1 KNOWLEDGE, LEARNING, AND SECTORIAL AND LOCAL CONTEXTS

Within the scope of neo-Schumpeterian literature, there is consensus that knowledge is essential to establish competitive advantages both for firms and regions, as it is related to the generation of innovative capacities and the skills needed to be on the cutting edge of technological advantage. In this direction, several empirical studies corroborate this understanding. Such studies provide evidence “[...] that the combination of different types of knowledge is positively related to firms’ performance and the transformation of regional paths” (PLECHERO; GRILLITSCH, 2021, p. 3).

In this context, knowledge is viewed as a key asset, and learning as a fundamental element. In other words, while knowledge is viewed as the main resource, learning is seen as the most relevant process (LUNDVALL, 1992). This happens because it is this learning process that generates and integrates specialized knowledge, making it possible to innovate.

The literature describes different types of learning processes, which can be united in two large categories: internal and external learning (MALERBA, 1992). Privileging one or more internal or external learning sources will depend on the firm’s type and size, but in any case, it will involve an ongoing and interactive process influenced by the environment in which the company is inserted (FREEMAN, 1996).

Internal learning is connected to the firm’s main functions – R&D, production, marketing, and organization. Although external learning cannot replace the internal process, it can speed it up or change its direction. The different types of internal learning are the following: learning by using (ROSENBERG, 1982), learning by doing (ARROW, 1962), and learning by searching. External learning comprises the following types: learning by imitating, learning by interacting (LUNDVALL, 1988), and learning by cooperating. Firms are characterized by the different combinations of internal and external learning they implement.

Among these processes, learning by interaction is worth highlighting (LUNDVALL, 1988). This involves forward and backward interactions with sources of knowledge such as suppliers and clients. In the analysis of LPSs, this type of learning is relevant both because of its importance for relations and exchanges among different agents, which is consistent with the proposed investigation based on the analysis of localized production systems, and because it involves a specific relation, the one between producers and clients, a key relation in the footwear machinery industry herein analyzed.

The neo-Schumpeterian literature makes it clear that the processes of innovation and diffusion of new technologies are not independent; on the contrary, they are part of the same process. Thus, innovation involves ongoing and progressive changes, introduced during diffusion, through different learning processes. Hence, the advances during the diffusion process are also the result of clients’ contributions and experiences. The literature acknowledges the importance of both producers and users in the development of technological capacities and demonstrates that the nature and intensity of those interactions, as well as the synergy between producer and customer, are indispensable for a successful innovation process (LUNDVALL, 1988; FREEMAN, 1996).

In addition, taking into account thesectorial specificity of the machinerymanufacturers we are discussing, it is important to underline the process of learning by using, as stressed by Rosenberg (1982) in the literature. For thisauthor, this type of learning is especiallyrelevant in the case of capital goods, as their performance can only be measured after a long period of experience; as well, their characteristics can only be understood after prolonged and intense use.

Because of this understanding, several authors, such as Pavitt (1984); Dosi (1988); Klevorick et al. (1995); Archibugi (2001); Malerba (2005; 2007; 2011); Malerba and Nelson(2011); and Castellacci (2008; 2009) discuss those sectorial differences.

Pavitt (1984), in his classical sectorial taxonomy, identified four main groups of sectors: supplier-dominated, specialized suppliers, scale-intensive, and science-based. To build this classification, he analyzed a set of factors for each sector: dominant technological regime, structural characteristics, origin of the innovation, types of results, forms of appropriation, and possibilities of technological diversification. Pavitt (1984) considers that firms tend to behave in a similar way when they are in the same sector, as productive determinants are similar. Hence, observed intrasectorial routines are similar and allow us to project analyses for a firm to the sector as a whole.

Among the groups proposed by this author, we highlight specialized suppliers, whose attributes help the discussion in this paper. In general, this group includes capital goods and especially machines and equipment. It is characterized by high economies of scope, high supply diversification generally concentrated in medium-sized firms with capacity for product innovation. For the most part, these are enterprises that develop close relations with customers (other firms), have high internal capabilities, and specifically master design and engineering technologies.

In line with discussing the role of sectoral patterns with a view to local systems, Gertler and Wolfe (2010) emphasize that the geography of knowledge flows will also exhibit characteristic patterns by sector. According to them, in those industries where innovation to be driven by need to solve specific problems arising in the interaction with customers and suppliers (such as the development of specialized machinery), tacit knowledge tends to be more important. As a result, given the central role of tacit knowledge and face-to-face interactions with customers and suppliers, they argue that learning and localized knowledge flows are very relevant. Also for Asheim, Grillitsch and Trippl (2017), the knowledge acquired through the application or combination of existing knowledge, typically in interactive learning processes with customers or suppliers, has a strong tacit and context-specific dimension.

Based on a large study of clusters in different regions of Canada, Gertler and Wolfe (2010) propose a typology on the “geography of knowledge flows". From there, they confirm the understanding that in the industrial sectors of capital goods "intra-cluster relations promote the local circulation of knowledge, underpinning the learning process that enables firms to succeed at innovation".

