2.1 Introduction

The adoption of the Paris Agreement in December 2015 was portrayed by then United Nations (UN) Secretary-General Ban Ki-moon as ‘a resounding success for multilateralism’ (UNFCCC, 2015) – after so many years of uncertainty had passed that many had begun to fear that the United Nations Framework Convention on Climate Change (UNFCCC) had become permanently gridlocked. Paris seemingly reaffirmed the centrality of the regime established by the UNFCCC in the international governance of climate change, and its ability to adapt to new challenges.

Although the UNFCCC can be viewed as a form of ‘monocentric’ governance (Cole, Reference Cole2015; see also Chapter 7), in the three decades of intergovernmental efforts to address climate change, it has become increasingly clear that it operates as part of a polycentric governance system. Due to the physical and socio-economic interconnections between climate change and a range of other issue areas, institutional overlaps between the climate regime and other international institutions from other domains such as trade and investment, human rights, other environmental issues (e.g. ozone depletion and biodiversity loss) and specific sectors (e.g. aviation and maritime shipping) are inevitable. Scholars have variously pointed to the ‘fragmentation’ of international institutions in this issue area (Biermann et al., Reference Biermann, Pattberg, van Asselt and Zelli2009; Zelli, Reference Zelli2011a), to the existence of a ‘regime complex for climate change’ (Keohane and Victor, Reference Keohane and Victor2011) and to ‘experimentalist governance’ (Sabel and Victor, Reference Sabel and Victor2017). In essence, all these terms recognise the increasingly polycentric nature of climate governance.

In this chapter, we systematically sketch the domain of international climate change governance from the angle of polycentricity, focusing on intergovernmental multilateral institutions. We pursue two objectives: characterising this governance system as polycentric, and then discussing to what extent certain manifestations of polycentricity have already materialised in this system. With regard to the first objective, this chapter begins by qualifying the claim that the UN climate regime is ‘monocentric’. This is followed by an overview of governance through several other intergovernmental regimes and organisations. In doing so, we illustrate how international climate governance itself can be characterised in terms of the first part of the essential definition of polycentric governance offered in Chapter 1 – namely one exhibiting multiple governing authorities that function independently from each other and set rules and norms pertaining to climate change. Next, and addressing our second objective, we assess the extent to which the domain of international climate governance exhibits the suggested features of polycentric climate governance outlined in Chapter 1. The conclusions summarise our main findings.

2.2 International Climate Governance by the UNFCCC

Much ink has been spilt by those seeking to describe the evolution of the international climate regime (e.g. Gupta, Reference Gupta2014; Bodansky, Brunnée and Rajamani, Reference Bodansky, Brunnée and Rajamani2017). We certainly know much more about its limitations (e.g. Rayner, Reference Rayner2010; Victor, Reference Keohane and Victor2011) than we did 25 years ago. We now know, for instance, that although countries can set lofty long-term objectives (e.g. the goals to keep global temperature increases to well below 2°C and to pursue efforts to stay below 1.5°C), this does not mean that when combined, the individual targets or pledges for the short and medium term made by countries will fulfil those goals (e.g. Rogelj et al., Reference Rogelj, den Elzen and Höhne2016). We know that differentiation between developed and developing countries has been a recurring challenge for the regime, often resulting in ‘dysfunctional North-South politics’ (Depledge and Yamin, Reference Depledge, Yamin, Helm and Hepburn2009: 443; see also Chapter 18). We know too that although innovations in the regime have been possible, as witnessed for instance by the introduction of market-based mechanisms such as the Clean Development Mechanism, the rules of those mechanisms have had to be carefully designed to prevent countries and private actors from abusing the system (Wara, Reference Wara2007). We also know that even though a compliance mechanism was incorporated into the regime through its Kyoto Protocol, it was not able to induce Canada, a country that was significantly off target and that ultimately withdrew, to comply (Zahar, Reference Zahar2015). We further know that reaching any agreement amongst more than 190 very diverse parties can be incredibly challenging, as was most visibly underscored by the failure to adopt the Copenhagen Accord in 2009. And finally, we have certainly learned how hard it can be to craft a regime that can keep one of the world’s largest greenhouse gas emitters, the United States, fully on board.

Yet these limitations are all too often ascribed to a rather simplistic characterisation of the climate regime as ‘top-down’ and ‘monocentric’. Specifically, the approach adopted by the Kyoto Protocol is often wrongly referred to as a quintessential example of top-down international governance (e.g. Rayner, Reference Rayner2010). Under this model, legally binding targets and timetables are set to achieve a common objective in a coordinated fashion, and targets are backed by a strong system of monitoring and enforcement in the form of reporting, review and a mechanism to address non-compliance (Hare et al., Reference Hare, Stockwell, Flachsland and Oberthür2010: 601).

However, the Protocol never fitted neatly into this ideal type: its legally binding targets and timetables were not imposed ‘from above’, but rather based on what countries were willing to put forward at the time; the Protocol’s common objective of 5.2 per cent greenhouse gas emission reductions between 1990 and 2008–2012 was simply the result of adding up those commitments. Moreover, although the Protocol strengthened the reporting and review system of the UNFCCC and put in place a compliance mechanism, the strength of either mechanism is debatable (Oberthür, Reference Oberthür2014; Zahar, Reference Zahar2015). Conversely, the Copenhagen Accord is often seen as an example of a ‘bottom-up’ approach, characterised by limited or no global coordination, with countries’ efforts based on what they are willing to unilaterally commit to, with no strong international mechanism to hold them to account (Hare et al., Reference Hare, Stockwell, Flachsland and Oberthür2010: 609). Yet this characterisation is also overdrawn. While the Copenhagen Accord asked countries to make unilateral emission reduction pledges that were not the outcome of multilateral negotiations, the Cancún Agreements anchored the Accord’s pledge-and-review system in the UNFCCC by elaborating the international reporting and review system developed under the Convention.

In short, the climate regime has always been a hybrid of top-down and bottom-up elements, though it is fair to say that elements of bottom-up climate governance – such as non-legally binding pledges – have gradually moved to the fore. The Paris Agreement both exemplifies and formalises this shift, effectively extending it out to the post-2020 period (Bodansky, Reference Bodansky2016). Under the Agreement, countries are no longer subject to legally binding emission reductions as developed countries were under the Kyoto Protocol; instead, the system pins its hopes on a series of procedural obligations and an institutional mechanism to ratchet up national ambitions over time (Bodansky, Reference Bodansky2016; Rajamani, Reference Rajamani2016).

This ambition mechanism is expected to function roughly as follows: (1) a long-term temperature goal (to stay below a temperature increase of 2°C and to pursue efforts to stay below 1.5°C) and a goal of net zero carbon emissions between 2050 and 2100 determine the ‘direction of travel’; (2) countries submit new pledges (known as ‘nationally determined contributions’, or NDCs) in five-yearly cycles; (3) new NDCs will have to go beyond previous ones and have to reflect a country’s highest possible ambition; and (4) countries’ efforts are subject to various types of (periodic) review, including a review of implementation through an ‘enhanced transparency framework’ (see also Chapter 12); a review of compliance through an implementation and compliance mechanism; and a review of overall progress through a five-yearly ‘global stocktake’, starting in 2023. Through an iterative process of submitting and reviewing NDCs, it is hoped, the international community will eventually achieve the Agreement’s long-term objectives.

Like the Copenhagen Accord before it, the Paris Agreement is not purely monocentric. But to what extent can the wider international climate governance architecture be considered polycentric? In a first step towards answering this question, the next section shows that the UNFCCC is not the only multilateral international institution addressing climate change.

2.3 International Climate Governance beyond the UNFCCC

Under the definition of polycentric governance put forward in Chapter 1, multiple centres of decision-making authority govern the same problem. In the domain of international governance, this can be observed in practice, with a variety of international institutions beyond the UNFCCC governing climate change directly and indirectly. To illustrate this diversity, this section reviews the main intergovernmental regimes that have begun to address climate change, looking specifically at international environmental, economic, human rights and sectoral institutions.

2.4 Polycentricity in International Climate Governance

The previous sections show that the domain of international climate governance is characterised by multiple institutions governing the same problems. This section now turns to our second objective. We discuss to what extent the five propositions on implications of polycentricity put forward in Chapter 1 – local action, mutual adjustment, experimentation, building trust and overarching rules – have materialised in the domain of international governance.

2.5 Conclusions

This chapter has shown that the domain of international climate governance displays some of the features of polycentric governance. With reference to the definition outlined in Chapter 1, we can observe multiple decision-making units (i.e. various intergovernmental regimes) that have overlapping jurisdictions and that are not in a hierarchical relationship with each other.

Focusing more specifically on some of the propositions put forward in Chapter 1, there are indications that actions at lower levels of governance are driven by the international level, but we still cannot say to what extent international institutions drive local action compared to ‘local’ drivers (but see Chapter 9), and further understanding is needed of the specific mechanisms through which international governance drives action by non-state and subnational actors. Moreover, actors involved in different international regimes seek to manage areas of overlap through activities that amount to ‘mutual adjustment’, but there is a dearth of research on why mutual adjustment occurs in some cases but not in others. The international climate regime can also be said to be the source of some international governance experiments and, more broadly, be seen as setting the stage for governance experiments at other levels (van Asselt et al., Reference van Asselt, Huitema, Jordan, Turnheim, Kivimaa and Berkhout.2018). The regime may further act as a ‘trust catalyst’ by offering a venue for regular deliberation and establishing a system for reporting and review. However, its trust-building capacity is primarily limited to state-based actions, as its transparency arrangements do not extend to actions by non-state and subnational actors. Finally, while an overarching set of rules can be said to have emerged through the UNFCCC, it has been made first and foremost by states for states. The extent to which there is a set of overarching rules applying to all actors in the system of polycentric climate governance – as well as the contents of those rules – remains an open question.

