Co-creating justice in housing energy transitions through energy living labs

3.1 Methodological approach

The analysis focuses on four Dutch ULL cases from the Just PREPARE research project (https://justprepare.nl/), hereafter referred to as ELLs, as context-specific experiments in neighbourhood-scale housing energy transitions. The study adopts a comparative case study methodology embedded in an action research project, drawing on semi-structured interviews and field notes from ELLs’ group sessions. The comparative approach emphasises a processual and relational analysis, tracing dynamics across multiple contexts and analytical scales to identify patterns that enable theoretical transferability (Bartlett & Vavrus 2017). Investigating multiple cases enables systematic comparisons and data richness, providing a holistic understanding of social phenomena within their specific contexts (Gustafsson 2017) and supporting evidence-based strategies and actionable, context-sensitive interventions. The action research orientation reflects the situational inquiry of ULLs, where researchers actively participate to address local practical problems (Chevalier 2019). It aims to explain and improve systems by producing knowledge relevant to both theory and practice, through collaborative processes of joint problem identification, analysis, action and reflection (Collatto et al. 2018).

3.2 Case selection and data collection methods

The study focuses on four ELLs implemented in Amsterdam Zuidoost, Rotterdam Bospolder-Tussendijken (BoTu), Nijmegen-Dukenburg, and Gemert. These cases were purposely selected in line with the Just PREPARE project’s focus on Dutch vulnerable neighbourhoods undergoing sustainable heating transition and/or housing energy renovation, for which dedicated ULLs were established to engage local communities and institutions in advancing a just transition. Inclusion criteria were: a) programme readiness (ongoing or imminent heating transition/energy renovation); b) socio-spatial vulnerability (low-income households, inefficient housing stock, energy poverty); c) institutional access and community initiatives (commitment of municipal and housing actors to collaborate, and active resident organisations); and d) variation (diversity in built fabric, tenure mix and governance arrangements). All cases meet these criteria, albeit with varying salience.

As resident and stakeholder engagement is tailored to local conditions (e.g. ELL session formats, thematic agendas), research approaches and interaction strategies differ across sites. Variation is inherent to participatory transition arenas and is treated as a source of analytical insight, rather than inconsistency (Lukkarinen et al. 2025). Comparability is thus constructed through a common focus on ELLs’ initial phases, tracing procedural justice dynamics across diverse research–practice settings.

Table 1 specifies data collection methods and sources, differentiating primary materials analysed in this study from supplementary datasets from parallel studies, used to contextualise neighbourhoods and local energy transitions. For transparency and traceability, anonymised reference codes were assigned to interviews and field notes (Table S1 in the supplemental data online) cited in the findings.

Researcher positionality differed across cases: Amsterdam and Rotterdam were primarily covered by authors DR and CL, and Gemert and Nijmegen by authors SVW and PDK. As Dutch was the working language, authors CL, SVW and PDK, who are Dutch speakers, participated in ELL sessions as observers/occasional co-facilitators, whereas author DR remained non-embedded and led the cross-case synthesis without direct involvement in sessions. While embeddedness improves access and contextual understanding, it also risks role-related bias. Incorporating external analysts mitigated this risk through analytic distance and an embedded external double-coding approach (Ortloff et al. 2023). Interpretative claims were reviewed by all researchers, with discrepancies resolved through negotiated consensus. Field notes, taken during and after sessions without a formal template, combine descriptive chronologies with reflexive commentary. Dutch excerpts were translated into English and back-checked by bilingual authors.

3.3 Analytical framework and data analysis

The empirical data were analysed through an abductive, thematic approach combining inductive coding with a deductive overlay. The deductive frame draws on established ULL methodology guidelines, which outline key ULL components and levels (ENoLL et al. 2025; Zalokar et al. 2021), and on procedural justice considerations, constituting an overarching theme tracked across cases’ dynamics. The semi-structured interviews with ELLs’ facilitators offered insights into ELL structures, local challenges and stakeholders’ dynamics, further enriched by the authors’ reflective observations from direct involvement in ELLs’ activities.

