“We are the state”: Cities’ quest for recognition in climate diplomacy

Abstract

Students of urban studies have long recognized cities as key sites, but also as critical actors, of global governance. However, they are only beginning to explore how cities’ infrastructural power can be translated into political influence and a distinct form of recognition within the exclusive fora of multilateral conferences and UN institutions. This research seeks to answer this question by applying a relational framework of global agency drawn from International Political Sociology. I mobilize event ethnography at climate COP28 in Dubai (2023)—where subnational governments gained unprecedented recognition, namely through the Local Climate Action Summit and the Coalition for High Ambition Multilevel Partnerships—to analyze how the Local Governments and Municipal Authorities (LGMA) constituency to the United Nations Framework Convention on Climate Change, which I define as a subnational community of practice, seeks recognition for itself and its members by deploying “hybrid” diplomatic practices that oscillate between “multilevel”diplomacy and transnational advocacy. By strategically balancing challenges to the centralized structure of multilateral diplomacy with advocacy practices that align with the UN’s binary distinction between “state” and “non-state” actors, the LGMA works to differentiate cities and other subnationals from “civil society” while promoting the redistribution of power and financial resources to the local level. This attempt marks a new phase in the historical mobilization of cities and city networks of great importance for urban studies: a shift from merely seeking “recognition” in global governance to advancing a progressive reform of the multilateral system into a genuinely “multilevel” structure that would empower subnationals as governmental actors and integral components of the “state.”

Introduction

When the United Nations started holding climate talks back in 1992 and for many years afterwards, conventional thinking was that climate change was a challenge that only national governments could solve (…) But city and regional leaders looked at the problem differently.

Michael Bloomberg’s statement on the stage of the Al Waha Theatre at Expo City Dubai is met with applause, marking a historic moment for the Local Governments and Municipal Authorities (LGMA) constituency to the United Nations Framework Convention on Climate Change (UNFCCC). Over two decades of persistent advocacy by urban stakeholders and local leaders (Acuto et al., 2024) have culminated in this moment, as the very first Local Climate Action Summit (LCAS) takes place on the second day of the 28th Conference of the Parties to the UNFCCC (COP28). For the first time, a subnational summit is part of the official agenda of a climate COP, held right in the heart of the UN-governed Blue Zone, where high-level segments and interstate negotiations unfold.

Yet, the LCAS’s importance should not be overstated. Indeed, a week and a half later, after the pomp of the summit had faded and the delegation of mayors funded by Bloomberg Philanthropies had left Dubai, the remaining LGMA members found themselves facing state negotiators who were far less willing to insert references to multilevel climate governance into the final outcome texts of the COP than might have been expected given the unprecedented recognition awarded to cities at the start of the conference. COP28 was therefore a somewhat paradoxical event: cities briefly received exceptional visibility but not formal acknowledgement of their distinct status as political and governmental actors within the UNFCCC process.

The events of November 30 to December 12, 2023, at COP28 reveal a central paradox of multilateral governance: local governmental actors with the capacity and legitimacy to implement global agendas often lack a seat at the table (Martinez, 2023). Despite having established their authority as global actors (Curtis and Acuto, 2018) and climate governors (Bulkeley and Schroeder, 2012; Gordon, 2020) in their own right within various transnational governance initiatives (Bernstein and Hoffmann, 2018) and city networks (Jakobi et al., 2025), local authorities are typically indiscriminately grouped with other non-state actors in global governance fields and therefore excluded from the multilateral diplomatic institutions that underpin the international system stricto sensu. In the United Nations’ legal terminology, they are variously categorized as “non-Party stakeholders” or “NGO constituencies,” alongside civil society NGOs and businesses, which are broad, undifferentiated categories grouping together all actors not recognized as sovereign states under international law.

This research thus proposes shifting the focus from analyzing how city actors and their networks acquire and exercise authority in global climate governance writ large to examining whether and how they can be meaningfully integrated into climate multilateralism and diplomacy. Such integration would entail transcending the current separation between transnational and polycentric governance processes and the multilateral, Party-driven interstate negotiations that continue to determine global policies today. It would also involve representation of local authorities as a constituency distinct from the others: not a private or civil society advocacy group, but local decision makers, co-designers, and co-implementers of climate policy alongside central state authorities: that is, as political and “state” actors in their own right. Such a transition from cities’ recognition as relevant stakeholders, first achieved in the Paris Agreement, to their integration within a reformed, multilevel climate multilateralism was one of the LGMA’s most ambitious objectives going to Dubai. It furthermore poses a direct challenge to one of international orders’ most fundamental institutions: multilateral diplomacy as a centralized practice gatekept by national governments. As Acuto et al. (2023, 2024) have recently noted in their review of the climate field and the broader UN system, the representation of cities in multilateral diplomatic fields and their advocacy strategies toward UN bodies and state Parties remain underexplored topics in the city diplomacy and urban governance literatures—a gap this research aims to fill.

