Does a cluster promote innovation and productivity in its firms?¹

ram RAM. Revista de Administração Mackenzie RAM, Rev. Adm. Mackenzie 1678-6971 Editora Mackenzie; Universidade Presbiteriana Mackenzie Resumo Objetivo: A acirrada competição em um mercado globalizado obriga as empresas a adotarem estratégias inovadoras para obter vantagem competitiva sobre seus concorrentes. No entanto, inovar é uma tarefa difícil de realizar isoladamente. Uma forma de amenizar esse isolamento é separar as organizações em clusters setoriais, facilitando a busca por inovação e produtividade. Nessa perspectiva, esta pesquisa tem como objetivo comparar a relação entre o desempenho de inovação e a produtividade das empresas brasileiras e, especificamente, se elas estão inseridas em clusters, considerando suas estratégias de pesquisa e desenvolvimento (P&D) interno e externo implementadas. Originalidade/valor: Este artigo expande a compreensão da literatura sobre inovação aberta, testando o papel complementar do P&D interno e externo para a implementação de inovação em empresas usando uma condição contextual relevante: sua presença ou não em um cluster. Assim, o estudo ajuda a aumentar a compreensão de uma série de mecanismos pelos quais o cluster ajuda a promover a inovação e a produtividade. Design/metodologia/abordagem: Para testar as hipóteses, o estudo utilizou modelagem de equações estruturais multigrupo em uma amostra de 5.581 empresas, com 1.878 participantes de um cluster e 3.703 não participantes. Resultados: Os resultados suportam um impacto positivo de P&D externo e interno no desempenho da inovação e sustentam a noção de que o desempenho da inovação, independentemente de a empresa fazer ou não parte de um cluster, impacta positivamente na sua produtividade. Por outro lado, P&D externo e interno mostraram-se complementares apenas para empresas dentro de um cluster. INTRODUCTION The effectiveness of industrial clusters as promoters of innovation and performance for firms has recently been questioned because the digital transformation has made the diffusion of knowledge much more accessible over long distances. Porter (1998) highlighted how easy it had become for firms to obtain resources, such as capital, technology, and inputs, among others, with just one mouse click. Besides, globalization implies that most firms can compete for consumers worldwide, regardless of the physical distance (Levitt, 1993). However, successful clusters do not let the literature neglect the strategic importance of location and proximity for some industries. Some of these examples are the leather industry in Italy; the wine cluster in California (Porter, 1998) and textile companies in North and South Carolina, in the United States; and automotive companies in south Germany (Levitt, 1993). What do these success stories have in common? Are location and proximity factors that influence the achievement of a competitive advantage by these firms in a globalized market? According to Porter (1998), a cluster is the geographic concentration of companies from the same industry, related industries, and other support institutions (for example, research centers, universities, and government support agencies). Aside from being physically close to one another, these organizations have similar goals and participate in a system that allows competition and simultaneous cooperation (Nalebluff et al., 1996) in search of the development of productive, innovative, and commercial capacities to stand out in national and global markets. Many successful and well-known clusters exist. Silicon Valley is known worldwide for the generation of new technologies and counts on the presence of major companies, such as Google and Apple (Souza, 2019). Among its most striking features, the Boston biotechnology cluster has more than 80 academic centers to support research and provide the necessary human resources for the firms installed in that location (Cluster Mapping, 2019). Beyond the borders of the United States, we will mention three additional technology clusters: London, Israel, and, more recently, Paris, the latter of which started in 2013 (Ragalado, 2013). Firms in clusters can achieve competitive advantage by participating in an innovative environment, fostering their innovativeness (Mascena et al., 2013). Developing innovation activities is an arduous task for organizations. It usually requires high investments, high-risk activities, and increased management costs to monitor activities and measure their results. Collaboration strategies are alternatives to face these challenges (Schilling, 2017). They are boosted by being situated in a cluster environment that facilitates proximity to research institutes, universities, competitors, suppliers, specialized labor, and other relevant knowledge sources for innovation (Porter, 2000). This proximity increases specialization and quality (Porter, 2000), formal and informal exchanges (Ozer & Zhang, 2015) that expedite knowledge transfer (Eisingerich et al., 2010; Guo et al., 2020), and trust among participants (Ozer & Zhang, 2015). Silicon Valley, for instance, promotes a strong relationship network between participants (Saxenian, 1994). It encourages the necessary opening to discuss the problems common to all stakeholders and seeks joint solutions, allowing the community’s interest to surpass the individual interests of each firm (Saxenian, 1994). In China, technological learning and participants’ knowledge spillovers boost the emergence of successful clusters (Guo & Guo, 2011; Guo et al., 2020). These arguments indicate that participation in clusters is still a condition that improves firms’ innovativeness and competitive advantage. The literature has explored the relationship between firms’ participation in clusters and innovation. Several authors state that clusters incentive the development of internal R&D capabilities (Eisingerich et al., 2010; Porter, 1998) and, consequently, provide more innovative firms (Bittencourt et al., 2019; Lai et al., 2014), which, usually, achieve higher productivity and performance (Greco et al., 2021; Griffith et al., 2006; Paula & Silva, 2019). Some of the reasons are related to resource and knowledge trades between competitors stimulated by the proximity (Castro, 2015; Ozer & Zhang, 2015), evidencing how such arrangements promote open innovation (Chesbrough, 2003). However, other studies focus on some problems caused by clusters - such as excessive resource redundancy (Boschma, 2005; Presutti et al., 2019) and misappropriation of innovation by firms other than the innovator, caused by unintended knowledge spillovers (Cassiman & Veugelers, 2006) - to justify that they can be detrimental to innovative efforts. Despite this divergence in academia, there is a lack of empirical studies comparing the relationship between open innovation activities, innovativeness, and performance of firms participants and non-participants of clusters. This study intends to fill this gap by answering the following research question: to what degree does participation in an industrial cluster influence the relationship between a firm’s innovation activities (especially R&D), innovation performance (IP), and productivity? The article’s contribution is twofold. Regarding the innovation management theory, it extends the understanding of open innovation by testing some of its mainstream theories using a new contextual condition: the firm’s presence in a cluster. These theories are based on the strategic alliances’ literature (Gulati, 1998; Powell et al., 1996; Zaheer et al., 2010), which discusses the positive influence of external knowledge sources for collaborating with different partners on the innovation outputs (Hagedoorn, 1993; Hagedoorn & Wang, 2012; Oerlemans et al., 2013) and on the absorptive capacity literature (Cohen & Levinthal, 1990; Zahra & George, 2002), which reinforces the existence of complementarity between external knowledge (from strategic alliances) and internal knowledge (mainly from R&D) (Cassiman & Veugelers, 2006; Paula & Silva, 2018a). For the cluster literature, in turn, the contribution is that, beyond the already recognized influence of the presence in clusters to improve innovation (Eisingerich et al., 2010; Ozer & Zhang, 2015; Mascena et al., 2013) and financial performance (Nalebluff et al., 1996; Newlands, 2003; Porter, 1998, 2000; Saxenian, 1994), this study helps to understand specific mechanisms through which the cluster helps to promote innovation and productivity. The structure of the article is as follows: it starts with a literature review, presenting the concepts and theories concerning open innovation activities, IP, productivity, and the influence of the cluster, proposing the theoretical model and hypotheses to answer the research question. Next, the method section describes the data source and the constructs’ rationale and explains the statistical methods used. Then, we present the results, followed by the discussion section, and, finally, the conclusions, including the study’s contributions, limitations, and suggestions for future research. LITERATURE REVIEW Internal and external R&D and innovation performance A firm’s R&D can be carried out within walls (also called internal R&D) or outside its boundaries by its partners, technology suppliers, or absorbed from knowledge spillovers from other organizations (all of which are called external R&D). The contemporary open innovation approach (Chesbrough, 2003) presupposes that integrating both internal and external R&D is fundamental to succeed (Cassiman & Veugelers, 2006; Paula & Silva, 2018a). According to Berchicci (2013), external R&D refers to an organization’s exposure to external partners, allowing the outsourcing of R&D projects and technologies. It can take different forms, including contracts, strategic alliances with other firms, partnerships with universities, among others (Fey & Birkinshaw, 2005). External R&D has a substantial impact on IP. In accordance with Chesbrough (2003), a firm would not