The nexus between the cultural and creative industries and the Sustainable Development Goals: a network perspective

ABSTRACT Scholars and policymakers have widely claimed that the cultural and creative industries (CCIs) provide positive knowledge externalities that can help address sustainable development challenges, yet questions remain about the pathways through which this occurs. In this study, we hypothesise that several features of knowledge networks in the CCIs relate to a location’s sustainable development outcomes. We use data of ownership networks between 22,455 cultural heritage-related firms across 292 cities in China to empirically test our hypotheses. We find that the density of the CCI network has a positive relation with a city’s performance in terms of several Sustainable Development Goal measures. Moreover, the scale of local CCIs has an inverted ‘U’-shaped relationship with a city’s sustainability performance. Finally, a city’s degree of trans-local ties has an inverted ‘U’-shaped relation with a city’s sustainability performance.


INTRODUCTION
At the 74th United Nations (UN) General Assembly, 2021 was declared the International Year of Creative Economy for Sustainable Development, an initiative that calls attention to the role of the cultural and creative industries (CCIs) in achieving the UN's Sustainable Development Goals (SDGs) (UN, 2019).This announcement emphasised the critical link that the UN perceives to exist between the CCIs and sustainability, notably through knowledge creation and dissemination.A quotation by Thangavel Palanivel, Deputy Director of the Human Development Report Office of the UN Development Programme, captured this sentiment: [c]ultural and creative industries, which include arts and crafts, advertising, design, entertainment, architecture, books, media and software, have become a vital force in accelerating human development.They empower people to take ownership of their own development and stimulate the innovation that can drive inclusive sustainable growth.(Palanivel, 2019, para. 1) The UN based its support for the CCIs on extant scholarship that points towards the industry's generation of positive knowledge externalities that help address sustainability challenges.Several authors argue that the CCIs are among the most creative economic activities that foster innovative solutions for a more sustainable economy (Burford et al., 2013;Hawkes, 2001; United Nations Educational, Scientific and Cultural Organization (UNESCO), 2013).Other scholars indicate that the CCIs facilitate other dimensions of sustainability by providing employment opportunities to industries such as tourism, retail and transportation, especially for marginalised groups and regions (Li et al., 2021;Pagan et al., 2020).A third set of researchers suggest that the CCIs shape intergenerational relationships, ways of living together and human-nature interaction (Auclair & Fairclough, 2015;UNESCO, 2001), which build cohesive communities through the generation of localised knowledge (Cattaneo et al., 2020).A common policy recommendation that comes out of these studies is that governments should support the CCIs in their quest for a sustainable future.
A question that in this discussion has flown under the radar, and which should be of interest to policymakers, is under which circumstances the CCIs generate the optimal knowledge externalities to support sustainability performance.Is the relation between the CCIs and sustainability merely a function of the scale of the cultural and creative sectors?Or do structural features of the CCIs play a major role in the nexus between the CCIs and sustainability performance?Gaining deeper insights into these questions can help governments better target their actions to promote the CCIs for sustainability.
In this paper, we take steps in this direction by examining how the structure of knowledge networks in the CCIs matters in the nexus between the CCIs and sustainability.For this purpose, we build on an influential literature that has shown that inter-firm networks act as a fundamental channel for the transmission of tacit knowledge both within and across knowledge-intensive industries, generating localised knowledge spillovers that spur economic growth (Giuliani, 2013;Owen-Smith & Powell, 2004).We extend this literature in two ways.First, we analyse the role of knowledge networks in the CCIs, and specifically the role of local versus trans-local knowledge linkages, which in the CCI literature has received limited attention (Asheim et al., 2011;Bathelt et al., 2004;Martin & Moodysson, 2011).Second, we investigate how the structure of knowledge networks relates to various sustainability outcomes, which is different from most knowledge network studies that focus on innovation outcomes (Phelps et al., 2012).
We specifically aim to answer three questions: • How is the scale of a city's CCIs associated with its progress towards achieving SDGs 8-12?
• What are the implications of the CCI network structure on the city's progress towards these SDGs?
• Do trans-local partnerships complement the local network in achieving these SDGs?
We empirically study these questions by focusing on knowledge networks in China's cultural heritage industry and evaluating their relationship with performance measures related to SDGs 8-12, which we argue to be particularly relevant to the CCIs.Cultural heritage refers to monuments, groups of buildings, and archaeological sites that bear universal value from historical, artistic, scientific, aesthetic, ethnological or anthropological perspectives (UNESCO, 1972).It fosters local creativity and innovation in urban communities that are essential resources for every type of activity in the CCIs (Cerisola, 2019).It also attracts creative talents who then form knowledge communities and hubs where the exchange of knowledge and ideas becomes possible (Pratt, 2008).To maximise the potential of cultural heritage for development, public and private sectors need to collectively raise awareness of the contribution of cultural heritage, to invest in the preservation of heritage sites, and to collaborate in networks to manage and operate cultural resources (Nocca, 2017;UN, 2017).
We conduct our analysis using a network dataset involving 22,455 organisations related to 1355 cultural heritage sites in 292 Chinese cities.We find that cities with a denser CCI network make more progress towards achieving SDGs 8-12.We also show that the scale of local CCIs contributes to a city's SDG score, but that the relation is an inverted 'U'-shape.Finally, we document an inverted 'U'-shaped relationship between a city's degree of translocal linkages and its SDG performance.
Our study makes several contributions.First, it explicitly incorporates the role of network characteristics of the CCIs in the analysis of the nexus between the CCIs and SDGs.As emphasised by SDG 17, sustainable development requires collective action by multiple parties and cannot be realised by individual entities working in silos without any collaboration (Nilsson et al., 2016;UN, 2022).This study illustrates how local and trans-local knowledge dynamics collectively contribute to achieve the different goals (Petti et al., 2020;Salvia et al., 2019).Second, by focusing on knowledge networks in the CCIs across Chinese cities, we contribute to the discussion of the localisation of SDGs, especially for cities in emerging economies (United Nations Development Programme (UNDP) & CASS, 2020; Xu et al., 2020).Cities with different resources and capabilities often face unique sustainability challenges and develop diverse strategies and solutions, and this study shows how the heterogenous structure of cities' CCI networks contribute to local SDGs.

