Cluster advantage and firm performance: a concluding remark

Cluster advantage and firm performance: a concluding remark This book does not represent only an up-to-dated analysis on the industrial districts and clusters (ID/C) theory, involving several experts coming from different academic experiences and European country-origins (Spain- ch. 2; ch.5; Ch.6; ch. 8; ch. 10; Ch. 13; Italy – ch. 2; ch.4; ch.7; ch.9; ch.11; ch. 12; ch.14; ch.15; Germany – ch.2; UK – ch. 4; ch.11, Austria – ch.15, Poland – ch.3; US- ch.6), it is also a tentative to settle the issue of cluster into a more deep investigation focused at the micro level: How firms originate within ID/Cs, how they develop, and how they decline. The main idea is to provoke an analytical shift towards the unit of analysis: not just the forest, the cluster, but the individual trees, the firm, that form the forest, and the aggregated subunits of bushes, clearings, and the mix of varieties that lives in symbiosis. Clusters: a cross-boundary concept The first interesting contribution to the analysis of the cluster concept derives by an historical long-term exploration of the literature. Ch. 7, applying the bibliographic analysis to the evolution of ID/C literature, guides us to further study this cross-boundary concept, that confirms the trend towards management-related topics, where innovation and firm performance are the leading issues. The analysis is based on 8,381 articles and 829 journals. Analysis is splitted in two periods (1985-1999 and 2000-2013). The first most relevant journals in the literature, in terms of number of the articles published, were in the first period Strategic Management, Journal of Environmental Planning, Regional Studies, Academy of Management Review, and Research Policy, while, in the second period, we can observe Research Policy, Regional Studies, European Planning Studies, Strategic Management, and the International Journal of Technology Management. We see that the general boundaries of the discipline are extensively involving management studies. In those years we moved from a local development approach and from sociological related issues to management-related topics, and from the field of Economic and Geography to the fields of Management and Innovation. The content analysis of the ID/C literature, divided in three periods, is moving from the concepts of a) flexibility, location, flexible accumulation, embeddedness, to b) knowledge spillovers, patent citations, proximity, knowledge transfer and tacit knowledge, and to c) knowledge spillovers, knowledge transfer, patent citations, structural holes, and innovation systems. The scholarly debate that characterised the 1980s was developing around the creation of sustainable growth path based on ID/C and flexible small firms models. Product differentiation and innovation allowed the construction of post- mass production systems. During 1990s Silicon-like ID/C brought a growing interest in the origins and dynamics of production networks. Several studies supported the new approaches towards a co-evolutionary view of firms, firms networks, social structures, and local institutions. In the period 2000-2009 literature becomes increasingly focused on innovation and knowledge spillovers, together with absorptive capacity. It was discovered that territorial proximity triggered innovation and collective learning, allowing leading innovators to develop endogenous growth paths. In this context, university-industry relations, venture capitalists, technological transfer via interactions clients-suppliers-subcontractors, showed how ID/C may develop distinctive organisational knowledge and dynamic capabilities. The analysis referred to the literature of 2010-2013, illustrates the existing mix between firm-level and cluster-level studies, where knowledge absorption from external sources starts to become relevant as the issue of international strategic alliances, and ID/C internationalisation. Industrial districts and cluster typologies There are many ways of classifying ID/Cs according to the process(es) through which cluster benefits are produced. In this book we have considered the Marshallian notion of industrial district as a synonymous of the Porterian notion of cluster. In Belussi (1996) and Belussi (2015) we have put forward a careful examination of the differences and similarities. Marshallian districts were characterised during 1920s by the presence of traditional and low-tech sectors such as wool and cotton (see the Lancashire case discussed by Belussi and Caldari, 2009), footwear, or engineering industries based on artisanal skills, such as Sheffield in the production of cutlery automobile manufacturing. The formation of ID/C continued also in the post II world war in advanced countries, in the same sectors which declined in Britain after 1930s, but that emerged in Italy, Spain, France and Germany (in textile clothing, footwear, packaging, tiles production, furniture, automobile, wine industry, etc.), through the agglomeration of flexible SMEs. In the last 30 years, in contrast, many researchers have envisaged also a new phenomenon: the formation of high-tech ID/C: in ICT and software (see the paradigmatic case of Silicon Valley Saxenian, 1994), biotech-pharma, finance, pharmaceuticals, finance, advertizing, and media sectors (Karlsson, 2008). There are other ways of classifying ID/C without just referring to their sectoral dominance (in manufacturing, service sectors, or agriculture). ID/C can be characterized by whether the goods and services that they produce are in fast or slow growing sectors nationally or internationally, or, again, by the nature of the labor force skills at their core (low-skilled or