Initially, it must be highlighted that the contemporary institutional thought is essentially divided into two main theoretical lines, the first of which is known as Old Institutional Economics (OIE). As stated in Aguilar Filho & Silva Filho (2010), this line helped disseminate the approach of economic analysis, and sought to study economic phenomena employing analytic tools from other sciences, such as psychology, grounding itself on the influence of technologic developments, social and historical facts, and practices of collective norms and values. Old Institutionalism, as it may also be known, is guided by initial thinkers, such as Commons (1959), Mitchell (1967) and Veblen (1974).

The second theoretical line, our focus in this study, is New Institutional Economics (NIE), stemming from the studies of Coase (1937), Williamson (1985) and North (1990). This framework originates from the previous theoretical line, presented earlier, but differs from it on methodology and concepts due to being based on conventional economic thinking, related to understanding of the influence of the institutional matrix on the economic environment (Aguilar Filho & Silva Filho, 2010). Joskow (2003) posits that New Institutional Economics is clearly different from Old Institutional Economics, emphasizing that NIE seeks to better explain economic phenomena so that economic policies are able to better serve social interests (Joskow, 2003).

In this context, New Institutional Economics has been a topic of much debate in the scientific community. It emerged in the 1930s, with Ronald Coase’s foundational article The Nature of the Firm (1937), when contributions in economic theory, which aided and directed theories of today’s NIE, began. As stated by Coase (1937), the firm could be created in an uncertain configuration of manufacturing activities and, for that, the analysis is developed following two subjective coordination models: the market and the firm (Zylbersztajn, 1995).

Coase (1937) discussed the dynamism provided by competition in the market, where the costs or enacting different coordination strategies would diverge enough for one strategy to be clearly preferable to another. As a consequence, these costs, which do not originate from production processes, would depend on the how transactions happen. These costs could have two distinct natures: the costs of finding the relevant prices in the market or the costs of negotiating and making a contract (Zylbersztajn, 1995).

As we have demonstrated so far, the matter of structures of governance has been debated by theoretical lines within New Institutional Economics (NIE), specifically the frameworks of Transaction Costs Theory (TCT) and Measurement Costs Theory (MCT). However, governance structures have also recently been a topic of analysis for further comprehension under the approach of Resource-Based View (RBV) (Argyres & Zenger, 2010).

RBV originated from the will to study firms internally as a way to understand why some firms have a better performance than others. From this perspective, Penrose (1959) showed that firms are heterogeneous and have different performances, and competition arises from that in imperfect markets. The author then defined the firm as a collection of resources, under internal direction, for production of goods and services which are sold in markets for a profit.

Durand et al. (2017) highlight that interest on the competitive heterogeneity between firms motivated studies on their origins. Consequently, there is a research continuum at different degrees of analysis, in which classic predictors of performance variation, be them micro or macroeconomic in nature, become the relevant focal phenomena. Due to the diversity of origins for competitive heterogeneity and to the width of interrelations among such origins, the focus on phenomena grows and leads to development and increase of new paths of research.

Therefore, since Penrose, there have been other author who follow similar thinking, such as Porter (1980; 1985), who sees strategic advantage as stemming from differentiation, cost leadership and focus. Nelson & Winter (1982), in turn, devised the evolutionary economic theory, connected to the emergence of the contingency theory. During the same period, Wernerfelt (1984) analysed firms as a collection of resources in the form of tangible and intangible assets. Assets were only seen in a concrete manner under RBV perspective after the work of Wernerfelt (1984), when the author created an analogy between the traditional view, based in products, related to Porter’s (1985) competitive model, and a resource-based view (Augusto et al., 2013).

Wernerfelt (1984) notes that firms must be analysed in terms of their resources, and not in terms of their product, as indicated in Porter’s (1985) model. Thus, the term ‘resource-based view’ was initially presented in his paper ‘A Resource-Based View of the Firm’ (1984), as an attempt to formulate a theory regarding the competitive advantage based on the resources an organization controls and employs to compete in the market (Barney & Clark, 2007).

Due to this initial development, Wernerfelt (1984) is seen as the first author in specialized literature to employ an RBV approach, inspired by the work of Penrose (Saes, 2009). With Penrose (1959), Wernerfelt (1984), Barney (1991) and Peteraf (1993) as precursors, this strategy-focused approach presents a new theoretical perspective looking at the firm, and not the industry (Foss, 2005; Saes, 2009). This developed the theoretical foundation for RBV and, at the same time, the theory has increasingly been used as grounds to show the relevance of strategic resources in clarifying the boundaries of the firm.

