Search results

The present study seeks to answer the following research question: what is the profile of the academic production related to the interorganizational networks in the period between 2006 and 2016? Thus, this study aims to characterize the academic production about the subject interorganizational networks available in national journals with Concept “A” (Qualis Capes), in the period between 2006 and 2016.

This paper uses national journals with Concept “A” of the Qualis classification (2016) for journal selection. In total, 12 “A” concept journals were identified. However, it was decided to analyze ten of them. The procedures suggested by Crossan and Apaydin (2010) for conducting bibliometric studies were adopted. It has been identified that 77 articles were published in eight journals. The R 3.3.2 and R Studio 1.0.136 software were used. The IGRAPH 0.5.5-2 extension (package) was used to analyze graphs and co-authorship networks (Csárdi and Nepusz, 2006). This extension is able to manipulate networks with millions of vertices and edges and provides a series of functions to analyze the properties of social networks, such as subnetworks, intermediation, centrality, among other characteristics (Csárdi and Nepusz, 2006). Correspondence analysis (CA) was also performed. CA is a multivariate exploratory technique that converts a data matrix into a graphical representation, so that rows and columns are represented by points in a graph (Greenacre and Hastie, 1987). This extension is dedicated to the multivariate analysis of data and allows the manipulation of different types of variables (quantitative or categorical). In the present research, multiple CA (MCA) was applied – indicated when the elements are described as categorical variables (Lê et al., 2008). The characteristics considered for carrying out MCA were the “main term”, “research approach”, “type of research”, “constructs” and “research strategies”. By using the FactoMineR 1.34 extension, the hierarchical clustering on principal components (HCPC) function was used (Husson et al., 2007; Lê et al., 2008). This function allows creating clusters from the characteristics of the articles analyzed and highlights the justifications for the groupings created. The function allows forming as many clusters the researcher wishes, being of its attribution to analyze a division that best represents the characteristics of the data (Husson, Josse, and Pagès, 2010). Husson et al. (2010) suggest that an analysis should be performed from the hierarchical tree, thus the number of clusters can be defined considering the overall appearance (or shape) of the tree formed. At last, a word cloud was created using the Wordcloud 2.5 extension (Fellows, 2013). The noticed advantage of using this extension is that it does not separate the terms that form a keyword when generating the cloud. It has been used for the keywords of the 77 articles analyzed; however, it has been decided to keep those that presented frequency greater than or equal to two. By avoiding occasional terms, a more intelligible cloud was obtained.