This research hence contributes to a new strand of literature on “globally engaged cities” (Gordon and Ljungkvist, 2022) and “city network secretariats” (Lecavalier and Gordon, 2020) seeking to assert themselves as foreign policy actors and norm entrepreneurs (Jakobi et al., 2025) in the global governance of migration (Stürner-Siovitz, 2022; Rother, this issue) and the sustainable development goals (Martinez, 2023), among other domains. However, while this research has identified the sovereign barrier to cities’ participation in multilateralism, it has devoted less attention to how it might be overcome (but see Martinez, 2023; Stürner-Siovitz, 2022).

This article brings this question to the forefront by highlighting a recent empirical development in cities’ collective global mobilization: the emergence, between 2015 and 2025, of what I call “the subnational community of practice” (Lamontagne, forthcoming), to borrow Emanuel Adler’s seminal concept (Adler et al., 2024). I argue that, in recent years, dozens of practitioners from the most prominent global city networks—working in partnership with UN-Habitat, Bloomberg Philanthropies, the European Committee of the Regions, and regional governments and their networks—have coalesced into an advocacy community that is particularly active and visible in the climate field, but also in other fields of global environmental and sustainability governance, signaling a new era of global urban mobilization aimed at transforming multilateralism into a genuinely “multilevel” structure (GTF, 2024). This study therefore contributes to urban studies by arguing in favor of a shift in our empirical and theoretical conceptualization of global city actorness: from a focus on cities and city network secretariats as discrete actors seeking global recognition to an analysis of the practices through which they collectively pursue not only recognition, but the incremental reform of the very multilateral institutions and diplomatic practices that currently constrain their policy-making agency, both domestically and internationally, as well as the transnational flows constituting them as “spatio-temporal fixes […] or nodal points […] where global connectivity is articulated, managed, materialized” (Keller and Friedrichs, this issue: 6).

Hence, in keeping with this special issue’s relational approach and its focus on the patterns, practices, and impacts of global city agency, I examine how the subnational community of practice positions the LGMA as a constituency relative to other non-Party major groups, the UNFCCC secretariat, and state Parties in a specific governance “situation” (Hofferberth and Lambach, 2023)—that of COP28 Dubai in the climate “field of global governance” (Sending et al., 2015)—by combining practices of (multilevel) diplomacy with practices of transnational advocacy. Relational approaches in international relations (IR) demonstrate that recognition processes unfold over time, both incrementally and at critical junctures: in this vein, I consider COP28 not as case study representative of the broader universe of city diplomacy, but rather as a critical site of observation (Aykut et al., 2024) in which the subnational community’s strategies and the structural and contingent constraints it faces in its quest for multilateral inclusion—and change—were made particularly visible. The present analysis is based on 162 hours of ethnographic fieldwork within the Blue Zone of COP28 in Dubai from November 30 to December 12, 2023, as well as on primary digital sources produced by the LGMA ¹ and 9 semi-structured individual interviews with its members. It also builds on insights gathered over two years of ethnographic research observing the community across the fields of biodiversity, climate, and SDG governance at five United Nations conferences and two subnational global summits. Overall, I demonstrate that by strategically balancing challenges to the centralized structure of multilateral diplomacy with advocacy practices that align with the UN’s distinction between “state” and “non-state” actors, the LGMA simultaneously defers to the established international order while seeking to change the very practices of diplomacy and centralized sovereign governance that support it.

Theoretical framework: Subnational agency and relationalism in IR

How can we assess the progress of local governments’ quest for recognition as political actors in climate diplomacy? The new relationalism literature in IR provides useful analytical tools for understanding how new aspiring global governors emerge as actors in and through their relationships with other global agents and established incumbents of world order: the central argument this special issue aims to drive home.

First, I build on theoretical and empirical insights from the literature in practice theory, which offers tools to capture the fluid and undetermined processes through which “non-state” actors compete for authority and reconfigure authority relationships within multilateral and global fields dominated by sovereign states (Sending et al., 2015). The climate field is thus constituted by hierarchical authority relationships between a multiplicity of actors, all vying to influence its dominant practices and norms. The central stake of the game is the determination of who can legitimately sit at the negotiation table to define global norms and targets for climate mitigation, adaptation, and financing (Hughes, 2015).

Within this global governance field, as in most others, established practices of multilateral diplomacy reinforce the status of sovereign governments as the primary actors in international politics (Pouliot and Thérien, 2018), limiting the range of possible practices and modalities of engagement for all other actors to tightly monitored practices of transnational advocacy (Davies, 2019) or “NGO diplomacy” (Betsill and Corell, 2008). These mechanisms enable “observer” stakeholders of all stripes to engage in the COP process through the “constituency” system: nine stakeholder groups, each with their own formally designated focal point organization, enable environmental NGOs, businesses, women, and other actors—such as the LGMA constituency—to coordinate among themselves to take collective advocacy stands, which they can then communicate to the UNFCCC secretariat, the host country COP Presidency team, and state Parties through pre-approved channels, such as short interventions at the end of COP plenary meetings or written submissions (Kuyper and Bäckstrand, 2016).

