Technological proximity and the role of joint steering committees on contractual dispute resolution complexity in joint ventures

Technological proximity and the role of joint

steering committees on contractual dispute

resolution complexity in joint ventures

M a s t e r T h e s i s

Jason J. van Heyningen

j.j.van.heyningen@student.rug.nl

University of Groningen

Faculty of Economics and Business MSc BA – Change Management

Supervisor: Dr. Marvin Hanisch

Co-assessor: Dr. Hille Bruns

June 22, 2020

Joint ventures are renowned for providing firms with the ability to integrate external knowledge and technology through the establishment of a new firm. Previous research in this field has acknowledged challenges in preserving proprietary knowledge and that these exchanges require appropriate contractual governance to address these issues when they emerge. Dispute resolution mechanisms, particularly, offer a logic on how disputes are settled, which results in different approaches depending on who is involved. This thesis aims to investigate the relationship of technological similarities – referred to as ‘technological proximity’ – on different contractual alternatives of dispute resolution, which range in complexity. Building on the merits of transaction costs economics, I propose that a higher extent of technological proximity will be associated with a higher probability that partners will contractually include more complex dispute resolution alternatives. There has been a prevalence in research pertaining to other contractual devices of governance, which are more adaptive and see features which may guide partners in interactions and mitigate exchange hazards. This thesis also aims to understand whether such a device, in the form of a joint steering committee, disincentivizes partners from contractually including more complex dispute resolution mechanisms through the committee’s ability to pre-emptively resolve disputes. Using a unique sample of 257 joint venture contracts and the USPTO patent database, the results do not find empirical evidence that supports these propositions. This research contributes to alliance governance literature by recognizing the role of technological proximity as an antecedent of contractual dispute resolution, provides a methodical approach to examine contractual dispute resolution, and sheds light on the role of joint steering committees in establishing the foregoing.

Table of content

1 Introduction ... 4

2 Theoretical background and hypotheses ... 6

2.1 Joint ventures and contractual governance ... 6

2.2 Types and complexity of contractual dispute resolutions ... 8

2.3 Technological proximity and knowledge spill overs in joint ventures ... 9

2.4 Joint steering committees ... 11

3 Methodology ... 14

3.1 Empirical setting ... 14

3.2 Data collection and sample ... 15

3.3 Measures ... 16

3.3.1 Dependent variable: Dispute resolution complexity ... 16

3.3.2 Independent variable: Technological proximity ... 18

3.3.3 Moderating variable: Presence of a joint steering committee ... 19

3.3.4 Control variables ... 20

4 Analytical method ... 21

5 Results ... 22

5.1 Descriptive statistics and correlations ... 22

5.2 Regression results and hypothesis testing ... 22

6 Discussion... 27

6.1 Theoretical contributions and implications ... 27

6.2 Managerial implications ... 29

6.3 Limitations and future research... 29

1

Introduction

In the face of competitive pressures, firms may seek resources and knowledge outside the borders of a firm (Powell & Koput, 1996), and therefore seek alliances to combine resources and strengths (Kogut, 1988). Joint ventures, above other modes of inter-firm exchanges and transactions, enable firms to acquire knowledge in the form of a joint venture firm (Buckley & Casson, 1996). The nature of a joint venture allows more tactic and embedded knowledge to be accessed due to the characteristics of a firm, as a separate legal entity is formed (Kogut and Zander, 1992). Particularly in high-technology industries where firms pursue activities such as R&D and production (Wang, Hsu & Fang, 2008), firms seek access to complementary resources to gain economies of scale and to shorten development times (Powell, Koput & Smith-Doerr, 1996).

In spite of the benefits of establishing a joint venture, the inter-firm relationship houses a substantial amount of risk, as it exposes the parties to each other’s opportunism (Williamson, 1985). A prominent reason for this is that partners in joint ventures may differ in their cooperative and competitive intentions (Tiessen & Linton, 2000). Hence, as firms seek to acquire external knowledge, they also face challenges in retaining valuable and proprietary knowledge (Giarratana & Mariani, 2014). The incentives to absorb and consequently internalize a partner’s knowledge, may be incongruent in the inter-firm relationship (Hamel, Doz, & Prahalad, 1989). In the case that such opportunism goes unchecked, a dispute between the partners may arise (Lumineau & Malhotra, 2011).

In the anticipation of potential opportunism and subsequent conflicts, firms impose governance mechanisms to both facilitate cooperation and mitigate risks of the relationship (Ring & van de Ven, 1994). In joint ventures in particular, the structure and process of exchange is governed by the contract, which also includes how order is brought to the relationship in the advent of a conflict (Luo, 2002). Hence, a rigorous governance design should anticipate potential disputes arising during the relationship (Williamson, 1985). Thereby, firms may have different approaches to resolve such disputes (Brett, Goldberg & Ury, 1990). These dispute resolution approaches may take the form of a bi-lateral, consensual approach or a tri-lateral, distributive approach (Devarakonda, Klijn, Reuer & Duplat, 2019). If these disputes are of a bad nature, e.g., when a party displays signs of opportunistic behaviour, parties may refer to a tri-lateral approach and thus involve external parties, such as an arbitration commission, to resolve the conflict and impose sanctions on the non-compliant party (Brett et al. 1990). Mechanisms involving third parties thereby involve more complexity and hostility over less complex methods to resolve the dispute bi-laterally (Koolwijk, 2006).

In particular, literature on alliance governance has studied contractual and firm related characteristics as antecedents of dispute resolution mechanisms (e.g., Lin & Germain, 1998; Devarakonda, Klijn, Reuer, & Duplat, 2019). In addition to this, prior research has recognized the challenges firms face in preserving proprietary knowledge when partners in alliances possess similar technological capabilities (e.g., Mowery, Oxley & Silverman, 1998; Devarakonda & Reuer, 2018). However, to my best understanding (and based on this research), literature has not studied technological similarities as an antecedent of dispute resolution mechanisms in alliance literature.

