The UN Security Council, normative legitimacy and the challenge of specificity

ABSTRACT This paper discusses how the general and abstract concept of legitimacy applies to international institutions, using the United Nations Security Council as an example. We argue that the evaluation of the Security Council’s legitimacy requires considering three significant and interrelated aspects: its purpose, competences, and procedural standards. We consider two possible interpretations of the Security Council’s purpose: on the one hand, maintaining peace and security, and, on the other, ensuring broader respect for human rights. Both of these purposes are minimally morally acceptable for legitimacy. Second, we distinguish between three different competences of the UNSC: 1) the decision-making competence, 2) the quasi-legislative competence, and 3) the referral competence. On this basis, we argue that different procedural standards are required to legitimise these competences, which leads to a more differentiated understanding of the Security Council’s legitimacy. While maintaining that the membership structure of the Council is a severe problem for its legitimacy, we suggest other procedural standards that can help to improve its overall legitimacy, which include broad transparency, deliberation, and the revisability of the very terms of accountability themselves.


Introduction
The United Nations Security Council's (UNSC) legitimacy has been severely criticised for a variety of reasons. The most prominent critiques are centred on the anachronistic composition of its membership structure and its inefficiency in sustaining international peace and security in the face of long-lasting and deadly conflicts, such as most recently in Syria. But what does it mean to question the legitimacy of an international institution, and when are these legitimacy critiques justified? Legitimacy can be understood as an empirical or a normative concept. There have been different studies examining the sociological legitimacy of the UNSC, as a legitimacy belief or as authority influencing states' behaviour (e.g. Binder & Heupel, 2015;Cronin & Hurd, 2008;Hurd, 1999;. The literature on the normative legitimacy of international institutions is booming. Since Dahl's (1999) critical contribution, political theorists have developed and refined the standards that confer legitimacy on international institutions (e.g. Buchanan, 2013;Buchanan & Keohane, 2006;Grant & Keohane, 2005;Macdonald, 2016). In this literature, the concept of legitimacy is often understood as the 'right to rule' (cf. Raz, 1986). Recent contributions have sought to render the traditionally statecentred understanding of this concept applicable to international institutions (e.g. Besson, 2014;Buchanan & Keohane, 2006;Tasioulas, 2010). However, most accounts of legitimacy remain highly abstract, which makes their application to a specific international institution difficult.
This paper seeks to further clarify the concept of international legitimacy and show how it can be applied to the UNSC specifically. To this end, we draw on the 'meta-coordination view' of legitimacy (Buchanan, 2013, pp. 174-195;. The paper offers an illustration of this account of international legitimacy and contributes to a more applied understanding of legitimacy theory. In the first part, we outline our account of the legitimacy of international institutions. We argue that international institutions have to fulfil different standards depending on their purpose and competences in order to provide sufficient, content-independent reasons to comply and not to interfere in order to be legitimate. Therefore, we examine three significant and interrelated aspects of the UNSC: first, the purpose of the UNSC; second, its institutional competences; and third, its procedural standards. In the second part, we reconstruct two diverging but minimally acceptable readings of the purpose of the UNSC in the relevant literature in international relations and international law, namely international peace and security on the one hand, and the protection of human rights on the other. For this analysis, we concentrate on the question of whether the purpose of the UNSC fulfils minimal moral standards and leave aside the questions of its effectiveness and comparative benefits vis-à-vis other feasible institutions. In the third part of the paper, we distinguish three different competences of the UNSC: 1) decision-making competence; 2) quasilegislative competence; and 3) referral competence. Finally, in the fourth part, we argue for reforms of the procedural legitimacy standards of the UNSC on the basis of its purpose and competences. We start by discussing membership amendments on the basis of delegation and participation models of accountability. For reasons of feasibility, we shift the focus to other procedural standards for the outlined competences and argue that the procedural standards for the newer ones are particularly lacking.

The concept and standards of legitimacy for international institutions
The concept of legitimacy as the 'right to rule' is considered to provide sufficient, content-independent reason for compliance. 1 The meta-coordination view is based on the understanding that coercion is not sufficient to make institutions efficient, but that voluntary compliance is required, at least to some degree. Buchanan argues that before coordination through an institution is possible, it is necessary to converge on how to decide which of the different possible institutional alternatives we should coordinate our behaviour around (Buchanan, 2013(Buchanan, , pp. 174-195, 2018). The meta-coordination view understands legitimacy as addressing this problem by providing standards that are reason-based, falling into the space between self-interest and justice. Legitimacy forms part of nonideal theory as it provides institutional standards under circumstances of profound disagreement about justice but where coordination through institutions is necessary to move towards justice (Buchanan & Keohane, 2006, p. 412). The metacoordination view places most of its weight on assessable procedural standards and is therefore epistemically more accessible than a Razian account (Raz, 1986) that relies on objective reasons over which disagreement might arise.

