Etzkowitz and Ranga (2010) explain that the use of the Triple Helix may be useful to drive the evolution of less developed regions. The model details the path of that evolution, allowing for the adaptation of successful practices to the circumstances of such regions. Thus, there is support for comparing both the Brazilian and Chilean clusters, which is the procedure employed in this research effort. The Triple Helix structure is also useful in order to analyze the emergence of innovation and the development of regions and businesses. In this matter, Ranga and Etzkowitz (2013) observe that innovations do not come from the dyad business-government, as occurred in industrial societies, but from a triad within the current knowledge economy.

Zheng and Harris (2007) conclude that the Triple Helix model is relatively recent and more research required to fully understand the consequences. Yuwawutto et al. (2010) point out to its importance to developing countries, emphasizing its power to bring efficiency and competitiveness to firms. Cunha and Neves (2008) have applied the Triple Helix in a cluster in Brazil and concluded that joint actions among academia-business-government are already yielding positive results to competitiveness.

There exists a set of institutions that can be classified as hybrid. They bring together elements of industry, academia and government and may or may not be engaged in R&D. Some examples in this category include multidisciplinary research centers, the consortia between industry and academia toward research, university offices that work on technology transfers, research labs that belonging to firms, support institutions such as science parks and incubators, institutions that provide financial support to start-ups, angel investor networks and funds that supply capital to start-up companies (Etzkowitz & Ranga, 2010).

Papagiannidis, Li, Etzkowitz, and Clouser (2009) have reached a conclusion about the convenience of the Triple Helix model regarding the analysis of business alliances and clusters. As universities become progressively more engaged in entrepreneurial activities in addition to their existing research and teaching role, they have become resource providers to businesses. The role of government has been changing as well: in addition to its regulatory action, it has been promoting innovation, facilitating greater flexibility within the legal setting via tax breaks, loans and grants.

Lundberg (2013) and Todeva (2013) confirm the importance of the government in supporting innovation and entrepreneurship through the facilitation of investments in activities that generate knowledge. Additionally, in the studied cases, the rotation of roles between government, industry and the university sector has been confirmed. This situation is the apex of cooperation among the helices.

The importance of the Triple Helix model to innovation, entrepreneurship and development is recognized by Carayannis and Campbell (2012). They proposed a model that extends the Triple Helix to a Quadruple and Quintuple Helix model, also comprising media, culture and civil society, as well as the dimension of the natural environment, respectively drawing on the ideas of Etzkowitz and Leydesdorff’s (1995, 2000). Although we acknowledge this evolution, the preferred option was to use the Triple Helix as it has been further consolidated and thus possesses greater operational viability than Carayannis and Campbell´s (2012) proposition.

We also acknowledge the existence of criticism regarding the Triple Helix model; for example, Power and Malmberg (2008) view as inappropriate the emphasis placed on scientific aspects in Triple Helix accounts of innovation systems over social and economic processes operating within global knowledge frontiers. Even traditional Triple Helix model authors have suggested that globalization erodes local university-industry-government relations and thus can be expected to have had an increased differentiation in national systems since the mid-1990s. This globalization effect is more pronounced in developed countries than in developing ones (Ye, Yu, & Leydesdorff, 2013).

Recently there have been publications that have further developed ideas regarding the Triple Helix model, such as that by Etzkowitz (2012), in which the importance of permeability across academic–industry–government boundaries is discussed. The present paper has not employed such advanced notions of the model.

2.2 – Model developed by Zaccarelli et al. (2008)

Zacarelli et al.’s (2008) model aims at explaining the origin of cluster competitiveness through the presence and respective intensity level of 11 factors that are explained in the following sections. To each one of these factors a metric is proposed. This metric indicates whether each factor is present in the cluster and its intensity.

2.2.1 – The 11 competitiveness factors

Zaccarelli et al.’s (2008) model proposes 11 factors in order to analyze business cluster competitiveness: (1) Geographic concentration; (2) Scope of viable and relevant businesses; (3) Firm specialization; (4) Balance without privileged positions; (5) Complementarity due to by-product utilization; (6) Cooperation among cluster firms; (7) Selective substitution of firms; (8) Uniformity in technological prowess; (9) Community culture adapted to the cluster; (10) Evolutionary character through new technology introduction; and (11) Result-oriented strategy originating in the cluster. The existence of the nine first factors is viable only with self-organization. However, for factors (10) and (11) to occur, the cluster must have its own governance. Table 1, elaborated by Sarturi et al. (2016), displays the operational definition of Zaccarelli et al. (2008) for the 11 competitiveness factors employed to study the wine clusters in Brazil and Chile and Zaccarelli et al.’s (2008) explanation regarding the relationship of each competitiveness factor in relation to cluster competitiveness (Sarturi et al., 2016).

