Designing regional innovation systems in transitional economies: A creative ecosystem approach

Regional Innovation Systems (RIS) have become a policy panacea for states looking to develop their economies. However, much of the research on RIS is from Western developed economies, which have established infrastructures and institutional governance networks. Yet, in transitional economies, the growth rate in some parts of the economy is so rapid that policy makers and institutions are unable to record changes as they occur, including the emergence of new economic sectors, such as the creative industries. This results in knowledge gaps, leading to an inability to understand, identify or react to the needs of those nascent sectors. Our research paper, through the analysis of creative industries in Vietnam, proves that taking a creative ecosystems approach to designing RIS will bridge these knowledge gaps by providing a mechanism through which information can be collated and fed into the policy process. Our paper facilitates this process by developing a model to understand the characteristics of the creative ecosystem and as well as a flipped model of policy diffusion to allow bottom-up development of policy in transitional economies. We then discuss how these models can be used by policy makers to design a more informed RIS to meet the needs of the sector.


| INTRODUCTION
The role of the creative economy is widely understood in developed economies (Chapain & Comunian, 2010;Florida, 2002;Mellander et al., 2013), but very little has been researched in transitional countries, especially in Asia. Therefore, the contributions and impact of creative industries in this context remains unclear (Fahmi et al., 2017). This paper analyses the impact of creative industries in transitional contexts, in particular the economic, social and economic values this sector creates, and in doing so it addresses a number of the core themes of this special issue. The paper uses insight from Vietnam, a country in the Global South that is seldom featured in the wider academic literature. It also takes the literature on innovation systems out of their comfort zone by exploring a sector (the creative industries) that has provided the conditions for innovation whilst operating outside of formal governance structures.
We use the term transitional (Ateljević & Trivić, 2016; rather than developing economies as the economic context of Vietnam has moved beyond that of a developing economy. We define a transitional economy as one where the economy is rapidly changing: it is moving from being a developing one and it is displaying some characteristics that are ascribable to developed economies, but it has not completed the transition yet. In a transitional economy, the rapid rate of growth in parts of the economy and society has created a situation whereby government cannot keep up with the pace of change, resulting in policy and infrastructure vacuums. For example, Vietnam's considerable rise up the Global Innovation Index, coupled with a move from lower income toward lowermiddle-income country status, could give an impression of an economy and policy landscape that is bearing fruit. However, as various sources indicate, such as the World Bank and Oxfam reports (Nguyen Tran, 2017), significant parts of the Vietnamese economy and society are being left behind. In this transitional setting, sectors considered less strategic from an investment, property and land perspective, such as the creative industries, are not being supported by policies designed for them. Subsequently, progress is hindered .
In many developed countries, Regional Innovation Systems (RIS) have become something of a policy panacea used to address regional imbalances in national economies. These theories have been used across various sectors, as well as extensively in the creative industries (Trippl et al., 2013). However, the very nature of transitional economies, as described above, exposes gaps in both infrastructure and knowledge of the economy, making it more difficult to design appropriate RIS in these settings.
We will show in this paper that, when the RIS model is used in transitional countries weaknesses in the model become apparent. In particular, there is a considerable vacuum in understanding the value and impact of emerging, such as the creative industries. To overcome this, scholars and policy makers need to adopt a new approach. We contribute to the creation of a new framework by bringing together and using as a continuum two concepts to analyze value creation: RIS and what we define as creative ecosystems.
Whilst the literature on ecosystems largely focuses on identifying economic value creation (Adner & Kapoor, 2009;Clarysse et al., 2014), in a creative ecosystem social and cultural values are as important as economic values.
In this paper, we will identify the values produced by the creative sector in a transitional setting, as well as the ascertainment of policy requirements to achieve further growth.
RIS and creative ecosystems should not be seen as competing concepts but as complementary: only using these two concepts in tandem can we explore and identify the different kinds of values being established by the creative industries and achieve the appropriate development of infrastructures.
Whilst RIS allows the identification of economic value, we suggest using a creative ecosystem approach to analyze social and cultural values, which are crucial to the socio-economic development of the country as demonstrated in . We define a creative ecosystem as a network of dynamic relationships between actors (intended as individuals as well as institutions and infrastructures) whose goal is to create economic, social and cultural values through creative and innovative activities. It is vital that any policy framework aimed to support the sector is designed to facilitate the economic, social, and cultural values and their impact on the community.
Taking RIS out of their usual North American or Eurpean contexts and into transitional settings uncovers a weakness in their construction. Although the intention in much of the RIS literature is for a bottom-up approach to policy making, when it is used to create appropriate infrastructures in a transitional setting that lacks of dedicated institutions, the is the danger that a RIS approach becomes a topdown approach to policy making. This would result in not meeting the needs of the sectors they are designed for . To avoid this eventuality, policy makers need to understand the creative ecosystems that exist in their settings in order to be able to support them. Therefore, we do not dismiss RIS models, but we propose to adapt them to transitional contexts using creative ecosystems as a lens. This will facilitate their bottom-up construction through what we term as a "flipped model of policy diffusion" (see Figure 2 in Section 7).
Our contribution in this paper is three-fold. First, we determine the characteristics and the role of the creative ecosystem in a transitional economy. We find that the creative ecosystem is a network that acts as a fundamental mechanism for creative organizations to work despite the lack and absence of an institutional infrastructure. We define a creative ecosystem as a complex network of actors whose interactions produce innovative and creative processes and outcomes, which are shaped by economic, social, and cultural values.