The role of CCI scale
The CCIs are 'industries whose principal purpose is the production or reproduction, promotion, distribution, or commercialisation of goods, services, and activities of a cultural, artistic, or heritage value' (UNESCO & World Bank, 2021, p. 8).Extant scholarship identifies three channels through which CCIs can contribute to local sustainable development: a direct effect, an indirect facilitating role, and an overarching transformative principle (Dessein et al., 2015;Soini & Dessein, 2016).
First, the CCIs can directly contribute to a sustainable society through its close relationship with innovation and creativity (Burford et al., 2013;Nocca, 2017;UNESCO, 2013).A growing number of academics thus maintain that culture should be viewed as the fourth pillar of sustainability besides economic, social, environmental dimensions (Hawkes, 2001;Soini & Birkeland, 2014).The main reason why the CCIs have this direct effect is their knowledge-intensive nature that generates localised knowledge externalities which help produce innovative solutions for several grand challenges such as societal resilience, digitalisation and equality (Štreimikienė & Kačerauskas, 2020;UNESCO & World Bank, 2021).The CCIs create dynamic knowledge processes which depend on knowledge flows (Asheim et al., 2011;Bathelt & Cohendet, 2014).As economic activities become more complex and collaborative, networks extend beyond individual enterprises, and the boundaries between different sectors become less clear (Turkina et al., 2021).For instance, in the CCIs, the upstream activities such as heritage management, performing arts and visual arts become intertwined with downstream areas including design and media production through digital technologies (Handke & Towse, 2013).When knowledge from different domains is exchanged more frequently, this sparks artistic, scientific and economic creativity and leads to innovation (Cohendet & Simon, 2017 Second, the CCIs can act as an indirect facilitator for other pillars of sustainability by providing economic viability of development and by promoting social inclusion (Luckman, 2015).The CCIs employ a large number of young talents, generate 3% of global gross domestic product (GDP) with a fast growth rate, and contribute to the formation of creative cities (Cerisola, 2019;UN, 2019).Moreover, the CCIs also create employment opportunities for marginalised communities and groups in less creative sectors such as tourism, retail and transportation (Li et al., 2021;Pagan et al., 2020).It is estimated that one creative job in the CCIs generates 1.7 non-creative jobs (UNESCO & World Bank, 2021).As Hosagrahar (2017, para. 1) indicates, 'Placing culture at the heart of development policies is the only way to ensure a human-centred, inclusive and equitable development.'Third, culture is the collection of intergenerational relationships, lifestyles, ways of living together, value systems, traditions and beliefs, human-nature interaction, and many other critical components of our social systems (Auclair & Fairclough, 2015;UNESCO, 2001).It influences the overarching worldview of how people live with each other and interact with the materialised world, and thus can be viewed as the specific presentation of a sustainable development trajectory.The analysis of the CCIs helps to understand the nature, the principles, and the processes of how to use cultural capital, preserve heritage, and balance production and consumption, since development itself is 'a cultural process' (Soini & Dessein, 2016).
The lifestyle-shaping role of the CCIs is reflected in numerous policies that governments have developed to encourage the CCIs to use cultural resources and to transform cultural heritage sites into creative hubs, open spaces for creative activities, knowledge intermediaries and business support networks (Pratt, 2009).Thus they indirectly contribute to the formation of responsible and sustainable production and consumption behaviours (Fazlagic & Skikiewicz, 2019).Therefore, the development of the CCIs represents a commitment to responsibly utilising 'a stock of cultural capital that has been inherited from previous generations and can be handed onto future generations' (Soini & Birkeland, 2014, p. 216).
Nonetheless, there are limits in the extent to which CCIs can contribute to local sustainable development which has been underappreciated in the literature.The overconcentration of the CCIs in a location may generate challenges that dampen the link between the CCIs and sustainable development (Kacerauskas et al., 2021).First, as the scale of local CCIs increases, cities may enter a 'success trap' where an excessive scale of similar activities eliminates creativity and hampers innovation (Junni et al., 2013;Sobel et al., 2007).Second, a larger scale of cultural activities is sometimes associated with social problems such as gentrification, damage of social cohesion, dissolution of local communities, especially when developers witness the success of culture-and creativity-driven development trajectories and step in to carry out commercial regeneration plans (Betlej & Kacerauskas, 2021;Pratt, 2008;Wang, 2009).Considering both the opportunities and constraints related to the CCIs, we propose our first hypothesis: Hypothesis 1: The scale of local CCIs has an inverted 'U'-shaped relationship with a city's performance in sustainable development.