Table 1
Operational definition of each competitiveness factor and its relationship to competitiveness

Adapted from Sarturi et al. (2016); Zaccarelli et al. (2008).

The business cluster model proposed by Zaccarelli et al. (2008) has its strategic approach based on the conception of supra-enterprise governance, in which the cluster is understood as “the exercise of the strategy-oriented influence of supra-enterprise entities, facing the vitality of the cluster, composing competitiveness and the aggregate result and affecting all of the organizations comprising the supra-enterprise system” (Zaccarelli et al.,2008, p. 52). Although this model has mostly been used in studies published in Brazil and in business clusters located within the country, it has proven useful (Siqueira, Gerth, & Boaventura, 2011; Santos, Boaventura, & Telles, 2012; Pereira, Sarturi, Boaventura, & Polo, 2014). Sarturi et al.’s. (2016) study shows that it can be useful in comparing business clusters in different countries.

The explanation for the existence of business clusters is presented in three steps: (a) Comprehension that clusters are self-evolving systems, capable of having a strategic orientation; (b) The understanding that the constitution of such systems is based upon strategic thinking; (c) Comprehension that the basis for the existence and operation of a cluster reflects observable evidence of the competitive advantage that exists over firms operating outside the cluster.

According to Zaccarelli et al. (2008), two ideas are key to understanding the model: (a) Self-organization; (b) Supra-enterprise governance. The first one, self-organization, has an evolving and spontaneous nature. It results from the systemic effects that arise from the relationships established within a supra-enterprise entity, characterized by the development of increasingly complex connections over time. The second one, governance, works as the supra-enterprise entity, adopting a strategic nature in business clusters.

Cluster competitiveness is based on the 11 aforementioned factors in this article. These, in addition to showing the specific effects generated by the system, suggest that there is a cluster competitive advantage. The factor which is key in characterizing the existence of a cluster is the geographical concentration of firms within the same industry in a contained area. Without this concentration, there is no evidence of the cluster’s existence. Factors 10 and 11 only occur with the presence of supra-enterprise governance and thus possess a strategy-oriented nature.

Fensterseifer and Rastoin (2013) adopt a resource-based multidisciplinary perspective to analyze clusters, as they embody several factors that influence resources and hence their internal process of competitive advantage creation. One of the several research directions that emerge from this work, on the empirical front, is the comparative analysis of wine clusters in different regions and countries. Grimstad (2011) also compares clusters and points to an obvious weakness in the resource-based theory (RBV) through the omission of some important issues for businesses in the future as the basis for gaining competitive advantage will increasingly be based on such socially complex and tacit capabilities not observed in most RBV based competitiveness models.

2.3 – Conceptual model proposed

Figure 1
Examined relationship among Triple Helix elements, competitiveness factors and business competitiveness.
Prepared by the authors

Data was collected from secondary data sources. The source used herein is the research work of Sarturi et al. (2016), which has comparatively examined the competitiveness of wine clusters in Serra Gaucha in Brazil and Valle del Maule in Chile.

Zaccarelli et al. (2008) propose metrics to analyze cluster competitiveness. Sarturi et al. (2016) use these metrics to state that one of their own paper’s contributions is a methodological one, as they propose metrics for the analysis of the competitiveness in agribusiness clusters. The metrics suggested in the Zaccarelli et al. (2008) model serve as guidelines or suggestions, but they may require adaptation or even replacement depending on the specific characteristics of the cluster studied. The results achieved by Sarturi et al. (2016) using these metrics are employed to analyze the influence of Triple Helix factors, as proposed in the objectives of this research.

Sarturi et al. (2016), aiming at the operationalization of their own study, have analyzed the metrics proposed by Zaccarelli et al. (2008), which meet the peculiarities of the study objectives. It is worth noting that, for the analysis of some competitiveness factors, two metrics have been used, such as for competitiveness factor 9. In this case, the result is the average of the cluster performance in both metrics. The metrics used in the analysis are described in the paragraphs that follow. Sarturi et al. (2016) analyzed metrics used by Siqueira et al. (2011) and Santos et al. (2012) in order to propose their own.