Second, I build upon Hofferberth and Lambach’s (2023) relational framework, which shows that non-state actors in world politics primarily rely on recognition from established incumbents within concrete and repeated “situations” of global governance (such as global conferences) to build—and potentially institutionalize—their own status and authority. Recognition, as emphasized by scholars of urban diplomacy like Gordon (2020), is the key mechanism through which authority, and thus agency, is granted to aspiring actors such as cities. Analytically, it is therefore crucial to identify the incumbent actors in the climate field from whom non-state actors in this field, including the LGMA, must seek recognition to improve their status and enhance their political agency. These incumbents notably include the official representatives of state Parties, the host country COP Presidency, and the UNFCCC secretariat.

I study the LGMA’s position in the climate field and the specific “situation” of COP28 by conceptualizing it as a transnational “community of practice” (Adler et al., 2024). This community assembles a geographically and professionally diverse group of stakeholders from various global organizations and domestic jurisdictions: from “urban professionals” and city network representatives to UN-Habitat officers, academics, mayors, governors, city councilors, subnational government civil servants, and even representatives of philanthropies. As I elaborate elsewhere (see Lamontagne, forthcoming), these practitioners have developed a common identity as a group brought together by shared knowledge on “multilevel governance,” “localization,” and “sustainable urbanization” (GCoM, 2022) and bound by a common repertoire of diplomatic and advocacy practices, which they seek to consolidate among their members and disseminate to other communities of practice and fields of governance (or government) they encounter in national, local, and international politics. Importantly, while the existing literature on city diplomacy and actorness emphasizes the agency of global cities and city network secretariats, as well as local leaders’ mobilization within a global “municipal movement” (Martinez, 2023), I draw attention to a growing community of transnational practitioners that has gradually emerged within this movement and across networked local governance efforts for the specific purpose of petitioning the UN and state Parties for international political status, the localization of global finance, and a multilevel reform of multilateral institutions. By theoretically underscoring diplomatic and advocacy practices jointly developed by this community, I signal the beginning of a new phase in the historical evolution of city diplomacy and global actorness: one in which cities and regions, in seeking political recognition as a distinct constituency of “state” actors, are also attempting to break through the barrier that currently separates transnational governance from traditional interstate multilateralism (GTF, 2024). At climate COPs, this joint effort is led by a core team of staff members and local leaders from several key organizations, which coordinate on behalf of the constituency, ² with the staff from ICLEI–Local Governments for Sustainability World Secretariat in Bonn, the formally designated focal point organization of the LGMA at UNFCCC, facilitating the community’s coordination and joint messaging on the ground.

Methodology: Event ethnography at COP28

Methodologically, I take a page out of the urban studies playbook by foregrounding temporal as well as spatial dynamics in my analysis of relational processes at COP28, using political geographers’ “event ethnography” (Koch, 2023), a methodological approach also increasingly used in the study of environmental agreement making in the IR subfield of global environmental governance (Suiseeya and Zanotti, 2023).

Applied to COPs (Aykut et al., 2024), this methodology enables researchers to grasp the temporal and spatial conditions within which micro-level practices and relational dynamics of recognition/non-recognition occur, as well as to show how those practices structure the competition between “state,” “non-state,” and subnational “state” actors to increase not only their influence or authority over the main expected political outcome of each conference—that is, the production of consensually adopted decisions following negotiations over text—but also their overall recognition as political actors within the climate field. As an annually recurring “megaevent” (Aykut et al., 2024) and the climate field’s central decision-making body, the COPs can serve as ideal “sites” for examining how the long-term structural organization of the field interacts with the contingent dynamics of the conference itself, shaping recognition outcomes for stakeholder constituencies in ways that cannot be predicted in advance but can only be traced during and after the fact.

With the convening of the first-ever LCAS within the official COP agenda, COP28 was regarded by the LGMA as a critical milestone in its decades-long pursuit of recognition in the climate field. Accordingly, this analysis treats COP28 as a salient site of observation rather than a “case” in the traditional sense: a strategic lens through which to examine subnational agency dynamics that may also span multiple governance fields beyond climate, and to understand how the interplay of structure, agency, and contingency shaped the subnational community’s practices, strategies, and recognition outcomes both at a particular moment in time and over the longue durée.

