Local Action with a Global Vision: The Transformative Potential of Food Social Enterprises in Australia

There is an urgent need to make food systems more sustainable and resilient. Such a transformation goes beyond technological innovation and requires economic and social change. Research interest in the transformative potential of community level action has increased. Food social enterprises often operate at the community level and consist of not-for-profit organisations that aim to make a positive contribution to social justice and environmental sustainability. The question addressed in this paper is whether these social enterprises are limited to isolated improvements or have the capacity to transform food systems more widely. This paper uses a multi-dimensional framework (involving the social setting, operational models, governance, and institutional context) to analyse the transformative potential of eight food social enterprises in the Australian cities of Brisbane and Melbourne. The analysis indicates that these enterprises create social networks, pursue agendas aligned with a global vision of transformation, and include a diversity of stakeholders. Their operational models are consistent with the goals of environmental sustainability and social justice. Their governance involves equality, transparency, and flexibility. In the institutional context, support from public policy is limited and there is a need to improve their engagement with governmental actors. While food social enterprises are well placed in the quest to make food systems more consistent with ecological dynamics and social justice, they need to engage in greater advocacy for institutional change in order to maximise their transformative potential.


Introduction
Current food systems are often contributing to social inequality, failing to provide adequate nutrition, and causing environmental degradation [1]. Food systems need to transform into more sustainable and resilient structures in order to cope with the impacts of climate change, connect with ecological dynamics, and be more socially just [2,3]. Such transformation needs to occur at all levels, ranging from individual food consumption habits to more systemic changes, such as de-coupling from fossil fuels [4,5]. This requires a holistic vision that seeks opportunities for improvement in the whole life cycle of food, and includes social, cultural and environmental aspects.
Government policies often focus on the technological development of food production and the reduction of negative externalities, which do not create a context for major change [6,7]. While

Transformation, Food Systems and Niche Initiatives
Transformation is defined by socio-ecological systems thinking as fundamental shifts that create global environmental change [23]. Transformation is required when socio-ecological systems are entangled in undesired states [24] that can be composed by persistent ecological, economic, or social structures that are unsustainable [25,26]. Folke et al. [25] point out that transformation in socio-ecological systems does not arise in a vacuum and that crises may be an opportunity for pushing systems into more favourable states.
Individuals and social groups have the capacity to innovate and ultimately transform systems [26]. Values, practices, and power relations can be changed, leading to system transformation [27]. These social changes need to create positive interactions with natural capital and ecosystem services to achieve sustainability [27]. Changing the different social components of systems, from personal values to wider institutions, demands the engagement of various actors, structures and processes [25,27]. A fundamental component of transformation is a shift in governance systems that makes the desired change mainstream and common practice [28]. Socially driven transformation, therefore, starts with novelty and innovation, creating a new trajectory that will be consolidated by institutionalisation [25,27]. Institutionalisation involves routinization, scaling up the change, and overcoming resistance from recalcitrant powerful stakeholders [27].
The transformation of food systems may be a gradual process because of their scale and complexity [21]. The dimensions proposed by Loh and Agyeman [21] as pertinent for transformation are society, economy, and governance. In terms of the societal dimension, food consumption values need to be transformed in order to encourage the adoption of diets that are healthier, more sustainable and connected to cultural and traditional practices [5]. These values need to articulate ideas related to environmental sustainability and social equality, as well as local cultural principles, in order to support the transformative process [5,29]. The involvement of a diversity of stakeholders is also important [29,30]. These stakeholders will influence which actions are taken and can limit or boost transformation [29,30]. Communities with diverse stakeholders are also valuable for creating the social networks needed to expand the reach of transformative values beyond one socio-economic group [14,31].
The economic dimension of food system transformation requires less hierarchy (in terms of risk-sharing and decision making), more transparency, and the embrace of environmental sustainability [21,29,31]. Hebinck et al. [29] discuss the importance of flexibility, inclusivity, and dynamicity in transformation. The economic dimension needs to be comprehensive, fostering environmental resilience and creating equality [14]. Economically marginalised players (such as food insecure communities and small farmers) also need to be accommodated [31].
In terms of the governance dimension, food systems need to be more reflexive and participatory in order to respond to particular local contexts and incorporate solutions more effectively [4]. Once again, a diversity of actors is necessary to ensure democratic governance [4]. The kind of relationship that governance actors have with different food system players will remove or contribute to obstacles to shifts in institutional regimes [23,28].
Niche initiatives have the potential to transform food systems in a bottom-up manner that improves the responsiveness to social needs [29], and food social enterprises are one example of agents of change [19,[32][33][34]. The not-for-profit character of social enterprises does not impede them from being part of the market economy [11,12] and they aim to change markets from within by adopting values related to sustainability and resilience [32,33]. Frequently, social enterprises will engage in food systems transformation by adopting new operational models or influencing society's values by fostering notions of citizenship and solidarity, improving the knowledge of food systems, and building on traditions [5,31,35]. They also need to influence government policy [29] but there is a tension between adopting more radical or moderate approaches in the engagement with institutional regimes [36]. It is recognised that the small scale of social enterprises limits their capacity to exert widespread transformative potential [13,21]. Food social enterprises, however, can serve as an incubator for transformation pathways [13,19,21,31]. It is important, therefore, to investigate what aspects of food social enterprises are contributing or hindering transformation [35].

