The rising stars of social innovations: How do local governments facilitate citizen initiatives to thrive? The case of waste management in Brussels and Hong Kong

Waste management is an important issue in cities. The achievement of a zero-waste future relies on the collective actions of individual households. Indeed, citizens are initiating social innovations in waste management in their communities. However, citizen initiatives (CIs) encounter strategic and operational barriers in the process of social innovation. They often seek support from local governments to overcome these barriers. These initiatives need a “ facilitative ” local government that is responsive and enables the initiatives to thrive without too much interference from them. Yet, there is a lack of conceptual and empirical understanding of such a new facilitative role of local governments. Furthermore, we do not know whether and how this facilitative role differs between different types of CIs and between different policy contexts. Based on 24 in-depth interviews with CIs and government officials, this study explores the various facilitative practices offered by the Brussels and Hong Kong local governments, explains patterns of facilitative practices between different types of CIs, and clarifies how divergence in policy contexts influences what, why, and how much the local government can facilitate CIs. This study finds that common facilitative practices include financial and administrative assistance, but certain barriers requir-ing actions from the government and other actors remain unresolved. Local governments need to review their facilitative role to provide more effective support to CIs as new agents of sustainable urban development at the community level.


| INTRODUCTION
The emergence of wicked environmental problems such as climate change and resource scarcity calls for collaboration between multiple stakeholders from different levels and sectors of society, including citizens (Bekkers et al., 2014). Former British Prime Minister David Cameron paid particular attention to citizens' role in solving wicked societal problems and proposed the concept of "Big Society" in his manifest (Cameron, 2010). It emphasises the idea of community empowerment, social entrepreneurship and liberalism, encouraging citizens to take an active role in managing their communities. Sharing the responsibility of public service provision with citizens is also seen as a solution to financial constraints in the public budget. It also avoids problems to appoint market operators on quality control, coverage, and coordination (Healey, 2015).
Meanwhile, citizens are motivated to steer changes in their community. Self-organised informal citizen initiatives (CIs) often arise from dissatisfaction with governments' actions (Edelenbos & van Meekerk, 2011). They seek alternative solutions to the current institutional arrangements to better address social challenges (Bekkers et al., 2014). CIs mobilise resources to drive social innovation independent of the local authorities, and have their power and freedom to act outside of the existing public service system for the collective benefits of the communities. The processes and interventions of CIs are described as social innovations (Mulgan et al., 2007;Seyfang & Haxeltine, 2012). CIs are observed to play a role in various environmental issues such as renewable energy, community green space development, and urban waste management, to name a few (Mattijssen et al., 2018;Mees et al., 2019;van der Schoor & Scholtens, 2015). CIs take on some responsibilities for the provision of environmental services that were previously exclusively provided by public authorities. While some scholars have criticised the delegation of responsibility to citizens in public service provision as a neoliberal governance strategy due to budget cuts and as false promises to co-produce public services with citizens (e.g., Kleinhans, 2017;Lister, 2015), others perceive it as the empowerment of citizens and realisation of participatory democracy (e.g., Bakker et al., 2012;Edelenbos & van Meekerk, 2011).
CIs are not isolated actors in social innovation. They seek new relationships with local governments to overcome strategic and operational barriers, which cannot be solved by themselves (Jakobsen & Andersen, 2013). For instance, a lack of financial resources and management skills may limit the capacity of CIs to grow (Mulgan et al., 2007). The emergence of CIs has gained attention from scholars in the past decade. Some scholars study the characteristics of CIs, how they pursue social innovation and which barriers they have encountered (Mattijssen et al., 2018;Middlemiss & Parrish, 2010;Seyfang & Haxeltine, 2012). Bekkers et al. (2014) made a theoretical sketch of the enabling conditions for selforganisation and the changing role of governments in support of CIs.