Geographical concentration. For this competitiveness factor, Sarturi et al. (2016) used two metrics: the demographic density of the companies and the number of municipalities involved in the cluster. The metric “demographic density of companies” is measured by dividing the number of companies in the cluster by the city area, which had already been applied by Siqueira et al. (2011) (cited in Sarturi et al., 2016). This metric was used to meet the theoretical proposition of Zaccarelli et al. (2008) that the ideal geographical concentration is the largest possible. It is worth noting that, for this study, the metric was adapted to the context of the cluster, and hence the calculation is made by dividing the number of wineries by the total cluster area (Sarturi et al., 2016).

The metric “number of municipalities involved” was used by Santos et al. (2012) and is in line with the theoretical proposition that a cluster must preferably be located in only one municipality (Zaccarelli et al., 2008). In this case, the cluster with the highest demographic density and situated in the fewest municipalities is considered to have the highest level of competitiveness for this competitiveness factor (Sarturi et al., 2016).

Scope of viable and relevant business. To study this competitiveness factor, Sarturi et al. (2016) used the model developed by Fensterseifer (2007), which presented a mapping of the activities involved in the Serra Gaucha wine cluster. This mapping was adopted to analyze the presence of production chain actors, as it is specific to a wine cluster, that is, it meets the peculiarities of the clusters under study, which would not be possible through the use of another classification. According to Fensterseifer (2007), the companies that make up a wine cluster are grape growers, winemakers, producers of seedlings, fertilizers, pesticides and herbicides, barrels, bottles, caps and corks, labels, machinery and equipment, educational and research bodies, funding, regulatory, inspection and coordination entities, specialized public relations companies, specialized trade publications, tourism offices, and food facilities/restaurants. The cluster with the highest level of competitiveness for this competitiveness factor will be that with the largest number of activities in the winemaking chain, as proposed by Fensterseifer (2007, cited in Sarturi et al., 2016).

Specialization of companies. To analyze this competitiveness factor, Sarturi et al. (2016) sought to identify the stages of the wine production chain that companies outsource, since it is understood that the more the activities that are outsourced, the greater the company specialization. A similar metric to this proposal was used by Santos et al. (2012); the difference is that these authors have analyzed the percentage of companies that outsource part of their production. For the identification of the activities, Sarturi et al. (2016) used the classification of Ferreira, Rosina, and Mochiutti (2010), which consider the production wine chain one of the most complex in agribusiness, containing 11 steps after grape-growing (input). To collect this information, Sarturi et al. (2016) visited the websites of the firms that compose both clusters, which, in most cases, provide information about the winemaking process and indicate any outsourced activities. As a result, the cluster that exhibits the largest number of stages within the production chain that are outsourced is considered the cluster with the highest level of competitiveness for this competitiveness factor plea (Sarturi et al., 2016).

Balance with no privileged positions. The evidence of the balance between companies in a cluster is that there are no significant differences in the size of companies (Zaccarelli et al., 2008). Santos et al. (2012), for example, used as a measure of this competitiveness factor the degree of homogeneity of the company size within the cluster. Sarturi et al. (2016) understand that the measure of size for the clusters of agribusiness may be the area for the growth of raw material, in this case grapes. Because of this Sarturi et al. (2016) use as a metric the coefficient of variation of hectares planted by the companies in order to determine whether there is equilibrium among the companies within the clusters. Therefore, the smaller the coefficient of variation of hectares planted, the greater is the balance between the companies in the clusters and consequently the higher the level of competitiveness. Thus, the cluster that presents the lowest variation coefficient will be considered as having the highest level of competitiveness. The data for this competitiveness factor was collected from company websites.

Complementarity through the use of by-products. To analyze this competitiveness factor, Sarturi et al. (2016) investigated the destination of winemaking leftovers by the firms. The winemaking process generates waste such as stalk, grape left overs and seeds (Makris et al., 2007, cited in Sarturi et al., 2016). As a result, the cluster with the highest level of competitiveness is that with the largest number of initiatives to recycle these leftovers.

