Key actors in community-driven social innovation in rural areas in the Nordic countries

Rural areas in the Nordic countries often face challenges such as demographic change and the closure of public services. To address this situation, some rural communities have developed solutions which can be characterised as social innovation. This paper analyses 18 such community-driven social innovation projects across the Nordic countries and seeks to examine the importance of different actors in the initiation and implementation phases of


Introduction
Rural areas in the Nordic countries have experienced a significant decline in population in recent decades. This often distorts the age, gender, and socio-economic balance by depleting the number of young, well-educated, and economically active people living in those rural areas (Grunfelder et al., 2020;Hörnström et al., 2015;Lindberg, 2017;Tietjen, 2017). Although Nordic countries are well known for their welfare state models (based on universal access to services such as healthcare and education), the high cost of sustaining these models for a widely dispersed and shrinking population remains a challenge (Copus et al., 2017;Hörnström et al., 2015;Shucksmith and DavidBrown, 2016). This shift also means that actors other than those from the public sector gain a larger role in rural development initiatives. In most countries increased centralisation and funding cuts have led to a decrease in public sector involvement in rural areas, but also a corresponding rise in civil society actors assuming responsibility for services, infrastructure and local assets (Bernal and Avella, 2019;Cheshire et al., 2015;Coffey and Polese, 1984;Fougère et al., 2017;Harris and Albury, 2009;Pestoff, 2012).
Despite the ongoing challenges that continue to exacerbate the ruralurban divide in Nordic countries (Baerenholdt, 2002;Grunfelder et al., 2020;Tietjen, 2017;Wittorff Tanvig, 2015), new solutions and projects are also emergingones that support rural community development.
Some of these could appropriately be described as social innovation (SI). While rural communities are connected to specific geographical areas where people live, work, shop and participate in recreational activities, SI activities are associated with new social interactions which reach beyond the geographic area of the rural community (Bosworth et al., 2016a(Bosworth et al., , 2016bHowaldt and Schwarz, 2017;Noack and Federwisch, 2019). New social interactions and collaborations involving actors from different geographical areas elevate the ability of rural communities to address unmet social needs and challenges through the establishment of SI and to create a transformative social change (Bernal and Avella, 2019;Bock, 2016;Howaldt et al., 2018;Richter, 2019;Ziegler, 2017).
SI is connected to qualities such as community participation and empowerment (Edwards-Schachter and Tams, 2013;Lindberg, 2017). This is the result of a collaborative process (one person alone cannot bring about a social innovation), and that process in turn brings together partners across sectors, disciplines and specialities from government, civil society and the private sector (Howaldt et al., 2018;Neumeier, 2012;Ziegler, 2017). In this way, SI shifts our perspective away from fixed actors in separate rural areas, towards a more fluid image of shifting actors, relations and functional networks operating across different geographical areas, beyond the local and the rural (Bock, 2012(Bock, , 2016Lang and Fink, 2019). There is also evidence that SI involves multiple institutions, norms, and practices, and that there is a connection between SI activities and those which are already very much part of the European Union (EU) policy concept of Community-Led Local Development (CLLD) and LEADER (Copus, 2016).
The main difference between rural development per se, and SI in rural development specifically, is that SI needs to be perceived as novel by the community involved in its creation (Neumeier, 2012). This perceived novelty can also be due to new partnerships and collaboration across different geographical areas. SI has been studied empirically and theoretically as a driver of social change. However, SI processes in rural areas throughout the Nordic countries have not been analysed in earlier research (Cajaiba-Santana, 2014;Howaldt et al., 2018;Nordic Council of Ministers, 2015). This study proposes to fill that gap by examining the dynamic among the actors involved in initiating and implementing SI projects in rural areas across the Nordic countries. Its contribution includes a systematic analysis of the SI process, based on studying actors' involvements and partnerships as they initiate and implement SI. It is guided by two research questions: a) How are community-driven SI projects initiated and implemented in rural areas in the Nordic countries? b) What are the roles of different actors at different stages of the innovation process?