The role of the CCI network
Although the scale of the CCIs is relevant to the local sustainable development, relational structures of actors in these industries often vary (Huggins & Thompson, 2014;Owen-Smith & Powell, 2004).Prior research suggests that social relationships among collocated firms contribute to localised interactive learning, especially in knowledge-based sectors such as the CCIs (Boschma, 2005;Malmberg & Maskell, 2006;Mattes, 2012).Interorganisational networks formed by CCI actors are the collective of knowledge flow channels embedded in these social relationships (Asheim et al., 2011;Wittel, 2001).
In CCI networks, nodes can be treated as the organisations that engage in cultural and creative activities.The equity ownership and investment relationship can represent the ties between these organisations.Equity ownership relationship is one of the strongest and most stable interorganisational network ties and it signals 'commitment and long-term coordination' (Mani & Moody, 2014, p. 1637;Nohria & Garcia-Pont, 1991).Therefore, they are more likely to facilitate interorganisational symbolic knowledge sharing that is necessary for innovation and creative activities in the CCIs.In this study, we refer to this network as 'CCI ownership network'.
Extant scholarship argues that networks play an important role in many CCI subsectors.Uzzi and Spiro (2005) show that networks formulated by musical artists in Broadway production groups foster creativity and generate higher economic returns.Martin and Moodysson (2011) observe that the media industry in Sweden is characterised by localised interorganisational networks that promote the exchange of context-specific symbolic knowledge.Lee (2015) notes that networks between entrepreneurs promote trust and thus contribute to interpersonal knowledge transfer in advertising, design, and ecommerce industries in South Korea.To summarise, a dense interorganisational network between CCI firms is generally considered to provide timely knowledge sharing and effective interorganisational learning (Ahuja, 2000;Schilling & Phelps, 2007).
In line with these studies, we analyse one of the most important structural features of a network: density (Provan et al., 2007).We argue that dense CCI ownership networks can contribute to sustainability, by generating positive knowledge externalities that can support sustainable development.This includes directly fostering knowledge sharing and innovation, indirectly empowering communities and offering more income sources, and facilitating sustainable production and consumption of firms and citizens.
Dense CCI ownership networks mean that there exist multiple pathways of communication, of knowledge sharing, and of resource access between business actors who use the local cultural resources.The denser a city's CCI network, the more creative inspiration actors will be able to gain (UNCTAD & UNDP, 2008).By forming more cohesive interorganisational collaboration relationships, these actors become familiar with each other, learn from each other more effectively, and develop absorptive capacity that is critical for innovation (Gilsing et al., 2008;Gulati et al., 2000).
Several studies have provided initial indications of such pathways.Cattaneo et al. (2020) observe that transformation projects turning traditional shared-living heritage sites into sustainable community spaces rely on cohesive collaboration networks between local actors, just as the cases of Tulou in China and Cascina in Italy have shown.Li et al. (2021) find that investments towards several related economic activities, such as arts and crafts, agriculture, and recreation and tourism, help small agricultural villages in China to develop creative industries.
Denser CCI ownership networks also facilitate the preservation of traditional arts and crafts which are rooted in cultural landscapes.Investing in cultural heritage indicates a circular economy practice that is aligned with sustainable production and consumption principles since most cultural-heritage-related investment projects embody a key component of reutilisation of local heritage (Foster, 2020).
Despite these positive effects, some scholars warn that overly dense interorganisational networks can become obstacles of knowledge generation (Bathelt et al., 2004;Broekel & Boschma, 2012;Uzzi, 1997).Innovation in networks require recombination of diverse knowledge from different actors (Fleming et al., 2007).Since knowledge circulating in excessively dense networks becomes similar and redundant, overly high density 'inhibits the existence and the utilisation of diversity, and hence of novelty value', leading to lock-in or the paradox of embeddedness (Gilsing et al., 2008(Gilsing et al., , p. 1721;;Uzzi, 1997).Overly dense CCI ownership networks are unlikely to escape these paradoxes since preserving and utilising cultural resources also require innovative solutions (Foster, 2020).
Network density is often associated with network closure, that is, trust and support are higher among network members, but norms of obligation and commitment also pose constraints on network members (Coleman, 1988;Pillai et al., 2017).Denser CCI ownership networks thus may signal overinvestment in a limited series of cultural resources, specialising in exploiting economic values from certain heritage, and reinforcing established development trajectory and losing the capability of switching to new growth tracks (Boschma, 2015;Plaza, 2008).These tendencies may suggest that an overly dense CCI ownership network constrains sustainable development.
Moreover, dense networks tend to exclude external members, which is at odds with one of the important contributions of CCI-related development: social inclusion and empowering marginalised groups.As culture-related investment ties become extensive in cities, they represent interests of shareholders, who are in most cases business actors from development sectors, rather than local communities, who are forced to lose identity, to change ways of living, and to face the consequences of social exclusion and gentrification (Qian & Li, 2017;Wang, 2009).We thus propose the following hypothesis: Hypothesis 2: The density of the CCI ownership network of a city has an inverted 'U'-shaped relationship with its sustainable development performance.

