Collaborative Governance for Urban Decarbonisation in Italy: Insights on Networked Capacity Building

Abstract

This article analyses how capacity building programmes interact with structural constraints in mission-oriented climate policy, focusing on the Italian pilot Let’sGOv (GOverning the Transition through Pilot Actions) within the EU Mission “100 Climate-Neutral and Smart Cities by 2030”. Using an iterative, reflexive methodology (document analysis, direct observation, and qualitative analysis of questionnaires, workshop outputs, and online training feedback), it examines how municipal actors experience and reinterpret capacity building across three coupled dimensions: internal organisational capacity, external stakeholder relations, and multilevel governance interfaces. The empirical setting is a network of nine Italian Mission Cities (Bergamo, Bologna, Florence, Milan, Padua, Parma, Prato, Rome, Turin) supported by technical partners. The bench-learning pathway combined barrier diagnosis, an intensive in-person workshop, and a codesigned online curriculum structured around three thematic clusters (engagement, data, climate finance). Findings indicate that persistent barriers—departmental silos, resource and time scarcity, rigid human resources and procurement routines, asymmetric data access, and regulatory instability—are not removed by capacity building; rather, they are progressively articulated, specified, and reframed into actionable organisational and policy demands. Bench-learning strengthens diagnostic and relational capacities and enables modest institutional innovations (templates, protocols, internal task forces, shared policy briefs), while “hard” governance infrastructures largely remain unchanged. The paper argues that networked capacity building contributes to the emergence of nascent, project-dependent multilevel interfaces only when it supports collective negotiation with national actors and translates local experimentation into durable multilevel interfaces, mitigating risks of projectification and downward responsibility shifting.

1. Introduction

Climate neutrality policies and targets in Europe have generated an unprecedented demand for capacity building in the public sector [1,2], an instrument, usually framed as able to improve skills and reframe institutional arrangements for urban planning and implementation of adaptation and mitigation actions. However, the concept is often mobilised in ways that might hide power relations and structural constraints. National and local governments are urged to strengthen their capacities [3] to design policies, coordinate actors, and monitor progress, while the systemic nature of climate transitions and the political conflicts they entail are often downplayed or depoliticised in policy discourse [4]. The assumption that more training and better procedures will close the “governance gap” remains dominant in EU climate policy documents, including those underpinning the Mission “100 Climate-Neutral and Smart Cities by 2030”.

Research on climate governance is attempting to challenge this perspective by working on transformative and reflexive governance [5,6,7] emphasising that transitions unfold in conditions of uncertainty, path dependency [8], and distributed control, where conventional steering and rational problem-solving often fail. Seen from the municipal side, “capacities” concern not only technical expertise but also the ability to navigate conflicts, to learn from experiments (and failures), questioning dominant problem framings to reconfigure institutional arrangements. Studies on urban transformative capacity show that organisational routines, interdepartmental coordination, political leadership, and multilevel power relations shape what cities can do, often more than the availability of specific tools or data [9,10]. Capacity building, if reduced to generic training or project-based support, risks reinforcing projectification [11,12] and reproducing existing gaps rather than enabling structural institutional change.

The debate on capacity building for ecological transition in the public sector is therefore undergoing a tension. On the one hand, there is growing recognition that administrations lack time, resources, staff, and skills to know, design, and implement integrated climate policies. On the other hand, empirical analyses of transformative governance point out that many capacity building initiatives do not address the legal, financial, and multilevel governance conditions that constrain action [10] and tend to individualise responsibility for systemic problems. In the context of the Cities Mission [13], research highlights a gap between the ambition of Climate City Contracts and the everyday realities of municipal bureaucracies. This includes uneven mandates, unstable funding, and unclear divisions of roles across levels of government [14,15,16]. In parallel, studies on urban climate governance have produced important evidence on how local administrations develop transformative and experimental capacities, and have documented barriers such as political short-termism, fragmented mandates, and projectification risks [7]. These contributions have advanced the understanding of what governance conditions enable, or inhibit, urban climate action. However, this has occurred through the lens of transition management, urban experimentation, or polycentric governance, without examining capacity building programmes themselves as the object of analysis or tracing how a structured, multi-city programme interacts with the everyday organisational constraints of participating administrations and whether it generates durable change. This article contributes to filling that gap by offering a process-level account of how a networked bench-learning programme unfolds within, and is shaped by, the structural constraints of Italian municipal governance.

This article addresses this gap by examining the work conducted under a national pilot project that explicitly targets governance capacity for ecological transition in the public sector: Horizon Europe project Let’sGOv (NZC-H2020-202209). The project has been developed with the nine Italian cities selected among the 100 Climate-Neutral and Smart Cities by 2030, with the aim of strengthening internal, external, and multilevel governance for climate neutrality. The project engaged the cities of Bergamo, Bologna, Florence, Milan, Padua, Parma, Prato, Rome, and Turin and three technical partners (AESS Agenzia per l’Energia e lo Sviluppo Sostenibile, Energy Center Politecnico di Torino, University of Bologna—UNIBO), with UNIBO being also the scientific coordinator, as part of a national response to the EU Cities Mission. The experience of the researchers in the project gave access to internal municipal debates, cross-city negotiations, and interactions with national actors that are rarely documented in the literature on mission-oriented climate policy. While the paper draws on insights emerging from the project, it develops a broader analytical framework, moving beyond the specific case to support a more general and holistic reflection.

The paper is organised as follows. Section 2 develops the theoretical and analytical framework, situating capacity building within the debates on bench-learning, transformative and reflexive governance, multilevel governance, and projectification. Section 3 sets out the research objectives, research questions (RQ), and the methodology. Section 4 presents the empirical findings from the Let’sGOv pilot through the three analytical dimensions of internal, external, and multilevel capacity. Section 5 discusses these findings in relation to mission-oriented climate governance, with particular attention to institutional change, projectification, and the emergence of multilevel interfaces. Section 6 concludes by synthesising the main contributions of the article, its limitations, and the implications for future research and policy

2. Theoretical and Analytical Framework

2.1. Capacity Building as Organisational Space: Constraints, Justice Tensions, and “Governance Traps” Beyond Training

Capacity building is widely recognised as a central means of implementation for global climate goals, as formalised in Article 11 of the Paris Agreement [17] and supported by the Sustainable Development Goals (particularly SDG 13). This means that debates on how cities learn and municipal officials experience capacity building activities cannot be separated from the broader institutional policy settings in which that learning is supposed to operate [18].