Justifying political power with regard to equal autonomy
On the meta-coordination view, legitimacy provides a risk-benefit analysis that includes moral considerations (Buchanan, 2013, p. 190). Yet what constitutes the relevant risks and benefits remains mostly undefined. In our opinion, institutions that exercise authority through rules that demand compliance generate special risks and benefits for the autonomy of their subjects. The conception of legitimacy that we propose is a specification of the meta-coordination view based on the fundamental notion of equal autonomy, including both individual and collective autonomy (e.g. Habermas, 1996, Chapter 3;Forst, 2012, pp. 125-137). However, other readingsfor example, those based on non-domination or equal consideration of interestsmight also be used. We focus on political institutions because they do not just narrow the option set of others, but also affect their status as norm-givers. Therefore, the claim to political power, i.e. authority, through rules is specifically normatively problematic. However, it is different from brute force insofar as it addresses those subjected to the rules and demands their compliance. In this sense, authorities enter into a form of discourse in which they strive to justify why their rules should count as sufficient, contentindependent reasons for action to autonomous subjects. On this basis, we define legitimacy as the requirement to provide sufficient, content-independent reasons to comply and not to interfere with an institution's rules, compatible with equal autonomy.
As a value of political institutions, legitimacy applies to institutions with rule-setting authority, including rule application and possibly some form of enforcement. In a system of rules, it is not only the enforcing agency that exercises political power, but the combination of the legislative, executive, and judiciary that creates the systematic capacity to restrict autonomy. Therefore, legitimacy judgements evaluate the justification of political power through both the setting and the application of rules, which refers to taking decisions that are supported by and embedded within a system of rules or a treaty. We understand rules as applicable to a class of individuals or organisations as general standards of behaviour. The international system is, however, highly fragmented, which means that these functions can be more dispersed, though they can also exhibit less of a separation of powers than within the state. The Security Council is such an institution because it operates within the framework of the UN Charter and issues resolutions that can claim to be binding on all UN Members.
Our concept of legitimacy also comprises a non-ideal dimension. Nonideal theory is especially relevant at the international level because it takes into account feasibility restrictions. Currently, we are faced with an international system structured in different states that are not all democratic. Assuming that, for normative and pragmatic reasons, it is not possible to force states either to become democracies or to establish a democratic, state-like federation without their consent, a fully democratic international system is not achievable in the short-or medium-term. Therefore, such an ideal alone cannot guide which institutions we should comply with. As such, legitimacy for international institutions needs to include non-ideal standards that consider how to move from current conditions towards an ideal. In a multilevel framework, international institutions cannot and should not be thought to compensate for a lack of democracy on the domestic level. However, taking the conditions of individual and collective autonomy as the basis for legitimacy establishes two minimal requirements for international institutions. First, taking human rights norms as a baseline for individual autonomy, an institution continuously or even systematically violating these norms cannot be legitimate. Second, international institutions should neither threaten democracy in already democratic states, nor prevent democratic development in non-democratic ones.

Graded standards of legitimacy
International institutions vary vastly in their purposes, the range of authority they claim in carrying out these purposes, and the effects they generate. Given such variety, which legitimacy standards are appropriate for international institutions? We argue that international institutions have to fulfil different standards of legitimacy depending on their specific purpose(s) and competences in rule-setting and application. Based on these parameters, legitimacy standards help international institutions to generate sufficient reasons for compliance with their rules by minimising the threat to the autonomy of those subjected to them.
First, the purpose of an institution is itself part of what makes it legitimate. Certain institutional purposes are not compatible with securing the conditions of equal autonomy regardless of their effects. For example, the purpose of a murder guild violates autonomy even if the guild does not actually kill anyone. Therefore, minimal moral acceptability should be regarded as a necessary requirement for legitimacy (cf. Buchanan & Keohane, 2006, Adams this issue). 2 Beyond this restriction, different purposes can be more or less important for the protection of autonomy and accordingly generate more or less weighty reasons for compliance. Furthermore, institutions should be judged on the basis of how effective they are in achieving their purpose. On the one hand, this is necessary for institutions to generate a benefit compared to non-institutional alternatives, in order to mitigate the risk of empowering an institution in the first place. This is often captured in terms of effectiveness. On the other hand, whether institutions also have to create benefits compared to other feasible institutional alternatives is debatable (e.g. Buchanan, 2013, p. 178). In our opinion, they need to create at least some additional benefits when compared with other clearly feasible institutions so as to be legitimate. Otherwise, legitimacy cannot achieve its meta-coordinative function. However, feasibility also restricts what can be demanded from an institution to be legitimate. Finally, an institution's purpose(s) and competences are closely connected since the competences needed for an institution heavily depend on its purpose. An institution's purpose also influences the kind of standards they should fulfil because the different functions (e.g. legislative, judicative or executive) aim at different benefits and create specific risks, which can be supported or constrained by different procedures.