Materials and methods
Our study was commissioned by the Nordic Working Group on Demography and Welfare under the Nordic Council of Ministers' Committee of Senior Officials for Regional Policies. The working group consisted of representatives from each of the Nordic countries. They provided context-specific input, accommodating perspectives arising both from welfare policy and rural development. They also guided research design, addressing the need for greater knowledge of successful initiatives and enabling factors for SI in rural areas.
The working group also had a specific interest in understanding how SI can address demographic challenges faced by rural communities across the Nordic countriesincluding ageing populations, outmigration (particularly of young people), and limited access to services. A joint Nordic analysis is of interest from a political perspective, in the context of Nordic cooperation; and because of similarities between Nordic welfare systems, as well as the socio-economic profiles of rural areas and planning systems (Getimis, 2012;Grunfelder et al., 2020;Knieling and Othengrafen, 2009).
The examples chosen for this study had to fulfil two criteria. First, they needed to involve a social process requiring significant levels of community spirit and cohesion and enhancing the capacity of a community to respond to future local challenges. Second, they had to differ from other rural development projects, in that they were perceived as novel within a local and institutional context.

Rural typology in the Nordic countries
All the social innovation initiatives selected for study were located in rural or remote areas and, as a result, help us understand SI in a Nordic rural context. We have used the urban-rural typology based on indicators and categorisations by the Organisation for Economic Cooperation and Development (OECD). These categorise rural areas according to local population density, the proximity of main cities, and the remoteness of communities compared with predominantly urban regions (Grunfelder et al., 2020). As illustrated in Map 1, all SI projects in this paper are located outside the capital region in their countries, in regions ranging from those which are predominantly rural to those characterised as intermediate. Communities located in both predominantly rural and intermediate regions can face similar challenges, including school closures, cuts in public transport, lack of employment opportunities, and the outmigration of young people.

Research design
The research design builds on practical examples to understand the foundation of community-driven SI projects across the Nordic countries. The study applies a qualitative approach to examine the space and time dimensions of the innovation processes. Semi-structured interviews have been conducted with local project leaders and/or others who had been central to the community-driven SI process, in order to gather empirical information. Where possible, additional background documents have been used to inform the analysis. A detailed narrative for each SI project is available online. 1 The combined narrative and dataacquisition approach used here was inspired by the "social innovation biography" methodology developed and tested in the EU-funded SIM-PACT project (Terstriep et al., 2015;Terstriep and Rehfeld, 2014;Totterdill et al., 2016). This approach was chosen because it provides an in-depth "understanding of the development paths, knowledge trajectories and stakeholders at the micro-level" (Terstriep and Rehfeld, 2014:1), and also because it pays specific attention to the relationships and contextual setting of each SI initiative.
For each biography, the initial data acquisition involved desk research, following a qualitative interview and a short questionnaire summarising the interviewee's assessment of the importance of the different actors' involvement in the initiation and implementation phases of the project. The answer to each question in the questionnaire produced a score ranging from 0 to 3, depending on its perceived importance of ideas, resources, and decision-making. The score of all SI examples was added up, in order to illustrate the involvement of different actors. In calculating the average for the 18 SI projects, a trend emerged, revealing the varied involvement of actors at different stages in the SI process. The questions covered the different roles of various participants (civil society, public sector and private sector) in generating ideas and inspiration, in providing resources (including labour), and in making decisions at two different stages of the SI process: the initiation and implementation phases.
The involvement of civil society actors was broken down into two categories. The term "civil society (organisation)" was used to indicate community involvement rooted in an institutional framework (e.g. a village association), whereas "civil society (community)" was used to indicate the general involvement of community members not directly a http://archive.nordregio.se/Publications/Publications-2016/Territorial- connected to an organisation. Public sector actors were also broken down into two groups. "Public sector (local)" referred to the nearest local government entity (in the Nordic countries this is usually the municipality), while "public sector (regional/national/international)", referred to public entities at the regional, national and international level. This division is based on the frequency of involvement by different public sector entities in SI activities. The data is presented in radar charts in the results section in order to support the findings of the qualitative analysis.