The analysis resulted in three thematic clusters, representing aggregated dimensions derived from synthesising theoretical insights with inductive coding. This analytical framework reflects the LLs’ multiscalar logic (Schuurman 2015), which provides analytical differentiation between macro (organisational structure), meso (project), and micro (user activity) levels. Accordingly, the identified themes address both the strategic development of ELLs and their embeddedness within context-specific experimental settings and multistakeholder dynamics:

a) The first theme captures the temporal trajectory of ELLs and their initiation and early-stage evolution. As both the research and the local housing energy transition initiatives are ongoing, the analysis is limited to initial set-up and diagnostic activities. Although LLs vary in structure, previous studies typically outline three to five key phases: i) preparation and definition, including problem framing and scoping; ii) ideation and co-creation, where solutions are collaboratively developed; and iii) evaluation, involving testing and refinement based on stakeholder feedback (Bridi et al. 2022; Kowaltowski et al. 2024; Liedtke et al. 2015). This study does not address the co-creation of interdisciplinary solutions, as these are planned for later phases of the research project.

The thematic clusters b) stakeholder constellations and collaboration dynamics and c) ELLs as interface for local energy and housing issues also allow situating the ELLs within the ULL framing provided by the quintuple helix model (Almeida & Deutsch 2025; Carayannis et al. 2018). The multifaceted interactions between citizens, industry, knowledge institutions and municipalities (corresponding to the first four helices) were examined for their implications for energy-related practices and research. This underscores research embeddedness in real-world energy transition contexts, while countering extractivist approaches through reflexivity, mutual learning, and knowledge co-production (Glassman & Erdem 2014).

Theme c), in turn, engages more directly with the fifth dimension of the natural environment, following Almeida and Deutsch’s (2025) reading, in which it is reflected in the sustainability and liveability concerns of transitioning urban ecosystems. From this perspective, the analysis calls for critical reflection on ELLs’ role in mediating this process.

3.4 Contextualising the case studies

The four case studies represent the Just PREPARE project’s primary empirical sites (Figure 1), with the ELLs combining practical interventions with action research. Table 2 summarises the sociocultural framing of each case, alongside residential typologies and sustainable transformations. Shared vulnerabilities coexist alongside contextual differences, informing the development of scalable, justice-oriented strategies applicable to other neighbourhoods and municipalities.

Figure 1

4.1 Initiation and evolution of the ELLs

The four ELLs were initiated under the mediation of the Just Prepare research project, aligned with its timeframe but with different starting conditions reflecting variations in neighbourhood dynamics and institutions’ cooperation (Figure 2).

Figure 2

In Rotterdam BoTu, the ELL is embedded in two ongoing programmes: the municipal Resilient BoTu 2028, aimed at strengthening the community’s social, economic and physical conditions, and the national Natural Gas-Free Districts, supporting municipalities in decarbonising neighbourhoods through local pilot projects. This institutional anchoring provided legitimacy, resources, and a shared framework, but also revealed its limits, as trust-building required patience and careful negotiation (INT-F1-2-A/R). These dynamics highlight procedural justice challenges to resident influence in decision-making within a highly institutionalised setting. Yet, with trust established, the ELL operates as a recognised platform integrated in the neighbourhood’s actor network, structured around a co-developed ‘learning calendar’.

By contrast, the Amsterdam ELL emerged in a fragmented setting. Its initial linkage with a prominent resident organisation dissolved owing to capacity constraints, leaving the lab without a clearly demarcated framework, operating on a more ad hoc basis, shaped by the district’s fluid social and political conditions (INT-F1-2-A/R). Activities included workshops and thematic group discussions with municipal and housing representatives. While this adaptive approach reveals emergent priorities and enables flexible participation, it also causes fluctuations in scope (SES-A).