innovate, whatever the intended type is, without using external knowledge. Currently, the technologies are complex and demand multiple partnerships with different types of partners that have the complementary resources needed to succeed in the innovation process (Cobeña et al., 2017, 2019). Among the main benefits expected by firms that perform external R&D activities, there is diluting the risks and costs of R&D activities, facilitating the transfer of tacit or codified knowledge from partners, and having access to complementary assets and capacities that are hard to develop internally (Faems et al., 2005). Internal R&D, in turn, is a set of learning-oriented activities conducted within walls (i.e., the development of scientific knowledge, learn-by-doing, R&D training) that accumulate knowledge and technological capabilities (Bell & Figueiredo, 2012; Malerba, 1992), which are the basis of innovation. Furthermore, it can also produce the necessary expertise to better take advantage of the innovation opportunities outside the organization’s boundaries (Hagedoorn & Wang, 2012). This capability is referred to in the literature as absorptive capacity (AC) (Cohen & Levinthal, 1990; Zahra & George, 2002). Bell and Figueiredo (2012) discussed the rationale for developing AC, stating that a complementary relationship between internal and external learning activities is based on two steps: in an ex-ante moment, internal learning efforts are necessary to form the knowledge base required to acquire external knowledge. Subsequently, similar efforts are needed to ensure effectiveness in absorbing the knowledge gained outside the walls. According to this rationale, internal R&D investments are necessary to improve the firm’s AC. Reinforcing these arguments, Kim (1997) stated that AC consists of two dimensions: the accumulated internal knowledge and the continuous learning effort (Kim, 1997). Developing these two items increases the firm’s capacity to identify, assimilate, and apply external knowledge (Cohen & Levinthal, 1990). Several studies have used investments in internal R&D as a proxy for AC, mainly reflecting the continuous learning effort dimension (Belussi et al., 2010; Berchicci, 2013; Cohen & Levinthal 1990). Regarding the dimension of accumulated internal knowledge, some proxies are employees’ level of education (Berchicci, 2013; Paula & Silva, 2018a) and patent stock (Dushnitsky & Lenox, 2005a, 2005b; Vanhaverbeke et al., 2015). Innovation performance and productivity According to Rattner (1967), productivity is the relationship between inputs and outputs and relates from the firm level to the level of the global economy. Increasing productivity means increasing the value added to a firm’s production resources, which is closely associated with the growth of the monetary production of a worker (King et al., 2014). Several studies have investigated the relationship between R&D efforts, IP, and productivity at the firm level (Griffith et al., 2006; Paula & Silva, 2019). Innovation, especially open innovation, is crucial for improving organizational performance due to a direct and measurable increase in productivity (Greco et al., 2021). Process innovation is the first innovation type commonly related to improvements in productivity (Paula & Silva, 2019). In many cases, it lowers the product or service costs by improving the productive process (Moutinho et al., 2015). According to Findik and Beyhan (2017), process innovation improves production reliability and quality, adding value to customers. Additionally, it increases production flexibility and capacity, reduces labor costs and improves work health and safety conditions. Similarly, product innovation generates the need to adapt the production process to new product lines (Utterback & Abernathy, 1975). Corroborating these ideas, Dabla-Norris et al. (2012) identified that firms that innovated by changing their product line or by adopting a new technology managed to increase their productivity by 8% when compared to firms that did not innovate. Moreover, Aldieri et al. (2021) found evidence that all types of innovation (product, process, organizational, and marketing), individually and conjointly, improved productivity in Italian manufacturing firms. The influence of the cluster on a firm’s open innovation activities The innovation management literature recognizes that clusters are influential factors in firms’ innovative capacity. Inhan et al. (2013) described several types of innovation that entrepreneurs stated they developed due to their insertion in a cluster. Among the most important ones, it was the launch of new products and innovations in their production processes. This innovativeness happened because firms that are part of clusters tend to have better conditions to develop their internal R&D capabilities, as they usually have access to more specialized and productive employees (Porter, 1998), which leads to the opportunity to attend the best universities and work in the best companies in a competitive environment and having the chance to be more productive. Local customers tend to be more sophisticated and demanding (Porter, 2000), naturally selecting more innovative firms to survive in this environment. Besides more advanced internal innovative capabilities, these firms benefit from more fluent knowledge flows inside clusters (Eisingerich et al., 2010), both intentional and non-intentional. The latter of these is called knowledge spillovers (Cardamone, 2018; Cassiman & Veugelers, 2002). Deliberate knowledge exchange between firms is facilitated by their proximity, according to Ozer and Zhang (2015). The authors stated that firms in a cluster tend to develop a common identity and a sense of belonging that encourages participation in industry events and the development and sharing of common tools, language, and business standards. The cluster also generates trust and reciprocity that facilitates information sharing and improves innovation. Furthermore, Castro (2015) highlighted that combining complementary knowledge of different actors is a stimulus to innovation in clusters. According to this study, innovation in clusters is also the result of the dynamics of peer interaction, which enables collaboration even in diverse and ambiguous environments. Per this argument, Bittencourt et al. (2019) stated that clusters generate more innovative companies. According to Lai et al. (2014), clusters influence IP in daily activities by bringing together similar sectors, attracting talents, and promoting information sharing that improves operational performance. These organizational formats also foster interaction between schools, the government, and upstream and downstream companies, generating knowledge and promoting innovation. Proximity, shared goals, and a sense of belonging encourage more competition in the network of institutions inside a cluster (Nalebluff et al., 1996; Ozer & Zhang, 2015; Saxenian, 1994). Regarding non-intentional knowledge spillovers, firms in clusters can watch rivals more closely and learn about new product features, designs, and marketing efforts (Ozer & Zhang, 2015). The proximity this arrangement promotes allows knowledge spillovers as a result of the increase in information circulation (Acs et al., 2017), employees’ job mobility (Fernandes & Ferreira, 2013), and leaked knowledge, due to collaboration with cluster partners (Cassiman & Veugelers, 2002). The effect of these knowledge spillovers in clusters is associated with firms’ positive IP (Cardamone, 2018). In addition to knowledge exchanges, geographical proximity can facilitate the development of joint projects to share costs and risks, especially for small and medium-sized enterprises (SMEs), which usually lack resources (Kapetaniou & Lee, 2019). However, some studies present a different point of view. Although there are clear advantages of the geographical proximity promoted by the clusters, there are some negative aspects. Too much geographical proximity may provoke an excessive redundancy of resources from too similar firms, causing lock-in problems (Boschma, 2005; Presutti et al., 2019). On the contrary, partnerships with firms from other places, especially other countries, can add value by allowing access to new markets, with the local partner helping to adapt products and services to local demands (Beers & Zand 2014; Garcia Martinez et al., 2017, 2019; Zhang et al., 2020). Geographically distant partners also bring more diverse knowledge that is possibly unavailable in the firm’s location (Ardito et al., 2019). Considering the contradictory arguments in favor and against proximity, some authors advocate that the relationship between partners’ geographical proximity has an inverted-U shape (Boschma, 2005; Leeuw et al., 2014). In the lower levels, the lock-in effect happens; in the higher levels, the firms are so distant that the costs of management and coordination of this relationship are too high. Alternatively, an average level of proximity would be ideal. Consequently, innovativeness is no guarantee of improvement for firms in a cluster. It is contingent on several variables, one of the most important being the firm’s AC (Beers & Zand, 2014; Kapetaniou & Lee, 2019; Leeuw et al., 2014; Presutti et al., 2019). As it corresponds to the ability of the firm to identify, absorb, use, and exploit external knowledge (Cohen & Levinthal, 1990; Zahra & George, 2002), it defines how effectively the knowledge flows from external sources to the firm both in the case of intentional cooperation and knowledge spillovers (Aldieri et al., 2018). Giuliani and Bell (2005), studying the knowledge flows inside a Chilean wine cluster, found that firms with a higher AC level can develop more links to exchange knowledge with other firms from the cluster. According to several authors, AC is strongly related to the firm’s intensity of efforts and accumulated knowledge in R&D (Cohen & Levinthal, 1990; Kim, 1997). This correlation between internal R&D and AC indicates a complementarity