In this study, I build upon the ample literature inspired by transaction cost economics and the role of contractual governance as a means to alleviate opportunistic behaviour in interfirm relationships (e.g., Luo 2002; Reuer & Arino, 2007; Hoetker & Mellewigt, 2009). I hereby aim to, firstly, answer the question whether the technological similarities of partners establishing a joint venture acts as an antecedent to the dispute resolution mechanisms in the contract. Specifically, whether an increasingly similar technological portfolio of the firms involved in a joint venture include more complex dispute resolution mechanisms as a safeguard to potential opportunistic behaviour, which may arise during the relationship.

Partners in joint ventures face the trade-off of writing more detailed and complex contracts, versus those who have the ability to maintain flexible in the face of unforeseen contingencies and issues during the cooperation phase (Macneil, 1978). In light of this, transaction cost economics suggests that firms may include alternative devices to diminish disturbances, which may upset exchange relationships (Williamson, 1975, 1985). Therefore, firms may incorporate adaptive mechanisms as opposed to writing more complex arrangements (Bajari & Radelis, 2001; Schepker, Oh, Martynov & Poppo, 2014). Joint steering committees are as such, an adaptive mechanism (Smith, 2005). These committees can, through their delegated contractual authority, perform monitoring and control functions which can mitigate moral hazards, such as opportunism and thereby resolve disputes before they escalate (Devarakonda & Reuer, 2016). The question therefore arises, whether the inclusion of adaptive mechanisms, in the form of a joint steering committee, spurs partners to veer away from more complex methods of resolving disputes.

Hence, to consider the above, I secondly aimed to understand whether the monitoring and controlling features of contractually delineated joint steering committees incentivise partners to include less complex dispute resolution mechanisms to pre-emptively mitigate opportunistic behaviour and resolve unforeseen disputes. Although previous literature has recognized the role of joint steering, committees in non-equity alliances in mitigating moral hazards (e.g, Devarakonda & Reuer, 2016, 2018), its role in joint ventures or other equity-based alliances has received little to no attention.

increasingly similar technological portfolios relate to more complex dispute resolution mechanisms in the contract. Additionally, there was no evidence that the role of a joint steering committee moderates this relationship. However, this thesis does provide some useful contribution to the alliance and joint venture literature.

This study contributes in several ways to alliance and joint venture literature. Firstly, this study extends research on the antecedents of dispute resolution (e.g, Lin & Germain, 1998; Lumineau & Malhotra, 2010; Devarakonda, Klijn, Reuer, & Duplat, 2019) by investigating the role of similar technological capabilities possessed by the partners. Secondly, this study enriches research on contractual dispute resolution mechanisms by considering different combinations and alternatives of approaches to resolve disputes. Lastly, this paper builds on research pertaining to administrative bodies in the form of steering committees (Reuer & Devarakonda, 2016, 2018), and seeks to extend this by understanding the committee’s relation to dispute resolution mechanisms in joint ventures. From a managerial perspective, this paper may offer insights into the approaches of dispute resolution in joint ventures as well as the role that joint steering committees can pose in mitigating disputes before they materialize.

This study is structured as follows: Firstly, the relevant background theory is addressed, which lays a groundwork on which the hypotheses are based. The ensuing section delineates the data collection process and methodological decisions. Next, I tested the proposed hypotheses and present the empirical results. After which, I discuss the findings of the analysis including theoretical and managerial implications as well as the limitations of the study and consequent future research suggestions.

2

Theoretical background and hypotheses

This section guides the reader through relevant background theory in this study. Firstly, I discuss advantages of joint ventures and how contractual governance structures the inter-firm relationship, and in addition how it is not able to fully protect firms from unforeseen disturbances. This is followed by theory on contractual dispute resolution and why different approaches and subsequent mechanisms, may be more applicable when partners face threats of opportunism. I then present theory on why joint ventures are particularly susceptive to a partner’s opportunism due to a high amount of knowledge exchanges which are harder to safeguard compared to other alliances, followed by the first hypothesis. Lastly, I go into the concept of joint steering committees and why its related features help mitigate opportunism. I then present my second hypothesis of why partners may opt for less complex methods of dispute resolution when such a committee is contractually included.

2.1 Joint ventures and contractual governance

Joint ventures, by their nature, have the ability to incite competition, stabilize profit levels, and cause structural changes in vertical integration (Harrigan, 1998). Additionally, by two firms combining forces, joint ventures can take advantage of economies of scale, overcome barriers when entering new markets, combine relevant knowledge, and alleviate nationalistic reactions when pursuing the entry of a foreign market (Hennart, 1988). As an overarching rationale, firms establish a joint venture to combine resources and strengths, whereby tactic and often organizationally embedded capabilities and knowledge, of each firm, can be accessed more readily (Kogut, 1988). These features have led scholars to research joint ventures from a variety of perspectives and theories.

Joint ventures have been conceptualized and analysed with the application of various economic theories, such as internationalization theory, transaction cost theory, game theory and agency theory (Nippa, M. & Reuer, 2019). For this research, and one which is of particular interest to this study, is the transaction cost economics (TCE) theory. This theory has been the dominant lens, through which viewed joint have been analyzed (Tsang, 2000; Nippa, M. & Reuer, 2019). Transaction costs refer to the costs of writing and consequently enforcing contracts. This is where, when discussing terms and contingencies, partners aim to increase dependence on another party to stabilize a relationship and administering a transaction (Williamson, 1985). According to TCE, two elements in particular are distinguished: joint ownership (and control) and the mutual commitment of resources (Kogut, 1988). In such inter-firm collaborations, the contract provides structure and process though which these elements are addressed (Williamson, 1985).

