Non-state actors in hybrid global climate governance: justice, legitimacy, and effectiveness in a post-Paris era

In this article, we outline the multifaceted roles played by non-state actors within the United Nations Framework Convention on Climate Change and place this within the wider landscape of global climate governance. In doing so, we look at both the formation and aftermath of the 2015 Paris Agreement. We argue that the Paris Agreement cements an architecture of hybrid multilateralism that enables and constrains non-state actor participation in global climate governance. We ﬂ esh out the constitutive features of hybrid multilateralism, enumer-ate the multiple positions non-state actors may employ under these conditions, and contend that non-state actors will play an increasingly important role in the post-Paris era. To substantiate these claims, we assess these shifts and ask how non-state actors may affect the legitimacy, justice, and effectiveness of the Paris Agreement.


INTRODUCTION
T he Paris Agreement now stands at the center of efforts by the international community to address the threats associated with climatic change. Within this Agreement-built upon the United Nations Framework Convention on Climate Change (UNFCCC)-non-state actors will play an increasingly important role. The presence and prominence of non-state actors within the Paris Agreement mirrors a broader shift across the international climate governance landscape in which nongovernmental organizations (NGOs), business groups, think tanks, trade unions, private governance arrangements, transnational networks, and substate authorities assume active roles in limiting the negative effects of global warming. 1,2 This conceptualization of non-state actors includes civil society, business, research groups, and substate authorities. We prefer this expansive definition as it fits alongside the definition of nonparty stakeholders employed by the UNFCCC (see also Ref 3).
In this article, we focus on how the Paris Agreement further deepens and complicates the connections between multilateralism and non-state action. It does so by creating an architecture that we call 'hybrid multilateralism' that splices together state and nonstate actors (on the usage of this term, see Ref 4). This hybrid arrangement emerges in the Paris Agreement through the adoption of two different governance traits: state-led action defined and stipulated by the parties through their own nationally determined contributions (NDCs) as well as efforts by the UNFCCC to orchestrate transnational climate efforts. In both instances, non-state actors are formally and informally woven into the Paris Agreement performing a range of different and increasingly important functions. Non-state actors will act as watchdogs of the NDCs enhancing transparency, facilitating the stocktakes, and pressuring for the ratcheting up of NDCs every 5 years. Likewise non-state actors will act as contributors and governing partners through orchestration as they are encouraged by the Agreement 'to scale up their climate actions, and [register] those actions in the Non-State Actor Zone for Climate Action platform'. 5 This process is coalesced further under the Global Climate Action Agenda (GCAA). 6 These hybridized governance traits complicateperhaps even render superfluous-traditional categorizations of 'top-down' and 'bottom-up' initiatives commonplace in the literature and policy practice.
The review proceeds as follows. We commence by outlining the differential roles non-state actors play in the polycentric system of global climate governance and link this to the UNFCCC specifically. Next, we discuss how non-state actors contributed to the formation of the Paris Agreement. We argue that the Paris Agreement establishes a hybrid architecture that amends previous roles and creates new opportunities for non-state actors vis-á-vis states. We flesh out the implications of this complex hybrid architecture by evaluating how non-state actors will contribute to justice, legitimacy, and effectiveness of the Agreement moving forward. These, in our view, are three key outcomes of state-non-state interaction, and they have each generated significant academic debate (on the importance of these criteria for functioning governance systems, see Ref 7; In terms of climate governance see Ref 8). We do so by highlighting three dimensions within each criterion to home in on relevant bodies of scholarship within each of them. For justice, we probe how non-state actors generate agency, gain access, and alter allocations. In terms of legitimacy, we focus on how nonstate actors promote participation, strengthen representation, and foster accountability. Finally, we look at how non-state actors can enhance the effectiveness of the Paris Agreement by enhancing transparency, augmenting compliance, and affecting outcomes. These dimensions enable us to unpack how non-state actor participation will be structured, facilitated, and (possibly) hampered as efforts are made to secure the broad goals of the Paris Agreement. While taking stock of pre-Paris literature, we also offer the first 'meta-review' of the post-Paris literature that has burgeoned over the past year in our discussion of justice, legitimacy, and effectiveness.

GLOBAL (CLIMATE) GOVERNANCE, THE UNFCCC, AND NON-STATE ACTORS
International politics is bewilderingly complex. 9 Although this is true of most issue areas, 10 it is an especially apt descriptor of global climate governance, which today operates as a dense regime complex populated by states, international organizations (IOs), and non-state actors. 11,12 Although states and IOs have long been established features of international affairs, it is only in the post-Cold War era that we have seen a noticeable rise in terms of the number and influence of non-state actors. 13 Accordingly, these agents have begun playing different roles in how governance beyond the state unfolds, especially in the climate realm. To elucidate the general roles of non-state actors in the lead-up to the Paris Agreement, we highlight how these agents have emerged as: co-contributors within formal multilateral negotiations; conductors and players in different orchestration efforts; partners in transnational networks; private governors; and outside protesters.
First, non-state actors now routinely seek to gain access to formal multilateral negotiations and IO activities. 14 Over the past 20 years in particular, we have seen a sharp increase in efforts by non-state actors to insert themselves in different stages of IO policy cycles such as agenda-setting/policy formulation, decision-making, implementation, as well as the monitoring and enforcement of agreements. 14,15 Within the field of climate governance, non-state actors have been increasingly active within the UNFCCC, 16  Second, states and IOs have begun shifting their mode of governance from regulation to orchestration: attempts by multilateral actors to steer the efforts of other state and non-state actors through soft power. 19,20 Although prevalent across issue spaces, the utility of this concept of orchestration was borne out in the lead up to Paris. For example, the United Nations Secretary General (UNSG) organized the 2014 Climate Summit in the lead up to the Lima COP to mobilize non-state commitments 21 and the 2012 High-Level Political Forum on Sustainable Development has been discussed as an orchestration effort. 22 Moreover global cities are increasingly brought within orchestration efforts by the UNFCCC and the World Bank. 23 For an analysis of orchestration within climate governance-notably in terms of aviation regulation-see Ref 24. Third, non-state actors have been centrally involved in climate governance in the form of transnational networks, epistemic communities, public-private partnerships (PPPs), and multi-stakeholder partnerships (on epistemic communities, see Ref 26 and on PPPs and multi-stakeholder efforts, see Ref 27). The number of networks involved in UNFCCC activities has continued to rise over the past two decades.
Fourth, non-state actors have increasingly taken it upon themselves to become governors in climate politics. These private governance arrangements usually take the form of certification schemes and/or global standard-setting. The empirical uptick in private governance is most evident in the sub-fields of forest and marine sustainability as enacted by the Forest Stewardship Council and the Marine Stewardship Council. 28,29 These efforts at private governorship can often be explained in terms of a 'supply' and 'demand' model as private rules are adopted (and diffuse) across the climate regime complex as private actors seek to fill governance gaps. 30 Finally, non-state (civil society) actors are often engaged in activist efforts. This most frequently takes the form of rallies and actions, intended to gain influence through media attention and by disrupting 'politics as usual'. 31 Although protests are common in transnational politics, they have a central place in climate governance as we saw at the Copenhagen COP in 2009 in which civil society turned out en masse to contest the exclusionary practices of negotiations. On December 12, 2009, some 100,000 participants publically resisted the disenfranchisement of certain perspectives in negotiations. 32 The importance of protests continued in the run up to Paris with hundreds of thousands of individuals taking part in Global Climate Marches. 33 Due to the attacks in Paris preceding the COP, protests were banned during the 2 weeks of negotiations. 34 Nevertheless, 'Peoples Marches' for climate change continue to be organized as we enter the post-Paris period. 35 Non-state Actors in the UNFCCC Before discussing the emergence of the Paris Agreement and its hybrid architecture in more depth, we will focus our review on the role of non-state actors within the UNFCCC. We do so by discussing how non-state actors contribute to party delegations, through the constituency system, and through the organization of side events.