Similar to the works of Saes (2009) and Foss (2005), the search for integration between different strategy approaches is a movement which originated in the 1990s. Durand et al. (2017) points out that in that decade and in the current century there has been an extension of conceptual and theoretical diversification driving the open, flexible and pluralist nature of the field, in such a way that a theoretical framework with several tentative contributions is already presented, including Langlois (1992) and Zylbersztajn (2009) (who discusses integration between TCT and MCT). There are also studies which propose arguments on the grounds of resource capability and logic as a path to complementarity (between TCT and RBV), such as Jacobides & Winter (2005), Argyres & Zenger (2010), Crook et al. (2013), Foss & Foss (2004) Tebboune & Urquhart (2016), Ghozzi et al. (2016), and others such as Argyres (1996), Poppo & Zenger (1998), Leiblein & Miller (2003) and Gulbrandsen et al. (2017). Following the same perspective, this study seeks to discuss this theoretical direction, looking at the interaction of the three theories and discussing the propositions that concern the decision to integrate.

Even though the premises of Transaction Costs Theory (TCT) are widely known, Resource-Based View researchers defend that, surely, RBV is not feasible without empirical data. Nonetheless, as it is presented today, this approach is more explanatory than predictive. In other words, it allows an analyst to tell the ex-post story of a firm’s success, or why its boundaries are set where they are, but it is considerably weaker in regards to predicting future success or future patterns of economic organization (Foss, 2005; Argyres, 1996). To such end, this section will discuss papers that studied the Resource-Based View in relation to determining the boundaries of the firm, seeking to propose the integration between this strategic theory and other theories within New Institutional Economics (TCT and MCT).

Firstly, one must consider institutional aspects which relate Resource-Based View (RBV) and New Institutional Economics (NIE). According do Oliver (1997), considering the meaning of ‘institution’, the accumulation and choice of resources are conditioned to decisions either in the endogenous or in the exogenous sphere of the organisation (Foss, 2005). Similarly, North (1990) states that in this environment, actors must regularly make decisions based on incomplete information and process in using inadequate mental models which result in inefficient paths. That said, choices in the endogenous sphere are guided by the intention of rationality and aimed towards efficiency, effectiveness and profit (Oliver, 1997).

In turn, exogenous factors are strategic aspects of the market, involving the power of buyers and bidders, magnitude of competition and market structure of the product. Such factors, as per Oliver (1997, p. 698), “influence what resources are selected, as well as how they are selected and deployed.” In this manner, decision-making regarding possible alternatives is conditioned, apart from other factors, by the expansion strategy, the availability and the allocation of the resources required for decisions of this scope, as well as by the institutional environment that permeates the organization.

Concerning institutions, Dorobantu et al. (2017) brought once again non-market strategy studies to the foreground, providing a typology based on New Institutional Economics, which summarizes and aims to supersede previous research. The authors' expectation is to offer a more robust and associated set of decisive factors in governance decision-making for companies facing weak institutions, whether locally or internationally. Their governance-based model focuses on three major non-market strategy divisions: organizational, collective and political.

In turn, Argyres (1996), in his empirical study that analyzes some companies' 'make-or-buy' decisions, sets out from the proposition that firms vertically integrate those activities in which they believe they have more production experience (capabilities) than potential suppliers, outsourcing those activities in which they have lower capabilities, except in cases where there is an explicit intention, in the long term, to develop such capabilities internally (Argyres, 1996).

According to Argyres (1996), differently from transaction costs logic, a capabilities-focused approach is unable to make predictions, offering only ex-post explanations. The author found in his research some cases where transaction cost aspects would foresee vertical integration, but the firm opted for outsourcing, presenting a reasoning consistent with the argument of relative capabilities. Also, there were situations in which, at different stages of production, it was possible to find activities carried out internally and outsourced activities, decisions that always had as motivation both transaction costs and determinants related to capabilities (Argyres, 1996).

In posterior research, Argyres & Zenger (2010) argue that their proposal is clear: firms internally govern activities and assets when they possess comparative capability, outsourcing activities when they possess comparative incompetence. Barney (1999, p. 138) affirms that some firms are simply better at doing some things, and, as a result, “the capabilities possessed by a firm and by its potential partners often should have a significant impact on boundary decisions.” In a similar manner, Jacobides & Winter (2005) assert that, when seeking to understand whether or not firms are integrated, it is fundamental to analyze the allocation of their productive capability.