high-skilled), or by the average wages paid by local firms, or again, by their export performance (Simmie, 2008). Finally, ID/C can be characterized by being located in urban or peripheral areas (Feldman and Audretsch, 1999). Another set of ID/C differences (Wolman and Hincapie, 2014) may have to do with the extent to which clusters are consciously organized at local, or regional level, through the creation of cluster organizations induced by the intervention of cluster policies (with human intervention aimed to create, build upon, or improve a cluster) or whether their functioning is just explained by pure market forces that occur naturally. Following the Markusen’ typology (1986), ID/Cs may be spontaneous (Marshallian, or with activity aggregated among one or few leading firms: hub-and-spoke) or planned by government (science-based clusters localised in science parks) or again, deriving by the activity of MNEs entry in developing countries. In the latter case they are called by Markusen’ satellite clusters. In fact, the Markusen’ typology based her theoretical framework on the size of the firms that are part of the ID/C, their linkages and networks within and across the district, and the distribution of power among firms. In contrast, Gordon and McCann (2000) and Iammarino and McCann (2006) have posited three basic models of cluster processes which are looking more generally to the modality of agglomeration, the sectorial specialisation, the network activity among firms, and the level of social embeddedness. They distinguish between process of agglomeration (territorial concentration), clustering (specialized concentration inter-firm linkages) or “distrectualisation” (historical specialized concentration showing social embeddedness), presenting three models: • Pure agglomeration economies • Industrial complex • clusters with social networks They classify agglomeration, clusters (localized inter-firms transactions), and industrial districts (Italianate model of social integration) as radical different types of local systems. However, their typology is rigid and static. In own view local systems can evolve from one type to another (cluster ↔ district; district ↔ cluster). For instance, many industrial districts localised in South of Europe, after the recourse to delocalisation strategies (Sammarra and Belussi, 2009), to global supply chains (Gereffi, et al., 2005), and having suffered from the 2007 crisis, have radically transformed their industrial structure and we have observed a diminishing of cooperation, social benevolence, trust, and mutual support, and a radical emergence of leading large firms, with the entry of MNEs (Belussi and Hervas-Oliver, 2017; Belussi, Caloffi, and SeditaThe , 2018). This has clearly blurred the difference between the idea of industrial district and cluster. Considering this literature, Ch. 11 offers a reflexion on the endogenous rerouting and longevity of ID/Cs. In their critical review, the focus mainly on the analysis of radical knowledge creation in ID/Cs, as the main element distinguishing the historical evolution of this territorialised form of development, which can be historically described through the mark I, mark II, and mark III typology. Following Belussi and Pilotti, (2002), Belussi and De Propris (2013), and Bellandi, and De Propris (2015). Mark I relates to a complex socio-economic adaptive system characterized by a path accumulation localized technical knowledge and decentralized industrial creativity (Bellandi, 1996). While Mark I represents the typical Marshallian district, Mark II is the result of the reemergence of ID/Cs during 1980s in a context of flexible specialization and robust transition capacities, and processes of learning by doing, using, and interacting (Asheim, 2000). In Mark III ID/Cs should avoid rigid specialization traps, exhaustion of innovation thrust, and lock-in clashes with constantly increasing innovation capacity of global competitors. Exploring and exploiting new global knowledge they have to overcome inertia, rents-seeking behaviors, and coordination problems. Knowledge, here, is more codified and may come from gatekeepers or trans-local anchor firms (Belussi, 2015). The district effect revisited is the focus of Ch.3, where an empirical study is presented regarding Spanish local systems and districts. The study of innovative firms is based on data regarding patents and utility models (mainly designs) registered during the period 2001-2005. Local systems are characterised as industrial districts, as manufacturing areas where large firms predominate, as not specialised manufacturing areas, and as large metropolitan areas. Counting on average the number of innovations per area, the most intensive innovative type of local systems results to be the “pure” Marshallian districts, with 446 innovations per million of employees, followed by the metropolitan areas, with 427 innovations per million employees. The third position is conquered by manufacturing areas where large firms predominate with 366 innovations per million of employees. Weighted patents (considering the costs of application for obtaining a patents among the different offices: national, European, and wipo), give the predominance to large metropolitan areas (178 innovations per million of employees), and then to ID, with 135 innovations per million of employees, while in the third position we find manufacturing areas where large firms predominate (127 innovation per million of employees). Therefore, in relations with their findings, both considering unweighted patents and weighted patents, and estimating the contribution of many innovation-related variables, the authors are able to demonstrate that