Party Delegations
There is no formal rule as to the size or makeup of a party delegation. Practice varies widely, and mostly mirrors the relative wealth and concern for climate change of the respective country. Delegation size has ranged from under a handful (e.g., some African countries) to over a thousand members (e.g., Brazil, US). The resulting capacity gap limits poor countries' negotiating power and makes their participation in each of the many sessions typically running in parallel less effective. The makeup of a delegation is also varied, with some countries sending particularly large representations from business associations (Brazil), local government, and youth (Canada) or science and academia (Russia). There is also a noteworthy increase in the number of non-state members in developing country delegations in the past decade or so (e.g., Gabon and Guyana). 36 At COP17 in 2011, for example, some 70% of delegations included at least one non-state actor representative and 18% of delegations were non-state representatives. 37 The proportion of women in national delegations has been found to experience a modest but consistent growth, albeit varying strongly across countries. It tends to be higher in countries that enjoy a higher level of development and a higher degree of political gender equality. 38 However, the participation of non-state actors in the negotiations can also be hampered or facilitated by informal practices. 39 States hold the right to close the door to meetings at the COP. A common explanation in the literature has been the functional efficiency hypothesis, where states only allow participation when it suits their interest and in particular during the agenda setting stage, while restricting in the more sensitive decision-making stages. When examining the practices under which circumstances non-state actors are allowed to participate, Nasiritousi and Linnér 13 concluded that this hypothesis only partly holds as explanation why some sessions are open to observers while others are held behind closed doors. They found that decisions on whether to open or close the door are influenced by standard operating practices, habits, and routines. In addition, 'states strategically seek to influence decisions on open/closed meetings depending on their individual political preferences on particular issues'. 16 States can actively use lobbying non-state actors to strengthen their own policy preferences. After the overview of observer organizations, the SBI encouraged an opening up and greater involvement of observers at these COP meetings.

Constituency System
To attend the COPs and intersessionals, non-state actors-if not on a party delegation-must be accredited with the UNFCCC. At the Paris climate negotiations, 1109 NGOs descended on Le Bourget bringing over 8000 individual participants. 40 Of the 30,000 actors who attended COP21, around one third were accredited as NGO or media observers. 40 When gaining accreditation, NGOs are asked to join one of nine existing constituency groups. These are the business and industry NGOs and environmental NGOs; local government and municipal authorities; the research and independent NGOs; trade union NGOs; Women and Gender; Youth NGOs; and farmers and agricultural NGOs (Farmers) (which currently still await full accreditation).
The ability of NGOs to attend the COPs and intersessionals-and therefore access different stages of the policy cycle such as agenda setting, policy formulation, and decision-making-requires accreditation. The importance of the constituency and accreditation system will persist into the post-Paris period. As attendance numbers have varied widely over the past 10 yearsrising at Copenhagen, Lima, and Paris-it seems likely that rising attendance will coincide with new submissions of NDCs (2020, 2025, 2030) and global stocktakes (2023, 2028, 2033) as they offer most scope for NGO impact.

Side Events and Exhibition Booths
The COPs and Intersessionals are comprised of formal sessions and what is typically referred to as 'side events.' The latter take place alongside the formal negotiations and are coordinated by the UNFCCC. Both national and observer delegations can apply to hold a side event. Typically, they highlight diverse climate change-related issues and are held in the form of panel discussions. Competition can be fierce. For example, at COP 16 in 2010, 249 side events were held and about 400 applications were submitted. 41 While there are several types of non-state engagements, the official UN side events at the COPs are possibly the most prominent. They provide a forum for interaction among national delegates, intergovernmental and observer organization representatives, and the media. They are considered by non-state actors to be an avenue for exerting influence on the negotiations. 42 A small set of side events is covered daily by a publication called the Earth Negotiations Bulletin on the Side, a service provided by the International Institute of Sustainable Development, usually at a cost, giving them extra visibility. 43 Likewise, non-state actors can apply to organize and run an exhibition booth. These are housed within the conference center and offer a chance for groups to showcase their actions and achievements and draw attention to divergent topics. Both side events and exhibition booths are applied for through the 'Side Events and Exhibits Online Registration System,' and must be approved by an accredited organization with the UNFCCC.