In this context, empirical research on RBV seems to support the logic of comparative capabilities, much like the study by Argyres (1996), described before, which presented several examples of a manufacturing firm where the relative capabilities of the firm seem to be decisive factors of the boundaries. Concerning the connection between RBV and vertical integration, Ghozzi et al. (2016) recently presented a perspective which emphasized the relevance of RBV rationale, establishing a two-step logic explanation. The first consists in knowing ‘why’ a determined governance structure (market, hybrid, hierarchy) was chosen, using a series of analytic items. The second refers to ‘how’ it was chosen, providing more subsidies in relation to the strategic development model, which is based on a combination of both competitive advantage and comparative resources (capabilities). Such a combination of factors indicates the vertical integration model, be it internal development or acquisition, depending on the relevance of the firm's need for superior knowledge.

Following the same thinking, other authors sought to empirically relate skills and specialized experience with vertical integration decisions. Poppo & Zenger (1998) looked into decisions about boundaries using a TCT approach combined with RBV and MCT explanations. The authors’ results indicated that a theory of the firm and a theory of boundary choice would be complex, demanding integration of transaction costs (TCT and MCT) and resource-based views (RBV).

It is noted that, as a result of the boundary choices made via RBV, Poppo & Zenger (1998) found that governance influences the development and transfer of capabilities and knowledge, seeking to understand how capabilities determine boundaries and limits influence capabilities. They also note that when valuable knowledge can be generated by company-specific language and routine development, hierarchy is preferred, but when company-specific routines and language generate impediments to the creation of valuable knowledge, markets are favored, Poppo & Zenger (1998) conclude.

Leiblein & Miller (2003) and Hoetker (2005), focusing on specialized experience, or specialization, examined these aspects in regards to boundary definition. Jacobides & Hitt (2005), on the other hand, analyzed the firm’s boundary setting within the mortgage banking industry, finding that firms with higher productive capabilities in a determined production phase carried out these activities internally. Thus, it can be understood that the basic idea that emerges from the application of resource-based logic in this context is the simple concept that firms determine whether to make or buy through a benchmarking process. As the comparative capabilities move, the limits of the firm are determined (Argyres & Zenger, 2010).

Meanwhile, other knowledge-based researchers, coming from a different perspective, have studied empirical findings previously seen in Transaction Costs Theory, stating that specialized activities are internalized for reasons of greater governance efficiency combined with greater asset specificity within the firm (Monteverde, 1995). They establish that the more specific an activity, the greater the use of firm-specific language and routines, and, therefore, the more efficient its internal governance will be. Poppo & Zenger (1998) claim that these researchers observed that this explanation entirely circumvents the possibility of opportunism, considering vertical integration a ‘creator of positives’ instead of an ‘avoider of negatives.’

Furthering the discussion, Augusto et al. (2013), in seeking complementarity between RBV and TCT, posit that opting for vertical integration from the perspective of learning is in line with the results of Ghoshal and Moran (1996), Poppo and Zenger (1998) and Argyres and Zenger (2010). Augusto et al. (2013) then suggest that these researchers argued that the choice of vertical integration can be based on the company's commitment to the development of skills, aiming at proactivity, learning and collaboration. Opportunism resulting from specific assets does not seem to be, therefore, the exclusive motivation for integration.

Furthermore, Augusto et al. (2013) confirmed Langlois’s (1992) proposition that the capabilities of the market and of organizations, accompanied by governance costs, determine the boundaries of the firm in the short term (Augusto et al., 2013). Still from the perspective of dynamic capabilities (DCs), as discussed in Langlois (1992), Gulbrandsen et al. (2017) dealt with DCs focusing on elements of knowledge, routines and resources of the company. The researchers presented in their results that DCs and TCT are simultaneously indispensable concerns when debating the decision between outsourcing and internalizing activities, pointing out the complementarity of these theories for the comprehension of vertical integration.

Other theorists have sought the complementarity of RBV and TCT, such as Jacobides & Winter (2005), who point out that after a disintegration, integration may occur again when there are more radical changes in new technologies, integrated resources become valuable, with prevalence of transactional practices that are often obsolete, thus leading to a phase of vertical reintegration. These authors thus suggest a complementarity between TCT and RBV, arguing that in order to understand capabilities, it is necessary to look at the roles of transaction costs and scope development. They also find that the distribution of resources (production capabilities) influences the vertical scope and competitive dynamics, and that the change in integration affects competition (Jacobides & Winter, 2005).

Tebboune and Urquhart (2016) used the TCT and RBV theories in order to predict the sourcing strategies employed by Netsourcing suppliers. They found that, for TCT, asset specificity and uncertainty motivated the internalization. In particular, the researchers noted that the uncertainty attribute has a stark influence in the boundaries, especially applying RBV to the analysis, since they found complementary resources in Netsourcing supply strategies as a motivation for internalization. In this situation, if the level of uncertainty linked to these resources is high, there is a tendency for complete vertical integration, aiming for complete control of the service.