When two or more parties participate in the establishment of a joint venture, a contract binds the parties respective rights and duties. The contract provides a mutually agreed upon, formal structure, which underlies all related goals, responsibilities, policies, and other specifications of the inter-firm relationship (Luo, 2002). The contract is central to the organization and management of exchange relations between firms (Reuer & Devarakonda, 2016) and can range from standard boilerplate layouts to highly contextualized terms and agreements (Schepker, Oh, Poppo & Martynov, 2014). Every joint venture contract aims to build the appropriate arrangements which optimally facilitate the processes of exchange among the firms, and prevent the leeway for opportunistic behaviour, prohibit moral hazards and protect proprietary knowledge (Hackett, 1993). Partners may therefore choose to write extensive contracts, which can reduce uncertainties in decision-making and risks stemming from opportunistic behaviour (Williamson, 1985; Brown, Dev & Lee, 2000). Every contract intends to essentially facilitate cooperation between partners and to prevent opportunistic behaviour (Luo, 2002).

However, contracts are not always complete, as parties are faced with a trade-off between the

ex-ante costs of writing more complete contracts and the ex-post inefficiencies, coupled with less

conflicts (Goldberg, 1976). Killing (1993) suggests that less specific contracts and terms are less able protect a partner’s resources and induce uncertainties by not adequately allowing firms to control opportunism. In a response to this, firms may incorporate specific contractual provisions which can pre-empt conflict and provide a basis to resolve disputes which may surface throughout the relationship (Crocker & Reynolds, 1993). To address the basis on which disputes are resolved, the next section discusses the function of contractual dispute resolutions and the approaches and subsequent mechanisms which firms may employ.

2.2 Types and complexity of contractual dispute resolutions

Although the previous discussion delineates some advantages of joint ventures, there is a prevalence of high mortality and poor performance amongst joint ventures (Bleek & Ernst, 1991; Harrigan, 1988), of which making-joint decisions is a key contributor (Geringer, 1988). In joint ventures, parent firms aim to build successful ventures by displaying mutual trust and commitment in the partnership (Beamish, 1988; Geringer, 1998). This mutual trust creates the expectation that parent firms will act in a favourable and non-exploitive manner (Bluckley & Casson, 1988). Parent firms can, however, have competitive intentions (Tiesen & Linton, 2000) in which they try to maximize their individual gains and act in unfavourable ways, such as hiding information. This might consequently taint their decision-making rationale (Lax & Sebenius, 1986). Subsequently, when partners show signs of opportunism or other moral hazards, disputes may arise (Williamson, 1985).

In anticipation of potential opportunism taking place, firms can contractually include governance mechanisms which mitigate relationship risks and facilitate cooperation (Ring & van de Ven, 1994). Contractual governance creates a ‘logic of appropriateness’ which offers a means through which bi-lateral behaviour can be assessed, including the responses to unwanted behaviours (March, 1994). A robust governance design recognizes the likelihood of a dispute arising through asymmetric and or self-beneficiary interests (Williamson, 1985) and the contract can include methods of brining order to the relationship (Luo, 2002). Hence, contracts usually contain specific provisions regarding sanctions which can be imposed if a party violates contractual agreements (Miller & Ross, 1975). In the instance that the parties cannot agree on such violations or other matters, they may refer to contractual mechanisms in which disputes can be addressed (Luo, 2002; Reuer, Klijn & Lioukas, 2014).

These conflict resolution alternatives differ in complexity and are likely to be influenced by the approach to the dispute resolution (Ury et al., 1988). Lumineau & Malhotra (2010) compare two common approaches in resolving inter-firm disputes which guide partners to choose bi-lateral or tri-lateral mechanisms: the interest-based approach and the rights-based approach (Brett, Goldberg & Ury, 1990). The interest-based approach is integrative, consensual, and problem-solving, and will seek to settle disputes through discussions, negotiations and mediations, and emphasize mutually acceptable alternatives to escalating the conflict (Ury et al., 1988). Such a compromising and problem-solving approach will not only help parties escape “deadlock” scenarios, but they will also lead to joint venture success in the long run (Friedmann and Beguin, 1971). In contrast, the rights-based approach, according to Brett et al. (1990), is mainly distributive, adversarial, and competitive and relies on an independent standard with sufficient legitimacy or fairness to determine who is right, i.e., it implies a win-lose outcome. The rights-based approach seeks to understand which party is right and which party is wrong, which can be exceedingly difficult to achieve and therefore often includes the participation of a costly, external party or institutions to determine this. Notably, these approaches do not have to be mutually exclusive. Parties may seek both routes to successfully approach a dispute (Ury et al., 1988).

In summary, firms pursuing a joint venture may include bi-lateral and or tri-lateral mechanisms in the contract to resolve any disputes which can arise once the cooperation has commenced. Such mechanisms differ in complexity, whereby tri-lateral, legalistic approaches portrait more hostility and are more complex as an external party is involved (Koolwijk, 2006), which may be required if the dispute relates to sensitive matters such as the mis-use of proprietary knowledge.