The European Green Deal, the “Fit for 55” package, and the Mission “100 Climate-Neutral and Smart Cities by 2030” (hereafter called Mission) all frame climate neutrality as a systemic objective that demands new skills and multi-actor coalitions across levels of government. The Mission places cities at the core of an urban-oriented policy that asks local administrations to move beyond sectoral interventions and to design integrated, coproduced Climate City Contracts (CCCs) within multilevel governance settings. CCCs [19] are presented as engines for this transformation, securing city-level capacity building to multilevel governance, long-term investment planning to co-create transition roadmaps with quintuple helix actors’ commitment to climate justice goals. CCCs are not legally binding instruments but represent an explicit political commitment, validated by the European Commission through the awarding of the Mission Label, which recognises the quality of the path undertaken, strengthens its visibility, and facilitates access to new funding and partnership opportunities. Each CCC is a dynamic and continuous process, a “living” document, that should be intended to adapt over time to changes in local conditions, political priorities, and available knowledge.

In practice, this turns CCCs into a key interface where expectations about local capacities and multilevel coordination are crystallised while periodically renegotiated. In the Mission framework, capacity building aims at intercepting and enhancing the ability of administrations to hold together ambitious climate objectives, reduced resources, dispersed regulatory frameworks [20], often the social conflicts that emerge around unequal exposure to climate risks and the uneven spatial distribution of transition benefits and burdens [21].

In parallel to the mission, national frameworks and regional strategies for climate and energy define binding targets and funding architectures that condition what local administrations can implement [22], therefore, what “capacity” means in practice. Seen from this angle, capacity building is as much about navigating constraints set elsewhere as it is about improving local skills.

In this context, research on capacity building moves beyond a narrow focus on training and technical assistance to question how public sector capacities are shaped by contextual factors, such as justice claims. Recent work on urban climate governance and the Cities Mission highlights both the enabling role of transnational municipal networks and contracts [23] and the risk of “governance traps”, where cities remain locked in weak institutional settings despite ambitious rhetoric [24]. The debate increasingly recognises that capacity building must be read as part of a broader transition for governance architecture, one that links European missions, national regulatory choices, and local institutional experimentation. In the public sector, this transition needs long-term sustainable capacity systems and human resource development and the extension of curricula to prepare future public servants for effective climate governance arrangements. The question is therefore not only what individual officials know but how whole administrations are positioned within this evolving architecture.

In this context, the emergence of bench-learning methodologies is interesting to explore as a device to activate multiple relationships among partners who are not positioned within hierarchical learning structures. Rather than assuming a distinction between those who teach and those who learn, it is based on a condition of equality, where value emerges through mutual exchange and shared reflection.

Bench-learning has emerged as an evolution of benchmarking, shifting the focus from performance comparison to collaborative learning and the adaptation of good practices. Early contributions define it as an organisational learning method that allows actors to understand and improve their own practices through the experiences of others, combining efficiency, collective learning, and broad participation [25]. More recent studies position bench-learning within the fields of knowledge management and organisational learning, highlighting its role in fostering openness, experimentation, and the integration of knowledge across organisational boundaries [26].

In contrast to benchmarking, which is typically based on quantitative indicators and retrospective comparison, bench-learning emphasises a continuous and qualitative learning process. It is also distinct from more generic forms of peer-learning. While peer-learning focuses on exchange among equals, bench-learning introduces a more structured and action-oriented approach, centred on the interpretation of good practices, the circulation of tacit knowledge, and the co-development of context-sensitive solutions. This makes it particularly suited to complex and uncertain environments, where standardised solutions are not easily transferable [27].

Applications in the public sector, and particularly in municipal contexts, show how bench-learning can support inter-organisational learning through networks, self-assessment processes, and iterative cycles of reflection and action. These approaches have been used, for instance, in European public administrations to improve performance while allowing local adaptation and to support decision-making in contexts characterised by institutional diversity [26].

In the context of the European Union (EU) Cities Mission, bench-learning provides a relevant methodological approach as it allows cities to engage with shared challenges while acknowledging contextual differences. Compared to more conventional peer-learning formats, it offers a more structured way to connect experience, reflection, and action and to move from the exchange of practices to their translation into organisational change. For this reason, it is particularly suited to addressing governance barriers, as it supports the articulation of concrete organisational and policy demands emerging from practice.

2.2. Transformative and Reflexive Governance as Lenses for Capacity Building

Debates on climate governance have progressively moved from incremental policy adjustments, transfer of technical skills, or the diffusion of predesigned solutions to questions of structural transformation. Work on transformative climate governance emphasises that climate change is embedded in sociotechnical systems whose stability relies on deep-rooted technologies, institutional routines, and assigned interests [6] and that these systems cannot be redirected through single sector measures [11]. Here, governance is not limited to formal institutions but includes the large configuration of actors, rules, discourses, and resources that shape how problems are defined and which solutions become thinkable and implementable. Transformative climate governance seeks to understand how such configurations can be reoriented over time and over different issues and which capacities public institutions need to support that reorientation. This body of work invites us to treat capacity building as part of system change [9] and not as a technical add on.

Against this backdrop, an emerging strand of research advocates a shift from traditional capacity building towards “transformative capacity building frameworks”: an evidence-informed [28], justice-oriented approach that equips practitioners and public institution staff to understand system interdependencies and the importance of long-term planning and orientation, to enhance institutional flexibility and cross-sectoral coordination, to learn and revise strategies to tackle context specific constraints before designing experimental interventions [29,30]. These studies show that many urban administrations remain trapped in rigid bureaucratic cycles, budgeting rules, and departmental isolation, even when they adopt ambitious climate targets. Transformative capacity building, in this view, concerns the development of institutional settings that imply a movement from skills transfer to collaborative problem setting, generating both solution-oriented position (based on evidence) and experimentation informed by evidence. This approach marks a shift from individual to relational institutional capacity that deals with uncertainty and conflict and connects local initiatives with wider shifts in policy and investment. The analysis of Let’sGOv’s capacity-building process speaks directly to this shift, as it examines how far a concrete programme can move from one model of capacity to another.