Methods
This research adopted a comparative case study approach [37,38] to analyse the transformative potential of food social enterprises in Australia. The strategy used for sampling was maximum variation [37] in order to gather a picture of the diverse range of food social enterprises. The sample consisted of two food hubs, two buyers' groups, a city farm, an urban agriculture initiative, a specialist retailer, and a pop-up market. These were mostly operated by a small group of people (between 3 and 4) and often rely on volunteer work. The initiatives were located in the City of Brisbane and four local government areas of the Melbourne Metropolitan Region. The data collection stage gathered information about (i) the emergence of the initiatives; (ii) operational practices; and, (iii) any external support. Twenty-two semi-structured interviews were conducted with founders or members of social enterprises between the period of July and October 2018 (thirteen in Brisbane and nine in Melbourne). Information and documents from social enterprise websites and social media pages were also used. Visits were conducted in the cases where the social enterprises had a site of local operation. Data from the site visits, interviews, documents and literature were triangulated. Ethical clearance was obtained from Griffith University Human Research Ethics Committee (GU Ref No.: 2017/790) prior to the commencement of the data collection.
The semi-structured interviews were recorded and subjected to a thematic content analysis following the procedures described by Byrne [39] using NVivo 11 [40]. The themes considered in the content analysis followed the conceptual framework presented in Table 1. The framework was initially based on the three dimensions proposed by Loh and Agyeman [21] for food systems transformation (social setting, operational model, and governance) and later complemented by other relevant aspects discussed in the literature. The social setting analysis looked at the goals and connections of social enterprises. This revealed the kind of transformation that the social enterprises were aiming for and the stakeholders involved [29][30][31]. The operational practices were analysed to determine whether they were centred on environmental sustainability and sharing [21]. The ability to replicate these practices was also analysed in order to assess their potential to be scaled up [4,41]. The governance dimension was analysed in terms of the degree of democratic processes, equality and transparency [21,29]. Finally, an additional dimension of institutional support emerged from the empirical research data and was added to the conceptual framework. This fourth dimension relates to the relationship between governmental actors and social enterprises [23,41]. Table 1. Conceptual framework for analysis of social enterprises' transformative potential.

Features Analysed Description
Social setting Leading actors The actors who lead the social enterprises [5,29].

Goals
The goals that social enterprises want to achieve and that reveal their desired transformation [5,29,30].

Stakeholders involved
The different groups that participate in the social enterprise [29][30][31]. These can be from the private sector, policy makers, academia, civil society, or groups that social enterprises are targeting (e.g., farmers, food insecure communities) [29].