of internal and external sources of knowledge (Cassiman & Veugelers, 2006; Paula & Silva, 2018a). Some of these arguments reveal that firms in clusters have better conditions to develop internal R&D, which would increase their AC. In parallel, they would be more exposed to external knowledge than firms that are not part of clusters, indicating a stronger positive relationship between external R&D, internal R&D, and IP. Other arguments, however, contradict these affirmatives by considering that geographical proximity is detrimental to innovation. Nevertheless, one consensus is that the AC helps to improve IP in this context. Considering the arguments just mentioned, this study defends the positive influence of clusters in the relationship between R&D and IP. Thus, we propose the following hypotheses: H1: The positive influence of a firm’s external R&D on its IP is more intense for clustered firms than for the non-clustered ones. H2: The positive influence of a firm’s internal R&D on its IP is more intense for clustered firms than for the non-clustered ones. H3: A firm’s internal R&D positively moderates the impact of external R&D on IP more intensely in clustered firms than in non-clustered ones. Similarly, clustered firms should feel the effect of innovation on productivity more strongly. Gaining competitive advantages usually increases productivity as it means an increase in revenues most of the time. Many authors suggested that clusters are sources of competitive advantage (Perry, 2005; Porter, 1990, 1998; Powell, 1987; Schmitz & Nadvi, 1999). There are several possible explanations for this relationship. Firms inside clusters tend to cooperate more, providing complementary resources and capabilities that bring competitive advantages. Additionally, the proximity increases the interfirm trust, which lowers transaction costs (Newlands, 2003). Considering these arguments, we formulate the following hypothesis. H4: The positive influence of a firm’s IP on its productivity is more intense for clustered firms than for the non-clustered ones. Figure 1 presents the conceptual model proposed in this paper, which contains all the hypotheses. Figure 1 Conceptual model Source: Elaborated by the authors. METHOD Data source and sample This study used data from the Brazilian Technological Innovation Survey (Pesquisa de Inovação Tecnológica - Pintec) of 2014 (Instituto Brasileiro de Geografia e Estatística - IBGE, 2019a) and the Annual Industrial Survey (Pesquisa Industrial Annual - PIA) of 2014, 2015, and 2016 (IBGE, 2019b). Pintec 2014 investigated R&D, innovation activities and outcomes of Bra­zilian firms in the period between 2012 and 2014. Productivity was calculated using variable indicators from PIA 2014, 2015, and 2016. It enabled us to have data from productivity with a two-year lag compared to innovation and R&D data. We used this lag because previous studies indicated that some time is needed for R&D and innovation activities to influence IP, as inventions created by combinations of internal and external knowledge achieved by R&D activities need to mature to become new products and services ready to reach the market with success (Paula & Silva, 2018b). To identify if the firm participated in a cluster or not, we used the employee numbers data from firms and cities from the Annual List of Social Information (Relação Anual de Informações Sociais - Rais) of 2019 (Ministério do Trabalho, 2020) to calculate a dummy variable described in the next section. The sample consisted of 5,581 Brazilian manufacturing firms that introduced product or process innovation from 2012 to 2014, had abandoned or suspended projects during this period or had an ongoing innovation project by the end of 2014. The sample was split into two groups: cluster, which aggregated 1,878 firms that participated in an industrial cluster, according to a method shown in the next section, adapted from Britto and Albuquerque (2000); and non-cluster, which aggregated 3,703 firms that were not part of a cluster. Description of the variables The model tested in the study was composed of external R&D, internal R&D, IP, productivity, and a moderation variable cluster, which will split the sample into two groups to train a multigroup structural equation modeling (SEM). Table 1 illustrates the variable definitions, calculations, and sources. Table 1 Variables list Construct Proxy Format Year Source Moderation variable Cluster 0 - no; 1 - yes. 2014 Rais 2019 (Ministério do Trabalho, 2020) Internal R&D PercSpenndR&D R&D expenses/total revenues. 2014 Pintec 2014 (IBGE, 2019a) PercSpendTrain Training expenses/total revenues. 2014 Pintec 2014 (IBGE, 2019a) PercEmpR&D Employees in R&D/total employees. 2014 Pintec 2014 (IBGE, 2019a) EducR&D (No. PhD * 3 + No. Msc * 2 + No. graduated) /total employees. 2014 Pintec 2014 (IBGE, 2019a) Innovation performance (IP) IProd 0 - no; 1 - yes. 2014 Pintec 2014 (IBGE, 2019a) Innovation performance (IP) InnovExt 0 - it did not introduce innovation; 1 - innovation to the firm; 2 - innovation to the country; 3 - innovation to the world. 2014 Pintec 2014 (IBGE, 2019a) InnovRad 0 - it did not introduce innovation; 1 - incremental innovation; 2 - radical innovaton. 2014 Pintec 2014 (IBGE, 2019a) PercRevInnov Sales of new products/total sales. 2014 Pintec 2014 (IBGE, 2019a) External R&D Cli (importance) 0 - not used; 1 - low; 2 - medium; 3 - high. 2014 Pintec 2014 (IBGE, 2019a) Supp (importance) 0 - not used; 1 - low; 2 - medium; 3 - high. 2014 Pintec 2014 (IBGE, 2019a) Comp (importance) 0 - not used; 1 - low; 2 - medium; 3 - high. 2014 Pintec 2014 (IBGE, 2019a) Consult (importance) 0 - not used; 1 - low; 2 - medium; 3 - high. 2014 Pintec 2014 (IBGE, 2019a) Univ (importance) 0 - not used; 1 - low; 2 - medium; 3 - high. 2014 Pintec 2014 (IBGE, 2019a) Productivity Productivity Average of 2014, 2015, and 2016’s total revenues/total employees. 2014, 2015, and 2016 PIA 2014, 2015, and 2016 (IBGE, 2019b) Source: Elaborated by the authors. External R&D and IP are reflective constructs. The model represents them as latent variables reflected by their proxies. Internal R&D, which also represents AC, was considered formative. According to Kim (1997), two conditions are necessary to AC, forming and not reflecting this concept: learning efforts and accumulated knowledge. Variables such as PercSpenndR&D and PercSpendTrain represent the learning efforts, while employees’ accumulated knowledge, EducR&D and PercEmpR&D, represent the accumulated knowledge. As SEM is not appropriate to deal with formative constructs, we proposed the identification of perpendicular factors in the space of the proxies of AC and calculated their resulting vector, which is the variable of AC: internal R&D. We conducted an exploratory factor analysis (EFA) with varimax rotation (Hair et al., 2006) that resulted in four factors, each weighted by one of the four proxies, supporting the notion that the four variables are approximately orthogonal. Therefore, we calculated the indicator of Internal R&D as the square root of the sum of the four squared variables. The moderation variable cluster calculates according to a method adapted from Britto and Albuquerque (2000), which consists of the following formulas (all data to calculate the formula is from RAIS 2019): (1)Q= participation of sector i in city j participation of sector i in Brazil If Q > 1 and number of firms > 10, cluster = 1; otherwise, cluster = 0. (02)Participation of sector i in region j= total employees of sector i in city j total employees in section j (03)Participation of sector i in region Brazil= total employees of sector i in Brazil total employees in Brazil Statistical method As the database came from different surveys, common-method bias was not an issue. In the first step, we conducted a multigroup confirmatory factor analysis (CFA) with both the latent variables (external R&D and IP), forcing all regression weights to be the same in both groups (cluster and non-cluster) to validate the measurement model. Then, we conducted a multigroup SEM to test the hypotheses. Minimum acceptable fit measures for both were a comparative fit index (CFI) higher than 0.95 and a root mean square error of approximation (RMSEA) lower than 0.07 (Hair et al., 2006). We used the mean-centering technique (Little et al., 2006) to represent the moderation of AC (internal R&D) on the relationship between external R&D and IP. This technique suggests creating a new construct representing the interaction between the indicator of internal R&D and the construct external R&D, which loaded the IP construct. The proxies of this construct are the products of internal R&D and all the factors of external R&D after applying the Z-score. The residuals of all proxies correlate in the SEM. RESULTS The descriptive statistics of the variables used in the model and others for both groups are presented in Table 2. Analysis of variance (Anova) was used to test average differences between groups and did not identify any significant difference for all variables. IP from clustered firms is higher than for the non-clustered counterparts. Considering the sample, 58% of the firms outside clusters and 68% in clusters introduced product innovation between 2012 and 2014. Besides this, clustered firms produce more radical innovation (1.01 versus 0.87) and more innovation to the market or to the world (1.02 versus 0.80). These data reinforce the assumption that firms that participate in clusters have better conditions to innovate. Table 2 Descriptive statistics Non-cluster(n = 3,703) Cluster (n = 1,878) Variable Description Average Standard deviation (SD) Average SD IProd The firm introduced product innovation. 0.58 0.49 0.68 0.47 InnovExt Innovation to the firm, to the market, or to the world. 0.80 0.81 1.02 0.89 InnovRad Innovation is incremental or radical. 0.87 0.83 1.01 0.81 SpendR&D(BRL x 1,000) Total R&D expenses. 1,213.52 9,269.98 5,968.24 66,238.18 SpendTrain(BRL x 1,000) Total training expenses. 53.90 869.16 105.31 1,321.02 TotalEmpR&D Total R&D employees. 