In the evolving European and national governance architecture for climate neutrality, this transformative perspective reframes capacity building as the gradual construction of analytical competences, e.g., the ability to measure and monitor urban emissions overlapped to urban vulnerability indicators, that enable public institutions to negotiate their own priorities and align climate action with questions of justice and territorial inequality. Analytical capacity here affects whose vulnerabilities become visible and which territories are prioritised, avoiding neutral positioning.

Reflexive governance adds a focus on how knowledge, power, and uncertainty are handled inside these arrangements [5]. Several authors insist that under conditions of high uncertainty and contested values, climate governance cannot rely on linear planning and optimisation [31] but should require continuous questioning of goals and instruments, attention to side effects, and deliberation about trade-offs. Instruments such as strategic experiments, adaptive policy mixes, and participatory assessment are seen as means to open governance to learning and contestation without reducing the debate around predefined solutions. From this angle, capacity building must support reflexivity inside institutions; however, as pointed out by Voss and Bornemann, reflexive governance [31] risks remaining primarily concerned with problem-solving through experimentation and learning at the focal level of governance arrangements, while neglecting agonism [32], asymmetric power relations, and avoiding embedding processes within broader political contexts. Transformative capacity building must instead foster the ability of public officials to engage with criticism, to recognise conflicts as informative, and to adjust their own categories and routines considering new evidence and arguments [28]. For mission-oriented programmes, this implies treating training sessions and peer-exchanges as political arenas where definitions of problems, solutions, and legitimate expertise are open to contestation, not merely as technical exercises. Integrating these critiques, reflexive governance raises a warning for capacity building in climate neutrality missions, highlighting the no-neutrality of such activities. They can instead be arenas where problems, solutions, and legitimate expertise are constantly negotiated, keeping debates on goals and pathways open and linking local experimental practices to ongoing conflicts and structural reforms at higher levels. Otherwise, capacity building may slide into technocratic routines that stabilise arrangements under the language of participation and learning, a tension that becomes particularly evident in urban climate experiments.

2.3. Urban Experiments and Multilevel Governance: Scaling, Feedback Loops, Power Asymmetries, and Projectification

Urban climate experiments, including living labs and mission-oriented pilots, have been presented as concrete arenas where transformative and reflexive governance can take shape [33,34]. These experiments can generate new coalitions and make visible overlooked dimensions of climate vulnerability or injustice. However, they can also become highly technocratic, having a tendency to cluster in specific sociotechnical systems, confined to demonstration sites, and taking the form of discrete, project-based interventions that leave institutional logics untouched. The literature shows that experiments tend to remain isolated when they are not connected to processes of institutional scaling [35] and when they do not challenge underlying power relations and dominant narratives [36]. Reflexive and transformative governance viewpoints also insist on the need to link experimentation to broader policy and institutional reforms and to treat experiments as occasions for collective learning [37] about system dynamics and political choices. This is where the multilevel dimension becomes crucial to correct the limits of urban experimentation and to build the vertical feedback loops, regulatory interfaces that would allow experiments to contribute meaningfully to systemic climate governance.

Multilevel governance and polycentricity draw attention to the vertical and horizontal relations that condition cities’ action towards climate neutrality [38]. Italian cities operate in an energy landscape where competencies and resources are owned by the state, regions, and national regulators, while municipalities retain limited fiscal autonomy and formal powers over energy infrastructures [39,40]. Energy systems are treated as national assets, and planning for grids is largely isolated from urban policy, even though cities are expected to deliver climate neutrality. Climate governance literature documents how responsibilities and resources are distributed and misaligned across these levels, leaving local administrations with extensive implementation duties but limited influence over regulatory and financial frameworks [3,41]. It is commonly shared that multiple centres of decision-making can enhance experimentation and learning [42] and can reduce the risks of centralised failure; they risk hiding accountability and reproducing inequalities [43] as uneven capacities to set agendas, define problems, and shape outcomes become frequently overlooked or treated as external to governance design. In this context, multilevel capacity building cannot be reduced to better coordination. However, it needs to address how Italian cities might gain durable roles in shaping climate and energy rules and investment decisions and how Mission instruments such as Climate City Contracts can be used to renegotiate the current division of authority rather than only adding new responsibilities at the local level. These findings caution against celebrations of local and transnational experimentation and call for closer attention to how multilevel coordination and power asymmetries condition what experiments can achieve and whose interests they serve.

In parallel, recent contributions have examined the increasing role of projectification [44,45] in urban governance and sustainability transitions. The term refers to the growing dominance of a “project logic” as a way of organising action, particularly in contexts of urban change. This logic operates at multiple levels: in the structuring of funding schemes around time-bound projects with predefined objectives, in the use of temporary organisational forms within public governance, and more broadly, in the diffusion of the project as a taken-for-granted way of framing action.

In the context of urban transitions, project-based funding and governance tend to impose short time horizons, predefined outputs, and measurable results. While this can support coordination and delivery, it also tends to orient action towards what is achievable within project constraints, rather than what is required for systemic change. This dynamic has been described as a form of incrementalism, where interventions remain limited in scope and ambition.

A central issue concerns the distinction between project and experimental logics, which are often conflated in practice. Project logic relies on planning, clear objectives, and assumptions of control and transferability. Experimental approaches, by contrast, are oriented towards learning, uncertainty, and context-specific adaptation and may involve contestation and redefinition of goals. When experiments are framed as projects, their open-ended and exploratory nature is often reduced, limiting their potential contribution to transformative change.

The literature also highlights several critical effects associated with projectification. These include forms of institutional amnesia, as knowledge produced within temporary projects is not retained or embedded in organisational routines; processes of depoliticisation, where complex issues are reframed as technical tasks; and a diffusion of responsibility across fragmented and temporary arrangements. Taken together, these dynamics can weaken the capacity of urban governance systems to sustain long-term transformation.

In response, some authors propose moving beyond project-based approaches by strengthening connections between projects, experiments, and permanent institutional structures. This includes developing forms of cumulative learning across initiatives and creating organisational arrangements that can retain and build on project-based knowledge over time.

This tension is especially evident in EU policy frameworks such as the Mission “100 Climate-Neutral and Smart Cities”, which combine a strong reliance on project-based instruments with the ambition of achieving systemic and long-term transformations. This raises the need to critically assess how projectification affects local governance capacities and the ability of cities to move beyond incremental change.