The role of trans-local ties
Although knowledge actors are embedded in local contexts, they nevertheless rely on linkages with organisations in other regions to access resources in distant knowledge pools (Bathelt & Cohendet, 2014).These trans-local ties, often cited as 'pipelines', act as important knowledge sources for resilient regional development (Bathelt et al., 2004;Boschma, 2015).First, trans-local knowledge circulation helps identify what knowledge is needed for creativity and innovation and where to locate relevant complementary knowledge sources since 'local knowledge landscape is never complete' and useful knowledge is often distributed across regions (Cao et al., 2022;Maskell, 2014, p. 893).Second, trans-local ties help bring new perspectives to avoid knowledge lock-in of geographical proximate organisations (Boschma, 2005).Moreover, for cultural firms, the absence of ties to other regions means disconnection with mainstream industries and insufficient market for cultural products and services (Chapain & Comunian, 2010).
The benefits of trans-local knowledge pipelines are evident in several CCI subsectors.For European video game developers, for example, the engagement in global innovation networks, through which both industrial knowledge and market information are transmitted, is equally important as local symbolic knowledge sharing (Chaminade et al., 2021).Music industry players in North-West England maintain ties to other regions to validate the symbolic knowledge, to circulate product information, and alleviate market barriers for cultural players from small cities (Watson, 2020).Film industries of the United States, Canada and Germany maintain connections with other film production locations (Coe, 2000;Mossig, 2008;Scott, 2002).While the media industry in Leipzig suffers from the lack of pipelines to external partners and markets and becomes stagnated (Bathelt, 2005).
Trans-local ownership ties reveal the flow of investments and act as an important pipeline-building mechanism since they 'form a basic infrastructure for the development of further material, human capital and knowledge flows over time and provide more permanent trans-local connections between distant locations' (Bathelt & Li, 2014, p. 47;Mani & Moody, 2014).Therefore, in the CCIs, we expect that trans-local ownership ties also contribute to improved sustainability performance.
Here again, it should be noted that an excessive degree of trans-local pipelines may also create constraints.First, maintaining trans-local tie requires investments and time, which raise barriers for local small actors who lack financial support (Bathelt et al., 2004;Throsby, 2015).Second, local actors need to possess sufficient absorptive capacity to urn knowledge into innovation ingredients (Bathelt, 2005;Malmberg & Maskell, 2006).As the number of trans-local ties becomes disproportionately high, the balance between knowledge access and knowledge absorption is more difficult to maintain.Third, trans-local investments are often developed by organisations who are less sensitive to local communities and thus may pose threats on local cohesion, as illustrated by numerous cultural heritage regeneration projects (Qian & Li, 2017).These conflicts in turn jeopardise the sustainability of local communities (Pratt, 2009).Therefore, we propose the following hypothesis: Hypothesis 3: The number of trans-local ties of a city has an inverted 'U'-shaped relationship with its sustainable development performance.
Scholars suggest that local networks and trans-local ties should not be viewed in isolation.Rather, the interaction of the two generates conditions for the growth of cities (Bathelt et al., 2004;Boschma, 2015).Dense local networks provide a favourable environment for collocated actors to develop shared understanding of knowledge transferred from distant knowledge pools and trans-local ties reduce the possibility of regional knowledge overload by pumping fresh ideas and introducing alternatives of established knowledge routines, a combination that is termed 'mutually reinforcing' (Malmberg & Maskell, 2006).Moreover, recent studies on buzz-and-pipelines indicate that trans-local ties foster local buzz, an important local context that knowledge flow and knowledge accumulation are extensive, which in turn contribute to local development (Henn & Bathelt, 2018).
In the CCIs, trans-local investment ties help build more inclusive local networks and more balanced systems of cultural production through the flexible and timely circulation of capital generated in other regions.Interregional ownership ties open wider markets for local cultural production and help upgrade cultural resources for the city and this in turn enhances the role of local networks (Sasaki, 2010).Trans-local investment ties also provide valuable opportunities for trans-local labour supply and mobility and financial resources that are critical to address local creative talent employment precarity problems, which are often due to the lack of resources and the shortage of financial support (Farr-Wharton et al., 2015).Therefore, we argue that the interaction between local networks and trans-local ties benefits the sustainability performance of cities.