So far, only a few scholarly works have discussed the facilitative practices of local governments in support of such initiatives. For instance, Bakker et al. (2012) combined various theoretical settings to analyse how local governments can facilitate CIs through providing networks and process management. Igalla et al. (2020) proved that government support has a significant positive influence on the performances of CIs, but a further elaboration on facilitative practices offered by the government was absent. Government facilitation is believed to be beneficial for achieving CIs goals (Bekkers et al., 2014;Healey, 2015), but little is known about how local governments facilitate CIs on the ground. Moreover, existing literature has discussed the facilitative role of local governments generically, but it has not yet put sufficient emphasis on how such role might be influenced by different barriers encountered by CIs that use different approaches in social innovation. It is necessary to classify types of CIs for two reasons. First, CIs aiming at achieving different social outputs may not use the same approach. Thus, they encounter different sets of barriers in various phases of innovation and hence require different facilitation from the local government. Second, it allows a systematic study on the facilitative role of local governments, which is fundamental for developing theories on this topic in the future. Mees et al. (2019) developed the ladder of government participation in CIs, distinguishing different roles of local governments, including the facilitative role, and corresponding practices. They suggested a further scholarly investigation into specific roles that local governments take on in relation to specific CIs, and into how such facilitative roles may change over time depending on the need and pace of CIs that they support. Igalla et al. (2020) argued that government support can be vital in certain phases of CIs and called for further research on how government support relates to characteristics of CIs. This article enriches empirical research on CIs and facilitative practices of local governments to support CIs and sheds some lights on how to facilitate CIs more strategically.
This paper aims to explore the facilitative practices of local governments for different types of CIs for waste management in two distinct policy contexts: Brussels and Hong Kong. Waste management is a common challenge for every local government. A World Bank report estimated over 2.01 billion tonnes of Municipal Solid Waste (MSW) were produced in 2016 and warned that it is expected to grow to 3.40 billion tonnes in 2050 as the population doubles (Kaza et al., 2018). Improper waste management poses threats to public health and the environment, which can be costly to society and the economy (UNEP, 2016). As citizens are household waste producers, the reduction of waste and the achievement of a zero-waste future relies heavily on the partnership with communities and individual households for collective action (Robbins & Rowe, 2002). CIs demonstrate great potential in initiating and implementing waste management innovations from the bottom up. In contrast to a top-down approach, CIs are capable of developing innovations according to the characteristics of the community, which allows them to respond better to the local situation and interests of the community (Hoppe et al., 2015). CIs working on waste management cover a wide variety of waste types using different approaches, such as the sharing of goods, repairing of broken items, and awareness-raising campaigns (Angelidou & Psaltoglou, 2017).
This study proceeds as follow. Section 2 presents the conceptual framework, which defines and classifies CIs, as well as gives an overview of facilitative practices offered by local governments derived from a review of different studies. Section 3 describes the methodology. Section 4 demonstrates different facilitative practices offered to CIs by the two local governments and clarifies the similarities and differences in relation to different types of CIs and differences in policy contexts. The study ends with a discussion and a critical reflection on the facilitative role of the local government.
2 | CONCEPTUAL FRAMEWORK 2.1 | CI as an actor of social innovation CIs are self-organised community-based social innovations that aim to strengthen local communities (Schartinger et al., 2019). Although the action arena is at the local level, CIs do not limit their scope to specific local problems. Global societal problems such as environmental pollution and climate change are often addressed (;Schartinger et al., 2019). Citizens' engagement is rooted in the sense of community stewardship that goes beyond personal interest (Krasny & Tidball, 2012). In this study, a CI is defined as a self-organised body composed of a coalition of citizens and social interest groups that carry out their social innovations for the common good. The concept of social innovation can be understood by breaking it into two parts: "Social" refers to the output or outcome that the innovation wants to produce. In contrast to business innovation, social innovation is motivated by social well-being rather than profit maximisation (BEPA, 2011;Mulgan et al., 2007). "Innovation" refers to both the process and the output dimensions. The essence of innovation is to think and do differently from the existing practices, and to design alternative solutions that are more efficient, effective, and sustainable (Brown & Osborne, 2013;Phills et al., 2008). Innovation can be both physical and non-materialistic, such as products, services, principles, and models (BEPA, 2011;Davies et al., 2012). To accumulate social capital for social innovation, relevant stakeholders from different backgrounds who are involved in the development and the adoption of innovation become co-creators, calling for new configurations of organisations and relationships to tackle social issues (Schartinger et al., 2019;Voorberg & Bekkers, 2018).