The Gemert ELL built on existing relationships within a rural care and well-being alliance. Local SHAs, already active in renovations and committed to the energy transition, sought to strengthen connections with knowledge institutes, which had been barely involved in the area (INT-F3GN). Conceived as a praktijkleerwerkplaats (practical learning workshop), or ‘spark hub’ for student–researcher collaboration on housing and care issues (SES-G), it was initially imagined as a physical hub with a cohesive group (INT-F3GN). Practical constraints shifted it into a flexible and open network-based model (INT-F5GN). Stakeholder turnover and limited continuity resulted in what participants described as a ‘restart’ of the ELL, prompting a re-evaluation of its purpose and composition, while fostering accessible and reflective forms of collaboration amid changing dynamics (INT-F3GN).

In Nijmegen, the ELL began without previous connections among local institutions, requiring new collaborations from the outset (INT-F3GN). Researchers pre-selected the Dukenburg district with a SHA, municipality, and social welfare organisation, already engaged in previous projects. However, finding suitable stakeholders took time: without established relationships, amid turnover and competing agendas, participation fluctuated, with some organisations often questioning the benefits of involvement (INT-F4N, SES-N). Over time, the ELL evolved from a discussion forum to a platform organising low-threshold activities to meet residents and (in)directly address energy issues, creating two levels of learning: how to engage residents so their voices were heard, and how organisations could collaborate more effectively (INT-F3GN, SES-N).

4.2 Stakeholder constellations and collaboration dynamics

All four ELLs engaged similar stakeholders (Table 3), with participation levels varying but consistently including housing and energy actors alongside resident-oriented social organisations.

Collaboration dynamics were shaped by both the types of ELLs’ activities and the strength of local networks. In Rotterdam, a ‘close-knit’ group developed solid relationships, with the municipality playing a strong leading role both in the energy project and the ELL sessions. A key ambition was to reflect on the social impact of local collaborations involving formal and informal institutions (INT-F1-2-A/R). Despite the intention of having sociotechnical expertise at the table, the collective need was inherently social. Stakeholders recognised that moving beyond formal responsibilities enabled cross-role interactions, fostering integrated approaches and collective ownership of the transition (Manktelow et al. 2023). For the SHA and the energy company, a recurring concern was demonstrating the social value of neighbourhood engagement to back offices, which typically demand quantifiable results (SES-R). Early trust-building by facilitators eased researchers’ entry and acceptance, with the ELL providing a softer landing in the neighbourhood (INT-F1-2-A/R). Researchers, in turn, facilitated a ‘learning calendar’ to collectively explore questions on collaboration, energy justice and the transferability of successful practices in other contexts, linking on-the-ground activities with organisational learning (see Figure 3).

Figure 3

Collaboration in the Amsterdam ELL unfolded through a fluid approach, as researchers continually sought alternative local actors and modes of engagement. Inconsistent participation prevented the formation of a cohesive multistakeholder group, reflecting the fragmented nature of energy governance in Amsterdam Zuidoost (SES-A). The lab repeatedly realigned after failed pathways, ultimately turning to SHAs as key actors. While this proved the value of ELL’s adaptive capacity, it also underscores the complexity of pursuing a just transition in a large district with unstable dynamics (INT-F1-2-A/R). Stakeholders experienced how the absence of a strong collaborative platform or long-term shared vision hindered consistent planning and meaningful resident participation. Meetings and workshops were often improvised, yet researchers’ mediation created space for explicit discussions of energy justice, though this seldom translated into concrete neighbourhood actions. Tensions between local social organisations and the municipality further revealed how contextual and political dynamics shape collaboration, residents’ representations, and the perceived urgency of action research (SES-A).