3. Objectives and Methodology

The literature shows that current debates on urban governance are shaped by a set of recurring challenges. In particular, contributions converge on three main aspects. First, capacity building needs to be understood as an organisational process, not only as training. Second, it can play a role in supporting more reflexive and transformative forms of governance. Third, the European project-based approach introduces specific risks linked to projectification. Existing literature on urban climate governance pays increasing attention to capacity building but offers limited empirical evidence on how it translates into organisational change and engages with governance barriers, especially in contexts shaped by project-based policy instruments such as the EU Cities Mission.

Drawing on this, the paper aims to contribute to the understanding of urban climate governance by analysing a group of Italian cities involved in the EU Mission. The focus is on how capacity building processes relate to the barriers that cities face in advancing the climate transition, contributing to a more critical understanding of the promises and limits of capacity building in mission-oriented climate policy, with a qualitative embedded process evaluation.

The paper is based on two main assumptions. First, capacity building can support ecological transition only when it is oriented towards changes in organisational routines, regulatory interfaces, and multilevel coordination. Otherwise, it risks adding new tasks to already overstretched administrations. Second, governance barriers remain a key entry point for municipalities, particularly in the Italian context, to structure and accelerate climate action.

On this basis, the article addresses two research questions:

RQ1. 

How do municipal actors experience and interpret capacity building processes in relation to organisational change, external relations, and multilevel climate governance?

RQ2. 

Under which conditions can a networked bench-learning programme contribute to addressing governance barriers and strengthening capacities across organisational, inter-institutional, and multilevel dimensions?

The article adopts a qualitative, embedded process evaluation of a capacity-building programme. It combines direct observation, document analysis, and iterative survey data collected within a network of nine Italian municipalities involved in the EU Cities Mission. The empirical material includes questionnaires, workshops, and pilot training activities.

On this basis, the paper develops an analytical perspective and makes three main contributions. First, it provides empirical evidence on how municipal actors experience and reframe governance barriers through networked learning processes. In particular, it shows how these barriers move from general problem statements to more defined organisational and policy demands.

Second, the paper develops and applies a three-dimensional analytical framework based on internal, external, and multilevel capacities. This framework is used to assess where capacity building contributes to changes in institutional arrangements and where its effects remain limited.

Third, the paper reflects on the conditions under which networked capacity building can support the emergence of multilevel coordination. This includes a critical discussion of projectification dynamics and of the limits of the findings, given the absence of formal responses at the national level.

Overall, the paper contributes to a more grounded understanding of what capacity building can and cannot achieve in the context of mission-oriented climate policy.

The authors took part in the codesign of the programme, in preparatory meetings, in the intensive workshop, and in online training sessions. This dual role as project partners and researchers provided access to internal municipal conversations, emerging governance practices, and cross-city negotiations. At the same time, this proximity carries risks of confirmation bias, for which we adopted several mitigation strategies. Role separation was maintained throughout the project: facilitation tasks (e.g., workshop moderation, training logistics) were carried out by the broader consortium, while the analytical work reported in this article was conducted by the authors, working with independent field notes and coding protocols.

During the activities, it was possible to observe how municipal staff described their governance gaps and constraints and how they reacted to proposed capacity building tools and formats.

The multisite character of Let’sGOv, spanning different cities and on internal, external, and multilevel capacities of governance, allowed the study to trace convergences and divergences across the different Italian institutional contexts and to identify patterns that are relevant for debates on national climate governance (Figure 1).

The direct observation was further divided into three core phases: barrier analysis (named Phase 1) carried out mainly through the questionnaire instrument; the organisation of an intensive in-person workshop of two days in Bologna in the beginning of 2024 (named Phase 2), and an online bench-learning curriculum between the end of 2024 and the beginning of 2025 (named Phase 3). From a methodological perspective, the barrier analysis presented in this paper is designed to address the specific objectives of the study. For this reason, the analysis is conducted at an aggregated level across the group of cities, rather than focusing on the specificities of individual cases. This choice allows for the identification of recurring governance barriers and shared patterns within the network. This approach entails a deliberate limitation. It does not aim to provide an in-depth account of city-specific conditions, which would require a different research design. A more detailed analysis at the individual city level is therefore left for future research.

Phase 1 was devoted to gathering information on barriers evolutions. The questionnaire phase was organised with three different iterations, as

Table 1. Scheme of the data collection phases.

Iteration Name	Period of Ref.	Method Used	Types of Questions
Preliminary phase	Jun–Aug 2022	Feedback collection on a word file	Identification of main barriers (internal, external), identification of priority themes to focus on to address the transition acceleration.
First Questionnaire	Aug–Sep 2023	Online survey	Evolution of the main barriers (internal, external, multi-level), identification of key challenges for the future of the transition.
Intermediate Questionnaire	May 2024	Online survey	Evolution of the main barriers (internal, external, multi-level), perceived evolution of key challenges for the future of the transition.
Follower Questionnaire	Aug–Sep 2024	Online survey	Identification of main barriers (internal, external), identification of key challenges for the future of the transition.

details. The first iteration was a preliminary scouting of the main barriers and identification of common aspects to work on, before the starting of the project (Preliminary Phase). The second iteration was conducted through the method of the questionnaire (First Questionnaire). Only partner cities were able to access the questionnaire. The final iteration was conducted again using the questionnaire methodology (Intermediate Questionnaire). At the end of the project, a final questionnaire was sent to cities to gather feedback on the entire programme issued.

In each questionnaire, all nine Mission Cities responded, typically through internally designated municipal officers per city (9–18 individual responses per wave, depending on whether cities submitted a single consolidated response or multiple departmental inputs). The questionnaire instruments included a mix of closed-ended items (e.g., Likert-scale ratings of barrier severity across the three capacity dimensions: internal, external, and multilevel) and open-ended prompts. Instruments were administered digitally via Microsoft Forms to invited participants only. The Follower Cities questionnaire (24 additional cities, August–September 2024) used a simplified version of the same instrument and served solely to triangulate the commonality of barriers identified by the nine Mission Cities.

In total, the study engaged with 24 Italian cities, which reflects two distinct groups with different analytical roles in this study. The nine Italian Mission Cities constitute the primary unit of analysis: they participated in all three phases of the bench-learning pathway and provide the empirical basis for the findings. The additional 15 “follower cities,” which completed only the final questionnaire instrument, are a supplementary validation layer, allowing the research to probe whether the barriers and dynamics observed among the Mission Cities are recognisable more broadly across the Italian municipal landscape.