Second, our paper contributes to the debate initiated by Brown et al. (2016) on the failure of policy makers to account for local conditions: policy makers need to account for the local ecosystems, rather than imposing a top-down implementation of a RIS. Whilst Brown et al. (2016) find that the innovation system implemented in Scotland on an existing ecosystem is not functional because it suffers from institutional thinness, we theorize that a creative ecosystem that has yet to be incorporated into any meaningful policy framework or intervention in the Vietnamese context can survive despite institutional and policy absence. In fact, we assert that a creative ecosystem thickness compensates for institutional thinness (or even absence) in the creative economy that is often manifested in these transitional countries.
Third, in order to understand the value of creative and cultural industries in the Vietnamese economy, inferring from the data analysis, we suggest that policy makers cannot rely only on produced economic value, which is the most used metric in Western studies and reports. Instead, they need to translate the economic, social, and cultural values that these industries create into policy frameworks to support the development and establishment of creative industries. To exemplify this approach, we design a flipped diffusion model that places creative ecosystems at the heart of any RIS.
The paper proceeds as follows. First, we briefly discuss the creative industries, the sector our research is based on, before moving on to review the literature on the systems of innovation and its critique provided by the RIS literature. Through our analysis of the literature, we analyze the issue of applying RIS in a transitional setting, as this approach does not identify the values being created nor the gaps in infrastructures. This problem, as explained, is overcome using a creative ecosystem approach, which we define in the discussion. We then provide an outline of our methodology. In the remainder of the paper, we fill the knowledge vacuum for policy makers regarding the creative industries in transitional economies by employing a creative ecosystems approach to explicate the three kinds of values (economic, social, and cultural) being produced by the creative industries in Vietnam, and by proposing policy recommendations for the development of a locally informed RIS suitable for the creative sector in transitional economies.

| CREATIVE INDUSTRIES AND INNOVATION SYSTEMS
This section outlines the main features of creative industries, the sectoral setting of our research. We then unpack RIS research tracing its development and departure from National Innovation Systems (NIS) and discussing its suitability from a transitional context and construct our case for adding a creative ecosystems perspective to strengthen the development of RIS models in a transitional context.
The creative industries have long been studied in regional studies. Florida's (2002) concept of the creative class has proved an attractive and persuasive one for policy makers looking to develop their economies through the cultural and creative industries. Mellander et al. (2013) argue that the creative class is a truly global phenomenon. However, there are three main problems with this theory: first, much of the work on this global spread remains focused on Western or developed economies and little research has been carried out in transitional settings; second, further theories of creative economies need to be developed ad hoc for these contexts, as they cannot be a translation of the Western concepts into the Asian context, as pointed out by Fahmi et al. (2017); and third, there is a misunderstanding of the motivations of supporting creative industries in non-Western contexts. In fact, Fahmi et al. (2017) argue that the motivation for applying a creative economy policy in developing 1 economies is different compared to developed economies, as developing countries tend to use creative economy to compete in production costs rather than knowledge creation or neighborhood regeneration. As Peck (2011) and Peck and Theodore (2015) demonstrate, there are significant dangers of "mobility mutation" in moving policy ideas that have worked in one setting to another without first understanding why they worked in the original context and what the context is into which they are going to be implemented.
In developed economies, it is common for both academics and policy makers to use the notion of systemic views of innovation (Edquist, 2009) to spark creativity and innovation in the economy, whereby the incidence of innovation in the economy is directly influenced by the institutions and infrastructures in place. However, we advocate that this is not working in transitional contexts, as neither the policy frameworks nor institutions are developing at the same pace as innovation and growth are occurring within sectors of the economy. This results in a situation where innovation and creativity hotspots can be mismanaged by policy makers who are unable to react to the rapidly changing landscape they are dealing with. This ultimately can result in innovation systems being proposed that are not suitable for their location. Therefore, as we will explain in the discussion, we propose to use the notion of creative ecosystems to facilitate an understanding of the sector. Here the social, cultural and institutional factors are locally mobilized (since creative industries tend to be geographically clustered) to create operational mechanisms and ensure that they are designed and developed bottom up rather than top down.
In the next section, we review the literature on innovation systems, highlighting some of their weaknesses in non-Western contexts, and putting forward the idea, explored in the analysis, that a creative ecosystems approach will enhance RIS and ensure they are fit for purpose for creative industries in transitional settings.

| Regional innovation systems
This paper is primarily concerned advancing the theory of RIS in transitional settings by adding an ecosystems approach. In order to analyse the RIS, we first present the National Innovation Systems (NIS), the literature that underpins some of the concepts of Innovation Systems literature. Lundvall (1992, p. 12) defines a National System of Innovation (NIS) in a "broad" sense, encompassing many aspects of the economic structure and institutional setup affecting learning, searching and exploring: the production system, the marketing system, and the finance system. Niosi et al. (1993) add legal and commercial aspects to their understanding of a NIS. Williams and Edge (1996) argue that a systemic view of innovation is necessary as innovation (both in terms of the kind of innovation and the rate at which it occurs) is inevitably influenced by economic, institutional and cultural factors, and that these are best understood at the national level. Lundvall and Borras (2005) identify three types of policy that national governments could pursue to help economic growth: a science policy, focused on the production of new scientific knowledge; a technology policy, focused on increasing and exploiting for profit new technical products and knowledge; and finally, an innovation policy, focused on attempts to increase the rate of innovation in the economy as a whole. These three forms represent the foundations of a NIS constituted by legal frameworks, firms (especially large multinationals), policy organizations, and education systems (Groenewegen & Steen, 2006). Groenewegen and Steen (2006) propose a layered model of NIS which is heavily reliant on national institutions and formal institutional arrangements. However, viewing innovation from a national perspective is problematic as economic activity is not evenly distributed and the nature and size of innovative industries can differ significantly from place to place within a State (Ashiem & Gertler, 2005;Malerba, 2004;Nelson & Rosenberg, 1993).