2.3 Technological proximity and knowledge spill overs in joint ventures

Creating and commercializing products in a timely and cost-effective way can increase one’s competitive position, which is particularly present in technology-intensive industries (Buckley & Casson, 1996). In response to competitive pressures and to build capabilities, firms may therefore look for resources and capabilities outside their internal prowess (Powell & Koput, 1996). To strengthen this competitiveness, firms that are co-operatively pursuing collaborative development and production of a product, are likely to choose partners that have some level of similar technological capabilities (Cantwell & Barrera, 1996). Likewise, when technology-based capabilities are internalized by one firm in an alliance, the absorption process would demand that this firm has sufficient in-house expertise to accommodate and compliment the technology development activities (Cohen & Levinthal). Consequently, alliances as a source of resources and information has proven to have a positive link on patenting and innovation (Ahuja, 2000), where joint ventures, such as other forms of alliances, form superior means of sharing and transferring technological capabilities over spot contracts (Kogut, 1988).

given the moral hazard problems that are coupled with sharing knowledge (Sampson, 2007). Collaborative vehicles, such as joint ventures, particularly carry risks of partner appropriation of the shared knowledge (Madhok, & Tallman, 1998). A predominant reason for this is that joint ventures create a new firm, which in turn has its own hierarchical structure. Kogut and Zander (1992) argue, that because of the characteristics of a firm, knowledge sharing (particularly complex or tactic knowledge) happens more easily in firms than between firms. In joint ventures, partners allow exposure of each other’s routines, skills, and proprietary know-how to combine strengths. By combining existing knowledge with spill overs from the adjacent partners, new technological capabilities can be created (Khanna, Gulati, & Nohria, 1998). However, this knowledge can dissipate and consequently be appropriated within joint ventures along a myriad of paths. For instance, by opening private communication channels such as email communication, bi-lateral knowledge can be accessed. Additionally, firms might share resources such as documents, blueprints, memos, intern alia (Appleyard, 1996). In joint ventures particularly, employee rotation and the ensuing contact amongst technical employees is a prevalent method of knowledge sharing. In joint ventures, where technological development is part of the venture’s activities, partners typically supply technical staff from the parent organization. When these technical staff members rotate back to their parent organization, as they might occasionally do, they take this gained knowledge with them (Sampson, 2007). Hence, by making resources and knowledge accessible with increased transparency, critical know-how might permeate through such points of contact (Giarratana & Mariani, 2014).

In establishing a joint venture, firms may differ in their intentions to absorb, and consequently internalize a partner’s knowledge (Hamel, Doz, & Prahalad, 1989). Therefore, partners in inter-organizational exchanges, such as joint ventures, need to manage the access of their proprietary knowledge in order to prevent leakage (Oxley & Sampson, 2007). Furthermore, there might be an incongruence of what a firm is willing to share versus what it is able to share (Sampson, 2007). As channels open, through which valuable knowledge can be captured by the adject firm, and where incentives may not be congruent, the opportunities to absorb spill overs for private benefit are increased (Khanna, Gulati, & Nohria, 1998). Hence, as a result of learning and acquiring a partner’s knowledge, joint ventures can become unstable (Beamish & Inkpen’s, 1997). This possibility of instability, coupled with fears of opportunistic behaviour, may raise tensions and disputes within the relationship which require appropriate contractual means to be addressed.

Dispute resolution mechanisms as a safeguard for opportunistic behaviour

complexity can include confidentiality provisions and other provisions which can restrict partners’ use of proprietary knowledge. Specifically, provisions of how unforeseen disputes are resolved can hinder the misuse of sensitive technological knowledge (Lerner & Malmendier, 2010).

As the foregoing theory points out, when partners in alliances seek to collaborate in knowledge-intensive activities, such as R&D and production, there is a necessity to share proprietary knowledge to create value. In this process of bi-laterally acquiring knowledge of the other firm, joint ventures can become unstable. Partners may have different intentions before or during the inter-firm collaboration, whereby a partner may seek to internalize the knowledge of the other for self-beneficiary reasons. Consistent with transaction cost economics arguments, I argue that partners will negotiate appropriate contractual provisions in the joint venture contract if there is an increased possibility and incentive to misappropriate the associated knowledge spill overs. In the case such behaviour is noticed by either party, the possibility of a dispute arising among the parties is then a likely occurrence. In such a scenario, I contend that partners would refer the dispute to an external party to address the concerns and subsequently rule – with a rights-based approach – the appropriate ramifications to the party which shows intentions of opportunism. Hence, I argue that when there is an increased possibility to misappropriate, due to high levels of similar technological expertise and capabilities, partners are more likely to contractually include dispute resolution mechanisms which hold the authority to assign appropriate consequences to the non-compliant party. I thus derive the following hypothesis:

Hypothesis 1: When partners in joint ventures foresee disputes arising due to increasingly similar technological proximity, partners are more likely to include more complex dispute resolution mechanisms in the contract.

2.4 Joint steering committees

writing more detailed contracts to deal with future contingencies (Bajari & Radelis, 2001; Schepker, Oh, Martynov & Poppo, 2014).

Research on adaptive, ex-post governance mechanics has studied how partners in inter-firm collaborations incorporate features which can sustain a satisfactory relationship beyond the ex-ante contractual contingencies (e.g., Poppo & Zenger, 2002; Schepker, Oh, Poppo, & Martynov, 2014; Reuer & Devarakonda, 2016). Partners may construct broad guidelines or specific responses to certain contingencies which can provide assistance in adjusting to contingencies (Argyres & Silverman, 2004). Additionally, when partners define decision rights in the contracts, partners can then, ex-ante, allocate authority to who can decide on these contingencies (Arruñada, Garicano, & V´azquez, 2001). Partners may also contractually decide to dedicate an adaptive interface which can mitigate information barriers to adaptation by facilitating joint decision-making and sharing information (Palay, 1984). Firms can contractually establish governing bodies which can act as coordination and monitoring interfaces. Such governing bodies in highly relationship-intensive exchanges, such as joint ventures, can facilitate information sharing and therefore increase problem solving, adapting, mutual trust, and mitigating exchange hazards. Such a governing body can provide a dedicated interface between partners which can guide interactions, address contingencies and potential conflicts as they surface through the course of the collaboration (Reuer & Devarakonda, 2016).