5.35 25.53 19.51 136.02 PercEmpR&D Percentage of employees working in R&D activities. 1.49% 4.77% 1.78% 4.45% PercSpendR&D Percentage of revenue invested in R&D. 0.79% 6.93% 1.09% 8.31% PercSpendTrain Percentage of revenue invested in training. 0.11% 0.66% 0.10% 1.03% EducR&D Level of education of employees in R&D. 0.23 0.41 0.34 0.48 PercRevInnov Percentage of sales revenue generated by innovations. 17.29 27.77 21.50 30.62 Cli How important the clients are for innovation activities. 0.43 1.00 0.61 1.15 Supp How important the suppliers are for innovation activities. 0.50 1.06 0.67 1.17 Comp How important the competitors are for innovation activities. 0.16 0.58 0.22 0.68 Consult How important the consulting firms are for innovation activities. 0.23 0.70 0.33 0.81 Univ How important the universities are for innovation activities. 0.23 0.72 0.35 0.87 SalesRev2014 (BRL x 1,000,000) Sales revenue in 2014. 224 1,440.00 588.00 5,870.00 NumEmp2014 Number of employees at the end of 2014. 490.55 2,439.30 795.49 2,191.64 Productivity(BRL x 1,000,000) 2014 sales revenue/number of employees in 2014. 0.321 0.593 0.443 0.626 Source: Elaborated by the authors. Another interesting fact concerns internal R&D activities. With 1.49% and 1.78% of employees in R&D, respectively, non-cluster and cluster firms have the following investments: in R&D, it varies from 0.79% (non-cluster) to 1.09% (cluster); in R&D training expenses, this value is around 0.1% in both groups. The level of education of employees in clustered firms is also higher (0.34 versus 0.23). Therefore, considering the variables of internal R&D, clustered firms score higher in all except for R&D training expenses, which is similar for both groups. Analyzing the importance of different partner types for innovation, which reflects external R&D activities, clustered firms also score higher than the non-clustered ones considering all types of partners: clients, with 0.61 versus 0.43; suppliers with 0.67 versus 0.50; competitors, with 0.22 versus 0.16; consultants, with 0.33 versus 0.23; and universities, with 0.35 versus 0.23. We can verify the importance of vertical alliances, with suppliers being considered the most important, followed by clients for both groups. Horizontal alliances (those with competitors) are considered less important and surpassed by the ones with consulting firms, universities, and research institutes. Lastly, productivity is also higher for clustered firms (0.443 versus 0.321). If R&D activities and IP are higher for this group, and IP, as this paper advocates, positively influences productivity, this study predicted that clustered firms have superior productivity than the non-clustered ones. The results of the multigroup CFA are presented in Table 3. Convergent validity was acceptable for both the latent variables, as all average variance extracted (AVE) was higher than 0.5 (Fornell & Larcker, 1981), and the standardized regression weights mainly were higher than 0.7 (Nunally & Bernstein, 1978). Discriminant validity was also achieved, with the construct AVE being higher than the squared correlation between the constructs, in addition to nomological validity, with this correlation being significant but low (Hair et al., 2006). Table 3 CFA results Non-cluster (n = 3,703) Construct Proxy Std. R.W. SE p-value AVE CR Innovation Performance PercRevInnov 0.526 0.689 0.895 Innovation Performance InnovRad 0.892 0.049 *** Innovation Performance InnovExt 0.834 0.046 *** Innovation Performance IProd 0.995 0.054 *** External R&D - alliances Cli 0.914 0.039 *** 0.570 0.837 External R&D - alliances Supp 0.833 0.037 *** External R&D - alliances Comp 0.619 0.033 *** External R&D - alliances Univ 0.608 Cluster (n = 1,878) Construct Proxy Std. R.W. SE p-value AVE CR Innovation Performance PercRevInnov 0.485 0.655 0.878 Innovation Performance InnovRad 0.875 0.074 *** Innovation Performance InnovExt 0.806 0.075 *** Innovation Performance IProd 0.985 0.079 *** External R&D - alliances Cli 0.900 0.050 *** 0.567 0.835 External R&D - alliances Supp 0.847 0.047 *** External R&D - alliances Comp 0.604 0.044 *** External R&D - alliances Univ 0.614 Source: Elaborated by the authors. *** p < 0.01. The SEM analysis (see Table 4) determined that all the relationships proposed by the model are significant for both groups (t-test presenting p < 0.05 for all relationships). External R&D positively influenced IP for the cluster and non-cluster groups (0.197 and 0.168, respectively). The relationship between internal R&D and IP was positive and significant in all cases (0.046 for non-cluster and 0.023 for cluster). The moderation of internal R&D on the relationship between external R&D and IP, in turn, was different when comparing both groups. For clustered firms, the relationship was 0.012 and significant. In the case of the non-clustered ones, the relationship, although significant, was negative (-0.038), indicating a negative moderation of AC in the influence of external R&D on IP for firms in this group. The influence of IP on productivity was also positive for both groups (0.211 for non-cluster and 0.320 for cluster). Table 4 SEM results Non-cluster(n = 3,703) Cluster (n = 1,878) Relationship S. R. W. p-value* S. R. W. p-value* External R&D → IP 0.168 0.000 0.197 0.000 Internal R&D → IP 0.046 0.000 0.023 0.003 External R&D x internal R&D → IP -0.038 0.000 0.012 0.024 IP → productivity 0.211 0.000 0.320 0.000 Source: Elaborated by the authors. * Statistical significance with p < 0.05 (t-test). Considering the hypotheses, which stated that all the previous relationships are more intensely felt in clustered than in non-clustered firms, only H3 was supported, and all the others were rejected. An analysis of the confidence intervals of the four regression weights, calculated with p-value = 0.05 according to the t distribution (see Table 5), indicated that among all of them, only external R&D x internal R&D → IP had a statistically significant difference considering both groups, while the others did not (high­lighting that a statistically significant difference between groups happens when the confidence intervals do not overlap). Consequently, the only difference the results indicated was that AC, represented by the indicator of internal R&D, positively moderates the relationship between external R&D and IP in clusters. At the same time, this moderation is negative outside clusters. Table 5 Confidence intervals Non-cluster (n = 3,703) Relationship Low High Sig. Hypotheses test External R&D → IP 0.137 0.199 ** H1: rejected Internal R&D → IP 0.032 0.060 ** H2: rejected External R&D x internal R&D → IP -0.050 -0.026 ** H3: supported IP → productivity 0.142 0.280 ** H4: rejected Cluster (n = 1,878) Relationship Low High Sig. External R&D → IP 0.150 0.244 ** Internal R&D → IP 0.007 0.039 ** External R&D x internal R&D → IP 0.002 0.022 ** IP → productivity 0.216 0.424 ** Source: Elaborated by the authors. ** p-value = 0.05 (t distribution). DISCUSSION The results indicated that the development of internal R&D and external R&D does not directly promote higher IP in Brazilian manufacturing firms participating in a cluster, as Table 4 shows. These results ensure the influence of geographical proximity (as high as possible inside a cluster) on IP remains an open topic of research. The literature diverges from this result, with some studies indicating that the impact is positive (Cardamone, 2018), some suggesting that it is negative (Presutti et al., 2019), and others saying that low and high levels of proximity are harmful while an intermediate level is more favorable (Boschma, 2005; Leeuw et al., 2014). In this study, the direct effects of both R&D sources separately on IP were positive, and their differences were not statistically significant, as demonstrated in Table 5, contradicting all the streams mentioned above. The influence of IP on productivity was also similar for both groups. These results and the close average of firms’ productivity in both groups contrast with the conclusions of Arimoto et al. (2014). The authors compared the productivity of firms in and out of clusters in the silk industry. They concluded that firms in clusters are more productive due to the concentration of the best firms from the industry and related sectors, which would also influence innovation. The results also diverge from Cavalcante et al. (2015), who found a statistically significant positive relationship between R&D, innovation, and productivity in Brazilian firms in clusters. The only difference in the comparison between both groups concerns the moderation of internal R&D on the relationship between external R&D and IP. This relationship tests the hypothesis based on the AC theory (Cohen & Levinthal, 1990; Zahra & George, 2002), with the internal R&D construct reflecting this capability (Cohen & Levinthal, 1990). The moderation was positive in the group of firms participating in clusters and negative in non-clustered firms. This finding indicates that the relationship between internal and external R&D is complementary (Paula & Silva, 2018a; Cassiman & Veugelers, 2006) in clusters, while these sources of R&D are substitutes (Berchicci, 2013) outside them. These results are consistent with the nature of the clusters and the geographical proximity brought about by this type of arrangement. Clusters facilitate knowledge transfer (Eisingerich et al., 2010; Guo et al., 2020), to which the firms’ AC is central. Furthermore, participation in clusters stimulates the development of relative AC (Lane & Lubatkin, 1998) among the participant firms. This phenomenon occurs because the proximity favors mutual observation and informal exchanges that promote the development of shared concepts, tools, language, and standards of business conduct, increasing trust and reciprocity, which facilitate information sharing and innovation (Ozer & Zhang, 2015). As stated by Bittencourt et al. (2019, p. 657), the “process of assimilation, transformation, and application of knowledge can be linked