NON-STATE ACTORS AND THE PARIS AGREEMENT: A HYBRID ARCHITECTURE
The Lead-up to Paris The 1992 UNFCCC treaty text makes no reference to either non-state actors or observer organizations. Perhaps it indirectly acknowledged the role of scientists to determine the threshold toward 'dangerous anthropogenic interference with the climate system.' In the main, however, observer organization input into the UNFCCC was first discussed just prior to the Kyoto Conference. 44 The Kyoto Protocol makes reference of nongovernmental bodies (Article 13). Already at COP3, nonparty participants outnumbered Party delegates. The UNFCCC originally followed the so-called convention-protocol approach 45 spearheaded by the ozone regime with little to no formal recognition of non-state actors.
Up until COP15 in Copenhagen, states had tried to negotiate a successor to the Kyoto Protocol that would similarly rest upon a set of top-down targets and timetables. After the stalemate of Copenhagen, the participation of non-state actors in the climate negotiations was far from certain. Some parties questioned the value of inviting observer organization. The UNFCCC initiated a review process of their participation. 16 Yet, out of the reevaluation of international climate regime also grew a heightened interest in new transnational initiatives, where non-state and substate actors were vital to spur novel forms of climate actions and to rejuvenate the UNFCCC agenda. 4 The number and type of nonstate actors involved in international governance expanded significantly in the period between Copenhagen and Paris. The UN negotiations involved an increasing variety of NGOs, trade unions, business, women's and youth organizations, cities and regions, indigenous people communities and different religious groups. These groups assumed a wide range of roles, from idea providers to implementers, from knowledge brokers to watchdogs, from lobbyist to representing marginal voices. As their mandates and importance grew, so did their potentiality for authority in the international climate politics. 30,46,47 At the same time, however, gridlock in Copenhagen meant that states began increasingly looking for a way to build flexibility into a treaty as a means to promote agreement and compliance. A commonly demanded alternative was to shift toward some form of minilateralism (see for instance Ref 48 and for an expanded version see Ref 49): limiting the number of negotiating states to make agreement on climate change more tractable in ways sensitive to the recognition that climate change is a (super) wicked problem. 50 So prominent became this logic that, at one point, David Roberts argued that minilateralism was the 'conventional wisdom' for addressing climate change in developed countries. 51 Although the precise proposal varied between advocates, three different (though overlapping) logics underpin this move. First, an increased bargaining efficiency argument suggests that by reducing the number of negotiating states to those most capable of reducing global emissions, this would supposedly create more preference-overlap between negotiating states and thus foster agreement. Second, a club-based model suggests that a smaller group could craft rules and incentives to encourage climate action and reduce the risk that a small set of oppositional countries can veto a decision. Finally, a legitimacy-based claim emerged that a small group of responsible (in both senses of the word) actors should take the lead on climate action in ways that recognize both the complicity and importance of having major powers engaged in efforts to avert global climate disasters (for an excellent summary of these positions and the relevant proponents of each see Ref 52). Although often unsaid, these minilateral proposals would have restricted non-state actors to formal negotiations in hopes of securing agreement faster. However, among UNFCCC participants in general, the minilateral approaches never gained traction as legitimate replacements of the globally focused multilateral negotiations. 53 Ultimately, then, minilateralism is not the route adopted by the international community (for an assessment of the direction taken from a minilateralist point of view see Ref 54). The Paris Agreement instead opted for what we describe here as a hybrid model: displacing top-down and bottom-up efforts, it rests upon inclusion and voluntary commitments of all states with myriad roles for non-state actors. By the start of the Paris conference 181 parties had stipulated their intended NDCs with a following seven documented during the fortnight of COP21. With these commitments, the Paris Agreement needed only to formalize these positions and use them as a platform for future engagement on mitigation, adaptation, and finance. Despite some last minute hiccups in negotiation, an Agreement was eventually reached and, on November 4, 2016, the Paris Agreement entered into force when 55 countries representing at least 55% of carbon emissions ratified the treaty (for a discussion on the path to Paris see Ref 34). Donald Trump, who has now withdrawn the US from the Paris Agreement, was elected President the day after (on the potential benefits of a US withdrawal, see Ref 55). The showcasing of the many non-state initiatives, such as the under 2 C coalition of regions, cities and companies encompassing about a third of the global emissions, together with China's declaration of taking the leadership helm, became emblematic for the resolve to stick to the Paris Agreement.

A Hybrid Agreement
What shape did the Paris Agreement take, and how does it relate to multilateralism and non-state actors? The emerging literature on the Paris Agreement uses similar nomenclature to describe how the key elements fit together into a comprehensible whole. Johan Rockström et al. suggest that 'the hybrid make-up of the PA is the result of pragmatic political design'. 56 Harro van Asselt et al. argue that 'the hybrid model of international climate policy embodied in the Paris Agreement requires countries to deliver their NDCs and to progressively increase collective and individual efforts over time.' 57 Similarly Dan Bodansky and Elliot Diringer noted in the lead up to Paris that the Durban platform, combined with policy shifts at the Lima and Warsaw COPs, meant that a hybrid model would ensue at Paris. 58 Radoslav Dimitrov argues that the Paris Agreement 'is a hybrid that enshrines both bottom-up and top-down approaches to global climate governance.' 59 Michele Stua characterizes the Paris Agreement as a 'hybrid, holistic, harmonized' model of multilevel climate governance for the supply, demand, and exchange of mitigation outcomes. 60 Finally, Karin Bäckstrand et al. describe the Paris Agreement in terms of a hybrid architecture. 61 While there seems to be convergence on this language, precisely what is 'hybrid' about the Agreement remains underspecified. In order to provide conceptual precision for the literature and undertake a critical review of the promises and pitfall of the Agreement moving forward, we discuss how the Agreement is a hybrid architecture in light of non-state actors (Box 1).
Beyond Top-Down and Bottom-Up: NDCs and Non-state Actors through NDCs and orchestration. This shift unsettles categorizations of top-down and bottom-up activity, as well as issues of legality, voluntarism, and other established concepts in climate governance. Generally speaking, top-down refers to the relative authority of the actors who seek to produce an intended outcome. The more centrally located to the authoritative power, the closer to the top. The concept of topdown in the Kyoto Protocol context often referred to the decisions which should be implemented through command and control. Alternately, bottom-up implementation refers to action originating from the target groups intended to implement the policy. Policy decision, either through statutes, executive orders, or court decisions at both nation-state and substate level, should spur implementation of the general goals of the agreement, in addition to company policies and other non-state initiatives. In the UNFCCC bottom-up draws attention to the usage of voluntary contributions to implement the Protocol or the Framework Convention. Top-down and bottom-up then implies a hierarchical relationship in terms of actions initiated by actors with relative decisionmaking authority, and does not refer specifically to modes of decision making (legal, non-binding etc.). 62 As the locus of authority is changing in the post-Paris climate governance, the top-down/bottom-up distinction becomes harder to establish.
In large measure, this is because the Paris Agreement consolidates the bottom-up approach to target setting that emerged in the aftermath of Copenhagen, where developed countries agreed to provide national mitigation pledges. These pledges are not legally binding, but the framework for monitoring them does have this status. 63 This process entails several commitments. First, in 2018, there will be a facilitative dialogue on mitigation. Second, according to Article 4.9 of the Paris Agreement, from 2020 a new or updated NDC will be required that outlines commitments for the next period. These must be renewed every 5 years, and according to Article 4.3, each submission must build upon the previous to reflect the individual state's 'highest possible ambition, reflecting its common but differentiated responsibilities and respective capabilities, in the light of different national circumstances'. Finally, building on the facilitative dialogue in 2018, from 2023 onward every 5 years there will be a global stocktake on mitigation, adaptation, and finance. This stocktaking-enshrined in Article 14-is designed to track progress of NDC implementation. 3 In addition to these formal procedures, Article 13 also formulates a 'transparency framework' building supposedly on the Kyoto Protocol's model. This framework is supposed to ensure that states begin harmonizing the formulation of their NDCs through the usage of similar metrics and with a common format. The details of this Framework will be negotiated in 2018. 64 Both the transparency framework and the global stocktake will engage non-state actors directly. This will most likely take the form of non-state actors engaging in monitoring, review, and verification (MRV) of state NDCs, and feeding this information in to the global stocktake process as well as the compliance mechanism (Article 15). 65,66 Especially concerning the NDCs of Least Developed Countries-who may lack the means of costly MRV-non-state actors will prove vital.
Although NDCs combine top-down and bottom-up practices in some ways, this binary distinction is not very helpful for understanding the hybridity of international climate governance after Paris. As noted above, legal bindingness is not restricted to top-down governance and bottom-up with voluntarism. After the Trump administration's announcement to withdraw the US as a Party to the Paris Agreement, states such as Hawaii, California, and Colorado, and cities, such as New York, Los Angeles, Atlanta, Salt Lake City, and Pittsburgh,