Social networks
The capacity of social enterprises to improve bonding and bridging within and between communities with shared values [14].
Operational model Risk-sharing The more equal distribution of risks and benefits [21,31].

Environmental sustainability
Food provision and consumption that reduce environmental damage such as: resource depletion, land degradation, the loss of biodiversity, pollution, waste, and green-house gases emissions [4].
Ability to replicate operational models The ability to reproduce social enterprises in different contexts and domains [4,41].

Transparency and flexibility
Models that are open to scrutiny, allow more flexibility in operations, and encourage autonomy in problem solving [29,42].
Equality in decision-making Democratic decision-making processes that are inclusive and improve participation by stakeholders [21,31].

Institutional context
Enabling policies The support that social enterprises receive from governments (e.g., grants) [29].

Engagement with government
The connections and relationship that food social enterprises have with governmental actors that indicate their capacity to create institutional change [4,29].

Results and Discussion
The eight initiatives included in this study form a diverse picture of food social enterprises in the Australian context. Their average time of existence is nine years, with the youngest one being established in 2016 and the oldest in 1994. The size of the initiatives was small-scale, even for the ones that have been operating for longer periods of time. Their significance, however, is on the achievement of positive environmental and socio-economic outcomes that distinguish them from the dominant food market players. Table 2 summarises the main activities of the initiatives that participated in this study. The following sections discuss the food social enterprises' transformative potential in the dimensions of social setting, operational model, governance, and institutional context. It is important to emphasise that the analysis is limited by the perspective of social enterprise members that were the focus of this study. The city farm is used for growing food, educational activities, and sells produce in a market garden system. It also provides compost facilities to the general public.

Brisbane
Farming team manager

Community garden facilitator
The initiative facilitates the implementation of productive community gardens on private and public lands and connects landowners with people who wish to grow food.

Melbourne General manager; and Projects manager
Pop-up market This initiative runs affordable fruit and vegetable markets once per-week in five different neighbourhoods. Melbourne Operations manager; and Accounts and communications.