Phase 2 consisted of an intensive in-person workshop that brought together city representatives to present governance initiatives (i.e., climate assemblies, one-stop-shops, and energy communities) and to discuss obstacles and enabling factors. Fieldnotes and workshop outputs were analysed to trace how individual perceptions were confirmed, contested, or reframed in a collective setting. The intensive workshop was designed to facilitate an operational and structured exchange among the nine cities involved in the project, focusing on selected practices currently implemented within their climate transition pathways.

Phase 3 consisted of a cycle of online pilot trainings focused on energy infrastructure, energy communities, and monitoring. Here, the analysis concentrated on how participants linked their local situations to national regulations and expert inputs and how they positioned their administrations within wider multilevel governance debates. The online pilot training was codesigned with the cities through an iterative process of proposing themes and voting on the most challenging but also urgent ones. The process allowed the identification of six key topics, spanning from the electrification of the built environment to the monitoring of climate transition projects.

These three steps generated complementary evidence on how capacity building activities were experienced and interpreted by municipal actors.

The empirical material used in this article consists of (a) the baseline and intermediate questionnaires on governance barriers and anticipated challenges; (b) synthesis reports from the bench-learning workshop; and (c) documentation and feedback from the online training sessions. The analysis focused on how municipal actors articulated barriers and needs, how these articulations evolved over time, and how they connected to internal, external, and multilevel capacities.

Analytically, the study proceeds through an iterative, reflexive thematic analysis operationalised as follows. The unit of analysis is the individual questionnaire response, aggregated at the city level for cross-case comparison. Thematic codes were first generated from the questionnaire responses, grouping recurring problem framings and organisational descriptions. These were progressively aligned with the three analytical dimensions (internal, external, and multilevel capacities), following a constant comparison logic across cities and data collection waves. Where possible, emerging themes were cross validated against the workshop outputs and training feedback, which provided an additional layer of triangulation. Direct quotations included in the Section 4 were selected to illustrate recurring patterns, and the selection prioritised variation across cities and departmental roles.

4. Results

4.1. Analytical Framework for Let’sGOv: Internal, External, and Multilevel Capacities

The analysis is structured around three interconnected dimensions of capacity building: internal organisational arrangements, external relations with local stakeholders and system actors, and multilevel interfaces with regional, national, and European institutions.

Table 2. Synthesis of the discussion framework, main concepts, and their relevance for the research questions.

Perspective	Capacity Building	Main Risks	Relevance for RQ1	Relevance for RQ2
Transformative governance. Structural change in socio-technical systems	Should reconfigure institutional settings: from skills transfer to collaborative problem-setting, long-term orientation, cross-sectoral coordination, and attention to justice	Capacity reduced to training; projectification; ambitious targets without changing budget rules, HR regimes, mandates	Helps read staff frustration when new tasks are added without changing structures; explains why officials experience a gap between rhetoric and organisational reality	Bench-learning is relevant only if it feeds into changes in mandates, procedures, investment priorities, not just “learning events”
Reflexive governance. How institutions handle uncertainty, knowledge, conflict, and learning	Should foster critical self-questioning, openness to contestation, and revision of goals and instruments through experiments and deliberation	Focus on problem-solving within existing arrangements; underplays power asymmetries and agonistic conflict; risk of technocratic “learning”	Reflects how officials use capacity-building spaces to question routines and name conflicts but also how these spaces can feel constrained	Bench-learning contributes when it is designed as a political arena where problem framings and expertise are negotiable, not as neutral best-practice exchange
Urban climate experimentations. Pilots, living labs, mission projects as sites where new practices and coalitions are tested	Seen as practical tools for learning-by-doing, demonstrating solutions, and building coalitions around specific interventions	Remaining isolated, short-term and project-based; clustering in “safe” areas; failing to challenge underlying logics or to scale institutionally	Helps explain why staff may value concrete pilots but also perceive them as parallel to ordinary work, with limited impact on everyday constraints	Bench-learning matters when it links local experiments to institutional scaling strategies and to shared diagnosis of structural barriers, not when it only showcases projects
Multi-level governance/polycentricity. Vertical and horizontal relations between EU, state, regions, municipalities, utilities, and other actors	Should build cities’ ability to navigate and influence regulations, funding schemes, and data/infrastructure governance across levels	Celebrating decentralisation while leaving agenda-setting power at higher levels; obscuring accountability; reproducing territorial inequalities	Frames how municipal actors experience external and multi-level constraints as part of capacity	Bench-learning is effective when it helps cities act collectively towards national/EU actors, creating real interfaces for rule and resource renegotiation

summarises the relationship between these dimensions, the theoretical concepts discussed above, and the two research questions. The links between the analytical framework, the core concepts, and the research questions are synthesised in Table 2.

This integrated lens underpins the analysis of Let’sGOv, guiding the questions posed to the pilot: how municipal actors experience capacity building activities in each of these fields and under which conditions a networked bench-learning programme can contribute to capacities that matter for the emergence of a national climate governance architecture for climate neutrality.

The following sections apply this framework to the Italian case, showing where capacity building efforts begin to shift these three dimensions and where they remain constrained.

Let’sGOv (“GOverning the Transition through Pilot Actions”) is a Horizon Europe pilot developed within the EU Mission “100 Climate-Neutral and Smart Cities by 2030” as a shared initiative of the nine Italian Mission Cities. The project targets the governance side of the climate and energy transition starting from the shared diagnosis that internal silos, weak participation mechanisms, and fragile multilevel coordination are major bottlenecks for local climate action. Moreover, it seeks to experiment with new ways of organising, negotiating, and learning across administrations and levels of government. The project architecture is articulated in five work packages that together construct a multilayered governance laboratory. A first package designs and runs a cross city-bench-learning programme that combines an initial survey of governance barriers, an intensive in-person workshop, and a digital training curriculum, with continuous monitoring of internal, external, and multilevel obstacles throughout the pilot. A second package organises three thematic clusters (engagement, data, finance) to map stakeholders and data needs, codesign toolkits (on participation, data agreements, innovative financial models), and prepare approaches, tools, and methods to be tested in the participating cities and beyond. A third package supports the finetuning and implementation of test beds in each city, aligning local experimentations with the learning and toolkits, and documents their pathways and governance barriers. The project is completed by a sensemaking framework, shared key performance indicators and evaluation briefs that link city-level experiments, cluster work, and network-level learning, with attention to both impacts and financial sustainability.

