ساختار سیستم های بین صنعت و انتشار نوآوری : مورد اسپانیا
کد مقاله | سال انتشار | تعداد صفحات مقاله انگلیسی |
---|---|---|
2287 | 2012 | 20 صفحه PDF |
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Technological Forecasting and Social Change, Volume 79, Issue 8, October 2012, Pages 1548–1567
چکیده انگلیسی
This paper focuses on the role of inter-industrial structures and the position of economic sectors in them for the diffusion of knowledge and innovation. Network Theory and Social Network Analysis have been applied to analyze the structure of the Spanish Input–output system and its evolution over a thirty-five-year period. The structural analysis conducted tests the existence of a Scale-free topology and also includes the identification of sectors acting as hubs or super-spreaders, which make up the core of the system. Scale-free networks correspond to structures that allow for faster and more efficient diffusion processes that are enhanced when initiated in hubs. As a concluding remark, this paper puts forward a proposal for interventions to attain a higher incidence in the national innovative capacity and in the development process.
مقدمه انگلیسی
Even if Network Theory (NT) is still a novel methodology in economics [1] and [2], it is increasingly accepted that the economy is a complex system that requires deep systemic analyses, starting from its topological characteristics [3], [4] and [5]. More precisely, understanding the structure and dynamics of economic networks requires the study of the structural properties of the underlying interaction networks and their dynamics [6]. This paper analyzes the structure and evolution of inter-industry systems in Spain, from NT and Social Network Analysis (SNA). It adopts a systemic approach and the structural change definition stated in Saviotti and Gaffard [7]: “In the past, the concept of structural change has been interpreted in the economics literature as a change in the weights of different sectors. However, today it is increasingly evident that a broader concept of structural change is required. In a systemic framework, structural change can be defined as a change in the structure of the economic system, that is, in its components and their interactions”. When this structural view is applied to innovation processes, it is assumed that actors acquire and develop overlapping and diverse knowledge resources through interactions with other actors, and that the newly acquired knowledge can be converted into new products, patents and other tangible forms [8]. Knowledge is more likely to be transferred between organizations that make chains or systems than through independent organizations [9]. In general terms, knowledge flows between two actors are made easier when the actors are embedded in a dense network of third-party connections. This is the case of economic sectors embedded in dense production networks and exchanging knowledge and innovation. Although the processes of production and of innovation differ in important respects “they are also mutually interdependent” [10]. In accordance with Hauknes “At the firm level, it should be evident that for most firms, their relations with customers, competitors and suppliers are the most significant links to their environment, to the extent that these agents constitute the major dimensions of this environment. It is not unlikely that these immediate relations shape the major learning modes for a majority of firms” [11]. In many innovation surveys, as in OECD surveys [12], firms indicate that suppliers, customers and competitors are ‘highly important’ sources of knowledge for innovation. User–producer interactions, joined through inter-industry linkages, play a fundamental role providing embodied knowledge flows in incremental innovation and in the learning process [10], [13] and [14]. In the same line, Schmookler [15] points out that “the best way to improve an industry's technology is often to improve the inputs it buys from other industries”. Even if not all user–producer relationships promote innovative activities [10], all of them constitute opportunities to increase the efficiency of policy interventions. The ability to take advantage of those opportunities depends on the structure of production networks. Structure matters, but there is a diversity of them that shapes dense networks, and Scale-free is outstanding among them. The role of topology in the study of the diffusion of innovations and of the effectiveness of innovation strategies is emphasized by the picture emerging from the system of economic interactions. At the same time, an innovation flow may die out on the same network immediately or persist for a considerable time, depending on the sector where it was originated. We know that our knowledge about the interactions that allow and promote innovation flows will improve by going in depth into inter-industrial structures, with its policy implications. The availability of a considerably long time series of Input–output Tables (IOT) should not be passed by. A structural analysis of inter-industry networks would contribute to the understanding of how innovation flows, to identifying highly connected sectors, named hubs or super-spreaders, that speed up the process, and to improving the design of interventions. In the case of Spain this is particularly relevant because of its innovation backwardness, its efficiency problems and the lack of effectiveness of its innovation policies. Research on the impact of innovation on productivity growth and on other economic variables, by Input–Output (IO) analysis, was initiated by Terleckyj and Scherer, who