Our starting point is that different types of CIs encounter different barriers and thus require different kinds of facilitative practices of the local government. Likewise, we assume that facilitative practices of local governments will not only differ according to the type of CI but also according to the phase of social innovation (Ansell & Gash, 2007;Healey, 2015). Recent studies have proposed different typologies of social innovation. For instance, differentiating social innovation by the societal domain that drives the innovation and the degree of interaction with broader society (Schartinger et al., 2019); and by the relationship between CIs and government bodies (Edelenbos et al., 2018). This study borrows the three-type classification from the Bureau of European Policy Advisers (BEPA) report because of its universal applicability (see Figure 1). BEPA was a former Directorate-General of the European Commission, preparing the EU for future challenges and opportunities (European Political Strategy Centre, 2019). Its classification is suitable for analysing social innovations from different backgrounds (schools, citizens, businesses, or government) and different levels (community, local, regional, or global) (BEPA, 2011). A flexible borderline between different sectors and scales is important because social innovation emphasises the change in roles and relationships among various actors. Besides, each type of social innovation provides a distinctive type of social output, which potentially leads to a variation of facilitative practices needed from the government. Moreover, as the typology from BEPA (2011) is not tailor-made for a specific social issue, it is applicable to study social innovation in waste management.
Type A initiatives aim at meeting social needs that are neglected or insufficiently provided by the state or the market. The definition of "social" here is meeting the need of the vulnerable groups in society. This type of initiative does not seek financial gain, and even if there is a profit, the principle is to re-invest in achieving their missions. For example, Food Grace is a Hong Kong CI that collects leftover food from stores and redistributes it to low-income families. Type B initiatives tackle social challenges by creating economic value to social well-being, redirecting the market towards society. Solving social problems becomes an economic opportunity, generating productivity and economic value for the society.
For example, Vert dÍris from Brussels produces compost from organic waste and sells locally grown crops and compost to households and restaurants. Lastly, type C initiatives aim at reshaping society by changing the organisation of institutions and the relationships of actors. They promote change in fundamental values, organisational structures, and division of responsibilities, which steer society to a more participatory one. For example, WORMS asbl from Brussels offers composting coaching and training workshops for individual citizens and CIs, which takes over the responsibility of the local governmental institute and empowers others to innovate.
Regarding social innovation phases, the four-phase model from Mulgan et al. (2007) is adopted (see Figure 1). Although there are studies that distinguish even more phases (e.g., Murray et al., 2010), distinction into four phases is regarded as sufficiently detailed and parsimonious to capture differences in facilitative practices. Social innovation begins from spotting a social problem. Social innovators identify potential solutions and adjust them in the light of experiences (phase 1). Then, a few promising ideas are developed further as prototypes. These prototypes are tested and improved until a working solution is found (phase 2). Next, the working solution is promoted in society and scaled up (phase 3). The momentum of social innovation is continued after it has become the new mainstream solution for a social problem (phase 4).
The purpose of classifying CIs is for the analytical ability to distinguish ideal types. CIs may want to generate more than one type of social output in a real-life context. Similarly, overlap of innovation phases may exist. In this study, CIs are grouped by the dominant type of social output they want to produce and the innovation phase they identified themselves with during the interview.

| Facilitative practices of the local government
To better respond to the social challenges at the local level, many scholars suggested governmental officials to adopt a facilitative political leadership, which promotes interaction and communication among different actors, including the public (Bussu & Bartels, 2014;Bussu & Galanti, 2018;Ford & Green, 2012). In this study, facilitative practices of local governments refer to actions that support social innovations by CIs. This section summarises empirical observations and theoretical arguments from the literature on the facilitative practices of local governments across different contexts. We identified five main categories of facilitative practices in the emerging literature: (i) financial assistance; (ii) technical assistance; (iii) capacity building; (iv) networking; and (v) flexibility in rules and procedures.