Second, the kind and degree of competences influences the legitimacy standards required for international institutions. The risk to both individual and collective autonomy created by international institutions ultimately depends on the political power of the particular institution in question. Therefore, graded legitimacy standards should be applied. This means that the more political power, i.e. rule-setting and application competences, an international institution wields, the more demanding procedural legitimacy standards this institution has to meet. Different dimensions of political power determine an institution's capacity to affect the autonomy of its subjects. In particular, these dimensions include the scope of its rules, its domain sensitivity, and its rule applicability. 3 The legitimacy burden for institutions increases if any of these dimensions is extended. The first two dimensions concern the width and depth of the institution's rule-setting and application domain. The accumulation of ruleseven concerning relatively technical issuescan restrict autonomy. Third, the domain sensitivity or depth considers the issues that can be affected by the institution and how important these are for autonomy. The applicability of rules can be direct or indirect. Since directly applicable rules do not require the agreement of states for their implementation, they remove the protection level of domestic institutions for individuals and collective autonomy.
While Buchanan and Keohane (2006) propose minimal standards for the legitimacy of international institutions, our approach seeks to determine specific standards for specific institutions depending on their competences. As Bodansky puts it, 'the basis of legitimacy of the Security Council may similarly vary depending on whether it is making a decision in an individual case or acting in a quasi-legislative capacity by promulgating more general rules ' (2008, p. 316). The higher the competences, the stronger the procedural standards required to protect against the risks of undermining the autonomy of all UN Member States and citizens in their capacity for self-determination. The traditional way of thinking about how to make competence delegation compatible with autonomy is through accountability standards. Transparency should be regarded as a general legitimacy standard since it is a precondition for accountability. Furthermore, attaching standards to the creation or pedigree of an institution is another way to secure accountability. As far as international institutions are concerned, state consent is the main standard used in trying to achieve this. In this paper, however, we will not discuss the normative advantages and issues related to this standard. Instead, we will focus on the revisability of the terms of accountability, which is particularly pertinent if an institution's competences change.
Third, the overall legitimacy of an institution results from a sufficient balance of reasons between the risks and benefits for autonomy. On the one hand, the risk is created by the institution's political power, i.e. its competences and the process of its creation. On the other, three factors contribute to the positive reasons for compliance: first, the importance of its purpose for autonomy; second, its comparative benefits in achieving said purpose; third, the justification of its competences through procedural legitimacy standards.
In other words, an institution with a very important purpose but only weak procedural standards might still provide sufficient reasons to be legitimate. For example, Christiano argues that institutions that fulfil morally mandatory aims do not require state consent (2012 and in this issue). Vice versa, procedural standards may provide reasons to comply with institutions that do not function wholly efficiently or efficaciously. The overall sufficiency of reasons to comply constitutes legitimacy as a threshold. However, the feasibility of institutional alternatives relevant for both efficiency and procedural standards can shift this threshold considerably (to the point where the non-institutional option would prevail). Since epistemic uncertainties, particularly regarding feasibility, need to be taken into account, it is still important to consider which procedural standards help to shift the balance in favour of legitimacy.
In the next section, we evaluate the purpose of the UNSC by discussing its importance for autonomy and minimal moral acceptability. Discussing its effectiveness and comparative benefits would require an in-depth empirical analysis of the UNSC's performance, which goes beyond the scope of the present paper. That said, it is clear that the UNSC faces some notable challenges in this respect. One of the largest problems of the UNSC's effectiveness is not its action, but its inaction (cf. Berdal, 2003, p. 26;Buchanan & Keohane, 2011, pp. 48, 51). The ability to control the unauthorised use of forcein particular by the USremains another central issue (cf. Malone, 2004;Weiss, 2003). 4 Finally, since the UNSC forms part of a bigger organisation, the checks and balances in the UN system, in particular its relationship to the General Assembly (GA), are also relevant for its legitimacy but go beyond the scope of this paper. Due to these restrictions, we cannot provide an all-things-considered assessment of the UNSC's legitimacy. Rather, we focus on the UNSC's competences in order to determine procedural standards of legitimacy that can contribute to improving the balance of reasons.