Gemert ELL adopted a core–shell structure: researchers, municipality, and the SHA forming the core, while technical staff, energy coaches and social professionals joined depending on the session’s focus (INT-F5G). Collaboration developed slowly and required several restarts owing to personnel changes, but the small core group provided stability, even though they prioritised practical, day-to-day issues rather than larger strategic questions (INT-F6G). Sessions with shell actors occasionally disrupted group dynamics and trust, entrenching thinking within one’s profession instead of addressing transcending issues together (INT-F3GN). Nevertheless, meetings were valued by stakeholders as spaces for reflection and ‘sparring’, with complementary perspectives between social and technical advisers (INT-F6G). By attending most ELL sessions, researchers connected with local professionals and community histories, values and concerns, while also using the venue to share preliminary findings. However, tensions emerge when interests misalign: when professionals seek prescriptive, quantitative outputs to address day-to-day issues or focus narrowly on resident behaviour while overlooking systemic factors. Such dynamics hinder the introduction of perspectives that contextualise resident (energy) behaviour or question professional assumptions, normative standards around heating, and institutional practices. The ELL thus supported mediation by making divergences explicit and opening them up for collective reflection (SES-G).

Nijmegen’s dynamics were more outward-oriented, with the ELL prioritising neighbourhood activities linked to residents’ values (health, nature, affordability) over structured reflective sessions. Its network expanded through snowballing, engaging smaller organisations, residents’ groups, local energy coaches and fixing services. Discussion depth and continuity were challenged as professionals frequently joined and left, with SHAs unevenly engaged, often preferring to be approached with fully formed plans rather than co-developing ideas (INT-F4N). The social work organisation became more active once the local energy transition turned contentious, and as the ELL shifted on organising activities. This pragmatic orientation is appreciated for offering residents relational events, facilitating engagement with daily concerns; however, it limits deeper reflections or cross-organisational exchange (INT-F3GN). As in the other cases, the ELL brought researchers closer to lived realities. A PhD researcher initially spent one day a week at the municipal energy consultation service in the local shopping centre, gaining insights into residents’ concerns and district dynamics as people stopped by spontaneously during shopping trips. The subsequent closure of this location raised reflections on how being physically there makes it easier to connect to the neighbourhood and share energy information, while also emphasising the researcher’s position as an ‘outsider’ (SES-N). More broadly, researchers gained a deeper appreciation of the complexity of conducting research on politically sensitive transitions, where navigating tensions requires careful coordination and trust-building.

4.3 ELLs as interface for local energy and housing issues

The ELLs were established in contexts with local commitment to energy renovation and transition, where projects were either already underway or planned. Their shared ambition to support these transitions by fostering resident acceptance and integrating their needs provides cohesion and orientation for ELL activities. Yet the Amsterdam case illustrates how the crisis of a ‘unifying big energy project’, combined with the area’s scale, challenged inclusive participation (INT-F1-2-A/R).

Across cases, the ELLs opened channels for direct and indirect engagement with neighbourhood realities and for justice-oriented discussions, sensitising stakeholders to local needs and to collaboration values. Where the energy transition project was perceived as successful, as in BoTu, the ELL focused on understanding how and why the coalition functioned effectively. Although its pilot character, high costs and context-specificity limit replicability in other vulnerable settings, certain collaborative practices can be effectively incorporated into organisational routines and transferred elsewhere (INT-F1-2-A/R). To this end, BoTu stakeholders co-developed a learning calendar addressing both the layered sociotechnical impacts of renovation and ways to sustain collaboration beyond the project’s lifespan. They emphasise both structural supports (e.g. consistent neighbourhood meetings, sufficient funding) and interpersonal qualities such as empathy and role flexibility, which help hold the coalition together and advance the energy project on schedule (SES-R). A central challenge for the ELL was reconciling municipal decarbonisation objectives with residents’ everyday concerns, such as mould and safety. Stakeholder stress ‘meeting people where they are’, noting that resolving immediate issues, such as housing maintenance or billing problems, contributes to broader impacts, including long-term trust and meaningful engagement (INT-F1-2-A/R, SES-R).