In conclusion, researchers need to move well beyond the idea that there is or has ever been a single ‘monocentric’ international climate regime. International climate governance emanates from a variety of international regimes, suggesting that this domain in itself is already polycentric. Moreover, as this chapter has shown, the domain of international governance at least partly confirms some of the propositions on polycentric climate governance. What is still needed, however, is a better and more systematic understanding of how exactly international regimes – and the UNFCCC in particular – function in relation to the other domains within the broader polycentric governance system, and where the limits of the suggested positive implications are. A polycentric perspective suggests that existing work conducted by international policy researchers on the linkages within and between other domains be accelerated.

3.1 Introduction

The landscape of climate governance has changed considerably in the past decades. From being dominated by scientists on the Intergovernmental Panel on Climate Change (IPCC) and national governments under the auspices of the United Nations Framework Convention on Climate Change (UNFCCC), climate governance is now populated by actors and institutions ranging from businesses, local governments and civil society organisations, to novel hybrid forms including offsetting standards, emissions registries, carbon-labelling schemes and collaborations between cities (Hoffmann, Reference Hoffmann2011; Bulkeley et al., Reference Bulkeley, Andonova and Backstrand2012; Bulkeley et al., Reference Bulkeley, Andonova and Betsill2014; Hale, Reference Hale2016).

The theory of polycentric governance attempts to explain this dynamic scene by offering a more holistic and inclusive view of climate governance. Chapter 1 identifies three defining features of polycentric governance: (1) it operates at multiple centres of decision-making authority with overlapping jurisdictions, which (2) interact through a process of mutual adjustment and with (3) their interactions generating a regularised pattern of overarching social order. However, some of this scholarship often underappreciates or even entirely neglects the role of the state in polycentric governance. For instance, emphasising the lack of hierarchy in polycentric systems, some scholars suggest that states cannot, or will not, be relied on, because a multitude of other actors will provide alternative mechanisms and solutions (Skelcher, Reference Skelcher2005). It is also argued that engaged and autonomous non-governmental actors can enhance the state’s capacity to deliver (Hooghe and Marks, Reference Hooghe and Marks2003; Newig and Fritsch, Reference Newig and Fritsch2009; Bixler, Reference Bixler2014). The underlying argument is that states are often weak and distant from the societies they govern, and that by providing autonomy to alternative authorities, there is an increase of trust, which in turn improves accountability.

Others acknowledge the importance of the state and of actions taken at the national level. Within a polycentric approach to climate change, Elinor Ostrom asserted that ‘solutions negotiated at a global level, if not backed up by a variety of efforts at national, regional, and local levels … are not guaranteed to work well’ (Ostrom, Reference Ostrom2009: 4). Nation states and their governments are, thus, part of an ‘increasingly diversified structure of climate governance, with its multiple actors’ (Dorsch and Flachsland, Reference Dorsch and Flachsland2017: 47).

We set out to understand the role of the state in an ever-more polycentric setting. Is it simply one of many actors in a non-hierarchical structure, its functions replaceable by those of other actors? Or does it maintain a unique position? As a starting point, we unpack some of the state’s more relevant characteristics. We focus on the functions of states’ domestic governmental institutions performed by their three branches – the legislative, executive and judicial – and on their interactions with subnational governments, individuals and civil society groups at the national level. These particular features of domestic political structures make the state a polycentric actor in itself, acting within a broader, polycentric environment. In specifying the roles of the state in polycentric climate governance, we examine states as a particular polycentric domain, where state institutions and social actors interact, and we focus our attention on how climate change is – and at times is not – scaled down from national to subnational governments, and from governmental to non-governmental spheres.

In terms of their capabilities, states are both rule-makers and rule-enforcers. States with legitimate democratic mandates represent collective interests and have the power to grant and deny other actors their liberties (e.g. by imprisoning them), to collect and distribute money and to regulate financial flows. States are also significant economic actors, with global public expenditure amounting to an average of 17 per cent of gross domestic product (World Bank, 2017). While no other actor in society can challenge the formal political mandate of the state (Peters and Pierre, Reference Peters and Pierre2016), in societies governed by the rule of law, with an independent and impartial judiciary, states can be held accountable for their actions and lack of actions. Moreover, states’ legal frameworks, which are rooted in actions taken in administrative, legislative and judicial settings, are then augmented by rule-making decisions taken by individuals in particular settings (Ostrom, Reference Ostrom2005: 20).

With respect to climate change, states are a central focal point for the implementation of mitigation and adaptation efforts. This role was further acknowledged by the Paris Agreement (2015) when it made states responsible for formulating, reporting and updating their nationally determined contributions (NDCs). States are also the organisations that are expected to implement the policies to give effect to NDCs (Purdon, Reference Purdon2015), and thus promote the changes in societal processes that will allow climate action and sustainable development to move forward. In turn, these actions will influence patterns of consumption and production, encouraging investments in low-carbon technologies, etc. (Boasson, Reference Boasson2015). In a nutshell, states stand out distinctly among the vast number and types of actors in the world of polycentric climate governance.

We approach the participation of the state in polycentric climate governance by focusing on two roles. The first is regulating, defined as ‘the intentional activity of attempting to control, order or influence the behaviour of others’ (Black, Reference Black2002: 19);¹ this role is carried out by the legislative, executive and judiciary branches of government. The second is mobilising others (such as subnational units of the state and non-state actors) to act. Courts carry out both these roles by holding the state and other actors accountable to regulatory frameworks, and by ruling on cases which set norms and directions for all actors to follow.

In highlighting these two roles, we take the position that governing (by national governments) is not opposed to governance (by non-state actors), but rather a fundamental building block that establishes structures and frameworks that interact extensively with other actors in the wider, polycentric climate governance landscape. Overall, we assert that, by providing increased regulation and mobilisation, domestic governmental institutions contribute to – and enhance – polycentric climate governance.

3.2 Regulation: Rule-Making by the Legislative and Executive Branches

Using their capacity to set and enforce rules, in the past two decades, legislatures and administrations have been developing, passing and implementing climate legislation and policies (Lachapelle and Paterson, Reference Lachapelle and Paterson2013). Since the Kyoto Protocol was adopted in 1997, climate legislation and policymaking has been on a steady rise, and the number of climate laws and policies has increased twentyfold, nearly doubling every five years (Nachmany et al., Reference Nachmany, Fankhauser, Setzer and Averchenkova2017b). According to the ‘Climate Change Laws of the World’ database,² by mid-2017 there were approximately 1,300 laws and policies in the 175 countries covered in the database. On average, states have almost eight relevant laws or policies; among the least developed countries the average is fewer than six per country, although they are catching up with the rest post-Paris (Nachmany et al., Reference Nachmany, Abeysinghe, Barakat, Averchenkova, Fankhauser and Nachmany.2017a).³ The rate of adoption of new laws and policies peaked during the period between 2009 and 2013 at approximately 100 per year. The rate dropped to around 40 new laws in 2016, as the existing body of laws already covers substantive ground.

Since the quantity of laws and policies does not necessarily reflect the quality, depth or even the breadth of the climate actions they govern, it is worth noting some of their characteristics.

3.3 Mobilisation: Supporting Action by Non-state Actors

In addition to the formal rule-setting capabilities discussed earlier, states are also suited to create and facilitate non-state action. As Kahler (Reference Kahler2017) argues, states ‘remain prominent governors, setting boundaries and benchmarks as well as engaging as partners with an enlarged and diverse universe of actors’. Similarly, Peter and Pierre’s (Reference Peters and Pierre2016: 5) definition of ‘government’ takes into consideration both the formal structures of the public sector and the set of actors exercising state power (a state-centric conception of governance), as well as the interaction with – and mobilisation of – other actors in society to perform key governance tasks. States, thus, can mobilise or ‘orchestrate’ actions across levels of government as well as across types of actors (Hale, Reference Hale2016; see also Chapters 4 and 11).

The idea of states mobilising non-state action is clearly spelt out in the Paris Agreement (2015). Recognising the polycentric nature of the system, the Paris Agreement acknowledges that climate action cannot and should not be taken by states alone. The Agreement specifies that states will operate in a coordinated manner to enhance public- and private-sector participation in the implementation of the NDCs (Article 6). It also recognises that climate adaptation is a challenge with local, subnational, national, regional and international dimensions, and requires the state to take those into account when formulating adaptation strategies (Article 7).

But mobilisation of non-state actors by the state is a hugely difficult task to perform, not least because of the ‘increasing complexity of society, and the limited effectiveness of traditional policy instruments to shape social behaviour and markets in the desired directions’ (Peters and Pierre, Reference Peters and Pierre2016: 11). As a result, the interaction between state and non-state actors may be complementary (Andonova et al., Reference Andonova, Hale and Roger2017) and reinforcing (Roger, Hale and Andonova, Reference Roger, Hale and Andonova2017), with states addressing weak capacities and low accountability of non-state action (Widerberg and Pattberg, Reference Widerberg and Pattberg2015; see also Chapter 10), yet it can also be contradictory (Cao and Ward, Reference Cao and Ward2017). Acknowledging these difficulties, we now turn to examine how, in a polycentric setting and from a domestic perspective, national governments may mobilise subnational and non-governmental climate action.

3.4 Regulation and Mobilisation: Judicial Law Enforcement and Challenging the State

The three branches of the state – legislative, executive and judicial – interact amongst themselves in multiple ways. A functioning judicial system dedicated to the rule of law contributes to ensuring that the state guarantees civil and political rights (Slaughter, Reference Slaughter1995: 511). In the context of climate change, the courts play a double role, both enforcing existing climate laws and policies and directing action by state and non-state actors. As Peel and Osofsky (Reference Peel and Osofsky2015) argue, litigation is a forum for enforcement and interpretation of the law, as well as a site of potential regulatory development. Used strategically, litigation offers another possible response to inadequate lawmaking activity by governments and also prompts wider policy change. This dual role of the courts in climate litigation – enforcing the law and challenging the state and large emitters – illustrates polycentricity in action within the state.⁶

Climate litigation is a growing phenomenon. In the past years, in many countries, local and regional authorities, businesses, NGOs and individuals have been involved in climate litigation. There have been nearly 700 cases of climate litigation in the United States, and more than 250 court cases across 25 other jurisdictions.⁷ Governments have been the defendants in most of these cases. For instance, in the 25 jurisdictions for which data are available, excluding the United States, 79 per cent of the cases are against governments. Corporations are the second most common defendants (13 per cent of cases). Previous research similarly suggests that in the United States, the government has also been the defendant in the majority of cases relating to climate change (Markell and Ruhl, Reference Markell and Ruhl2012). Out of the 201 cases filed prior to 2010, governments (federal, state and/or municipal) were named as defendants or co-defendants 204 times.⁸ Corporations were defendants in 45 cases.

