. This lack of ‘normative quality control’ of states is flawed in the same way as Althusian subsidiarity, as long as public international law is unable to specify further requirements for which states should receive such standing. The state-centric versions of subsidiarity seem at a loss to defend this general presumption.
Mattias Kumm and others have argued that state consent has become less central in international law more generally, which is
no longer firmly grounded in the specific consent of states and its interpretation and enforcement is no longer primarily left to states. Contemporary international law has expanded its scope, loosened its link to state consent and strengthened compulsory adjudication and enforcement mechanisms.Footnote 49
State-centric subsidiarity of the Althusian or confederal ilk might lament such alleged developments toward a reduced role for state consent. I submit that more plausible conceptions of subsidiarity may instead welcome this trend – if the substantive contents of contemporary international law satisfy the appropriate normative standards. Whether these developments are consistent with subsidiarity or other proposed standards of normative legitimacy is in part a matter of which alternative legislative mechanisms that might replace state consent, and whether there are other, better ways to control such authority – such as accountability mechanisms or constitutionalized guarantees and checks.
Treaties should be interpreted in favour of the signatories, rather than of individuals
Another development in international law affected by the central role of states in the creation of international law concerns treaty interpretation. At least two issues merit mention. The old doctrine of ‘restrictive interpretation in favour of state sovereignty’ entailed that treaties should be interpreted so as to minimize the restrictions on state sovereignty:Footnote 50
If the wording of a treaty provision is not clear, in choosing between several admissible interpretations, the one which involves the minimum of obligations for the Parties should be adopted.Footnote 51
This interpretive principle finds support in several of the more state-centric conceptions of subsidiarity, such as the Althusian and the confederal. These theories would thus also lament the shift away from such restrictive interpretation of international law, witnessed by the absence of this doctrine in the Vienna Convention. Interpretations now instead focus more on the object and purpose of the specific treaty.Footnote 52 Under the ‘restrictive interpretation’ doctrine treaties may well end up benefitting individuals, especially when a sufficient number of the states are democratic, or otherwise responsive to the interests of their citizens. These consequences are thus more likely in the EU insofar as all member states are democracies. But problems of a mismatch between the interests of states and those of their citizens do arise – even in democracies – partly since interpretations are largely left to the judiciary, and treaty negotiations that could check such interpretations are typically handled by the executive of each state. These processes are often opaque and insufficiently accountable to the domestic electorate or the legislature. This aspect of the ‘democratic deficit’ of the EU is common to much treaty making and interpretation, leaving the judiciary and the executive somewhat unchecked to pursue other interests than those of the citizenry at large. The Lisbon Treaty requirement that the EU should ratify the European Convention on Human Rights will help reduce such risks. Still, interpretations of agreements that benefit individuals to the disadvantage of their or other states are not to be expected.
State-centric conceptions of subsidiarity may support this priority of the interests of states when it comes to interpretation of treaties. The risk of tyranny from the centre in the name of the interests of individuals are paramount for the confederal conception, and Althusian theories may insist that the local needs are better identified and secured by local authorities, protected against interference into domestic matters. Cosmopolitan normative individualism and several of the other conceptions of subsidiarity will question precisely why international institutions should aim to promote the interests of states instead of the interests of individuals. They will therefore be more favourable in principle toward stronger human rights protections against citizens’ own government, and welcome the end of ‘restrictive interpretation’. From this perspective the challenge is not how to defend this ‘piercing of sovereignty’, but rather how to set up reliable international human rights authorities that are themselves not likely to be abused. Thus the Catholic and liberal contractualist theories will agree with the Althusian and confederal conception about these real risks of abuse of central authority – but the former insist that other risks are also paramount.
The second issue of interpretation where state-centric conceptions of subsidiarity may favour sovereignty but at the cost of normative credibility concerns the practice to accord the states a ‘margin of appreciation’ when it comes to determining compliance. This practice, mainly associated with the ECtHR, grants states the final authority to determine whether certain policies are in compliance with the ECHR – thus expressing the centrality of states.Footnote 53 Several authors note that this practice is no longer limited to the ECtHR, but is also found in the International Court of Justice. Some argue that it should be adopted for international law more generally.Footnote 54 Again, the Althusian and confederal conceptions of subsidiarity will support an expansive margin of appreciation, while the Catholic and liberal contractualist accounts will be more concerned to also recognize the need to constrain the margin.