industrial districts cannot be considered weak innovators. Also Ch. 9 represents another industrial district “exercise”, presenting a qualitative and quantitative analysis of the long-term development of the footwear industry in Italy and Turkey, focusing in particular on four main industrial districts/clusters (one in Italy and three in Turkey). Agglomeration benefits appear to exist in the various initial stages of the ID/Cs life cycle (Belussi and Sedita, 2009), but not for final phase of the main “mature” ID localized in Italy: the Montebelluna cluster, that now has taken the form of a multi-localized cluster in Timisoara and China, where many former small firms are now large homegrown multinationals. In Montebelluna homegrown multinational firms established after the 1990s (Tecnica, Geox, Alpinestars, Aku, etc.). During cluster emergence the presence of a specialized local labour market, and the formation of a district atmosphere, characterized by the circulation of ideas among entrepreneurs, was a common feature, as described by Marshall (1920). Later on, the subsequent stage of cluster development is driven by the ability of some leading firms to connect the cluster (and its internal supply chains) to external markets, to global knowledge sources and to a global supply chain). In addition, a large firms heterogeneity predominate: not all firms show an accelerated pattern of growth after being located in the cluster. Apart from the life cycle, the four clusters differed also in terms of the economic external environment (mature vs. emerging fast-growing countries), for the existence of countries-specific institutions (among which labour regulations and environmental protection), for innovation intensity (high innovative clusters vs. imitative clusters), and the political frame (free market policies vs. defensive-barriers to import policies). In the Istanbul cluster, the leading role is played by large Turkish retail chains, which are also producers, but which buy 40-50% of their sales from other Turkish firms mainly located in Turkey clusters. The most dynamic Turkish leading firms is Zylan, which recently entered the Montebelluna district with a greenfield investment, focused on prototype design for the Turkish production. Zylan has also acquired the brand Lumberjack from Canguro (an Italian firm based in Verona, that went into bankrupt), together with its distribution nets. This means that globalisation is now creating firms networks among global districts, beyond the existence of global value chains. A more theoretical chapter, written on similar theoretical research questions, is Ch. 5. How ID/Cs evolve? Are leading firms and gatekeepers feeding the process of introduction in firms of new technologies and original breakthrough innovations? Leading companies impose their technological trajectories on firms in their orchestrated network. In ID/Cs, therefore, leading firms are mainly responsible for upgrading industrial districts shaping a district’s learning process (Lorenzoni and Lipparini, 1999), as long as knowledge upgrading is incremental. A fact that can promote lock-in in the long term, but that make ID/Cs to extend their stages of growth. The authors hypothesize that new radical knowledge in ID/Cs is introduced not by incumbent, but by new firms, which rejuvenate old rigid trajectories. They state, thus, that the entrance of new firms brings new knowledge, renewing the existent technological path, and favoring district evolution. Therefore, disruption in ID/Cs often needs knowledge coming from outside the sector-specific technology developed in the cluster. To conclude, first, in ID/Cs, leading incumbents demonstrate predominantly an orientation towards the creation of incremental-sustaining knowledge, but they do not create important breakthrough; second, radical disruption can be expected to be led by new firms and not by incumbents or technological gatekeepers; third, disruptive ideas must come from other industries, non-related technological fields, and they must be based on external linkages, forming in this way new technological trajectories which may renew clusters. Clearly, this theoretical approach deserves further elaborations with the support of empirical evidence, and helps to design new lines of future research. The role of collective actors and local/regional policies in ID/Cs upgrading and path renewing ID/Cs evolve because the influence of spontaneous changes and deliberate collective actions. Ch. 3,10, and 15 contribute to develop this line of reflection. In Ch. 3, the case of the Polish boiler-making cluster, in the region of Wielkopolska, illustrates how a cluster organization supported by the EU founding has organised several cooperative innovative activity, in research and in the adoption of more ecological standards, among the SMEs belonging to the ID/Cs. The top-down assistance of a cluster organization has also played a distinguished role in promoting the internationalisation process of the cluster. For years, local supporting organisations have been focused on providing to ID/Cs firms specialized services, fostering innovation. Nowadays, thanks to the increasing connectivity, they have become knowledge catalyzers and gatekeepers of knowledge of knowledge, mediating between local and extra-cluster firms. This is the main contribution developed in Ch. 10. Using data collected in the Toy valley in Spain, this chapter analyzes brokerage behavior. Firms and supporting organizations exchange different types of knowledge (technical and market-knowledge) in different ways. Endorsing micro-level polymorphism in clusters, this study verifies that cluster actors perform diverse roles when transferring