Hence, the debate on innovation systems moved from national systems to regional or territorial models in an attempt to overcome some of the inherent weaknesses of national models (Moulaert & Mehmood, 2010;Pino & Ortega, 2018). Jessop (2004) introduces an institutional turn to the RIS literature in addition to the more functional national systems approach. RIS models view innovation systems as being dependent on territorially or sectorially informed networks of invested actors and communities in the innovation process (Fiore et al., 2011;Keating et al., 2009;Phillips et al., 2015). Cooke (2001) outlines the conditions needed for higher innovation potential in regional and territorial systems, which are the establishment of trust and co-operative working methods among economic actors. A collaborative and networked approach is essential in a RIS if it is inclusive and if innovation is driven bottom up rather than top down. There is an inherent assumption here, however, that those institutions are in place to facilitate this.
In a bottom-up approach, networking is essential for capacity building since learning processes are social and highly interactive; consequently, innovators, firms, and educators need to co-operate to co-develop the skills necessary for the local economy to thrive (Cooke, 2001;Pekkarinen & Harmaakorpi, 2006). Policy makers can support this process and achieve higher regional innovation potential through a series of policies aimed at increasing local influence on infrastructure, institutions, and spending taxation locally (Pino & Ortega, 2018). This process would support the creation of conditions for inclusive organizational settings, harmonious labor relations, and university-industry strategies to produce the skills needed to promote and stimulate innovation (Cooke, 2001). All of this leads RIS proponents to suggest the need for an active State or local governance structure with the power and the influence to enable these conditions to flourish.
Moving on to RIS analysis in less economically successful contexts, Trippl et al. (2013) discuss ways of exploring regions and places with less obviously developed innovation and governance systems. Their work has important implications in our data analysis; however, once again, the cases they use are drawn almost exclusively from Europe. They propose viewing arrangements using the lens of organizational and institutional thickness and thinness (linking back to Amin's (1999) earlier work). Their case studies demonstrate that when using this approach, three broad types of underdeveloped innovation systems emerge in the European context: (a) institutionally thick but organizationally thin regions; (b) organizationally thick but institutionally thin regions; and (c) organizationally and institutionally thin regions. The first of these are places where co-operation and institutional support are strong but there is a lack of research-led organizations to take advantage of this fact. In organizationally thick but institutionally thin regions a number of strong organizations exist in an institutional vacuum (both culturally and governmentally). Finally, regions with both organizational and institutional thinness are those where both the organization and the institutional setup are lacking in innovation potential. These typologies are certainly worth bearing in mind when considering our cases, especially in relation to what might constitute "thickness" in a transitional context; and it involves an assumption that these governance or institutional elements are crucial to explaining the relative success or failure of innovation. As Brown et al. (2016) show, successful innovation is possible in an ecosystem setting without strong institutional support, and indeed if institutions are imposed on that ecosystem they can ultimately damage it.
Since RIS do have a more reflective logic (Geddes, 2005) than national systems, they can overcome some of the rigidity of NIS to respond to local needs. However, since they focus on institutions, they have been implemented top down in a prescriptive fashion. This results in systems where the Vietnamese Government sets the economic agenda without first understanding their local contexts. An example of this issue can be seen in the United Kingdom where the government asked regions to produce local industrial strategies that had to dovetail with the nationally produced one rather than develop a national strategy informed by local contexts. Within a transitional setting this approach means that emerging sectors and creative ecosystems in local economies get left out by the national and regional strategies due to the knowledge and infrastructure vacuums that define them. Policy makers need a way of addressing those vacuums in order to better design RIS in transitional settings (authors, forthcoming).
Both national and regional systems rely on the diffusion of policy to organizations, institutions and places. As Cairney (2012) shows, policy diffusion is reliant on the ability of policy makers and institutions within a governance structure to "communicate" with their networks. However, in our transitional research setting, the organizations and institutions are still developing, or are still to be developed, making the diffusion more difficult, as there are very little infrastructures in place. If a structure is imposed without understanding the local system--as in Brown et al.'s (2016) case--then it is hard to build trust in a system that Cooke (2001) and others argue as being so crucial to the success of innovation policy.
The focus on institutions in the innovation systems literature has led us to identify two challenges for both policy makers and actors in a transitional setting, which may be mitigated by a creative ecosystem approach. First, there is a tendency (as we have shown) to overstate the importance of institutions in stimulating innovation. Second, this in turn has led to the failure of innovation policies being attributed to institutional thinness or a lack of institutions. The problem with the institutional approach (Jessop, 2004) here is the assumption that innovation may be absent without institutions. As we will show in our institutionally absent setting, innovation flourishes and we need a new approach to analyze the creative economy. Therefore, in this paper, we take a creative ecosystems approach to analyze and understand innovation and creativity as it occurs outside the usual policy and institutional frameworks one might expect to see in a developed economy setting. The institutional turn (Jessop, 2004) taken in developed contexts will not work where institutions are either absent or weak. Instead this approach will allow us to understand what is happening on the ground and to outline what institutions, policies and infrastructures are needed to support local innovation. We propose to construct RIS for transitional economy settings designed from a bottom-up perspective rather than top down.

