We are grateful to Jeff Dunoff and Antje Wiener for organising this symposium on behalf of Global Constitutionalism and for the contributors for taking our concept of contested multilateralism seriously and writing such thoughtful essays in response.

Our article on ‘contested multilateralism’, which appeared in the 9(4) Review of International Organizations (December 2014) argued that multilateralism is not essentially cooperative and characterised by integrated rules. Instead, established multilateral institutions are often challenged through the use of other multilateral institutions, as states and non-state actors seek to transform institutional rules or decision-making procedures. Such strategies may rely on formal or informal multilateralism, and may be launched in conjunction with or separate from unilateral or bilateral threats. Regardless of approach, contested multilateralism promotes policy and institutional change in ways that are quite distinct from unilateralism or bilateralism.

More precisely, the phenomenon of contested multilateralism (CM) occurs when coalitions of states and/or non-state actors either shift their focus from one existing institution to another or create a new multilateral institution to compete with existing ones. Such challenges can be observed across many different issue areas in international relations, including issue areas such as counterterrorism and non-proliferation that are central to state security. Our paper discusses a number of instances of contested multilateralism, including challenges to established institutions like the UN Security Council, the World Intellectual Property Organization, and the World Health Organization (WHO), as well as episodes of institutional creation like the International Renewable Energy Agency and the Proliferation Security Initiative. For each case we identify the challenged institution; the challenging coalition; the principal sources of dissatisfaction with the existing institutions; whether the issue focused on institutional missions, rules or practices; whether a new institution was created (‘regime creation’) or an existing one given more powers (‘regime shifting’); and the ultimate effect on the regime complex for the relevant issue-area.

‘Contested multilateralism’ is a conceptual and descriptive paper, identifying and naming a distinctive phenomenon in contemporary world politics. It does not seek to provide a tested explanation of variation in CM across issue areas, although at the end of the article we suggest, but do not test, some hypotheses about the conditions under which we expect it to occur.

The central question for this symposium concerns the relevance of our concept of contested multilateralism to the concept of Global Constitutionalism.

As Andreas Follesdal explains, the identifying characteristics of Constitutionalism are institutions that establish and limit the exercise of political authority, by codifying institutions, creating legislative powers, and limiting governmental powers through provisions to assure the accountability of rulers to those subject to rules. Consistently with this conception, Christian Kreuder-Sonnen and Bernhard Zangl define the constitutionalism of multilateral institutions as ‘the degree to which their exercise of authority is legally constrained in terms of democratic participation and judicial accountability’ (p. 328, in this issue). Normatively, democratic legal theorists believe that constitutions as such are valuable since they provide stability and stabilise expectations (Follesdal: p. 306, in this issue) and that they do so within a framework of accountability that limits the abuse of power.^{Footnote 1}

Empirically, it is clear that no comprehensive constitutional institutions operate at the global level, but also that in some domains such institutions exist, constituting what Neil Walker calls post-national constitutionalism as plurality, which we characterise as ‘pluralistic global constitutionalism’. That is, the key defensible empirical proposition of Global Constitutionalism is that some ‘islands’ of global constitutionalism exist in world politics, associated with certain treaty regimes that have become institutionalised and legalised, and with the European Union.^{Footnote 2} But none of the large nuclear-armed countries – China, Russia, or the United States – is an active participant in such constitutionalised regimes.

These normative and empirical propositions of Global Constitutionalism are logically unrelated: one could believe that pluralistic global constitutionalism exists and is desirable, exists and is undesirable, does not meaningfully exist but would be desirable, or that it does not meaningfully exist and would not be desirable. It appears, however, that most scholars of Global Constitutionalism believe both in the core empirical and normative propositions associated with this point of view.

Andreas Follesdal is principally concerned with the normative implications of contested multilateralism for global constitutionalism, which he argues can be either positive or negative depending on the situation. We agree. Our article does not address these issues since its focus is conceptual and empirical. As he points out, however, from a normative standpoint, contested multilateralism is not inherently undesirable. Disrupting the institutional status quo may weaken institutional constraints and destabilise expectations, but if new authorities are more legitimate and better able to accomplish policy goals, this decrease in constitutionalism may be normatively desirable. When states and non-state actors chose to create the Global Alliance on Vaccines and Immunization (GAVI) outside of the WHO, effectively wrestling away control, they did so to allow stronger public–private partnerships; as a result, GAVI has been much more effective at delivering vaccines to children in developing countries.^{Footnote 3} Surely this relative weakening in the institutional order is normatively offset by the improvement in outcomes.