When partners establish such a governing interface within contracts, it suggests that partners foresee potential adjustments to contingencies, but are unsure of the exact nature of these in the contracting stage (Reuer & Devarakonda, 2016). A common construction of such a governing interface takes place in the form of a joint steering committee, which provides various advantages to complex collaborative vehicles such as joint ventures. Joint steering committees can be established contractually, which have similar traits to the board with a defined scope of authority (Smith, 2005). Steering committees can serve several important adaptation and monitoring roles, through its ability to reduce information barriers and facilitate exchanges (Barnard, 1938). Firstly, joint steering committees can act as a structure which overlays a formal administrative layer over more operational management in alliances. Thereby, with their respective associated contractual authority, rules of engagement and interaction can be formulated and subsequently monitored (Liebeskind, 1997). Secondly, joint steering committees can facilitate the exchange of knowledge amongst parties in alliances, by for example adopting mutually agreed upon codes which may enhance the efficiency of communication (Sampson, 2007). Thirdly, joint steering committees can guide interactions in the light of unforeseen contingencies and subsequently respond to these efficiently and thereby mitigate the chance that small conflicts materialize into larger disputes that involve third parties (Reuer & Devarakonda, 2016).

Moderating effect of joint steering committees

contingencies. The presence of a committee can help facilitate a satisfactory relationship through a defined and dedicated interface by promoting and coordinating adaption (Reuer & Devarakonda, 2016).

A key source of adaptions concerns, which is particularly relevant for high-tech alliances, is the need to adapt to the spill over leakage of knowledge (Oxley & Sampson, 2004). In the presence of a means to facilitate and guide interactions through a governing body, partners can manage the limits of cooperation and provide guidance on the loss and diffusion of proprietary knowledge (Reuer & Devarakonda, 2016). These abilities can become particularly important when partners possess similar technological expertise and capabilities (Katila, Rosenberger, & Eisenhardt, 2008). Because joint steering committees possess the authority by virtue of the contract, they can exert their influence to engage in organizing activities by for example establishing communication channels and impose approvals (Liebeskind, 1997). Therefore, partners who ex-ante contractually include a joint steering committee, can expect less opportunistic behaviour during the collaboration phase (Reuer & Devarakonda, 2016, 2018).

As joint steering committees feature monitoring and controlling functions to foresee potential hazards which might otherwise go unnoticed, opportunistic behaviour can mor readily be detected and acted upon. Joint steering committees can enable partners to coordinate adaption in the advent of disturbances which may otherwise strain partners and their relationship or lead to costly disputes involving external parties such as arbitration or litigation. Therefore, I argue that when partners include joint steering committees in the contract, it mitigates the ex-ante incentive and opportunity of partners in joint ventures to seek opportunistic behaviour and the appropriation of knowledge. Specifically, joint steering committees can resolve disputes as they surface during the inter-firm relationship. Therefore, partners may omit more complex and hostile methods of dispute resolution mechanisms to facilitate conflicts. I thus derive the following hypothesis:

Based on the previously theorized hypotheses (H1 and H2), Figure 1 depicts the relationships in a conceptual model.

3

Methodology

In the following section, the empirical setting, data collection and sample, measures, and results are explained in detail. The empirical setting explains the industries included in the sample and why they are relevant. The data collection and sample describe how relevant contracts were found and analysed and how patent filings data was collected. The measures section delineates how the collected data was transformed into variables which were employed for statistical analysis. The control variables and related measures are also explained. Throughout the methodology, I elaborate on the statistics choices I have made.

3.1 Empirical setting

The empirical setting of this study was comprised of two industries. The data collection of joint venture contracts and their respective analysis was part of an ongoing collection process. The industries that represent the sample are U.S. biopharmaceutical joint ventures and joint ventures in the high-tech industry, the latter being my and a fellow student’s addition to the dataset. The biopharmaceutical industry lends itself, for various reasons, to why it is suitable for this study. Firstly, the sector increasingly relies on collaborations, such as joint ventures, for innovation (Hagedoorn, 2002). Secondly, it is a very knowledge-intensive sector (Dong & Yang, 2016). Thirdly, due to the high levels of knowledge transfer in the industry, intellectual property and leakage concerns are prevalent (Hagedoorn & Hesen, 2007; Arti, 2014). To complement the dataset, the high-tech industry was chosen as it shows similar characteristics, such as a very large number of alliances in the form of joint ventures. Moreover, these collaborations often include R&D, production, and supply activities (Wang, Hsu & Fang, 2008), and therefore also sees a high degree of proprietary knowledge exchanges.

Dispute resolution complexity Technological proximity

+

_

Figure I: Conceptual model

Control variables

1. JV scope 2. Relative firm size

3.2 Data collection and sample

The existing dataset contained 225 analysed joint venture contracts of the biopharmaceutical industry. In order to collect sufficient and relevant joint venture contracts in the high-tech industry, the Securities and Exchange Commission (SEC) EDGAR database was firstly consulted. The SEC collects a vast number of filings from U.S. and international securities markets. The EDGAR database provides a tool to find companies and their respective fillings. The database assigns every to firm, it has received/gathered fillings from, a Standard Industrial Classification (SIC) code. Table 5 of Philips et al. (2009) provides an overview of SIC code combinations which represent the closest match to their benchmark of what is considered high-tech, by analysing firm descriptions and footnotes. By searching for these SIC codes in the EDGAR database, I researched these firms to see which has joint ventures by conducting a Google search. In order to find the contracts, I googled both names of the companies in the joint venture plus “sec joint venture”, as that was the most time-effective method to find the contracts, given that they are all publicly available.