Cooperation among companies. As the metric for this competitiveness factor, Sarturi et al. (2016) adopt the number of wine cooperatives in the cluster, with the requirement that these consist of members of the cluster itself and that they concentrate their efforts on selling their products, as opposed to cooperatives and associations that focus on local development. This qualification aims at meeting the theoretical proposition of Zaccarelli et al. (2008) that the cooperation among companies consists of the level of spontaneous and voluntary collaboration practiced within the cluster. It is understood, therefore, that the presence of this type of cooperative indicates the existence of relationships of cooperation among the companies comprising a cluster. The cluster that contains the largest number of such cooperatives will be considered that with the highest level of competitiveness for this competitiveness factor.

Selective replacement of firms. For this competitiveness factor, Sarturi et al. (2016) used as a metric the percentage of new businesses in the sector. This metric is similar to that one used in the study by Siqueira et al. (2011). The difference is that the latter study used the absolute number of companies, while Sarturi et al. (2016) applied the percentage of new companies because it is understood that a relative metric can more reliably reflect the competitive difference between both clusters for this competitiveness factor. As in Siqueira et al. (2011), the lack of information regarding the closure of companies made the metric employed in the study more simple than that proposed by Zaccarelli et al. (2008). The cluster with the highest percentage of new companies will be considered to achieve a higher level of competitiveness for this competitiveness factor.

Uniformity of the technological level. To analyze this competitiveness factor, Sarturi et al. (2016) used the metric originally proposed by Zaccarelli et al. (2008), that is, “the presence of inferior technologies,” but with some adaptations. The first adaptation in the Sarturi et al. (2016) study it was not possible to quantitatively analyze the percentage of inferior technologies within the clusters, as proposed by Zaccarelli et al. (2008), given the absence of information in this regard. Considering this absence, Sarturi et al. (2016) have qualitatively analyzed only the presence or absence of inferior technologies. The second adaptation was to divide the analysis of the technologies used into two categories. The first category refers to technologies used in grape-growing activity, while the second refers to the technology used in the winemaking process. In this competitiveness factor, the cluster that indicates the greatest uniformity in these two categories was considered to have the highest level of competitiveness.

Community culture adapted to the cluster. For the analysis of this competitiveness factor, Sarturi et al. (2016) adopted two metrics: the percentage of individuals in the region associated with the cluster and the earliest date of grape-growing activity in the area. The first metric had already been used by two previous studies: Siqueira et al. (2011) and Santos et al. (2012). The second metric was proposed in order to consider the peculiarities of an agribusiness cluster because, as in the case of this study, the wine culture in both clusters has developed due to the process of colonization. For this competitiveness factor, the cluster with the highest percentage of individuals associated with it and the earliest date of grape-growing activity will be considered as having the highest level of competitiveness.

Evolutionary nature due to the introduction of new technologies. The metric used to analyze the competitiveness factor was the number of institutions that operate in the cluster supporting technological research and development. This metric was adopted because it was understood that these institutions may contribute to the development and performance of governance. For this competitiveness factor, the cluster with the highest number of institutions of this nature will be considered as that with the highest level of competitiveness.

Cluster-oriented result strategy. For this analysis, Sarturi et al. (2016) used two metrics: first the registration of the geographical indication (GI), and second the number of exporting firms. The first metric attempts to measure the effort to differentiate products developed within clusters, because the use of geographical indicators offers a strong suggestion of a potential differentiation of products from a specific region (Skuras & Vakrou, 2002, cited in Sarturi et al., 2016), and the second metric refers to efforts regarding market expansion. According to the Ministry of Agriculture (2012), cited in Sarturi et al. (2016), GI registration is assigned to products or services that are characteristic of their place of origin, assigning them reputation, intrinsic value and identity, which distinguishes them from similar products or services available on the market. There are two modalities of the GI: “indication of origin” and “denomination of origin (DO).” The cluster with the highest level of competitiveness in this competitiveness factor will be that with the oldest record of GI, the largest territorial coverage and the greater number of exporting companies.

3.2 Analysis

The analysis is qualitative. Each one of the competitiveness factors proposed by Zaccarelli et al. (2008) is analyzed in terms of how they are influenced by universities, industry and government. To analyze this influence, the results obtained by Sarturi et al. (2016) results are examined in detail with the objective of understanding which helix is influencing the competitiveness factor. How each helix influences the factor is also examined in the terms proposed in the study objectives. The present study uses logical validity (Gibbert, Ruigrok, & Wicki, 2008; Yin, 1994) in analyzing the causal relationships between variables and results. Thus, we endeavor to present a plausible causal argument with logical reasoning that is both powerful and compelling enough to defend the research conclusions.