Research setting
To empirically evaluate our hypotheses, we have collected network data on the cultural heritage sector in China.The cultural heritage sector is a central part of the CCIs since cultural heritage is 'the origin of all forms of arts and the soul of cultural and creative industries' (UNCTAD & UNDP, 2008, p. 14).Conserving cultural heritage is fundamentally important to the CCIs because cultural production and creative services constantly draw resources and inspiration from it, as Scott (2010, p. 123) noted: [a]n added ingredient in this rich creative mix of inter-firm networks and local labour market relationships is place itself, not only as a spatial aggregation of industrial capabilities and skills, but also as a stockpile of knowledge, traditions, memories, and images.
In our study, we investigate the extent to which cultural heritage networks represent an important knowledge creation mechanism in Chinese cities that helps support sustainable development.
China is an excellent research context due to its rich cultural resources and long tradition of utilising local cultural heritage (WHC, 2021).According to the National Administration of Cultural Heritage (NCHA), as of 2021, China had 5292 'major historical and cultural sites protected at the national level' in more than 300 cities (NCHA, 2021).Cultural heritage sites are managed and operated by various government agencies, by public institutions, by state-owned tourism firms, or by private cultural firms at the local level.
Moreover, China is known for actively integrating transnational development agendas into its own national and local strategies.To implement the Millennium Development Goals, the predecessor of the SDGs, China has already integrated various aspects of sustainability into its 12th Five-Year Plan and built more than 100 sustainable development experimental zones that focus on localised sustainable urban development policies and strategies (State Council, 2016).The Outline of the 13th Five-Year Plan is also committed to the 'active implementation of the 2030 Agenda for Sustainable Development' and regions and cities have formulated roadmaps and localised annual plans accordingly (State Council, 2016, p. 7) We gave each city a composite score to measure the progress in achieving the cultural SDGs during the period 2015-19.Following prior studies, we used a four-step method that includes indicator selection, data collection, rescaling and normalising, and score aggregation (Xu et al., 2020).The appendix describes the four-step procedure in greater detail.First, we listed indicators used in prior studies, such as the Global SDG Indicator Framework, the reports on the US and European cities, and previous subnational reports focusing on China, and removed indicators with no comparable data at the Chinese city level (Lafortune et al., 2019;Lynch et al., 2019;Sachs et al., 2018;TUSDG & WWF, 2020;UN, 2022;Xu et al., 2020).This generated 34 indicators for SDGs 8-12.Second, we consulted publicly available databases such as statistical yearbooks and government websites to obtain data for these indicators.Third, we set the upper and lower bounds for these indicators, normalised them using the max-min method to obtain scores that range from 0 to 100, and calculated the mean of each indicator across the period 2015-19.Finally, we used the arithmetic mean of the 34 indicators as the composite score of a city's cultural SDGs.Table 1 documents the indicators and identifies top-performing cities.

Explanatory variables 3.3.1. Scale of the CCIs
The scale of a city's CCIs was measured as the number of CCI firms per capita.We consulted the Classification of Culture and Related Industries used by the National Bureau of Statistics of China (NBS) and identified firms in the National Enterprise Credit Information Publicity System (NECIPS) with industry codes of the classification as CCI firms (NBS, 2018). 1

Cultural heritage ownership network density
We relied on public records to identify the operators of cultural heritage sites, including social media pages of cultural heritage operators, the websites of government agencies, and the reports of CCI firms, the databases of NCHA and the NECIPS.Overall, we identified 2114 organisations related to 1355 cultural heritage sites in the NECIPS.
To collect data on the ownership ties of these 2114 organisations, we retrieved information on their shareholders and on the firms in which they invest from publicly available sources including the NECIPS, company reports, and social media pages of cultural heritage operators and local governments.An ownership tie was identified if an equity ownership relationship exists between firms.Following prior research, we did not distinguish the percentage of ownership but treated the ties as binary as exist or absent (Turkina & Van Assche, 2018).Since there is no consensus on how many steps of relationships should be included in this research setting, we followed the 'six degrees of separation' rule popularised by Milgram (1967) and Watts and Strogatz (1998).That is, we searched for shareholders and subsidiaries of the 1146 cultural heritage operators whose investment ties were within the range of six steps.This approach allowed us to develop an ownership network of 22,455 organisations with 21,898 ties between them during the period 2015-19.Finally, we used registered addresses to assign organisations to specific cities. Figure 1 depicts cultural heritage ownership networks of four representative cities in different regions.
The traditional network density measure calculates the proportion of existing ties relative to all possible network ties.However, this measure often does not accurately reflect the actual network density since it generates a low density value for larger networks and a high density value for smaller networks (De Nooy et al., 2018).Since the network sizes of the cities in our sample vary significantly, we used average degree of nodes in a city to measure how densely connected a network is, which is not dependent on network size in its calculation (De Nooy et al., 2018).
We constructed the variable in two steps.First, we counted degree centrality, namely number of direct ties, of each organisation in a city.Second, we calculated the arithmetic mean of all organisations' degree centrality and used it as the proxy of network density.