Social Setting
The analysis of the social setting dimension considered who were the leading actors in the initiatives, their goals, the stakeholders involved in their operations, and evidence of their capacity to create/enhance social networks. Table 3 presents a summary of the results of the eight social enterprises analysed.
In terms of leading actors, half of the social enterprises were started by one individual who had experience with, and knowledge of, food systems and who saw an opportunity for addressing issues faced by their local community. The other half consisted of groups of activists or former members of other social enterprises who created organisations based on their own volunteer work in order to improve their local communities. Thus, all social enterprises were founded by people who had a deep involvement with their food systems. Founders and members of initiatives analysed in this study were seeking something more than the fulfilment of professional goals by starting a social enterprise: they had a personal and ethical commitment to the improvement of food systems. These results tally with the findings of Sonnino and Griggs-Trevarthen [33] where food social enterprises in the UK were founded by individuals or small groups motivated by an interest in improving their community.
The goals of these food social enterprises centred on access to organic food, environmental sustainability, social justice, and community building. Other studies have also identified a more holistic vision that is not focussed on economic profit in similar initiatives from different places [12,18,32]. These goals are reflected in practices such as: offering produce that is affordable, seasonal, local and diverse; supporting agroecology farming; transparency in the origins of the food; and, the fair payment of farmers for their produce. The goals of social enterprises are different from the mainstream 'food from nowhere' regime, where cheapness, convenience, and invisibility of food origin are emphasised [2]. One individual who used to be a farmer The people who engaged with the initiatives studied did not always share the goals of the social enterprises. Being able to change the values and behaviours of actors in order to keep them engaged is, therefore, important for the survival of these initiatives [42]. Interviewees described a tendency of people engaging first with the intention of eating organic food or local produce, and later developing other values that led them to question their former practices. Examples of changes include the consumption of seasonal food, a different perception on food prices, and different expectations of food appearance and taste. Consumers from the Brisbane food hub reported that they now refuse to buy fresh produce in supermarkets. One mentioned that the uniformity of food in supermarkets feels "weird", whereas the vegetables and fruits of the food hub look more "real" due to "different sizes and a little bit of dirt". He also added that "you can actually feel that it comes from the earth, or from a tree".
The impact of the goals adopted by the Brisbane food hub on people is evident, however, the founder suggested that there is still room for improvement in terms of education and the dissemination of the food hub's goals to the community: "We're still a long way from where we would like us to be, particularly around the pricing side of things. There's still that dominant economic paradigm of market pricing so prevalent in farmers' minds and in our minds [ . . . ] So there is a massive amount of work to be done there. There is a massive amount of work to be done in terms of educating people on how to cook with all this seasonal produce, how to get back food literacy and cooking skills." [Founder of the Brisbane food hub] The third feature of the social setting relates to which stakeholders from the urban food system are involved with the social enterprises. The social enterprises investigated engage with a range of urban food system actors at different levels. The engagement of social enterprises with the private sector was the most marginal one, possibly due to different goals relating to economic profits. Only the urban agriculture initiative from Melbourne engaged with the industry by helping private contractors to establish community gardens in their lots as a form of temporary land use before construction activities. The involvement with government actors was present in four initiatives but was linked more to bureaucratic issues rather than their core mission. Five social enterprises had established connections with academics. The highest level of external participation came from civil society actors in the form of activist groups and other food social enterprises.
The involvement of target groups varied amongst the social enterprises. While the urban agriculture initiative involved people interested in urban gardening, the Melbourne pop-up market was not able to directly involve food insecure communities in its planning and operation. Initiatives that had farmers as the main beneficiary group (food hubs and buyers' groups) had them involved at some level in an informal manner. Farmers were able to discuss prices and produce availability, but none held a permanent position in the social enterprises. The specialist retailer involved farmers in a similar fashion but had formal procedures for dealing with the buyers' groups who were their main consumers.
Clark et al. [31] suggested that social enterprises seek to accommodate marginalised groups within current contexts, rather than promote transformation. In contrast, the cases studied here transformed the context for marginalised groups. Food hubs, for example, did not only aim to include small farmers in the trading system, they also appropriately valued their products. The Melbourne pop-up market not only allowed food access to vulnerable communities, it supported culturally appropriate diets by offering produce that is not available in most food outlets. What could be improved in these two examples is the greater inclusion of the target groups in the management of the social enterprises. This could enhance the empowerment of these groups and increase their transformative potential. Hebinck et al. [29] also reported the difficulty of having adequate participation of target groups and suggested that this was a limiting factor in the transformation process.
The transformative potential of social enterprises could also be higher if conventional farmers were better incorporated as a target group and provided with the opportunity to engage with the values of sustainability and social justice [42]. A stronger focus on helping conventional farmers to transition to agroecological practices could generate further benefits but only participants from the Brisbane food hub indicated that they had facilitated such a group.
The last feature of the social setting is the increase in social networks through community building. Social enterprises ran events, activities and workshops, and provided spaces for community interaction. In addition, initiatives like the Melbourne pop-up market and the Brisbane city farm functioned as a permanent open space for relationship building. Community building was also identified in other studies [14,18,33]. Interviewees from buyers' groups spoke of their wish to know people who shared similar values and be part of their local community. This community can also act as a support network (e.g., a space to talk about personal issues). One buyers' group member exposed her feelings around the individualistic structures of current societies that she believed their buyers' group was trying to counteract: "What happens when we become more self-sufficient, efficient, and convenient, we lose a lot of these community connections that are really valuable for a whole range of reasons. I'm quite passionate about this whole convenience argument and things being self-contained, it creates a whole ripple of other patterns, isolation, anxiety." [Member of a Melbourne buyers' group] Apart from enhancing the sense of belonging, these initiatives have also allowed members to explore their collective power. The communities created by social enterprises allow their members to act as what Carolan [43] defines as an "active food citizen". Active food citizens are interested in challenging routines, understanding, and practices around food, and believe in collective power [43]. Being part of a group that reinforces their values helps members of social enterprises to develop an active food citizen role [19,44]. One example of collective action is the cultivation of food on verges by members of the Melbourne buyers' group. Another is the commercialisation of products made by members from the Brisbane buyers' group, such as cakes and jams. This is creating new avenues for food selling and enhancing the local processing of food, something mentioned by Blay-Palmer et al. [14] as necessary for transformation. A third example is to grow food for the local community, as a manager of the Melbourne urban agriculture initiative said: "Part of their kind of stated goals is that they want to grow for the broader community and not just for the members of the garden. So, they are aware that their area also has some levels of social deprivation and they want to grow in excess to be able to feed some of that need." [Manager from the Melbourne urban agriculture initiative] The social networks of initiatives were also present in their sources of support. Both buyers' groups relied on the provision of space by local businesses to run their activities. The Melbourne pop-up market was developing a network of collection points for food boxes in the community as a way of increasing food access and avoiding waste. The system of voluntary collection points has also been used by food hubs to expand their geographical range. Interviewees reported connections between social enterprises that share goals and provide resources for cooperation. The Melbourne urban agriculture initiative often counted on the help of other social enterprises to conduct workshops and activities. The development of these networks is particularly valuable to establish collaborations that can be actioned when dealing with challenges and promoting knowledge sharing [14].