Two readings of the UNSC's purpose: beyond the letter of the UN charter
How should the purpose of an institution be defined? Formally, the purpose of the UNSC is defined by the UN Charter. Since our aim here is not to produce an in-depth analysis of the effectiveness of the UNSC, but rather to provide a normative evaluation of its purpose, we start with the explicit purpose as stated in Article 24 of the UN Charter, namely the Council's 'primary responsibility for the maintenance of international peace and stability'. This purpose sets it apart from the League of Nations' unspecified mandate. This purpose has to be interpreted in light of its evolutive practice. While initially aiming at resolving inter-state conflicts, the UNSC has resorted to Chapter VII of the UN Charter to address intra-state conflicts or non-military issues, whether national, regional or global. The scope of the UNSC's mandate in this domain is 'broad, if not virtually unfettered or unlimited, other than by the purpose and principles of the Charter, themselves broadly defined' (Shraga, 2011, p. 12).
Discussing how one should understand the UNSC's purpose of 'maintaining and restoring international peace and stability' in more detail is therefore necessary to assess its importance and the requirement of minimal moral acceptability. We suggest that the UNSC's evolving practice pursues two broad goals: (1) the 'classical' maintenance of peace and stability, and (2) the protection of human rights. Clearly, these two purposes are not mutually exclusive.
One reading confines the purpose to the UNSC's founding but evolutive aim of the maintenance of international peace and security. Two classical types of resolutions grounded in Chapter VI more concretely illustrate this first reading: peace-keeping and peace-building operations. The former, for instance, comprises the military observation of ceasefires following inter-state wars, while the latter, for instance, includes the reintegration of former soldiers into society or the rebuilding of rule of law institutions (legislative, judiciary, executive). Clearly, securing the interest(s) of states in security, peace, sovereignty, and selfdetermination remains the core normative value behind this reading. Interestingly, David Bosco suggests dividing this initial broad reading into two more specific sub-readings: the concert one, and the governance model. The governance model suggests that, moving from the original and formal mandate of preventing and managing inter-state conflict, the UNSC's purpose is to effectively address the external and novel threats to international peace and security, such as epidemic diseases or climate change. The 'securitisation' of the HIV/AIDS epidemic on the UNSC's agendasuch as the most recent 2011 resolution on the role of peace-keepers as vectors of HIV (Res. 1983 (2011))would be an illustrative example. Bosco does not, however, circumscribe the domain of 'governance' itself in clear terms (2014, p. 546). In contrast, the concert variant exclusively focuses on facilitating diplomatic negotiation between the five permanent members of the UNSC (United States, United Kingdom, France, Russia, and China) and therefore restricts the interest(s) at stake to this restricted circle: 'a concert perspective shifts the focus from the body's ability to resolve external challenges to its impact on relations between permanent members' (Bosco, 2014, p. 546). This concert view is reinforced by a consensualist decision-making process rather than a majority or super-majority: 'among the permanent members, the veto ensures that the Council is consensus-based' (Bosco, 2014, pp. 548-549).
At the other end of the spectrum, a far more challenging reading of the UNSC's purpose extends the purpose to the protection of human rightsa view that has been more developed by legal scholars. We suggest viewing this aim as individuals-oriented as it is individuals, but not necessarily states, that benefit from this much wider purpose. Although the UNSC is not a party to the main human rights treaties (such as the ICCPR), the unavoided atrocities of the post-Cold War era (former Yugoslavia, Rwanda, East Timor, and Sierra Leone), the subsequent formalisation of the Responsibility to Protect doctrine (Res. 1674(Res. (2006) and the expansion of UN-mandated bodies (ECOSOC, Human Rights Council, UN treaty bodies) have carved out a novel role for the UNSC to address intra-state or internal conflicts and use the language of human rights to articulate policies. Scholars of the UNSC speak of a 'normative shift' (Hassler, 2012 p. 28) or 'paradigm shift' (Rodley, 2015, p. 776) despite the fierce debate on the legality of humanitarian intervention. Legally, one can argue that human rightsthrough their growing recognition and implementation by Member Stateshave become general principles of international law (in the meaning of Article 38(1) of the ICJ Statute) applicable to international organisations that exercise 'governmental authority' (Sarooshi, 1999, p. 16). Endorsing the states' implementation efforts would then fuel the very definition of international peace and security. As Shraga puts it, 'a conceptual link was thus established between human rights and international peace and security, or between their serious, systematic and massive violations and the existence of a threat to the peace' (2011, p. 12).