As in Rotterdam, where bilingual energy coaches and the local social organisations played a strategic role in reaching residents ‘behind the front door’, the Gemert and Nijmegen ELLs relied on mapping exercises to gain a nuanced understanding of the local context (see Figure 4). These exercises combined quantitative data, such as energy labels, income levels and housing typologies, with the situated knowledge of professionals from the social domain, SHAs and municipal planning teams. Beyond producing an integrated picture of technical and social conditions, the process is generative: it allows participants to exchange perspectives, identify suitable locations for household-level fieldwork, and highlight areas of diversity (e.g. social housing alongside low-income homeowners, mixed generations, and culturally diverse residents living in similar apartment blocks) (SES-G, SES-N).

Figure 4

In Gemert, ELL researchers also collaborated with a municipal initiative that provided small-scale energy fixes to low-income households, allowing researchers to accompany mechanics on home visits and gain valuable insight into residents’ everyday energy practices (SES-G). Similarly, the Nijmegen ELL organised low-threshold activities (in)directly related to the energy transition, with careful attention to location, themes, and participating organisations (see Figure 5). ELL stakeholders in Nijmegen and BoTu work with asset-based community development (Kretzmann & McKnight 1996), which assumes that activities should be rooted in the neighbourhood’s organisational capacity, needs and interests. So, rather than starting from the energy transition, a theme often distant from residents’ priorities, the ELL built on existing initiatives and concerns such as love for nature, health, social cohesion and living costs.

Figure 5

However, discussions remain on whether the ELL risks projecting its own values and interpretations onto residents, and on how the absence of governance institutions limits the extent to which residents’ voices influence decision-making processes beyond the ELL (INT-F4N).

In parallel with practical outings, reflective discussions unfolded across the cases. As examples, Amsterdam ELL organised a thematic group explicitly focusing on energy justice. Starting with its normative conceptualisation of distributional, procedural and recognition justice allowed stakeholders to collectively explore their situated meanings in respective context(s), developing a shared understanding of the concept before moving to practical implications and interventions (INT-F1-2-A/R). An initial normative framing of energy justice thus appears a necessary and strategic step towards meaningful operationalisation of justice-in-practice in urban energy transitions (Ricci et al. 2025).

In Gemert sessions, recurring themes of tenant ownership and participation indirectly bring justice issues to the table. Stakeholders tended to frame ownership as tenants taking initiative in the energy transition. However, only idealised or large-scale forms of ownership and participation appear valued; smaller, practical actions by tenants, such as modification to internal partition walls or turning off the noisy ventilation system, are often overlooked as a sign of ownership. Discussions also raised questions of resource distribution (e.g. energy fixers, subsidies for efficient appliances), who should be prioritised for support, and whether professionals should assume full responsibility or encourage self-reliance. Additional concerns include privacy regulations limiting information-sharing and interorganisational exchanges (e.g. energy coaches unable to report social issues), the rising cost of solar panels and residents’ reluctance to adopt them, and whether the ELL genuinely fosters a just transition or risks remaining a symbolic forum (SES-G).

5.1 Supporting procedural justice through innovation and learning

In all four cases, the ELLs operate predominantly as social laboratories, focusing on multiple complex energy transition challenges: actor interests, housing quality, energy practices and vulnerable group involvement. Unlike conventional, technology-driven LLs that prioritise product and service development, the ELLs emphasise socio-spatial considerations and collective perspectives, thereby advancing social innovation (Franz 2015; Hoppe & De Vries 2019; Scholl & Kemp 2016). The innovation of the ELLs lies in complementing local housing decarbonisation plans and projects with collaborative networks committed to community-sensitive governance (Blezer et al. 2024) and energy justice, echoing Almeida and Deutsch’s (2025) argument that ULLs must embed ecological and social dimensions within urban innovation. This integration not only supports just transitions (Sovacool et al. 2023) but also helps overcome renovation bottlenecks in technical planning through trust-building, stakeholder information exchange and context-sensitive understanding of residents’ building practices (Prieto et al. 2023).