Recognition and political influence in the field can be inductively drawn out in a number of ways. A first tangible sign of recognition are verbal acknowledgments or explicit support of a constituency by incumbent actors through speech acts. An event stronger indicator is the influence of an advocacy group over the outcome of the interstate negotiation (Betsill and Corell, 2008: ch. 1), that is their ability to get at least some of their preferred “language” inserted into the negotiation text. Furthermore, as I have just mentioned, recognition processes carry a temporal dimension. In other words, as Hofferberth and Lambach (2023) observe, agency seeking unfolds over time through a succession of circumscribed situations that may accumulate to build recognition—notably by generating “momentum” (Koch, 2023: 6)—or, conversely, result in the breakdown of what remain largely open-ended and non-teleological sequences. I structure my empirical analysis around two temporal sequences at COP28: the first week of the COP (November 30 – December 6), which, as is customary, centered on technical negotiations and side-events; and the second week, when the attention of Parties and other participants was intensely focused on the tense negotiations between national ministers regarding the highly politicized insertion of an explicit reference to fossil fuels in the final text of the first Global Stocktake (GST).

Other important signs of a constituency or group’s social status are the placement of its organized or dedicated events within the COP program, their modalities of access to different rooms within the COP28 venue, and the location of events they have convened or organized within the Blue Zone, all of which are intrinsically connected to the spatialization of social dynamics at climate COPs (see Suiseeya and Zanotti, 2023). Global governance fields materialize within the concrete venues of global conferences like the COPs, which revolve around interstate negotiations and the securitized and exclusive conference rooms and plenary halls in which they occur. At COP28, the Blue Zone venue—set within Expo City Dubai, a 3.5 square-kilometer mixed-use development promoted as a “sustainable” 15-minute city—was organized into two primary social spaces. The first comprised the areas dedicated to interstate negotiations, UNFCCC-mandated events, and COP Presidency agenda events, which were largely held in the conference rooms, plenary halls, and theaters located near the venue’s entrance (which correspond to zones B1, B8, and B2 on the COP28 venue map ³ ). The second encompassed the areas allocated for country, NGO, IGO, and civil society pavilions and events, mostly situated in the two petal-shaped “wings” of the venue (see zones B2 to B7). The ability to convene, attend, or participate in events held in the first space constituted a concrete indicator of inclusion and recognition for any agent or constituency participating in the global climate governance field.

I have attended all of the events described in my empirical analysis in person, but of course the “ethnographic vision” is always but a partial one (Koch, 2023: 4), which can however be complemented by interviews—my research strategy was to follow the LGMA’s agenda of main events and daily internal coordination meetings throughout the COP, all of which were publicly accessible to Blue Zone-accredited observers like myself (I bore a yellow “observer” badge representing my university). However, I was not privy to any bilateral meetings the LGMA held with Party (i.e. national) delegates or the UNFCCC secretariat behind closed doors, so I cannot account for other, more discreet forms of recognition the community of practice may have received from field incumbents away from the public eye. As an non-member of the LGMA constituency and community of practice, I focus my analysis on three key moments of the LGMA’s diplomacy at COP28: the opening ceremony of the LCAS, the second Ministerial Meeting on Urbanization and Climate Change, and the community of practice’s advocacy work during the second week of the COP. ⁴

Analysis: The Local Governments and Municipal Authorities community of practice at COP28

Week 1 (November 30 to December 6, 2023): Cities (finally) take center stage

The Local Climate Action Summit: Cities on the frontlines and center stage

According to the UNFCCC, COP28 “was the biggest of its kind,” with “some 85,000 participants, including more than 150 Heads of State and Government” (UNFCCC, 2024). From the beginning, COP28 President Sultan Al Jaber made multistakeholder partnerships and collaboration between state and non-state actors central to the conference’s messaging. He was praised by many observer constituencies for fostering one of the most “inclusive” and “proactive” presidencies ever witnessed in climate diplomacy (Interview, October 2023, Bonn; Interview, February 2024, online). Sultan Al Jaber’s interest in urbanization and multilevel governance was evident through his professional and personal ties with former New York Mayor and United Nations Special Envoy on Climate Ambition and Solutions Michael Bloomberg, a key figure in global climate advocacy and a founder of the Global Covenant of Mayors for Climate & Energy (GCoM, 2022), a core LGMA member since 2017.

The LCAS was announced by Bloomberg Philanthropies and the COP28 President in September 2023. Bloomberg funded travel and lodging for hundreds of mayors from around the world, creating an unprecedentedly large subnational delegation of over 500 subnational leaders to a climate COP (COP28 UAE, 2023), including a delegation of 12 mayors who were invited to attend the high-level World Climate Action Summit of World Leaders (WCAS) in person. The LCAS was a two-day affair: the opening ceremony was held in the evening of December 1 at the Al Waha Theater (the same venue where the WCAS had taken place an hour prior, in Zone B8), while a series of side-events in the form of panel discussions on urbanization and multilevel action, open to all accredited observers and held in various locations in the wings of Expo City, took place all throughout the day of December 2. The format of the 90-minute opening ceremony was overall very similar to other municipal movement events and city summits I have attended in recent years: after speeches by Antonio Guterres, Sultan Al Jaber, Michael Bloomberg, Xie Zhenhua (China’s Special Envoy for Climate Change), and Maroš Šefčovič (Executive Vice President of the European Commission), two panels with mayors and stakeholders, including Tokyo Governor Yuriko Koike and Paris Mayor Anne Hidalgo, emphasized cities’ role on the “frontlines” of the climate crisis. They argued that cities bear the cost of delayed action and are thus the most motivated to act swiftly. As Claudia López, Mayor of Bogotá, stated: “Centralization of powers at the national level is a recipe for disaster.”