Thus, geographical proximity gains importance as a relevant element in the process of creating and disseminating innovation because when we take into account that the processes of knowledge generation and innovation are interactive and localized, we understand that exchanges among agents in the local space favor and maximize those processes. Furthermore, it is inferred that the specific local institutional framework, through particular learning mechanisms, generates qualitatively different innovation processes (CASSIOLATO; LASTRES, 1999; MORGAN, 2004; MATOS et al., 2017).

This idea is corroborated by the understanding that economic processes are socially rooted, and that most knowledge is created and reproduced through various forms of social interaction (MASKELL; MALMBERG, 1999, p. 171; MASKELL, 2001). Thus, knowledge is socially embedded and the learning process is predominantly interactive (LAM, 1998, p. 4). For this reason, in the framework of evolutionary thinking, innovation should be understood as resulting from this interactive learning process between firms and their environment. Hence, the environment in which the agents are immersed, with its cultural, socioeconomic and political characteristics, interferes with and influences the form in which knowledge is generated and circulated among different actors, in the same way that the possibility of this exchange and knowledge sharing is maximized and stimulated by the physical proximity of those agents.

Therefore, local systems are adequate as units of analysis, as they are considered interactive learning spaces (AROCENA; SUTZ, 2000; LASTRES; CASSIOLATO; MACIEL, 2003; LASTRES et al., 2020). To this extent, we can affirm that local communication channels facilitate knowledge diffusion, making it possible for local enterprises to obtain advantages in innovation performance, as informal networks have an important role in the emergence of production agglomerates and in interactive learning and, as a consequence, in local development (DAHL; PEDERSEN, 2003, p. 2). Likewise, when local potentialities and noncommercial interdependence are emphasized, it is demonstrated that the socio-institutional framework, intercompany communication, and the local learning process play a decisive role in the innovation and growth process (BATHELT; MALMBERG; MASKELL, 2002, p. 10).

2 METHOD

A descriptive and qualitative research was carried out based on the collection of primary data, through two surveys that were carried out in each set of firms in two local systems of manufacturers of footwear production machines. We chose to do a comparative survey, taking into account the economic relevance of the two LPS, both for the productive sector investigated and for the regions where they are located. The first LPS surveyed was Vigevano (Italy) between November 2007 and February 2008; and the second was Vale do Rio dos Sinos (Brazil) between August and November 2008.

The starting point to investigate the Vigevano firms was setting up a databank of footwear machine producers located in that region, based on the register kept by the National Association of Italian Manufacturers of Footwear, Leathergoods, Tannery Machines and Accessories (ASSOMAC). A population of 35 firms was identified and we were able to survey 28 of the firms (80%). The remaining seven did not want to participate in the survey.

Likewise, the starting point to survey the Vale do Rio dos Sinos producers was setting up a databank of footwear machine manufacturers located in that area. We used two information sources: the register of companies affiliated to the Brazilian Association of Machinery and Equipment Industries for Leather, Footwear and Allied Trades (ABRAMEQ) and the catalog of the participants in the main fair in this sector, the “International Fair of Leathergoods, Chemical Products, Components and Accessories, Equipment and Machines for Footwear and Tanneries” (FIMEC) in 2007 and 2008. Twenty-four firms were identified. Respondents totaled 19 firms, i.e., 79% of the population was surveyed.

A structured questionnaire was applied in face-to-face interviews in all investigated firms. The questionnaire was organized in three parts: a) firm's general information; b) information about the performance of R&D activities and the technicians who perform these activities; and c) information about the flows of information and knowledge. The validation of the questionnaire was done in two stages: a) an evaluation by industry experts and academics specialized in the research's field of knowledge; and b) face-to-face application in three firms of each LPS investigated in the exploratory stage of the field research.

The structured questionnaire was designed based on previous studies that had already applied the Social Network Analysis (SNA) method in the investigation of LPS, as Giuliani (2005); and Morrison and Rabellotti (2005). It is important notice that it was used in the questionnaire the roster-recall7tool which comprises the provision of a list of all network members to each network actor (survey respondent), so that the respondent mentions with whom he or she has a relationship. This tool was applied in the third part of the structured questionnaire and with this it was possible to structure the networks that represent the horizontal relations, analyzed in the next section.

In terms of indicators used in network analysis, the most relevant for this study are density (the amount of ties among actors, as the ratio between existing linkages and possible ones) and reciprocity of ties between actors, which were calculated for the analysis of horizontal relationships.

For each interview, the researcher went to the firm's facilities - both in Brazil and Italy - and, in most cases, as a complement to the interview, a visit was made to the factory facilities. Interviews were done with the person responsible for the product development and improvement process in each firm. The interviews lasted an average of 1h30minutes.

