You can't tell me what to do! Why should states comply with international institutions?

Increased international coordination after the Second World War saw both the creation of more multilateral institutions but also the deepening of existing institutions' authority. Since then, many international institutions have faced criticism from both civil society and state representatives (e.g. Zürn et al., 2012) which has intensified in recent years. Yet, addressing global problems such as poverty or climate change requires greater international collaboration than ever. So, how should we evaluate the authority of international institutions that demand compliance? When are such institutions legitimate? The question of state legitimacy has been at the core of political philosophy, and the concept and standards of legitimacy in respect of international institutions have recently garnered much attention (e.g. Adams et al., 2020; Besson, 2014; Buchanan & Keohane, 2006; Christiano, 2012; Sandven & Scherz, 2022). However, one aspect has largely been neglected in this debate, namely how and when legitimate authority is able to bind not only individuals but also states.

The concept of legitimacy is traditionally applied to state institutions, though questions of legitimacy also arise for international institutions as soon as they demand compliance. International institutions seek to bind and therefore demand compliance primarily of states through their rules or agreements, while they do so of individuals only secondarily through their states. It remains unclear how normative conceptions of legitimacy apply to states. Therefore, it is important to understand how legitimacy is applicable internationally to bind states.

This paper tackles precisely this question: Under what circumstances should states comply with international institutions? It is generally assumed that legitimate authority can demand compliance of subjected individuals also in cases when compliance conflicts (or is seen to conflict) with their self-interest according to normative considerations such as the common good, rights, or moral reasons. Yet, for states, such normative considerations are often seen as naïve and thus quickly abandoned for realist international relations theories. Therefore, the legitimate authority of international institutions is often challenged on the basis of state sovereignty. On the other hand, if the normative side is taken seriously, state consent is criticized for as a legitimacy standard. Can states be bound in order to solve global problems or to comply with human rights even without their consent? If individuals have a right to “personal pursuits” (Tan, 2004) based on their freedom or autonomy, then states might have a similar right to decide for themselves and only be subjected to the authority of international institutions to which they have explicitly consented. However, the use of such “domestic analogies” has often been criticized (e.g. Bull, 1977). Therefore, we need to understand what reasons for actions apply to states and what role they play in binding states or granting them immunity from international institutions' decisions. This paper seeks to provide a coherent normative account of reasons for states to comply with international institutions.

The first part of the paper addresses the conceptual question of how legitimate authority is thought to bind its subjects. It starts from legitimate authority as content-independent and exclusionary reason-giving and outlines conceptual issues with the instrumental conception of legitimacy. Proposing an autonomy-based conception of legitimacy (see also Scherz, 2021; Scherz & Zysset, 2020), it argues that self-regarding reasons cannot be the basis for authority but that there is a right to personal pursuits for issues that only affect oneself. The second part turns to how such reasons apply to states. It discusses whether states also have a right to personal pursuits that can ground a domain of sovereignty, in which they are immune to the claimed authority of international institutions. It argues that such a domain of sovereignty exists but is restricted by other-regarding reasons concerning both their respective citizens and others outside the state. The paper addresses an objection to this conception, namely that democratic states owe less compliance to international institutions than nondemocratic ones. Finally, it discusses what relevance different purposes of international institutions and state consent have for the legitimacy of international institutions.

The paper aims to determine the implications for international institutions' legitimacy if we accept a Kantian autonomy-based account or a modified, noninstrumental service conception. In particular, the autonomy-based conception of legitimacy leads to the conclusion that all states have reasons to join and comply with multilateral arrangements that ensure the equal autonomy of their own citizens but also the autonomy of other states' peoples.

The increasing authority that is claimed by supranational, transnational, and international institutions raises the question of what makes international institutions legitimate.¹ There is a growing literature on why states comply with such institutions (e.g. Checkel, 2001; Hurd, 1999; Tallberg & Zürn, 2019). Jan Hurd distinguishes between coercion, self-interest, and legitimacy as reasons or motivation for compliance. Here, legitimacy is understood as “the normative belief by an actor that a rule or institution ought to be obeyed” (Hurd, 1999, p. 381). This sociological (or descriptive) approach conceptualizes legitimacy as the belief in the justified authority of a certain international institution.² From this debate about sociological legitimacy, the relevance of specific compliance audiences such as states or civil society groups is well-known (e.g. Bodansky, 2013). In fact, in the case of international courts, attentive publics may primarily consist of various elite actors such as national governments, courts, or legal communities (e.g. Pollack, 2018). Yet, how are we to think about different compliance constituencies from the perspective of normative legitimacy? Why should states comply? Are individuals and states bound by different reasons? This question is highly relevant for the legitimacy of many international institutions, as they are first and foremost addressing states in their decisions.

The question about normative legitimacy has traditionally been phrased in terms of whether an authority has not only de facto authority but a “right to rule” often associated with a corresponding obligation to obey on the side of those subjected to said rule (Raz, 1986).³ In other words, such a right to rule addresses the question of how political power can be justified, that is, how to set and apply binding rules to which compliance is normatively demanded. Since international institutions seek to bind primarily states, it is important to understand how legitimacy is applicable internationally to do so. Before turning to states, however, it is necessary to understand how different conceptions of legitimate authority are thought to bind their subjects. This is crucial to understanding how reasons are generally thought to generate legitimate authority and because the reasons to comply for states may depend on what these reasons are for individuals. To do so, I characterize the concept of legitimacy as a normative power: revisiting one of the most prominent conceptions of legitimacy, namely the service conception, and outlining the alternative of an autonomy-based conception of legitimacy. This conception is shared by many liberal and republican views.

How do these conditions of legitimacy apply to international institutions? Are there underlying duties for states and can states claim a similar right to personal pursuits in order to protect their sovereignty? In order to understand how exclusionary content-independent reasons to comply can be generated by international institutions, to gain authority over states, it is important to understand what reasons apply to states in general. This is not a simple question and requires a more extensive discussion than the format of a paper can accommodate. This section focuses on four aspects: (1) It outlines how the distinction between self- and other-regarding applies to states; (2) It argues that other-regarding reasons can establish a general requirement to enter into multilateral arrangements; (3) It addresses the objection that the respect for political autonomy of democracies causes the autonomy-based conception to acknowledge mainly authority over nondemocratic states; (4) It discusses how different purposes and state consent influence the legitimacy of international institutions.

An account of the legitimacy of international institutions needs to clarify what reasons apply to states in order to ground the requirement for their compliance. I have proposed an autonomy-based conception of legitimacy that is applicable on both the individual and the state level. Reasons that apply to states are based on those of individuals, but both an institutional and people-centered understanding of the states leads to more other-regarding reasons for states. Therefore, the domain of self-regarding reasons which cannot serve as grounds for authority is more restricted for states than for individuals. As to whether these reasons are different for democratic and nondemocratic states, I have argued that reasons for states are generally the same as they are grounded in background duties to respect the equal autonomy of others that apply to all states. The autonomy-based conception rejects the idea of subject fragmentation regarding political authority also on the international level. Rather it explains why states have reasons to enter multilateral institutions that ensure rule of law and impartiality as requirements of the autonomy-based account. Finally, since neither individuals nor states are alone on an island, the autonomy-based account's focus on disagreement provides an understanding of why authorization mechanisms such as consent and democracy are important for legitimate authority.