The breakdown of the municipality as caring platform: lessons for co-design and co-learning in the age of platform capitalism

ABSTRACT If municipalities were the caring platforms of the 19-20th century sharing economy, how does care manifest in civic structures of the current period? We consider how platforms – from the local initiatives of communities transforming neighbourhoods, to the city, in the form of the local authority – are involved, trusted and/or relied on the design of shared services and amenities for the public good. We use contrasting cases of interaction between local government and civil society organisations in Sweden and the UK to explore trends in public service provision. We look at how care can manifest between state and citizens and at the roles that co-design and co-learning play in developing contextually sensitive opportunities for caring platforms. In this way, we seek to learn from platforms in transition about the importance of co-learning in political and structural contexts and make recommendations for the co-design of (digital) platforms to care with and for civil society.


Introduction
Municipalities can be seen as the caring platforms of the 19-20 th century 'sharing economy'. They were structures created by European societies concerned to manage resources and redistribute finances in equitable and effective fashion, for the public good. These platforms were seats of local determination, provision, administration and mutual solidarity (Schuyt 1998), voted into being by the citizenry of the area to meet communal needs. A different impetus now dominates. We are seeing the meeting of network economics and neoliberalism in 'platform capitalism' (Srnicek 2016), a blend of novel infrastructure and politics that challenges existing models of socio-economic engagement to produce global financialised monocultures. Instead of investing in civil society through the older structures of the municipality, neoliberal politicians have begun to downsize the State, replacing politics with metrics and allowing the free market to dominate. The growth of a significant public sector in many parts of Europe (and beyond) has professionalised and ultimately hidden the co-created and concerned nature of the undertaking behind our well-established civic platforms. These changes have introduced the opportunity for (digital) business models alongside public sector services, as well as peer-to-peer exchanges of service and service delivery mechanisms based on the close collaboration between the public sector and civil society, i.e. co-production.
The notion of co-production was first coined to highlight how informal collaborations between public service providers and citizens can improve the quality of public service provision (Ostrom 1996). During the last decade, co-production has been developing as an explicit strategy that aims at directly engaging citizens, and, more generally, third sector actors, in delivering and maintaining the public good (Brandsen and Pestoff 2006). With changes in funding and political will, uncomfortable tensions have developed in the rhetoric and practices of including and enabling citizens; co-production processes are cited both as a tool of empowerment and as a means of making harsh austerity agendas operational (Voorberg 2017). The UK's Big Society initiative (c. 2010) was a case in point: 'Voluntary action is valued in the rhetoric, and deprived of funding in practice.' (Barker 2012).
In this context, we might expect relations between the public sector and citizens to be in flux too. Through an analysis of how platforms for the public good are constructed and evolve in different geographical, policy and cultural contexts, we show how care can (but do not necessarily) emerge from co-design and co-learning efforts.

Two welfare systems embracing co-production
Different welfare state models entail differences in the relationship between the public sector and civil society. In this paper, this plays out in how they embrace co-production as a strategy for generating and maintaining the public good. Sweden has an inclusive 'social democratic welfare state regime' (Esping-Andersen 1990), while, particularly in England, there has been tight central control and erosion of public service provision (Leach et al 2018).
In the UK, co-production has been mainly a strategy to reduce welfare costs. As cuts and diminishing municipal authority have left each council holding a range of duties as provider or facilitator, but insufficient budget to meet many former responsibilities, the expectation is that communities will rally to take over services. Yet, recent cuts in agencies funding civil and community initiatives have also damaged these third and voluntary sector activities (cf Civil Society Futures 2018).
Like the UK, the Swedish public sector has been strongly influenced by managerial and private sector logics in the last 30 years (Ivarsson Westerberg 2014). However, this has not entailed a progressive erosion of the role of the public sector, which still delivers welfare services even though in the frame of a strict economic efficiency logic (Johansson, Arvidson, and Johansson 2015). Private actors and civil society who traditionally have been portrayed as peripheral in the Swedish welfare system (Esping-Andersen 1990;Kumlin 2001) are still playing a marginal role (Johansson, Arvidson, and Johansson 2015), but, in the last years, co-production is gaining momentum as a way to tackle complex societal issues. 1 3. Care, platforms, co-design and co-learning