It soon becomes apparent that these two readings fulfil the requirement of minimal moral acceptability set out above. Both aim at securing a basic form of personal and collective autonomy through the protection from state-sponsored atrocities against individuals (personal autonomy) and/or by ensuring the state's capacity to exercise external sovereignty through non-interference (collective autonomy). In fact, these purposes are more than merely acceptable; they are of fundamental importance in creating and preserving the conditions for autonomy. However, one may interpret the purpose of an institution too charitably if one only takes into account official statements and regulations. Those in power might use institutions to further their interests while framing the purpose of the institution as a public good. It is important to keep this issue in mind while discussing how the institution's purpose is actually implemented. A further caveat is in order here: the minimal condition for equal autonomy in fulfilling its purpose means that an institution should at least not create new, or sustain severe existing, inequalities. Whether the UNSC fulfils this condition remains highly questionable. 5 The changing competences of the UNSC The political power of an institution understood as rule-setting and applying competences have to be justified to those subjected to its rules in order to be legitimate. Since higher competences pose a higher risk to autonomy, stronger procedural safeguards are required. We understand 'competences' as the legal mandate that is conferred on the institution by means of treaty law and its interpretation in practice. This concept captures the principled authority that the institution claims and the de facto use of authority that may overstep these claims. It appears that the terms of the treaty have been filled by the situations to which it has been applied, which are documented in the resolutions that the UNSC is empowered to generate.
On this basis, we distinguish three different competences of the UNSC. First, the UNSC has the narrow and exclusive competence to determine the creation, deployment, and use of military force in international relations in a system structured by the default principle of non-intervention in domestic affairs (UN Charter Article 2(4)). The procedural conditions for resorting to military force are enshrined in Chapter VII of the UN Charter. It must be borne in mind, however, that Chapter VI offers various non-military measures that should be primarily exhausted (peace settlement measures). The collective security system serves an interest that all Member States share, namely the protection of their autonomy vis-à-vis external interferencesan interest that the UN Charter refers to as 'the territorial integrity or political independence of any state' (Article 2(4)). Now, Member States can resort to military force without the UNSC's authorisation, but only within the scope fixed by the same charter, namely self-defence (Article 51). Once the UNSC is seized, however, the right to self-defence is suspended. The political power that the Security Council exercises by adopting resolutions with a primarily executive character is what we call the decision-making competence of the UNSC. The Charter also establishes corresponding responsibilities on the part of its subjects (all UN Member States) in Article 25: 'the Members of the United Nations agree to accept and carry out the decisions of the UNSC in accordance with the present Charter'. This means that the decisions and resolutions of the UNSC enjoy the status of binding obligations upon all UN Member States as a matter of international law. Within this competence, the UNSC can resort to a variety of softer tools to undertake its mandate. Chapter VI of the Charter encourages and recommends peaceful settlements of disputes (Articles 33-38), while Chapter VII of the Charter allows for a range of sanctions: forcible sanctions (Article 40) or economic and other non-forcible sanctions (Article 41). In accordance with Article 28 of the Charter, the UNSC can create a variety of subsidiary bodies as needed for the performance of its functions. These include committees, working groups, investigative bodies, international tribunals, peacekeeping operations, political missions, and the Peacebuilding Commission as an advisory, subsidiary organ.
Second, its competences apply to a broad and indeterminate domain, namely all threats to international peace and securitya domain that the UNSC has the exclusive (and legally recognised) latitude to define. As Bosco puts it, 'the Charter makes no geographic or qualitative distinction between potential disruptions to the peace and makes clear that the Council can investigate any dispute it deems dangerous to peace and security' (2014, p. 546). This indeterminacy leads to what we refer to as the quasi-legislative competence conferred to the UNSC by the UN Charter itself. This means that the UNSC can create law through resolutions that include general and abstract norms incurring binding obligations on states (cf. Tsagourias, 2011). There is ongoing debate about how far this quasi-legislative competence goes and whether it reaches beyond the issues of terrorism and weapons of mass destruction (e.g. Res. 1373Res. (2001 and 1540 (2004); (cf. Neusüss, 2008;Hassler, 2012, p. 16). However, these two cases show that the UNSC has the capacity to act through competences that are broad and that resemble those of a legislative body.  (2002). However, the referral enables prosecutions in situations outside the Court's treaty-based territorial and national jurisdiction. While we cannot address this referral competence and practice in detail here, suffice it to say that it signals the growing role of the UNSC in addressing basic interests of individuals at the expense of state authority. Indeed, international crimes concern the large-scale and systematic crimes committed by individuals belonging to or in position of public authority (Zysset, 2016).
In conclusion, the UNSC has developed its rule-setting and applying competences far beyond a narrow understanding of collective security limited to non-interference and international peace. Even though its resolutions have no direct effect in the legal sensemeaning that its decisions are not immediately valid in the domestic legal orderand although the UNSC's decisions leave a wide degree of discretion to states to determine the appropriate course of action, its potential capacity to restrict state autonomy is severe. The UNSC issues resolutions that are directly binding on UN members, and even though its competences may not concern a very wide set of issues, the issues in question lie at the heart of autonomy considerations. First, security or the protection of basic human rights is a fundamental precondition for individual autonomy. Second, the UNSC's competences also have the capacity to severely restrict collective autonomy by allowing the use of force or by issuing other sanctions against sovereign states. Therefore, the legitimacy of the UNSC requires (besides the effectiveness in pursuing its purpose) fulfilling procedural standards so as to minimise the danger that these far-reaching competences pose to autonomy. It should be noted that the discussed competences might require procedural standards that contradict one another.