This initial method yielded limited results. Specifically, when I was able to find firms which currently have or previously had joint ventures, I was often not able to find the related joint venture contract, whilst these firms did file other information as 10-Q, 10-K, 8-K and other filings. This left me with seeking and finding a method which yielded sufficient and relevant contracts. Therefore, I sought publicly available databases which contained joint venture contracts. LawInsider.com provides a large database of publicly available legal contracts, which was the primary source from where I collected contracts. To find joint-ventures in the high-technology sector, I conducted a large number of searches including terms such as “technology”, “semiconductors” amongst a host of other terms until I was unable to find any new contracts which seemed appropriate for analysis. In total, an additional 109 high-tech contracts were allocated and analysed. After dropping duplicates and not-suitable contracts, the dataset totalled 257 contracts between 1981 and 2019. Table I presents an overview of the respective number of contracts of both sectors.

TABLE I

Sectors included in data set

The next step in the data collection process was to extract datapoints from the additional contracts as defined in the existing dataset containing the biopharmaceutical joint ventures. For this study, the data points of interest were those pertaining to the dependent variable – the dispute resolution method –

Sector

Freq.

Percentage

Healthcare 159 61.87%

High-tech 98 38.13%

delineated in the contracts. The dataset includes whether governing bodies, such as the board and the joint steering committee, have stated methods of internal dispute referral and/or which party or by what method the final decision-making was attributed to. Additionally, contracts include a general dispute resolution clause in the contract, which states the method of resolving disputes in relation the contract, which might also include referrals. Finally, the dataset also included whether a joint steering committee is present in the contract.

For the purpose of this study, I exclusively looked at the general contract dispute resolution clause and its respective mechanism(s) of dispute resolution. This is because the inclusion of dispute resolution mechanisms of governing bodies such the board and JSC would complicate the method of analysis given the relatively low number of observations. Additionally, the presence of a contractual dispute resolution section was the most prevalent throughout the sample (81%).

For the independent variable technological proximity, a similar approach as Sampson (2007) was followed. For each firm in the dataset, I assign the respective id(s) from the PatentsView.org database, which derives its data from the US Patent and Trademark Office (USPTO). This data includes the patent technology classification, which is key to calculating the technological proximity. As patents are often not assigned to subsidiaries, the granted patents of the entire firm were collected. This was done to prevent an obscured and incomplete measure of technological capabilities and consequently, biased parameter estimates (Kennedy, 2003).

Finally, for the control variables, I used additional data points which were derived from the contracts. Additionally, public sources such as a firms’ annual reports were also used. All the collected data was exported to STATA 16 for further refinement and subsequent data analysis which is elaborated on in the following sections.

3.3 Measures

To test the hypotheses theorized in this study, variables had to be constructed from the collected data. This section describes the variables present in the analysis and what measurement was opted for each. The variables dispute resolution complexity (dependent), technological proximity (independent), joint

steering committee (moderator), and control variables are elaborated on in this section.

3.3.1 Dependent variable: Dispute resolution complexity

Bi-lateral forms of dispute resolution have been codified during the data collection as referrals. Thereby, dispute resolution clauses may have several levels of referrals. To illustrate this with an example, a contract might state when a dispute arises among the involved parties, the dispute is firstly referred to the board of the joint venture. Then, if the board is not able to decide on the respective issue or conflict at hand, it may be referred to the CEO’s of the parent organizations, who may then discuss the matter. Finally, the dispute resolution clause will almost always include who has the final decision-making authority over the dispute. This can then be decided either internally (e.g., the one party or a governing body such as the JSC) or through external tri-lateral mechanisms (arbitration).

Amongst the three most common forms of tri-lateral dispute resolution mechanisms mediation, arbitration and litigation - arbitration offers the speediest, most flexible, and economical way to proceed, which allows parties to sustain a business relationship in contractual agreements. The benefits also derive from the decision to legally commit to the final arbitration settlement, resulting in the non-prevailing party having little freedom to appeal (Bonn, 1972). For the purpose of this study, I chose arbitration as the tri-lateral mechanism in determining dispute resolution complexity, as it is the most frequently used tri-lateral mechanisms in the analysed contracts (mediation - 13%, litigation - 3% and arbitration - 67%, not containing tri-lateral mechanism - 17%). This mechanism also contains the appropriate legal repercussions when determining the non-prevailing party (Bonn, 1972). Arbitration compared to internal referrals or negotiations increases the level of hostility and cost in resolving disputes (Koolwijk, 2006). Additionally, legalistic approaches to dispute resolution such as arbitration can have more corrosive implications for the interfirm relationship, as it often implies a ‘win or lose’ situation (Lin & Germain, 1998), which may be considered preferable in the case of opportunistic behaviour. Bi-lateral, or ‘low cost’ options for dispute resolution in the form of ‘alternative dispute resolutions’ has been found to be considerably less costly than its legalistic counterparts (Brett, Barsness & Goldberg, 1996).

Hence, I measure dispute resolution complexity as a categorical value represented by 0, 1, 2 or 3. Although the methods of bi-lateral mechanisms might also differ in complexity, attributing a weighted value or alternative measurement to this variable would be subjective in nature and might therefore impair construct validity.

3.3.2 Independent variable: Technological proximity

Technological proximity, otherwise referred to as “technological diversity”, “technological overlap”, or “technological relatedness” has been used as an indicator to measure the extent to which firms in inter-firm relationships possess similar technological capabilities (e.g., Mowery, Oxley & Silverman, 1998; Sampson, 2007; Colombo & Rabbiosi, 2014; Colombo & Piva, 2018). The measure of this variable follows that of Sampson (2007). The method she employs measures the extent to which partners patent products in the same technological classes (Jaffe, 1986), and therefore provides an adequate proxy for stating the relative technological position of two partners. This method provides a distinct advantage above categorizing the end products, as two similar products can underlie very different technologies (Sampson, 2007). For example, if two firms venture into a collaboration in the manufacturing of automobiles, where one partner specializes in electric motors and the other in combustion engines, the technological capabilities under the product category “engines” does not reveal such capabilities.