Tebboune & Urquhart (2016) found that each theory would not be able to independently explain the full extent of the decisions made by the evaluated companies. However, when these theories are analyzed in a complementary way, their explanatory power can be considerably improved. The authors showed that TCT and RBV were not presented as competing theories to understand the same phenomenon from the same perspective, but acted in a complementary way to explain the same phenomenon from different perspectives (Tebboune & Urquhart, 2016).

Along the same lines, Ghozzi et al. (2016) addressed the complementarity between TCT and RBV in order to understand the influences of the adoption of non-GMO standards in poultry supply chain governance structures in France and Italy. The research presented an important conceptual contribution. Within the TCT framework, the concept of opportunistic behavior was discussed with the insertion of an intermediate degree of opportunistic behavior, whereas the RBV framework indicated a stepwise approach, associating the strategic relevance of activities to potential superior knowledge (Ghozzi et al., 2016).

The authors indicate the complementarity between TCT and RBV to support the analysis of the decision of governance structure, besides verifying that the theoretical relevance is subject to transaction presented as well as to the mode of governance that is being analyzed. This is based on the findings that TCT clarifies the choice of governance in the relationship between producer and processor, while the RBV presents more consistent explanations in the relationships between processors and distributors. Considering this, Ghozzi et al. (2016) suggested that an independent theoretical perspective may not be sufficient for a complete understanding of the firm's boundaries, thus expanding the points of investigation on determining factors which must be complementarily analyzed for this decision.

Going further in the context of integration between TCT and RBV, Gulbrandsen et al. (2017) studied the complementarity between Transaction Costs and Dynamic Capabilities (resources used to extend, modify, alter and/or create common resources), using these approaches to explain decisions regarding vertical integration in the hydroelectric power sector, in relation to the provision of maintenance services and technical upgrades. Employing the TCT approach, the researchers considered transaction costs, specific investments and supplier reliability; regarding the capabilities view, they used the concept of Dynamic Capability (DC), but with a focus on similarity of capabilities, meaning the level at which the company can develop the same resources as the supplier, which are internally necessary (Gulbrandsen et al., 2017).

Presenting their results, Gulbrandsen et al. (2017), when applying the integrated model of transaction costs, capabilities and resources, found an explanation for vertical integration in 63% of the analyzed cases, pointing to a broad explanatory power within the model. In view of this, the authors consider that in addition to its assumptions, TCT should also consider at which level the organization has means to internally develop the necessary capabilities, which is shown to be the impetus for boundary decisions, given that by using existing capabilities, purchasing organizations can save on transaction costs in contractual relations and on their production costs, thus directly impacting on internalization.

Thus, Gulbrandsen et al. (2017) suggest integration between the paradigms of transaction costs and capabilities and resources, because, to the extent that they address each other's gaps, the theories are complementary. The authors posit that their findings provide an empirical basis for the value of using these two approaches with the firm in focus, pondering that each perspective is relevant but insufficient to independently explain decision of boundaries. From their results, Gulbrandsen et al. (2017) point out that more empirical research is needed, which can combine the perspectives of TCT and capabilities and resources, with respect to predictive investigations of the vertical structure (Gulbrandsen et al., 2017).

For Argyres and Zenger (2010), TCT and RBV present quite similar views, in conceptual terms, regarding the definition of the firm's boundaries. Concerning RBV, the authors propose that it is necessary to reflect on how to handle the matter of which resources are complementary in order to create a competitive advantage. Meanwhile, TCT inquires which of these complementary resources will be directed to remain in common ownership of the firm and which will be independently owned. Argyres and Zenger (2010) state that they, along with several other theorists, share the idea that if capability differences between a firm and its potential suppliers play a key role in the firm's governance definitions, it is largely possible that transaction costs are somewhat influential in the origins of such differences.

As a result, Argyres and Zenger (2012), as in their previous papers, argue that the resources are very relevant due to the transaction costs that arise from their generation, purchase, sale and governance. In this perspective, Argyres and Zenger (2010; 2012) show that RBV supports management in clarifying what resources are necessary for the firm's strategy, and TCT has the role of providing information regarding direction in the supply and organization of these resources. Additionally, TCT has as its central concern the matter of which governance options promote the development of which types of capabilities. The authors then advise researchers to look for these associations between capabilities and transaction costs, in addition to which forms of contractual safeguards allow the transfer of tacit knowledge and encourage the development of the desired routines, making it possible to understand the insertion of property rights, which are inherent in the Measurement Costs Theory (MCT) (Argyres & Zenger, 2010; 2012).