Trans-local ties
We defined an investment tie that extends beyond a city as a trans-local tie.For instance, if a firm has four subsidiaries in its home city and one subsidiary in another city, we assign the value '1' to the firm.Then we counted the total number of trans-local ties of all firms in a city and divided the number by the city's population to obtain the trans-local ties per capita.

Control variables
We incorporated several control variables to capture the differences in local cultural resource inventory, basic economic settings, and heritage network features.
Logged city area and logged GDP account for differences in resources, capabilities, and opportunities for utilising The nexus between the cultural and creative industries and the Sustainable Development Goals: a network perspective these resources since we are interested in tangible cultural heritages that are essentially immovable (Cerisola, 2019).
We also added two variables to control for the size of cities and the region in which they are located: city size (logged population of people residing in the metropolitan area of a city) and city region (dummy variable with East as reference and 1-3 representing Northeast, West and Central China) (China Statistics Press, 2023;NBS, 2022).Moreover, we introduced logged number of cultural heritage sites and percentage of open cultural heritage sites to reflect the inventory and utilisation of cultural resources.We identified the proportion of cultural heritage sites managed by professional organisations and open to public in the total heritage sites.Among more than 5,000 cultural heritage sites in China, many are under maintenance by the NCHA and not open to public audiences.
Since tourism is closely related to cultural heritage, we introduced scale of tourism firms by dividing the number of tourism firms by the city population (Loulanski & Loulanski, 2011;Su & Lin, 2014).We also introduced logged average registered capital of cultural heritage related firms of a city to indicate overall resources of nodes in networks.Finally, we calculated the centralisation of the cultural heritage ownership network since it represented the level of concentration of network actors in a city (Provan et al., 2007).

Model specification
We used multiple regression models to validate our research hypotheses.We also conducted a series of additional tests to check the robustness of these results.First, we analysed the nexus between the main variables and each of the five cultural SDGs separately.Second, we verified if core and supporting CCIs have a different relation with sustainable development outcomes by using the Classification of Culture and Related Industries which identifies industries such as tourism, retail, and transportation related to culture heritage as supporting domains (UNESCO, 2009).

RESULTS
Tables 2 and 3 report the descriptive statistics.We did not observe serious multicollinearity problem since all variables' variance inflation factor (VIF) values were below 10.
Table 4 shows our main regression results.Model 1 only includes the control variables.Model 2 adds our three variables of interest: scale of the CCIs, local network density and trans-local ties.Models 3-5 adds interaction terms for the three variables respectively.Model 6 adds the interaction between local networks and trans-local ties.
Hypothesis 1 proposes that the scale of a city's local CCIs has an inverted 'U'-shaped relationship with the city's sustainable development performance.In model 3 of Table 4, the positive coefficient of the variable and the negative coefficient of the squared term suggest that this is the case.Figure 2 shows the curve estimation between the scale of the CCIs and the cultural SDGs with 95% confidence interval.4 suggests network density is indeed relevant to city sustainable development.However, the squared term for the variable is not negative, thus rejecting the hypothesis.There are several possible explanations.First, the benefits of local ownership networks may be culturally sensitive and contingent on context (Broekel & Boschma, 2012).Chinese society is often classified as highly cohesive and thus dense networks may be more effective in promoting collaboration and building cohesive communities (Cattaneo et al., 2020;Li et al., 2021).This may help explain why our results fail to show a dampening of the positive relation between density and sustainability performance.It is also possible that the curvilinear argument is more prevalent in the innovation-related dimensions of the SDGs, but not with the other dimensions (Le Blanc, 2015;Pradhan et al., 2017).We will explore this possibility in our robustness tests by analysing the relation between network density and several separate cultural SDGs, some that are more innovation-related and others that are not.
Hypothesis 3 argues that the number of trans-local ties has an inverted 'U'-shaped relationship with the cultural SDGs.The coefficients of the variable and its squared term in model 5 of Table 4 provide support for the  The nexus between the cultural and creative industries and the Sustainable Development Goals: a network perspective 849 hypothesis. Figure 3 shows the curve estimation between the scale of the trans-local ties and the cultural SDGs with 95% confidence interval.Hypothesis 4, the interaction between local and translocal networks, is not supported.It is possible that our empirical analysis only reveals the long-term investment relationships in creating local buzz and trans-local pipelines.While the buzz-and-pipeline arguments also point out that knowledge access and transfer can be realised through temporary ties such as business conferences and projects combined with other favourable conditions for proximities (Boschma, 2005;Broekel & Boschma, 2012;Maskell, 2014).Considering that local networks and trans-local ties contribute to local sustainability independently, we acknowledge that there could be different types of strategies to benefit from network embeddedness and adapt to local contexts (Wu & Wu, 2016).
Several control variables reveal statistically significant relationships with the local cultural SDG performance.City-level GDP positively relates to local SDGs, which is in line with expectations.In terms of regional differences, compared to cities in the other three less developed regions, cities in East China generally reveal higher scores of the SDGs.
The more cultural heritage sites a city has, the lower score it receives on the cultural SDG progress.This negative relationship is aligned with observations in the prior literature that the mere existence of cultural heritage is insufficient to foster local sustainable development (Cerisola, 2019;Kourtit & Nijkamp, 2019).The concentration of cultural heritage sites in a city also indicates the potential conflicts between commercialisation and preservation of heritage.These conflicts become more complex when more heritage sites are involved since cities only possess limited resources that need to be distributed to different sectors.
To validate the robustness of our results, we have in Table 5 analysed the relation between our variables of interest and the various cultural SDGs.For SDG 8 (model R1), the results confirm Hypotheses 1 and 3, but are not aligned with Hypotheses 2 and 4. For SDG 9, which is related to industry, innovation and infrastructure, the results in model R2 are aligned with all four hypotheses.For SDG 10-12, none of the hypotheses is confirmed.These results suggest that the network perspective that we have developed in this paper explains particularly well how the CCIs support sustainable development through the generation of knowledge externalities, but less so in the social inclusion and the value systems shaping.
Finally, we investigated in R6 of Table 5 whether the proportion of core CCIs in the cultural heritage networks influences the sustainability outcomes.While the regression results confirm the benchmark findings, they do not find a discernible influence of the proportion of core CCI firms on sustainability outcomes.