BOX 1 EXPLAINING THE EMERGENCE OF THE HYBRID ARCHITECTURE
Although we focus on assessing the justice, legitimacy, and effectiveness of the Paris Agreement with respect to non-state actors, it will be equally important for scholars to explain why this model was chosen and how it was enacted. This will likely entail both qualitative work (archival, interviews) and quantitative analysis to explain state preferences. This work should also focus on explaining how non-state actors mattered in bringing this situation to fruition by inserting themselves in policy processes, keeping global warming high on the international agenda, and lobbying domestic governments to work on intergovernmental relations between the COPs (for an initial analysis on the importance of non-state actors in the formation of the Paris Agreement see Michael Jacobs work on civil society pressure 72 and on the importance of domestic preferences in shaping international climate action see the recent special by Charles Roger, Thomas Hale and Liliana Andonova 73 and the associated special issue contributions).
have through executive order or other decisions decided to commit to the Paris Agreement. Even if these commitments are still unfolding, many of them may be seen as legally binding bottom-up initiatives contributing to the Paris Agreement.

The UNFCCC and Orchestration
Beyond the merging of top-down and bottom-upand legal bindingness alongside voluntary contributions with the NDCs-the Paris Agreement is also hybrid in terms UNFCCC orchestration and nonstate actor engagement. 67 The basic premise underpinning orchestration is that states and IOs-lacking hard-and-fast means of securing compliance-seek to mobilize or even catalyze the efforts of third parties in pursuit of some governance goal. This is typically cashed out in terms of an orchestrator-intermediarytarget model: orchestrators enlist intermediary third parties to impact targets in pursuit of some governance goal. Orchestration therefore moves away from principal-agent models of governance and instead relies upon the provision of material or ideational resources as a means of moving intermediaries-and ultimately targets-toward particular actions and goals. For instance, Hale and Roger discuss how both states and IOs can orchestrate sub-and non-state action. The World Bank, to take one case, orchestrated the Global Gas Flaring Reduction Partnership that sets out a number of rules that oil companies and other concerned actors should adopt. This initiative engaged a number of intermediaries-state governments, oil companies, non-state watchdogs, and other IO bureaucraciesand pushed these actors to alter their own or other target behavior (Box 2). 68 Orchestration in the UNFCCC is most clearly exemplified by the Lima-Paris Action Agenda (LPAA) and the related Non-State Actor Zone for Climate Action (NAZCA) Portal. Both of these initiatives were launched by a quartet of actors: the Peruvian Presidency at COP20, the French Presidency of COP21, the Executive Office of the United Nation Secretary-General, and the UNFCCC Secretariat. Since that time, at Marrakech, the LPAA is being rebranded as the GCAA. LPAA, or now the GCAA, was framed as the 'fourth pillar' of the Paris Agreement (alongside national pledges, the financing package, and the negotiated agreement). It showcases non-state climate action across 12 thematic fields. The Paris Agreement solidifies the UNFCCC as orchestrator of the transnational (non-state) actors. It does so by mandating the continuation of NAZCA (para. 118), the new GCAA (para. 117), and the high-level champions (para. 122). Since Marrakesh, there has been a concerted shift by the High-Level Champions to consolidate these orchestration efforts under the NAZCA and especially GCAA label. This has entailed releasing a 'Roadmap for Global Climate Action,' showcasing non-state efforts, and calling for non-state actor contributions in the aforementioned Roadmap. For instance, at the Marrakesh COP in 2017, the High-Level Champions hosted an event designed to discuss the role of nonstate actors in monitoring and tracking NDCs, contributing to technical expert meetings, and participating in mitigation/adaptation efforts. 69 Yet these efforts to orchestrate non-state actors come with their own set of complications. Even though nonparty stakeholders are invited to scale-up their commitments, how will these efforts BOX 2

THE ROOTS AND RELEVANCE OF ORCHESTRATION
Orchestration has become a key concept in the governance of climate change. It is worth noting that orchestration efforts seen in Paris have their roots in the review that the Subsidiary Body of Implementation (SBI) requested after Copenhagen on the future role of observer organizations in the negotiations. While the admittance of the number of observer participants was restricted, the number of organizations continued to rise. The report concluded that the Chairs at the different sessions and negotiation strands should 'make greater use of observer input' 25 (para 23). To this end their participation at the COPs should be facilitated, for example, by host countries finding suitable venues and Parties were encouraged 'to further engage stakeholders at the national level, including information dissemination and consultation' (para 23).
be related to NDCs? Who will be responsible for measuring non-state commitments and ensuring their implementation? Will the high-level champions be neutral promoters of non-state climate action, or partisan actors with their own agenda? The hybrid architecture thus complicates the relationship between multilateralism-states and the UNFCCC Secretariat-and transnational non-state actors. 70