Care
Structures that support the welfare of citizens are what we here call caring platforms and we argue their design can play down or stress caring features. We define care as 1) an active response to others' circumstances, 2) a degree of passion or enthusiasm, 3) an underlying relation of mutual habitation (after Puig de la Bellacasa 2012). Thus, care is something that pre-exists intervention, but which can be supported through system design that privileges relations of reciprocal accountability and mutual commitment and which encourages reflexive engagement among citizens (caring). Alongside this more general set of definitions, we note that individuals can care for something/someone (i.e. look after) and care about something/someone (i.e. feel passionate towards), but the two do not have to coincide. In other words, it is possible to have care duties but no interest in conducting them, or to feel strongly on an issue and have no influence. Only caring about necessarily relates to one's sense of what matters or is meaningful.
Our particular concern is an ethos of shared and collective responsibility for each other and how that is expressed or discouraged through the mechanisms of citizen-state interaction and, particularly, co-designed interventions. Welfare structures have developed as a way of enhancing collective and mutual care within societies (e.g. Avram et al. 2017). Yet, this grounding ethos has been progressively forgotten, becoming invisible in public sector structures and regulations (Schuyt 1998).
A last observation, that 'an ethics of care cannot be about a realm of normative obligations but rather about thick, impure involvement in a world where the question of how to care needs to be posed' (Puig de la Bellacasa 2017, p.6), points to our method here of being part of the action, rather than the research team watching developments from afar. In attending to two very different contexts of action, we may seem to be asking two different questions, but what links our analysis is care for how the public good is ensured, from being heavily managed moment-to-moment by state infrastructure to becoming, through co-production, something emerging through new forms of relationships between the public sector and citizens. Giving or taking the power to cocreate instruments of management, to determine what care can be administered and how care is understood, is then, itself, a primary act of care for the public goodi.e. care can be exhibited in building structures for care, seen in the efforts, described here, to negotiate the socio-economic structures of local government and in offering or taking the opportunity to co-design. In this paper, we attempt to show how this can be undertaken by a caring municipality or self-administered by a reflexive citizenship.
We give examples of how co-design allows for care to be co-produced, noting that the actions of municipalities have a bearing on the different possible collaborative constellations involving citizens and/or whether it is even possible to run initiatives that are self-organised.

Platforms
Gawer (2009) suggests that a digital platform 'acts as a foundation upon which other firms can develop complementary products, technologies or services' (p2). Jegou and Manzini (2008) define enabling platforms 'as a system of material and immaterial elements (such as technologies, infrastructures, legal framework and modes of governance and policy making)' (p. 179). Thus, platforms are relational and infrastructural (Star and Ruhleder 1996) and we understand them here as the sociotechnical infrastructure-supporting welfare. In as much as a platform is a support, it is a form that, digital or not, seems well suited to provide care and this informs our understanding of caring platforms in the context of municipalities and civil society as sociotechnical structures that can be designed to administer care or to promote the articulation and development of a care that pre-exists explicit manifestations. We also note that platforms can support platforms in a system of interdependencies.

Co-design and co-learning
Devising services together is co-design. Designing and delivering them together is coproduction (Coote et al 2010). Co-design, when applied to co-production, goes beyond the simpler types of design participation (that focus on outcomes) to include stakeholders in conceiving of issues and processes. Figure 1 shows the degrees of engagement that different commitments entail, with a transition between user-centred design, where users inform outcomes, and actual co-design, where some element of the process and/or outcome is at stake. More advanced levels of participation stimulate shared commitment around the issue and process at stake, often leading to co-ownership (Light et al. 2013;Seravalli 2014). Figure 2 links codesign to mutual learning. Mutual learning is discussed as a rationale (Simonsen and Robertson 2012) and a methodological stand (Bratteteig et al. 2012) in participatory design, meaning that participants' learning is seen as something emerging from and through co-design processes. By introducing multiple stakeholders, co-design processes not only bring together different knowledges, but also create opportunities for collective articulation and mutual understanding.
We distinguish mutual learning, a welcome product alongside a design activity, from the notion of 'co-learning', which we understand as a collaborative effort explicitly aimed at creating the conditions for learning together and in which collaborative making, when present, is instrumental to that learning (e.g. Boys 2017, in DiSalvo et al. 2017). We understand co-learning as an explicit and structured collaborative reflective process (see Figure 3). Co-learning, then, is related to co-research approaches, such as participatory action research (e.g. Senge and Scharmer 2006, on communities). However, while design research (Rodgers and Yee 2014) and action research (Reason and Bradbury 2006) stress coming into a situation to change it, co- learning is rather about making collaborative learning opportunities that attend to a specific issue and/or situation through the co-design of and participation in informative encounters and reflective engagement opportunities.
Literature focusing on citizens' participation discusses collaborative learning as a key characteristic of higher levels of participation (Pretty 1995;Collins and Ison 2006). When engaging in collaboratively defining and responding to issues, participants are learning together about the issue at stake, as well as about each other's perspectives and positions (ibid.). Collaborative learning might lead to new, shared understandings, but also it changes participants' positions and reciprocal relationships. Whether collaborative learning is at play (and how), thus, becomes key in determining the quality of participatory processes (Collins and Ison 2006). From a co-design perspective, this raises the importance of paying attention to if and how mutual learning is at play in codesign processes (DiSalvo et al. 2017), but it also means we must consider if and how co-design might be instrumental to collaborative learning.
Through the cases, we further articulate how co-design, co-ownership and colearning might play a role in the (co-)design of caring platforms.