In some cases, climate litigation aims to drive climate action in countries that lack comprehensive policies or legislation to address climate change. Plaintiffs hope that their claims will fill a governance gap in the short term and spur legislation and regulation in the longer term (Setzer and Bangalore, Reference Setzer, Bangalore, Averchenkova, Fankhauser and Nachmany.2017). A favourable court decision could allow national or subnational governments to regulate greenhouse gas emissions and implement climate policies, even when there is no specific legislation. In the United States, litigation has been driven by the absence of a comprehensive federal legislation that addresses climate change. In this context, court decisions might even replace the need for legislation. For example, the ruling in Massachusetts v. United States (US) Environmental Protection Agency by the Supreme Court in 2007 not only created a legal basis for regulating carbon dioxide emissions but also formed the basis for a bilateral deal with China, and the Obama government’s participation in the Paris Agreement (Carnwath, Reference Carnwath2016).

In other cases, lawsuits are brought to enhance climate action in countries that already have climate regulation in place, and is geared to interpret or enforce existing legislation. An example is the case of Ashgar Leghari v. Federation of Pakistan in 2015, in which the national government was found to have failed to implement its climate policy. Another example is Urgenda Foundation v. Kingdom of the Netherlands; in a 2015 decision, the District Court of The Hague ruled that the Dutch government is required to reduce its emissions by at least 25 per cent by the end of 2020 compared to 1990 levels.

However, the capacity of courts to contribute to effective climate governance should not be overstated. In addition to these cases where climate litigation is brought as a means to strengthen climate action, litigation can equally be used to oppose climate laws and policies, most commonly because such instruments affect private commercial interests (Hilson, Reference Hilson, Fracchia and Occhiena2010). For example, coal companies opposing regulatory emissions reductions have used the courts to challenge clean energy measures. Even if an examination of the outcomes of climate litigation suggests that so far courts are mostly strengthening, rather than hindering, climate regulation (Setzer and Bangalore, Reference Setzer, Bangalore, Averchenkova, Fankhauser and Nachmany.2017), in the lawsuits so far identified in jurisdictions outside of the United States, 40 per cent of the cases were brought by corporations against governments and government agencies, but also against NGOs and individuals.

Viewing litigation as an appropriate site for regulatory development to address climate change is also controversial. Sine argue that strategic climate litigation has been largely political, having no plausible legal basis or chance of success (Zahar, 2015: 24). Courts and tribunals still have to consider whether the law can and should recognise climate change as a problem and respond to it (Fisher, Scotford and Barritt, Reference Fisher, Scotford and Barritt2017: 184). Procedural questions over separation of powers, legal standing, jurisdiction or the scope of permissible review also constitute significant obstacles to cases in many jurisdictions.

Another concern is that the majority of cases taken thus far to courts have not addressed climate policies and legislation or wider emission reductions. Instead, lawsuits have aimed at specific projects (e.g. coal-fired power plants, wind farms or coastal homes), commonly brought under land use and planning laws, or at details regarding the implementation of existing climate policies. As with other climate governance initiatives in polycentric systems, lawsuits dealing with specific projects at the local level have seen more success, while ambitious attempts to promote significant mitigation still constitute the minority of cases. The few examples of successful strategic climate litigation cases are Massachusetts v. US Environmental Protection Agency and the Urgenda case, which push for more aggressive national climate change mitigation policies, and Coalition for Responsible Regulation v. US Environmental Protection Agency and West Virginia v. US Environmental Protection Agency, which challenge the legal bases for US mitigation policy.

But while climate change litigation may not provide the whole answer to the problem of climate change, it is increasingly clear that it will be an important part of the answer (Peel and Osofsky, Reference Peel and Osofsky2015). Despite some limitations, rather than simply a forum for enforcement, courts are a site of potential regulatory development of the law (Peel, Godden and Keenan, Reference Peel, Godden and Keenan2012). New strategic cases brought by NGOs, local authorities and public prosecutors involve a great deal of experimentation. Although so far there are few cases in which the judiciary has improved existing regulatory outcomes, in a polycentric climate governance scenario, courts are likely to continue being used to pressure for future regulatory decision-making to be more responsive to climate change (Peel and Osofsky, Reference Peel and Osofsky2015: 308). As Ostrom (Reference Ostrom2005) acknowledged, the rule of law depends on actions taken by the state, as well as by individuals, and all of these actors are potentially involved in lawsuits dealing with climate change. Climate litigation is a potentially powerful mechanism offered by the state, which allows non-state actors to hold governments to account for insufficient lawmaking, and corporations for current and historical emissions. In addition, instances of strategic litigation that seek to push for more aggressive mitigation policy have been initiated particularly since 2015; this is likely to be a growing trend.⁹

3.5 Conclusions

Through legislative, executive and judiciary branches, national governments remain key actors in the changing climate governance landscape, particularly in the post-Paris period, in which there is an increased reliance on states’ ambitions and on their capacity to establish and implement ambitious policies, mobilising subnational and non-governmental actors.

This chapter has explored the roles of the state in the context of polycentric climate governance, asking if functions performed by the state (a polycentric actor in itself) can coexist with the logic of polycentric governance. At first, it is difficult to envision how the built-in hierarchy fits in the deeply complex and dynamic polycentric setting. Yet, the unique role of states requires some theoretical reconciliation with the logic of non-hierarchical polycentric governance. Without challenging the concept of polycentric approach to climate governance, we claim that states and their governments play a central role, which cannot be filled by any combination of non-state climate activities. National regulation is unique in that it sets rules and a trajectory for other actors. Also, overarching rules can potentially promote effective coordination at the societal level. But this does not imply a hierarchy of importance, as the concerted action of other actors is required more than ever. This is why, in practice, the extent and quality of coordination should remain an empirical question (McGinnis, Reference McGinnis2016), and not part of the basic definition of polycentric governance.

In this context, a polycentric approach to climate governance should be able to accommodate governmental action intertwined with non-governmental, as well as governmental units at different levels, competing and cooperating, interacting and learning from one another (Cole, Reference Cole2015). Nevertheless, effectively implementing rules and mobilising others to action are difficult tasks to perform, and even harder to measure. The challenges to shape social behaviour and markets towards a low-carbon economy are many and varied. With that, state regulation and the mobilisation of subnational and non-governmental actors in future years is likely to encounter only varying degrees of success.

4.1 Introduction

Over the past decade, a key dynamic in climate politics has been the emergence and growth of transnational climate change governance (TCCG) (Abbott, Reference Abbott2012; Bulkeley et al., Reference Bulkeley, Andonova and Betsill2014), which has played an important part in the shift from the monocentric regime established by the United Nations Framework Convention on Climate Change (UNFCCC) to an increasingly polycentric system of climate change governance (Ostrom, Reference Ostrom2010; see also Chapter 2). Transnational governance is typically understood as efforts to authoritatively steer society by a range of actors – including civil society organisations, subnational governments and companies – operating across international borders (Rosenau and Czempiel, Reference Rosenau and Czempiel1992). TCCG takes on many different forms, including carbon-trading mechanisms, labelling and certification schemes, emissions registries, voluntary corporate reporting and urban planning (Andonova, Betsill and Bulkeley, Reference Andonova, Betsill and Bulkeley2009; Hoffmann, Reference Hoffmann2011; Bulkeley et al., Reference Bulkeley, Andonova and Betsill2014; Hale and Roger, Reference Hale and Roger2014). TCCG often includes novel arrangements, techniques, measures and interventions designed to respond to climate change. TCCG initiatives are by definition the product of ‘local’ self-organisation (where ‘local’ is understood to mean action within the context of a particular setting), and they tend to interact both with each other and with other forms of governance such as the UNFCCC and national-level arrangements (Roger, Hale and Andonova, Reference Roger, Hale and Andonova2017) in a process akin to mutual adjustment in polycentricity theory. TCCG has been described as a form of experimental governance (Hoffmann, Reference Hoffmann2011; see also Chapter 6), though the extent to which this is producing learning across the diverse universe of TCCG remains moot.

The emergence of TCCG and its gathering momentum through the 2000s reflected the growing engagement of a diverse array of actors with climate change, the ease of establishing transnational connections and the stalemate within the multilateral climate change regime. In some accounts, its emergence is firmly linked to the deficit of climate governance and leadership at the level of the international regime and the nation state (Roger et al., Reference Roger, Hale and Andonova2017). For others, TCCG as a form of governance innovation has more diffuse causes. These include: broader trends in the fragmentation of authority; diverse motivations amongst those actors who initiated climate governance, including cities, non-governmental organisations (NGOs) and corporate actors; as well as the evolution of the climate issue itself from a singular environmental issue into many diverse realms, including carbon trading, the development of new forms of energy supply, forestry and so forth (Bulkeley et al., Reference Bulkeley, Andonova and Betsill2014). As TCCG has grown, our aggregate knowledge of the scale and scope of TCCG has increased. Significantly, several interrelated databases have been developed to map TCCG initiatives and the subnational and non-state actors that engage in them (Hoffmann, Reference Hoffmann2011; Bulkeley et al., Reference Bulkeley, Andonova and Betsill2014; Widerberg and Pattberg, Reference Widerberg and Pattberg2015; Hsu et al., Reference Hsu, Cheng, Weinfurter, Xu and Yick2016; Roger et al., Reference Roger, Hale and Andonova2017). Drawing on this evidence base, in this chapter we review TCCG and identify its most salient features. With the development of the 2015 Paris Agreement, we see a significant shift in the extent and positioning of TCCG. Rather than remaining a relatively marginal form of climate governance, TCCG has come to be recognised and integrated within the multilateral climate change regime complex (Hale, Reference Hale2016). At the same time, TCCG provides new arenas for contesting what climate governance entails. We detail how TCCG and UNFCCC politics have become increasingly intertwined through the Paris Agreement and suggest that this evolution can be captured through an appreciation of the development of polycentric climate governance as a whole.