Foss (2005) sought more specifically, within this discussion, a path through which the integration supported in his study could be formulated, emphasizing that in the context of economy of property rights (MCT), a theoretical insertion for organizational economics is presented, in a way it would be able to complement and deepen RBV by improving the understanding of resources and adding new theoretical perspectives to the value of resources.

In addition, in regards to strategic choice (RBV), flaws which are clearly related to the transaction can be found. Market failures arise as transaction costs are introduced into the perfect world of Coase theorem (1960). As a result, choices such as organizational arrangements become important, as with positive transaction costs, different arrangements lead to different results in terms of value created (Foss, 2005). Saes (2009) and Foss & Foss (2004) argue that, independently, none of the theories (TCT, MCT or RBV) is capable of bringing together the essential elements of sustainable competitive advantages (SCA), adding that TCT and MCT cannot properly introduce limited rationality, which would make it possible to understand the heterogeneity of firms, while RBV presents difficulty, in regards to methodology, to elucidate the existence of economic organizations.

In the same direction, Foss (2005) clarifies that in this complementarity proposal there is the point that resources are not given, but rather result from saving in operations with transaction costs. For that, resources which are considered equal, in different companies, may be economically distinct, resulting in that businesses do not have the same conditions to protect important attributes. Foss (2005) highlights resources that may warrant advantages for businesses, increasing the creation of value, and that may be appropriated through minimization of transaction costs (Williamson, 1996; Barzel, 1997). The author thus proposes that such resources are relevant origins of heterogeneity and competitive advantages in several industries. However, as per Foss (2005), this point was largely neglected in RBV, considering that this occurs because such resources are only perceived in a scenario with positive transaction costs, while RBV has not yet explicitly adopted the point of view of transaction costs (Foss, 2005).

With regard to identifying and promoting theoretical and empirical developments that integrate multiple theoretical currents, Gans & Ryall (2017) demonstrate progress with value capture theory. This theory allows access to ‘blind spots’ in the main themes to analyze the heterogeneity of performance among organizations, encompassing works such as Porter's ‘five forces’ (1979), Resource-Based View (RBV), and Transaction Costs Theory (TCT). Simultaneously, the value capture model provides a theoretical framework in which key points and empirical tests of these prevailing theories can be evaluated and integrated. By addressing a broad – and potentially measurable – concept of value, the model provides support for integrating and comparing shareholder and stakeholder approaches interested in the organization's objectives.

Saes (2009) indicates this as a relevant and favorable topic for the integration of these theories, which had already been presented by Foss & Foss (2004), but somewhat neglected by RBV in relation to value creation and its capture. These theorists assume the existence of a large differential between the value of the quasi-rents to be generated and the way they will be distributed. In a situation of transaction between different resource owners, there may be loss of rents, leading to dissipation of value and few stimuli for its subsequent creation. In Nickerson (2003), RBV does not account for the costs of internal hierarchies in the organization, allowing a disparity for high-cost strategies of value capture. In addition, with RBV, managers may fail to capture value if their unique resources are not valuable enough for clients (Nickerson, 2003; Saes, 2009; Ghozzi et al., 2016).

This means that the value which a resource-owner can create is subject to the property rights (defined or otherwise) that they own (Foss, 2005). For Saes (2009), including transaction costs gives the strategies suggested by RBV new perspectives, since the choice of the governance structure will lead to a situation of greater (or lesser) appropriation of rents. It is also understood that, in case of difficulty of protection, the structure may tend towards vertical integration.

Following this line, Saes (2009) understands that, in order to integrate these theories (TCT, MCT and RBV), there is the realization that, when considering transaction costs, a strategy that formulates barriers to entry is sustained. Regarding the firm's internal resources (RBV), these can be understood as specific assets and thus studied from the NIE approach (TCT and MCT). Besides that, TCT and MCT consider resources as a set of attributes that define property rights. Therefore, the way attributes are collected to constitute a resource depends on transaction costs (Saes, 2009; Foss, 2005).

Therefore, as suggested by Argyres & Zenger (2010; 2012), research focused on this type of issue certainly leads to discoveries about the relationships between governance choice and capability development. In the same way, this goal will be pursued in section four, for an analytical proposal of complementarity between TCT, MCT and RBV, aiming for the consistent association of these theories in the decision of vertical integration.