This scheme translates into three interconnected levels. At network level, it builds a structured collaboration among the nine cities, supported by a Memorandum of Understanding, signed in 2022 with the national Ministry of Infrastructure, and by a bench-learning programme. At cluster level, it organises activities around three levers considered crucial for energy-related emissions reduction where cities face similar governance barriers but have developed different practices. At city level, it links this shared work to concrete pilot actions, such as energy communities, one-stop-shops, data platforms, or new public–private agreements, which serve both local purposes and the collective learning ambition of the network. This article focuses mostly on the network level of this architecture, making the Let’sGOv deployment the empirical field in which the interaction between capacity building efforts and structural constraints was observed, and in which the extent to which a national city network can operate as a laboratory for climate governance was assessed.

4.2. Internal Governance and Organisational Capacity

Internal silos and weak coordination emerged as the most persistent barriers across all nine cities. In the questionnaires, almost all administrations described fragmented responsibilities, limited transversal collaboration, and an overload of project-related duties concentrated on a small group of recurring officers as structural, ongoing condition, and shortcoming. The bench-learning work made visible how these issues are experienced in everyday routines and how they vary across departments. Respondents captured these conditions noting a “great difficulty in getting the priority of climate transition recognised within the various Directorates, where it is put in second place compared to contingent priorities.”

Moreover, the first phase of bench-learning showed that internal climate governance relationships are often translated into very tangible tensions. Examples include the following: difficulty in coordinating procurement and Italy’s Minimum Environmental Criteria (Criteri Ambientali Minimi, CAM) for green public procurement requirements, misalignment between climate objectives and human resources (HR) or information technology (IT) procedures, and lack of shared tools to track processes’ responsibilities on energy-related files. Participants described climate tasks as an “extra layer” added on top of existing responsibilities, with limited or absent formal recognition in job descriptions or quality evaluation systems. This reinforces a perception of climate sensitive work as project-based and dependent on a few motivated individuals, mainly working in dedicated departments. As a result, “the sustainability and stability of these groups, for example after drafting the Climate City Contract and at the end of specific projects, will also depend on the willingness and availability of the individual officers to participate and act as connectors” (from the questionnaires). At the same time, several cities report first attempts to institutionalise cross-sectoral spaces, such as interdepartmental task forces, climate offices, political dedicated councillors (such as the city of Bologna’s Councillor for Ecological Transition and the Climate Pact, or the City of Florence Climate Direction Office, directly linked to the mayor office), or steering groups linked to Sustainable Energy and Climate Action Plan (SECAP) and/or Climate City Contract processes, which Let’sGOv takes as starting points.

The Intensive Workshop phase made this heterogeneity explicit and was a critical moment of further reflection of the evolution of the governance barriers and the production of alliances to overcome them. When presenting peer case studies such as climate assemblies’ pathways, energy communities implementation processes, one-stop-shops, or internal task forces, city teams were asked to map both outputs and internal procedures: which departments were involved, who convened whom, how decisions were made, and what the resource allocation process was. This exercise revealed that organisational learning is often informal and person-based, relying on personal networks and shadow coordination. Bench-learning does not solve these structural constraints, but it allows participants to see that others face similar problems and to name recurrent patterns: the central role of a few brokers who translate between political leadership and technical departments, the difficulty of maintaining transversal knowledge groups beyond project timelines, or the fragility of arrangements when key individuals move or elections change priorities.

Furthermore, an important aspect marked as crucial need from the questionnaires and workshop discussions is training and transversal skills development, stressing that basic climate and energy literacy is uneven within administrations and that many colleagues perceive the topic as too complex or technical. At the same time, participants expressed fatigue with generic training offers and asked for formats that are closely tied to ongoing work (e.g., permitting photovoltaic (PV) systems in historic centres, data-sharing agreements with distribution system operators (DSOs), design of MoUs—Memoranda of Understandings—for energy communities). The online sessions responded to this by inviting national agencies, grid operators, and expert cities to discuss specific bottlenecks while keeping a focus on governance implications. Feedback collected after the sessions suggests that city staff valued this situated training but also that they struggled to involve colleagues outside the climate or project teams due to time constraints and the lack of internal incentives. This points to the limits of external capacity building offers when internal HR and management systems do not recognise transversal learning as key. Despite these constraints, small incremental changes were observed over the project’s two-year duration. In the intermediate survey and in the lessons-learnt report, some cities reported that climate objectives have been integrated into performance plans, that cross-sectoral groups created for Let’sGOv or related projects have been maintained, or that new internal procedures for data requests or pilot selection have been formalised. Others noted that the project has mainly reinforced existing awareness without yet changing embedded routines. Overall, internal barriers such as limited resources, rigid procedures, and political precariousness remain, but they are now more clearly articulated and, in some cases, translated into concrete organisational demands (e.g., a dedicated coordination unit, clearer mandates for climate offices, necessity to monitor training results, shared templates for MoUs).

4.3. External and Multilevel Governance Capacity

On the external governance side, the project confirmed that relations with higher-level governments and key system actors are considered crucial but also challenging to implement and demanding to maintain. In all phases of the bench-learning, the nine cities identified interactions (or lack thereof) with national-level ministries, regional regulators, and private grid operators as primary barriers to local energy transition. They often described a misalignment between national rules and local trajectories: e.g., burdensome authorisation procedures for renewable plants, uncertainty and continuous shift around the regulation of energy communities, dispersed responsibilities for data access, and rigid constraints on the use of public finance for long-term investments. These issues appear repeatedly in open responses to questionnaires and in workshop debates, where participants highlighted that many bottlenecks would require dedicated changes and commitments in national frameworks.