Regarding TCCG as part of polycentric climate governance has significant consequences for how we explore the phenomenon and evaluate its impacts and implications. Rather than analysing singular initiatives, it suggests the onus is on understanding the interactions between individual initiatives and the wider governance complex of which they are a part. In the final part of this chapter, we consider three such arenas – clean energy, carbon markets and fossil fuel divestment – and examine the forms of governance innovation that are emerging in the transnational domain. While early forms of TCCG tended to share the same ideological positioning (thus enabling the building of trust across initiatives, a key dynamic in polycentric governance), we find that the transnational arena today is characterised by both centripetal and centrifugal forces. Many TCCG initiatives now explicitly align themselves with goals and frameworks embedded in the UNFCCC regime. At the same time, TCCG is becoming a more contested political domain in which actors challenge those goals and frameworks in search of alternative forms of climate action. In conclusion, we reflect on the implications of our discussion for the development of this area of research and our understanding of polycentric climate governance.

4.2 Constructing TCCG: Experimenting with an Alternative Approach to Governing?

Transnational efforts around climate change in the 1990s predominantly began not as governance efforts but as attempts to influence the state-centric global response to climate change (Newell, Reference Newell2000; Betsill and Corell, Reference Betsill and Corell2008). These actors (NGOs, corporations, regions, provinces, etc.) were actively engaged in the multilateral negotiations and considered themselves either governance takers (having to implement the directives that came from the multilateral process) or governance influencers (seeking to shift the trajectory and substance of multilateral treaty-making). One exception to this picture was found in the work of cities (see Chapter 5), which actively formed networks intended to directly govern climate change.

After the adoption of the Kyoto Protocol, many of these actors shifted towards attempts to engage in transnational climate change governance. In part this was because of what they saw as inadequate progress within the Kyoto Protocol itself (Depledge, Reference Depledge2006), but in part it was because the overarching rules created by the Kyoto Protocol (e.g. emissions trading to support the Kyoto targets) offered an enabling environment for their growth. This was noted first by those examining municipal climate governance efforts through transnational networks (Bulkeley and Betsill, Reference Bulkeley and Betsill2003), but came to be seen as much more widespread (Andonova et al., Reference Andonova, Betsill and Bulkeley2009; Hoffmann, Reference Hoffmann2011). During this period, actors began to experiment with alternative responses to the issue of climate change in ways that cut across traditional divides between actors and scales. Through these efforts, TCCG was becoming an alternative form of global climate governance, independent in many crucial ways from state and multilateral climate governance. By the mid-2000s, there were two coexisting and interrelated realms of the global response to climate change – the multilateral arena and an emergent TCCG arena. Initial efforts at understanding TCCG revealed that it is widespread, but also patterned in particular ways. Indeed, rather than consisting of a random assortment of initiatives only tied together by an externally imposed analytic definition, TCCG – like the broader polycentric climate governance system of which it is a part – displays self-organisation and significant order even though it is not centrally organised.

Three elements of this ordering are particularly prominent – functional, geographical and ideological. Functionally, early studies revealed that TCCG initiatives produced innovative governance arrangements, but the novelty had limits. Hoffmann (Reference Hoffmann2011) uncovered four types of governance prominent in the TCCG world – networking, capacity building, voluntary action and accountable action. Bulkeley et al. (Reference Bulkeley, Andonova and Betsill2014) explored these varied functions in terms of the patterns through which diverse public and private actors institutionalise TCCG initiatives and create authoritative governance arrangements. Furthermore, while TCCG initiatives take on a wide range of climate-related issues, there is clustering around four topics: energy, carbon markets/finance, biodiversity and sequestration and infrastructure (Bulkeley et al., Reference Bulkeley, Andonova and Betsill2014).

The geographical dispersion of TCCG initiatives is also uneven. While actors in the global North have been the dominant initiators of TCCG, this broad generalisation obscures significant regional variation in TCCG activity (Bulkeley et al., Reference Bulkeley, Andonova and Betsill2014). Despite a large proportion (75–90 per cent) of TCCG initiatives aiming to operate in developing countries (UNFCCC, 2016; Chan et al., 2018), developing country–based actors lead only a tiny fraction of these initiatives. Northern-based actors lead 70–90 per cent of initiatives (Widerberg and Pattberg, Reference Widerberg and Pattberg2015; Hsu et al., Reference Hsu, Cheng, Weinfurter, Xu and Yick2016) and 64–84 per cent of participating actors come from developed countries (Galvanizing the Groundswell of Climate Actions, 2015). Actors from Africa and Asia are particularly underrepresented (Bulkeley et al., Reference Bulkeley, Andonova and Betsill2014; Hsu et al., Reference Hsu, Cheng, Weinfurter, Xu and Yick2016). The role of the global South in TCCG remains a key area of ongoing research (Newell and Bulkeley, Reference Newell and Bulkeley2017; Chan et al., Reference Chan, Falkner, Goldberg and van Asselt2018). So far it is unclear whether, as a component of the wider system of polycentric governance, TCCG delivers ‘the achievement of more effective, equitable, and sustainable outcomes’ (Ostrom, Reference Ostrom2010: 552).

Finally, ideological patterns are prominent in terms of the underlying worldview across TCCG initiatives and legitimating discourses. What Bernstein (Reference Bernstein2002) dubs ‘liberal environmentalism’, a notion that sustainability efforts are dependent on or have to be compatible with economic growth, permeates the TCCG world (Bernstein et al., Reference Bernstein, Betsill, Hoffmann and Paterson2010). In addition, TCCG initiatives follow relatively similar strategies of formal or informal institutionalisation to generate the authority to govern in the absence of the more traditional legal authority that state-based governance efforts possess.

As TCCG activities have expanded and academic interest in them has grown, analysis has shifted from examining their emergence, substance and functioning to considering the extent and kinds of impacts that TCCG initiatives have individually and collectively generated. Put simply, do they achieve their objectives? Do they have second-order effects on other actors or on national policies? A number of approaches to assessing the impact of TCCG are now available. Some focus on direct impacts (what individual initiatives accomplish themselves) measured in terms of quantitative emissions reductions goals (Widerberg and Pattberg, Reference Widerberg and Pattberg2015). Others argue for a process-based evaluation (Chan and Pauw, Reference Chan and Pauw2014: 33) like a ‘function-output fit’ approach to assess outputs against stated goals of TCCG initiatives (Chan et al., Reference Chan, van Asselt and Hale2015: 45; see also Chan et al., Reference Chan, Falkner, Goldberg and van Asselt2018). Much of this existing literature, however, focuses on its potential contributions rather than its actual performance and effects. For example, Michaelowa and Michaelowa (Reference Michaelowa and Michaelowa2017) analyse climate partnerships to understand whether they have design features that would allow them to effectively mitigate emissions independently of national policies. One step closer to impact, Chan et al. (Reference Chan, Falkner, Goldberg and van Asselt2018) look at what activities climate partnerships undertake to see if they are producing the kinds of outputs that are likely to lead to impact. Literature on the related area of partnerships for sustainable development suggests that effectiveness may vary considerably across TCCG initiatives (Szulecki, Pattberg and Biermann, Reference Szulecki, Pattberg and Biermann2011; Pattberg et al., Reference Pattberg, Biermann, Chan and Mert2012).

Complementing these attempts to directly measure impacts are proposals to evaluate TCCG initiatives on the basis of indirect impacts – how much they contribute to broader transformations (van der Ven, Bernstein and Hoffmann, Reference van der Ven, Bernstein and Hoffmann2017). This approach considers that the key effects of TCCG initiatives are likely to be catalytic and political – contributing to normative change, building the capacities of political actors and altering coalition-building and conflict dynamics (see Chapter 14) – in addition to, or even instead of, quantifiable emissions reductions. Measuring indirect effects is thus a matter of monitoring the political dynamics that initiatives entail over time (Chan et al., Reference Chan, van Asselt and Hale2015). Evidence suggests that TCCG initiatives are now woven into the fabric of global climate change governance, and interact with United Nations–based multilateral treaties and national government policy systems in important ways (Betsill et al., Reference Betsill, Dubash and Paterson2015) such that they provided an important foundation for the Paris Agreement (Hale, Reference Hale2016). Cao and Ward (Reference Cao and Ward2017) even speculate that growing transnational networks created by TCCG will fundamentally alter the policy preferences of the nation states enmeshed in them. TCCG then – through experimentation, network-building and establishing trust between actors across the climate governance complex – can prepare the ground for the formal recognition and incorporation of the efforts of non-state actors under the umbrella of the multilateral regime. Rather than operating in isolation or in parallel, therefore, we suggest that we should consider the multilateral process and TCCG as part of an evolving polycentric climate governance system. We turn now to considering how this phenomenon has evolved in relation to the shifting multilateral regime and the 2015 Paris Agreement through the formation of a global climate governance complex, before examining specific developments within TCCG since Paris.