The analytic effort conducted in this research is presented in a scheme in figure 2 as follows. It is important to note that there are two levels of analysis: the First Level compares how each one of the helices influences the competitiveness factors of each cluster. The Second Level compares the clusters in terms of how they are influenced by the helices.

Figure 1
Examined relationship among Triple Helix elements, competitiveness factors and business competitiveness

At the first level of analysis we selected those factors that are clearly influenced by the three helices. The competitiveness factors were subsequently categorized into two groups: group A includes those factors influenced by one or two helices, while group B is comprised of factors influenced by the three helices. Such a classification was implemented by taking into consideration episodes and institutions that might have suggested the influence of a particular helix on the competitiveness factor.

The first analyzed element for examining the influence of helices on competitiveness factors was the presence of episodes and institutions that might suggest such influence. Diverse sources that may or may not confirm it have also been consulted. Our paper has attempted to go beyond Sarturi et al.´s (2016) observations on competitiveness factors by introducing the use of the Triple Helix theory. However, found evidence should not be considered conclusive and thus represents a methodological constraint in this paper. Such constraints however do not invalidate the study since it possesses an exploratory nature as a result of the aforementioned reasons.

Once the presence of episodes and institutions was verified, the existence of the influence exerted by helices on the analyzed competitiveness factor was consequently observed. Each competitiveness factor in group A was then analyzed regarding the influence coming from the helices and their respective explanations are listed. A similar procedure was carried out in group B. Then, the analysis once again explored the competitiveness factors and the influence from the helices was compared to the results of Sarturi et al. (2016) about which cluster (Brazilian or Chilean one) shows higher competitiveness for that factor. An attempt, based on these results, was made to explain how each helix influences the competitiveness factor.

The presence of episodes and institutions has been discussed in the literature even when it has not been corroborated by other sources (triangulation). Sarturi et al. (2016) mention Marshall (1890), who suggests that the presence of episodes such as the concentration of firms specialized in different steps of the production process would allow for the occurrence of externalities. They also mention the presence of activities related to by-product reutilization to point out the existence of complementarity within the cluster; a similar observation to that of Zaccarelli et al. (2008), who cite the movie industry in Hollywood as evidencing complementarity due to the presence of business ventures. Sarturi et al. (2016, p. 200) also state “that the presence of this type of cooperative indicates the existence of relationships of cooperation among the companies in the cluster”.

Other published studies may also be highlighted, such as that undertaken by Asheim and Coenen´s (2005), which agrees with the idea that a top-level university and scientific parks are essential foundations for the growth of a cluster. Albino, Carbonara, and Giannocaro (2006) also use the presence of specialized services to point to the occurrence of learning processes in the area. There is also the work of Affuso, Capello, and Fratesi (2011), which discusses the lack of opportunity to obtain any direct measure of the industrial competitive strategies, an indirect approach is adopted with the presence of certain factors that may be observed.

It is possible to group into two different categories the 11 competitiveness factors proposed by Zaccarelli et al. (2008): (a) Group A, which includes those factors influenced by one or two helices; (b) Group B, comprised of factors influenced by the three helices.

4.1.1 Triple Helix influence on each one of the competitiveness factors – Factors influenced by one or two helices

The following list outlines the factors which are influenced by one or two helices - (1), (3), (4), (5), (7), (8) and (9) – with the respective explanations.

(1) Geographical concentration: in addition to the clear presence of industry-related issues, such as the existence of firms and the presence of business associations in both clusters, that are considered important in the literature (Beebe et al., 2013), it is possible to argue that the presence of research and teaching institutions contributes to intensify the geographic concentration even further. However, it is not possible to assert that the presence of government contributes to the existence or reinforcement of this factor. Of course, there is always the possibility of indirect influence of government through funding research and teaching institutions, as can be observed in the literature (Cusmano, Morrison, & Rabellotti, 2010). Incentives from the municipalities through the offering of tax breaks to business is one such example. Nevertheless, for the present study, there is not enough information to characterize that the possible tax breaks offered were relevant to geographic concentration.