After the interviews conducted, the data were tabulated and relevant characteristics of the two groups of firms analyzed were described, such as the frequency of firms by size, by localization, by product specialization, and general characteristics of the R&D activities that the firms carry out. In the case of data regarding information and knowledge flows, the data were tabulated using UCINET software, with which it was possible to generate the networks' representation (figures) and calculate two indicators (density and reciprocity) for the horizontal relations.

3 KNOWLEDGE FLOWS AND THE PROCESS FOR GENERATING TECHNOLOGICAL INNOVATIONS

The data collected in the field research concerning the characteristics of the firms investigated, their R&D activities, and the horizontal relations they establish to seek knowledge are reported in subsection 4.1. The results concerning the vertical relations established by the firms to obtain knowledge and generate innovations are reported in section 4.2. This last section presents, mainly, the results obtained from the use of UCINET to represent the networks.

3.1 INTERNAL STRUCTURE AND HORIZONTAL RELATIONS

Beginning with the group of firms in Vigevano, it's noteworthy that footwear machinery manufacturers are mostly small-scale producers: 82% of the 28 firms interviewed have up to 35 employees. The three largest firms have 200, 93, and 87 employees, respectively.

The size of the firms surveyed in the Vale do Rio dos Sinos region reinforces the characteristic of the footwear machinery manufacturing sector as being mainly formed by small-scale and medium-sized firms. Out of 19 companies, 16 are small-scale and three medium-sized. There is no record of a large company in any of the surveyed areas.

In geographical terms, all the Italian firms surveyed originated in the city of Vigevano where they still remain. Most (22) have productive activities just in Vigevano. Six companies also produce in other places – mainly China and India. The situation is similar for the group of firms in the Vale do Rio dos Sinos region, where all of them originated, developed, and are still operating. Unlike their Italian counterparts, the Brazilian companies surveyed produce exclusively in Vale do Rio dos Sinos.

Regarding the firms’ productive activity, we highlight as a characteristic the specialization in producing certain classes of machines, organized according to the stage of the shoe production process, such as: modeling and cutting, preparing and stitching the upper, assembling and making the sole, among others. In the case of the Italian firms, ASSOMAC has a typology with eight classes of machines and classifies their firms accordingly. The machines produced are standard or customized. The group of 28 firms in Vigevano includes manufacturers of different classes of machines. The majority of the firms are specialized in producing more than one class.

We highlight three factors that characterize the two groups of firms investigated: 1) they are essentially formed by small-scale firms; 2) both groups of firms are rooted in their original localities, in the sense that the majority still produce exclusively where they were born, despite important moves by foreign competition, such as the entrance of China into the market; and finally 3) they are firms specialized in producing classes of machines; this means complementarity of products among firms in the same local system, indicating a priori low intragroup competition.

Concerning the process of generating technological innovations, the vast majority of Vigevano firms stated that they have adequate internal structures to carry out their own research and development activities (R&D). Few firms share this activity or seek structure and knowledge (equipment or technical personnel) from other firms or institutions related to the sector, through formal or informal relations. The vast majority of these structures are informal and are made up of a relatively small number of technicians specialized in the improvement and innovation of machinery – 24 firms reported to have up to five technicians. This is an interesting characteristic of this sector, together with the fact that those technicians do not have university degrees and the majority have been working in the same firm for over ten years.

Most of the Vale do Rio dos Sinos firms stated that they carry out R&D without a formal department. They think their actual structures are adequate to ensure innovation. These characteristics are consistent with the small size of those firms, as well as with the pattern for the group of specialized suppliers, who often explore opportunities to innovate through informal activities of project improvement, usually with little R&D activity.

An important element in the process of generating innovations is the level of formal education of technicians dedicated to R&D. These actors play a central role in identifying clients’ needs and translating them into an improved or new product. They are employed for quite some time by the same firm; in general, they have a fair amount of experience, but not necessarily formal qualifications. It is important to notice that there is little mobility among local and extra-local firms.

In Vigevano, we identified a total of 105 technicians working in the 28 companies surveyed. Seventy-six percent of them have no university degree. A similar situation was found in Vale do Rio dos Sinos, where 12 firms stated they had technicians with university degrees and 18 worked with technical personnel without such a degree. In this latter LPS, the group of technicians is made up of 89 professionals, 74% without a university degree.

This is certainly an indication that tacit knowledge is very important for the development of new products, as knowledge acquisition takes place more through experience in the execution of activities and solution of technological problems (learning by doing) – often in the same firm – than through formal knowledge (university education, for example).

Concerning the types of innovations, it was identified mostly the introduction of incremental product innovations, resulting mainly from demands from the major customers: footwear manufacturers, who have a central role in the production chains in which these productive segments are inserted. Product innovation introduced by the Brazilian firms is focused on facilitating the use of the machine by its operator, instead of automation. On the other hand, product innovation in Vigevano firms generally emphasizes the issue of quality in the shoemaking process, equipment automation to favor greater standardization in the final product, and thus the possibility of production in small batches (typical of this industry in Italy). Italian firms introduced earlier safety items in their machinery.