Methodology: design research and case study analysis
The authors were involved in the cases described as (co-)design researchers, adopting a research through design approach (Koskinen et al. 2011), in which the direct engagement in design processes is used to generate academic knowledge. In both cases, knowledge creation closely involved the people we were working with, who became engaged not only in the actual co-design processes, but in reflecting on the unfolding and outcomes of these processes.
We use these studies to learn from platforms in transition. They explore how codesign, co-ownership and co-learning can be a central tenet in developing caring platforms with, and in spite of, local government. In line with Flyvbjerg (2006), we do not claim that comparing the studies produces universal knowledge about the differences between contexts. Rather, we use our examples to articulate contextdependent knowledge about the (co)-design of caring platforms in these two different places. More detail about the processes in each study is given in the descriptions below.

Case study one: ReTuren, breakdowns in caring platforms for waste prevention
The first case is the Swedish waste prevention municipal service ReTuren, co-designed to encourage citizens to reduce the amount of waste they produce. The case shows how alliances between the public sector and civil society can establish platforms for caring about waste prevention. It also reveals how existing structures and public sector attitudes may hinder the flourishing of these platforms.
Waste management in Sweden is regulated by EU and national policies, but it is on a municipal level that services are developed and driven. Consequently, Swedish municipalities are responsible for promoting waste prevention among citizens (Naturvårdsverket 2016). Waste prevention represents a break in traditional waste management. Municipal waste departments struggle in delivering waste prevention services, since they are 'locked in' by existing waste management infrastructure; lucrative business models around waste handling; lack of confidence in being able to deliver waste preventions services; legal and economic frameworks (Svingstedt and Corvellec 2018). Some of these hindrances emerge in looking at ReTuren's history.
ReTuren has three functions: a service for waste disposal; a shop where people can exchange used things for free; and a workshop to repair and upcycle things. ReTuren was set up by the municipal waste department in collaboration with one of the authors and a local makerspace (an NGO). Later on, it involved residents and initiatives from the neighbourhood; civil servants addressing local area development; and the regional company working with waste processing. The service had a pilot period (2015)(2016) and then went through a major reorganisation that led to its actual organisational model (2017 on), where the cultural department has a leading role (Table 1).  The department's ambition in setting up and running the pilot was to co-design the outcome (i.e. the service); however, along the way, the process and the issue at stake (i.e. how to promote waste prevention) was also co-designed. The co-design process focused on the involvement of local residents and the progressive refinement of the service in relation to local conditions (see Seravalli, Eriksen, and Hillgren 2017). Rather than developing a full concept for ReTuren and then proceeding to its implementation, the process started from a loose framework. By prototyping functions and activities, the department progressively refined the service with its users while running it. With the researchers, ReTuren staff developed four kinds of co-design actions.
The first action entailed organising different public events to test possible activities, as well as refine the purpose of the service in collaboration with residents, local initiatives and civil servants working with local area development. One such event was the 'Colourful Notes' festival, involving local schoolchildren, the neighbourhood department, ReTuren and a recording studio ( Figure 4). Children repaired and upcycled old pianos, which were placed in the neighbourhood square for one month and used for a number of organised and spontaneous concerts. The festival sought to reclaim the square, often used for criminal activities, as space for all the people in the neighbourhood. The festival became a way for ReTuren to develop relationships and reach out to residents that otherwise would have been difficult to engage. It also revealed how waste prevention aspirations and local concerns could be synchronised. Alongside bigger events like the festival, ongoing initiatives, like open workshops for repairing and upcycling, were organised. This allowed staff to prototype and refine the service with residents, as well as learning how waste prevention could be promoted in that specific neighbourhood.
The second action was engaging residents in refining more mundane features. For example, users got the opportunity to decide ReTuren opening times by voting.
A third action was the involvement of local collaborators in the ongoing evaluation of ReTuren. Input from users and residents was gathered through questionnaires and conversations in the everyday running of ReTuren and after activities and events. Regular meetings were organised with key actors (like NGOs representatives, civil servants working with local area development, staff from the local library) to discuss shared initiatives, common issues and the development of the service.
Last, seven months after beginning the pilot, three workshopsorganised by the codesign researchersbrought together the managers of the different organisations. These workshops aimed at collaboratively developing a strategy for the continuation of ReTuren after the pilot.
The involvement of residents and local actors in developing, deciding on and delivering the service led to a strong sense of co-ownership over ReTuren, which became apparent when a major breakdown hit the service.
Ten months after its official opening (and after some disagreements between ReTuren staff and local gangs) the waste department managers decided to terminate the service. The managers recognised ReTuren's value and achievements. However, they also highlighted that the department did not have the competences to run a staffed service. And legislation was unclear about to what extent waste fees could be used for waste prevention initiatives.
Everyone else responded strongly to this decision. Residents protested to the waste department and collected signatures that were sent to politicians. Eventually, local civil servants managed to convince the waste department managers to engage in a collaborative process to redesign the organisational model of ReTuren and redistribute responsibilities (with the cultural department taking over main responsibilities regarding staff and workplace safety).
The process of reorganising ReTuren was made possible by the strong sense of coownership that had developed around the service. It engaged ReTuren staff, ReTuren project manager, local civil servants and co-design researchers, as well as the managers of the different organisations. In this process, alongside looking for a shared solution, there was also a major effort made in co-learning about the different actors' possibilities and constraints.