(3) Firm specialization: in terms of this factor, the government extends its influence in the larger context as seen in the number of different existing businesses. This can occur in either a form of taxation that may inhibit or spur new firm creation and/or activities in the value chain that can be vertically integrated. In the literature, in Chilean clusters the occurrence of vertically disintegrated firms in Chilean clusters mean more specialized was more often observed in the case of subsidiaries (Giuliani & Bell, 2005).

(4) Balance without privileged positions: it is not possible to assert government and academic influence in the existence and intensity of this factor. Thus, only the presence of industry is evident here. It was observed in the literature, that in one Chilean cluster innovation-related knowledge is transferred within clusters in a strikingly uneven and selective way (Giuliani, 2013).

(5) Complementarity due to by-product utilization: again, for this factor it is not possible to assert a government presence. In addition to the industry, there could be an argument in favor of academic presence – through the creation of new winemaking processes and techniques – that could spur new firm creation. However, in the context of the present study, no information confirming the presence of academia could be obtained.

(7) Selective substitution of firms: this factor seems to be exclusively industry-related, because neither the presence of government or academia was verified here. The university sector can influence the creation of new firms, but not their replacement.

(8) Uniformity in technological prowess: this is another factor on which there seems to be only the

influence of industry. Academia can influence in the opposite direction, because often new technologies are created in the university environment which can instead lead to a technological imbalance instead. As mentioned in factor (4), in one Chilean cluster innovation-related knowledge is transferred in clusters in a strikingly uneven and selective way (Giuliani, 2013).

(9) Community culture adapted to the cluster: this factor could only be positively verified in the presence of industry.

As all of Zaccarelli et al.’s (2008) competitiveness factors require self-organization to occur, industry is likely to be involved in all of them

4.1.2 Triple Helix influence on each one of the competitiveness factors – Factors influenced by the three helices

The factors influenced by the Triple Helix are: Factor 2, Scope of viable and relevant businesses; Factor 6, Cooperation among cluster firms; Factor 10, Evolutionary character through new technology introductions; and Factor 11, Result-oriented strategy originating in the cluster. A brief explanation of each Triple Helix influence on these four factors follows.

(2) Scope of viable and relevant businesses: in terms of this factor, the influence of the Triple Helix is justified by the fact that, in addition to the evident need for industry influence, government contribution facilitates or impedes new firm creation. Academia is required to educate the workforce for a variety of activities.

(4) Cooperation among cluster firms: in addition to lack of cooperation viability in the absence of business, the creation of cooperatives cannot exist without facilitation from government. Often, the creation of cooperatives brings together the university and/or local teaching institutions. In truth, there is a second dimension of academic influence because these entities are state-owned or, even when private, receive resources from government as indicated in other factors.

(10) Evolutionary character through new technology introduction: new technologies are often created within universities which then transfer knowledge to industry. Thus, there is industry involvement because for this factor to be present, self-organization and supra-enterprise governance are needed. Here, the indirect influence of government can be highlighted, as in factor 6, because universities and teaching institutions, even when private, receive resources from government.

(11) Result-oriented strategy originating in the cluster: in addition to the self-organization and

supra-enterprise governance needed to make the presence of this factor viable, the metrics proposed to this factor – geographic indication and number of export firms – depend upon government involvement. In the case of the former, government must set the rules, while for the latter government bodies are required support the promotion of exports. The university sector is involved in both cases, indicating the parameters for regulation and supplying a qualified workforce to firms that export.

4.1.3 – Triple Helix influence in the proposed levels of analysis

The factors influenced by the Triple Helix are: Factor 2, Scope of viable and relevant businesses; Factor 6, Cooperation among cluster firms; Factor 10, Evolutionary character through new technology introductions; and Factor 11, Result-oriented strategy originating in the cluster. A brief explanation of each Triple Helix influence on these four factors follows.The factors influenced by the Triple Helix are: Factor 2, Scope of viable and relevant businesses; Factor 6, Cooperation among cluster firms; Factor 10, Evolutionary character through new technology introductions; and Factor 11, Result-oriented strategy originating in the cluster. A brief explanation of each Triple Helix influence on these four factors follows.

Table 3
Presence of teaching and research institutions in the analyzed clusters

Adapted from Sarturi et al. (2016).

In the case of government, its influence is more evident in factor 10. Some of the institutions are state-owned in Brazil, for instance IFRS, EMBRAPA, and FEPAGRO. At the second level of analysis, which compares how the two clusters are influenced by the helices, the results are as follows: (a) the Brazilian cluster has an advantage in factors 1, 6, and 7; while (b) the Chilean cluster sees an advantage in factors 2, 4, 10, and 11 (Sarturi et al., 2016).