Procedural standards of the UNSC's legitimacy
What legitimacy standards are implied by the range of the UNSC's competences discussed here? The Council has often been criticised, and reforms of both its membership structure and its procedures have been demanded. The first issue concerns in particular the veto of the permanent members. 6 Even though all members on the Council formally have one vote according to Article 27(1), the necessity to include 'the concurring votes of the permanent members' (Article 27 (3)) gives them the power to block decisions. Besides the veto, further questions regarding the representativeness of the UNSC, in a world that is fundamentally different from the one of 1945, have been raised (e.g. Caron, 1993;Fassbender, 1998;Hassler, 2012). In this section, we first analyse the issue of the Council's membership structure. In a second step, we discuss how procedural standards regarding the decision-making process that go beyond the questions of membership may justify the three competences of the UNSC outlined above.
UNSC membership structure: delegation or participation?
Due to the restricted size and structure of permanent and elected members, the UNSC does not represent all UN Member States equally. Yet what kind of representation is required of this particular institution? We suggest that representation is important as it creates accountability as a mechanism to limit political power. Grant and Keohane define accountability as a process in which 'some actors have the right to hold other actors to a set of standards, to judge whether they have fulfilled their responsibilities in light of these standards, and to impose sanctions if they determine that these responsibilities have not been met ' (2005, p. 26). Accountability, therefore, always includes elements of ex post control. Grant and Keohane further suggest distinguishing between participation and delegation models of accountability. Both models allow for various degrees of discretion versus ex ante instructions. However, they differ on the question of who is entitled to hold the powerful accountable: in the participation model, it is those affected, while in the delegation model, it is those entrusting them with powers. Representation can be understood in both ways as participation-or delegation-centred accountability. International institutions are often based on the delegation model. Therefore, what is at stake here is not a lack of accountability, but rather to whom those institutions are accountablein the case of the delegation model, powerful states (cf. Grant & Keohane, 2005). From a normative perspective, the protection of autonomy through accountability standards is only possible if all subjected to the rules of an institution can hold it accountable. 7 Therefore, the participation model is the relevant legitimacy standard in order to guarantee meaningful accountability.
On the delegation model, the inequalities in the Security Council's membership structure are not incompatible with accountability. The Member States have delegated political power by consenting to an international treaty and to the corresponding procedures to check the authority holders. While this corresponds to the general standard of state consent in international relations, several normative problems connected with this understanding of accountability follow. First, the question of origin or source legitimacy arisesin particular for non-democratic states, but also for the international system providing sufficiently just background conditions (cf. Ronzoni, 2009) under which consent can be understood to be normatively binding (e.g. Buchanan, 2013;Christiano, 2010). Second, even though state consent can, to some degree, function as a protection for collective autonomy, it remains questionable whether these inequalities are compatible with autonomy. Third, in the case of the UN membership, the one-time consent of a state will bind all its future generations, even if the competences of the Security Council change.
Regarding the representativeness of the UNSC, the participation modelbased on holding institutions accountable to those subjected to its rulesrequires modifying the membership structure of the UNSC. Reform proposals concern mainly the enlargement of formal membership (permanent or non-permanent) and the improvement of representativeness (e.g. by more accurate regional distribution). As Hurd (2008) points out, these suggestions involve a trade-off between increasing the legitimacy of the UNSC for some, while reducing it for others. While this may be the case for the sociological legitimacy that individual Member States exhibit either as a belief in the UNSC's legitimacy or in terms of compliance, it is not true for the normative legitimacy of the Council. For example, an improved regional representation can increase the Council's legitimacy for all, not just those who 'benefit' from the representation. The UNSC's current structure is particularly problematic from the perspective of accountability to those subjected, since the Council's agenda mainly concerns developed countries while four of the five permanent members 'are from the Westernized, northern hemisphere, and even China's non-industrial status is rapidly changing' (Hassler, 2012, p. 93).
In conclusion, the participation model is normatively more robust than the delegation model because it is based on the all-subjected principle that seeks to promote the autonomy of all equally and not just one of the delegating states. It follows that the membership structure should be changed. Yet the argument of contributions has been used to justify the current structure of the UNSC. Historically, the veto was essential for bringing on board the states that could provide the necessary military power to realise the Council's purpose (Hassler, 2012, p. 35). Even if the international power structure has drastically changed since its foundation, the inclusion of the major powers is probably still the only way to bind them if compliance with the prohibition of use of force is the ultimate goal. This is a question of effectiveness and other feasible institutional alternatives. Since the participation of the powerful states is seen as necessary to maintain collective security, the UNSC's membership structure reproduces rather than overcomes the power inequalities of the states in the international system. This means that in the current setting, states' self-interest makes the UNSC a highly power-based forum.