NON-STATE ACTORS IN THE POST-PARIS ERA: JUSTICE, LEGITIMACY, AND EFFECTIVENESS
This architecture of the Paris Agreement will have a major bearing upon the ability of states and non-state actors to tackle the demands wrought by global warming. In order to think about how this hybrid architecture may work in practice, we focus on assessing how non-state actors may contribute in terms of justice, legitimacy, and effectiveness. These are important topics that the post-Paris literature has begun assessing, albeit unsystematically (on the importance of pledge-and-review in a post-Paris context see Ref 71). We discuss these three features, note connections between them, and conclude by suggesting that nonstate actors will need to play many diverse roles if the Paris Agreement will approach its lofty aspirations.

Justice
The hybrid multilateral characteristic of non-state actor engagement in the UNFCCC under the Paris Agreement continues to pose questions of justice and equity when it comes to deciding who gets what, when, and how. Sites of climate governance are crowded, highly contested, and often disjointed from one another, in particular across multiple levels of governance. 74 This has led to conflict between global goals and on-the-ground realities, which is also played out between global justice discourses and local perceptions of justice. 75 Unless the root causes of ongoing inequality between the haves and have-nots, and the developed and developing worlds, are addressed, climate governance post-Paris will continue to result in harm done and in violations to the human rights of local communities. 76,77 How such matters of justice are dealt with (or not) will be a crucial determinant of the effectiveness of the post-Paris climate regime, according to Chuks Okereke and Philip Coventry. 78 While references to climate justice, human rights and equity are included in the Paris Agreement's preamble, they are not elaborated on in its substantive provisions. 5 Addressing such issues is important not only because it is relevant to those who will be most affected by climate change, but also because justice analysis is essential for understanding the dynamics behind political claims, actions, and trade-offs. 79 In this section, we elaborate on key elements of agency, access and allocation in relation to justice post-Paris.

Agency
Agency is concerned with who makes decisions on behalf of whom (e.g., Refs 13,80). It is widely acknowledged that climate governance no longer rests solely with states, if it ever has. 78 Hierarchical forms of governance have been replaced with a more complex polycentrist and plurilateralist world 'order' as non-state actors play increasingly visible and influential roles and top-down/bottom-up modalities are subverted (e.g., Refs 81,82). However, procedural injustices over inadequately representing the views and voices of the (adversely) affected remain a major shortcoming in climate governance, as noted in the section on legitimacy. Many have documented the ways in which reducing emissions from deforestation and forest degradation in developing countries (REDD +) projects have been designed and implemented without due involvement of local stakeholders (e.g., . This also plays out in the (still) limited participation of observer organizations, in particular of the marginalized, during COP proceedings. 77 Many argue that civil society participation must continue to be improved if the democratic legitimacy of environmental governance is to be strengthened. 78 The role and status of observer organizations in the UNFCCC have indeed changed over the years. The Paris Agreement dedicates one of six sections of its preambular text to nonparty stakeholders. The verdict may still be out with regard to the optimal level of non-state participation in intergovernmental climate proceedings post-Paris. Questions, such as to what extent too much civil society participation might contribute to paralysis of negotiations (and a threat to national sovereignty) 37 and whether non-state actors under this system are able to influence decisions more than minimally, 86 in particular when the policy outcomes affect them directly, are still debated.
Formal interaction and discussion between national delegations and observer organizations have in the past not been directly facilitated by the UNFCCC beyond perhaps constituencies making statements in the high-level segment of each COP and special meetings for party-observer exchanges hosted by the UNFCCC Secretariat. This can happen through side events, thematic days and exhibit spaces; yet, effective channels for bringing alternative discourses, such as on justice and the rights of Mother Earth, to the negotiating tables have been limited. 43 Although such discourses might surface in informal discussions such as the side events, discourses of market-based ecological modernization 87 and technocratic rationalization 74 remain dominant in the UNFCCC. Perhaps the 'Multi-stakeholder Dialogue-Local Communities and Indigenous Peoples.' held during SB46 in May 2017, is a departure from this. It is the first UNFCCC meeting cochaired between a party delegate and an observer representative and driven by the desire of Indigenous Peoples to enhance their status from just observer to being granted decision-making rights (similar to Article 5 under the CBD perhaps). Discussions included comments acknowledging Indigenous Peoples not only as stakeholders but as 'right holders' and the need to enhance interconnectedness between different knowledge systems (author's observation and notes from on-site participation).

Access
Next we discuss access, in particular just (or unjust) access to benefits and rights. 80 The Paris Agreement, like other UNFCCC decisions, is driven by technocratic and market-oriented rationales that serve to produce economic efficiencies and market-driven ecological modernization. This paradigm is unsympathetic to alternative approaches, knowledge, values, and experiences, 88 thus erecting further barriers for indigenous groups, for example, to access their benefits and rights under the UNFCCC. 89 Indigenous peoples, with the help of NGOs, succeeded in their outcry during COP14 in Poznan in 2008 against having the plural of peoples dropped from negotiating texts. Its subsequent reinstatement acknowledged their status as peoples who share collective rights and responsibilities under the UN General Assembly resolution on the United Nations Declaration on the Rights of Indigenous Peoples. These collective rights are legally and normatively distinct from individual rights as they reflect different social, custodial and kinship obligations of indigenous peoples. 77 While the UNFCCC and the Kyoto Protocol had made no mention of any marginalized or vulnerable subgroups, referring merely to the vulnerability of states, such as small-island states or Least Developed Countries, the Paris Agreement does make multiple references to such groups deepening the emergence of a hybrid arrangement. Formal acknowledgment of such rights and benefits comes with no guarantee they are always translated into action, however. The REDD+ safeguard provisions are a case in point. They were adopted to ensure that REDD+ activities do no harm to people or the environment. Safeguards that should be promoted and supported when undertaking REDD+ activities include the recognition of knowledge and rights of indigenous peoples and local communities as well as full and effective participation of relevant stakeholders, in particular indigenous peoples and local communities. In practice, however, there are numerous accounts of cases where these have been paid lip service to, at best (e.g., Refs 84,86,90).