Figure II: Dispute Resolution Complexity Quadrant

Arbitration Yes No Bi -l at eral m echani sm (s ) Yes No 0

(No arbitration nor bi-lateral mechanism(s))

2

(Arbitration and no bi-lateral mechanism(s))

1

(Bi-lateral mechanism(s) and no arbitration)

3

(Both bi-lateral mechanism(s) and

To construct this variable, I generated each joint ventures’ partners’ technological portfolio by measuring the distribution of patent across patent classifications before the date of the commencement of the joint venture. For example, if firm A and B have patents classified to three categories, the dataset would be constructed as following:

This distribution is then captured by a multidimensional vector, Ca = (Ca1 … Cas), where Cas represents the number of patents assigned to firm A in patent class s. The technological proximity is then calculated as following:

𝑇𝑒𝑐ℎ𝑛𝑜𝑙𝑜𝑔𝑖𝑐𝑎𝑙 𝑝𝑟𝑜𝑥𝑖𝑚𝑖𝑡𝑦 = 𝐶𝑎𝐶𝑏 ′ √(𝐶𝑎𝐶𝑎 ′)(𝐶𝑏𝐶𝑏′)

Where a ≠ b. This results in range for the variable from 0 to 1, where 1 represents the greatest possible value for technological proximity between partner firms A and B. Not all firms in the dataset are either listed in the USPTO or have any registered patents. In the dataset, 170 observations do not have patents assigned to them. In these cases, these values are represented by 0 (i.e. lowest possible technological proximity value), to avoid undefined results. Sampson (2007) distinguishes a few drawbacks that this method might encounter which are also relevant here. Such as, that some classes might be more similar than others in terms of the technology they represent and that new classes might emerge when existing classes fail to categorize technologies adequately.

3.3.3 Moderating variable: Presence of a joint steering committee

The moderating variable in this relationship constitutes whether a joint steering committee has been contractually established with oversight activities in the joint venture. Therefore, the joint steering committee is a binary variable. The variable equals one when the contract includes such a governing body and is otherwise equal to zero. For this study, we contend that the partners have valid reasons to establish such a committee. It draws on research which suggests that potential moral hazards, such as opportunism and misappropriation that may arise in the presence of high technological proximity of partners. This can be alleviated through the monitoring functions that joint steering committees exert (Oxley & Sampson, 2004).

id class1_firmA class2_firmA class3_firmA class1_firmB class2_firmB class2_firmB

3.3.4 Control variables

I introduced control variables for several features of partners and other contractual elements which potentially explain the contractually delineated dispute resolution complexity, and which may also relate the independent variable technological proximity. These controls are:

Joint venture scope. Partners that establish a joint venture may choose to conduct a wide range

of activities, whereas others might only combine strengths on a few key areas. Partners that have a diverse range of activities in the joint venture, are more likely to open more communication channels and subsequently share more information (Appleyard, 1996). Additionally, the more knowledge that is transferred to the joint venture, the more technical staff are required to rotate to the joint venture and back (Sampson, 2007). Hence, the more interdependent, extensive, and complex the scope of activities, the higher the chance of opportunistic behaviour (Oxley & Sampson, 2004). The contract often contains what joint activities the partners will conduct. To measure this, seven different activities were generalized which include for example R&D, manufacturing, and marketing from the contracts. The more activities outlined in the contracts, the higher the measure for joint venture scope.

Relative firm size. Larger firms may venture into collaborations such as joint ventures with

smaller, more entrepreneurial firms. These smaller firms may lack the capabilities to scale and commercialize their innovate technologies and therefore form alliances to with larger firms to create value (Colombo & Piva, 2018). Thereby, cutting-edge technology may unintentionally leak over to the other partner and be misappropriated to exploit the knowledge for its own focal business (Alvarez and Barney, 2001). Moreover, the ability to assimilate knowledge from partners may also be related to the size of the firm (Dong & Yang, 2016). Therefore, smaller firms may opt for more contractual complexity to safeguard their knowledge (Colombo & Piva, 2018). The relative firm size was calculated by dividing the number of employees of the smallest firm by the largest firm in each joint venture and subtracting this number off one. Thus, 1 means that the relative size difference between the partners is very high.

Deal size. To measure this, I used the equity size stipulated in the contract (in millions of

dollars).

Ownership Asymmetry. In joint ventures, ownership stake has been measured as a determinant of control and dominance, implying that there is less potential for conflict as decision can be made by the dominant party (Killing, 1983). However, evidence in literature has found this to be not often true (e.g., Geringer & Hébert, 1989). Alternatively, the ownership stake may suggest a measure of commitment and involvement, whereby the more the ownership is shared, the more stability is present due to equal involvement of each partner (Beamish, 1985; Mjoen & Tallman, 1997). To account for this, I constructed a binary variable which equals 1 if equity stakes are asymmetrical (i.e., not 50-50), and otherwise is equal to 0.

Prior joint venture experience. Partners may have previous experience with joint ventures,

for this, I created a dummy variable which is equal to 1 if either party as previous joint ventures experience prior to the data of the respective joint venture.

Contract length. Partners may push for more complex contracts to safeguard their knowledge

assets. A more complex contract can include provisions that stipulate solutions to foreseeable contingencies as well as processes for resolving unforeseen outcomes (Poppo & Zenger, 2002). To measure this, the word count of the contract is used.

Sector dummy. The extent that partners’ markets overlap may raise concerns of opportunistic

behaviour, due to goal conflict or knowledge misappropriation (e.g., Oxley & Sampson, 2004). Due to the competitive nature of firms in similar industries, partners may act in ways which is self-beneficiary (Lax & Sebenius, 1986). To control for this, I followed previous research in grouping the quite extensive array of industries in the sample into two sectors: healthcare and high-tech (omitted) to facilitate the analysis.