This kind of political discourse has long been central to cities’ efforts for recognition as global climate leaders (Gordon, this issue), but it is typically heard during city network summits or parallel subnational summits organized on the side of UN conferences, not on the main stage of a COP. ⁴ Thus, in the end, the most notable aspect of the LCAS was that it took place at all, and that its opening remarks were delivered by prominent climate field incumbents who are typically very difficult to secure even as attendees at a city-focused event. The Al-Waha had a capacity for 500 attendees, which roughly matched the number of local leaders in Dubai (Interview, February 2024, online): the UNFCCC Secretariat had imposed a last-minute ticketing system, implying high demand and limited access, but in the end most WCAS attendees and national representatives had already left the Theatre by the time the LCAS began, and not all invited subnational guests attended, leaving last-minute tickets available for academic and NGO observers like myself unaffiliated with the LGMA.

When asked a year later during an online interview to name some of the LGMA constituency’s most significant global recognition milestones in recent years, a senior official from one of its core organizations responded without hesitation: the LCAS (Interview, February 2025, online). Nevertheless, the real question is whether the LCAS performance left a lasting impression on other stakeholders and field incumbents beyond the constituency, enough to build momentum toward greater recognition of subnational governments as political actors in global climate governance, as notably reflected in final negotiation texts or country plenary speeches. The answer, as we shall see, is rather mixed.

The second Ministerial Meeting on Urbanization and Climate Change and the Coalition for High Ambition Multilevel Partnerships

The second edition of the Ministerial Meeting on Urbanization and Climate Change (first held at COP27 in Sharm El-Sheikh in 2022) took place at 9:30 am on December 6, co-convened by the COP28 High-Level Champion and UN-Habitat, a UN organization closely tied to the LGMA through its mandate and role in promoting the 2016 New Urban Agenda. Its aim was to “reinforce mechanisms for multi-level climate action to achieve the targets of the Paris Agreement” and “increase and accelerate deployment of climate finance for cities and local governments” (COP28 UAE, 2023). Eight interstate and multistakeholder initiatives, targeting such sectors as waste management and totaling around US$500 million in financial commitments for “city-focused climate investment,” were launched at the Ministerial (COP28 UAE, 2023).

The Ministerial on Urbanization was held as part of the COP28 Presidency agenda, in the Al-Ghafat Plenary Hall, a venue that had hosted numerous interstate plenaries and other high-level events in zone B1. However, despite the LGMA’s requests, the Ministerial is not yet a mandated UNFCCC event and remains outside the official program; its yearly inclusion in the COP agenda is not “mandatory,” meaning it is not secured by a formal COP decision. Although this second edition was the first to produce a set of “recommendations” for Parties, no formal mechanism exists for these to feed into the UNFCCC process or inform Party negotiators; a limitation the LGMA still hoped to address at COP30 in Belém two years later (Interview, April 2024, online; COP30 Belém, fieldwork notes, November 2025).

Unlike the LCAS, the Ministerial was widely accessible to observers of all badge ranks, pending registration on the UN-Habitat website. Incumbent speakers included COP28 High-Level Champion Razan Mubarak, Maimunah Mohd Sharif, Executive Director of UN-Habitat, Diana Urge-Vorsatz, Vice Chair of the Intergovernmental Panel on Climate Change, and Makati (Philippines) Mayor Abigail Binay, speaking for subnational leaders. While less prestigious than the LCAS lineup—which had included the UN Secretary-General and Party representatives from China and the EU—the Ministerial opened on a high note when national ministers were invited on stage for a “family photo” alongside municipal and regional leaders; journalists crowded the stage, cameras clicked, and dozens of attendees stood on chairs or plenary tables to capture the moment.

According to the COP28 UAE December 7 press release (2023), “40 national ministers” and over “280 LGMA representatives” attended the Ministerial. While these numbers may initially appear modest given the 198 Parties to the UNFCCC, they were considered “pretty good,” as one LGMA interviewee noted (Interview, February 2024, online), especially for an event solely dedicated to cities and considering how overbooked ministers are during COPs. The observer section at the back of plenary hall stayed mostly full throughout the meeting, while many Party seats in the front rows were left vacant or saw ministers arriving late and leaving immediately after their remarks; a common occurrence in high-level global meetings.

Following the opening segment, two sessions—one on multilevel climate action and another on localizing climate finance—featured ministers delivering two-minute statements on their countries’ policies on sustainable urbanization and intergovernmental cooperation under the Paris Agreement. Each ministerial speech was immediately followed by a two-minute response from a mayor or governor, personally invited by their minister to offer the local perspective.