Framing community-driven SI in rural development research
In the literature, SI is related to what in recent years has become known as "neo-endogenous" growth. In very broad terms, this follows the rejection of the kind of top-down (exogenous) policies common at the end of the twentieth century, and the purely locally-based "endogenous" approaches that followed. Neo-endogenous development, by contrast, requires a fine balance between local initiative and resources on the one hand, and appropriate inputs of capital, expertise and sources of innovation on the other. These may be best accessed through networks that extend out into the wider world (Bock, 2016;Bosworth, 2010;Dubois et al., 2017;Ray, 2006;Shucksmith, 2016).
Neo-endogenous local development initiatives are thus only possible through meaningful interaction between local areas and their wider environments (Bosworth et al., 2016a(Bosworth et al., , 2016b. Similarly, "social innovation can never be analysed as belonging only to 'its' place, the place where it was generated, but [must be conceived] as occurring within a complex web of spatial interconnections" (MacCallum and Haddock, 2016:7), and "it is evident that social innovation requires networking and building of relations across the borders of the place in question" (Bock, 2016:569).
However, certain preconditions need to be in place before SI projects can lead to change. These include human and social capital, the presence of social networks that bind the community together, and outreach to potential external partners (Cajaiba-Santana, 2014;Lang and Fink, 2019;Neumeier, 2012;Unceta et al., 2017). Other key factors for the success of SI initiatives include the establishment of new partnerships within the community, and the role of local facilitators (Moulaert et al., 2005(Moulaert et al., , 2013. The success of an SI project also depends upon the degree of benefit it offers, its ease in terms of application, the foreseeability of its results, and local willingness to participate (Neumeier, 2017;Terstriep et al., 2015).

Defining SI in community-driven rural development
SI is a "contested concept" (Bock, 2016:2). Research in this field offers many definitions; some of them vague and inclusive, others "tighter" and more specific (Bock, 2012). According to Bock (2016), the term "social innovation" can be used to describe a change in "social relations, people's behaviour, norms and values" and can therefore "refer to the effort, method, result or change initiated by collaborative actions" (Neumeier, 2012:54). This means that the SI project is generated by a social process rather than by an individual (Bosworth et al., 2016a(Bosworth et al., , 2016bLindberg, 2017).
An SI process may or may not lead to the creation of a social enterprise (Copus et al., 2017). The main difference between a social enterprise and a commercial enterprise is that the commercial enterprise is driven primarily by profit-making through acts of trade, while the social enterprise takes the initiative in identifying and solving social challenges in a community (Crisan-mitra et al., 2011). Often, an SI initiative is triggered by a social need rather than by the pursuit of a distant goal (Cajaiba-Santana, 2014;Neumeier, 2012), and the people involved in the process of its initiation and development perceive the SI project to be new within their context. The process usually also entails a change in attitudes, behaviour and perceptions (Neumeier, 2012).
The collaborative process involved is also acknowledged in the most frequently quoted definition of SI, namely: "Social innovations are innovations that are social in both their ends and their means" (Caulier-- Grice et al., 2010;European Comission, 2011;Mulgan, 2007). This indicates that the process is as important as the result, and that the overall purpose of SI is to meet social needs, to create new collaborative models or social relationships that are good for society, and to enhance society's capacity to take action (Caulier-Grice et al., 2010;Neumeier, 2012;Totterdill et al., 2017) Almost half of the initiatives mapped in a global study on SI show a direct user or beneficiary involvement. A central finding in this global mapping is the identification of the local competences and capacities involved in acting together (Howaldt et al., 2018).
Collaboration on SI is about creating social assets rather than material outcomes. This definition acknowledges both the process (as in the means) and the outcome (as in the ends) of SI. By building on this understanding, SI is defined in this paper as a process that leads to novel, community-produced solutions to social problems.