| BACKGROUND TO THE CASE
Over the last 20 years, Vietnam has transformed its economy from State-planned to market-based (Truong & Quang, 2007), and this process has seen the creation of a number of privately owned enterprises and a reduction in reliance on the State to provide employment (Truong & Quang, 2007). In particular, following the introduction of the Doi Moi Policy in 1986, Vietnam has witnessed a rapid transformation of its economy which has allowed it to move from being one of the poorest countries in the world to being a middle-income country, lifting over 40 million people out of poverty. This growth can be attributed to the opening of the economy, building infrastructure, introducing private enterprise, and attracting foreign direct investment that has contributed to the creation of new markets and the modernization of industry (Cameron et al., 2018). The strategy of the Government of the Socialist Republic of Vietnam is to move toward a democratic, open and innovative economy, as delineated in the strategy "Towards Prosperity, Equity, Creativity and Democracy" (World Bank Group, & Ministry of Planning and Investment of Vietnam, 2016). This strategy seems successful, as the Global Innovation Index reports that Vietnam has demonstrated persistent performance for innovation in recent years.
Whilst the progress made to date has largely been in traditional areas of the economy and industry, Vietnam does not have the infrastructure in place to promote steady growth in the creative economy as designed in the government resolution. For example, there is no institutional contemporary art education, no critical thinking courses at school, very few pioneering curatorial programs, and rare innovation courses for SMEs to develop critical and creative thinking. Nevertheless, there are creative spaces and creative organizations working in a creative ecosystem, which we were tasked with mapping as part of a British Council funded research project. As we carried out this mapping exercise, research questions emerged concerning the nature of innovation networks and the values being created. Thus, in this paper we investigate what kind of values creative industries in Vietnam are producing, how innovation occurs within the ecosystem that has developed, and what support (in terms of policies, infrastructures, and institutions) do they need to survive and thrive.

| MATERIALS AND METHODS
The data presented in this paper were drawn from a qualitative study carried out over an 18-month period in 2017 and 2018. Fifty-five interviews took place with a total of 92 individuals, which included private sector creative organizations, involving creatives, managers, funders and hub directors. In addition, 12 group interviews were conducted with Ministries in Hanoi (Ministry of Culture, Sports, and Tourism and Ministry of Technology and Information), in which 37 individuals participated. Our analysis also covered social media, analyzing 112 Facebook pages of creative organizations (Facebook is the most popular social media in Vietnam). In the field, the first author took notes and digitalized them at the end of each day. The notes were uploaded to a virtual diary that was made available to the research team to review and discuss the progress of the fieldwork, and to question and probe for additional information.
The fieldwork was carried out as we became engaged in a research project consisting of mapping the Vietnamese creative economy. The British Council Vietnam acted as a gatekeeper and organized the initial interviews. In addition, we developed a network of Vietnamese contacts who provided further assistance in locating policy makers willing to be interviewed. Vietnam is an emergent transitional economy, which has undergone a rapid change from a completely State Planned model to a quasi-free market system, albeit still retaining a one-party structure. The Vietnamese creative industries are collectively one of the last sectors of the economy to be freed up from overt State control. Until this research was conducted in 2017, little had been done in terms of measuring and assessing the scale of the industry and its impact on the wider Vietnamese economy. Our mapping research found that at least 3.8 million people are working in the creative economies in Hanoi, the capital of Vietnam, and Ho Chi Minh City, the biggest city, alone, thus they are already a significant part of the country's burgeoning economy. As Vietnam continues to move up the Global Innovation Index, it will play an ever-increasing role.
During the initial phase of the research, interviews and observations concentrated on the creative sector itself. After the first round of interviews and preliminary analysis, we interviewed Ministerial teams from Vietnamese governmental bodies. Cross-case analysis of this data subsequently led to the development of the creative ecosystem presented in the findings and discussion of this paper. Through this, we were able to assess what aspects of the creative ecosystem and creative organizations already existed, what was not in place according to creative people's experience, and what could be further supported to enhance creative people's innovation rate and impact on society.
We selected creative organizations in Hanoi and Ho Chi Minh City, and across the creative sectors using Florida's (2014) classification triangulated with what Vietnamese creative organizations defined as creative. In this way, we used an approach grounded in the literature and adapted to the local context.
The interview protocol for the creative organizations was divided into five parts: (a) New product/service development; (b) Care (which was aimed at capturing social impact and motivation); (c) Diffusion; (d) Values; and (e) Context. We also collected from archives, 124 documents including: organizations' reports, ministerial papers, financial data published in economic papers, Vietnamese and industry magazines and reports, World Bank reports; and also gathered information through direct observations and social media. Gasparin conducted 4 months" of direct observations aimed at understanding the cultural, social and economic contexts, individual and organizational struggles, tensions and aspirations, and the modus operandi of organizations. This research was conducted 5 days a week from July to October 2017, and the data collected from the interviews and observations along with the field diary were then subjected to a coding data analysis process.