DISCUSSION
This study provides two theoretical contributions.First, we highlight the important role that the CCIs play in   The nexus between the cultural and creative industries and the Sustainable Development Goals: a network perspective 851 REGIONAL STUDIES improving a city's performance in cultural SDGs.Second, we extend the study of local and trans-local knowledge dynamics to the relationships between the CCIs and sustainability.
Prior studies have illustrated that cultural heritage is a precious resource for local development by promoting creativity and innovation (Cerisola, 2019).This study extends this line of research by arguing the specific mechanisms of utilising cultural resources and explaining the critical role played by the CCIs, which directly draw inspiration from cultural heritage and function in a sustainable way (Foster, 2020;Pagan et al., 2020).The findings suggest that the CCIs deserve to be called an 'industry of tomorrow' given its contribution in building innovative and sustainable societies (UNESCO, 2021, p. 1).We also explore the limits associated with the creativity-oriented development and argue that as the CCIs grow the benefits diminish, especially when large similar economic activities become obstacles of innovation and related profit-pursuing sectors are involved (Kacerauskas et al., 2021;Pratt, 2008).The results of the additional test on SDG 8 further support the dynamics of local versus trans-local knowledge creation by identifying the curvilinear trends and the mutual reinforcing mechanisms of local networks and trans-local ties (Malmberg & Maskell, 2006).Moreover, our additional analyses reveal interesting nuances and trade-offs between different dimensions of the SDGs.Although both the scale and the network of the CCIs contribute to sustainable city building, the emphasis of SDG 11, culture-based development is not always beneficial for social inclusion and reduction of inequality, the focus of SDG 10.This result indicates that there are indeed risks associated with excessive culture led development trajectory such as gentrification, inequality, and loss of identity (Pratt, 2009;Qian & Li, 2017).
Second, we explicitly incorporate the business sector, a key player contributing to sustainability, into the discussion of the cultural SDGs (Pfisterer & Van Tulder, 2021).This study thus addresses the limitation of the SDG framework that the role of the business sectors is often unaddressed (Scheyvens et al., 2016).Not only do we explain how agglomeration of CCI firms can contribute to local development, but we also explore the implications of the structure and distribution of CCI networks within and across cities.Therefore, this study complements existing case study insights by providing more generalisable evidence of how businesses contribute to achieving the SDGs.By applying the buzz-and-pipeline arguments to activities related to cultural heritage in cultural sectors, this study extends the local and trans-local knowledge creation dynamics to innovation and creativity based sustainable development trajectory.
This study has several policy implications.First, our results suggest that government support for the CCIs at the city level can help the implementation of the SDG framework locally, especially when it remains in its infancy.Moreover, our focus on networks indicates that not only the scale of cultural activities matters, but also the network relationships between cultural business actors play an important role.These findings suggest that government agencies should go beyond supporting the attraction of CCI firms to cities and also develop policies that encourage collaboration and interaction in the CCIs both locally and trans-locally to foster dense networks and to access necessary resources in distant locations.
Second, there are only limited studies on the localisation of the SDGs at city level, especially in emerging markets, and we therefore believe that it can spur significant interest among policy professionals (UNDP & CASS, 2020; Xu et al., 2020).Since many aspects of the sustainability challenges are localised, it is critical to monitor local SDGs and investigate the effects of local business activities on sustainability (Petti et al., 2020;Xu et al., 2020).Policy agendas focusing on the localisation and the monitoring of the SDGs could benefit from the insights of this study by acknowledging the different possibilities of achieving sustainability and by exploring the effectiveness of local development initiatives (Forestier & Kim, 2020).Our results indicate that translocal investments into the CCIs can be particularly beneficial for several SDGs such as the ones related to innovation, economic development, and sustainable urban communities.Carefully exploring differentiated initiatives tailored to local conditions is equally important.For instance, although eastern cities generally outperform the inner land cities in the The nexus between the cultural and creative industries and the Sustainable Development Goals: a network perspective