Operational Model
This section analyses the food social enterprises' operational models in relation to risk-sharing, environmental sustainability, and the ability to replicate their operational models. Table 4 presents a summary of the results for the eight cases studied. Risk-sharing was mainly linked by interviewees to the financial instability of social enterprises. Four initiatives relied on government grants or land concessions to start operations but were now running independently. Members of these initiatives considered that reliance on government funding placed them in a vulnerable position and exposed them to the risk of detrimental policy changes. The other four initiatives used the community as a source of finance during their start-up phase. Buyers' groups from both Brisbane and Melbourne relied on member subscriptions. The Brisbane specialist retailer and the purchase of the Brisbane food hub warehouse also used community funding. In general, community funding has proven to be a useful strategy for social enterprises to overcome financial constraints [21]. The use of community funding and membership structures also devolves the financial risk to a larger group of people. One founder from the Brisbane specialist retailer explained the process: The second feature of the operational model dimension is the concern with environmental sustainability. All food social enterprises investigated were making efforts to reduce the use of plastics, food loss and waste. Above all, the main characteristic which can be linked to environmental sustainability was the commitment of social enterprises with agroecology farming. Agroecology farming as a technique incorporates traditional practices to produce food using resources and interactions that occur in a natural ecosystem [44,45]. It has as main principles decoupling from fossil fuels, abandoning the use of agrochemicals, and fostering biodiversity [5,[44][45][46]]. An agroecological system combines natural vegetation and animals with minimal human interference [45]. The offering of seasonal products by the initiatives is also associated with agroecology farming. Promoting that out of season produce should not be consumed avoids the need to ship food over long distances and reduces interference in natural cycles. Therefore, by adopting agroecology farming, social enterprises are significantly reducing their environmental impacts.
The practices adopted by social enterprises are capable of increasing resilience as well as sustainability. The community building, the increase in flexibility, social networks and trust, and the democratic decision-making resonate with the descriptions of resilient food systems [3,[47][48][49][50]. The resilience-building of social enterprises allows them not only to emerge but also to persist in their urban food systems, while resisting the pressures of globalised markets. Their operational practices may cope well with unexpected shocks and stresses, such as extreme weather events and climate change [18,47].
The members of the Brisbane food hub have experienced an extreme weather event and identified their initiative as having performed a central role in the recovery of their urban food system. The State of Queensland had a large proportion of its land area flooded in January 2011 caused by a series of extreme rainfall events [51,52]. The severity of the floods caused loss of farmlands, several road closures, damage to infrastructure and resources, the flooding of homes and businesses, and deaths [52]. Typical food routes supplying Brisbane were disrupted and the city's main food warehouses were inundated. To cope with this event, the Brisbane food hub used direct connections with small farmers and their social networks, which resulted in the social enterprise being one of the few places that had fresh food available: "I suppose we have not done anything really special; we just know now that we would use social media a lot more, social media made it all happen. So, we did sandwiches here for instance, for other army teams, and then we just put out a Facebook message and five or six cars would turn, load up the sandwiches, and take them to the places that needed food. So, it did not take any special effort, it just happened." [Founder of Brisbane food hub] The final feature is the ability of social enterprises to replicate themselves, something that is essential for increasing their transformative potential. The participants in this study were more interested in the creation of new initiatives rather than the expansion of existing ones in order to increase the diversity and participation of different actors. Replication was achieved by the Melbourne pop-up market, the Melbourne urban agriculture initiative, and the Brisbane specialist retailer. The Brisbane specialist retailer has supported the replication of the buyers' groups model. They partnered with the developer of an online platform for bulk purchasing to make it easier for local communities to create new buyers' groups. An increase in the number of buyers' groups can result in greater demand for the specialist retailer and more people having access to affordable organic food. Replication of the specialist retailer model itself was also desired by its founder but has not been achieved to date. Their vision was of a network of independent specialist retailers across Australia. The history of four of the social enterprises that were started by former members of other local food initiatives also indicated that both replication and the development of new models were possible. Previous experience in food initiatives and having connections with relevant stakeholders has helped those initiatives to establish themselves. It is a different challenge, however, to start initiatives without being part of the food social enterprises environment.