The findings were presented at the National Assembly in April 2019 (the equivalent of Parliament) and feedback collected in August-September 2019. Observations were recorded in the form of field notes that were written up later. Gasparin conducted site visits, informal conversations, made observations on new projects, service launches, and social gatherings, including six event launches, which focused on: exhibition opening, a British Council social enterprise event, the launch of a crowdfunding campaign, and talks with an artist.
In the following sections, we draw upon our analyzes of the Creative Industries in Vietnam to produce three models which allow us to apply and design RIS in a transitional context. As we note in the discussion below, when researching incidence of innovation and creativity in these sectors, we found an absence of the institutional structures and innovation systems that one would normally expect to underpin development in a sector or economy. Nevertheless, creativity and innovation are present in Vietnam. We observed that different ecosystems had emerged in each of the creative communities we studied. Due to the absence of institutions but presence of ecosystems, Western based RIS theory is not an effective lens for examining the innovation potential of these communities, nor for capturing the values created in these communities. We are not claiming that RIS theory cannot be applied in transitional contexts when building support structures. However, we argue that, the model needs to take the local conditions and the ecosystems into account before adding institutions to the RIS and indeed NIS, especially when innovation institutions are absent.
In this paper, we present the creative ecosystem (Figure 1) that we found underpinning the sector. From this, we identify the different kinds of values (social, cultural, and economic) being created by creative organizations, which further allows us to highlight the issues and challenges they face in creating these values. This leads us to propose a flipped model of policy development (Figure 2) which could be used as a mechanism to channel those challenges into policy development, and from here propose a RIS suitable for the sample we were dealing with (Figure 3). We use these models to propose a set of policy recommendations that emerged from the data. We do not have sufficient space here to go into detail on each policy area, and we have explored this in depth in .

| REFLECTIONS FROM THE FIELD
In this section, we use our data to build a model of the creative sector ecosystem in Vietnam. We then use this model to analyze and identify the innovation and value being produced by creative industries F I G U R E 1 Creative ecosystems in Vietnam in Vietnam, which in turn leads to policy recommendations for a stronger, more locally informed RIS to be constructed.
In Vietnam, policy makers are not aware of the RIS models, nor do they have a grasp of the activities, processes, and values (economic, social, and cultural values) created by creative industries as they have not collected data/evidence on their impact. Interestingly, we found that, despite the lack of policies, creative ecosystems have nevertheless developed. Creative industries have found ways around institutional absence to co-create, co-innovate and co-develop their industries and, crucially, their communities.
This contradicts the assumption of the "institutional turn" (Jessop, 2004) in the RIS literature, which states that scholars and policy makers should use institutions as the lens through which to analyze the innovative capacity of a region. Whilst Trippl et al. (2013) discuss the implications of institutional thickness and thinness on innovation potential in their chosen cases, we witnessed institutional absence in both Hanoi and Ho Chi Minh City. The governance institutions that RIS models assume would normally be expected to support economic development, and would be expected to create trust between the public and private spheres, are not developed in the Vietnamese economy. The lack of trust is manifested in a palpable atmosphere in which efforts by government to get involved are met with suspicion by creative organizations, as they see government's role as a tax collector and a regulator rather than stimulator. We noticed a fierce resistance from organizations to be part of online surveys, as they were afraid (despite the ethical form and the letter being presented in the survey in Vietnamese and English) that we were collecting data on behalf of the government, and they might need to shut down. This problem was not manifested in the face-to-face interviews as we gained the trust of the respondent.
During our interviews with policy makers, we realized that institutional absence makes policy diffusion much harder as there are no allocated institutions supporting policy implementation. This F I G U R E 2 Translating ecosystems into a flipped model of policy diffusion | 11 GASPARIN ANd QUINN absence also bars the co-creation of policy to realize higher potential RIS which, according to the literature, are needed for growth.
The Western concepts of government's economic development role and of governance rather than government is absent in Vietnam. Government produces policy and strategy, and works with policy consultants and its own departments, but not with local creative communities. In order to survive, the creative communities prefer to work with each other rather than with the government in creating new products. Hence, innovation and creativity exist within a creative ecosystem, which has emerged in the spaces created by the switch from a form of ##state-planned economic system to a quasi-capitalist system.
Thus, in a transitional economy both national and regional innovation policy can be improved by acting as a continuum from the ground up (creative ecosystem, feeding into a regional system, feeding into a national system--see Figure 2 in Section 7 for a visual representation of this) and helping RIS to overcome challenges in a non-Western setting. Working in this way, we ensure that any institutions that are eventually incorporated into a RIS (see Figure 3 in Section 7) are only done so once the need for them to be there is clearly established by the creative ecosystem.
Whilst the literature largely focuses on identifying economic value creation (Adner & Kapoor, 2009;Clarysse et al., 2014), our coding process highlighted that the creative ecosystem creates three types of value: social, economic, and cultural. Realizing the existence of these three values and supporting ecosystem development will allow policy makers to translate the ecosystem that is already in place and the areas that need strengthening into polices that allow the sector to grow.
From the coding, we theorize two concepts that are used in the analysis: creative ecosystem and values. We define a creative ecosystem as a complex network of actors whose interactions produce innovative and creative processes and outcomes, which are shaped by economic, social, and cultural values. The social, economic, and cultural values that we found in the creative ecosystem are represented in Figure 1.