Financial assistance was defined as one of the most important factors to enable social innovation by The Economist Intelligence Unit (The Economist, 2013). Public funding is spent on high-risk innovations in priority areas (Mulgan et al., 2007). Other than funding innovation directly, intermediary bodies such as innovation incubators and accelerators for developing social innovation projects are financially supported too (Mulgan et al., 2007).
Technical assistance refers to providing expertise, knowledge, technology, and tools (Bekkers et al., 2013;Wang et al., 2014). As many environmental issues are technical in nature, local public administrators can facilitate the development and implementation of community sustainability practices by acquiring technical support from professionals (Wang et al., 2014). Bekkers et al. (2013) added that the advancement of Information and Communication Technology (ICT) allows the government to facilitate information exchange and mobilise knowledge for innovation among different actors.
Facilitation of capacity building refers to supporting organisations to build, maintain, and evaluate the skills and resources to reach their goals (Oluwaseyi & Author, 2018). Scholars suggested that social innovation can be accelerated by building capacity within the government, CIs, and the public. For instance, the government can increase its capacity for social innovation by familiarising itself with social innovation processes and building trust among other stakeholders (Ansell & Gash, 2007). It can help CIs by sharing skills that enhance their performance and durability (Healey, 2015). Lastly, the government can increase public acceptance of social innovation by nourishing social entrepreneurship alongside traditional school subjects (BEPA, 2011).
The government can facilitate the networking of CIs with different actors, such as experts, the business sector, government officials, and other CIs (Hoppe et al., 2015). Enlarging the network of CIs not only enables the sharing and dissemination of knowledge and experience but also provides opportunities for actors to collaborate and pool resources, which can strengthen the collective capacity to drive social innovation (Bekkers et al., 2013;BEPA, 2011).
Social innovation can be promoted if relevant policies are integrated at different levels of government, as this can maximise the efficiency, impact, and acceptance of an innovation (BEPA, 2011). When an initiative needs support from the higher-level government, procedures can be over complicated for non-experts, thereby discouraging social innovation at a lower level (van der Schoor & Scholtens, 2015). Therefore, when social innovation is well integrated into a government's policy coordination, it can offer more flexibility in rules and procedures to CIs.
In facilitative leadership literature, government officials can facilitate partnership with CIs by accepting and valuing opinions. Previous research found that public inputs are weakly linked to internal policy changes because policymakers may fear losing control and power to citizens (Bovaird & Loeffler, 2012;Edelenbos, 2005). Partnership implies a non-hierarchical and innovative approach to collaboration, and leaders need to be open-minded to new ways of thinking and to empathise differences among different actors (Ford & Green, 2012).
Facilitative leaders can foster change within the government by promoting an idea, by gathering political support, and by reducing resistance to change (Bekkers et al., 2013).
In sum, Table 1 below is used to inform and structure our analysis of which kinds of facilitative practices are offered by local governments to reduce or remove the barriers of CIs.

| METHOD
To obtain an in-depth exploration of the facilitative practices of local governments for CIs in waste management, this study used an embedded multiple case study design (Yin, 2009). In total, 13 CIs were selected as case units in the two cities of Brussels and Hong Kong.
These 13 CIs cover all types and phases of social innovation (see Appendix 1). Such a comparative design enabled us to analyse and compare the facilitative practices of local governments in two very distinct policy contexts. The logic of a "most different" comparative design was followed (Burnham, Lutz, Grant, & Layton-Henry, 2008;Pickvance, 2001) to explore similarities and differences in facilitative practices under different circumstances and test the generalizability of findings.
The two cities of Brussels and Hong Kong were selected for their dif- The 13 CIs were selected for their coverage of the three types and four social innovation phases (see Section 2.1). More experienced CIs were chosen because they are more likely to have gone through different innovation phases. Hence, they can add value to our study. These CIs were found from databases of facilitation programs offered by the two local governments. Two of the same three types of CIs were studied in each city. Lastly, the selected CIs must have received or intended to receive facilitation from local governments, to ensure that they have a clear understanding of the facilitative role of local governments.