Results and analysis
The results section examines the 18 community-driven SI projects addressing demographic challenges faced in rural and remote areas of the Nordic Region. These 18 examples can be divided into three broad themes: local fora for community development and capacity building; creative approaches to service provision; and the provision of community spaces and activities. These examples, with short descriptions, are shown in Table 1.
As illustrated in Table 1, each SI initiative belongs to one or two categories. Our analysis uses examples of different SIs to illustrate key points from each of the categories, and to reveal nuances in the development process for these three categories. There are some differences in the way the actors are involved in the process, depending on the category. All the new "community development fora", "local services" and "community spaces and activities" initiatives illustrate a significant local development which mitigates the impact of demographic change.
Even though the SI examples here are embedded differently, depending upon the local context (be they associations, social enterprises, co-operatives or committees), they can still facilitate or provide a mitigating effect in relation to rural challenges such as outmigration and an ageing population by offering new activities to stimulate local development. All the projects have a novel approach and succeed in establishing a new form of collaborative model: one that involves public and third sectors, commercial providers and service users finding new ways for the different actors to work together.

Getting started
The initiation phase includes everything from the initial idea right up to the start of what could be considered regular operations. In all cases, SI initiatives were triggered by regional vulnerabilities, such as the demographic shifts and challenges posed by shrinking public budgets. The challenges themselves were varied. They included service closures (or the threat of closure), ageing populations and population decline, local economic restructuring, and local authority mergers. Although these factors have each been significant drivers, it is important to acknowledge that such conditions, in and of themselves, do not necessarily result in community-driven SI. Each one of our examples illustrates a capacity to develop ideas, collect resources and manage decision-making. This leads on to the initiation phase, which is characterised by starting-up a new service, the formation of a new association, the opening of a new facility, or the commencement of a new activity. Fig. 1 shows the involvement of different types of actor in the initiation phase. It sums up our assessment of different actors producing ideas, decision-making, and collecting resources in each of the 18 SI projects. The involvement of the different actors is valued on a scale from 0 to 3 by the key people involved in each SI initiative. By adding up the total value and subsequently calculating an average, Fig. 1 illustrates the perceived involvement of civil society (community), civil society (organisation), public sectorlocal, public sectorregional/national/ international, and private sector for all of the18 examples.
Civil society (in the shape of the community) together with the local public sector, are the most important actors during the initiation phase. In most cases, informants articulate this as a process that engages people from the community, and they underline the importance of connections between members of the community, civil society, and local government/the public sector. This engagement of people in the community in  the initiation phase can also be described as an endogenous development process.

Ideas for the initiation phase
The source of ideas is most likely to be individual members of the community, but local associations and the local public sector also play a role in many of these SI examples. For initiatives resulting in 'community space and activities' and 'local fora for community development' members of the community were the key actorsthe ones who came up with the ideas. One example is the Skovgård model, which provides community space and activities based on efforts by members of the local community to reopen a local shop and hotel, while also building an inclusive community and involving disabled people actively in work and social life. Without the vision of members of the community (including some particularly driven, passionate individuals), this idea would never have been realised. A similar trend is seen in those SI projects which established 'local fora for community development'. One example in this category is Free State Lucky Naeroset, where residents of a village in Norway decided to set up their own form of local government. They used local cultural activities and roleplay, so that the values of humour, imagination, originality and courage were tapped in order to support a process to reverse the decline in population and to keep a local school open (a venture in which they succeeded).
In the SI examples which offered additional service provision, the local public sector was the most engaged actor during the idea phase. This type of SI initiative often focuses on end-user involvement and participation. In the example known as May I Help You, the idea emerged from workshops in the local community and hosted by the local public sector of developing a new way of bringing young peopleparticularly those at risk of social exclusiontogether with elderly people who needed assistance with everyday tasks. Similarly, the Ramsjö Public Meals Programme was set in motion by local authority actors through an initiative to involve end-users in a thorough needs-assessment, with the aim of developing new service solutions. This process eventually led to a new approach to public meals provision, which has subsequently supported job creation through local entrepreneurship and enhanced the quality of the service provided.
For all projects, interviewees mentioned that a supportive attitude towards growing new project ideas among community members is an important factor in both the idea phase and the development phase. When experimenting with new modes of working, the social environment is also of great importance to the prospects of the idea developing into an actual, functioning project.