4.3 Reforming the Global Climate Governance Complex: Before and after the Paris Agreement

There was a sharp expansion of attention to the role of TCCG activity in the broader regime around the UNFCCC Conference of the Parties (COP) in Paris (Hale, Reference Hale2016). This increase resulted from a variety of factors, including greater mobilisation of civil society, heavier media attention and, critically, the efforts of the United Nations and national governments to ‘orchestrate’ such actions (Hale, Reference Hale2016; see also Chapter 11). This orchestration, and thus the shifting terms of engagement between the multilateral regime and the realm of TCCG, reflects an ongoing process of evolution within the multilateral regime itself. Since the 2009 Copenhagen COP, the climate regime has evolved in interesting and unexpected ways, which has been characterised as a shift from a gridlocked ‘regulatory’ regime to a ‘catalytic’ regime (Falkner, Reference Falkner2016; Hale, Reference Hale2016; Keohane and Oppenheimer, Reference Keohane and Oppenheimer2016). The UNFCCC process has brought climate action from cities, companies, civil society groups and other subnational/non-state actors into its understanding of the ways in which climate change can and should be governed (Figure 4.1).

Figure 4.1 The growing role of sub/non-state climate action in the climate regime.

In September 2014, UN Secretary-General Ban Ki-moon’s Climate Summit brought heads of state together with business leaders, mayors and others to announce bold actions on climate. The Secretary-General’s office had spent months in advance of the summit working to orchestrate multi-stakeholder initiatives on climate change as a way to motivate countries to increase their own ambition (Hale and Roger, Reference Hale and Roger2014). This dynamic was repeated two months later at the High-level Action Day at COP20, held in Lima, Peru, which provided significant impetus to existing TCCG initiatives. It was at this time that the UNFCCC, under the auspices of the Peruvian hosts, created its online Non-state Actor Zone for Climate Action (NAZCA) portal to track climate action by cities, businesses and other subnational/non-state actors. While this portal identifies the action being taken by individual actors, much of what is reported actually takes place in forms of TCCG. In parallel, the United Nations Environment Programme’s (UNEP) Climate Initiatives Platform specifically monitors transnational initiatives. This effort to track and profile subnational/non-state climate activities and TCCG initiatives on an ongoing basis has therefore been central to the attempt to organise and coordinate TCCG in relation to the multilateral regime.

Throughout 2015, the governments of Peru and France, in partnership with the UNFCCC secretariat and the UN Secretary-General, worked to mobilise additional action and initiatives from all sectors of society. This ‘Lima-Paris Action Agenda’, as the programme was called, eventually came to include more than 10,000 individual commitments, many of which were aligned to TCCG initiatives. It was declared a ‘fourth pillar’ of the Paris climate conference (alongside the national pledges, the climate finance package and the negotiated agreement itself), and cited as a critical driver of the successful outcome. Instead of being relegated to the sidelines, local and regional governments, the private sector and other actors were showcased at a series of thematic days throughout the COP, and celebrated in a star-studded Action Day. This conscious effort by international organisations and governments to bring subnational and non-state actors more closely into the process was reflected and augmented by the countries meeting in Paris. In a major departure, governments in Paris instituted the NAZCA portal as an ongoing system to track, support and accelerate subnational/non-state climate action going forward. They appointed two ‘high-level champions’ to catalyse bottom-up climate action. They mandated that a high-level event be held at every future COP for subnational/non-state actors to announce new commitments and report on progress. And they decided to link the ‘Action Agenda’ to the technical process in the negotiations through which countries consider new policy options they might adopt, so that subnational/non-state action can inform national policy and vice versa. These initiatives were further institutionalised at COP22 in Marrakech, through the Marrakech Partnership (UNFCCC, 2016), especially via the formation of a support unit in the UNFCCC secretariat to coordinate the process, bolstered by a hybrid support network envisioned to include a mix of governments, representatives of city and business networks, international organisations and other actors. This new link between the intergovernmental sphere and the subnational and transnational spheres sets, in many ways, a unique precedent in global governance.

4.4 The Dynamics of TCCG Post-Paris

There is, then, an impressive level of activity within the transnational realm now being recorded that could have very significant impacts. Yet the place of TCCG within a broader polycentric climate governance system means that understanding this phenomenon requires moving beyond the analysis of individual initiatives towards an analysis of the ways in which initiatives are interacting both with one another and with other aspects of the climate regime. In short, innovation within the transnational realm can only be evaluated in terms of its position and dynamic within the broader governance landscape. Betsill et al. (Reference Betsill, Dubash and Paterson2015) usefully distinguish between ‘divisions of labour’ and ‘catalytic’ linkages. The former refers to types of interaction where two or more organisations might be attempting to govern a specific aspect of climate change, and the question is whether and how to coordinate their activities to remove unnecessary duplication, avoid contradictions between them, and so on. The latter refers to ways in which two or more governance initiatives may create effects that interact, for example, between the information disclosure from investor initiatives like CDP (formerly, the Carbon Disclosure Project) and carbon price initiatives by governments via carbon markets. Such interactions may then create synergies, realising improvements in climate change responses beyond which each could individually achieve, or of course conflicts, with one undermining the other, with for example some economists arguing that renewable energy targets undermine carbon-pricing initiatives.

Of course, a significant problem of studying TCCG in general, but especially these interactions, is the dynamism of transnational governance. In this section, we discuss three areas where transnational governance has been changing especially rapidly and in which divisions of labour and catalytic links are visible – clean energy, carbon markets and fossil fuel divestment. Each area demonstrates how polycentric climate governance now entails the intertwining of TCCG and the multilateral regime, but that the relationship is not singular. On the contrary, we observe both complementarities and contestation.

4.5 Conclusions

After two decades, TCCG has come to be recognised as a substantive arena of climate governance in both academic and policy circles. If the establishment of TCCG was forged through, and in some senses required its distinction from, the multilateral climate process, the recent history of climate governance dominated by the creation and aftermath of the Paris Agreement has witnessed stronger interactions between these arenas as TCCG becomes both formally recognised and orchestrated by actors within the UNFCCC. However, the ‘inherent messiness’ of these interactions when seen alongside the known deficiencies of polycentricity (Biermann et al., Reference Biermann, Pattberg, van Asselt and Zelli2009; Jordan et al., Reference Jordan, Huitema and Hildén2015) requires further research.

Despite its increased prevalence and profile, it is important to remember that TCCG remains a far from universal phenomenon. The North–South gap in both participation and action implies, for instance, that developing country–based actors do not have a similar impact on the definition of objectives. This might in turn undermine political support for effective engagement of non-state and subnational actors in the UNFCCC, even when they play a crucial role in the provision of additional means necessary to meet targets in NDCs by developing countries. Especially when such imbalances in participation serve to lend weight to Northern framings and initiatives on climate governance, they may encounter opposition from Southern governments to their recognition and inclusion under the umbrella of multilateral climate governance. At the same time, it is important to recognise that the partiality of our picture of TCCG may be a result of how it is defined and observed, and that although our understanding of TCCG has advanced considerably in recent years, mapping and understanding the phenomenon continues to present significant challenges, which in turn creates a need for future work.

This chapter has also identified the issue of evaluating the impact of TCCG as a significant challenge. While existing studies provide helpful information regarding the process through which partnerships might have impact, we need more systematic studies of the actual outcomes and effects of partnerships to fully assess their critical role in global climate governance (van der Ven et al., Reference van der Ven, Bernstein and Hoffmann2017). The most important aspect of this in relation to polycentric climate governance is to think about effectiveness in relation to the interactions across different initiatives. In a polycentric system, there are traditional forms of ‘orchestration’ of interactions between different specific sites or practices of governance (Hale and Roger, Reference Hale and Roger2014), but also forms of mutual adjustment (Ostrom, Reference Ostrom and McGinnis.1999). These interactions and linkages are only beginning to be studied (Hale and Roger, Reference Hale and Roger2014; Betsill et al., Reference Betsill, Dubash and Paterson2015; Hickmann, Reference Hickmann2015; see also Chapter 10).

While we are starting to build a picture of the kinds of TCCG innovation that have emerged in the post-Paris era, our understanding of the ways in which specific forms of TCCG are taking shape remains relatively limited. Cases of clean energy, carbon markets and divestment reveal rather different patterns and forms of interaction between the multilateral regime and TCCG initiatives, revealing both centrifugal and centripetal dynamics. For example, many of the transnational city and business initiatives now frame themselves explicitly as contributing to the Paris Agreement, rather than as alternatives to the UNFCCC process (which was not always the case). They are adopting some of the intergovernmental goals (like the 1.5°C target) and are finding frameworks for coordination through the Marrakech Partnership and elsewhere. At the same time (and to the extent that divestment becomes more widely adopted and starts to have effects on the legitimacy of fossil fuel companies), this suggests that TCCG, and by extension polycentric climate governance as a whole, may be becoming more of a contested field, where the interactions are not only functional but properly political. Divestment arguably constitutes a true ‘innovation’ (Jordan and Huitema, Reference Jordan and Huitema2014) in climate governance through the delegitimation of fossil fuels, such that in some contexts, the burden of proof is on those seeking to argue why we should invest in new fossil fuel infrastructure rather than on those promoting clean energy. The conflicts over pipelines in North America or new-build coal mines in Europe and Australia seem to provide some evidence in favour of this – those promoting pipelines are subject to increasing amounts of scrutiny where the presumption is no longer automatically in favour of the construction of new high-carbon infrastructure. This revival of conflict in climate governance is a reminder that underlying the technical, almost managerial language of polycentric climate governance, as with any similar concept, are deep conflicts of interest and vision at the heart of climate change politics.

5.1 Introduction

Cities and local communities will play a key role in climate change adaptation and mitigation (Bulkeley and Betsill, Reference Bulkeley and Betsill2003; Parnell, Reference Parnell2016; Jayne and Ward, Reference Jayne and Ward2017). Already in Local Agenda 21 (UNCED, 1992), adopted at the Earth Summit in Rio de Janeiro in 1992, they were recognised and explicitly mentioned as an important site for climate action. Fast forward to the mid-2010s: the Climate Summit for Local Leaders was hosted, in parallel to the Paris Conference of the Parties (COP) in 2015. This event was attended by many urban leaders and gained much recognition in the climate negotiations that resulted in the Paris Agreement. At COP22 in Marrakech in 2016, the parallel Climate Summit for Local and Regional Leaders was held. Again, this event provided cities and other local actors with an opportunity to influence international climate change negotiations. Similarly, cities were a central focus of the United Nations’ Sustainable Development Goals of 2015. Meanwhile, the New Urban Agenda resulting from the bi-decennial HABITAT Conference in 2016 has a strong focus on the role of cities in climate change mitigation and adaptation (United Nations, 2016).