Before going further in the analysis, it is important to examine that there are four elements that synthesize the generation of innovation by the firms investigated: There are more similarities than differences between the two groups of firms, although they are in localities with distinct characteristics in terms of the quality of the footwear industry final product; the prevailing explanation for this item is the pattern of firm size and sector to which the firms belong. Informal internal structure, in terms of departmentalization and budgetary planning, and a streamlined structure. The central role played by technicians in the process of generating innovations, through their tacit knowledge and experience. Incremental product innovations; but it is important registered that the Vigevano group of firms more frequently develop products with greater aggregate value than the Vale do Rio dos Sinos group.

Finally, concerning the external elements related to knowledge exchange through horizontal relations, i.e., with firms of the same segment, it was observed that in both groups of firms surveyed the networks formed had low density. The characteristics of these relations could be summarized as follows:

- In the case of the group of firms operating in Vigevano and Vale do Rio dos Sinos, knowledge flow densities were identical and, coincidently, very low (4%); which means that only 4% of the possible linkages were present in the network; - Both networks showed null reciprocity in the linkages among actors; - Only a small group of firms participated in the networks. In the case of Vigevano, this group was made up of 13 firms, out of a total of 28; in the case of Vale do Rio dos Sinos, the group had seven firms, out of a total of 19. The explanations for this reality are the following: a) Little competition among firms of the production system does not ensure low market opportunism because of low private knowledge appropriability by those firms. This makes it possible to copy products, even by firms that are not direct competitors at the time, as commented before. In general, firms consider that the risk of this occurring is very high. Hence, locally there is a “potential competition” that inhibits firms from meeting to exchange technological knowledge (although this has been observed in the two groups, it was more evident in the Vale do Rio dos Sinos); b) It was also reported as an explanation for this situation the fact that the firms’ productive specialty has diverse knowledge requirements to solve technological problems; although in the same vicinity they are specialized in different products and knowledge; hence, they would unlikely find the needed technological solutions. Thus, we inferred that geographical proximity among firms was not sufficient to stimulate a dense knowledge flow in horizontal relations; c) An explanation for this situation can also be found in the sectorial logic for generating innovations, as firms reported that the knowledge needed to solve technological problems is found through relations with their clients and suppliers, as described in the next section. This finding also corroborates what is pointed by the literature (Lundvall, 1988, Lundvall et al., 2002; and Asheim, Grillitsch and Trippl, 2017) by contributing to the reinforcement of important regularities that helps to explain the firm's innovation generation process.

3.2 VERTICAL RELATIONS AND THE GENERATING INNOVATIONS

In both agglomerates, the firms highlighted that the main information needed to further product development or improvement processes comes directly from their clients, footwear producers, or from resellers, who communicate with footwear manufacturers. Hence, it is the identification of users’ technological needs – the footwear producers – that drives forward the technological innovation process. Information obtained from specialized suppliers is also relevant and refers especially to the technical characteristics of equipment and materials for the machines.

So, survey results emphasized the importance of vertical relations, as compared to horizontal relations. Knowledge flows based on vertical relations have proved to be an important element in the firms’ external knowledge process, corroborating the sectorial pattern for specialized suppliers (Pavitt, 1984; Malerba, 2005; 2011). The latter are quite dependent on relations with clients and other suppliers, even when there is a strong influence of the “geographical proximity” element for firms in the same production segment. These findings corroborate the literature. As highlighted earlier, both Gertler and Wolfe (2010) and Asheim, Grillitsch, and Trippl (2017) emphasize that knowledge arising from interactive learning processes with customers or suppliers has a strong tacit dimension resulting from local characteristics.

To examine vertical relations, the footwear machinery manufacturers were questioned about their main clients (footwear producers) and suppliers (mechanical and electronic components, and software) with whom they exchange important information and knowledge for the footwear machinery innovation process. Clients and suppliers were classified according to the regions where they are located. Figure 1 presents the vertical relations established by firms in Vigevano.

The node format represents the different types of actors: squares are the firms that responded to the survey, listed from A to AE, triangles represent the suppliers, and diamonds represent the clients. To understand correctly Figure 1, it is important to keep in mind that the representation of clients and suppliers is not necessarily of individual actors, as in the case of respondent firms, but also of a category of information and knowledge sources relevant to the generation of innovation in the surveyed companies. The latter is also characterized by its geographical location, as explained later on. Colors in the charts represent the intensity of the linkages of each node with the others: red nodes have the largest number of linkages, and most of the respondent firms belong to this group; followed by black nodes, where there is only one surveyed firm (P); and finally the blue nodes, representing the firms X, Z, and B that did not report on their relations with clients and suppliers and, thus, are not connected to the network.