Resources for the initiation phase
Resources include knowledge, networks, funding and (volunteer) labour. Several interviewees also mention knowledge and funding from the local public sector to get initiatives off the ground. This was seen as crucial in providing the support needed to gain knowledge of legal procedures and to extend the social network involved. Those SI projects setting up and registering an association often combined membership fees and other fundraising activities to support their work financially. In a few examples, financial support was also received from the regional and/or national/international public sector (e.g. national or regional development grants, or the EU funded Leader programme).
In addition to knowledge, networks and funding, an important input during the initiation phase is volunteer labour. Active village associations and their volunteers are crucial, because many SI projects rely on the work of volunteers over a prolonged period. A committed "core" of volunteers is therefore often regarded as a necessity. Having active local inhabitants and associations involved is supportive for SI activities. The type and scale of the business and of the investments involved would rarely be profitable using a conventional business model, and as a result these social initiatives are heavily reliant on volunteers. Volunteers are, indeed, the single largest asset for many community-driven SI projects. Some such initiatives take the form of specific activities created through personal encounters, organising workshops, trust-building exercises, and so on. Several of the initiatives (e.g. Skovgård Hotel, Tag Del/ Participate, Röstånga Together) have thrived by raising share capital/ crowdfunding. In many examples, therefore, a crucial step towards moving from the idea stage to the start-up phase of the SI project is the willingness of local members of the community to invest their time and effort in it.
In some cases, the private sector also provides key resources (such as sponsorship, grants or staff time) to organise events and activities that help to develop an initiative. However, the SI examples cited in this paper indicate that this is mainly provided when the key people in the community who are involved in the SI are also part of running a business. In rural communities, people often have several responsibilities (e. g. chairperson of the local handball club, mayor and parent coordinator) and most of our SI examples also reveal a high degree of dependence upon a few passionate souls leading the way during the initiation phase. In this initiation phase, the resources may consist of a combination of voluntary and charitable fundraising, public sector grants, fees charged for membership, or fees paid for services such as consultancy, catering, cultural activities and events.

Decision-making in the initiation phase
Decision-making is defined as the ability to make decisions driving a process towards results. The decisions can involve use of financial resources, labour, work and meeting places, infrastructure, structure of collaboration, events, and media outreach. Civil society organisations are the most likely to play a key role in decision-making, however, for several of the SI projects studied, members of the community were involved in the initial decision-making and then the local project manager collectively led the way forward through workshops and consultations. More generally, the results of the initiation phase illustrate the tendency of members of the community to band together and organise. All these SI examples went through a process of formalising their initiatives. This formalisation process mainly took on the most common legal formssuch as associations, civil society organisations, social enterprises, public-private partnerships, committees, community development trusts and co-operatives.
To emphasise the social value of SIs, it is necessary for them to specify, in some way, that the potential profits will be reinvested in order to generate even greater social valuerather than just turn a profit. Only one of the 18 SI examples had a high level of private sector involvement in decision-making. This SI is in Tranøy in Northern Norway, where a partnership between a social entrepreneur and the local council led to a company being established that built 36 homes for disadvantaged people and refugees, followed by a cleaning company that provided jobs for six refugees.
Decision-making in the initiation phase usually involves considering which legal entity it is best to establish, so that appropriate support can be given to the next step in implementing the SI project. In several cases, the community was part of this decision-making process. Those individuals leading the initiation of an SI initiative locally usually also take on a key function in the civil society organisations set up. This step is a precondition for starting to get the work recognised as an SI activity, and to provide foundations for the implementation phase.