854
Yang Gao et al.

REGIONAL STUDIES
SDGs, it is possible to pursue the SDGs through strategies relying on trans-local collaborations, just as the case of Nyingchi and Karamay cities show (Department of Tourism Development, 2017;Zhang & Gao, 2005).To summarise, implementing effective cultural policies that draw inspiration from local cultural resources, building dense local networks, and actively engaging external partners and markets are critical factors for local sustainability.
Our study also contributes to our understanding which SDGs are more closely related to the CCIs and cultural heritage networks.Our results reveal that although the CCIs and networks contribute to the overall cultural SDGs, they have differentiated roles in each of these goals and future studies can further explore these nuances (Le Blanc, 2015;Pradhan et al., 2017;Sachs et al., 2019).
This study also bears some limitations in terms of conceptualisation, data availability, causal reasoning, and generalisability.First, due to data availability, we focused on culture heritage ownership networks in cities and ignored other types of interorganisational networks such as supply chain networks and strategic alliances (Provan et al., 2007).Prior research suggests that embeddedness in supply chain networks both contributes to innovation of the CCIs and stimulates innovation in other sectors (Bakhshi & McVittie, 2009;Shafi et al., 2019).Also, when exploring the structure of networks, we simplified the interorganisational networks in different cities, conceptualised investment ties as undirected, and did not consider weights of ties, which lose some of the possibilities of discovering interesting network dynamics (Phelps et al., 2012).
Second, we only included tangible cultural heritage in our analysis and left out intangible cultural heritage and natural heritage, which are also important factors in local sustainability development.The emphasis of the cultural SDGs limited our analysis mainly to the scope of social, economic, and cultural sustainability while the discussion of the environmental dimension is absent (Forestier & Kim, 2020).Although this approach is aligned with prior studies that chose the SDGs based on relevance, we believe a more comprehensive research design could provide better understanding of relationships between local actors' behaviours and the SDGs (Stafford-Smith et al., 2017).
Third, we do not propose strict causal relationships in our study and thus the results of the study should be viewed as tentative.Although we introduced several control variables and collected data on explanatory variables, this research design does not rule out the possibility of reverse causality.
Finally, results of this single-country and single-industry study can be contrasted with outcomes from more comprehensive international and cross-country study initiatives to see whether the empirical results hold up in different contexts.We leave this for future work.

CONCLUSIONS
Although it is widely acknowledged that cultural heritage and economic activities related to culture and creative are relevant for sustainable development, we lack in-depth analysis of how the agglomeration and the structure of the CCIs matter for sustainability performance.To address this limitation, we adopted a network perspective and built on the buzz-and-pipeline approach to study the relation between culture heritage ownership networks and city-level SDG performance.We argued that a city's scale of the CCIs, the structure of the cultural heritage ownership networks, and trans-local ties contribute to the overall progress in SDGs 8-12 and tested the propositions on 292 cities in China.Our study provides new insights on the nexus between cultural activities and sustainability and offer practical implications on how to build sustainable cities by encouraging partnership and interaction in the CCIs.

Hypothesis 4 :
Trans-local ties positively moderate the relationship between local ownership network density and sustainable development.
an inverted 'U'-shaped relationship between the cultural heritage ownership network density and the cultural SDGs.The positive and significant coefficient for local network density in model 4 of Table

Figure 1 .
Figure 1.Cultural and creative industry (CCI) ownership networks and basic cultural heritage information of four cities.

Figure 2 .
Figure 2. Relationship between the scale of local cultural and creative industries (CCIs) and the cultural Sustainable Development Goals (SDGs) (with 95% confidence interval).

Figure 3 .
Figure 3. Relationship between the trans-local ties and the cultural Sustainable Development Goals (SDGs) (with 95% confidence interval).
. In the implementation of the SDGs, government agencies provided institutional support for innovative solutions for urban development challenges in different cities and formulated various policies in innovation, agriculture, information technology, poverty reduction, health, biodiversity and climate change (State Council, 2016; UNDP & CASS, 2020).nexus between the cultural and creative industries and the Sustainable Development Goals: a network perspective 845 REGIONAL STUDIES 3.2.Dependent variable To develop a measure of a Chinese city's sustainability performance, we focused on SDGs 8-12 which theoretically speaking should benefit most from stronger CCIs.This includes SDG 9, which focuses on industry, innovation and infrastructure and thus more likely benefits from knowledge externalities generated by the CCIs.It comprises SDG 11, which addresses the sustainable cities and communities.It includes SDG 8 on decent work and economic growth, which can be boosted by the CCIs' provision of employment opportunities, empowerment of marginalised groups and contribution to equitable societies.It includes SDG 10, which focuses on reducing inequalities.And it comprises SDG 12 on responsible consumption and production due to the fundamental role that the CCIs have of shaping value systems.Taken together, we categorise SDGs 8-12 as the 'cultural SDGs'.

Table 1 .
Cultural Sustainable Development Goal (SDG) indicators and the cultural SDG score for the top 10 performing cities.

Table 3 .
Correlations of the variables.