Of course, the normative implications of CM are more troubling when challengers articulate competing visions that run counter to liberal values. After 9/11, states worked together through a variety of institutions to establish principles and norms on combating terrorism. As European countries and civil society actors worked through the UN to increase the emphasis on pairing counterterrorism efforts with human rights protections, however, China, Russia, and the central Asian countries used the Shanghai Cooperation Organization (SCO) to articulate an alternative vision. The SCO employs an expansive definition of terrorism that conflates terrorist acts with the co-equal ‘evils’ of separatism and extremism; such rules have been used to justify violations of international human rights treaties.^{Footnote 4} From a normative standpoint, this type of CM both weakens the constitutional order and produces worse results.

Would such a challenge have occurred in a world where CM was less common? Perhaps not. It may very well be the case that even if CM is often normatively desirable because it deepens cooperation or allows traditionally underrepresented actors more influence over world politics, it also creates the conditions for converse action. Pluralism may beget fragmentation, and fragmentation may beget pluralism.

Gráinne de Búrca asks what contested multilateralism adds to these discussions. She thinks that the answer is our stress on conflict, and she argues against this emphasis, suggesting that several of our examples of CM do not involve conflict over goals. Challenges may be to the incomplete realisation of agreed goals (as in the health cases) or to the failure of an institution to emphasise sufficiently, in the view of challengers, one aspect of the overall goals (the creation of the International Renewable Energy Agency to promote renewable energy).

The substantive point is well taken, but our article did not claim that contested multilateralism is about conflict over goals. Instead, we focus on challenges to established missions, practices, and rules. Both regime shifting and regime creation typically focus on alleged flaws in institutional missions, rules, or practices to attain a general set of objectives such as health, energy security, economic development, or the imposition of effective constraints on terrorism or nuclear proliferation.

With this clarification, however, Professor de Búrca’s challenge remains. What is the value of the concept of contested multilateralism? Are we merely creating a new term to describe familiar phenomena? Since doing so would be more confusing than productive, it is important for us to clarify what the value of the concept is. Fortunately, Neil Walker comes to our rescue.

First, Professor Walker emphasises (pp. 316–17, in this issue) that contested multilateralism can be distinguished from certain abstract formulations of constitutionalising processes by its emphasis on agency. We are interested in how coalitions of specific actors – especially but not exclusively states – pursue strategies to attain their objectives through CM.

Second, as Walker also points out, CM calls attention to dynamics:

This emphasis on dynamics is particularly important for how we understand the current institutional order. Institutions do not arise in a vacuum, and can only be understood fully in relation to each other. Moreover, the institutional status quo as it currently exists may actually be in a state of flux. Thus we should understand the Asian Infrastructure Investment Bank (AIIB) not only as an independent entity or a reaction to the World Bank’s failure to reform its voting system, but also as an actor with the potential to codify or upend existing norms of development banking.

Third, and most important, CM facilitates movement away from comprehensive generalisations to genuinely comparative analysis, by issue-area or region, of patterns of contestation, or lack thereof. Descriptively, we want to understand how these patterns of contestation vary across domains. Table 1 of our 2014 paper is designed to facilitate descriptive comparison by asking, about each of our cases, what the principal sources of dissatisfaction were; whether the focus was on rules, practices, or mission; and what the implications of CM were for the relevant regime complex. As Walker points out, our paper does not attempt to explain contested multilateralism, but specifying descriptive variation would be critical for such an explanatory inquiry, since the key questions for social scientists are always under what conditions one type of behavior or another, or one type of outcome or another, occur. Contested multilateralism is an important concept because it can enable social scientists to move beyond generalisations about fragmentation to explore, in comparative fashion, the specific strategies and processes that generate fragmentation, or whose absence enables relatively integrated regimes to exist.

Kreuder-Sonnen and Zangl ask a further question, which is not addressed in our paper but is important: what are the consequences of contested multilateralism? We will respond to their interesting points, but we first need to deal with a terminological issue.