Allocation
Finally, we consider the just (or injust) allocation of climate-related responsibilities and financial compensations, including the emerging norm of loss and damage. 91 The principle of common but differentiated responsibilities and respective capabilities, in particular historical responsibility, has perhaps been the most significant underlying principle of the international climate regime. After the long era of rigid differentiation of countries according to their status as an Annex I or non-Annex I country (which featured a number of anomalies), the Paris Agreement has blurred this divide and instead refocused attention on levels of ambition and willingness to act on the part of states. It has granted non-state actors roles in reviewing the ambition of countries' NDCs, thereby blurring the divide between state and nonstate actors also. 92 The just allocation of climate finance, to both states faced with disaster relief and adaptation needs and non-state actors as capable partners in the delivery of relief and adaptation, is another area that will likely see further hybridization of climate governance post-Paris. Having promised the generation of 100 billion USD in annual flows to developing countries in need, the international community is faced with the challenge of mobilizing additional private finance, and bridging public-private investments, as well. 93 All three pillars of Article 2 of the Paris Agreement-mitigation, adaptation, and financewill therefore hardly be implementable unless nonstate actors are part of the various global, national, and local level efforts.

Legitimacy
Legitimacy is a crucial element of any governance system, determining whether actors find rules acceptable and rightful. Will non-state actors improve the legitimacy of the Paris Agreement? If so, how? Here we focus on participation, representation, and accountability as key elements necessary to generate legitimacy of the Paris Agreement in the eyes of different stakeholders (following work in the field of international political theory and international relations, these three values were chosen for their normative importance and susceptibility to rigorous empirical analysis. For an example, see Ref 94).

Participation
Participation is a key element of legitimacy: it ensures that stakeholders are given a say in how rules and regulations are formed. The Paris Agreement is predicated on universal participation of states and broad inclusion of non-state actors across both axes of its hybrid structure. 95 By turning up at COPs in huge numbers as well as lobbying governments at home, non-state actors had an impact on several key elements of the Paris Agreement including the laudable (if problematic) 1.5 target (on why this target is laudable but likely unattainable, see Refs 96,97) as well as the 'loss and damage' provision. 98 There is also an emerging literature on the importance of non-state actors in global climate governance, focusing on how domestic features (i.e., societal cleavages) condition meaningful participation (see for instance Ref 99).
But here we want to highlight a normative concern surrounding participation, especially in terms of orchestration. The efforts of the UNFCCC and states to bring non-state actors within the fold of formal climate governance potentially undermine the contestatory potential of civil society. 100 This criticismrelated to issues of agency above-brings to mind Foucault's notion of governmentality in which actors are conditioned to work in service of governors, thus depleting their critical potential. The literature on non-state participation post-Paris has barely begun to think through these kinds of normative questions. 101 However one recent contribution by van den Ven et al. to the post-Paris literature has begun stressing that the evaluative tools for orchestration by the UNFCCC on orchestration platforms (such as those run by Ecofys) remains too narrow, focusing on limited conceptions of 'value' and thus depriving non-state actors of a broader range of participatory goals. 102 This mirrors problematic forms of knowledge exclusion discussed in the justice section. Given that non-state actors are increasingly brought into the hybrid architecture through monitoring of state NDCs and through orchestration efforts, both scholars and practitioners should bear in mind the importance of maintaining space for authentic deliberation and participation by non-state actors (authentic deliberation implies a regulative ideal important to gauge whether actors exchange justification under conditions of non-coercion. While never fully attainable, governance arrangements can fall closer or further from this ideal, and thus remains a useful device for understanding how agents interact and for providing a normative yardstick to judge crucial element of legitimacy (uncoerced participation, meaningful representation, governmentality, etc.). On the importance of deliberation in climate governance, see Refs 103,104).

Representation
While efforts at orchestration might have the potential for near universal scope, it is of course impossible for all stakeholders to participate directly in formal multilateral negotiations. Within the UNFCCC this puts the question of representation at the forefront (for discussions in the lead-up to Paris see Ref 105). Those who gain accreditation have to 'make present' that which is absent: the views, ideas, and preferences of those they are supposed to represent. As non-state actors have gained increasing influence in climate negotiations (not to mention the impact of non-state actors as private governors or in networks), the legitimacy of representational claims need to be scrutinized. 106 This will be especially important in the post-Paris hybrid context as NGOs, businesses, and local governments will be asked to report on the impact of NDCs. This reporting will necessarily have to take consideration of how actors on the ground are experiencing mitigation and adaptation efforts (on the importance of representation in collective climate governance efforts see Ref 107). Representative efforts will therefore need to be authentic-actually embodying the views of affected stakeholders-if the legitimacy of the Paris Agreement is to be maintained. 108 This follows closely from the questions of agency and access above, in which the lived experience of those peoples 'on the ground' need to be given due consideration. Without authentic (i.e., accurate and faithful) attempts at representation in hybrid multilateralism by non-and substate actors, justice and legitimacy will suffer. 109 Yet evaluating the authenticity of representational efforts is tricky business. And here we see an unfortunate fissure opening up between empirical studies of representation-especially in the climate field-and political theory. Work in political theory now routinely recognizes that representation is a two-way process: not only do representatives reflect the views of those they claim to represent, but they are also essential in 'constructing' those views. 110 This complicates democratic notions of legitimate representation because the representative is no longer purely responsive to the represented, but actually shapes the represented wants. In the field of climate governance, we know of no study that takes this constructivist view of representation seriously. Yet finding ways to determine whether representative claims are authentic will prove vital for legitimacy. 111 This is especially true moving into the post-Paris period in which non-state actors play varied roles, such as monitoring NDCs, lobbying governments at the COPs, and partaking in orchestration efforts as both intermediary and target.