Twenty-three face-to-face interviews (and one written response) and content analysis of 86 relevant policy documents were the primary data collection methods. Respondents of CIs were the founders or project leaders. They were highly involved in their initiatives and experienced in designing and implementing social innovation in their communities. We also interviewed government officials as the representatives of various government departments that offer facilitative programs to CIs. These government officials were contact points for CIs. Fifteen interview invitations were sent (seven from Brussels and eight from Hong Kong). Ten government officials were successfully interviewed, and one provided a written response.  (Mulgan et al., 2007; The Economist, 2013) Technical assistance (Bekkers et al., 2013;Wang et al., 2014) Capacity building of: • The government (Ansell & Gash, 2007; BEPA, 2011) • CIs (Healey, 2015) • The society (BEPA, 2011) Networking support (Bekkers et al., 2013;BEPA, 2011;Hoppe et al., 2015) Flexibility in rules and procedures (BEPA, 2011; van der Schoor & Scholtens, 2015) Accept and value opinions (Bekkers et al., 2013;Bovaird & Loeffler, 2012;Ford & Green, 2012) independent of the local government. Some governmental departments offered financial support to intermediaries to provide facilitation to CIs.
An overview of the facilitative practices for the different types of CIs is provided in Table 3. An overview of the facilitative practices of the two cities of Brussels and Hong Kong is provided in Table 4. In the following paragraphs, each category of facilitative practice is discussed in more detail.
Financial assistance is a common facilitative practice offered to all types of CIs (see Table 3). The value of a grant ranged from €6000 to 35,000 and HKD 541,704 to 2,691,406 (equivalent to €61,927-307,681) per project in Brussels and Hong Kong, respectively (see

| Facilitative practices in relation to the barriers of the different CIs
Section 4.1 and Table 3 show that local governments provided different facilitative practices to different types of CIs. We will now explain T A B L E 4 Facilitative practices offered by Brussels and Hong Kong

Brussels Hong Kong
Financial assistance -The value of a grant ranged from €6000-35,000 -Provided in several instalments -Some project calls provided money at once and required CIs to return the unspent money -The value of a grant ranged from €61,927-307,681 -Provided in several instalments Administrative assistance -Application documents and guidelines were available online. Inquiry email and hotline were provided.
Technical assistance -Provided by government officials -Knowledge and skills especially related to a circular economy -Provided by intermediaries -Knowledge and skills related to social enterprises.

Capacity building -Experience-sharing workshop with other CIs
Networking support -Created an ecosystem with actors working in sustainable economy and relevant government departments for type B initiatives. -Increase CIs' visibility through government website and social media -Annual Zero-waste fair to showcase innovation projects -Provide contact information of relevant projects and government departments.
-To type B initiatives through intermediaries -Provide contact information of relevant projects and government departments Flexibility in rules -Small deviation of the project from the original plan is allowed.
Accept and value opinions -Frontline government officials collected opinions from CIs during regular project evaluation meetings.
-Opinions were formally collected through irregular inspection of the funding schemes by the Audit Commission this by analysing the barriers that these CIs encountered. Since each type of CI encountered different barriers during their innovation, they may require a distinct form of facilitation. Financial and administrative assistance are common facilitative practices of both local governments for all types of CIs (see Table 3).
This is logical because a lack of financial resources is the common barrier among different CIs in phase 2 (testing pilots) and 3 (scaling-up) (see Table 5). Rents and salaries were high and unaffordable, especially for non-profit making type A and C initiatives. Flexibility in rules and acceptance of opinions from CIs were also offered to very limited extent, which seems to be insufficient to remove barriers such as the lack of supportive legislation and policies.
For example, the regulation on organic waste in Brussels did not distinguish between animal waste and kitchen waste. Organic waste was considered dangerous, so CIs were not allowed to take the organic waste out of the waste stream for other purposes such as community composting, thereby limiting innovation in organic waste management. In Hong Kong, food donors hesitate to give away leftover food to CIs as no legislation exempts them from their responsibility when food receivers are harmed unexpectedly after consuming the food.
Solving these barriers require more flexibility in rules and acceptance of opinions to trigger changes at higher government levels.
Some barriers appear to remain unresolved by local governments.