Implementation and day-to-day operations
The previous section uncovered the fact that, in the initiation phase, there is often a high level of dependence on community members, civil society organisations and the local public sector. They are the important actors in getting an initiative off the ground. Demographic changes, including outmigration, ageing and the closure (or threat of closure) of services such as local schools, can all be conditions that generate impetus for the initiation of an SI project.
This section will focus on the implementation phase. Its aim is to understand the way SI projects operate in practice, particularly in relation to the contributions made by different actors. Fig. 2 shows the degree of involvement of different types of actors in the implementation and operation phase.
As Fig. 2 demonstrates, there is a substantial shift in the type of actors who are likely to contribute to the implementation and operation stage of a project, compared to those who will contribute to the initiation phase. The most substantial changes are the increased significance of the role of civil society organisations and the larger role played by the private sector. To understand the nature of these shifts, it is important to emphasise that it is not necessarily the case that new actors from civil society organisations and the private sector have come on board at this stage in the process. In fact, in many cases, it is the same community actors who are involved but the nature of their involvement changes, due either to the establishment of a civil society organisation, or that of a social enterprise.

Ideas for the implementation and day to day operation
Ideas in the implementation and day-to-day operational phase change in character as the focus now moves to the ideas required for daily problem-solving, rather than for developing a new concept. In nearly all our SI examples, it is seen as crucial that a project manager plans, guides and supports a well-designed and well-structured approach to the task of implementation, in order to ensure a continuous, smooth process. A clear allocation of tasks, time and competences is important for the SI project to fully establish itself in the field. A key competence is the ability to communicate the contribution of SI by developing ideas for good and effective outreach stories. Several of the SIs made good progress after attracting media attention that lead to publicity for the projects outside its local area in the implementation phase.
Ideas generated during the implementation phase also includes new approaches to establish platforms for collaboration. These can include both traditional purchaser-provider models and other partnerships developing new services together. One example of this is Sorø Senior Service in Denmark, where a volunteer group collaborates with the municipality (local public sector) providing a grocery shopping service for elderly residents. The idea emerged as there was already a social network locally and through a conversation with some of the elders' they saw a need for some of them to get groceries delivered to their home. The social network in the group is an important factor for the volunteers' participation, but the service would not have been developed without the idea to ask for a meeting place and the possibility to receive petrol money from the municipality.
Another idea for a new collaboration model is the Rovaniemi Area Committee Model, funded by Rovaniemi municipality in Finland. Here, the area committees are responsible for the provision of certain public services (education, culture, sports, youth services, healthcare information, day care, home care and local development), as well as budgeting for these services. The underlying idea is that there is a need for knowledge about local conditions and needs, so that effective solutions can be found for facing the challenges of service provision in rural and remote areas with ageing populations.
In general, during the implementation phase, the ideas generated are usually about new models of collaboration, generating day-to-day operating structures problem-solving, business planning, and outreach activities.

Resources for the implementation phase and day-to-day operations
Ensuring revenues from several different sources is an important strategy during the implementation phase. Most SI projects therefore focus on multiple sources of income and diverse economic activities.
Those SI examples established as an association were more likely to depend upon a membership fee, combined with annual public-sector grants. In the Faroe Islands, the Association of Outer Islands was established to continue a network developed through a project initiated by the public sector. Over the years it has depended on public grants to provide support for its activities on behalf of the outer islands, however, the last few years they also generated some income from membership fees which also can be an efficient way for an association to become financially viable.
All the social enterprises in the study are funded by a combination of voluntary and charitable fundraising, public sector grants, and fees charged for services. For three of them, Skovgård Hotel, Viesimo Social Enterprise Co-operative, and Röstånga Together, the most important income is fees levied for service provision. However, for Viesimo Social Enterprise Co-operative, and Röstånga Together the income from service provision is also complemented with charitable fundraising and volunteer labour.
For the SI examples that enhance service provision, the available resources appear to remain constant. One way these resources are procured is by providing a service on behalf of a local council. For example, Offerdal Healthcare Centre, a resident-owned co-operative, provides healthcare services in a rural region, is contracted by the municipal council to deliver its healthcare services. The municipal council has public procurement principles based on quality and other values rather than just cost. This enabled the healthcare centre to obtain a public procurement contract. Similarly, the service centre in Kalix Ö vre Bygd signed an agreement with Kalix Municipality to allow for the outsourcing of tasks associated with homecare service provision for the centre.