Accountability
While participation and representation are critical, they most directly related to legitimacy when coupled with accountability: participating stakeholders-and their representatives-need to hold those who make decisions accountable. And alongside transparency (discussed below), accountability has become a central theme of discussions surrounding the hybrid Paris Agreement along both axes. 112 On one hand, the implementation of NDCs will require states to be accountable to the commitments they make. This, as we have outlined, will require non-state actors to monitor, verify, and report on implementation efforts-highlighting the linkage between legitimacy efforts and effectiveness. But it also means that nonstate actors will need to actively hold states to account through naming-and-shaming efforts between global stocktakes as well as at COPs (as we see currently with CAN's 'Fossil of the Day' awards). However, this watchdog role will be complicated until a transparency framework is in place and there is some convergence on how states should report NDCs. 113 We should also be wary that this watchdog function increases the likelihood of governmentality-that non-state actors are used in service of government functions.
It is in relation to orchestration, however, that efforts at accountability are more difficult to conceptualize. Orchestration, as a mode of governance, employs soft forms of steering by an orchestrator to mobilize intermediaries in order to influence targets. This lack of principal-agent dynamic muddles accountability relationships: who is setting goals in orchestration efforts? 67 Who is responsible for verifying that rules are implemented to reach those goals? 114 Who is responsible for ensuring that targets actually live up to their commitments? This last question intersects with debates over how accountable non-state actors need to be themselves, and to whom. 115 Just as importantly, though, the proliferation of orchestration efforts by the UNFCCC and other actors runs the risk of undermining accountability even further: if the same actor can pledge a climate effort to multiple orchestration efforts, how do we avoid double-counting? Will states end up counting multiple commitments within their NDCs? The emergence of orchestration efforts such as NAZCA, LPAA, the GCAA as well as those by other IOs (the UN Secretary General, the World Bank, HLPF) and subnational bodies raises serious accountability issues. Several pieces in the post-Paris period have begun noting how a lack of data from these orchestration initiatives will hamper accountability. 116,117 These difficulties will in turn increase the importance of non-state actors in orchestration as watchdogs of these various portals. For instance, the Carbonn Climate Registry, the CDP, the Covenant of Mayors, and many others will monitor and evaluate orchestration efforts. Given that the contributions of nonstate actors through orchestration portals and other cooperative initiatives is often said to be necessary to close an emission gap in the Paris Agreement, maintaining accountability will be vital to the legitimacy and effectiveness of the Paris Agreement. This should entail accountability of both the orchestrator (to set up and maintain portals with quantifiable and comparable information to help avoid double counting) and accountability of targets (to ensure they make efforts to live up to their commitments). Through the GCAA the High-Level Champions have begun to gather state and non-state submissions on how this accountability could be strengthened in practice. Finding the balance between more orchestrated contributions (needed to tackle global warming), on one hand, and the ability to monitor said contributions through accountability mechanisms (needed for effective action) on the other, will prove paramount under hybrid conditions.

Effectiveness
Whether non-state actors' participation enhances or impedes the effectiveness of the Paris agreement is a third crucial element in viewing the significance of non-state activity. Effectiveness of an international agreement can be defined as the function of the ambition and stringency of Parties' commitments in combination with the levels of state's participation and their compliance with what has been agreed. Greater stringency of Parties' obligations increases effectiveness only to the extent it is not hampering participation or compliance to a greater proportion. Likewise, if ambitions are attenuated, greater participation may not improve the effectiveness. 4 For example, in a study of the 1991 Paris Principles on the Design of National Human Rights Institutions, Linos and Pegram show that having noncommittal language weakened efforts in the compliance phase, in particular in authoritarian states. 118 The weak legal character of the Paris Agreement could lack effectiveness in the absence of strong instruments of compliance-one element of the effectiveness identity. On the other hand, the hybrid agreement may have spurred ambition and participation, the other two elements of the identity. In this section, we discuss how the literature has addressed nonstate actor effectiveness in relation to transparency, compliance and outcomes.

Transparency
The Paris Agreement includes a mechanism for scaling up Parties' contributions beyond the first 5-year period starting 2020. The first round of NDCs will be reviewed and new and more ambitious contributions will be encouraged. A facilitative dialogue shall explore possibilities to enhance effort to fulfill the Paris Agreement, including how information that is necessary for clarity, transparency, and understanding countries contributions can be enhanced. This is an area where non-state actors are expected to provide important contributions.
The transparency framework is a core feature of the Paris Agreement. 119 It shall provide a 'clear understanding of climate change actions.' This includes clarity and tracking of progress on both mitigation and adaptation contributions, including 'good practices, priorities, need and gaps' (Article 13.5). More specific roles for non-state engagement are laid out in the context of 'Enhanced action prior to 2020.' Parties are encouraged to cooperate with non-state actors in technical examination processes, which include sharing experiences and suggestions in addition to facilitating implementation. Non-state actors are expected to show-case the large-scale momentum of an on-going change, for example, registering new and updated initiatives on the NAZCA platform. 59,70 The LPAA independent assessment report-Galvanizing the Groundswell of Climate Actionsconcludes that the process has been 'highly effective at mobilizing initiatives, and giving them increased visibility and enhanced recognition.' This foundation, the report argues, ensures that the supportive function of non-state actors will remain after COP21. 'Elements of such supportive environment would include a system to track progress of climate initiatives, an increased focus on underrepresented action areas, and greater recognition to initiatives that are not yet recognized, especially those pertaining to adaptation' (p. 25). 120 Contributing to the transparency framework would be an important contribution to a fundamental part of the Paris Agreement. Based on pre-Paris agreement literature, this is a function that non-state actors can be expected to assume through a range of governance functions, such as shaping rules, principles and norms, providing information, capacity building, mobilizing public engagement, representing public opinion and including marginalized voices, in addition to the evaluation and monitoring of compliance.

Compliance
As part of the transparency framework, the Paris Agreement involves global stocktaking events every 5 years starting in 2023, where progress of compliance is assessed. However, how progress shall be assessed-in particular in relation to the vast flora of non-state contributions-remains a challenge. One example from the lessons from a prominent climate instrument-REDD+-illustrates the implications. Its objectives-to reduce deforestation and forest degradation while safeguarding conservation, sustainable management of forests, and enhancing forest carbon stocks in developing countries-is now enshrined in the Paris Agreement's Article 5. The REDD+ case reveals diverging views of means of effectiveness. Donor countries understood the aim to enhancing transparency about existing actions and finances, thus identifying gaps and needs. The Coalition of Rainforest Nations, on the other hand, maintained the scaling up actions and finance meant delivering direct increase of funding (Gupta et al. conclude that REDD+ as a PPP has delivered in terms of transparency, participation, and knowledge sharing and coordination, but it has not been able to fulfill the official aims of scaling up neither actions nor finance). 121 Differences in views on what contribution to compliance entails is one challenge. Another is how it shall be measured. Chan et al. argue: 'it is not very likely that the organizers of the high-level event would have the capacity to comparatively assess the performance of an extraordinarily large and growing set of non-State actions.' 122 These authors conclude that demonstrating the contribution of non-state actions at the stocktaking events and Technical Examination Processes can be biased in the selection of cases. Thus, the assessment of non-state contribution to compliance at present and its potential for future actions may be skewed. In particular if it remains unclear how progress shall be assessed, what type of non-state contributions are to be included, and which metrics shall be used for these actions. The hybrid architecture of Paris does not provide clear guidance on this challenge, but the next few years of interactions between states and non-state actors will determine what is being measured and how.