to absorptive capacity”, and clusters stimulate the formation of learning relationships and partnerships that promote innovative potential in institutions through interactions that generate knowledge and innovation (Bittencourt et al., 2019). This argument was reinforced by Ikram et al. (2018), which stated that the relationship between industries, government, and universities (also called Triple Helix, Ikram et al., 2018) within clusters creates a business environment that integrates companies, thus favoring innovation. Another possible explanation for the positive influence of AC solely in the group of firms in clusters is that the geographical proximity between companies operating in the same segment, which occurs naturally in clusters, can also facilitate knowledge spillovers (Cassiman & Veugelers, 2002). A high AC may allow these firms to take advantage of spillovers with more intensity (Aldieri et al., 2018; Caldas et al., 2019; Cardamone, 2018) than firms that are not in clusters, obtaining superior performance and, consequently, competitive advantage. CONCLUSIONS This study could verify that internal and external R&D are complementary in Brazilian manufacturing firms that operate in clusters, while they are substitutes in firms outside clusters. These results support the hypothesis that external R&D is available for every firm and helps to promote innovation in both contexts. However, participation in a cluster facilitates interaction and trust between firms from the same or correlated sectors with a level of knowledge redundancy sufficient for their relative AC to be higher on average than in the case of firms outside a cluster. Therefore, clustered firms’ AC is more effective in promoting open innovation activities, in which internal and external knowledge should be combined, helping to improve their IP. The influence of the cluster on the effectiveness of AC to promote open innovation and improve IP of firms implies that other factors - besides the effort they put into learning and their previously accumulated knowledge (Kim, 1997) - are relevant in the development of AC. Geographical proximity could be one crucial factor. Another one could be the sense of belonging that facilitates identification and commitment among firms in clusters (Ozer & Zhang, 2015), which help to build more effective communication channels. These factors might promote gains in the relative AC (Lane & Lubatkin, 1998), thus increasing their ability to interact effectively and conduct innovation activities. Considering the results, we can extract relevant contributions. The main theoretical contribution is the significant evidence of the contingent nature of open innovation and AC theories. The theory presupposes a complementary relationship between internal and external R&D to improve a firm’s IP. However, our results indicate that they are complementary for clustered firms and substitutes for non-clustered firms. As a practical contribution, we can mention that managers from clustered and non-clustered firms should have distinct priorities when deciding on their innovation strategies. Clustered firms should invest conjointly in internal R&D and in their alliances with other firms inside the cluster to improve external R&D, and their combination potentiates IP and productivity. Non-clustered firms, in turn, should choose whether they want to invest in internal R&D (if they have enough monetary resources) or prioritize external R&D, by investing in forming alliances with partners that possess the resources they lack. We can point out several limitations in the study. The method adopted by Britto and Albuquerque (2000), besides considering the participation (in terms of employees) of a sector, uses the proportion of industries that are suppliers of machinery, equipment, and processes for the focal industry. However, we did not have access to information about these suppliers’ industries and only used the sector’s participation in the calculation. Another limitation is that we worked with Brazilian manufacturing firms, making the results country and industry-specific. Another limitation is that Pintec is a survey based on the impressions and feelings of interviewees, which can cause bias to the results. Also, as not all firms answered Pintec and PIA, and the non-respondents are not identifiable, the sample was not probabilistic. Finally, it is a limitation that the only control variable used in the analysis was firm participation in a cluster, with this comparison being the study’s main contribution. We did not consider other control variables commonly used in the innovation management literature, such as firm size, location, age, and sector. It can be a source of model misspecification, and testing the model with additional control variables is an opportunity for future studies. However, Brazil is a relevant emerging country in Latin America, part of the group composed of Brazil, Russia, India, China and South Africa (Brics), so the results may presumably be generalized for Latin American and other emerging economies. For future studies, we suggest using the complete method of Britto and Albuquerque (2000) to classify clusters more precisely. Also, there is an opportunity to expand the analysis to other Latin America countries, to the countries that compose the Brics, or other developing countries. Lastly, we suggest expanding the study to other industries, such as service and agribusiness, and the control by sector could be more precise, using, for instance, the industrial classification of Pavitt (1984). 1 This work was supported by the Brazilian National Council for Scientific and Technological Development (Conselho Nacional de Desenvolvimento Científico e Tecnológico - CNPq), under the No. 305706/2015-2,406559/2018-0, and by the Foundation for Research Support of the State of Rio de Janeiro (Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro - Faperj), under the Nos. 26/210.277/2019(248665), E-26/201.409/2021(260810), E-26/202.864/2018 (239233). REFERENCES Acs, Z. J., Stam, E., Audretsch, D. B., & O’Connor, A. (2017). The lineages of the entrepreneurial ecosystem approach. Small Business Economics, 49(1), 1-10. https://doi.org/10.1007/s11187-017-9864-8 Acs Z. J. Stam E. Audretsch D. B. O’Connor A. 2017 The lineages of the entrepreneurial ecosystem approach Small Business Economics 49 1 1 10 https://doi.org/10.1007/s11187-017-9864-8 Aldieri, L., Barra, C., Paolo Vinci, C., & Zotti, R. (2021). The joint impact of different types of innovation on firm’s productivity: Evidence from Italy. Economics of Innovation and New Technology, 30(2), 151-182. https://doi.org/10.1080/10438599.2019.1685211 Aldieri L. Barra C. Paolo Vinci C. Zotti R. 2021 The joint impact of different types of innovation on firm’s productivity: Evidence from Italy Economics of Innovation and New Technology 30 2 151 182 https://doi.org/10.1080/10438599.2019.1685211 Aldieri, L., Sena, V., & Vinci, C. P. (2018). Domestic R&D spillovers and absorptive capacity: Some evidence for US, Europe and Japan. International Journal of Production Economics, 198, 38-49. https://doi.org/10.1016/j.ijpe.2018.01.015 Aldieri L. Sena V. Vinci C. P. 2018 Domestic R&D spillovers and absorptive capacity: Some evidence for US, Europe and Japan International Journal of Production Economics 198 38 49 https://doi.org/10.1016/j.ijpe.2018.01.015 Ardito, L., Peruffo, E., & Natalicchio, A. (2019). The relationships between the internationalization of alliance portfolio diversity, individual incentives, and innovation ambidexterity: A microfoundational approach. Technological Forecasting and Social Change, 148, 119714. https://doi.org/10.1016/j.techfore.2019.119714 Ardito L. Peruffo E. Natalicchio A. 2019 The relationships between the internationalization of alliance portfolio diversity, individual incentives, and innovation ambidexterity: A microfoundational approach Technological Forecasting and Social Change 148 119714 https://doi.org/10.1016/j.techfore.2019.119714 Arimoto, Y., Nakajima, K., & Okazaki, T. (2014). Sources of productivity improvement in industrial clusters: The case of the prewar Japanese silk-reeling industry. Regional Science and Urban Economics, 46, 27-41. https://doi.org/10.1016/j.regsciurbeco.2014.02.004 Arimoto Y. Nakajima K. Okazaki T. 2014 Sources of productivity improvement in industrial clusters: The case of the prewar Japanese silk-reeling industry Regional Science and Urban Economics 46 27 41 https://doi.org/10.1016/j.regsciurbeco.2014.02.004 Beers, C. & Zand, F. (2014). R&D cooperation, partner diversity, and innovation performance: An empirical analysis. Journal of Product Innovation Management, 31(2), 292-312. https://doi.org/10.1111/jpim.12096 Beers C. Zand F. 2014 R&D cooperation, partner diversity, and innovation performance: An empirical analysis Journal of Product Innovation Management 31 2 292 312 https://doi.org/10.1111/jpim.12096 Bell, M., & Figueiredo, P. N. (2012). Innovation capability building and learning mechanisms in latecomer firms: Recent empirical contributions and implications for research. Canadian Journal of Development Studies, 33(1), 14-40. https://doi.org/10.1080/02255189.2012.677168 Bell M. Figueiredo P. N. 2012 Innovation capability building and learning mechanisms in latecomer firms: Recent empirical contributions and implications for research Canadian Journal of Development Studies 33 1 14 40 https://doi.org/10.1080/02255189.2012.677168 Belussi, F., Sammarra, A., & Sedita, S. R. (2010). Learning at the boundaries in an “open regional innovation system”: A focus on firms’ innovation strategies in the Emilia Romagna life science industry. Research Policy, 39(6), 710-721. https://doi.org/10.1016/j.respol.2010.01.014 Belussi F. Sammarra A. Sedita S. R. 2010 Learning at the boundaries in an “open regional innovation system”: A focus on firms’ innovation strategies in the Emilia Romagna life science industry Research Policy 39 6 710 721 https://doi.org/10.1016/j.respol.2010.01.014 Berchicci, L. (2013). Towards an