In reading their essay, we were somewhat confused by their categories of ‘positive’ and ‘negative’ consequences since they seemed, until the last paragraph of their essay, to conflate normative with empirical issues. Empirically, ‘positive’ and ‘negative’ refer to whether CM promotes Global Constitutionalism or degrades it: that is, whether CM increases legal constraints on authority (the right-hand column of their Table 1, at p. 331, in this issue) or decreases these constraints. Normatively, ‘positive’ and ‘negative’ refer to whether the evaluator approves of a certain direction of change in view of his or her values. Unless we assume the absolute superiority of Global Constitutionalism over other institutional arrangements, the normative and empirical judgments will not necessarily be the same, so confusion arises from using ‘positive’ and ‘negative’ in both empirical and normative ways. Only in the last paragraph of their essay do the authors point out that there are values other than constitutionalism often at stake when we observe contested multilateralism. Under some circumstances these values might be so important that they would justify a less constitutional approach. Since the ‘positive’ and ‘negative’ language is therefore confusing, when we are referring to the empirical effects of CM, we will shift to the language of ‘pro-constitutional’ and ‘anti-constitutional’.

To us the core of the contribution by Kreuder-Sonnen and Zangl is their concept of ‘polity-centred contested multilateralism,’ in contrast to ‘policy-centred’ CM. In polity-centred CM, ‘the source of dissatisfaction is structural aspects of an institution’s political order’ (p. 330, in this issue). This is a great idea. We wish we had thought of this concept ourselves and will use it (with appropriate credit) in the future.

Kreuder-Sonnen and Zangl distinguish ‘top-down’ polity contestation – led by states – from ‘bottom-up’ contestation, led by social actors. They view top-down contestation as typically anti-constitutional because states generally seek to relax constraints on the discretionary use of power by institutions they control. The AIIB and the European Stability Mechanism (ESM), which is not bound by the EU legal system, are cases in point for them. In contrast, Kreuder-Sonnen and Zangl view bottom-up as likely to be pro-constitutional, since societal actors will typically favour more constraints on the wielding of authority and more accountability. Their examples are the contestation of UN Security Council anti-terrorism regulations, and the societal critiques of the WHO’s lack of transparency in issuing recommendations during the swine flu episode of 2009–10. We view both of these propositions as plausible hypotheses that it would be worthwhile to test. Even if these generalisations hold, there may be exceptions, which would be interesting.

In both of their cases of bottom-up contestation, the societal actors demanded more accountability after the fact, and the results were at least marginal increases in constitutionalisation. But the two cases of top-down contestation were very different from one another. World Bank development lending had not suffered a crisis in the years before 2016, rather China and its allies wanted to carry out development lending under their own control, perhaps with fewer restraints. Whether the AIIB turns out to be a clear case of anti-constitutionalism will depend on the degree to which the World Bank is forced to lower its standards or relinquish control to the AIIB. In this regard, the April 2016 co-financing framework between the World Bank and the AIIB seems a sign of partnership rather than rivalry.

The other case of top-down contestation is the ESM, about which it is even more difficult to make normative judgments. Unlike the AIIB, the ESM arose in response to a crisis – specifically, the financial and banking crises in the EU for which the established legalised mechanisms seemed insufficient. While the ESM may be anti-constitutionalism, the clear tradeoff between constitutionalism and effectiveness makes its normative consequences less clear.

These illuminating contributions have reinforced our view that the concept of contested multilateralism provides a useful way to describe certain contemporary patterns of state and non-state action that have the effect of making international regimes less integrated and accentuating the fragmentation of international law. The phenomenon of contested multilateralism challenges teleological interpretations of Global Constitutionalism as a view of the world. Certainly, the development of global constitutionalism, insofar as it is occurring, is far from linear; and backsliding sometimes occurs. Furthermore, the vicissitudes of global constitutionalism derive not principally from legal debates but from political action: states, societal groups, and institutions – international as well as transnational – seeking to further their interests and realise their values. Changes in global constitutionalism occur through an emphatically political process.

We conclude by reiterating our thanks to the Editors for convening this symposium, and to all five contributors for their perceptive comments.