This format marked the Ministerial’s most significant innovation: a rare instance of multilevel diplomacy (Criekemans, 2018) within a multilateral setting. As a hybrid diplomatic practice, it disrupts the convention of a single, unified state voice by revealing the internal distribution of power typically concealed on the global stage. That mayors and subnational leaders could sit beside their national ministers at their country’s desk and speak in their own capacity as political actors and heads of government in a COP plenary hall was a striking and symbolically significant precedent. Yet, this hybrid form of diplomacy still deferred to the hierarchy between state and subnational governments by having ministers speak first and decide which subnational leader would follow. Nonetheless, the notion that a state might carry multiple voices—even if aligned, as they were here—represents a new and potentially transformative idea in multilateral diplomacy. If adopted more widely, such practices could help bridge the climate governance “implementation gap by addressing the chronic exclusion of subnational and local governments from the development and implementation of international environmental (and other global) policies (GCoM, 2022).

Another notable subversion of established practice came through the launch of the Coalition for High Ambition Multilevel Partnerships (CHAMP), a COP28 Presidency initiative co-sponsored with Bloomberg Philanthropies and first announced during the LCAS opening on December 1. The CHAMP Pledge invites endorser Parties to “enhance cooperation, where applicable and appropriate, with our subnational governments in the planning, financing, implementation, and monitoring of climate strategies” (LGMA, 2024). What sets CHAMP apart from other Presidency initiatives at COP28 is its ambition not merely to shift state behavior in a specific policy sector, but to prompt central governments to establish long-term partnerships and coordination mechanisms with their local governments—many of which currently do not exist—with the goal of bolstering climate action across all levels and improving national climate plans’ (i.e. the Nationally Determined Contributions, or NDCs) ambition and implementation. This invitation carries potential implications for intergovernmental relations and power-sharing arrangements in sovereign states that extend well beyond climate policy, and according to the LGMA and its partners, is critical in helping to bridge the global environmental governance “implementation gap” (GCoM, 2022). During the Ministerial, it was announced that CHAMP had secured endorsements from 62 countries—a result described by one interviewee as moderately successful, though far from the most popular COP28 Presidency initiative launched in Dubai (Interview, September 2024, Montreal). By the end of COP28, the number had risen to 72, and two years later, in November 2025, the initiative had attracted six additional endorsers, including, prominently, the European Union and the United Kingdom, as a result of sustained LGMA advocacy throughout that period (see also Lamontagne, forthcoming).

Held on December 6, the final day of COP28’s first week, the Ministerial represented one of the LGMA’s last meaningful opportunities to make a lasting impression before negotiations intensified in week 2. But multilevel diplomacy has its limits: for a group of “non-sovereign” actors to adopt a properly political position—potentially opposing the positions defended by national governments—it must do so through the authorized advocacy channels of the COP constituency system. As we will see in the second part of this analysis, the momentum for multilevel action built during COP28’s first week did not seamlessly carry into the negotiations of week 2.

Week 2: (December 8 to 12, 2023)

LGMA advocacy at COP28: Cities’ misrepresentation as civil society stakeholders

The LGMA’s December 9 daily briefing, held on the second day of week 2 at COP28, opened with the launch of the constituency’s “Step-Up!” social media campaign, requesting that “UNFCCC Parties (…) state clearly and explicitly in the GST outcome texts the need to STEP UP multilevel action in the new NDCs towards 2025 (…).” Social media posts, hashtags, and emails sent to Parties via the constituency’s ICLEI focal point were widely disseminated in an endeavor to sustain the momentum generated during week 1 and to keep cities and, critically, multilevel action at the forefront of ministers’ and negotiators’ attention. On December 1, the COP28 Presidency had secured a major diplomatic victory with the adoption of the highly contentious “Loss and Damage Fund” decision by Parties—a win on the coattails of which the LGMA had also achieved a notable gain, as the text included a commitment to direct access to the fund for subnational governments. However, securing references to multilevel governance in the other key political decision of the COP—the GST—was proving far more difficult, despite 20 Parties having made references to multilevel governance and urbanization in their submissions to the UNFCCC Secretariat on the GST outputs in the months preceding COP28.

The briefing was held at the Multilevel Action and Urbanization Pavilion (MAUP), the LGMA’s headquarters and event space which had been officially dubbed the “Global Stage of Cities and Regions at COP28.” It had hosted dozens of panel side-events organized by LGMA members and partners for the duration of the COP, all livestreamed on ICLEI Global’s YouTube channel for the broader public and constituency members not in Duba. ⁶ Much like at the LCAS, these events served as platforms for local leaders to project and perform both the local and global identities of their cities. In format and practice, however, they also closely resembled the panels organized by NGOs and other civil society and private actors at COP. Located on the second floor of Building 88 at the far end of the Western wing of the Blue Zone (see zone B7 on the COP28 venue map), a temporary structure built for COP28 which housed hundreds of other NGO, IGO, and constituency pavilions, the MAUP was highly accessible and well attended by observers, but it was somewhat difficult to locate and too far from the negotiation rooms to be convenient for most national delegates and negotiators to visit.