ReTuren as caring/cared for platform
The upcycling station can be considered a caring platform that supports the emergence of collaborative ways to care about waste (prevention), by engaging civil servants and residents. It has achieved this by nurturing passion and enthusiasm among individuals and by supporting mutual understanding of the interdependences that exist between institutions and individuals in aiming towards waste prevention. In a nutshell, it highlights how waste prevention requires shared effort from municipalities and citizens and that that effort requires care from both sides.
The service's journey also reveals how ReTuren was deeply caredabout by different interests. It is notable that those less engaged with the development processes brought other concerns to bear in deciding its future. Closing down a service about which many professionals and residents had just come to care deeply could be seen as an act of extreme bureaucracy: the very absence of care. Yet, the decision to terminate came from managers being deeply concerned about ReTuren staff safety.
The aftermath of the threatened service termination reveals the different forms of care at play. Finding a way to progress past these obstacles to revive the future of the platform shows both sophisticated negotiation and an embrace of caring, whether out of respect or pragmatism. It shows the wider platform of the municipality functioning at a local and nuanced level to support, care for and care about broader sustainability initiatives and citizens' will.
In the Swedish context of a powerful public sector, ReTuren is able to challenge the idea of domineering authorities, recognising the need to engage citizens and their competences for achieving societal goals. Yet, the decision to terminate the service by the waste department reveals a further need in pursuing caring platforms: of providing municipalities (not least waste departments) with frameworks and competences that allow them not only to experiment with closer relationships with citizens, but maintain them over time.
5.1.2. Co-designing platforms, co-ownership and co-learning The example above reveals how co-design can nurture care by fostering appropriation and co-ownership through involvement. Yet, the near-termination of the upcycling service shows the limits of the way co-design was applied in this case. The managers were reasonably updated about what was happening at ReTuren, but they were not involved in learning among the people engaged in the everyday running of the service. The co-design process did not succeed in making the managers aware of the strong commitment of civil servants and citizens to the service.
In tracing how learning developed throughout the pilot, the researchers found that it was connected to the development of caring practices about waste prevention, but not to how such practices related to the different actors' organisations. Co-learning developed in ad-hoc moments, very much in the everyday interactions and collaborations among the people involved on the ground. The occasions on which people at different levels within the organisations sat together to develop a strategy for ReTuren were not enough to learn about each organisation's possibilities and constraints.
6. Case study two: grassroots co-learning platforms in the post-public sector In the study above, the municipal platform is negotiating its stake in a platform coproduced with residents. In Britain, much co-design is in response to recent cuts in local government (Coote 2010). The two-part study, below, features community initiatives that grew in the absence of larger-scale caring platforms. Instead of collaborating with governmental institutions, some civil society is producing its own platforms, scrutinised here as part of academic-community research into the effects of reduced public sector support (Table 2).