The various accounts will all agree that the rampant value pluralism and variations in natural and social conditions across states globally does counsel a certain leeway concerning how states should best respect and promote various objectives – including human rights. The room for discretion is especially important with several partly conflicting objectives, ranging from conflicts among human rights to conflicts between human rights and other important objectives. The different theories of subsidiarity will disagree more with regard to how and where to draw the limits of such a margin, partly due to different assessments of risks. Michael Walzer gives expression to a related ambivalence thus:
[the disagreement] has to do with the respect we are prepared to accord and the room we are prepared to yield to the political process itself, with all its messiness and uncertainty, its inevitable compromises, and its frequent brutality. It has to do with the range of outcomes we are prepared to tolerate, to accept as presumptively legitimate, though not necessarily to endorse.Footnote 55
Three cosmopolitan normative individualist reasons for constraining the margin is firstly, of course, to ensure the objectives of the treaties – in the case of human rights treaties: the protection and promotion of individuals’ human rights against state inaction or worse. Even though they are also respectful of the values of local communities and shared ways of life, the less state-centric conceptions of subsidiarity will be concerned to constrain such variations, in order to also secure other vital interests of individuals. One of the central challenges at the global level is indeed the broad range in domestic values and cultures – also concerning human rights. Some such claims of divergence may be overdrawn – witness the discussion some years ago about ‘Asian values’ versus human rights. Still, the appropriate response to all violators of treaties whose defence is simply that the treaty is counter to their own values is hardly leniency.Footnote 56 Again, the presumption that states will be sufficiently respectful of their inhabitants’ rights may be stronger in democracies than in more autocratic governments. The implication may be that such margins of appreciation should be broader for democracies – e.g., within in the EU – than for non-democratic states. The assessment of the democratic quality of any given state is no easy task, however – and underscores the risk of according courts and other treaty bodies too much discretion. The second reason to constrain the margin of appreciation is precisely to reduce the risk of domination by the courts and treaty bodies. Flexibility in interpretation and adjudication by these bodies can be abused in the absence of mechanisms to ensure that their authority is exercised in pursuit of the stated objectives. The third reason to maintain a narrow margin is to protect the courts and treaty bodies against more powerful states. Writing about the ECtHR, Roland MacDonald notes: ‘The margin of appreciation gives the flexibility needed to avoid damaging confrontations between the Court and Contracting States over their respective spheres of authority.’Footnote 57 The general risk is that the margin of appreciation doctrine may be abused not by the courts, but by powerful signatories. There is a real risk that the powerful do as they will, and the weak do as they must. Thus the ECtHR may find itself the weak party when confronting a powerful signatory such as the EU. The risks of abuse of power may be even greater in the global sphere, with weaker treaty bodies.
The upshot is that whilst state-centric conceptions of subsidiarity are likely to favour granting states a broad margin of discretion, those conceptions more favoured by cosmopolitan normative individualism are more sceptical – while agreeing that the risk of granting treaty bodies authority is real and merit institutional responses.
Conclusion
What are we to conclude, then, about the claim that several aspects of international law are justified as based on a principle of subsidiarity? The upshot of the arguments laid out above is that such a claim is difficult to maintain. While a principle of subsidiarity can be found to support several important state-centric features of international law, this version of subsidiarity suffers from weaknesses related to it being state-centric. The veto power of all states regardless of their normative legitimacy, and the leeway granted states when assessing their compliance appear to lack normative justification. A principle of subsidiarity worth respecting according to cosmopolitan normative individualism would seem committed to the fostering of individuals’ interests rather than those of states, and require that we assess states – and conceptions of subsidiarity – in such terms. States’ claims to sovereign authority should depend on whether their system of governance is sufficiently responsive to the best interests of their citizens and those of human beings in general. And their authority should be delineated and vetted by normative standards of legitimacy, instead of reflecting their current power in our system of states. Thus, attempts to justify several state-centric aspects of public international law by such appeals to subsidiarity are found wanting. However, it is worth noting that considerations of subsidiarity may provide strong arguments against centralized global government, and instead support a revised system of states whose sovereignty is limited and conditional on whether the state actually does respect and promote individuals’ well-being – perhaps enjoying a certain margin of appreciation.
This article has sought to argue that ‘the’ principle of subsidiarity cannot provide legitimacy or contribute to defensible structures in international law, including human rights law, on its own. While in general agreement with those who counsel some role for subsidiarity considerations, this paper cautions that the debates require attention to more items than are sometimes noted. The principle of subsidiarity stands in need of substantive interpretation, which must be guided by normative considerations. While a principle of subsidiarity may prove a helpful ‘constitutional principle’ in international human rights law and other international law, many crucial aspects require much further attention, including the standing of states. In particular, state-centric subsidiarity has several weaknesses that seems to also inform international law. The Lisbon Treaty version of subsidiarity avoids some of these problems by how it expresses and embeds its principle of subsidiarity: Only to member states and a Union that satisfy human rights standards.
State-centric principles of subsidiarity assume with insufficient argument that states and their interests should determine the scope of international institutions’ authority. The more plausible versions of subsidiarity canvassed above insist that ultimately, even the authority of the states is justified on the basis that such powers benefit individual persons’ interests better than alternative institutional structures. If this is accepted, we should hold Lisbon subsidiarity and state-centric subsidiarity to such standards, namely asking what sort of sovereignty states should enjoy, over what scope of decisions. We might then support stricter requirements on states in good standing, and stronger monitoring and sanctioning bodies to protect and promote human rights.
‘The’ principle of subsidiarity cannot on its own provide legitimacy or contribute to a defensible allocation of authority between national and international institutions, e.g., regarding human rights law. Appeals to subsidiarity are too vague, and require attention to more items – including the standing of states, whether centre action is prohibited or required, and who should decide such issues. The more plausible versions of subsidiarity insist that ultimately, these questions are answered in light of which arrangement benefits individual persons’ interests better than the alternatives. Unrestricted sovereignty and state consent are not obvious parts of the solution, but explicitly conditional or qualified sovereignty may well be.
Acknowledgements
A precursor to this paper was discussed at a workshop on ‘Global Governance as Public Authority: Structures, Contestation, and Normative Change’ 15–16 April 2011 at the Hertie School of Government, Berlin. I am grateful to the participants and organizers, especially Nico Krisch, and to Birgit Schlütter and Geir Ulfstein. I also thank two anonymous referees for constructive suggestions. This paper is written under the auspices of the ERC project MultiRights available online at <www.multirights.net> and the Centre for the Study of Mind in Nature (CSMN).