From these results, it is possible to discuss the cause of the advantages evident in the Brazilian cluster in factor 6 and the Chilean cluster in factors 2 and 11, as well as factor 10 where both clusters are tied. All of these factors - 2, 6, 10, and 11 - are influenced by the Triple Helix. In the case of factor 6, the advantage of the Brazilian cluster is due to the existence of a higher number of cooperatives than found in the Chilean cluster. A larger number of cooperatives is thus understood as representative of a higher level of cooperation.

In the Valle del Maule, there is the Loncomilla cooperative that is concentrated with more than 100 associates, mostly small- and medium-sized producers. At Serra Gaucha, cooperatives with a similar profile were found. One of the examples is Nova Aliança, with approximately 800 associated families; Aurora, with more than one thousand associated families, and Pompeia, with approximately 260 associated families (Sarturi et al., 2016).

Thus, the Brazilian cluster tends to show a higher level of cooperation than its Chilean counterpart once it is comprised of a larger number of cooperatives (Sarturi et al., 2016). The existence of cooperatives is mostly influenced by industry, allowing for the conclusion that the influence of this Helix element on this factor is stronger in the Brazilian cluster than in the Chilean.

In the case of factors 2 and 11, in which the Chilean cluster has an advantage, and in factor 10, in which there is a balance between both, the following was observed:

- Factor 2. Scope of viable and relevant businesses. The Chilean cluster seemed to be more competitive in this factor, since there are no makers of bottles and corks at Serra Gaucha (Sarturi et al., 2016). Wilks (2006) and Sarturi et al. (2016) highlighted in a previous investigation that there are only two large-scale bottle suppliers in Brazil, with the beer industry as the main customer. In the case of corks, these are supplied by five Brazilian makers which produce them using raw material imported from Portugal and Spain. This item accounts for a large cost outlay for Brazilian wineries.

- Factor 10. Evolutionary character through new technology introduction. Chart 4 contains a list of institutions that run activities related to research and technological development. Both clusters are assessed as similar for this item, although the Chilean may be considered superior due to the slightly larger number of institutions involved (seven versus six for the Brazilian).

- Factor 11. Result-oriented strategy originating in the cluster. In both metrics utilized to analyze this factor, the Chilean cluster has an advantage over the Brazilian one. The Valle del Maule cluster has had its origin denomination since 1995, according to the decree 464 (Decreto 464, 1995, cited in Sarturi et al., 2016), which encompasses the cluster as a whole. On the other hand, the geographic indication initiatives for the Brazilian cluster seem isolated, because the origin denomination dates from 2012 and is restricted to the Vale dos Vinhedos and not to Serra Gaucha as a whole. In the same fashion, for the second metric, Valle del Maule seems to be more competitive, because the number of firms that export their products is larger than seen in the Brazilian cluster. It should be observed that the internationalization process of Brazilian wine firms is recent (Sarturi et al., 2016), although exports have recently increased (Instituto Brasileiro do Vinho (IBRAVIN), cited in Bouças, 2016). Brazilian wineries export numbers in 2016, up to October, show an increase of 33% compared to the previous year in the same period. The United States of America and United Kingdom are two of the top three export destinations.

The table below presents the competitiveness factors in relation to the Triple Helix by comparing the Brazilian and the Chilean clusters.

Table 4
Competitiveness factors in the Brazilian and Chilean wine clusters in relation to the Triple Helix

Elaborated by the authors with data from Sarturi et al. (2016)

In the case of factor 2, industry is the predominant Helix element, since new business creation is dependent upon it. However, the role performed by government and the university sector is relevant because the former can incentive business creation through policies, and academia is key in new technology creation due to the fact that it drives new business creation when new technology is transferred to start-up firms. In factor 6, once again industry is the predominant Helix element. However, such initiatives have governmental support and involve the university sector in many instances. Regarding factor 10, the predominant Helix element is academia, with its role in creating and introducing new technologies. Industry plays an important role here as well because it introduces innovation within business processes. Government obviously also plays a role as a regulator in this factor. Industry is once more the predominant Helix element in factor 11. In relation to the metrics for this factor, industry plays a key role in the initiatives of denomination of origin, with the government playing a meaningful role, and the university sector only a supporting one.