Concerning the categories of clients and suppliers and their intensity of linkages in the knowledge flow, we note that the largest number of relations is established among clients and suppliersgeographically close, as follows: suppliers and clients from Vigevano and other Italian regions (FVigIta and CVigIta), as well as suppliers and clients from other Italian regions (F Itália and C Itália). This highlights the relevance of locality and the country where the agglomerate of firms is located, corroborating that the localized knowledge flows are very relevant, as analysed by Gertlher anda Wolfe (2010). Another fact seem from Fig. 1 also corroborates this statement: agents farther located from surveyed firms in geographical terms are the least consulted in the network 12.

It is important to take into account in our analysis information from firms in the agglomerate that are producing footwear machinery in other places in addition to Vigevano because they bring forth elements that help to explain behaviors. The five firms producing machines abroad are the following: A, AB, O, D, and AC – O manufactures in India and all others in China. Out of these five, three (AB, O, and AC) reported they have as important sources of knowledge clients and suppliers from several parts of the world; out of a total 17 linkages (lines) connecting surveyed firms to clients and suppliers outside Italy, nine connected the three above-mentioned firms.

It is important to compare the network of vertical knowledge flows with the network representing horizontal knowledge flows. Figure 2 presents the two networks identified for Vigevano in the field survey, allowing for that comparison.

The representation of the network of horizontal knowledge flows is shown on the left side of Figure 2. We observe that the network is quite sparse, as only a minority of firms participated in it and there are few connections among companies. Blue nodes indicate firms with more connections in the network, red nodes represent firms with fewer connections, and black nodes are the companies that neither seek nor are sought out for the exchange of technological knowledge. This is not the case in the network of vertical relations, where all firms have connections, except for three that did not respond to questions on this topic (X, Z, and B).

Comparing the two networks, we note:

1) Firms have more relations with clients and suppliers (vertical relations) than among themselves (horizontal relations). This finding from the research reported here is relevant because it allows us to make a comparison and conclude that the knowledge flows from horizontal relationships are less utilized by the firms investigated than the knowledge flows from vertical relationships.

2) The production of machinery in countries other than Italy (Vigevano) partly explains the vertical relations established with clients and suppliers from other countries; thus, we understand that there is a relationship between manufacturing machinery outside the home country and the search for sources of external knowledge for the innovation process among clients and suppliers abroad.

3) Another relevant finding is that there is specialization in the search for external knowledge: a large group of firms considers the vertical relations with clients and suppliers from the agglomerate locality and its country as the main source of external knowledge; another group of smaller companies exchanges knowledge through both vertical and horizontal relations, privileging local and national firms in the vertical network (Italian clients and suppliers); and lastly, the third group of firms has vertical relations with local and national clients and suppliers, but also from abroad (in general, these companies do not have horizontal relations in the agglomerate). On this topic, we highlight the low degree of openness of the agglomerate firms, given the few linkages with agents from abroad.

The footwear machinery manufacturers in the Vale do Rio dos Sinos agglomerate were also questioned about their main clients (footwear producers) and suppliers (mechanical and electronic components, and software) with whom they exchange important information and knowledge for their product innovation processes. Figure 3 presents the vertical relations established by these firms.

Formats in Fig. 3 represent the different actors: squares are respondent firms and are listed from 1 to 19. Triangles represent suppliers, and diamonds represent the clients. Similar to Vigevano, the representations of clients and suppliers do not necessarily refer to individual actors, as in the case of respondent firms, but can also represent a category of information and knowledge source, which was also categorized by its geographical location. Colors in the drawings represent the intensity of linkages of each node with the others: red for those with the largest number of linkages, followed by black nodes, grey, and finally pink (firm 2, which provided no information on its relations with clients and suppliers and hence is disconnected from the network).

Figure 3 also provides interesting information regarding clients and suppliers chosen by the Vale do Rio dos Sinos firms. Similar to the research conducted at Vigevano's LPS, we note that there are strong linkages with agents located up to the country’s geographical limit where the agglomerate is located: the most consulted clients and suppliers are represented in red; clients are from Brazil (CliBrasil) and suppliers from Rio Grande do Sul state (ForRS) and Brazil (ForBrasil). Reinforcing the importance of geographical proximity, other clients and suppliers located even closer to the agglomerate were also cited, but with fewer network linkages (black nodes), such as suppliers from Novo Hamburgo (ForNH), clients and suppliers from Vale do Rio dos Sinos (CliVRS and ForVRS), and clients from Rio Grande do Sul state (CliRS). This highlights a pattern of vertical relations whose characteristic element is the geographical proximity to clients and suppliers. Lastly, all network actors outside Brazil are suppliers from Europe or the USA. This raises two issues:

Horizontal relations are represented on the left of Figure 4. We note a quite rarefied network, the same as in Vigevano. Blue nodes are the firms with the largest number of linkages in the network, red nodes have fewer linkages, and black nodes refer to all companies that do not consult with others and are not consulted for knowledge exchange through horizontal relations. We also note that all firms in the network of horizontal relations are also in the network of vertical relations, such as the red nodes (8, 1, and 12) or black nodes (15, 4, 11, and 18), i.e., they are in the two groups with the most intensive relations.