assume that R&D is indirectly incorporated by purchasing intermediate inputs [16], [17] and [18]. For DeBresson [19] IOT can serve as economic maps that indicate which are the paths of least resistance for the industrial diffusion of the technologies when the analysis is focused on market relationships and on the accumulation of technological knowledge, through experience based on the circulation of goods and services and on the process of learning by doing. “In order for a new technology already adopted by industry i to be subsequently adopted by industry j, it is preferable that industry j be in direct contact, as a client or as a supplier, with industry i. In other words, the two industries must be directly linked in an input–output table by a supply–demand relationship”. DeBresson remarks that the information embodied in IO interactions is particularly useful for the analysis of the productive structure of the whole economy [19], [20] and [21]. A new research line was opened up by Leoncini et al. [22] with the identification and study of Technological Systems by combining IO and R&D data and applying NA [23], [24], [25] and [26]. According to Montresor and Vittucci [25], IO coefficients crucially affect learning by interacting and the entailed knowledge networks that firms establish in innovating. More specifically, IO matrices map inter-sector flows of goods and services which shape the inter-sector diffusion of innovation by channeling and driving both embodied and disembodied innovation flows and the knowledge embedded in the exchanged goods and services. Their work fits into a wider field that considers that organizations acquire knowledge through interactions with other actors, making chains and systems and explaining that knowledge and innovation spread through intermediate trade linkages [8], [9], [25], [27], [28] and [29]. Our argument is also in line with other relevant pieces of research. This is the case of the literature of systems of innovation studying system failures, with the focus on missing connections to support knowledge processes through interactive learning [30], [31], [32] and [33]. A system failure policy implies that the framework conditions for a better diffusion and adoption process taking place across the structure of economic activities should be set. Actors supplying knowledge and innovation through sales and also users and consumers of goods, receiving information and probably adopting innovations through them should be taken into account. The present paper is placed in the above literature both for its objective and its methodology. However, its focus differs as it is the structure of inter-industrial systems. We do not focus on direct relationships between two particular sectors but on the chains and sub-systems that are making up the whole inter-industrial structure. The structure and evolution of intermediate trade relationships in Spain in the period 1970–2005 using IOT are analyzed because the constituting networks that represent inter-industrial systems push production systems into the open [34] and [35] and act as a platform for interactions that ease learning and the processes of knowledge and innovation diffusion. The study of its structure implies a first necessary step, not addressed in the literature, before studying more specific topics affected by it. It is valuable not only for scholars but also for policy makers because its results relate to productivity and the national innovative capacity, and hence to the enhancement of development [36]. This paper raises the following questions: Does the inter-industry network in Spain show a Scale-free topology where a core and a periphery can be identified? How has it evolved in the period 1970–2005? Can hubs, or super-spreaders, sectors be identified? Are there specific strategies that can be proposed from a relational analysis to improve the diffusion of ideas, knowledge and innovation? By answering these questions this paper aims to fill a void in the literature by studying the relevance of structures for diffusion processes, particularly for the diffusion of innovations through inter-sectoral interactions. In doing so, this paper is methodologically coherent, as the economy is viewed as a complex system, and systemic methodologies are applied (SNA and NT). Following this structural view, the Scale-free topology of inter-industry networks has been analyzed and a core and a periphery have been identified; results also indicate that the core–periphery structure is consolidated and that there is a set of sectors in the core with a permanent character in the period considered. Hubs have also been identified, opening up a discussion on the suitability of the sectors which innovation policies are being directed at. Through these results, this paper offers novel contributions to the methods of identifying core–periphery networks, the analysis of innovation flows between economic sectors by using IO data, the design of public and private interventions that would enhance a more efficient diffusion of knowledge and innovation, and the methods to verify whether the selected sectors in innovation programs are the most appropriate in terms of scope and speed when a systemic effect is intended. The paper is organized as follows: Section 2 explains the most relevant theoretical matters and the methodology followed; Section 3 presents the Spanish context in the period analyzed and the data used; Section 4 contains the empirical analysis conducted to examine the structural evolution and the core–periphery structure of IO networks; the final section gives the conclusions.