In view of the theoretical argument presented, the aim was to establish a discussion of the theories belonging to the New Institutional Economics (TCT and MCT) along with the Resource-Based View Theory (RBV), belonging to Strategic Theory. As both seek to study the performance of the firm, on the one hand the theories on the microanalytical side of the NIE focus on efficiency, on the other hand, the Strategic Theory of RBV focuses on strategies for sustainable competitive advantages. Although they come from different perspectives and work independently, each theory has its own rationale to justify the decision of vertical integration, and so one can seek complementarity, since both have congruent objectives. In the same sense, Foss & Hallberg (2017) emphasized the importance of focusing on the origins of approaches and the potential benefits of combining assumptions from adjacent disciplines, in a highly aware and rigorous manner, as we sought in this essay.

Faced with this, the proposition of complementarity comes from the observation of the proposed literatures of the three theories until now, noting that, according to Foss (2005), none of them independently gives enough support to fully explain the option of integration. Tebboune & Urquhart (2016) consider that the managers responsible for boundary decision making may consider that a greater number of analysis units will help them to make better decisions. This section, as a result, seeks to bring about a discussion which posits that, from the complementarity of their different aspects, the theoretical limitations of each approach to explain vertical integration can be expanded through the union of the three theories (Foss, 2005; Foss & Hallberg, 2017).

The analysis was based on the Transaction Costs Theory (TCT) and was represented by Williamson (1985); the Measurement Costs Theory (MCT), with its creator Barzel (2005); coupled with the Resource-Based View (RBV) approach, introduced by Penrose (1959). By analyzing the proposition of complementarity, in order to understand its different assumptions, we tried to contrast them in what concerns their theoretical and empirical aspects, showing how their different perspectives can seek vertical integration, justifying it and being determinant for it to occur.

It should be noted that the proposition for complementarity of the TCT and MCT approaches, based on the study by Zylbersztajn (2009), is considered for this study. It was also based on the studies by Foss (2005), Saes (2009), Augusto et al. (2017) and Augusto et al. (2018), all of which proposed the complementarity between TCT, MCT and VBR. These propositions were widely elucidated and disseminated by academic research, bringing relevant points to the theoretical formulation of this work. Moreover, as has been discussed, the papers that dealt separately only with the integration of TCT and RBV are considered here, in addition to the theoretical review widely discussed on their complementary aspects.

Considering what has been exposed above, the following thought arises for the theoretical proposition of this paper: ‘the decision of vertical integration for protection of specific (TCT) and difficult to measure (MCT) assets provides only indications, that is, justifications for the adoption of integration, because even in this favorable scenario to adopt vertical integration, if the organization does not have the necessary resources and capabilities (RBV) for integration, it will not have the means to carry it out.’ Moreover, the literature of RBV, in dealing with the theme, also did not clarify how these resources can be determinant for the integration process, that is, essential for it to occur, whether or not based on the justifications of the theories of TCT and MCT.

It is therefore suggested that, to the extent that RBV can support the integration decision, resources and capabilities are essential in this process. In the same manner, one must ask ‘why, in some cases, are organizations that in theory should integrate not integrated in practice?’ In this work, we intended to clarify that one possible explanation for the non-integration is the lack of resources and capabilities.

Thus, it is assumed that in the case of companies where vertical integration is indicated but they do not have the resources and capabilities to do so, they need to look for other organizational forms, such as contracts or act in the market. With the lack of resources, they end up adopting these other forms of governance structures, which can be considered less efficient. Although it is the most efficient form among the viable alternatives, it may not be the best choice for your type of company, and it will not have the protection of resources and competitive advantage sustained by theories, which in the long run may result in inefficient paths.

Then, from the discussion presented so far, the arguments of complementarity on its general theoretical aspects (section 4.1.1) as well as vertical integration (section 4.1.2) have been demonstrated. These justify the following proposition, in which we seek to expose considerations of complementarity in determining vertical integration by means of the constructs of each approach.

Therefore, the proposition, next to the sub propositions in the theoretical scheme formulated, have the following acronyms as correspondences: (A) refers to TCT (Asset Specificity); (B) refers to MCT (Measurement); (C) indicates the Governance Structure (Vertical Integration); (D) refers to RBV (Capabilities and Resources). In view of the above, this study has the following proposition: "Vertical integration is justified when high specificity and difficult-to-measure assets are traded, but its viability is conditioned to the availability of resources and capabilities."

In decomposing this main proposition, given the indications of complementarity in Bacharach (1989) and Suddaby (2010), the interactions between the main categories were raised: (1.1) asset specificity and measurement difficulty level; (1.2) asset specificity; measurement difficulty level; and capabilities and resources. The sub-propositions are in Table 1 and in Figure 2.

Table 1
Theoretical proposition of complementarity for vertical integration

Source: Developed by the authors.

From the relationship presented between the theories on the condition of complementarity for the decision by the integration structure (Table 1), it was possible to propose the formulation of the theoretical scheme by the proposition and sub-propositions (Figure 2).