Relations with local stakeholders and private actors are another component of external capacity building. In the questionnaires and workshop discussions, cities report progress in engaging condominium administrators, banks, professional associations, and civil society organisations around energy communities and building renovation, often through one-stop-shops or energy desks. At the same time, they underlined difficulties in maintaining these alliances over time, in managing expectations and negotiating fair distribution of benefits, especially in contexts of energy poverty. Bench-learning addressed these issues mainly through peer-exchange: cities presented different models of public–private agreements, governance of Renewable Energy Communities (RECs), and local partnership arrangements, while participants discussed legal risks, trust-building, and the role of municipalities as guarantors. The online sessions on innovative finance and public–private partnerships further explored how local administrations can use their power without overstepping legal limits or assuming unsustainable risks.

Data and finance emerged as crosscutting external and multilevel issues, especially regarding the lack of appropriate and interoperable energy data as a major obstacle for monitoring and communicating climate actions. The participants reported asymmetric relations with utilities, who control key datasets and infrastructures but are not obliged or incentivised to share information in forms usable for local governance. There is an “absence of data-sharing protocols and updated databases on urban energy data, held by various entities including the Region, the national management energy agency, etc.” and “difficulty with the energy utility in retrieving consumption data in a precise way, due to privacy and competition issues” (cit. from surveys). The data cluster of Let’sGOv responded to this request by focusing training on data agreements, sharing experiences of interoperable platforms and thematic digital twins, inviting cities and regional actors who have already negotiated such arrangements to share templates and strategies. Regarding the topic of finance, the project highlighted both the opportunities offered by NRRP (National Recovery and Resilience Plan) and EU funds and the difficulties of aligning short-term project funding with long-term investment needs and of designing financial schemes that reach private buildings and vulnerable households. The subsequent discussions pointed to the need for national schemes that recognise municipalities as strategic partners in deploying innovative financial instruments and not only as mere project applicants.

The multilevel dimension became more visible towards the end of the pilot, when the nine Mission Cities used Let’sGOv as a platform to address national institutions through joint messages and to position themselves as a collective speaker with a unique, situated point of view on the challenges of project and action implementation. On this assumption, the project, while it did not have a formal mandate to reform national regulation, committed to create limited multilevel interfaces that began to address this gap. It built on the existing Memorandum of Understanding between the nine Mission Cities and the Italian Ministry of Infrastructure and Sustainable Mobility and used the bench-learning outputs to feed into joint policy messages. The policy brief produced at the end of the pilot, for example, synthesised local experiences with one-stop-shops and energy communities and translated them into requests for more stable support schemes and more agile procedures for authorisations and grid connections. In this specific case, capacity building was able to include learning to collectively address messages to national actors through shared demands.

The activation of a “follower” network of additional Italian cities reinforced this movement, highlighting that governance capacity building can also mean building horizontal coalitions that can negotiate more effectively with ministries and regulators. At the same time, the project team made clear that these steps are initial and fragile: coordination among cities still depends on project structures and external funding, and currently, there are no formalised national mechanisms that ensure continuity of this dialogue beyond the Mission framework.

Overall, the results suggest that Let’sGOv did not remove internal, external, and multilevel barriers, but it made them more visible and collectively legible. The convergence in barriers’ framing across different cities is a finding pointing to the systemic character of Italian municipal climate governance constraint. At the same time, the capacity to create a shared language and generate institutional innovations in how cities organise internally, engage stakeholders, and relate to higher and lower levels of government varies across cities, reflecting differences in institutional starting points.

5. Discussion

The findings of Let’sGOv bench-learning activities confirmed that capacity building for ecological transition is still moving through institutional constraints. On the one hand, internal disconnections and structural fragmentations persisted throughout the pilot; on the other, the recognition, naming, and definition of such barriers changed radically. It was observed that municipal staff moved from an abstract understanding of climate challenges to the identification of specific friction points (e.g., lack of shared templates and guidance for community governance, unclear mandates for climate offices, asymmetric relationships with grid operators, rigid procedures for innovative green finance) to be translated into tangible organisational and political demands for climate governance. This result marks a shift towards the ability to articulate diagnosis, toward achieving the capacity to recognise signals and alarms, prior to the emergence of certain challenges, stimulated by a relational capacity. Transformative and reflexive governance scholarship argues that transitions require both the ability to “see” system interdependencies and structural drivers of problems [29] and to alter the rules, routines, and resource flows that lock systems in place [10]. In Let’sGOv, the bench-learning pathway clearly strengthens the first dimension, while the “hard wiring” of administrations largely remained untouched, aligning to the concerns that reflexive governance often stops at the level of discursive problematisation [46] without shifting underlying power relations.

As a result, capacity building designed as a relational and diagnostic process can open spaces for the implementation of reflexive governance [5]. Reflexive governance, as Voß and Bornemann argue [31], tends to operate through the opening of deliberative spaces and the cultivation of systemic awareness but risks remaining confined to discursive reframing when it fails to engage with the political conflicts and power asymmetries that structure institutional arrangements. In this regard, the bench-learning programme, based on peer-exchange and targeting governance barriers, allowed participants to question their own routines and to expose tensions between ambition and resources (or lack thereof). This analytical self-observation confirmed also that bottlenecks regarding climate governance cannot be only solved locally, as they might generate parallel infrastructures. Much like the urban experiments analysed by Bulkeley and Castán Broto [35] and the climate governance experiments discussed by Sengers and colleagues [36], Let’sGOv creates spaces of experimentation and peer-learning that, however, sit alongside ordinary administrative circuits. These infrastructures generate valuable knowledge, but they depend on project funding, a small team of motivated officers, and temporary task forces. From a transformative governance standpoint, the key question is therefore not whether and how cities learn but if they gain levers to renegotiate their position in multilevel energy and climate regimes. The gap between articulating a problem and changing the institutional conditions that generate it is precisely what Let’sGOv highlights as the core challenge for mission-oriented capacity building.

A further tension emerges at the collective level, where Mäntysalo and colleagues’ framework of agonistic communication [32] offers a useful interpretive lens. The nine cities did not converge on shared demands through consensus: the bench-learning process involved negotiation, contestation, and differential positioning among cities with unequal institutional resources and political mandates. The production of a joint policy brief, or the activation of the follower city network, should therefore be read as a provisional achievement of agonistic coordination, a working agreement among actors with divergent interests and capacities, held together by the shared pressure of the Mission framework and the project timeline. This reflects the structural difficulty of sustaining multilevel interfaces in the absence of formalised mechanisms, and it qualifies the extent to which the emerging network can function as a durable interlocutor with national institutions.