Decisionmaking in the implementation phase and day to day operation
When making decisions, many of the SI initiatives act in partnership with local public sector or other support organisations. Nearly all of the SIs were able to employ a part-time or full-time project manager to oversee implementation of the concept, the continuous engagement of members of the community, the building of local capacity, the management of a business plan and administration, and so on. Informants emphasise the importance of day-to-day decision-making for efficient implementation, maintaining local engagement and structuring the workload. This is crucial for the civil society organisations in the implementation and day-to-day operation phases of the project.
The social enterprises that were established were in some cases initiated by the public sector, while in other cases they were galvanised by members of the community. In Finland, the public sector initiated Viesimo Social Enterprise Co-operative, a social enterprise, and later passed it over to a local civil society organisation to run and manage its implementation and day-to-day operations. Another example is Röstånga Together, a community-owned social enterprise initiated by citizens after the end of a Leader programme project. Through the community's collective effort, they established an enterprise which provided local development in terms of public services, cultural activity and businesses.
Several initiatives that became social enterprises helped maintain services otherwise threatened with closure due to restructuring, or the drawing back of the state. However, here it is important to point out that the legal and tax framework for social enterprises varies between countries. In recent years, different Nordic countries have adopted a variety of approaches to try to make the legal and business environment friendlier for SI projects and social entrepreneurship. In Finland, for example, a Social Enterprise Mark was introduced as a label in 2011, in order to make social enterprises more distinctive in the marketplace. In Denmark, a law 2 setting up a national register of social enterprises was passed in 2014. In Sweden, there is a similar register, but only of social enterprises working with labour market integration for migrants (Nordic Council of Ministers, 2015;Wilkinson, 2015).
The role of the public sector in decision-making is significantly reduced as projects proceed from the initiation phase to the implementation phase. Overall, the role of the local public sector in implementing the SI examples studied here ranges from no involvement at all, to the provision of minor funding, to the provision of free premises for the development of partnerships with civil society actors or social enterprises. Public sector involvement in implementation is not uniform. However, as noted above, the role of the sector during the implementation phase appears minor compared with that of the initiation phase.