Outcomes
The Intergovernmental Panel on Climate Change's fifth assessment report distinguishes between three forms of effectiveness: economic effectiveness, cost effectiveness and environmental effectiveness. All three address the environmental outcomes policies and institutions. Environmental effectiveness refers to 'whether policies achieve intended goals in reducing emissions or other pressures on the environment or in improving measured environmental quality,' whereas economic effectiveness is used to capture 'the impact of policies on the overall economy.' Cost effectiveness, which is included in economic effectiveness refers to 'the principle of attaining a given level of environmental performance at lowest aggregate cost.' 123 The literature on the effectiveness of nonstate participation in climate governance often refers to environmental effectiveness. This could be achieved by contributing to concrete emission reductions. 6,21 Other motives for non-state contributions could be cost effectiveness by taking on monitoring tasks otherwise performed by government agencies or complement the public funding in climate finance to deliver the committed annual 100 US billion 93 or other adaptation finance. 113 Further, they can achieve economic effectiveness, for example, by involving businesses and cities in greening the economy.
Not only are non-state actors being asked to help monitor others, to spur cost effectiveness, from a state perspective, they are also asked to raise finance and implement changes themselves. The Paris Agreement addresses a number of mitigation, adaptation, finance, and sustainable development objectives. So while non-state actions are expected to contribute to the reduction of greenhouse gases, enhancing adaptation, and raising finance, the Paris Agreement also sets out to 'strengthen the global response to the threat of climate change, in the context of sustainable development and efforts to eradicate poverty' (Article 2). 5 In the lead up to Paris, Blok et al. argued that 21 major initiatives involving a wide range of nonstate actors would have the capacity to spur greenhouse-gas emission reductions at the scale of 10 gigatonnes of carbon dioxide equivalent by 2020. Such contributions, including emission reduction of the top 1000 companies, supply-chain emission reductions, actions by green financial institutions, major cities initiative and subnational governments as well as voluntary-offset companies and consumers, would put the world on track of limiting global temperature increase to 2 C above preindustrial levels, according to the authors. 124 The high expectations on non-state actors remain, although some concerns have been raised that relying on non-state initiatives may provide states with an excuse to shy away from responsibility. 59,70 However, there is little systematic research on the outcome performance of organized non-state and subnational contributions to international efforts. Chan et al. analyzed 25 out of 52 climate actions in both developed and developing countries that were launched at a premeeting to Paris: the 2014 UN Climate Summit in New York. It was the first major initiative to formalize the non-state contribution to the Paris process. They assessed to what extent the production of outputs 1 year after were likely to deliver expected social and environmental impacts. The authors concluded that the output performance is higher than expected after 1 year compared to studies of similar actions at the World Summit for Sustainable Development in 2002.
Yet, they found varieties between policy areas. For example, energy actions had been fulfilled to a greater extent than actions for creating resilience. Partly, this may be explained by the actions aiming at enhancing resilience were in fact launched in conjunction to the meeting, whereas energy actions draw on ongoing initiatives and existing competences within participating organizations. 122 Nevertheless, it points at the importance of distinguishing between the different goals of the Paris Agreement when assessing effectiveness of non-state actions. Chan et al. ask if actions related to resilience may take longer time or encounter more obstacles. Also, the effectiveness of initiatives in the energy sector may appear high, whereas it in fact has been initiated long before it was packaged as a new initiative. They also found substantial disparities between performance in developing and developed countries: 'While many actions target low-income and lower-middle-income economies, the implementation gap in these countries remains greater' compared to OECD countries. 122 This highlights a core issue for the means of effectiveness of climate actions. Although there are few systematic empirical studies, they point to a clear connection between high performance and access to financial and organizational capacity. 120 About half of the New York meeting initiatives had sufficient staff or secretariats, budgets, work plans, and monitoring capability. 122 Yet a substantial part lacked the critical capacity, including many of the arrangements that carried on since COP 20. If the lack of the necessary organizational features in many of the nonstate initiatives persists, it may hamper many of the Post-Paris contributions. 120

CONCLUSION
This article has had three objectives. First, we have documented the myriad roles non-state actors have played in global climate governance leading up to, and in the aftermath of, the Paris Agreement. Understanding past roles will be important in assessing both potentialities and progress in non-state actor activity as we move in to the post-Paris period.
Second, we have argued that the Paris Agreement embodies a 'hybrid architecture' that amalgamates together state and non-state interactions. This term is becoming commonplace in the post-Paris literature, and we have provided much-needed specificity by unpacking the hybrid nature of the Agreement across two axes. On one hand, a legal status of binding reviews in the protocol is counterpointed by nonbinding state NDCs. However, these commitments are underpinned by both binding and nonbinding decisions also on substate level. On the other, the orchestration efforts of the UNFCCC are directed at (and often met by) a groundswell of climate decisions as well as implementation efforts by both state and substate actors. Across both axes non-state actors have come, and will continue, to play increasingly important roles. Undertaking a meta-review of the post-Paris literature, we have systematized and documented these changes in terms of how non-state actors will influence the justice, legitimacy, and effectiveness of the Paris Agreement.
Finally, this review leaves us with some important conclusions. While the pre-Paris literature on non-state actors often focused on whether non-state actors could affect intergovernmental relations, the debate will now focus on how-and under what conditions-non-state actors matter. In light of this, the interrelationships between justice, legitimacy, and effectiveness-touched upon in this review-will become more salient. Questions of justice-such as access-bear heavily upon legitimacy. Likewise inclusion in decision making has been shown to increase compliance. Finally, effectiveness is crucial for (some considerations of ) justice by providing the agreed outcome. These are just a few of the ways in which justice, legitimacy, and effectiveness intersect. As the post-Paris literature unfolds, unpacking these trade-offs and symbioses-especially in the complex hybrid architecture ingrained in the Paris Agreement-will be imperative as we seek to meet the ambition goals necessary for tackling climate change.