When surveying these developments, one might easily assume that cities are already an integral part of international climate governance (see Chapter 4). Unfortunately, the reality is less positive. In international policymaking, cities are not recognised as formal actors – after all, cities are sites as well as actors when it comes to climate action. They still have to break through institutional boundaries to make themselves heard at international climate negotiations and be recognised in international agreements. The side events at the COPs are exactly that – side events, not formal parts of the negotiations – and the Sustainable Development Goals, for example, are not even referenced in the Paris Agreement. Moreover, the Paris Agreement does not explicitly refer to cities, urban geographies or local settlements as actors or sites of governing, but mentions ‘country-driven’ processes as the key principle for organising climate action (United Nations, 2015: Articles 7, 9 and 11). In short, there is much talk at the international level about the importance of urban climate governance, but little is done to empower cities – as actors – taking meaningful action, nor is there much coordination or cohesion between the different international forums engaged with climate change governance in how they envisage the role of cities in climate action.

In response, cities themselves have become involved, as actors, in local and international climate governance interventions, experiments and networks (Hoffmann, Reference Hoffmann2011; van der Heijden, Reference van der Heijden2014; Bulkeley, Castán Broto and Edwards, Reference Bulkeley, Castán Broto and Edwards2015). This is illustrative of polycentric governance – albeit that cities and the networks they form can best be understood as units within a polycentric system rather than a specific domain (cf. Ostrom, Reference Ostrom2010). That is, acting as (partly) independent actors, city governments and other urban leaders have begun to organise themselves around specific urban climate challenges to better understand how these can best be addressed. They do so on regional, national and international scales, following more or less formalised rules. Thus, we see multiple governing authorities acting, as explained in this chapter, at different scales, and exercising considerable independence in making and implementing norms and rules – i.e. matching the essential definition of polycentric governance identified in Chapter 1 (see also Ostrom, Tiebout and Warren, Reference Ostrom, Tiebout and Warren1961; Ostrom, Reference Ostrom1990).

In what follows, three related topics are addressed to better explain the role of cities as units of polycentric climate governance. First, cities often set higher climate governance ambitions than the nation states they are in (Reckien et al., Reference Reckien, Flacke and Dawson2014). What explains this tendency of cities seeking to outperform and thus act independently of national governments? Second, cities are increasingly becoming sites and actors of experimentation with innovative governance instruments, including eco-financing and ‘urban laboratories’ (van der Heijden, Reference van der Heijden2016b). What drives cities to experiment with innovative governance instruments in the first place? Third, cities have begun to break out of traditional top-down, national-regional-local hierarchies and act in trans-local networks (Acuto and Rayner, Reference Acuto and Rayner2016). How do these networks seek to overcome regional and national barriers to climate governance, and what barriers do these networks raise themselves for cities in responding to climate change? Finally, whilst the literature on these three topics – and polycentric urban climate governance more broadly – has expanded rapidly since the early 2000s, it has a strong focus on a relatively small number of cities from the global North (Evans, Karvonen and Raven, Reference Evans, Karvonen and Raven2016). This chapter therefore concludes with a reflection on how applicable it is for all cities in the world – including, crucially, those in the global South. It also identifies what further research is required to understand and support the full potential that cities hold as actors in – and sites of – polycentric climate governance.

5.2 High Ambitions at the Local Level

From the early 2000s onwards, cities have been in a healthy competition to be at the forefront of emission reduction efforts. For example, Sydney aims to cut its emissions by 70 per cent from 2006 levels by 2030, and New York has set itself the goal of reducing its greenhouse gas emissions by 80 per cent below 2005 levels by 2050. What makes the ambitions of these cities – and others like them (C40 Research Team and Arup, 2014) – of particular interest is that they go above and beyond the ambitions set by their respective nation states. Indeed, Sydney and New York’s ambitions are more than double those of their respective countries. Comparing city-level emissions and reduction ambitions with those of nation states is somewhat like comparing apples and oranges (emissions from carbon-intensive sectors such as manufacturing and mining are normally not included in city emissions). Nevertheless, the size of this difference begs a question: why do cities set such ambitious mitigation targets in the first place?

In answer to this question, various reasons are highlighted in the literature. These can be clustered into four main themes: cities as a source and victim of climate change; cities as the low-hanging fruit in climate action; the rise of green growth and ecological modernisation thinking in cities; and national political support for urban climate action.

Starting with the first of these, cities are often considered both a key contributor to and a main victim of climate change. Most resources, including energy, are consumed in cities, and most wastes, including carbon emissions, are produced in cities. This makes cities – and particularly the high consumerist lifestyle that characterises modern urban life – a key contributor to climate change (Dodman, Reference Dodman2009). Because cities are often characterised by high population densities, and because cities represent the geographical epicentre of many economic activities, it will be in cities where climate change–related and other disasters will strike the hardest (IPCC, 2014). Seeking to prevent the devastating effects of such disasters, or simply seeking to save on the costs of operating cities by reducing waste or resource consumption, city governments around the world have implemented myriad regulatory interventions, subsidies and taxes to steer citizens towards more environmentally sustainable forms of living. A typical example is the emergency energy requirements introduced by the government of Tokyo in 2011. These were adopted in response to power shortages experienced from closing down all nuclear power plants after the Fukushima nuclear power plant incident. Whilst these emergency requirements aimed at relieving the electricity net, they had the positive side effect of considerable energy savings (and thus city-related carbon emission reductions), particularly from large offices. Many large office users continued their reduced energy consumption after the emergency requirements were lifted (Nishida, Hua and Okamoto, Reference Nishida, Hua and Okamoto2016).

Second, cities have access to much low-hanging fruit. Of all anthropogenic activities, it is only in constructing, maintaining and using cities (and particularly the built-up part of cities, or simply, buildings) that we see a unique combination of well-trialled, readily available technology and knowledge to achieve emission reductions at net-cost benefit and at a large scale (IPCC, 2014). In many areas – including manufacturing, agriculture and non-city transport – some of these conditions are also present, but not in the same, unique combination. In the United States, for example, possible building-related energy savings of up to 23 per cent are worth double the costs of upfront investments, with a return rate of ten years – $1.2 trillion can be saved if $520 billion is invested (McKinsey, 2009). Some studies even go so far as to forecast that fully carbon-neutral built environments can be achieved in the United States and China by applying all currently available technologies at a net economic gain (Lovins, Reference Lovins2013). Again, seeking to capitalise on such expected savings, city governments around the globe have been steering their citizens to forms of living that are less carbon-intensive than what is formally required by their national governments.

A third and related argument revolves around the paradigm of green growth or economic modernisation (Dryzek, Reference Dryzek2005). It is often argued, and sometimes empirically observed, that cities compete with each other to become the most climate-friendly city, seeking to attract investors and citizens that have a ‘green’ orientation (McCann, Reference McCann2013). The underlying assumption here is that city policymakers are mainly interested in economic prosperity, creating jobs and gaining votes by keeping citizens happy (Schragger, Reference Schragger2016). By creating an image of environmental sustainability and climate action and/or rewarding specific forms of investments, for instance reducing property taxes to encourage more energy-efficient buildings (van der Heijden, Reference van der Heijden2015), authorities seek to attract firms. This in itself can result in job creation. At the same time, creating an image of environmental sustainability and climate action may attract ‘creative’ people that may provide an additional boost to the economic competitiveness of a city (Florida, Reference Florida2005). Such images run the risk, however, of having a merely symbolic function, with cities being unable to live up to some of the high promises they make (Johnson, Toly and Schroeder, Reference Johnson, Toly and Schroeder2015).

A final argument, but one that is sometimes hidden between the lines, is that many cities have set climate change ambitions that are higher than those of the nation states they are in simply because they were actually mandated or supported by national governments to do so (Homsey and Warner, Reference Homsey and Warner2015; van der Heijden, Reference van der Heijden2017; see also Chapter 3). Despite its many flaws, Local Agenda 21 can be credited for recognising cities and their governments as an important level for climate action and addressing other societal problems. Following on from Agenda 21, national governments began requiring, supporting and promoting local action (Bulkeley and Betsill, Reference Bulkeley and Betsill2003; Jayne and Ward, Reference Jayne and Ward2017). Returning to the example of Sydney, in 2011 the Australian government launched the National Urban Policy (Australian Government, 2011). This policy required that all jurisdictions have in place the planning systems to deliver nine specific goals. These include better urban design, more environmentally sensitive new homes and offices and preparations for climate change and natural disasters (Albanese, Reference Albanese2013). Seeking compliance with this policy, Australian states and territories developed long-term regional and metropolitan plans and required cities to draw up strategic development plans and indicate how they were going to meet national requirements. Being exposed to pressure from higher levels of government as well as urban climate mitigation ambitions expressed by other cities resulted in a race to the top between Australian cities to set far-reaching carbon emission reduction ambitions (COAG, 2012). Therefore, even though cities may behave as partly independent actors in polycentric climate governance, the interactions between them and other actors should be borne in mind.