We use these concepts to unpack the creative ecosystem model and use it as a lens to analyze the creative industries in Vietnam. Focusing on this section of our literature framework allows us to better understand the needs of the industries and enables us to identify a range of differing values being created by the creative industries in Vietnam alongside areas that require further support. From this analytical focus on the creative ecosystem, we then propose a flipped model of policy diffusion, as well as policy recommendations to create a RIS that is appropriate for a transformational setting. F I G U R E 3 Regional innovation system for the creative industries in Vietnam 12 | GASPARIN ANd QUINN 6 | ANALYSIS: VALUE CREATION IN

CREATIVE ECOSYSTEMS
As described in the theoretical framework section, creative organizations create three types of values: social, economic and cultural. This demonstrates that the contributions creative industries make to society go beyond the economic sphere, and therefore policy makers would be mistaken in assuming that a (relatively) low economic return from a creative organization produces little value to its community and place. In the following sections we identify the values that are generated, as well as the struggles creative organizations experience.

| Social values
We refer to social values as the social benefits provided by creativity. According to the literature on creative industries, culture and creativity lead to urban regeneration (Evans & Shaw, 2004), social inclusion (Belfiore, 2002) and social benefits (Moreton, 2013). In Vietnam, we have observed that creative organizations are working closely with selected groups of society which are often marginalized or left behind, helping them to: improve their well-being, foster social and responsible behavior, develop soft skills, stimulate critical and reflexive thinking, address instances of human rights violations, stimulate creativity, create and cultivate positive emotions toward society, and provoke reflections and viewpoints on political issues.
Creative disciplines are not taught at school, which is problematic for the development of critical thinking in society (Goodwin & Sommervold, 2012). Therefore, creative organizations are teaching them in their hubs and courses, which is not surprising, as research by Nicholls and Murdock (2012) recognizes. Creative organizations focus on delivering training with the aim of creating more reflective innovators for tomorrow's society.
Almost all the creative organizations we interviewed are very sensitive toward issues concerning Vietnam's numerous ethnic minority groups and the divide between urban and rural areas. Therefore, they work to co-create a socially innovative, sustainable supply chain that involves ethnic minorities--for example, by buying raw materials to make new items. This process is emphasized especially among ethnic fashion designers. In Vietnam, there has been an emergence of slow designers that are going to use natural products from ethnic minorities and rural areas, or commercialize their products. Most of the socially innovative design organizations are working in co-creation settings, by developing their products closely with marginalized communities, involving them in the decision-making processes, providing ad hoc training to make them long-term independent from a single supplier, and paying a fair price for their work. One of the slow designers working with rural women is training them in basic accountancy and basic project management. During an interview, she commented on the importance of her work in holding communities together: The women of ethnic minorities work with me, we create a future for them to preserve the traditions. They are learning basic business skills, soft skills.
According to another slow designer, connecting the urban with the rural areas contributes to the preservation of the traditional family structure: The women can work at their place; in this way, they don't need to leave the villages and their families to go to the city to find jobs.

GASPARIN ANd QUINN
This is not an isolated case. In fact, the designers we interviewed are working with marginalized communities to create transferable skills and knowledge that will allow these communities to enjoy their traditional lifestyle, rather than moving to the cities and abandoning their families and their traditions. In this way, they not only preserve the traditions, but also contribute to the creation of an inclusive, open, democratic, and innovative society.
However, creative organizations are recognizing the values of the networks, but they are not yet benefitting from the positive externalities of being in hubs, nor do they have members who act as the main interface between members of their own network community and members of other communities, as shown in the study by Clement et al. (2018).

| Economic values
Creativity and culture can be used to lead economic development in cities (Florida, 2002;Mellander et al., 2013). We refer to the second set of values developed by creative organizations as economic values, which can be both external and internal (creating economic value) to the organization.
External value means creating economic value for society through taxes and net value, which can be calculated using cost-opportunity reasoning. This highlights that creative organizations produce value for the whole economy by supporting other businesses and creating a sustainable operational model that also results in economic return.
The creative organizations we interviewed have two main outcomes: artistic and commercial. The artistic outcomes have artistic value, but since there is no support from the government nor subsidies, in order to survive, they propose marketable products and services aimed at generating revenues. Examples of these include applying course fees, conducting consultancies, trading gadgets, hiring out their spaces for events and charging for tickets for exhibitions and concerts, generating income from cultural-based tourism activities that are offered as an extra; and selling beverages, food, and wines. The course fees are differentiated: when applied to companies, it is based on an estimated market-based price, when offered to people from disadvantaged backgrounds the price is dropped substantially and subsidized by the fees charged to companies.
These creative activities are socially innovative, as they seek to alleviate societal change alongside making a positive impact on financial performance (Mongelli & Rullani, 2017). Social innovation includes novel products and services that meet a social need . They are "diffused through organisations whose primary purposes are social" (Mulgan, 2006, p. 146), responding to changes in social relations (Bouchard, 2012) and seeking to contribute toward social and ecological resilience (Westley & Antadze, 2010). Furthermore, they generate a socio-economic impact for the whole of society, which is challenging to uncover within the traditional methods of data collection and economic value analysis, and it refers to negative and positive externalities.
In terms of reducing negative externalities, social impact is at times calculated using opportunity cost-reasoning (Schinckus, 2017): the social impact that creative organizations generate is based on the improvement of society by creating a reduction of the social costs that might be generated by the target population. An example of this was one of our cases where a creative hub developed art courses in order to encourage people to express their feelings and talk about mental health issues that are afflicting people. According to a group meeting that Gasparin attended, many participants felt that they overcame depression or other mental health issues by taking part in these groups, thus avoiding costly treatment in hospital or other facilities. The director of this creative hub commented: In the way that the economy and society is growing, it doesn't allow people to be themselves at all. It always, you know, appeals to them to be someone else… So we think we do need … that they really need a space so that they can be themselves even just for a short moment of time, so that they can be balanced, they can tell their thinking or emotion or stories, so that they can share, and because they share, they feel connected-with themselves.