Discussion of the importance of community-driven SI examples for rural development policy in the Nordic countries
Understanding the details of various actors' involvements in the different phases of our SI examples provides us with a better understanding of the conditions necessary to initiate and implement SI projects in rural areas. These SI projects in Nordic countries require a fine balance between, on one hand, local initiative, and resources, and on the other hand, the input of capital, expertise and sources of innovation accessed by networks stretching beyond the local community. The empirically specified SI examples in this study confirm the importance of changeable roles and show how the actors' activities shift from the initiation phase to the implementation phase.
The examples we have studied in this paper show that local public sector actors, along with civil society organisations and members of the community, each play key roles in facilitating the development processes involved. This supports our original contention that, at least in the Nordic context, SI is a hybrid phenomenon with no fixed sectoral boundaries (Caulier-Grice et al., 2010). While the growing interest in SI may, as Harris and Albury (2009) suggest, have some connection to "a growing recognition of the inherent limits of the state" (Harris and Albury, 2009:17), it is not necessarily synonymous with the complete withdrawal of the state from its responsibilities in the way that Bock (2016) implies. Instead, the preconditions for SI in the Nordic countries appear to be characterised by changing roles for different actors. In some cases, this may involve co-operation between civil society actors and the local public sector to build up new and creative ways to provide services or collaborative fora for local development. In other cases, it may involve people in a community banding together to create an initiative that meets a need that does not fall within the remit of the state.
The empirical examples illustrate how SI projects involving service provision have greater involvement by public sector actors than those involving community development fora, or community space and activities. While the local public sector often plays an important role in the initiation phase of our SI examples, corresponding involvement is far less in the implementation phase. Although some SI processes in this study are led by local authority actors, these same actors work closely with members of the community to develop services that respond as effectively as possible to local needs and conditions. For example, May I Help You, Rovaniemi Area Committee Model and Ramsjö Public Meals Programme are SI examples of local authority actors initiating a process in which end-users are involved in a thorough needs assessment in order to develop new services.
The results show that partnerships reaching beyond the community are important when it comes to reaping the full benefit of social innovation, and that this is in line with previous findings which demonstrate that SI requires interaction between many actors (Howaldt et al., 2018:13). From the 18 SI examples in our study, the collaboration involved results in different legal formswith eight establishing an association, six becoming social enterprises, two setting up as public-private co-operatives, and two becoming community development trusts. It is relevant to make a distinction between the initiation and implementation phases of SI activities. While there is a high influence of 'civil society as community' in the initiation phase, the 'civil society as organisation' is dominating the SI implementation phase.
Our analysis provides nuance to the picture, illustrating the variety of actors involved, as well as the importance of both the local public sector and social networks. The 18 SI examples demonstrate a similar trend across the Nordic countries regarding how the community-driven SIs are initiated and implemented. The actors involved in the process also illustrate a similar trend towards engagement. This implies that rural policy in all Nordic countries could benefit from enabling local actors to engage to a greater extent in regional, national and international networks. In several of the examples, the local authorities provided a supportive environment through education and training, with the aim of upskilling local people and providing access to spaces and infrastructure such as youth centres and community centres. This is the sort of thing that can help secure the legitimacy of SI initiatives, building that broader capacity and helping them thrive in the long term. Many SI examples rely on a few passionate individuals, and this entails a risk that must also be assessed as part of rural development policy. It is crucial to build up broad, local capacity to ensure that SIs continue to make a longterm contribution to a socially sustainable development.

Conclusions and implications
The contribution that this study makes is to provide a set of empirical examples illustrating a shift in the involvement of various actors in the initiation and implementation phases of community-driven SI projects. When considering all 18 SI examples from across the Nordic region, it would appear that civil society organisations play an important role throughout the process. During the initiation phase, however, our b L 148 Forslag til lov om registrerede socialøkonomiske virksomheder. analysis illustrates a more nuanced type of co-operation, one that involves members of the community, civil society organisations and the local public sector as the primary actors working together to develop ideas, find resources and manage decision-making.
Most of the examples describe how members of the community band together to develop an idea, and as part of that process established an association, a co-operative, a social enterprise or a committeedepending on the national context and the setting. A multiplicity of legal forms are, therefore, observed in SI projects in the Nordic countries. The resources deployed vary among the examples. Some are completely based on voluntary work; some are dependent on public sector support; some source their income from consultancy (e.g. catering or other services), and some charge membership fees. Resources can also take the form of networks and social connections outside the community, which provide advice, funding and/or publicity.
The implications of the results of this study point towards the need to support the community's capacity to engage in collaborative processes. Social innovation in rural areas in the Nordic countries is initiated by relevant actors getting together to begin a cumulative process which produces a project to address a local challenge. The ideas from our SI examples are, to a large extent, developed by community members as a response to demographic change, such as an ageing and declining population, as well as service closures (or the threat of closure), local economic restructuring, and local authority mergers., As such, the examples illustrate how community-driven SI initiatives make a positive contribution to addressing specific challenges faced by rural areas across the Nordic countries.