5.3 Experimental Urban Climate Governance and Innovative Governance Instruments

Around the globe, cities have also become highly active in experimenting with novel governance processes and innovative governance instruments to address local and trans-local climate challenges. This ‘experimental governance’ is characterised by iterative rounds of trialling governance instruments within a bounded jurisdiction or population, with the ambition to adapt the instruments based on lessons learnt and to ultimately scale it up to a larger jurisdiction or population (Hoffmann, Reference Hoffmann2011; Ansell and Bartenberger, Reference Ansell and Bartenberger2016). Scholars have identified hundreds of urban climate governance experiments ranging from very local ones to some at an international scale (Bai, Roberts and Chen, Reference Bai, Roberts and Chen2010; Bulkeley and Castán Broto, Reference Bulkeley and Castán Broto2013; van der Heijden, Reference van der Heijden2016b). Examples include the Chicago Sustainable Backyards programme that incentivises households to create water-efficient gardens, through to the international Transition Towns Network that provides tools and processes for citizens to take local climate action (van der Heijden, Reference van der Heijden2014). These experiments seek to act on barriers that stand in the way of effective urban climate action. Such barriers may be political or legal (such as the difficulty of mandating retrofits and upgrades for existing parts of cities), financial (such as split incentives between those who pay and those who gain from urban climate action), technological/behavioural (such as a mismatch between sustainable design and sustainable use of cities) and social (such as the risk of negatively affecting disadvantaged groups by requiring costly climate action) (van der Heijden, Reference van der Heijden2017).

The turn to experimental urban climate governance observed since the early 2000s is more than a pragmatic, local government–led approach to solving problems experienced in implementing national requirements (see Chapter 6). Urban climate governance experiments bring together local governments, private actors and civil society actors in formal and structured processes of developing, demonstrating and trialling new forms of authority and governance instruments to address climate challenges at the city level (Bulkeley et al., Reference Bulkeley, Castán Broto and Edwards2015). Scholars are confident about their ability to draw lessons from experiments about what governance interventions work, where and how, and to scale them up or extend them out across the city in question, and even to other cities and countries (Sassen, Reference Sassen, Johnson, Toly and Schroeder2015).

But what drives cities to experiment with innovative governance instruments in the first place? Again, the literature identifies various motivations. A first and somewhat structuralist understanding relates to the privatisation of (local) public service delivery that started in the 1970s (Hodge, Reference Hodge2000; van der Heijden, Reference van der Heijden2010), the ‘reinventing of government’ and implementation of new public management practices since the 1980s (Osborne and Gaebler, Reference Osborne and Gaebler1992; Hood, Reference Hood1995) and the larger shift from government to governance that has been documented since the late 1990s (Rhodes, Reference Rhodes1996, Reference Rhodes2007). City governments are no longer considered the executive branch of national governments, merely implementing national legislation and regulation (Pierre, Reference Pierre2011). They are increasingly expected to deliver local services themselves (or have local services delivered by others) in an effective and efficient manner, and have to be transparent about their actions and be fully accountable for these – for instance through ‘smart city’ rankings and urban climate indexes (López-Ruiz, Alfaro-Navarro and Nevado-Peña, Reference López-Ruiz, Alfaro-Navarro and Nevado-Peña2014). Facing these increasing expectations – and often assuming that satisfying them aids local economic development (an expectation that is not always based on sound evidence; see Schragger, Reference Schragger2016) – local governments then have little choice but to reach out to local private and civil society actors and search for innovative governance instruments. This is even more the case in a policy area like climate change, where city governments lack experience or prior knowledge about which interventions yield the most desirable outcomes.

Another literature assigns more agency to local governments, private and civil society actors. Rather than considering changing institutional and other structural conditions as forces that tie them together, it considers that all governments wish to be actively involved in addressing urban climate challenges in collaborative processes and experiments (Bingham, Reference Bingham2006; Hohn and Neuer, Reference Hohn and Neuer2006). This branch of the urban climate governance literature has very high hopes and expectations for the outcomes of these experiments (see Chapter 6). By involving a wide range of stakeholders in the development of governance instruments, their tacit knowledge can be used. This is expected to result in instruments that are ‘smarter’ than those developed by somewhat distant bureaucrats (Lobel, Reference Lobel and Levi-Faur2012). Also, by involving a range of stakeholders, instruments can be developed through a consensus-building process that allows for deeper reflection on the advantages and disadvantages of the instrument for the various parties involved. This is expected to bridge their diverse and sometimes competing views (Bulkeley and Mol, Reference Bulkeley and Mol2003). It is further expected to increase the acceptance of the instruments that are developed and implemented and, correspondingly, to improve compliance with them (Walters, Reference Walters2004). In terms of the design of the new governance instruments, scholars have focused on the move away from traditional deterrence-based, hard-law instruments that penalise non-compliance, such as building codes, to soft-law instruments that reward compliance and provide positive incentives. Such positive incentives come, for example, in the form of information, the ability to advertise compliant behaviour or some form of financial compensation (van der Heijden, Reference van der Heijden2016a). Scholars further point to a move away from mandatory governance instruments towards those that ask for voluntary commitments, again assuming that compliance is more likely when individuals and firms commit voluntarily to them (van der Heijden, Reference van der Heijden2014).

That being said, an emerging body of more empirically informed literature is rather more critical of the ability of cities to actually deliver on these normative expectations. It highlights that there is often a normative assumption in the urban governance literature that all experimentation is beneficial, and that whilst there is much talk about experiments and innovative instruments, their development and day-to-day performance are poorly understood (Johnson et al., Reference Johnson, Toly and Schroeder2015). The small empirical knowledge base highlights that challenges abound, and are particularly found when it comes to scaling-up and scaling-out experiments. For example, rules and regulations may lag behind to formalise experiments into urban policy, economic conditions and finance may work against scaling or the experimental setting may not fully reflect the real-world setting an instrument has to operate in (Bulkeley, Reference Bulkeley2013; Schroeder, Burch and Rayner, Reference Schroeder, Burch and Rayner2013). A specific risk associated with urban climate governance experiments is that they target frontrunners and not the majority of firms and citizens. Hence, there tends to be a mismatch between what climate action frontrunners can achieve and what ‘ordinary’ firms and citizens are willing to accept and are capable of delivering (van der Heijden, Reference van der Heijden2017).

In short, experiments are a popular focus for researchers and practitioners, but whether they will be successful in delivering governance instruments capable of quickly reducing carbon emissions and resource consumption at the city level remains an open question. In fact, many experiments have been found to result only in rather piecemeal solutions at best. Moreover, cities that are considered leading and lauded for their example-setting roles often are among the ones with the biggest environmental footprints (Johnson et al., Reference Johnson, Toly and Schroeder2015). More problematically, urban climate experimentation is sometimes used to justify a neo-liberal development agenda and not an especially environmentally or socially sustainable one at that (Evans et al., Reference Evans, Karvonen and Raven2016). For example, it is highly laudable that certain multinationals are collaborating with cities to experiment with new information technology solutions to reduce vehicle emissions or city-related energy consumption – so called smart cities. But questions need to be asked about whether they do so out of altruistic motivations or whether they see this as pilot projects for creating new markets for their products (van der Heijden, Reference van der Heijden2014). Of course, both could in principle be true – hence the desirability of assessing the performance of climate governance experiments against multiple criteria (see Chapter 14).

5.4 Trans-local Collaborations

Yet another manifestation of polycentric urban climate governance can be found in the ongoing growth of trans-local or city-to-city networks, as well as a growth of city-to-citizen and city-to-business networks (van der Heijden, Reference van der Heijden2016b). Whilst city networks, city collaborations, sister-city agreements and so on are not a fully novel development, the active networking of cities in the area of climate action stands out from earlier, somewhat more passive initiatives (Jayne and Ward, Reference Jayne and Ward2017). These active networks are important but informal bodies at trans-local and international levels, comprising formal bodies at the local level (Jordan and Turnpenny, Reference Jordan and Turnpenny2015). They allow cities to learn from each other, jointly experiment and seek governance solutions to urban climate problems and, perhaps most important, to bypass their national governments in the international arena. Three well-known city networks are ICLEI – Local Governments for Sustainability (originally the International Council for Local Environmental Initiatives), the C40 Cities Climate Leadership Group and the Covenant of Mayors for Climate and Energy. The first is an international network of more than 1,500 cities, towns and regions founded in 1990; the second is a network of more than 80 of the world’s largest cities founded in 2005; and the third is a network of more than 7,000 local and regional authorities (mostly from European countries) founded in 2008.

To what extent do these trans-local networks help overcome regional and national barriers to climate governance, and what barriers do these networks themselves raise for cities in responding to climate change? Sometimes a distinction is made between ‘first-wave’ and ‘second-wave’ networks. The first attempt made to push cities to act on climate change was made by ICLEI. It strongly focused on trialling and disseminating knowledge about technological solutions for climate mitigation. Following on from this, academics began writing ‘best practice’ books that were often linked to the then-popular notion of green growth and ecological modernisation. The first-wave city networks strongly revolved around creating knowledge for cities by cities (Jayne and Ward, Reference Jayne and Ward2017). C40 and the Covenant of Mayors can be considered ‘second-wave’ city networks. For these second-wave city networks, knowledge creation and dissemination is still important, but they also seek to have the voice of cities included in international climate negotiations (Johnson et al., Reference Johnson, Toly and Schroeder2015; it has been argued that first-wave cities are now engaged in this too). Representatives of ICLEI, C40 and the Covenant of Mayors were, for example, highly active at COP21 and COP22 (see earlier). Such international events allow cities to showcase their best practices, and challenge their nation states and others to go one step further in their commitments to climate action.

There is some evidence that city networks help overcome regional and national barriers to climate governance, including the difficulty of developing and implementing mandatory regulation and the lack of institutional capital in, particularly, smaller municipalities (van der Heijden, Reference van der Heijden2014). Progressive cities in less progressive nations may find like-minded cities in more progressive nations – there is an abundance of information available for members and non-members on the websites of these networks. By combining resources (funds, staff and so on), these networks are, in theory, capable of carrying out more rigorous experiments than cities can achieve on their own (Bansard, Pattberg and Widerberg, Reference Bansard, Pattberg and Widerberg2016). That said, even though such networks are reporting successes, it remains doubtful how valid these statements really are. The quality of data underlying the statements is sometimes questionable, simply because it is exceptionally difficult to measure reductions in carbon emissions or even energy consumption at the city level (Bulkeley, Reference Bulkeley2013). The networks might attract already well-performing cities rather than poor-performing ones and provide an unrepresentatively high willingness of cities to take climate action (van der Heijden, Reference van der Heijden2017). The reported successes might work in one city but not another. Thus, a big challenge for the climate networks is to find a balance between providing very general and very tailored information on governance interventions (Johnson et al., Reference Johnson, Toly and Schroeder2015). Finally, cities may seek to join these networks seeking co-benefits that may not always stem from a genuine concern about climate change. For example, by participating in the networks, cities hope to attract investors, new workers and residents (Brenner, Reference Brenner2004; Jonas, Gibbs and While, Reference Jonas, Gibbs and While2011).