The feasibility of changing the membership structure constitutes, therefore, a serious legitimacy challenge (e.g. Weiss, 2003). Keohane and Buchanan suggest the revisability of the terms of accountability as a legitimacy standard in itself. Based on the premise that there is reasonable disagreement about what accountability precisely entails, legitimacy cannot but depend on whether an institution facilitates 'principled, factually informed deliberation about the terms of accountability' (Buchanan & Keohane, 2006, p. 427). What does it mean that the terms of accountability must be revisable in light of critical reflection and discussion? For Buchanan and Keohane, this standard depends on feedback from external agents such as NGOs. In their own evaluation of the UNSC's legitimacy, Buchanan and Keohane (2011) evaluate the revisability of the UNSC's terms rather positively because its goals have been adapted. While we agree with this evaluation of the purpose, in our opinion this does not fully capture the revisability of the terms of accountability. Evaluating revisability should also concern the question of whether procedural standards are adaptableparticularly if the institutional competences change. In this respect, external actors' feedback is important and can be a good indicator of the need for change, although there will hardly be a unified view as to what about the terms needs to be adapted, and when. Therefore, we suggest that revising the terms of accountability should be considered one of the main standards for an international institution's legitimacy when its competences have changed. This form of revisability could also alleviate some of the problems of the delegation model (such as restricting the autonomy of future generations). Now, how can the UNSC's terms of accountability be revised? Article 108 (or in theory 109) of the Charter determines the procedure through which the UN Charter can be amended. In short, amendments have to be adopted by a vote of two thirds of the members of the General Assembly and ratified by two thirds, including the permanent members. This means that the terms of accountability can be adapted; however, the requirements are highly demanding. Since the enlargement from six to ten elected members in 1965, no reform proposals have been successful. In 1993, the UN General Assembly passed a resolution to set up an Open-ended Working Group on Council reform. 8 However, it is questionable whether the UN, under the current rules for change and with the increased number of members, is still able to change the Council's membership structure. Therefore, the possibility to revise the terms of accountability poses one of the most pressing problems for the procedural legitimacy of the UNSC.
Assuming that the feasibility of changing the membership structure is not a given in the short-to medium-term, this changes the question to that of whether circumventing the UNSC could generate an alternative institution or whether its structural deficit drops the balance of reasons to the level of the non-institutional alternative. For these questions, the importance of the institution's purpose and its effectiveness in achieving this end are essential. Polarisation in the Security Council may increase to the point where noninstitutional alternatives may be preferable. Yet, if stability considerations and the preservation of current international law matter, the transition costs resulting from abandoning the UNSC are significant. However, feasibility is not a static restraint; rather, it depends on the involved actors, timeframe, and practice (cf. Gilabert & Lawford-Smith, 2012, Erman and Kuyper this issue). Certain actions in the present may open up different options in the future. A particular problem in this regard arises if the institution itself contributes to keeping alternative institutions unfeasible. However, given that the purpose of the UNSC is of high importance, and assuming that a similarly efficient, alternative institution with a more appropriate membership structure is currently not feasible, the unequal membership structure of the Council is as such not a completely delegitimising feature. However, this does not mean that we should not try to change this structure or even seek to establish better alternatives. A more detailed discussion of feasibility and how the interplay of different actors in a multilevel system, such as states or citizens, is required to make these assessments for international institutions. This is of particular relevance, as the lack of political will on the part of certain powerful actors should not be taken to create binding obligations for others.
Alternatives to membership reforms: justifying the three competences of the UNSC The different competences of the UNSC, the expanded reach of the UN membership, and the outlined deficits of the UNSC's membership structure point to the need for additional procedural standards to improve the balance of reasons of the UNSC's overall legitimacy. Two widely discussed procedural standards regarding the internal working methods of the Council are transparency and deliberation.
Transparency is crucial to achieve any form of accountability. Reliable information on how the institution is actually performing must be provided. However, under the conditions of incongruence between the subjected and the accountability holders, 'broad transparency' (Buchanan & Keohane, 2006, p. 426), making information accessible to agents external to the institution, is required. In terms of the UNSC's transparency, several aspects stand out. First, the UNSC's institutional link to the UN General Assembly should be considered: Article 24(3) states that 'the Security Council shall submit annual and, when necessary, special reports to the General Assembly for its consideration'. Second, the Council President regularly briefs non-members and the press about private meetings (Weiss, 2003, p. 154). Third, the accessibility for NGOs has been improved through informal exchanges, e.g. with the International Committee of the Red Cross (ICRC) or the NGO Working Group on the UNSC. Additionally, the so-called Security Council Report, a private but publicly funded independent organisation, aims explicitly at advancing the transparency of the UNSC. 9 While these structures do not fulfil the perfect ideal of broad transparency and need to be strengthened further, they are in our opinion sufficient to create the basis for legitimacy.