Figure 2
Theoretical layout for analysis of the complementarity of vertical integration.

Source: own elaboration.

In view of the theoretical scheme (Figure 2), it is indicated that the transaction of high specificity assets (TCT=A) and the difficulty of measurement (MCT=B) only justify vertical integration (C), but the Resource-Based View (RBV=D), points out that the choice of agents also depends on the availability of capabilities and resources. It is emphasized that the idea of correlation between D and C outlined above is due to the proposition that, given that strategic capabilities and resources (D) confer competitive advantage, being in the form of a structure of vertical integration (D), this can foster the development (generate value) and at the same time the protection of the attributes of these capabilities and resources, making the competitive advantage also sustainable.

In this context, the question of how capability and resource constructs influence the choice of vertical integration in institutional arrangements by being integrated with transaction and measurement cost constructs is re-examined. This questioning is in line with studies such as that of Argyres (1996), who realized that, unlike the logic of transaction costs, the approach of capabilities is not enough to make predictions, but only to offer ex-post explanations.

Thus, regarding the interaction of the categories (1.1), asset specificity and measurement difficulty level, it is noticeable that the high asset specificity shows itself as a motivator for vertical integration, as it demonstrates an environment of more uncertainty, commanding vertical integration in an attempt for more control over these specific investments (assets) (Williamson, 1985; Tebboune & Urquhart, 2016; Gulbrandsen et al., 2017). When combined with the difficulty of measuring the attributes of assets and maintaining control over them, the fragility of verification and guarantees between supplier and final buyer is pointed out, which indicates that, by maintaining the contractual relationship, it is not possible to guarantee property rights and protect itself from the risk of loss of value in the transaction (Barzel, 1997; 2005). It is understood, therefore, that in the presence of opportunities to capture quasi-rents or the free-rife effect on the guarantee, it is possible that the conclusion of the contractual relationship may occur; therefore, united, these factors justify vertical integration as the most appropriate governance structure for the firm (Williamson,1985; Barzel, 2005). Consequently, the choice of vertical integration is justified when there are assets of high specificity (TCT), that demonstrate difficulty in having their attributes measured (MCT), together with difficulties in guaranteeing the (economic) rights of these.

However, although they justify verticalization, Argyres (1996) found in its studies some empirical cases where the transaction cost aspects would foresee vertical integration; regardless, the firm opted for outsourcing, with a justification that was consistent with the relative capability argument. Therefore, it demonstrates that only specific assets or the difficulty of measurement may not be independent reasons for integration and the contractual relationship may be maintained, confirming the sub-proposition ‘(1.1). The greater the specificity of the asset and the more difficult the measurement, the more the need to opt for vertical integration is justified, but vertical integration is not determined.’

At the same time, this study of Argyres (1996) identified cases in which, in the different stages of production, it was possible to find activities carried out internally (integration) and outsourced activities, having as motivation both the transaction costs as well as determinants related to capabilities (Argyres, 1996). Recently, Gulbrandsen et al. (2017) strengthened these same findings in their study, considering that apparently in empirical models of TCT there is a lack of elements related to capabilities, which are necessary to explain more fully the decision of the firm's boundaries. Thus, it is understood that the motivations of the theories of transaction costs and measurement costs have empirical validation in addition to showing greater comprehensiveness when complementing each other to explain the integration decision. However, for such explanation, it is also necessary to associate them simultaneously with the presence of capabilities and resources, in order to understand theoretically the aspects that make this decision feasible, that is, to determine the occurrence of vertical integration.

In this sense, the interaction of the three categories (1.2), namely asset specificity, measurement difficulty level and capabilities and resources, is sought. In relation to RBV, there is already the understanding that capabilities and strategic resources are considered in determining which governance structure to adopt, including vertical integration (Augusto et al., 2018; Argyres & Zenger, 2012; Saes, 2009; Ghozzi et al., 2016; Gulbrandsen et al., 2017). From this perspective, it is possible to notice some points. Some capabilities and strategic resources may be derived from what Foss & Stieglitz (2010) point out as path dependence, a stock of specific assets (TCT), accumulated and developed throughout the company's history, which are strategic resources and in this case are related and may be decisive for vertical integration. For the authors, the competitive advantages derive from these stocks of resources/assets specific to the firm, which need to be accumulated internally as manufacturing capabilities, technological knowledge and brand reputation.