The ELLs effectively served as arenas for reflexive self-evaluation, whether individual or collective. Through the regularity of sessions, self-evaluation evolves into iterative process learning (Bridi et al. 2022; Voytenko et al. 2016), making continuity in participation and turnover critical factors in ELL implementation. In particular, the Rotterdam ELL was deliberately set up to learn and evaluate how collective visions could inform local energy transition projects, with stakeholders regularly revisiting their action methods to better align them with neighbourhood dynamics. This process learning, situated and experiential, underpins more transformative learning outcomes (Evans et al. 2021; von Wirth et al. 2019), at the organisational and interorganisational scale, by stimulating ‘deeper’ or ‘higher-loop’ learning with continuous reflection on assumptions, power dynamics, and established strategies (Costa et al. 2022; Stenberg 2018). Activating such critical reflections in professionals helps counter single-minded views of residents’(energy) practices detached from their social contextual dynamics (Raven et al. 2021).

ELLs’ session themes further demonstrate the importance of making energy justice tangible. By addressing its dimensions explicitly or implicitly (e.g. Amsterdam, Gemert), energy justice emerges not as a predefined normative ideal but as a situated concept, embodied in practice and refined in relation to local realities. In this way, ELLs help to keep energy justice ‘in the loop’, fostering attentiveness and sensitivity to how it is continuously made and remade across settings, rather than assumed as fixed or universal application.

5.2 ELLs as relational arenas for enabling participation

In most cases, the ELLs set-up fostered a trusted network of municipalities, practitioners, researchers, and residents or their representatives. Despite limited direct resident participation in sessions, the involvement of social organisations safeguards the integration of local perspectives, supporting ELLs as intermediaries (Bouzarovski et al. 2023) while easing participation burdens for vulnerable groups (Gillard et al. 2017). This relational function reflects procedural justice by promoting residents’ visibility and co-creation of shared values and responsibilities (Bouzarovski et al. 2023; Bridi et al. 2022; Shejale et al. 2025). It lays foundations for reciprocal practice–practice, practice–science and institutional–lived realities support. Informal gatherings in Nijmegen and home visits in Gemert enabled direct interactions with residents in their sphere of proximity and daily life, while Rotterdam sessions on social impact stressed the value of cross-level and interdisciplinary collaboration.

The empirical analysis remains limited in tracing whether these relational dynamics and the interorganisational learning they foster effectively translate into equitable and inclusive actions in energy renovation practices. Given the multifaceted nature of these processes in vulnerable contexts (Tozer et al. 2023), parallel validations and monitoring are required.

‘Creating a good setting for participation’ is confirmed as essential for operationalising procedural justice and sustaining it throughout the process (Ricci et al. 2025). An effective strategy was to connect the ELLs to neighbourhoods with active stakeholders and initiatives rather than starting from a blank slate. This approach encourages participation through informal, spontaneous interactions rather than imposing it. The ELLs deliberately started with relationship-building and careful navigation of tensions, ensuring fairness and transparency in process initiation. The Amsterdam ELL, despite similar intent, demonstrated the challenges of achieving meaningful engagement in an extensive district lacking a cohesive governance structure and institutional stability (see Van Breugel et al. forthcoming). This underscores the critical role of spatial scale (Voytenko et al. 2016), suggesting that dedicated, project-scale ELL can usefully complement broader district-level ULLs in supporting vulnerable residents with technical choices, energy- and comfort-related issues. In this regard, the timing of ELL set-up also emerges as relevant: in all analysed cases, ELLs were established either after or in parallel with the implementation of energy-efficiency measures, rather than during the early decision-making phases, limiting the capacity of residents’ inputs to shape core decisions. This risks shifting participation towards acceptance of predefined solutions than their co-creation (Bridi et al. 2022). Nonetheless, the four cases show that, even if ELLs appear ‘late’, they can still foster process learning to support ongoing or future interventions, as seen in Rotterdam ELL. However, critical reflections are needed, as urban experiments often face the pilot paradox (Metze 2023): their value creation and activities frequently remain confined to (research) project boundaries/timeframes, hindering long-term impact.