2011-12
Study of groups co-designing workshops to share their codesigned grassroots cultural heritage projects.

2012-13
Study of groups making internet radio to share group process and learning.

2013-2018
Several grassroots co-design projects come to successful fruition, bringing in mainstream support.
Between 2010 and 2014, councils were made to cut their budgets by 48% and the cuts are still coming, affecting support for civil society initiatives (Civil Society Exchange 2018). One problem with removing formal infrastructure is an absence of initiatives to help communities learn from each other and spread insights (e.g. Botero et al. 2016;Light 2019). The work reported here explored how lessons about co-design could be shared, using co-learning processes.
The first part of the study explored grassroots co-design activity across four areas of England, bringing social activists together to share knowledge and understand how their context impacted on their ambitions and methods of engagement. Each group was already engaged in making a change in its community, envisaging a future through place-shaping, a practice defined as 'the creative use of powers and influence to promote the general well-being of a community and its citizens' (Lyons 2007, 3). 2 The collaborative project ran for a year in a bottom-up fashion, informing on grassroots cultural heritage work (see Light 2018). The style of co-design in the project involved issue, process and outcome (Figure 1), though the outcome was to be shared learning, not a material design. Having secured interest from local groups in four regions and won funding, the participating academic researchers handed over the design of a workshop in each area to a local organising committee linking community ventures. At each event, local civil society 'hosts' presented examples of their activities to a regional audience with visitors from the three other areas, then brought everyone together to consider local practices and priorities.
This produced four distinct orientations to the local municipality: in the multiplydeprived ex-mining area of East Cleveland, local authority representatives explored development opportunities with participants; in Sheffield, a city reinventing itself after the loss of its steel industry, a council arts officer showed people regeneration initiatives; in London, activities were ignored at the local authority tier, while, in Oxford, the work was in conflict with the city council, challenging the authority's view of what city priorities should be. This pattern reflected the local economics, with high unemployment characterising the first context and a long pedigree of ancient colleges and tourism giving prestige to the last (Table 3).
Both the platforms the groups imagined/adopted and how learning between groups played out are relevant here.
If we take a broad view of platforms as support infrastructure on which communities can build, these included a bombed sea-jetty turned into a tourist destination (successfully completed after a long campaign in the neighbourhood when the local council found sea defence funding for it: http://www.gazettelive.co.uk/news/teesside-news/ sealed-off-skinningrove-jetty-new-lease-8917601).A digital media archive, which later acquired supermarket sponsorship (see Light 2018); and a closed boatyard, threatening the wellbeing of Oxford's houseboat dwellers, which was eventually reinstated as part of a community asset deal with a property developer. All became an important part of local place-shaping, from an imaginary of what the place could become to the development of a new platform. In all these cases, prolonged bottom-up activity yielded support from other bodies sufficient to the size of the platform. The learning in these workshops was various. Most people had never visited the locations before and did not know about the activities being described, so the planning groups learned to incorporate increasingly detailed visits to inform visitors experientially (see Light 2018). For instance, in the first workshop, everyone sat in an ex-minemanager's house to discuss presentations. By the last workshop, in Oxford, participants travelled and slept on boats. Being on the boat gave insight into the dispute, the slow travel to other boatyards and the boarded-up boatyard at the centre of the conflict.
Further, the act of going to see related, but different, arrangements to develop local cultural heritage raised informative similarities and differences. People reflected more lucidly on their own ambitions through comparison with others' campaigns. And each small highly focused initiative could relate their practice to a bigger trend.
Feeling for others' situations was also a discernible outcome of meeting this way. The visitors from East Cleveland went to lengths to offer the Oxford boatyard campaigners their support on hearing of their battle, assuring them that if they had known about the campaign they would have joined forces. The subsequent presentation by the Oxford group included the national media coverage the campaign had received, leading the visitors to reflect, apologetically, that they had probably seen the news with no interest in the campaign until they knew the stakeholders. Being together created solidarity and concern for each other's circumstances, as well as an understanding of the importance of each platform to the local scene.
Based on these insights, a second project asked how people in different cities might use community internet radio as a platform to reflect on their own and others' achievements and learn from each other. Could this concern across contexts be mediated?
In this second project, three broadly-spread community groups were each given a deadline and sufficient funding to make a professional 15-min radio programme to share with other groups, capturing their purpose, issues and achievements (see Light et al. 2013). Here the co-design was of issue and outcome by the participating communities, and the process was constrained by the funders (in this case, academic) to include making a radio programme. All other aspects were open to choice by the community group and its leaders ( Table 4).
The first research question, as to whether any programmes would be made, was answered when each group competently met the deadline and made a programme (https://howwemadeithappen.org/): A craft group of older people spoke about the value of coming together to do crafts; two groups of museum volunteers interested in the oral history of their villages shared practices; a newly-formed women's group talked about the experience of being mothers and daughters. On evaluating their productions, the researchers (including civil society organisers) found the principal benefit to be the bonding that came from thinking about identity as part of planning and recording the programme. People enjoyed the process, got a sense of achievement from it and developed a sense of ownership (Light et al. 2013). The next question was whether others would engage with the programme beyond the producers themselves. It was found that the two other groups who had shared milestones and adventures through the project enjoyed the material. However, the study did not produce evidence that anyone beyond the participating groups would listen to the outputs. Although there was a broadcast, a website featuring the programmes (i.e. podcasts) and archiving with the UK's Community Media Association, there was no audience. Understandably, perhaps, little motivation to listen could be generated where there was no existing stake.
Again intimacy underlay interest and, without it, the broadcasts were just noise in a busy media landscape. Using the platform of internet radio was not enough, on its own, to act as an intermediary; it needed promotion, critical mass, a change in practices that would lead groups to seek out other groups, etc. Although we can contrast this with the success of other online learning platforms (Instructables.com, Transition Towns, etc.), the finding reminds us that lack of support for bottom-up co-design can leave each initiative isolated, whereas even a little resource for a connection could lead to mutual care.