In short, while the urban climate governance literature was initially positive about the opportunities provided by city networks and their potential to spur urban climate action, recently it has taken a more critical turn. Moving beyond questioning the successes reported by these networks, scholars have pointed out that they easily become ‘networks of pioneers for pioneers’ (Kern and Bulkeley, Reference Kern and Bulkeley2009). Rather than being all-inclusive, the networks run the risk of becoming exclusive clubs that only provide benefits (such as knowledge on urban climate action, or being represented in international climate change negotiations) to their members, somewhat at odds with some of the normative assumptions of polycentric theory (see Chapter 1). Others have highlighted that even members of a network do not always have equal access to all the benefits of membership (Lee, Reference Lee2015). For example, cities in the global North may find it easier to bear the costs of sending representatives to networking events than cities in the global South. An issue that has received less attention in the literature thus far is that these networks may produce an illusion of active and successful cities in the area of climate action (van der Heijden, Reference van der Heijden2017). While both ICLEI and C40, for example, proudly advertise the proportion of the global urban population that they affect – 25 and 15 per cent, respectively (C40, n.d.; ICLEI, n.d.) – it could just as well be argued that after three decades, many cities are still not members.

Furthermore, by looking at the urban governance experiments and innovative urban governance instruments that these networks consider illustrative of outstanding performance, it becomes clear that many only deliver quite moderate rather than transformative climate action. For example, the C40 network has an annual awards ceremony, the Climate Change Leadership Awards, to ‘[reward] important, innovative policies and programmes that reduce emissions and improve sustainability’ and to ‘recognize those successes, catalyze ambition, and share lessons with cities around the world’ (C40, n.d.). In 2013, one of these awards was given to 1200 Buildings in Melbourne, a programme that supports property owners in finding finance for building retrofits. At the time that it was awarded for being a ‘world-leading governance innovation for improved urban sustainability’ (C40, n.d.), only a mere five buildings had actually been retrofitted. In 2014, an award was made to the Amsterdam Climate and Investment Fund, a revolving loan fund. This was made to the city of Amsterdam for its ‘leading position in the transition to low-carbon cities’ (C40, n.d.), but at the time it had only issued some five loans, mainly to support highly conventional technological upgrades of buildings (see, for further examples, van der Heijden, Reference van der Heijden2017). If such action is among the best within the member cities, one may wonder what the rest are up to, and whether cities are really being truly challenged by their city networks to take ambitious climate actions.

There are, of course, good reasons for these city-to-city networks to provide their members with exclusive rewards, to put them in the spotlight in the international arena and to create a narrative of climate activity initiated and supported by them. The supply of networks is sufficient – to the extent that some have to compete for members. On a more positive note, showcasing good practice, however marginal, may spur other cities to take action too. But too much promise and too few results could just as easily backfire. For example, whilst ICLEI initially attracted many cities in the United States, substantial numbers have terminated their memberships as a result of changing political ideologies, interest group pressures and declining membership benefits (Krause, Yi and Feicock, Reference Krause, Yi and Feicock2015). It has been observed that some of these networks have over time become increasingly neo-liberal, seeking to expand and hold their membership base. Rather than a race to the top, there is a risk of a race to the bottom, in which the survival of the network becomes an end in itself (cf. Johnson et al., Reference Johnson, Toly and Schroeder2015). Put differently, the (dominant) city networks may even become a victim of their own success. With a growing membership base came a need to professionalise and formalise, but with the move from being voluntary networks of cities to being large and powerful city interest groups came all the problems that are typically found in bureaucratic organisations – turf wars, a focus on quantity over quality and managerialism (see further Wilson, Reference Wilson1989). That said, absent a benchmark of what constitutes ‘good performance’, it may be difficult for cities and their networks to spur city-level action that is genuinely transformative. Without that, it is also very difficult to assess the efficacy of cities as units in systems of polycentric governance.

5.5 Conclusions

This chapter has addressed polycentric urban climate governance in action. When stepping back and reflecting on the various examples and forms of polycentric urban climate governance discussed, a number of observations stand out. First, city governments often set higher climate action ambitions than the nation states they are in. Second, cities are very active in experimenting with novel climate governance instruments. In doing so, cities self-organise active collaborations with private and civil society actors. Third, cities participate in trans-local and often international networks to develop and share information about urban climate mitigation and adaptation, and seek to influence international climate negotiations. Policymakers, practitioners and academics often express high hopes for city governments and other urban leaders in addressing climate change. The forms and examples of (polycentric) urban climate governance discussed in this chapter are repeatedly used to argue that it will be cities rather than nation states that will take the most meaningful climate actions in the future (Barber, Reference Barber2013; Sassen, Reference Sassen, Johnson, Toly and Schroeder2015; Knieling, Reference Knieling2016). One could frame it even more dramatically than this, as did the former UN Secretary-General, quoted in the epigram of this chapter.

But how well-founded is Ban Ki-moon’s trust in the capacity of cities (including local governments and private and civil society actors) to take meaningful climate action? In line with other critical scholars (Johnson et al., Reference Johnson, Toly and Schroeder2015), this chapter urges some caution when considering cities ‘the key to addressing the global climate change problem’ (C40, n.d.; emphasis added). First, some care. The forms and examples of polycentric urban climate governance discussed in this chapter point to a growing reliance on private and other non-governmental actors in collaborative governance processes. Urban governance theory easily assumes that efficiency through such collaborations and democracy go hand in hand (Davies and Imbroscio, Reference Davies and Imbroscio2009). Yet the involvement of private and other non-governmental actors, particularly multinational companies, pushes urban climate governance further towards neo-liberalism and market-based interventions, and further strengthens the focus on technological fixes rather than behavioural change (Johnson et al., Reference Johnson, Toly and Schroeder2015). Also, with cities acting independently of their national governments, national climate policies no longer ensure that all citizens contribute equally to and benefit from climate action. This begs a need for stronger accountability systems (see Chapter 19), involving (perhaps elected) city officials who can represent and look after the interests of all citizens, especially those more vulnerable to its impacts (Pierre, Reference Pierre2011).

Second, some realism. Whilst the polycentric urban climate governance literature is burgeoning, scholars – myself included – only tend to focus on a handful of (leading) cities. More often than not, these are part of the three main, dominant city networks. The more active cities in these networks – the ones, incidentally, that receive the most scholarly attention – tend to be larger cities in the global North. Yet, whilst climate change is on the agenda of some of the larger cities, particularly in the global North, it should be remembered that in many others it is not: ‘[c]limate change simply remains “un-governed” in cities’ (Bulkeley, Reference Bulkeley2013: 104; see also Reckien et al., Reference Reckien, Flacke and Dawson2014). In short, we have substantial knowledge about polycentric urban climate governance in a small number of predominantly large cities in the global North, but we know very little about polycentric urban climate governance in general. Hence, we are not well equipped to determine how far cities are genuinely capable of self-organising (Johnson et al., Reference Johnson, Toly and Schroeder2015) as polycentric theory suggests. This is troubling for two reasons. First, urbanisation is taking place predominantly in the global South, particularly in Asia and Africa. Solutions that are found to ‘work’ in the global North are often found to generate less positive outcomes when transferred to the global South – or even have negative outcomes there (Gupta et al., Reference Gupta, Pfeffer, Verrest and Ros-Tonen2015; van der Heijden, Reference van der Heijden2017). Second, it remains unclear whether the trends visible in large cities are also found in smaller ones (Sassen, Reference Sassen2001). Smaller cities likely face different barriers than larger cities and have fewer capacities than their larger neighbours. Hence, solutions that are found to ‘work’ in larger cities may not easily transpose to smaller ones (Homsey and Warner, Reference Homsey and Warner2015).

Third, some downscaling of expectations. Following on from these points, the evidence base to support claims about the opportunities and constraints of (polycentric) urban climate governance is thin at best – and at worse may be imbued with a great deal of wishful thinking. There is no doubt that city governments and other local leaders (including private and civil society actors) are organising themselves around specific urban climate challenges to better understand how these can be addressed, following more or less formalised rules, and do so independently from national governments. It is particularly hopeful to see highly progressive cities in countries that are very conservative when it comes to taking climate action – for instance those that have (or had initially) not ratified the Kyoto Protocol or the Paris Agreement (Lee, Reference Lee2015). Yet the room cities have for climate action is shaped by the prevailing national, political and legal context (Johnson et al., Reference Johnson, Toly and Schroeder2015). In particular, the national, legal and policy setting hampers what cities can do locally in terms of self- and facilitative governance, limiting the possibilities for self-organisation by cities (Schroeder and Bulkeley, Reference Schroeder and Bulkeley2009; Schragger, Reference Schragger2016). Thus, some of the high hopes that have been expressed about the benefits of polycentric climate governance are not being borne out in practice. This begs the need for a more critical approach to studying polycentric urban climate governance than has hitherto been the case.

To conclude, it is difficult to determine whether polycentric urban climate governance will be ‘the key to addressing the global climate change problem’ (C40, n.d.). It is encouraging that city governments and other urban leaders have begun to reach out to each other, have begun to take climate action that reaches beyond action taken by nation states, and have not been unduly held back by the lack of formal (inter)national recognition. It is troublesome, however, that polycentric urban climate governance has been studied only in a relatively small number of cities, that we have a limited knowledge base about whether it really delivers on its promises, and that we have a very poor understanding of what this approach to governing means in areas with the very highest levels of urbanisation, namely those in the global South, and particularly Asia and Africa.