Governance
This section discusses the features of social enterprises that are related to equality in decision-making, transparency and flexibility. Table 5 presents a summary of the results. The equality in decision-making processes is built into the initiative's horizontal structures. The Brisbane city farm and specialist retailer used constitutions that ensure equal rights for participants. Five social enterprises had boards that included community representatives. Membership structures were used by both buyers' group, the city farm, and the Melbourne food hub to achieve higher community participation in the decision-making process. In these initiatives, all members were expected to vote and opinions were equally valued. A member of the Brisbane buyers' group expressed the importance of achieving consensus in decision-making: "The agreement around decision-making was that the aim is always consensus, and really encouraging people to say what they agree or not agree, help people get it out of the line if they are a bit unsure" [Member of the Brisbane buyers' group] As discussed previously, initiatives from Brisbane (the food hub and the specialist retailer) were making efforts to include farmers in their decision-making processes about which produce to offer and their prices, but no farmer was an official member. Including farmers would improve equality in decision-making. The inclusion of more diverse actors in the decision-making process could increase the likelihood of creating different economic models [53].
The second feature analysed was the existence of transparency and flexibility in social enterprises. The transparency of shorter supply chains allowed consumers of the initiatives to know the origin of food, how it was produced, and how much farmers were paid. This created a relationship of trust between social enterprises, farmers and consumers. In the buyers' group, the transparency of the procedures and rotation of roles helped to increase trust among members, something necessary in a system completely run by volunteers. Trust also allowed more flexibility, reduced bureaucracy, and eliminated, in some cases, the need for certifications. A member from the Brisbane city farm, for example, mentioned that they did not require certification of products. This model can help farmers who cannot afford certifications.

Institutional Context
The final dimension of the transformative potential of food social enterprises to be analysed is the institutional context. This dimension involves the existence of enabling policies that can support social enterprises, and their engagement with government actors (Table 6). Policy enablers are different for initiatives located in Brisbane and in Melbourne. Those in Brisbane received minimal government support, especially from local governments, and respondents did not see them as a source of change. The City of Brisbane currently does not have a local food policy or department, for example, and social enterprises have received support from their communities instead. This allows them to be more financially independent, something that is often difficult for social enterprises [21]. Table 6. Summary of social enterprises' features in the transformative dimension of institutional context.