Finally, in terms of vertical relations, Vigevano firms consult mainly with clients and suppliers from Vigevano or Italy in general, while Vale do Rio dos Sinos companies consult with clients and suppliers from Rio Grande do Sul state and Vale do Rio dos Sinos. Thus, we conclude that the two groups of firms obtain external knowledge mainly through interaction with local and regional clients and suppliers, or at the national level at the most.

We also highlight that for the Vale dos Rio dos Sinos network, no other clients beyond the national ones were cited. This situation is different from the one observed in the flow of technological knowledge in the Vigevano network, where other clients, besides Vigevano, were mentioned such as Pavia, Italy, and Europe, located in several parts of the world. Those firms in Vigevano play a role in knowledge exchange because they extended the knowledge flow by including foreign actors

In conclusion, while there is an openness of the firms' internal structures to different sources of knowledge, especially through relations with clients and suppliers, this openness is geographically limited. It is restricted to the main region of the agglomerate (Vigevano or Vale do Rio dos Sinos) or to areas relatively close to them (Italy, Rio Grande do Sul state, and Brazil). Thus, the surveyed generation of innovations has proved to be fairly specific to the firm, as technical knowledge remains compartmentalized within the company through its own technicians, and external knowledge needed to generate innovation is obtained most intensely from clients and suppliers.

4 FINAL CONSIDERATIONS

We can highlight that in comparison to horizontal relations, vertical relations between footwear machinery manufacturers and their clients and suppliers are more frequent in both production systems studied in Brazil and Italy.

This does not mean that the innovation process is not fairly specific to the firm, tied to its internal structure. The knowledge of professional technicians (who often spend their whole productive life in one single company), increasing their knowledge mainly through experimentation (vis-à-vis a process of formal education), has proven to be also important for generating product and process innovations. Among the internal learning mechanisms, the most relevant is learning by doing. Given the size of the companies that characterize these LPSs, mostly small-sized firms, the majority of them do not have formal R&D departments, assigning to technicians the key role in generating innovations. On the whole, it is not part of the firms’ routines to share their structures with other companies working in the same machinery segment or with institutions, either through formal or informal relations.

On the contrary, interactions with actors external to the firms were emphasized in the survey. Thus, respondents focused on informal vertical relations between footwear machinery producers and their clients and suppliers.

Thus, it was mainly through knowledge flows established with clients and suppliers that we verified the importance of geographical proximity between firms for the process of technological change in the sector – vis-à-vis relations with other machinery manufacturers from the same agglomerate or from others (horizontal relations). So, as observed by the literature, proximate partners, as identified in this study, help firms to exploit their existing knowledge by further deepening their understanding in specialized contexts (Balland; Boschma; Frenken, 2022). From this observation, it can also be concluded that learning by interaction is a key type of learning in both LPSs studied, and a central element to explain the firm’s innovative dynamics, generating a dense knowledge flow.

In both agglomerates, the informal participation of technicians from client and supplier firms is considered fundamental for the generation of innovations, as they favor the acquisition of important information and knowledge to build the technical “knowing how to do” and, thus, defines the firms’ innovation capacity. Hence, among external learning processes, learning by interaction was the most emphasized, and also explained the dynamics of generation of innovations in the surveyed firms.

The empirical study carried out, therefore, contributes to show that the two LPS studied use quite traditional knowledge sources for the generation of innovations, corroborating the characteristics highlighted by the literature, either the one that emphasizes the relevance of geographical proximity for knowledge flows in clusters, or the one that emphasizes the sectorial characteristics of proximity with clients and suppliers.

Taking into account these issues, the central contribution of this article is to be found in its empirical results, which demonstrate the following: (i) learning by interacting is the key form of learning in the two agglomerates, especially for the vertical relations, and (ii) geographical proximity is relevant, understood as the geographical closeness between local firms – producers, clients, and suppliers from Vigevano or Vale do Rio dos Sinos – and between footwear machinery manufacturers from the locality and the country where the agglomerates are located.

Considering these remarks, the two LPS analysed can be considered interactive learning spaces (AROCENA; SUTZ, 2000; LASTRES; CASSIOLATO: MACIEL, 2003; LASTRES et al., 2020). However, it is important to highlight that firms in these LPS have distinct roles in the knowledge flows, since the vertical relations among them reported that knowledge is transferred among a small and specific group of firms only. So, firms with distinct innovative dynamics and distinct roles in knowledge flows have been identified, countering the understanding that the firms in productive clusters have homogeneous characteristics and roles, as suggested by the traditional literature on Marshallian industrial. This has already been pointed out by other empirical studies such as Giuliani (2005) and Morrison and Rabelloti (2005).

It is worth mentioning the extent to which these field survey findings are consistent with the characteristics of the sectorial pattern of the segments producing equipment and machinery, identified as specialized suppliers. The reason is that firms with that sectorial pattern tend to be relatively small and operate in close contact with their users/clients, incorporating specialized and tacit knowledge into equipment design and manufacture. Formal R&D activity tends to be of low intensity. In turn, cumulative capacity building, as well as specific demands from users, tend to be important, as they result in innovation appropriability.