This means that creative organizations are proposing services that alleviate social problems (such as mental health issues), which usually are a cost to the government; in this way, the financial pressure is released from the government, as they provide a service for society.
Thus, we can consider these activities as socially innovative because they relate to areas of welfare that are not offered by the government (education of underprivileged persons, mental health etc.) and are designed to directly influence and improve the living conditions of marginalized and left-behind communities. This contributes to a significant improvement in the Vietnamese population's well-being. This is an important contribution to the economic system of the region, which goes beyond economic activities being generators of taxes. Thus, in RIS, creative businesses need to be recognized also for their contribution to the economical reduction of negative externalities (i.e., reduction in costs related to what is not in the economic system).
Creative organizations also increase positive externalities. In the case of organizations working toward positive outcomes in society (e.g., engaging minorities working in the traditional craft industry), the economic value of such initiatives relates to a gradual marketization of socio-cultural assets, in order to foster the slow evolution of cultural assets into a market by extending the concept of marketing.
In terms of generating economic value for organizations, managers struggle to capitalize this. Embedded in a not short-run profit-oriented environment, creative managers are aware of the importance of long-term analysis, although currently they do not have the means necessary or the skills to create a sustainable long-term business plan.

| Cultural values
We define cultural values as a set of beliefs, traditions, heritage, and ethical principles that guide and inspire a community . These values are contributing to the creation of a sustainable society, and the establishment of a shared identity. Because of the history of Vietnam, this is a relatively new but critical role being taken on by creative industries. The inclusion of design, creativity, and sustainability to encourage the consumption of arts and the exploration and preservation of Vietnamese visual identity and cultural heritage, is producing a burgeoning confidence in Vietnamese products. To better facilitate this, there is a growing need for creating space for arts, the development of education and income for marginalized communities, increased capacity building, freeing up emotional expression, and more tolerance for provision of cultural content that is not available through official channels. Creative organizations aim to not only preserve cultural values, but also question and mobilize them, especially those in the marginalized communities that are at risk of disappearing. Cultural identity in Vietnam is not well developed, nor respect for the work of the artists in terms of intellectual property, which is already found in Evans' (2009) research, as one respondent from the design sector commented: I told them [the students]: How to tell our foreigner friends 'That's from Vietnam' and represents something from Vietnam? That's important, that makes our work significant in the environment, because when we arrive, we are just copiers, and not really creative, and that's theft. It's not hard to be a copier. Whereas it takes time to be a real designer.
Vietnam has a long history of colonization and war, which have all but erased local cultural identity that nowadays is challenging to remember. In response, some creatives are working to overcome this by incorporating emotion and healing into their work: So in 2014, I cofounded [Name] with two of my friends, and we-our idea of Dao Tao is like a creative learning hub, right? We focus on creativity, connection and healing. That's the three basic and foundation of our values. Because we, what we saw is that those values is missing from the education and the education system, and there's so many cultural activities but they are not really nurture those values. There's-we need more creativity, we need more connection between people and between one with oneself, you know? You connect with yourself. And we need more healing because there's a lot of pain. We grow very fast and we focus on growing physically, and we growing on economy and on things on outside. The emotion and feeling is often being neglected, or you don't care about it at all.
The work of Moulaert and Mehmood (2010) on territorial innovation models shows how drawing the social and communal aspects of networks together can create better opportunities for innovation to flourish. Therefore, the creative sector is educating young generations toward team working, creativity, innovation, and culture, which have historically been absent from curriculums and public debates. They are also putting in place educational programs for the public as, according to one of the curators: The public doesn't care-it's not that they don't care… they don't understand. They don't take the time to stop and reflect. We need to invest time and resources to educate them.
The State's educational programs only use teaching techniques that train pupils to repeat/copy them (as prescribed in Confucianism). Instead, courses held by creative organizations are creative as they see art, creativity and innovation as societal glue to help overcome diffidence and build a collective consciousness, stimulate critical thinking, and imagination, and develop a physical and mental space for being irresponsible and looking for novel solutions. This behavior ultimately could lead to innovation and a general societal improvement.

| DESIGNING RIS USING CREATIVE ECOSYSTEMS
Ecosystem is not a new term in either innovation or regional studies, as the work of Brown et al. (2016) demonstrates, but it is perhaps an undertheorized term. Oh et al. (2016) call for a better theorization of the term innovation ecosystem. In this paper, we respond to that call by producing a clear outline and definition of creative ecosystems and the values that they produce. Better understanding and theorizing of creative ecosystems is essential for policy makers because they help plug the knowledge gaps that may exist. Although RIS remains useful in proposing formal solutions, they are weak at explaining policy absence, and remain too focused on top-down policy intervention and institutions, rather than analyzing what is happening on the ground. A creative ecosystem theory allows us to understand innovation in an institutionally absent setting. Creative ecosystem theory can be used to identify the value being created in transitional settings and the institutions that might be required to further support development in those transitional settings. Policy makers need to comprehend a creative ecosystem before they can put in place a RIS in and, as Brown et al. (2016) point out, where policy makers failed to recognize that the ecosystem and the innovation system failed. We would like to reiterate that we do not dismiss innovation system theories, but instead we propose that they can be viewed as a continuum of creative ecosystems, and the process of diffusion needs to be flipped. Innovation systems would be much stronger if they reversed the flow and started with an appreciation of local ecosystems before producing regional and then national structures to support those creative ecosystems.