open R&D system: Internal R&D investment, external knowledge acquisition and innovative performance. Research Policy, 42(1), 117-127. https://doi.org/10.1016/j.respol.2012.04.017 Berchicci L. 2013 Towards an open R&D system: Internal R&D investment, external knowledge acquisition and innovative performance Research Policy 42 1 117 127 https://doi.org/10.1016/j.respol.2012.04.017 Bittencourt, B. A., Zen, A. C., & Prévot, F. (2019). Capacidade de inovação dos clusters: Entendimento da inovação de redes geográficas de negócios. Revista Brasileira de Gestão Negócios, 21(4), 647-663. https://doi.org/10.7819/rbgn.v21i4.4016 Bittencourt B. A. Zen A. C. Prévot F. 2019 Capacidade de inovação dos clusters: Entendimento da inovação de redes geográficas de negócios Revista Brasileira de Gestão Negócios 21 4 647 663 https://doi.org/10.7819/rbgn.v21i4.4016 Boschma, R. (2005). Proximity and innovation: A critical assessment. Regional Studies, 39(1), 61-74. https://doi.org/10.1080/0034340052000320887 Boschma R. 2005 Proximity and innovation: A critical assessment Regional Studies 39 1 61 74 https://doi.org/10.1080/0034340052000320887 Britto, J., & Albuquerque, E. M. (2000). Estrutura e dinamismo de clusters industriais na economia brasileira: Uma análise comparativa. Revista Economia Ensaios, 15(1), 1-22. Britto J. Albuquerque E. M. 2000 Estrutura e dinamismo de clusters industriais na economia brasileira: Uma análise comparativa Revista Economia Ensaios 15 1 1 22 Caldas, L. F. P., Paula, F. O., & Macedo-Soares, T. D. L. A. (2019). Industry innovation spending and openness to collaboration as levers for firm performance. European Journal of Innovation Management, 22(4), 617-638. https://doi.org/10.1108/EJIM-04-2018-0075 Caldas L. F. P. Paula F. O. Macedo-Soares T. D. L. A. 2019 Industry innovation spending and openness to collaboration as levers for firm performance European Journal of Innovation Management 22 4 617 638 https://doi.org/10.1108/EJIM-04-2018-0075 Cardamone, P. (2018). Firm innovation and spillovers in Italy: Does geographical proximity matter? Letters in Spatial and Resource Sciences, 11(1), 1-16. https://doi.org/10.1007/s12076-017-0193-y Cardamone P. 2018 Firm innovation and spillovers in Italy: Does geographical proximity matter? Letters in Spatial and Resource Sciences 11 1 1 16 https://doi.org/10.1007/s12076-017-0193-y Cassiman, B., & Veugelers, R. (2002). R&D cooperation and spillovers: Some empirical evidence from Belgium. American Economic Review, 92(4), 1169-1184. https://doi.org/10.1257/00028280260344704 Cassiman B. Veugelers R. 2002 R&D cooperation and spillovers: Some empirical evidence from Belgium American Economic Review 92 4 1169 1184 https://doi.org/10.1257/00028280260344704 Cassiman, B., & Veugelers, R. (2006). In search of complementarity in innovation strategy: Internal R&D and external knowledge acquisition. Management Science, 52(1), 68-82. https://doi.org/10.1287/mnsc.1050.0470 Cassiman B. Veugelers R. 2006 In search of complementarity in innovation strategy: Internal R&D and external knowledge acquisition Management Science 52 1 68 82 https://doi.org/10.1287/mnsc.1050.0470 Castro, L. (2015). Strategizing across boundaries: Revisiting knowledge brokering activities in French innovation clusters. Journal of Knowledge Management, 19(5), 1048-1068. https://doi.org/10.1108/JKM-02-2015-0050 Castro L. 2015 Strategizing across boundaries: Revisiting knowledge brokering activities in French innovation clusters Journal of Knowledge Management 19 5 1048 1068 https://doi.org/10.1108/JKM-02-2015-0050 Cavalcante, L. R, Jacinto, P. A., & Negri, F. (2015). P&D, inovação e produtividade na indústria brasileira. In F. De Negri & L. R. Cavalcante (Eds.). Produtividade no Brasil: Desempenho e determinantes (pp. 43-68). ABDI; Ipea. Cavalcante L. R Jacinto P. A. Negri F. 2015 P&D, inovação e produtividade na indústria brasileira Negri F. De Cavalcante L. R. Produtividade no Brasil: Desempenho e determinantes 43 68 ABDI Ipea Chesbrough, H. W. (2003). Open innovation: The new imperative for creating and profiting from technology. Harvard Business Press. Chesbrough H. W. 2003 Open innovation: The new imperative for creating and profiting from technology Harvard Business Press Cluster Mapping. (2019). Cluster: A regional concentration of related industries. https://www.clustermapping.us/cluster/ Cluster Mapping 2019 Cluster: A regional concentration of related industries https://www.clustermapping.us/cluster/ Cobeña, M., Gallego, Á., & Casanueva, C. (2017). Heterogeneity, diversity and complementarity in alliance portfolios. European Management Journal, 35(4), 464-476. https://doi.org/10.1016/j.emj.2016.12.005 Cobeña M. Gallego Á. Casanueva C. 2017 Heterogeneity, diversity and complementarity in alliance portfolios European Management Journal 35 4 464 476 https://doi.org/10.1016/j.emj.2016.12.005 Cobeña, M., Gallego, Á., & Casanueva, C. (2019). Diversity in airline alliance portfolio configuration. Journal of Air Transport Management, 75, 16-26. https://doi.org/10.1016/j.jairtraman.2018.11.004 Cobeña M. Gallego Á. Casanueva C. 2019 Diversity in airline alliance portfolio configuration Journal of Air Transport Management 75 16 26 https://doi.org/10.1016/j.jairtraman.2018.11.004 Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35(Special Issue: Technology, Organizations, and Innovation), 128-152. https://doi.org/10.2307/2393553 Cohen W. M. Levinthal D. A. 1990 Absorptive capacity: A new perspective on learning and innovation Administrative Science Quarterly 35 Special Issue: Technology, Organizations, and Innovation 128 152 https://doi.org/10.2307/2393553 Dabla-Norris, E., Kersting, E. K., & Verdier, G. (2012). Firm productivity, innovation, and financial development. Southern Economic Journal, 79(2), 422-449. https://doi.org/10.4284/0038-4038-2011.201 Dabla-Norris E. Kersting E. K. Verdier G. 2012 Firm productivity, innovation, and financial development Southern Economic Journal 79 2 422 449 https://doi.org/10.4284/0038-4038-2011.201 Dushnitsky, G., & Lenox, M. J. (2005a). When do firms undertake R&D by investing in new ventures? Strategic Management Journal, 26(10), 947-965. https://doi.org/10.1002/smj.488 Dushnitsky G. Lenox M. J. 2005a When do firms undertake R&D by investing in new ventures? Strategic Management Journal 26 10 947 965 https://doi.org/10.1002/smj.488 Dushnitsky, G., & Lenox, M. J. (2005b). When do incumbents learn from entrepreneurial ventures? Corporate venture capital and investing firm innovation rates. Research Policy, 34(5), 615-639. https://doi.org/10.1016/j.respol.2005.01.017 Dushnitsky G. Lenox M. J. 2005b When do incumbents learn from entrepreneurial ventures? Corporate venture capital and investing firm innovation rates Research Policy 34 5 615 639 https://doi.org/10.1016/j.respol.2005.01.017 Eisingerich, A. B., Bell, S. J., & Tracey, P. (2010). How can clusters sustain performance? The role of network strength, network openness, and environmental uncertainty. Research Policy, 39(2), 239-253. https://doi.org/10.1016/j.respol.2009.12.007 Eisingerich A. B. Bell S. J. Tracey P. 2010 How can clusters sustain performance? The role of network strength, network openness, and environmental uncertainty Research Policy 39 2 239 253 https://doi.org/10.1016/j.respol.2009.12.007 Faems, D., Van Looy, B., & Debackere, K. (2005). Interorganizational collaboration and innovation: Toward a portfolio approach. Journal of Product Innovation Management, 22(3), 238-250. https://doi.org/10.1111/j.0737-6782.2005.00120.x Faems D. Van Looy B. Debackere K. 2005 Interorganizational collaboration and innovation: Toward a portfolio approach Journal of Product Innovation Management 22 3 238 250 https://doi.org/10.1111/j.0737-6782.2005.00120.x Fernandes, C. I., & Ferreira, J. J. (2013). Knowledge spillovers: Cooperation between universities and KIBS. R&D Management, 43(5), 461-472. https://doi.org/10.1111/radm.12024 Fernandes C. I. Ferreira J. J. 2013 Knowledge spillovers: Cooperation between universities and KIBS R&D Management 43 5 461 472 https://doi.org/10.1111/radm.12024 Fey, C. F., & Birkinshaw, J. (2005). External sources of knowledge, governance mode, and R&D performance. Journal of Management, 31(4), 597-621. https://doi.org/10.1177/0149206304272346 Fey C. F. Birkinshaw J. 2005 External sources of knowledge, governance mode, and R&D performance Journal of Management 31 4 597 621 https://doi.org/10.1177/0149206304272346 Findik, D., & Beyhan, B. (2017). A perceptual measure of innovation performance: Firm-level evidence from Turkey. International Journal of Innovation and Technology Management, 14(06), 1750038. https://doi.org/10.1142/S0219877017500389 Findik D. Beyhan B. 2017 A perceptual measure of innovation performance: Firm-level evidence from Turkey International Journal of Innovation and Technology Management 14 06 1750038 https://doi.org/10.1142/S0219877017500389 Fornell, C., & Larcker, D. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39-50. https://doi.org/10.1177/002224378101800104 Fornell C. Larcker D. 1981 Evaluating structural equation models with unobservable variables and measurement error Journal of Marketing Research 18 1 39 50 https://doi.org/10.1177/002224378101800104 Garcia Martinez, M., Zouaghi, F., & Sanchez Garcia, M. (2017). Capturing value from alliance portfolio diversity: The mediating role of R&D human capital in high and low tech industries. Technovation, 59, 55-67. https://doi.org/10.1016/j.technovation.2016.06.003 Garcia Martinez M. Zouaghi F. Sanchez Garcia M. 2017 Capturing value from alliance portfolio diversity: The mediating role of R&D human capital in high and low tech industries Technovation 59 55 67 https://doi.org/10.1016/j.technovation.2016.06.003 Garcia Martinez, M., Zouaghi, F., & Sanchez Garcia, M. (2019). Casting a wide net for innovation: Mediating effect of R&D human and social capital to unlock the value from alliance portfolio diversity. British Journal of Management, 30(4), 769-790. https://doi.org/10.1111/1467-8551.12310 Garcia Martinez M. Zouaghi F. Sanchez Garcia M. 2019 Casting a wide net for innovation: Mediating effect of R&D human and social capital to unlock the value from alliance portfolio diversity British Journal of Management 30 4 769 790 https://doi.org/10.1111/1467-8551.12310 Giuliani, E., & Bell, M. (2005). The micro-determinants of meso-level learning and innovation: Evidence from a Chilean wine cluster. Research Policy, 34(1), 47-68. https://doi.org/10.1016/j.respol.2004.10.008 Giuliani E. Bell M. 2005 The micro-determinants of meso-level learning and innovation: Evidence from a Chilean wine cluster Research Policy 34 1 47 68 https://doi.org/10.1016/j.respol.2004.10.008 Greco, M., Grimaldi, M., Locatelli, G., & Serafini, M. (2021). How does open innovation enhance productivity? An exploration in the construction ecosystem. Technological Forecasting and Social Change, 168, 120740. https://doi.org/10.1016/j.techfore.2021.120740 Greco M. Grimaldi M. Locatelli G. Serafini M. 2021 How does open innovation enhance productivity? An exploration in the construction ecosystem Technological Forecasting and Social Change 168 120740 https://doi.org/10.1016/j.techfore.2021.120740 Griffith, R., Huergo, E., Mairesse, J., & Peters, B. (2006). Innovation and productivity across four European countries. Oxford Review of Economic Policy, 22(4), 483-498. https://doi.org/10.1093/oxrep/grj028 Griffith R. Huergo E. Mairesse J. Peters B. 2006 Innovation and productivity across four European countries Oxford Review of Economic Policy 22 4 483 498 https://doi.org/10.1093/oxrep/grj028 Gulati, R. (1998). Alliances and networks. Strategic Management Journal, 19(4), 293-317. https://doi.org/10.1002/(SICI)1097-0266(199804)19:4<293::AID-SMJ982>3.0.CO;2-M Gulati R. 1998 Alliances and networks Strategic Management Journal 19 4 293 317 https://doi.org/10.1002/(SICI)1097-0266(199804)19:4<293::AID-SMJ982>3.0.CO;2-M Guo, B., & Guo, J. J. (2011). Patterns of technological learning within the knowledge systems of industrial clusters in emerging economies: Evidence from China. Technovation, 31(2-3), 87-104. https://doi.org/10.1016/j.technovation.2010.10.006 Guo B. Guo J. J. 2011 Patterns of technological learning within the knowledge systems of industrial clusters in emerging economies: Evidence from China Technovation 31 2-3 87 104 https://doi.org/10.1016/j.technovation.2010.10.006 Guo, J., Guo, B., Zhou, J., & Wu, X. (2020). How does the ambidexterity of technological learning routine affect firm innovation performance within industrial clusters? The moderating effects of knowledge attributes. Technological Forecasting and Social Change, 155, 119990. https://doi.org/10.1016/j.techfore.2020.119990 Guo J. Guo B. Zhou J. Wu X. 2020 How does the ambidexterity of technological learning routine affect firm innovation performance within industrial clusters? The moderating effects of knowledge attributes Technological Forecasting and Social Change 155 119990 https://doi.org/10.1016/j.techfore.2020.119990 Hagedoorn, J. (1993). Understanding the rationale of strategic technology partnering: Interorganizational modes of cooperation and sectoral dif­ferences. Strategic Management Journal, 14(5), 371-385. https://doi.org/10.1002/smj.4250140505 Hagedoorn J. 1993 Understanding the rationale of strategic technology partnering: Interorganizational modes of cooperation and sectoral dif­ferences Strategic Management Journal 14 5 371 385 https://doi.org/10.1002/smj.4250140505 Hagedoorn, J., & Wang, N. (2012). Is there complementarity or substitutability between internal and external R&D strategies? Research Policy, 41(6), 1072-1083. https://doi.org/10.1016/j.respol.2012.02.012 Hagedoorn J. Wang N. 2012 Is there complementarity or substitutability between internal and external R&D strategies? Research Policy 41 6 1072 1083 https://doi.org/10.1016/j.respol.2012.02.012 Hair, J., Black, W., Babin, B., Anderson, R., & Tatham, R. (2006). Multivariate data analysis. Pearson Prentice Hall. Hair J. Black W. Babin B. Anderson R. Tatham R. 2006 Multivariate data analysis Pearson Prentice Hall Ikram, A., Su, Q., Fiaz, M., & Rehman, R. U. (2018). Cluster strategy and supply chain management: The road to competitiveness for emerging economies. Benchmarking: An International Journal, 25(5), 1302-1318. https://doi.org/10.1108/BIJ-06-2015-0059 Ikram A. Su Q. Fiaz M. Rehman R. U. 2018 Cluster strategy and supply chain management: The road to competitiveness for emerging economies Benchmarking: An International Journal 25 5 1302 1318 https://doi.org/10.1108/BIJ-06-2015-0059 Inhan, L., Ferreira, J., Marques, C., & Rebelo, J. (2013). Paradoxo de inovação no cluster do vinho: O caso da região demarcada do Douro. Revista de Administração de Empresas, 53(3), 256-271. https://doi.org/10.1590/S0034-75902013000300004 Inhan L. Ferreira J. Marques C. Rebelo J. 2013 Paradoxo de inovação no cluster do vinho: O caso da região demarcada do Douro Revista de Administração de Empresas 53 3 256 271 https://doi.org/10.1590/S0034-75902013000300004 Instituto Brasileiro de Geografia e Estatística (IBGE) (2019a). Pesquisa de Inovação - PINTEC. https://www.ibge.gov.br/estatisticas/multidominio/ciencia-tecnologia-e-inovacao/9141-pesquisa-de-inovacao.html?=&t=destaques Instituto Brasileiro de Geografia e Estatística (IBGE) 2019a Pesquisa de Inovação - PINTEC https://www.ibge.gov.br/estatisticas/multidominio/ciencia-tecnologia-e-inovacao/9141-pesquisa-de-inovacao.html?=&t=destaques Instituto Brasileiro de Geografia e Estatística (IBGE) (2019b). Pesquisa Industrial Anual - Empresa. https://www.ibge.gov.br/estatisticas/economicas/industria/9042-pesquisa-industrial-anual.html?=&t=o-que-e Instituto Brasileiro de Geografia e Estatística (IBGE) 2019b Pesquisa Industrial Anual - Empresa https://www.ibge.gov.br/estatisticas/economicas/industria/9042-pesquisa-industrial-anual.html?=&t=o-que-e Kapetaniou, C., & Lee, S. H. (2019). Geographical proximity and open innovation of SMEs in Cyprus. Small Business Economics, 52(1), 261-276. https://doi.org/10.1007/s11187-018-0023-7 Kapetaniou C. Lee S. H. 2019 Geographical proximity and open innovation of SMEs in Cyprus Small Business Economics 52 1 261 276 https://doi.org/10.1007/s11187-018-0023-7 Kim, L. (1997). Imitation to innovation: The dynamics of Korea’s technological learning. Harvard Business Press. Kim L. 1997 Imitation to innovation: The dynamics of Korea’s technological learning Harvard Business Press King, N. C. D. O., Lima, E. P. D., & Costa, S. E. G. D. (2014). Produtividade sistêmica: Conceitos e aplicações. Production, 24(1), 160-176. https://doi.org/10.1590/S0103-65132013005000006 King N. C. D. O. Lima E. P. D. Costa S. E. G. D. 2014 Produtividade sistêmica: Conceitos e aplicações Production 24 1 160 176 https://doi.org/10.1590/S0103-65132013005000006 Lai, Y., Hsub, M., Lin, F., Chen, Y., & Lin, Y. (2014). The effects of industry cluster knowledge management on innovation performance. Journal of Business Research, 67, 734-739. https://doi.org/10.1016/j.jbusres.2013.11.036 Lai Y. Hsub M. Lin F. Chen Y. Lin Y. 2014 The effects of industry cluster knowledge management on innovation performance Journal of Business Research 67 734 739 https://doi.org/10.1016/j.jbusres.2013.11.036 Lane, P. J., & Lubatkin, M. (1998). Relative absorptive capacity and interorganizational learning. Strategic Management Journal, 19(5), 461-477. https://doi.org/10.1002/(SICI)1097-0266(199805)19:5<461::AID-SMJ953>3.0.CO;2-L Lane P. J. Lubatkin M. 1998 Relative absorptive capacity and interorganizational learning Strategic Management Journal 19 5 461 477 https://doi.org/10.1002/(SICI)1097-0266(199805)19:5<461::AID-SMJ953>3.0.CO;2-L Leeuw, T., Lokshin, B., & Duysters, G. (2014). Returns to alliance portfolio diversity: The relative effects of partner diversity on firm’s innovative performance and productivity. Journal of Business Research, 67(9) 1839-1849. https://doi.org/10.1016/j.jbusres.2013.12.005 Leeuw T. Lokshin B. Duysters G. 2014 Returns to alliance portfolio diversity: The relative effects of partner diversity on firm’s innovative performance and productivity Journal of Business Research 67 9 1839 1849 https://doi.org/10.1016/j.jbusres.2013.12.005 Levitt, T. (1993). The globalization of markets. In R. Z. Aliber, & R. W. Click (Orgs.). Readings in international business: A decision approach (pp. 249-266). Cambridge University Press. Levitt T. 1993 The globalization of markets Aliber R. Z. Click R. W. Readings in international business: A decision approach (pp 249 266 Cambridge University Press Little, T. D., Bovaird, J. A., & Widaman, K. F. (2006). On the merits of orthogonalizing powered and product terms: Implications for modeling interactions among latent variables. Structural Equation Modeling, 13(4), 497-519. https://doi.org/10.1207/s15328007sem1304_1 Little T. D. Bovaird J. A. Widaman K. F. 2006 On the merits of orthogonalizing powered and product terms: Implications for modeling interactions among latent variables Structural Equation Modeling 13 4 497 519 https://doi.org/10.1207/s15328007sem1304_1 Malerba, F. (1992). Learning by firms and incremental technical change. The Economic Journal, 102(413), 845-859. https://doi.org/10.2307/2234581 Malerba F. 1992 Learning by firms and incremental technical change The Economic Journal 102 413 845 859 https://doi.org/10.2307/2234581 Mascena, K. M. C., Figueiredo, F. C., & Boaventura, J. M. G. (2013). Clusters e APL’s: análise bibliométrica das publicações nacionais no período de 2000 a 2011. Revista de Administração de Empresas, 53(5), 454-468. https://doi.org/10.1590/S0034-75902013000500004 Mascena K. M. C. Figueiredo F. C. Boaventura J. M. G. 2013 Clusters e APL’s: análise bibliométrica das publicações nacionais no período de 2000 a 2011 Revista de Administração de Empresas 53 5 454 468 https://doi.org/10.1590/S0034-75902013000500004 Ministério do Trabalho (2020). PDET - Programa de Disseminação das Estatísticas do Trabalho. http://pdet.mte.gov.br/acesso-online-as-bases-de-dados Ministério do Trabalho 2020 PDET - Programa de Disseminação