As future studies, it is suggested to update the empirical research in the two investigated LPs, so that a longitudinal analysis can be done to identify changes and continuities in the knowledge flows established by the firms. In addition, it would be interesting to advance the discussions in the area by conducting empirical research in LPs from different productive sectors. This would make it possible to analyze regularities concerning the establishment of knowledge flows and learning processes of firms in productive clusters.

References

ABRAMEQ (Associação Brasileira Das Indústrias De Máquinas E Equipamentos). Apresentação do setor de máquinas e equipamentos para couro, calçados e afins. Novo Hamburgo, mimeo, 2001.

ÁLVAREZ-CASTAÑÓN, L.d.C. Footwear Cluster: A History of Entrepreneurship in Crisis. In: ALVARADO, A.A.; MONTIEL,Méndez, O.J. (Ed.). The History of Entrepreneurship in Mexico. Emerald Publishing Limited. Bingley. pp. 173-189. 2020. https://doi.org/10.1108/978-1-83909-171-120201005

ARCHIBUGI, D. Pavitt's Taxonomy Sixteen Years on: A Review Article. Econ. Innov. New Techn, v.10, n. 5, p. 415–425, 2001.

AROCENA, R.; SUTZ, J. (2000). Interactive learning spaces and development policies in Latin America. Copenhagen, DK: Druid ́s Summer Conference on the Learning Economy Rebild, 2000, pp. 15-17. Available on: http://www.druid.dk/.

ARROW, K. The economic implications of learning by doing. Review of Economic Studies, n. 29, p. 155-173, 1962.

ASSOMAC (Associazione Nazionale Costruttori Macchine Dell ́Area Pelle), Congiuntura 2006, Vigevano, 2007.

ASHEIM, B.; GRILLITSCH, M.; TRIPPL, M. Introduction: Combinatorial Knowledge Bases, Regional Innovation, and Development Dynamics. Economic Geography, 93 (5), p. 429-435, 2017. DOI: 10.1080/00130095.2017.1380775

BALLAND, P.-A.; BOSCHMA, R.; FRENKEN, K. Proximity, innovation and networks: A concise review and some next steps. In.: TORRE, André; GALLAUD, Delphine (Eds.). Handbook of Proximity Relations. Elgaronline, 2022. p. 70-80. DOI: 10.4337/9781786434784.

BATHELT, H.; MALMBERG, A.; MASKELL, P. Clusters and knowledge: Local Buzz, Global Pipelines and the Process of Knowledge Creation. DRUID Working Paper, p. 02-12, 2002. Available on: http://www.druid.dk.

BOSCHMA, R. Proximity and innovation – a critical assessment. Regional Studies, v. 39, n. 1, pp. 61-74, 2005.

BOSCHMA, R.; FRENKEN, K. Evolutionary Economic Geography. In: CLARK, Gordon L.; FELDAN, Maryann P.; GERTLER, Meric S.; WÓJCIK, Dariusz. The New Oxford Handbook of Economic Geography, p. 213-229, 2018. DOI: 10.1093/oxfordhb/9780198755609.013.11

CASSIOLATO, J. E.; LASTRES, H. M.M. (eds.). Globalização e inovação localizada. Brasília, DF: IBICT/MCT, 1999.

CASTELLACCI, F. Technological paradigms, regimes and trajectories: Manufacturing and service industries in a new taxonomy of sectoral patterns of innovation. Research Policy, v. 37, n. 6-7, pp. 978–994, 2008.

CASTELLACCI, F. The interactions between national systems and sectoral patterns of innovation: A cross-country analysis of Pavitt’s taxonomy. Journal of Evolutionary Economics, v. 19, issue 3, p. 321-347, 2009.

DAHL, M.S.; PEDERSEN, C.O.R. Knowledge flows through informal contacts in industrial clusters: myths or realities? DRUID Working Paper, n. 03-01, 2003. Available on: http://www.druid.dk/.

DOSI, G. ET AL. (Eds.). Technical change and economic theory. Londres: Pinter, 1988.

FREEMAN, C. Innovation and growth. In: DODGSON, M.; ROTHWELL, R. (eds.). The handbook of industrial innovation. Cheltenham: Edward Elgar, 1996, p. 78-93.

GARCIA, R.; OLIVEIRA, A. DE; MADEIRA, P. Bens Salário, Documento setorial: têxtil, vestuário e calçados. In: KUPFER, D. et al. (Eds.). Perspectiva do Investimento no Brasil. São Paulo: Instituto de Economia da UFRJ – Instituto de Economia da UNICAMP, 2010. Available in: https://www.eco.unicamp.br/neit/pesquisas/145-menu-principal/484-perspectivas-do-investimento-no-brasil.