The literature suggests that innovation systems are the determinants of the innovation process, which are "all important economic, social, political, organizational, institutional and other factors that influence the development, diffusion, and use of innovations" (Edquist, 2009, p. 182). However, they are made to work at the national level, although it is accepted that they have different geographical, sectorial, and activities boundaries, and they are defined as networks of private and public institutions whose interaction initiates, imports, and diffuses new technologies determined by political, cultural and economic policies (Freeman, 1987). Therefore, we suggest analyzing the data through local sensitivities. Because companies and individuals do not innovate in isolation, but within a network of actors, it is important to analyze the components of the network, or creative ecosystem.
As discussed in the literature review, too often policy diffusion models start from the point of view that a policy or an institutional framework can be imposed from above by a national or regional authority and then the initiatives and interventions will direct development, creativity, and innovation within the target area. Our data, together with Brown et al.'s (2016) work in Scotland, demonstrates that this is not the best approach to ensure that innovation at the ground level receives the appropriate level of support and the necessary means to make it flourish. To achieve these important objectives, the diffusion process needs to be flipped and to start by translating the values that local creative ecosystems create and the support requirements they put into a system to help underpin their existing strengths. Any institutional framework that is introduced should only be applied once a clear understanding of the local ecosystem has been established. As outlined in Figure 2, the flow of information should be reversed to begin with lessons drawn (Rose, 1991) from the creative ecosystem, which should then be fed into a RIS designed to support, not instigate, those ecosystems. These regional systems should then be collated into a NIS that is directly informed by the strengths and weaknesses of its economy.
Using creative ecosystems to feed into the flipped model of policy diffusion, we have drawn on the data presented in this paper to design a RIS based on the strengths and needs of the existing sector. The RIS we present is a deliberately broad template to allow for the specificities of each of the local creative communities we explored to come to the fore within their own RIS before they are fed into a National model. Our RIS is presented in Figure 3 and is split into three areas: institutions, infrastructure and specific policies needed to support and (further) grow the creative industries in Vietnam. At the outset the RIS must engage with and develop a number of organizations and emergent institutions including (but not limited to) Government, Local authorities, Creative organizations, Schools, Universities, and Banks. These should then be brought together to work with the organizations within the creative ecosystem to co-design the support structures needed for the ecosystem to flourish. Within this, policy makers should be given the scope to establish a "creative innovation system" framework, within which strategic priorities can be addressed in a coherent and effective manner and across different departments of the Ministry of Culture, Sport, and Tourism, as well as governmental bodies for economic planning. To facilitate this and to strengthen the infrastructure of the RIS, local policy makers should be updated and trained on the creative economy, as there are gaps in the interpretation of the law and in the application of censorship between central government and local offices. In addition, a review of investment in infrastructure to develop the creative economy, urban planning and heritage preservation is required to ensure it is fit for purpose. Further measures that need to be implemented to increase the infrastructure underpinning the RIS include improving the quality of data collected on creative industries by the State and Government Statistics Office. Current data sources significantly underestimate the size and scale of the private sector's involvement in cultural and creative industries. Finally, in terms of developing appropriate infrastructure, the Ministry of Education and Training needs to encourage Education and Higher Education Institutions to engage with the creative and cultural sectors, in order to reduce the disconnection between what creative businesses need from graduates and what universities are teaching them.
Specific policy requirements in the RIS initially include a policy to broaden the National Education Curriculum to incorporate humanistic subjects as well as art, design, and innovation management to help creatives as they enter the workplace and start their own businesses. Government should encourage the use and study of digital and artistic disciplines at school to ensure the school curriculum brings together art, design, technology, and computer science and that young people are able to enjoy greater opportunities to work creatively and with technologies. Turning to fiscal policies, tax relief schemes should be more accessible to creative businesses, which could open up opportunities for smaller digital firms and establish cross-disciplinary research knowledge exchange initiatives and further investment. Finally, copyright rules should be strengthened and implemented, balanced with educational purposes and thus leveraging a more creative world by promoting authors' rights. If authors and creators are to continue championing culture and promoting cultural diversity, they must be compensated fairly for the use of their work. The current failure to properly reward creators is limiting creative and cultural industry revenues as well as holding back growth and the ability to create jobs.

| CONCLUSIONS
This paper uses a creative ecosystems approach to take innovation systems out of their comfort zones and explore incidences of innovation and creativity in an institutionally absent setting. When examining the cultural and creative industries in the Global South (Vietnam) we found that, contrary to the expectations of some systemic views of innovation, innovation was thriving in the absence of formal governance and that innovation ecosystems had sprung up, bringing together creatives to support each other's initiatives, creating social, economic and cultural values.
These creative ecosystems produce multiple kinds of values for the economy and broader society, and therefore we posit that the usual model of policy diffusion within RIS needs to be flipped to ensure that the values and needs of local creative ecosystems can be translated into policy needs, can inform, and feed directly into the development of institutions and policies to underpin any NIS or RIS. Although our research is focused on the creative industries, the flipped model we present is transferable to other sectors within a transitional setting. Future research should be aimed at testing the use of the model in other sectors of a transitional economy.