By the morning of December 9, the second week of COP28 was in full swing, with many key negotiation tracks, particularly the GST, stalled between oil-producing countries and pioneering states pushing to “phase out” rather than merely “phase down” fossil fuels. The LGMA community of practice had spent the year preparing for these final days and had publicly endorsed a phase-out of fossil fuels, consistent with global city networks’ long-standing mitigation positions, which typically exceed the commitments made by national governments (Gordon, 2020). Its core group of practitioners had divided into 10 “working groups,” each tasked with following a particular negotiation agenda item of relevance to the constituency, which involved attending contact group negotiations and lobbying Party negotiators assigned to that item (Interview, March 2024, online).

This, of course, was easier said than done. COP negotiations are not always open to observers, and gaining access can be difficult, especially when observer participation is as high as it was in Dubai. UN security and UAE crowd management personnel gave priority access to conference rooms to pink “Party” badges, followed by yellow observer badges, the latter of which most LGMA practitioners held. Negotiators, while relatively accessible—they can be approached in hallways and are often willing to exchange contact details such as WhatsApp numbers (Interview, January 2024, online)—have limited time and varying discretion in adjusting the instructions they receive from their capital. This means they cannot always accept input from observers or constituencies, making advocacy outcomes highly uncertain on the ground. To maximize its chances, the LGMA had petitioned several other global fields and institutions—including the G20—in the year preceding COP28 to encourage sovereign governments to adopt its multilevel climate governance agenda, and had prepared a 54-page document compiling all relevant textual outcomes and verbal declarations supporting this agenda to substantiate its advocacy at COP28.

For the first time, the LGMA also prepared a joint position paper expressing the shared views of its members as a constituency, outlining 10 “asks” for Parties and the UNFCCC. ⁷ Two are particularly noteworthy: first, to “[f]ormalise subnational voices in the UNFCCC COP Agenda,” notably by making the Ministerial Meeting on Urbanization an annual “mandated event”; and second, to “[a]cknowledg[e] the urgent need to collaborate across all levels of government” in preparing NDCs in the COP28 outcome texts. To defend these positions and insert their preferred language into the text, the LGMA was constrained by the very structure of interstate multilateralism and the constituency system, forcing them to adopt many of the same advocacy practices as other constituencies within the highly competitive environment of the COP’s B1 and B8 zones, where they had no easier access to rooms or negotiators than other groups.

The literature on city diplomacy has noted that local governments, as subnational branches of the sovereign state, typically enjoy greater information and access to national administrations than non-state actors (see e.g. Amiri and Sevin, 2020). In the specific context of a COP, however, this advantage only partly applies. On the one hand, pre-existing connections to Party delegations and government remain the constituency’s primary advocacy channel both in the lead-up to COP throughout the year and during the conference itself. During their Daily Briefing coordination meetings every morning at 9:00 am, as well as on their private WhatsApp groups, LGMA practitioners updated each other on the progress of the various negotiation tracks they were following, and mandated those among them with citizenship, cultural, or linguistic ties to a given Party or group of Parties to approach their head of delegation or negotiators directly, even soliciting private bilateral meetings whenever deemed useful or possible. Another strategy was to mandate mayors and governors still onsite (of which there were few remaining after the activities of the previous week) to advocate directly to their national delegation on behalf of the constituency. Emails to all Party delegation heads were also sent by the focal point to widely diffuse the LGMA’s positions and demands regarding the inclusion by Parties of specific language in the negotiation texts.

On the other hand, there are important limits to intra-state channels of influence and hallway advocacy. For instance, months earlier, in October 2023, the Council of the EU had included a clear endorsement of subnational leadership and multilevel climate action (para. 43) within its Conclusions (the EU’s official position) in preparation for COP28 (EU, 2023: 17), which was a great victory for the LGMA. Yet, when representatives from ICLEI Europe and the EU Committee of the Regions, two core LGMA organizations, approached EU negotiators to encourage them to follow through on their commitment by placing the requested language into the GST text, they were told their demands were not considered a sufficiently high priority for EU negotiators. Throughout the last week of COP28, the EU delegation—along with many other “friends of multilevel governance” countries—also failed to provide strong endorsement to the constituency during negotiations, or even to mention them in passing in plenary speeches or at the November 8 head of delegations meeting on the GST, with the exception of Japan at the December 13 closing plenary.

Therefore, the momentum of the LCAS and Ministerial on Urbanization deflated very quickly between the two weeks of COP28. Parties’ attention was elsewhere. Despite this, and against unfavorable odds, the constituency succeeded in securing one important win in the final days of the negotiations (in addition to affirmative language on local adaptation in the Global Goal on Adaptation decision): a short sentence segment in paragraph 161 of the GST text, stating the following: “Urges Parties and non-Party stakeholders to join efforts to accelerate delivery through inclusive, multilevel, gender-responsive and cooperative action” (emphasis mine). It was the strongest reference to multilevel governance ever included in the operative (i.e. obligatory) clauses of a UNFCCC decision (Interview, January 2025, online), though it ultimately fell short of what the LGMA had requested in their position paper and call to Parties email.