نتیجه گیری انگلیسی
In Barabási [3] it is asserted that “interconnectivity is so fundamental to the behavior of complex systems that networks are here to stay”. Interconnectivity among different actors (persons, countries, firms, etc.) generally implies multiple and overlapping linkages making up complex networks and allowing for a wide range of flows (ideas, affects, goods, technology, etc.). Among the several disciplines that have demonstrated the relevance of interactions and networks, Sociology and Economics have emphasized the case of firms. Firms interact with other business and non business institutions, implying learning processes and also material and immaterial flows making up networks. This paper focuses on the structural analysis of the network made by the intermediate exchanges maintained by economic sectors. This constitutes a main relational dimension for firms as it generally implies repeated and also frequent and intense interactions. Considering an IOT as a map of relevant economic flows [19], this study examines the potential role of the Spanish inter-industrial structures in the diffusion of knowledge and innovation. Two complementary methodologies, SNA and NT, have been applied and novel contributions are offered to the methods of identifying core–periphery networks. The algorithm proposed in Borgatti and Everett [71] has been applied and a high correlation coefficient is obtained. Also a rigorous NT method has been used to analyze MSF structures and the existence of hub nodes has been tested. These nodes make a core sub-system of high density maintaining the whole network connected. This paper also adds to the literature on the analysis of innovation flows between economic sectors by using IO data. National IO matrices have been studied as interaction platforms where knowledge and innovation flows take place. Depending on these inter-industrial structures, innovation flow processes could be very different. In Spain, the inter-industry network shows a core–periphery structure with Scale-free properties. This implies that different sectors play quite diverse roles in diffusion processes. The evolution of the analyzed network also indicates that the core–periphery structure is consolidated in the period considered (1970–2005) and that there is a set of sectors in the core with a permanent character: Metal products, Electricity, Construction, Trade, Hotels and restaurants, Land transport, Financial intermediation and Business services. An in-depth view indicates that the core, where hubs or ‘superspreaders’ can be identified, is mainly made up by service sectors and, in2005, it comprised Business services, Wholesale trade, Post and telecommunications, Financial intermediation, Real estate, Land and auxiliary transport, Restaurants, Electricity, Construction, Chemicals and Metal products. This relational analysis offers valuable information when trying to promote innovation by considering the system's structural characteristics. Our results imply an advance in our understanding of how the design of public and private interventions would enhance a more efficient diffusion of knowledge and innovation. This includes methods to give priority to selected sectors in innovation programs when a systemic effect is intended and when the purpose is to increase its scope and speed. Interventions trying to foster innovation should, from this perspective, take into account the heterogeneity of economic sectors, in terms of their linkages and of their position in the inter-industry system. Certain linkages are more relevant than others for the diffusion process and the position of sectors conditions the final effect of innovation diffusions. This systemic information can be incorporated into the design and implementation of sectoral and innovation strategies, adding a criterion of rationality to the assignment of resources. Innovation fostered in the core sectors, when taking these as ‘superspreaders’, can flow faster and reach most of the sectors in the inter-industry system with more certainty. Owing to Spain's inter-industrial structure, special attention should be paid to the sectors identified in its core, particularly to Business services, because this appears as the main core sector and because of its capacity to transfer knowledge and innovation through its trade linkages. If this information is taken into consideration, the efficiency of national innovation policies could improve. The targeted sectors of the Spanish innovation policies, according to the European Commission [79], are food, agriculture and fisheries, in the first place, followed by biotechnology activities. Moreover, the several programs designed to foster innovation in Spain direct resources mainly to high technology sectors, measured generally in terms of R&D. The lack of a systemic view, detected in Spain, to select sectors for intervention, seems to be general, according to Jensen et al. [29]: “The tendency among policy makers to think in terms of the linear model of innovation and give priority to supporting R&D-activities in high technology sectors to the neglect of organizational learning and user-driven innovation is problematic. Equally, problematic are policies that give little attention to the strengthening of linkages to sources of codified knowledge for firms operating in traditional manufacturing sectors and services”. A systemic innovation strategy, when combined with the more generalized interventions, would imply considering the relevance of the linkages supported by core sectors with high technology sectors and also with knowledge and innovation institutions. In the case of Spain, this kind of strategy would lead to selecting the Chemical industry as a receiver of particular attention, because it has become increasingly relevant in the core group, it is a high technology content sector with high R&D expenditures and it is closely linked to biotechnology. The proposals raised in this paper could be considered in the selection of sectors and linkages that would receive preferential attention from public innovation policies and, in general, also in the processes of innovation decision making. The present analysis specifically suggests directing efforts towards: 1) The sectors identified in the core and, primarily, towards Business services and Chemical industry. We already know that product innovations, knowledge and learning, and efficiency increases taking place in the most central sectors reach further and faster, in the whole economic system. 2) The linkages that core sectors, and particularly Business services and Chemical industry maintain with other economic activities and with innovation and knowledge institutions. We are aware that policies specifically designed to increase connectivity, particularly with the most central and the most innovative sectors, clearly contribute to the improvement of competitiveness. 3) Sectors with competitive and also innovation problems not located in the core. Those sectors would require acting on their linkages with other sectors and coordinated interventions. As an example, this could be the case of Agriculture and its linkages with the sectors making up the Agro-food system (e.g. Vegetables industries) and with sectors located in the core of the whole system (e.g. Hotels and restaurants and also Chemical industry). The generalizability of our conclusions should be taken cautiously as we used data only for the case of Spain. Nevertheless, those results, and others offered by the network analysis literature, indicate that the core–periphery structure shows a high representative capacity of inter-industrial networks. Presumably, locations in a similar development situation to the Spanish one will show a similar structure. Then, the policy implications we have discussed could be widely useful. With regard to future research, we are already testing those statements and also the existence of a positive relationship between the core–periphery and Modular Scale-free topologies and economic development levels and their dynamics.