On the multilevel side, Let’sGOv seemed to be able to build vertical networks that were absent before: e.g., the policy brief on energy communities, one-stop-shops, and data access coproduced through the network were distributed and addressed to ministries and regulators, marking a tangible attempt to transfer local experiments into national governance; the activation of a followers city network and the explicit positioning of the nine cities as a collective entity and interlocutor suggested the emergence of a potential coalition around climate with negotiation power. In this light, the project can be read as an attempt to build “transformative capacity” [9] in a context where key decisions sit elsewhere.

This coordination was also aimed at preventing the risks of a polycentric governance, leading to dispersed experimentation often without orchestration, which might eventually lead to failure in both the local implementation and the multilevel grafting of measures. Nevertheless, the Mission Cities started to act as a multilevel intermediary [34], a national laboratory proposing to translate scattered local best practices into shared inquiries while creating accountability structures that link city pilots to national policy debates. However, the emerging of this informal negotiation infrastructure is precarious and project-dependent, it works under conditions of chronic underfunding and high regulatory rigidity [47], and it can now be tested against the lack of additional resources and beyond the pilot phase. Furthermore, the limited geographical representation of the nine cities might risk sending a message of exclusion to less organised territories, which are mostly located in the southern parts of Italy. Moreover, the reduced coordination and funding agreement might generate cities developing advanced narratives about transformative and reflexive capacity, while everyday work remains committed to compliance with national hegemonic regulations and short-term result achievements. This might result in missions becoming arenas where the responsibility for systemic failures is quietly shifted downwards: when cities cannot deliver, the problem is framed as lack of local capacity rather than as a misalignment between objectives and the structures that shape local action.

In this regard, the multilevel influence power of this network could be carefully curated and valued, building strong relationships around the three dimensions of governance to leverage national institutional reforms.

The project showed that planners and energy officers could use networked programmes to sharpen their diagnosis capacities (internal governance), to build coalitions across departments and across cities (internal–external governance), and construct more credible claims vis-à-vis national authorities (multilevel governance). The bench-learning materials, test bed reports, and policy briefs can give planners and policymakers specific and tailored language and examples to argue for internal reforms: e.g., the creation of climate coordination units, shared protocols for data access, standard models for governing energy communities, and one stop shops. On the other hand, the findings underline that without adjustments in legal frameworks, budgeting rules, and intergovernmental transfers, municipal departments will continue to be asked to “innovate” inside narrow boundaries.

This study has several limitations, most notably those related to geographic representation and the limited duration of the pilot. Concerning the geographic representation, as mentioned earlier, the nine Mission Cities participating in Let’sGOv are relatively large, administratively resourced, and politically committed to climate action. This selection reflects the composition of the EU Cities Mission’s Italian cohort, but it has important implications for the transferability of findings. We argue that the mechanisms identified in the study are potentially transferable to other municipal networks, including those comprising smaller or less-resourced cities. These mechanisms do not depend on city size per se but on the presence of motivated officers, prior engagement with Climate City Contract processes, and proximity to national and European policy networks. However, the enabling conditions under which these mechanisms operated in Let’sGOv may not be replicable everywhere. Smaller municipalities, and particularly those in Southern Italy, face greater fiscal constraints, less political continuity, weaker administrative capacity (fewer specialised staff, less access to EU project management expertise), and greater distance from the national networks and technical agencies that supported the Let’sGOv pilot. These municipalities are also less likely to have been involved in Climate City Contract processes or to have dedicated climate offices.

An important limitation concerns the time span of the pilot, too short to assess whether the small changes observed, such as formalised task forces or new data protocols, will be sustained or scaled. Many governance innovations documented in transition studies do not survive the end of project funding or changes in political leadership. Moreover, a fuller assessment would require linking the network level capacity building activities to concrete governance changes in specific pilots, tracing how shared diagnostics and tools are adapted or resisted in different local contexts.

Finally, the research findings are also shaped by the position of the authors as participant observer researchers involved in designing and facilitating the bench-learning pathway: this role enabled privileged access to internal conversations and emerging practices. While also inevitably influencing them, as the questions posed, the formats proposed, and the concepts introduced contributed to how cities narrated their governance gaps and possibilities. The analysis offered here should therefore be read as situated and partial, grounded in close engagement with actors and processes but also marked by the normative commitments and interpretative frames that come with such proximity.

6. Conclusions

The article has engaged with climate governance debates that call for reflexive and transformative approaches and has tested how these concepts can be applied to the everyday government and governance practice of Italian Mission Cities. The study has shown that building “capacity” is about opening decision-making spaces and challenging established routines to negotiate the structural constraints that shape climate-related municipal action. At the same time, it has highlighted the limits of reflexivity when such discussion spaces remain parallel to ordinary administration, precarious, and vulnerable to shifts in political leadership and funding.

Based on the empirical evidence generated through Let’sGOv, this article examines how municipal actors experience capacity building for ecological transition (RQ1) and under which conditions a networked bench-learning programme can contribute to developing governance capacities relevant to a national climate architecture (RQ2). Regarding the first question, the analysis shows that municipal staff did not experience capacity building as linear skills acquisition, nor as a technical juxtaposition of new competences or a way to solve bottlenecks. In fact, the persistent internal, external, and multilevel barriers were collectively diagnosed only to become more precisely articulated through peer-exchange and shared reflection. These findings suggest that the political interpretation of constraints became a precondition for interaction with the national level of government. However, the local, daily experience of climate work continues to rely on a few committed individuals, mostly due to the persistence of hard infrastructure, such as HR system, bureaucratic workloads, budget cycles, departmental mandates, and procurement procedures.

Regarding the second question, the observation of Let’sGOv bench-learning programme suggests that a networked programme of collaborating cities can begin to construct multilevel interfaces and collective negotiating capacity only if it creates a space in which the nine cities learn to speak as a collective interlocutor towards ministries and EU institutions, converting local observations into joint policy briefs and regulatory demands. This horizontal coalition-building can begin to fill a gap that Italian national climate governance has not addressed through formal mechanisms; however, it needs to escape its project-funding dependency and develop institutional mechanisms that ensure its continuation.

Future research should examine whether the follower city network activated by Let’sGOv develops its own governance dynamics or remains dependent on the original nine cities and how asymmetries between frontrunner and follower cities shape the politics of the network.