GERTLER, M. S.; WOLFE, D. A. Spaces of Knowledge Flows: Clusters in a Global Context. In: ASHEIM, B.; COOKE, P.; MARTIN, R. (Eds.) Clusters and Regional Development: Critical Reflections and Explorations. London: Routledge, 2010.

GIULIANI, E.; BELL, M. The Micro-Determinants of Meso-Level Learning and Innovation: evidence from a Chilean wine cluster. Research Policy, v. 34, n. 1, p. 47-68, 2005.

IBGE - Instituto Brasileiro de Geografia e Estatística. Pesquisa de Inovação Tecnológica. Rio de Janeiro: IBGE, 2005.

ISTAT - Istituto Nazionale di Statistica. 2001. Available on: http://www.istat.i/it.

KLEVORICK, A. K. et al. On the sources and significance of interindustry differences in technological opportunities. Research Policy, v. 24, p. 185-205, 1995.

LAM, A. (1998) Tacit Knowledge, Organisational Learning and Innovation: A Societal Perspective, DRUID Working Paper, n. 98-122, 1998. Available on: http://www.druid.dk/.

LASTRES, H. M. M.; CASSIOLATO, J. E.; MACIEL, M. L. Systems of innovation for development in the knowledge era. In: CASSIOLATO, J.E.; LASTRES, H.M.M.; MACIEL,M.L. (eds.). Systems of innovation and development: evidence from Brazil. Cheltenham, United Kingdom: Edward Elgar, 2003. In: SUAREZ, D.; ERBES, A.; BARLETTA, F. (Comp.). Teoría de la innovación: evolución, tendencias y desafíos: herramientas conceptuales para la enseñanza y el aprendizaje. Los Polvorines: Universidad Nacional de General Sarmiento; Madrid: EdicionesComplutense, p. 477-508, 2020

LUNDVALL, B-Å. Innovation as an interactive process: from user-producer interaction to the national system of innovation. In: DOSI, G. ET AL. (eds.). Technical change and economic theory. London: Pinter, p. 349-369, 1988.

LUNDVALL, B-Å. (ed.) National innovation systems: towards a theory of innovation and interactive learning. London: Pinter, 1992.

LUNDVALL, B-Å. et al. National systems of production, innovation and competence building. Research Policy, n. 31, p. 213-231, 2002.

MALERBA, F. Learning by firms and incremental technical change. The Economic Journal, v. 102, n. 413, pp. 845-859, 1992.

MALERBA, F. Sectoral systems of innovation: a framework for linking innovation to the knowledge base, structure and dynamics of sectors. Economics of Innovation and New Technology, v. 14, n. 1-2, p. 63-82, 2005.

MALERBA, F. Sectoral systems: how and why innovation differs across sectors. In: FAGERGERG, J.; MOWERY, D.; NELSON, R. (eds), The Oxford Handbook of Innovation. Oxford: Oxford University Press, p. 380-406, 2011.

MALERBA, F.; NELSON, R. Learning and catching up in different sectoral systems: evidence from six industries. Industrial and Corporate Change, v. 20, n. 6, p. 1645-1675, 2011.

MASKELL, P. Towards a knowledge-based theory of the geographical cluster. Industrial and Corporate Change, v. 10, n. 4, p. 921-943, 2001.

MASKELL, P.; MALMBERG, A. Localised learning and industrial competitiveness. Cambridge Journal of Economics, v. 23, n. 2, p. 167-185, 1999.

MATOS, M. P. et al. (Org.). Arranjos Produtivos Locais: referencial, experiências e políticas em 20 anos da RedeSist. Rio de Janeiro: e-papers, 2017.

MORGAN, K. The exaggerated death of geography: learning, proximity and territorial innovation systems. Journal of Economic Geography, v. 4, n. 1, p. 3-21, 2004.

MORRISON, Andrea; RABELLOTTI, Roberta. Inside the Black Box of “Industrial Atmosphere”: knowledge and information networks in an Italian wine local system. Quaderno n° 97, Dipartimento di Scienze Economiche e Metodi Quantitativi di Università del Piemonte Orientale, July, 2005.

PAVITT, K. Sectorial patterns of technical change: towards a taxonomy and a theory. Research Policy, v. 13, n. 6, p. 343-737, 1984.

PLECHERO, M.; GRILLITSCH, M. Stepping-up innovation in manufacturing firms: Knowledge combinations in an Italian local production system. Papers in Innovation Studies n. 2021/05. Centre for Innovation Research, Lund University, Sweden, 2021.

ROSENBERG, N. Lerning by using. In: Rosenberg, N., Inside de black box – technology and economics. Cambridge University Press, 1982, p. 120-140.

VILLAMIL RAMÍREZ, B. A. Proximity dimensions effects on innovation of footwear firms in Local Production Systems (LPS): case studies in Brazil and Colombia. 2019. Tese Universidade de São Paulo, São Paulo, 2019. Disponível em: http://www.teses.usp.br/teses/disponiveis/3/3136/tde-17072019-103323/pt-br.php.

IR