Food Social Enterprise Enabling Policies Engagement with Government
Specialist retailer Brisbane No governmental support. Relationships at the personal level with local council members.

Food hub Brisbane
No governmental support on day-to-day operations, but they had received state funding to renovate their kitchen.
Relationships at the personal level with local council members.

Buyers' group Brisbane
No governmental support.
No connection with government.

City farm Brisbane
No governmental support on day-to-day operations, but they had received a federal grant to build an office.
The initiative is often identified by local government officers to facilitate workshops on urban farming.

Urban agriculture Melbourne
Local government policies supported urban agriculture.
The initiative has helped local governments to establish community gardens and develop food policies or strategies.
Pop-up market Melbourne These were stablished with a grant from the City of Melbourne.
The initiative has collaborated with state community health department to create new pop-up markets.

Food hub Melbourne
A permanent flow of funding has ceased and now state funding is only available to develop new projects.
The initiative has local council representatives as members of their board, is recognised by the local government, and used as a reference for food business start-ups.
Buyers' group Melbourne Council funding was provided for buying equipment.
The initiative has relationships with members of the local council and participates in sustainability fairs.
The Melbourne Metropolitan region is composed of more than 30 local governments and the social enterprises studied were located in four different local government areas. All four local governments have policies that recognise the importance of developing local food systems and growing food in urban areas. Government grants have also been provided to some of the initiatives in the past. A number of researchers have argued that locally appropriate food policies are needed to transform urban food systems [4,14,42]. Institutional support, however, should be shaped in a way that does not limit the innovative capacity of initiatives and promotes community action [36]. Beyond recognition, it is also important to have local policies that encourage food growing in urban areas and stimulate new initiatives [14].
The engagement with government is also different for social enterprises in Brisbane and Melbourne. The local government policies in Melbourne see social enterprises as partners. The food hub had representatives from the local government on their board, while managers from the urban agriculture initiative support local governments with their expertise: By contrast, the Brisbane food hub created a local food strategy to fill the void left by local government, and their managers have lobbied for the council to develop a food policy. The Brisbane food hub founder suggested that a difference between mindsets might be what is hindering the scaling up process: "They have their own agenda who is driven by other messages that are coming from other places, so it has been really hard to [ . . . ] talk about the food system, or a new food system, something that addresses a lot of other social and environmental challenges, so it has been really hard to find traction, because of the silo mentality of government departments and all of that sort of specialisation." [Founder of the Brisbane food hub] The interest of the Brisbane food hub in engaging with government actors and creating changes for the whole society is crucial for enhancing the transformative potential of food social enterprises. Sage [5] mentions that isolation of the current urban food systems will result in benefits only for the few that engage with social enterprises. It is important that initiatives visualise themselves as a crucial actor to policy change. Advocacy for policy and regulatory change can also help social enterprises to reduce the risk of appropriation of the niche improvements created by them as a way of avoiding systemic changes. This 'parasitic' relationship was present in the study of Rut and Davies [36] where the government from Singapore was benefitting from developments created by social enterprises while postponing changes to policies. The description of the engagement that the Brisbane city farm has with Brisbane City Council suggested a similar context from the Singaporean study. Often the local government will bring visitors to the farm or ask for their help in doing workshops on urban farming, but no policy support or funding is provided. In addition, the Waste and Minimisation Department has placed the farm as a city compost hub but has provided minimal support for it: "Local residents register online and the work is actually done by us. The facilities they provided were the black bins and that was it, we are getting a lot of referrals, because a lot of the community composting hubs around Brisbane do not have the same capacity as we do, we are a very established organisation. We can actually make quality compost [ . . . ] so they might be referred to us and that is fine, I think it is great that people are able to recycle their food scraps, but there appears to be no practical support from the council. I would like to actually see a funded program, where someone actually gets paid once a week to look after the compost." [Manager of the Brisbane city farm] All levels of government need to be influenced by the ethos of social enterprises to create transformations. Even if social enterprises create change in local policies, Australian urban food systems are composed of actors that range from the local to the global scale. Transformation of urban food systems and food systems as a whole, therefore, requires that goals, policies, and regulations be re-shaped on a range of areas (e.g., environment, public health, trade) at all levels of government. One of the managers from the Brisbane food hub reported on a collective contribution to a national food plan that had no follow-up: "We did a big community engagement process around presenting the National Peoples Food Plan. So, groups around Australia got together and presented the government with our own policy, National Food Policy. But it just got ignored and the National Food Plan was hijacked by big corporate food, then the change of government happened and the whole thing just got put on the shelf. Nothing has happened since then..." [Manager from the Brisbane food hub] Urban food systems, and food systems more generally, are composed of many players and transformation cannot be created solely by social enterprises. Social enterprises are well placed to assist through their engagement with the local context, together with a commitment to continuous improvement and enhanced responsiveness to social and environmental issues. The engagement with communities on an agenda of social justice and environmental sustainability, together with the development of an economic activity framed by these principles, places them in a unique position. Social enterprises are a practical example for government actors of what a food system based on this agenda would look like. The reality is that many members feel overburdened and do not find the energy or time to advocate for institutional change [35]. It might also be the case that initiatives do not have the skills necessary for doing so. Initiatives should consider, however, if leaving advocacy for policy change aside will allow them to transform food systems.