Caring/cared for platforms?
The study above shows civil society leading larger organisations to support platforms that change local fortunes. Although Oxford City Council eventually stipulates that there must be a boatyard among the apartments that the property developer builds, it is only after nearly 10 years of impasse, protest and legal challenge. It takes the advocacy group in East Cleveland almost as long to influence the local council to provide resources for the jetty's restorationthen it comes from sea defence budgets, not an economic improvement. A lot of caring has gone into these initiatives by the time they become achievable, giving them an extra force in place-shaping. They harness local feeling and their success contributes to new meaning to the area. These are cared-for and cared-about platforms, as well as platforms for caring.
But we can see another form of care emerging in these narratives, generated by working together to share learning and recognise the value of one's activities. Especially where there are fewer societal benchmarks and support structures, such sharing and appreciating become part of what sustains volunteer effort and organisational persistence. It is this type of care that researchers sought to support and extend by introducing a platform for sharing practices in the shape of internet radio, a medium with low technical and cost barriers.

Co-designing platforms, co-ownership and co-learning
The academic researchers helped social activists learn from each other and used the colearning amongst those practising collaborative bottom-up design to glean insight about their processes and how they could be shared. Importantly, the chance to explain plans was enabling; it supported the choices made later in other projects and campaigns and gave extra conviction and vision to those telling their stories. This transitory platform of exchange was valuable for sharing insights about everything from relations with local authorities to apply for funding. It engendered caring among the activists in ways beyond those anticipated. It revealed how much emotional labour is involved in pushing for change and how welcome meeting others can be. The complexity of these exchanges makes it difficult to substitute interchange through another interface. Experiments using internet radio worked to inspire interest in related groups, but no real care. Interaction was missing. There was no discernible merit in attempting to store the learning in a more durational form; it merely became obsolescent before use.
However, the co-learning approach of giving multiple local organisers the budget to prepare a platform for sharing and reflecting (workshop, radio programme) worked well and could be replicated. While the platforms cannot be scaled easily with technology, the exchanges and reflection around them produce value. In sustained engagement to achieve citizen-led projects, such value may express itself as additional motivation. These are contexts where to keep going requires effort applied to civic structures, rather than supported by them.

Discussion: towards caring
The discussion summarises analytical insights about co-production of caring platforms, co-design and co-learning, then moves to formulating suggestions for the (co-)design of caring platforms.