For example, networking support to type A and C initiatives was minimal, because the local governments did not actively serve as a linking pin between the initiative and other actors. Only one funding scheme in Hong Kong had a formal networking platform connecting CIs with actors across sectors. Moreover, many CIs identified that the willingness to adopt social innovations in society is low. Innovation projects on waste management promote new ideas like sharing items instead Friend Action project from People Service Centre as an example. It collects leftover food from local markets and redistributes it to lowincome families. People did not trust that leftover food was safe to consume, so the participation rate was low. The lack of social willingness to change was a barrier to CIs in phase 2 and 3. Lastly, various types of facilitative practices were available throughout the funding period. There were hardly any patterns in each type of facilitation offered by local governments among different innovation phases, except for the fact that the content of technical assistance to type B initiatives changed as mentioned in Section 4.1.

| Facilitative practices in relation to different policy contexts
Although facilitative practices offered by the two local governments share many similarities (see Table 4), differences are manifested in the details regarding what, why, and how much they were provided. Both CIs and government officials in Hong Kong identified financial assistance as the main (in most cases, the only) facilitation provided. Often, submission of reports is the only way of communication between the local government and CIs after receiving the fund. In contrast, the facilitative practices from the Brussels' government were more diverse and comprehensive. This can be explained by the differences in the expected role of CIs in public policy. In the 5th regional waste plan, the Brussels' government "encourage(d) and support(ed) individual and collective initiatives of citizens who want to put into practice the changes towards sustainability" (Brussels Environment, 2018, p. 58).
The Smart City policy also pointed out that the local government encouraged its citizens and other actors to develop new urban dynamics through imagining new modes of organisation and producing new urban services (Government of Brussels, 2019). Both policies recognise that CIs provide local knowledge and new ideas, and therefore the local government is encouraged to develop facilitation to CIs.
However, in Hong Kong, the Blueprint for Sustainable Use of Resources acknowledged the existence of CIs for waste management without showing any intention to learn from CIs, or to integrate their work into the regional waste plan (Environment Bureau, 2013). Thus, the purpose of offering facilitation to CIs in Hong Kong is not to support social innovation on the systemic level but to pay for a changing lifestyle and to promote new policy (i.e., waste charge) at the community level. The relationship between the Brussels government and CIs is more horizontal, while the relationship between the Hong Kong government and CIs is more vertical. Since the Brussels government gives CIs a role in waste management policy and has a closer relationship with them, this may explain why the Brussels government is more willing to invest and offer a more diverse set of facilitation to CIs than that in Hong Kong. This study shows that the expected role of CIs in public policy influences policy strategies to facilitate CIs, affecting "what" is offered to them.
Both cities offered a unique set of facilitative practices to type B initiatives (see Table 3). Yet, the rationales of the two cities to offer facilitation to type B initiatives are different. The Brussels government positioned itself as an innovative and pioneering European Region on public policy to support a circular economy (Brussels Environment, 2016 (Cheng & So, 2015), leading to a lower demand for social innovation to reduce waste. Therefore, differences in the two policy contexts also influence "how much" support the local government can orchestrate and offer to CIs.

| DISCUSSION
This study has demonstrated that facilitative practices offered by local governments differ according to the type of CIs (see Table 3 Financial assistance is a common facilitative practice in both cities. This is especially essential for non-profit oriented initiatives (Types A & C). The two local governments provided non-financial facilitative practices like coaching and networking support as well.
The Hong Kong government partnered with intermediaries to expand the range of facilitative practices for CIs, which can be inspiring for other local governments even when the resources and knowledge to support CIs are limited. The role of intermediaries has been proven to increase the effectiveness of social innovation of CIs by maintaining the network with other initiatives and actors (Hargreaves et al., 2013), by facilitating diffusion and adoption of the innovation in society and the government (Hartley, 2008), and by facilitating communication between CIs and the authority (Davies et al., 2012). Therefore, non-financial related facilitative practices deserve more attention from both the local government and from scholars, to explore their potential and effectiveness in fulfilling the facilitative role of the local government.Even though existing literature has identified potential barriers encountered by social innovation initiatives in various innovation phases (BEPA, 2011;Murray et al., 2010), this study found that CIs generally did not find phases 1 and 4 challenging.