das Estatísticas do Trabalho http://pdet.mte.gov.br/acesso-online-as-bases-de-dados Moutinho, R., Au-Yong-Oliveira, M., Coelho, A., & Pires Manso, J. (2015). The role of regional innovation systems (RIS) in translating R&D investments into economic and employment growth. Journal of Technology Management & Innovation, 10(2), 9-23. http://dx.doi.org/10.4067/S0718-27242015000200002 Moutinho R. Au-Yong-Oliveira M. Coelho A. Pires Manso J. 2015 The role of regional innovation systems (RIS) in translating R&D investments into economic and employment growth Journal of Technology Management & Innovation 10 2 9 23 http://dx.doi.org/10.4067/S0718-27242015000200002 Nalebluff, B. J., Brandenburger, A., & Maulana, A. (1996) Co-opetition. Harper Collins Business. Nalebluff B. J. Brandenburger A. Maulana A. 1996 Co-opetition Harper Collins Business Newlands, D. (2003). Competition and cooperation in industrial clusters: The implications for public policy. European Planning Studies, 11(5), 521-532. https://doi.org/10.1080/09654310303649 Newlands D. 2003 Competition and cooperation in industrial clusters: The implications for public policy European Planning Studies 11 5 521 532 https://doi.org/10.1080/09654310303649 Nunally, J. C., & Bernstein, I. (1978). Psychometric Theory. McGraw-Hill. Nunally J. C. Bernstein I. 1978 Psychometric Theory McGraw-Hill Oerlemans, L. A., Knoben, J., & Pretorius, M. W. (2013). Alliance portfolio diversity, radical and incremental innovation: The moderating role of technology management. Technovation, 33(6-7), 234-246. https://doi.org/10.1016/j.technovation.2013.02.004 Oerlemans L. A. Knoben J. Pretorius M. W. 2013 Alliance portfolio diversity, radical and incremental innovation: The moderating role of technology management Technovation 33 6-7 234 246 https://doi.org/10.1016/j.technovation.2013.02.004 Ozer, M., & Zhang, W. (2015). The effects of geographic and network ties on exploitative and exploratory product innovation. Strategic Management Journal, 36(7), 1105-1114. https://doi.org/10.1002/smj.2263 Ozer M. Zhang W. 2015 The effects of geographic and network ties on exploitative and exploratory product innovation Strategic Management Journal 36 7 1105 1114 https://doi.org/10.1002/smj.2263 Paula, F. O., & Silva, J. D. (2018a). Balancing internal and external R&D strategies to improve innovation and financial performance. Brazilian Administration Review, 15(2), e170129. https://doi.org/10.1590/1807-7692bar2018170129 Paula F. O. Silva J. D. 2018a Balancing internal and external R&D strategies to improve innovation and financial performance Brazilian Administration Review 15 2 e170129 https://doi.org/10.1590/1807-7692bar2018170129 Paula, F. D. O., & Silva, J. F. D. (2018b). The impact of alliances and internal R&D on the firm’s innovation and financial performance. Brazilian Business Review, 15(6), 533-550. https://doi.org/10.15728/bbr.2018.15.6.2 Paula F. D. O. Silva J. F. D. 2018b The impact of alliances and internal R&D on the firm’s innovation and financial performance Brazilian Business Review 15 6 533 550 https://doi.org/10.15728/bbr.2018.15.6.2 Paula, F. D. O., & Silva, J. F. da (2019). The role of the appropriability mechanisms for the innovative success of Portuguese small and medium enterprises. International Journal of Innovation Management, 23(04), 1950032. https://doi.org/10.1142/S1363919619500324 Paula F. D. O. Silva J. F. da 2019 The role of the appropriability mechanisms for the innovative success of Portuguese small and medium enterprises International Journal of Innovation Management 23 04 1950032 https://doi.org/10.1142/S1363919619500324 Pavitt, K. (1984). Sectoral patterns of technical change: Towards a taxonomy and a theory. Research Policy, 13(6), 343-373. https://doi.org/10.1016/0048-7333(84)90018-0 Pavitt K. 1984 Sectoral patterns of technical change: Towards a taxonomy and a theory Research Policy 13 6 343 373 https://doi.org/10.1016/0048-7333(84)90018-0 Perry, M. (2005). Business clusters: An international perspective (Vol. 30). Routledge. Perry M. 2005 Business clusters: An international perspective 30 Routledge Porter, M. E. (1990). The competitive advantage of nations. Harvard Business Review, 68(2), 73-93. Porter M. E. 1990 The competitive advantage of nations Harvard Business Review 68 2 73 93 Porter, M. E. (1998). Clusters and the new economics of competition. Harvard Business Review, 76(6), 77-90. Porter M. E. 1998 Clusters and the new economics of competition Harvard Business Review 76 6 77 90 Porter, M. E. (2000). Location, competition, and economic development: Local clusters in a global economy. Economic Development Quarterly, 14(1), 15-34. https://doi.org/10.1177/089124240001400105 Porter M. E. 2000 Location, competition, and economic development: Local clusters in a global economy Economic Development Quarterly 14 1 15 34 https://doi.org/10.1177/089124240001400105 Powell, W. W. (1987). Hybrid organizational arrangements: New form or transitional development? California Management Review, 30(1), 67-87. https://doi.org/10.2307/41165267 Powell W. W. 1987 Hybrid organizational arrangements: New form or transitional development? California Management Review 30 1 67 87 https://doi.org/10.2307/41165267 Powell, W. W., Koput, K. W., & Smith-Doerr, L. (1996). Interorganizational collaboration and the locus of innovation: Networks of learning in biotechnology. Administrative Science Quarterly, 41(1), 116-145. https://doi.org/10.2307/2393988 Powell W. W. Koput K. W. Smith-Doerr L. 1996 Interorganizational collaboration and the locus of innovation: Networks of learning in biotechnology Administrative Science Quarterly 41 1 116 145 https://doi.org/10.2307/2393988 Presutti, M., Boari, C., Majocchi, A., & Molina-Morales, X. (2019). Distance to customers, absorptive capacity, and innovation in high-tech firms: The dark face of geographical proximity. Journal of Small Business Management, 57(2), 343-361. https://doi.org/10.1111/jsbm.12323 Presutti M. Boari C. Majocchi A. Molina-Morales X. 2019 Distance to customers, absorptive capacity, and innovation in high-tech firms: The dark face of geographical proximity Journal of Small Business Management 57 2 343 361 https://doi.org/10.1111/jsbm.12323 Ragalado, A. (2013, June 30). Infographic: The world’s technology hub. MIT Technology Review. https://www.technologyreview.com/s/517626/infographic-the-worlds-technology-hubs/. Ragalado A. 2013 June 30 Infographic: The world’s technology hub MIT Technology Review https://www.technologyreview.com/s/517626/infographic-the-worlds-technology-hubs/ Rattner, H. (1967). Produtividade e desenvolvimento. Revista de Administração de Empresas, 7(25), 53-78. Rattner H. 1967 Produtividade e desenvolvimento Revista de Administração de Empresas 7 25 53 78 Saxenian, A. L. (1994). Regional advantage: Culture and competition in Silicon Valley and Route 128. Harvard University Press. Saxenian A. L. 1994 Regional advantage: Culture and competition in Silicon Valley and Route 128 Harvard University Press Schilling, M. (2017). Strategic management of technological innovation (5th ed.). McGraw-Hill. Schilling M. 2017 Strategic management of technological innovation 5th McGraw-Hill Schmitz, H., & Nadvi, K. (1999). Industrial clusters in developing countries-clustering and industrialization: Introduction. World Development, 27(9), 1503-1514. https://doi.org/10.1016/S0305-750X(99)00072-8 Schmitz H. Nadvi K. 1999 Industrial clusters in developing countries-clustering and industrialization: Introduction World Development 27 9 1503 1514 https://doi.org/10.1016/S0305-750X(99)00072-8 Souza, T. (2019, June 17). Vale do Silício: Inovação muito além da tecnologia. Sebrae. https://sebraers.com.br/momento-da-empresa/vale-do-silicio-inovacao-muito-alem-da-tecnologia/ Souza T. 2019 June 17 Vale do Silício: Inovação muito além da tecnologia Sebrae https://sebraers.com.br/momento-da-empresa/vale-do-silicio-inovacao-muito-alem-da-tecnologia/ Utterback, J. M., & Abernathy, W. J. (1975). A dynamic model of process and product innovation. Omega, 3(6), 639-656. https://doi.org/10.1016/0305-0483(75)90068-7 Utterback J. M. Abernathy W. J. 1975 A dynamic model of process and product innovation Omega 3 6 639 656 https://doi.org/10.1016/0305-0483(75)90068-7 Vanhaverbeke, W., Belderbos, R., Duysters, G., & Beerkens, B. (2015). Technological performance and alliances over the industry life cycle: Evidence from the ASIC industry. Journal of Product Innovation Management, 32(4), 556-573. https://doi.org/10.1111/jpim.12253 Vanhaverbeke W. Belderbos R. Duysters G. Beerkens B. 2015 Technological performance and alliances over the industry life cycle: Evidence from the ASIC industry Journal of Product Innovation Management 32 4 556 573 https://doi.org/10.1111/jpim.12253 Zaheer, A., Gözübüyük, R., & Milanov, H. (2010). It’s the connections: The network perspective in interorganizational research. Academy of Management Perspectives, 24(1), 62-77. https://doi.org/10.5465/amp.24.1.62 Zaheer A. Gözübüyük R. Milanov H. 2010 It’s the connections: The network perspective in interorganizational research Academy of Management Perspectives 24 1 62 77 https://doi.org/10.5465/amp.24.1.62 Zahra, S. A., & George, G. (2002). Absorptive capacity: A review, reconceptualization, and extension. Academy of Management Review, 27(2), 185-203. https://doi.org/10.5465/amr.2002.6587995 Zahra S. A. George G. 2002 Absorptive capacity: A review, reconceptualization, and extension Academy of Management Review 27 2 185 203 https://doi.org/10.5465/amr.2002.6587995 Zhang, G., Tang, C., & Qi, Y. (2020). Alliance network diversity and innovation ambidexterity: The differential roles of industrial diversity, geographical diversity, and functional diversity. Sustainability, 12(3), 1041. https://doi.org/10.3390/su12031041 Zhang G. Tang C. Qi Y. 2020 Alliance network diversity and innovation ambidexterity: The differential roles of industrial diversity, geographical diversity, and functional diversity Sustainability 12 3 1041 https://doi.org/10.3390/su12031041