Breakdown as an opportunity for caring
A common thread runs through comparisons as to the role of the municipality, understood here as the major local platform charged with caring for the public good. In the first study, a community-based municipal service struggles for survival as council staff and local residents negotiate its future. The decisive moment comes between two council departments, when the more client-responsive cultural section takes over the running from the more functionalist waste management section. The survival of ReTuren depends on sorting out internal organisational elements, in other words, departmental culture is more important than host: both services reside in the same organisation. This reveals internal differences, but also a flexibility in the municipal structure that is itself an element of caring. There is a will in the municipality to care structurally (i.e. by implementing co-design and discharging its caring duties through engaging with local people), as well as in terms of caring for citizens and planet with good waste practices. The breakdown of the municipality as caring platform is only temporary and the protests of residents and professionals who care about ReTuren resurrect it. The local council shows itself to be one in which the values of the different platforms can coexist and build on each other, just as the infrastructural aspects do.
In the second, the focus is also on the struggle for viability and the Oxford boatyard campaign shows that sometimes co-operation with the local authority is less effective than resistance in protecting the environment for local people. Care may not manifest as kindness. This is not to say that local authorities and residents never co-create services in the UK. 3 But the breakdown here is not within the functions of the municipality, as in ReTuren; rather it reflects an overall decrease in supportive infrastructure. English councils can be slow to react to citizen initiatives because, as a by-product of national policy, they are struggling for identity. Civil society challenges go beyond facing bureaucracy to dealing with the neo-liberal alignment of local authorities' revenue-generating strategies, such as, in Oxford, selling public land for development and a tourist-focused approach that prioritises places as destinations.
In both studies, breakdowns reveal, teach us about and mobilise care. One could see the English example as a breakdown that makes space for citizen initiative and a recognition of mutual interdependence, or as a state abandoning its caring responsibilitiesthis being a subject of debate in the UK. Berlant (2016, 403), echoing Star and Ruhleder, comments that: 'institutional failure leading to infrastructural collapse [. . .] leads to a dynamic way to disturb the old logics, or analogics, that have institutionalised images of shared life'. In other words, making visible the hidden structures and processes of infrastructure offers the possibility, but not the inevitability, of repossession. The failure of something seen as valuablecombined with a perspective on its limitationscan enable people to step into the gap and develop new alliances. This view of limitationsthe waste handling context, the progressive dismantling of English municipalitiesalso creates awareness about the need for learning. At its best, this results in an opportunity to learn together across different organisations, structures and contexts.
This is not an argument for removing support, but for rethinking it to encourage confidence and meaningful engagement from both institutions and citizens (e.g. Wilson et al. 2018). In the two studies, we see highly contrasting situations, picked to exemplify the way that actors, values, tools and place can affect what co-design means for any context. Co-design creates the opportunities for encounter between different actors and types of expertise, but it is always the result of particular assemblages. The resultant learning need not be entirely orientated towards material outcomes and shared knowhow. The learning could be, as in the ReTuren study, about the importance of different orientations to care. It could be about how to generate the conditions for care to flourish in the moment. In the case of the UK study, much of what was shared between groups was learning about how to work collaboratively and endure despite infrastructures that resist or ignore this kind of grassroots local initiative.

Articulating the relationship between co-design, co-learning and care
We have already noted the degree of co-design we see operating in each study ( Figure  1). ReTuren was not conceived by residents; the concept pre-existed their involvement and came from civil servants. Yet, in its development, local residents and organisations had opportunities to intervene in both process and issue. Offering co-design resulted in a greater commitment to the issues and their outcome (by staff and community), but also created a small crisis for the municipality, since it ended up challenging what organisational structures allowed. In the English contexts, the projects pre-existed the research funding and the groups used the money for workshops to further their aims. In other words, in the second examples, both the community resources being developed and the co-learning platform of the linked workshops were the result of co-design of structure, process and outcome. So, it is evident the studies involved co-design. But does co-design lead to caring? Is it possible to reposition it in such a way that it will? The relationship is not so simple. Where accountability is denied, it is difficult to argue for care in any interdependent sense. But where accountability is distributed widely, as with higher degrees of codesign (Figure 2), a mutually caring relationship is not an automatic outcome. That said, involving people and trusting them leads towards sharing accountability and a sense of co-ownership and investmentit is, itself, a form of caring to make the opportunity to work together. It can be seen as enacting a democratic principle through co-design (Binder et al. 2015;Light 2015). In ReTuren, people's close involvement in the pilot led to the development of a service that fit the local context, but also a strong sense of co-ownership for that service. Similarly, it is clear from the stories of making radio programmes that it is useful beyond media outcomes to support the process of co-evolving activities and making things together. Producing the programmes bonded groups and generated mutual care, raising issues of identity and commitment. Cultivating a sense of co-ownership was key to nurturing participants and stakeholders in caring for and about these platforms.
Further, we found a connection between care and co-learning, i.e. an explicit and structured collaborative reflective process. Co-learning is about sharing of practices, needs and stories of success and failure with the aim of collective knowledge production around an issue. It is a form of care related to personal and organisational sustainability in difficult times. The stated goal of providing a means of reflection and opportunity to share practices is not the mutual learning by-product that comes from co-producing a service (e.g. Robertson et al 2014), but rather a codesigned learning project. Recognising this, the campaigners and project leaders in the English cases used the chance to address care through co-design as a type of collaborative learning about change and the potential of communities to enact it. We also see learning in the Swedish example. The pilot focused mainly on co-designing the service and the mutual learning focused on the waste prevention practices developed among people directly involved in the everyday running of ReTuren. In ReTuren's reorganisation, a collective effort to learn about each organisation's structure provided the opportunity to discuss and align different forms of caring, without people necessarily engaging in situated co-design or in each other's practice. Yet, as elsewhere, merely supporting co-learning might be not enough to lead to caring. The radio programme supported learning, but led to interest rather than caring. It was interesting to encounter the textural differences as researchers across the two projects and note the qualitative differences.