The present study was not able to verify if the journals analyzed by Andrighi et al. (2011) have influenced others to publish on the subject, as suggested by the Bradford’s Law. The standard “success breeds success”, suggested by the Bradford’s Law, was not confirmed. The so-called nuclear zone (Brookes, 1969; Novaretti et al., 2015) is composed of the journals Brazilian Administration Review (BAR), Revista de Administração Contemporânea (RAC), Revista de Administração Pública (RAP) and Revista Brasileira de Gestão de Negócios (RBGN). The journal RAC stands out for having been the one that has increased its annual average of publication in relation to the theme, when compared with the findings of Andrighi et al. (2011). The journals BAR and RBGN stand out because they are in the nuclear zone, even though they were not considered in the work of Andrighi et al. (2011). It should be noted that all the analyzed journals have in common the fact of addressing the themes of management and administration and, more specifically, making room for the “competitiveness” and “cooperation” constructs. These constructs are related to the theme of networks and were the most recurrent in the articles analyzed. “Cooperation” (29), “competitiveness” (27), “knowledge” (12), “learning” (6) and “trust” (3) were the “constructs” used to compose the 77 articles analyzed. In turn, “network” (49), “alliance” (18) and “cluster” (9) were the “main term” most used in the articles. This implies that the topic of cooperation is more linked to a vision of strategy. As occurred in the research of Andrighi et al. (2011), the term “network” is the most recurrent; in addition, the growth of space obtained by the term “alliance” stands out. The terms “network” and “alliance” were the most used by the articles, being predominant in 87 per cent of the research. In the present research, the predominance of the term “network” may have occurred because its concept is broader and it is used in the literature in different ways, even in contradictory ways (Andrighi et al., 2011; Schommer, 2001). In turn, the term “alliance” may have been recurrent because it has a wide dispersion of published issues, such as governance structure, cooperation, knowledge transfer and trust (Lima and Campos Filho, 2009). By using the HCPC function of the FactoMineR extension, the articles were grouped according to their characteristics, and then three clusters were formed. By analyzing the generated results, it is assumed that the division into three clusters was the one that best represented the data. Cluster 1 is characterized by descriptive, quantitative, half documentary and half survey research studies, being “cluster” the main term. Cluster 2 is characterized by exploratory case studies with qualitative–quantitative analyzes. Cluster 3 is characterized by theoretical tests. The Zipf’s law points out that a small group of words occurs many times; however, when considering the most recurrent words Networks (9), Strategic Alliances (8), Cooperation (8), Interorganizational Networks (8) and Alliances (6) show that they were present in only about 10 per cent of the works. Lotka’s Law, which states that few authors publish much and many authors publish little, was not confirmed. The authors who presented the highest number of publications, T. Diana L. v. A. de Macedo-Soares (6); Jorge Renato Verschoore (6); Alsones Balestrin (5); Douglas Wegner (4); Humberto Elias Garcia Lopes (4), participated in less than 10 per cent of the works. Thus, the authorship was characterized by many researchers publishing few works, what can be an effect of the behavior of these authors, who prefer to publish in network. The centrality of the relations between the authors was analyzed and, in addition, the intermediation points of the network were identified. The present study also analyzed all the references used by the 77 articles that compose the study. The main author of each of the references used was identified. Among the 30 identified authors, Yin and Hair Jr. stand out for books related to fundamentals and research methodologies. Borgatti and Eisenhardt developed research on the topic of interorganizational networks and also created works for methodological foundations. Powell was the most frequently mentioned author (28) and had more different works referenced (9). Powell stands out for the production of articles published in periodicals, not books. Porter’s situation is the opposite. Most of the quotations made to the author come from his books, especially the work “Competitive strategy” (Porter, 1980). All authors identified are foreigners, with the exception of Balestrin. Marshall, Polanyi, Granovetter and Williamson are authors of works considered seminal, being them, respectively, “Principles of economics” (Marshall, 1890), “Personal knowledge: towards a post critical philosophy” (Polanyi, 1958) and “The strength of weak ties” (Granovetter, 1973) and “Markets and hierarchies, analysis and antitrust implications” (Williamson, 1975).

Like all research, it has limitations. The first one derives from the selection criteria of the periodicals to be analyzed. The cut referring to the journals of greater impact excludes most of the national articles. These studies may contain important contributions to the knowledge of the national publication profile. In addition, the choice to analyze the journals disregards other types of work, such as books, scientific events, dissertations and thesis and reports. The choice of articles published in journals is based on the fact that these are a “certified knowledge”, as the studies are peer-reviewed, and in the case of the Qualis “A” stratum, a review of exogenous quality is supposed on this production. Despite its flaws, this system can be considered reliable to evaluate scientific knowledge (Bedeian, 2004; Shugan, 2007). The analysis of the most recent articles may have been hampered by a temporal issue. In addition, the choice of keywords, a necessary step, leaves out other studies. Another limitation refers to the fact that the articles have been analyzed and classified by the authors, which presupposes the use of their value judgments, at least to some extent. Other limitations refer to the bibliometric techniques employed. The main authors referenced in the studies were demonstrated, that is, those authors who have been used as a theoretical reference for studies of interorganizational networks. However, the circumstances under which these citations occurred were not analyzed. For example, an author may be quoted to use the contribution of his/her study, to be criticized or just to be another reference in the text. The lack of this analysis can be considered a fragility of the study.