CIs have a rich knowledge of the local situation, making them more capable of generating innovative ideas than the local government. They need to re-submit the application for funding to receive a range of facilitation after a period (usually 6 months to 1 year). Case studies from the United Kingdom and Ecuador proved that small-scale CIs suffered from governmental budget cuts, so some CIs saw governmental funds as "unreliable" (Healey, 2015;Johnson, 2009 This study found it important to recognise and integrate the contribution of CIs in public policy. Innovation in waste management that promotes radical change requires formal channels in the decision-making process to challenge the status-quo practice (Pollans, 2017).
The Brussels government, which regards CIs as partners for mutual learning in waste management policy, showed a closer relationship with CIs and more interest in learning from them when compared to the Hong Kong government. The literature on social innovation in the public sector has discussed various ways of citizen empowerment (Bakker et al., 2012;Davies et al., 2012), but assigning a role to CIs in public policy design and delivery as a means of citizen empowerment is a new idea, which can be an interesting topic for future research.
Finally, we would like to bring up some potential drawbacks that may arise from facilitating CIs. As CIs are self-organised, they are not elected bodies in the communities. The representativeness and legitimacy of such initiatives are questionable (Healey, 2015). Moreover, initiatives that are started by more passionate and skilled citizens are more likely to gain support from local governments (Mees et al., 2019). The needs of less well-off citizens may be neglected due to low coverage of CIs in deprived neighbourhoods. CIs tend to be over-represented in better-off neighbourhoods, and this may exacerbate existing inequalities in society (Healey, 2015;Mees et al., 2019). Both Healey (2015) and Mees et al. (2019) suggest that the local government should take on an additional role by coordinating and overseeing the performance of CIs and redistributing benefits to reduce inequality among them.
However, the CI representatives interviewed in this study expressed that this requires a high sensitivity of the local government because they do not want the authority to intervene and eventually take control over their innovations.

| CONCLUSION
This study has explored which kind of facilitative practices are offered by local governments to three different types of CIs in waste management in Brussels and Hong Kong. It confirms that local governments play an important role in providing facilitative practices to CIs. Financial assistance, administrative assistance, and acceptance of opinions are commonly offered to all types of CIs by both local governments.
Flexibility and networking support are available to a limited extent.
Networking opportunities with business actors, a capacity building program, and technical knowledge sharing are exclusively offered to CIs that create economic value to social output (type B initiatives).
Our analysis finds that patterns of facilitative practices offered per type of CI can be explained by the different barriers that they encountered. However, there are hardly any differences in facilitation offered by local governments among different innovation phases. In addition, barriers such as the lack of supportive legislation/policies and the lack of societal willingness to adopt social innovations remain unresolved in both cities. This implies that facilitative practices should not focus only on CIs, but also orchestrate actions from other actors such as the local government itself, the business sector, and the general public.
Although there are similarities in facilitative practices offered in the two cities, differences in policy contexts influence what, why, and how much facilitative practices are offered. For example, the Brussels regional waste plan affirms CIs as important contributors to local waste management. The local government has a higher risk acceptance towards social innovations than the Hong Kong government, which gives the Brussels government more interest to deepen and broaden "what" to offer to CIs. Besides, the Brussels government supports type B initiatives because it seeks new business models for a circular economy from CIs. By contrast, the Hong Kong government perceives type B initiatives as social enterprises and facilitates them according to social enterprise policies. It shows that the variation in policy can influence "why" facilitative practices are offered. Lastly, the Brussels government has more intra-governmental collaboration and stricter waste management regulation for the business sector than the Hong Kong government, which encourages different governmental departments and business actors to support social innovations. Thus, more resources are pooled for waste management by CIs, affecting "how much" facilitative practices are provided.
Further research on the facilitative role of local governments in other policy contexts and other environmental issues can bring insights from this study forward and test their generalizability.
Moreover, the policy context is only one of the factors accounting for the differences observed in the two local governments. Further research can be done to explore whether other factors such as the wider economic context and cultural differences can explain the findings. These ideas for future research can help develop a conceptual framework for facilitating social innovations by CIs in environmental governance, as guidance for local governments to develop strategies in this field.