Final Remarks
This article analysed the transformative potential of food social enterprises using a framework that included societal, operational, governance, and institutional dimensions. The social setting is where the transformative potential was strongest for all social enterprises. Their goals are concerned with creating benefits for their local communities by increasing access to nutritional food that is produced in a sustainable manner. Even though they operate at the local level, social enterprises are aligned with the global agenda of sustainability, climate change resilience, and social justice. Their engagement with different actors (e.g., academia, activists, government, and farmers) has allowed them to develop innovative operational models. A higher emphasis on the inclusion of farmers in their strategic process can potentially create even more innovation. Future research should investigate how farmers' inclusion can take place or if connections created by social enterprises have helped in this process. The main social transformation identified is the initiatives' capacity of creating social networks by developing communities of people who are more aware of food systems issues and open to a new food provision model. The capacity of social enterprise to influence individuals to shift from a passive consumer approach to an active citizenship one is a topic warranting future research.
In terms of operational models and governance, food social enterprises adopt practices that reflect their goals. The initiatives have a democratic structure that comprises a horizontal decision-making process and seeks benefits for social actors and ecosystems. Trust among actors eliminated the need for certifications and has enhanced the flexibility of operations. The operational practices adopted also helped to develop resilience in dealing with shocks such as extreme weather events and to persist in the face of pressures by globalised markets. Three of the social enterprises were able to replicate their models. The replication process could be facilitated with more government support. The environment of social enterprises also seems fertile ground for developing new initiatives. This is a valuable feature as transformation requires diverse solutions for different contexts [4,14].
In terms of institutional context and engagement with government, supportive local policies were more evident in the Melbourne region, as well as a constructive relationship with government. The social enterprises in both regions could also be more strongly involved in advocacy for policy change as a way to develop their transformative potential. Their goals of social justice will not be achieved by creating a better niche food supply alone. The engagement of initiatives with formal institutions is crucial [5,21]. The majority of social enterprises are not seeking government support and have proved to be able to survive without it. Nevertheless, they do need to consider partnerships with governmental actors in order to scale up their activities and create institutional transformation. Using the networks of support already present among social enterprises to assist with the process of pushing political agendas would be one way forward. The responsibility of policy change, however, does not need to be left solely to food social enterprises. Governments can also make efforts to incorporate food social enterprises in their policy agendas [4,42]. Future research can build on the multi-dimensional framework presented by this paper to analyse the transformative capacity of other urban food systems and their actors. This can both expose conflicts of pathways and agendas, and their suitability for dealing with the challenges of climate change, malnutrition, resource scarcity, and population growth.