Conclusion: The long game. Toward multilevel integration? COP30 Belém and beyond

From one perspective, the LCAS and the second edition of the Ministerial Meeting on Urbanization and Climate Change were historic events and critical milestones in cities’ decades-long quest for recognition as “state” actors in their own right in climate diplomacy. Never before had a high-level summit dedicated to subnational governments taken place on one of the central stages of a climate COP and as part of its official agenda. Cities were lauded as frontline actors of the climate crisis, much as they had been at COP21 in Paris nearly a decade earlier, and multilevel diplomatic and governance practices were promoted to a degree never seen before with the launch of the CHAMP. However, translating such performative and declaratory recognition into climate law and institutions proved much more difficult: COP28 yielded direct access to the Loss & Damage Fund for subnationals and progressive language on local adaptation, but also concluded with only a modest—albeit, historical—reference to “multilevel” governance in the GST decision, and did not result in further institutionalization of subnational authorities’ participation as policy makers in the UNFCCC process.

Studying the LGMA’s advocacy and diplomacy at COP28 through relational approaches in IR, combined with event ethnography as my primary method of investigation, has allowed me to demonstrate exactly why and how this mixed and somewhat paradoxical recognition outcome emerged. First, on the theoretical front, this research underscores and analytically leverages the separation and tension between two distinct and rarely intersecting social spaces of global (environmental) governance: the transnational space occupied by “non-state” actors and the multilateral space typically gatekept by states. Contributing to previous work on city diplomacy and global city actorness that has noted this divide (e.g. Gordon and Ljungkvist, 2022) without making it a central focus of analysis, this article demonstrates that it is cities’ positionality as a “stakeholder constituency” in the climate field—rather than their ontology as governmental “state” actors—that primarily constrains and shapes their inclusion within multilateral institutions. Demonstrating what makes cities uniquely indispensable to climate action and distinct from other stakeholders at the LCAS, the Multilevel Action & Urbanization Pavilion events, and the Urbanization Ministerial during the first week of COP28 ultimately did not exempt the LGMA from remaining bound to non-Party stakeholder advocacy practices and regular constituency participation channels to make their positions heard. As a result, their political demands were easily eclipsed by the fraught political context of the negotiations surrounding fossil fuels in the second week of the conference. I therefore contend that foregrounding the dual structure of global governance fields—multilateralism and transnationalism—as well as the ways diplomatic institutions and practices consistently misrepresent the sovereign state as a unitary, rather than a multilevel, entity on the international stage, should be central to the theorization of cities’ global agency going forward.

Second and related, I make an empirical and theoretical contribution to the city diplomacy and urban studies literatures by advancing a communities of practice approach (Adler et al., 2024; see also Lamontagne, forthcoming) to studying global city actorness, which can complement existing research on city networks (Acuto et al., 2024). I highlight the emergence of a “subnational community of practice” formed specifically to advocate for cities’ representation in multilateral diplomacy across multiple fields of global governance, including climate. I argue that focusing on their hybrid diplomacy and advocacy practices enables scholars to apprehend how they deploy collective agency in this dually structured space—an agency that cannot be reduced to the activities or interests of any single city or city network secretariat.

Beyond recognition, transcending the separation between multilateral and transnational governance through multilevel practices is a core aim of the subnational community. My use of event ethnography at COP28 allowed me to show how the structural division between state and non-state, Party and non-Party actors, is materialized in the physical and temporal organization of COPs, and how it interacts with subnational agency and the highly contingent dimensions of global events to shape recognition and multilateral reform outcomes. Further research could examine the subnational community’s practices at other events and in other fields of global governance beyond climate to assess whether they present in a similar way, whether the same structural constraints persist, and whether recognition obtained in one field can accumulate across—or translate to—other fields.

As of November 2025, the trend toward the integration of cities in climate multilateralism appears to be accelerating, although at a slower pace than the LGMA may have anticipated in the wake of CHAMP’s launch in Dubai. Just as the COP28 Presidency had, the COP30 Brazilian Presidency strongly endorsed the LGMA’s urbanization and multilevel governance agenda during the two years leading up to the conference, culminating in a second successful edition of the LCAS at the Local Leaders Forum, co-convened with Bloomberg Philanthropies, in Rio de Janeiro the week before. Despite this, anticipated breakthroughs at COP30 regarding the potential creation of a “work program on multilevel governance” and the institutionalization of the Ministerial Meeting on Urbanization as a UNFCCC-mandated event did not materialize. Nevertheless, recognition is a long game, and the subnational community of practice is not done playing.