This text was started talking about the dispersion of the studies on networks in the country. Previous work has been used, theoretically and empirically demonstrating this fact. Zipf’s Law applied to bibliometrics, as described by Guedes and Borschiver (2005), Novaretti et al. (2015) and Pao (1978), was not confirmed in this study, which seems to be an indicative fact that the research on this theme in Brazil presents fragmentation as an intrinsic characteristic. That is, it must remain fragmented, as this would be its own way to evolve. This is evident especially when comparing the study of Andrighi et al. (2011) and its results. With several but continuous temporal cut-outs, and the same keywords, the maintenance of this dispersion is evident. This is also a contribution of this study.

The study contributed to updating the research profile, mainly after the triennium 2013-2015 of Qualis Capes’ evaluations. It also added to the mapping of recent Brazilian academic production related to interorganizational networks, completing studies by Alves et al. (2013), Andrighi et al. (2011), Balestrin et al. (2010), Cunha and Carrieri (2003) and Mascena et al. (2013). Thus, it is believed that the research reached the proposed objectives, despite its limitations.

The present research is also justified by helping to understand the subject being useful for researchers, educators and students, in general, in the task of demonstrating gaps and opportunities of future researches and collaborating with the elaboration of a research agenda (Baumgartner and Pieters, 2003). The work has updated bibliometrics on the subject and allows comparisons with previous bibliometric studies (Alves et al., 2013; Andrighi et al., 2011; Balestrin et al., 2010; Cunha and Carrieri, 2003; Ferreira et al., 2014; Lima and Campos Filho, 2009; Mascena et al., 2013). It is believed that the present study differs from the others because of the analysis performed, the way the data were treated, with techniques that are rarely used simultaneously, going beyond the descriptive statistics.

This paper aims to examine the interrelation between two innovating strategies (product and process) on total factor productivity (TFP) growth and the dynamic linkages between these strategies, for Colombia. The authors first explore whether ex ante more productive firms are those that introduce innovations (the self-selection hypothesis) and if the introduction of innovations boosts TFP growth (the returns-to-innovation hypothesis). Second, the authors study the firm’s joint dynamic decision to implement process and/or product innovations. The authors use Colombian manufacturing data from the Annual Manufacturing and the Technological Development and Innovation Surveys.

This study uses a four-stage procedure. First, the authors estimate TFP using a modified version of Olley and Pakes (1996) and Levinsohn and Petrin (2003), proposed by De Loecker (2010), that implements an endogenous Markov process where past firm innovations are endogenized. This TFP would be estimated by GMM, Wooldridge (2009). Second, the authors use multivariate discrete choice models to test the self-selection hypothesis. Third, the authors explore, using multi-value treatment evaluation techniques, the life span of the impact of innovations on productivity growth (returns to innovation hypothesis). Fourth, the authors analyse the joint likelihood of implementing process and product innovations using dynamic panel data bivariate probit models.

The investigation reveals that the self-selection effect is notably more pronounced in the adoption of process innovations only, as opposed to the adoption of product innovations only or the simultaneous adoption of both process and product innovations. Moreover, our results uncover distinct temporal patterns concerning innovation returns. Specifically, process innovations yield immediate benefits, whereas implementing both product innovations only and jointly process and product innovations exhibit significant, albeit delayed, advantages. Finally, the analysis confirms the existence of dynamic interconnections between the adoption of process and product innovations.

The contribution of this work to the literature is manifold. First, the authors thoroughly investigate the relationship between the implementation of process and product innovations and productivity for Colombian manufacturing explicitly recognising that firms’ decisions of adopting product and process innovations are very likely interrelated. Therefore, the authors start exploring the self-selection and the returns to innovation hypotheses accounting for the fact that firms might implement process innovations only, product innovations only and both process and product innovations. In the analysis of the returns of innovation, the fact that firms may choose among a menu of three innovation strategies implies the use of evaluation methods for multi-value treatments. Second, the authors study the dynamic inter-linkages between the decisions to implement process and/or product innovations, that remains under studied, at least for emerging economies. Third, the estimation of TFP is performed using an endogenous Markov process, where past firms’ innovations are endogenized.