different knowledge. Market knowledge is brokered by a much more reduced set of actors, thereby suggesting more selective knowledge diffusion. In the cluster, technical knowledge is mediated by universities, by a technological institute, and by a local toy business association. This suggests that being a broker depends on certain micro-level characteristics. Several organisations are able to mix market and technical knowledge thanks to a wide number of relationships, helping to circumvent potential technological bias. Surprisingly, although limited to coordination and despite their technological focus, universities mediate both technical and business knowledge. While suppliers or toy manufacturers import knowledge from outside producers, local organizations mostly focus their gatekeeper activities on other local supporting organizations. Inspired by the recent literature on smart specialization policies, Ch. 15 examines 16 regional cases in which cluster polices have been recently developed, distinguishing among well developed, intermediated and less-developed regions. An interesting frame has been developed by the authors, which distinguishes between continuous and discontinuous – radical or breakthrough - pattern of change. Types of regional industrial path development Form of path development Key characteristics Change New path creation Rise of entirely new sectors deriving from breakthrough innovations New path entry of established industries Setting up of an established industry that is new for the region, often based on the inflow of FDI Path ramification Ramification-speciation of knowledge of existing industries into new but related ones industries Path upgrading and renewal Major change of an industrial path into a new direction based on new incremental/radical innovations or new organisational forms Continuity Path extension Continuation of existing industrial paths based on incremental innovation along established technological trajectories (danger of path exhaustion) Source: own compilation (modification on Trippl et al. 2016, 2017) The category of path extension reflects the continuation of an existing trajectory. Path upgrading and renewal, is related to the introduction of new incremental or radical innovations in ID/Cs. Path ramification relays on the introduction of new sectors through a process of “speciation”, through knowledge recombination. New path entry describes the setting of an already existing specialization in a region in the cluster (this sector is new for the region but is not new for the market). A new path creation (the emergence of a new specialisation based on breakthrough innovations) represents a truly novelty for the region and reminds the rerouting strategy discussed in Ch. 11. In the analysis of the application of smart specialisation policies, this chapter also discusses the issues of “prioritization” and “stakeholders involvement”.   References Asheim B. (2000). Industrial Districts: The Contributions of Marshall and Beyond, 2000 : Innovation Networks, Regions, and Globalization. In: Clark, G., Feldman, M. and Gertler, M. (Eds). The Oxford handbook of economic geography (pp.253-274). Oxford. Bellandi, M., & De Propris, L. (2015). Three generations of industrial districts. Investigaciones Regionales, (32), 75-87. Belussi, F. (2015). The international resilience of Italian industrial districts/clusters (ID/C) between knowledge re-shoring and manufacturing off (near)-shoring. Investigaciones Regionales, 32:89–113. Belussi, F. & Pilotti, L. (2002). The development of an explorative analytical model of knowledge creation, learning and innovation within the Italian industrial districts. Geografiska Annaler, 84, pp. 19-33. Belussi, F. & Sedita, S.R. 2009. Life cycle vs. multiple path dependency in industrial districts. European Planning Studies, 17(4): 505-528. Belussi, F., & De Propris, L. (2013). They are industrial districts, but not as we know them!. In: Giarratani F., Hewings G.J.D. & McCann P. (eds.), Handbook of Industry Studies and Economic Geography, Cheltenham: Edward Elgar Belussi, F., Caldari, K. (2009), “At the origin of the industrial district: Alfred Marshall and the Cambridge school”. Cambridge Journal of Economics. (33): 335-355. Feldman, M. and Audretsch, D. (1999). Innovation in cities: Science based diversity, specialization and localized competition. European Economic Review, 43, 409-429. Gereffi, G., Humphrey, J., Sturgeon, T. (2005). “The governance of global value chain”, Review of International Political Economy. 12(1): 78-104, Taylor & Francis, London. Gordon, I.R. and McCann, P. (2000). Industrial clusters: complexes, agglomeration and/or social networks?. Urban Studies, 37(3): 513-532. Iammarino, S. and McCann, P. (2006). The Structure and Evolution of Industrial Clusters: Transactions, Technology and Knowledge Spillovers. Research Policy, 35, 1018-1036. Markusen, A. (1996). Sticky places in slippery space: a typology of industrial districts. Oxford University Press. Porter , M. (2000). Locations, clusters and company strategy. European Planning Studies, 5 (1), 3-23. Rosenfeld, S. (1997). Bringing Business Clusters into the Mainstream of Economic Development. Economic Geography, 72 (3), 293-313. Saxenian, A. (1994). Regional advantage: Culture and competition in Silicon Valley and Route 128. Cambridge and London: Harvard University Press. Simmie, J. (2008). The contribution of clustering to innovation: from Porter I agglomeration to Porter II export based theories, In: Karlsson C. (ed.). Handbook of research on innovation and Clusters, Elgar, Cheltenham. Wolman H. and Hincapie D. (2014). Clusters and Cluster-Based Development. Economic Developed Quarterly, 29(2): 135-149.