(Co-)designing caring platforms, a focus on co-learning?
The cases in this paper highlight the role of existing structures and processes in designing caring platforms and, particularly, how processes cannot be abstracted from the specific context in which they are developed and used. It follows that attention must be given to how features, such as policy and organisational culture, enable or hinder opportunities for care through the provision of platforms.
We can observe from our studies that care does not scale as systems do. It is necessarily situated in these systems and a product of them, but it is not an automatic output. Since one cannot design care as such, we have to be careful in describing anything as a 'caring platform'. As we note, in describing structures that support citizens and protect the public good as caring platforms, design can play down or stress caring features; it can reveal the underlying care of interdependency more clearly and it can make the conditions for active forms of care to emerge. If motivated by a democratic will (rather than the delegation of all accountability), even an invitation to co-production can be considered a caring act. However, it might not be enough to ensure care. In other words, all our investigationsin different contexts and with different playersshow that intentions matter but also that caring has to be worked at/ for. Processes, such as co-design, and structures, such as jetties and boatyards, can enable the development of caring, but they do not supply it. And, beyond individual organisations and activities, a role is played by differences in geographical, cultural and policy context and how these impact, from municipal departments to grassroots enterprise.
We have not emphasised digital platforms, but the English study sheds light on the relation between digital systems, mediation and care. The examples show it is not sufficient to offer news of others' circumstances at a digital remove to inspire fellow feeling. We have only to compare the responses of the East Clevelanders to the Oxford campaigners, in person and (by their account) through national news media. The work of making radio programmes reinforces this message. Where affect and interest exist, the device of sharing stories worked well and could be envisioned for loosely linked networks. But the programmes were unable to turn idle interest into a matter of concern. It is evident that the scaling qualities of digital networks can only be harnessed for care as part of what Miskelly call 'sharing cultures' (2015, 2019), which bring people together to layer and mesh resources into 'relational assets' (ibid), rather than seeking to replace these elements.
Clearly, there is merit in co-designing platforms and benefit from the greater engagement, accountability and learning that comes when authorities work directly with citizens and their needs. However, we argue, mutual learning achieved on the way to an anticipated outcome may not be enough to promote care; the co-design process may need to focus more explicitly on how to support co-learning. Performing design together is not the whole story. Reflecting and learning together is key to giving visibility to the invisible (e.g. the role of underlying infrastructure in influencing and shaping platforms); to collaboratively articulating issues and to creating opportunities for caring and being cared for.

Conclusion
The paper looks at how care for the public good comes about in co-production initiatives in different geographical, policy and cultural contexts. It points to the opportunities and limits of co-design in supporting existing structures and relationships and furthering the development of a caring ethic in civic initiatives exploring new relationships between municipalities and citizens. Particularly, it highlights how, while it is not possible to (co-)design care as such, co-design can create the chance to consider care and conditions for caring in contexts with multiple authorities and beneficiaries by sharing power/control and fostering trust. The cases highlight the importance of co-learning as an explicit collective reflective process that articulates underlying conditions for caring, conditions that might be difficult to grasp and address simply through the mutual learning of co-designing.
In closing, we argue that caring platforms are not an inevitability, even when the will is good, but are something to be worked at as part of maintaining a meaningful ecology of service provision for the public good as network economics and neoliberal thinking drive forward the monocultures of platform capitalism.