Thus, this corresponds to the theory that internal strategic resources may be bundles of specific assets, and in their origins, there may be the presence of transaction costs (Foss, 2005). In the same sense, Jacobides & Winter (2005) suggest a complementarity, stating that for the understanding of capabilities, one should seek the roles of formation of transaction costs. Argyres & Zenger (2012) defend that beyond the understanding, considerations of transaction costs will outline the decisions of the companies about which resources to create, hold and abdicate, analyzed from the perspective of TCT, complementing RBV and TCT.

A few points stand out in this regard, for Argyres & Zenger (2012). There are several specific or specialized investments that would cause specific assets or activities and would demand governance safeguards. However, only some of these investments would generate added value or sustainable competitive advantage for the company, i.e. those assets that are specific and strategic. According to Saes (2009) and Crook et al. (2013), not every specific asset is a strategic resource. Then, those assets that are specific (TCT) and strategic (RBV) (valuable, rare, difficult to imitate or replace) are more related and dependent on vertical integration than those assets that are only specific (Crook et al., 2013).

Furthermore, for these capabilities and resources to be truly strategic, that is, to ensure sustainable competitive advantage, they must respect the conditions of non-mobility and heterogeneity, resulting in a superior income, compared to the competition. In this manner, with regard to the choice of the organizational form of vertical integration, the distinct capabilities and resources (RBV) of the companies should be taken into consideration, seeking sustainable competitive advantage, with the purpose of protecting these strategic capabilities and resources (MCT) (Saes, 2009).

As Argyres and Zenger (2012, p.1648) argue, in a scenario where heterogeneity (RBV) and specific investments (TCT) ‘the prescriptions of the two theories are quite consistent: firms integrate to create and protect value.’ So, when it comes to protection safeguards, the role of MCT is introduced. In this sense, Saes (2009) noted that decisions on how the firm structures itself involve the purpose of protecting and valuing the attributes of resources/capabilities (assets). Among the strategic resources/capabilities, there are those with attributes considered difficult to measure. Consequently, they are difficult to protect and ensure so that they are used only by the company, without being imitated by competitors. Thus, they seek to protect these resources (assets), because they give the firm a competitive advantage, which can become sustainable through the internalization of production (ex-post integration) (Barney, 1991; Ghozzi et al., 2016; Tebboune & Urquhart, 2016; Gulbrandsen et al., 2017).

Augusto et al. (2018) argue that there is a need to internalize production in situations where resources and capabilities, such as technological knowledge, are easily transferred if they are not well protected. As such, there is complementarity between strategic resources (RBV) and measurement (MCT), because within the logic of RBV vertical integration presents itself as a means of protection and control of resources seen as strategic for the creation of income superior to competitors, even if in economic balance (Argyres & Zenger, 2012; Peteraf, 1993). At the same time, Foss (2005) states that, given the costs, owners generally choose to control the ownership rights of the relevant attributes at different levels, and the value that may be appropriate for a resource will reflect this scenario. In a case of where it is difficult to delineate the property rights, vertical integration is sought.

In view of this, specific assets justify vertical integration, but alone they may not be determinant for it to occur (Gulbrandsen et al., 2017). In the same sense, just like TCT, MCT is unable to independently determine vertical integration due to the fact that attributes are difficult to measure. There will be scenarios in which organizations will not have ideal conditions to internalize their production, even if they face an environment of uncertainty (Tebboune & Urquhart, 2016), and know that it is necessary to avoid opportunistic captures, be them from quasi-rents or guarantees.

However, they may adopt verticalization if they observe, and are backed by their analytical bases, that the best way is integration and, at the same time, they also have the means (resources and capabilities) to achieve this internalization. Therefore, they would be faced with strategic conditions that could give them competitiveness, besides protecting and valuing the attributes of assets (resources), reducing transaction costs and achieving efficiency.

Williamson (1985) considers that in addition to transaction costs, it is possible that the comprehension of the adoption of vertical integration also encompasses strategic purpose. Thus, when a highly specific and strategic asset is complementary, vertical integration is more likely (Gulbrandsen et al., 2017). In addition, if the attributes of these assets (resources) that are specific and also strategic have a costly measurement because it is difficult to delineate their property rights, when aiming at the protection and increase in value of these attributes, it will also tend to the integration structure. By identifying this type of asset (resource), the decision-making process becomes more complete, showing that by aligning the theories it is possible to have more predictive power for verticalization.

Therefore, by aligning the constructs of the three theories analyzed, the sub-proposition ‘(1.2). The greater the specificity of the asset, the more difficult it is to measure, and the greater the availability of capabilities and resources, the greater the condition to opt for vertical integration is justified and determined’ is confirmed. Thus, as a complementarity one can think that the reduction of transaction costs (TCT and MCT) can explain which governance structures will be adopted, through the available resources, in an effort to protect and explore the attributes of the (assets) resources.