Discourse standards can increase the legitimacy of international institutions by enhancing both the quality and inclusiveness of the discussion (e.g. Johnstone, 2008;Steffek, 2003). From the start, the idea of the UNSC was to establish procedures that induce major military powers to consider one another's interests (Hassler, 2012, p. 31). In fact, the UNSC can also grant non-members access through Article 31: 'any Member of the United Nations which is not a member of the UNSC may participate, without vote, in the discussion of any question brought before the UNSC whenever the latter considers that the interests of that Member are specially affected'. 10 Furthermore, the number and scope of subsidiary bodies assisting the UNSC's functions have been continuously growing, and these serve to increase the Council's knowledge and transparency. In addition to countryspecific bodies, the themes of terrorism and international tribunals, for instance, have several dedicated bodies. 11 Since the majority of bodies are chaired by non-permanent members and operate by consensus, they provide significant opportunities for UN Member States to get informed and have their say in discussions relevant to the UNSC's work. Furthermore, the UNSC sends missions of representatives to countries in crisis to hear the views of those affected (Weiss, 2003, p. 154).
We hold that the decision-making competence can be regarded as partly legitimised through standards of transparency, but the latter should be strengthened through discursive improvements, in particular via restrictions on the use of the veto. The draft resolution presented by the Small Five discusses possible restrictions on the veto and mentions: i) explaining the reason for using the veto; and ii) abstaining from the use of the veto in the 'event of genocide, crimes against humanity and serious violations of international humanitarian law'. 12 The requirement to make reasons public improves the discourse without preventing the abuse of the veto in the strict sense; it pressures the permanent members to use their power responsibly and in accordance with reasons that are acceptable to all UN members. Therefore, such discourse-theoretical improvements with regard to the use of the veto specifically are highly conducive to the UNSC's legitimacy and should be implemented in order to counter the inequalities embodied in the veto. Two other issues that deserve further attention, since they are areas of possible improvement, are the penholder system and the short term length of the elected members (Martin, 2018). However, the revisability of the terms of accountability remains problematic, even more so for the other two (newer) competences.
The quasi-legislative competence enables the Council to restrict the autonomy of its subjects even further, as it creates the capacity to interfere with their self-legislation. Therefore, more equal participation would be required to counterbalance the risks created by this competence. The existing deliberative standards are insufficient to effectively protect autonomy with regard to the quasi-legislative competence as they only provide access to the discussion in the Council to non-members when the UNSC itself grants it. Since neither the abolition of the veto nor the increase in membership seem realisable in the short-or medium-term, these competences should rather be exercised by the UN General Assembly.
Finally, the referral competence of the Council opens up questions of impartiality and independence as this competence may affect not only the legitimacy of the UNSC, but also that of the ICC by changing its jurisdictional reach. Since the purpose of a court requires impartiality and independence, the political nature of these referrals and the bias in case selection of the ICC, 13 to which the former contributes, leads to the de-legitimation of the ICC. In our view, this also creates a legitimacy problem for the UNSC. Consequently, the procedural standards required for this competence and the political nature of the Council are simply not compatible.

Conclusion
The aim of this paper has been to show how the concept of legitimacy can be applied to specific international institutions on using the example of the UNSC. By specifying the meta-coordination view of legitimacy based on the protection of the autonomy of those subject to the political power of an institution exercised through its rule-setting and applying competences, we have identified key aspects that are relevant for applying legitimacy to international institutions. In order to assess the risks and benefits that an institution creates for autonomy, i.e. the balance of content-independent reasons to comply, one needs to fully consider the purpose and the competences of an institution, as well as its efficiency, the feasibility of institutional alternatives, and procedural standards.
Applying the concept of legitimacy to the UNSC, therefore, requires determining its purpose and competence. We have shown that the UNSC's original mandate of maintaining peace and stability, with a focus on noninterference through the collective security system, has extended to the protection of human rights. Both purposes are not only minimally morally acceptable but crucial to individual and collective autonomy. We have determined three different competences of the UNSCnamely the decisionmaking competence, the quasi-legislative competence, and the referral competence. We have argued that the legitimacy assessment of the UNSC can be differentiated along these competences insofar as they require different procedural standards. On this basis, we have recommended that either extended competences should be reduced or the procedural standards should be improved. In particular, the competences of quasilegislation and referral are rather problematic precisely because at present they are not sufficiently supported by procedural standards. Therefore, the broader purpose of the UNSC should be welcomed from the perspective of legitimacy, whereas its wider competences should be seen more critically.