Year dummies. To account for any year fixed effects, I include a dummy variable with a

timespan of ten years (1981-1985 omitted).

4

Analytical method

The dependent variable, dispute resolution complexity, is a categorical value which can take the values of 0, 1, 2, and 3 as per the quadrant. Important to note is that these numbers are merely codes and their magnitude cannot be interpreted. They merely form different combinations of dispute resolution mechanism, i.e., they are not equidistant. As the dependent variable contains more than two categories, a multinomial regression is therefore required. The independent variable technological proximity (and other control variables) takes the form of an alternative-invariant, or case-specific regressors, in that it varies per joint venture but does not vary based on the dispute resolution complexity. To illustrate with an example, several people with different levels of income may have the option to choose between two fruits. The income variable is not related to the fruit and is therefore alternative-invariant. However, the price of the fruit does vary based on what fruit it is and is therefore alternative-variant. Given that no variables in my model derive their value directly from the dispute resolution complexity, a multinomial logistic (logit) model best suits this regression over a conditional or mixed logistical model. A multinomial logistic regression is used to model nominal outcome variables, in which the log odds (probability) of the outcomes are modelled as a linear combination of the predictor variables. Using the previous example with fruit and income, if income takes the value of 0.001 in the regression, then a one unit increase of the income is associated a 0.001 increase in the relative odds of choosing one fruit over the base category, which is this example is the other fruit.

variables have rescaled to have a mean of zero and a standard deviation of one. Therefore, an increase of one for the continuous independent variables will infer an increase of one standard deviation on the relative log odds of choosing the respective alternative to no dispute resolution mechanism in the contract.

In my model, I compare the levels 1, 2, and 3 of dispute resolution complexity against 0 (no contractual dispute resolution). In that, I compare the relationship of technological proximity on the relative log odds that partners establishing a joint venture choose a more complex dispute resolution in the contract compared to no dispute resolution provision in the contract.

5

Results

In this section I present the statistical results of this study. Firstly, the descriptive statistics and correlations are shown. Secondly, the results of the multinomial logistic model are presented to test the hypotheses. Lastly, a figure with the marginal effects of the interaction variable between joint steering committee and technological proximity on one level of dispute resolution complexity is shown.

5.1 Descriptive statistics and correlations

Table II presents the correlation matrix of the descriptive statistics of this study. The table includes the mean values, standard deviations and the correlations between all variables present in the study. I note that a contractual dispute resolution provision was present in 81% of the contracts, a technological proximity indicator in 34% of the contracts and a joint steering committee in 35% of the contracts. The correlation matrix shows very weak correlations between the variables and no significant correlations. As there are no correlations which have high values, I therefore suspect no issues with multicollinearity.

5.2 Regression results and hypothesis testing

Table III presents the multinomial logistic regression results. The results are constructed as follows: Every model (1-3) shows the three alternatives of Dispute resolution complexity (1, 2 and 3 per the quadrant) and thus the increasing levels of Dispute resolution complexity (see Figure II for the distribution of dispute resolution complexity alternatives in the sample). Every alternative displays the relative log odds of each predicter variable on the different levels of Dispute resolution complexity. The first model includes only the control variables, the second model includes the independent variable

Technological proximity, and the third model includes the Joint steering committee and interaction

Model 1 – the base model – includes the control variables and their relationship on three alternatives for the independent variable Dispute resolution complexity. JV scope consistently displays a positive association throughout all three models on all alternatives for dispute resolution complexity. Notably,

JV scope is significantly associated with the first alternative for Dispute resolution complexity (bi-lateral

and no arbitration), with (p = 0.860, p < 0.01), (p = 0.844, p < 0.01), (p = 0.966, p < 0.01) respectively for model 1-3. This implies that an increase of one standard deviation in the contractually defined scope of the joint venture is positively associated with the probability that partners include bi-lateral mechanism(s) of dispute resolution in the contract compared to none. Ownership asymmetry shows a positive and weakly significant association on the inclusion of an arbitration provision in the contract in the first (p = 1.048, p < 0.05) and second model (p = 1.038, p < 0.05). Therefore, an asymmetric ownership stake in the joint venture (i.e., not 50-50) has a positive relationship on the probability of arbitration as the dispute resolution mechanism in the contract. Contract length shows a positive and significant relationship with more complex alternatives of dispute resolution complexity (2 and 3, p < 0.01 and p < 0.001 respectively) throughout all three models. Thus, a more detailed (and perhaps more complex) joint venture contract is positively associated with a higher relative probability of a more complex dispute resolution complexity, compared to none.

In model 2, I examine the relationship of the independent variable Technological proximity on the Dispute resolution complexity alternatives. Here, no significant results are shown and hence no inferences can be made. Hypothesis 1 suggests that Technological proximity has a positive relationship on the relative probability that the joint venture contract will include more complex alternatives of

Dispute resolution complexity compared to none, and thus that Technological proximity would have a

positive sign associated to the more complex alternatives. However, the results in the third – full specification model – indicates the opposite relationship. One standard deviation increase in

Technological proximity shows a negative and significant relationship with the relative probability of

the second alternative (p = -1.030, p < 0.01), and a negative but weakly significant relationship with the third, most complex, alterative (p = -0.585, p < 0.05). These results infer that the opposite of Hypothesis 1 is visible in this study, which means that with an adequate certainty of confidence, Hypothesis 1 is

FIGURE II Percentages of dispute resolution complexity in

rejected. Hence, my argument that when partners, prior to establishing a joint venture, possess increasingly similar technological capabilities, opt for more contractually complex dispute resolution mechanisms, is not supported.

Finally, I include the moderating variable Joint steering committee and test the interaction between a Joint steering committee and the Technological proximity of partners prior to the establishment of a joint venture. To be noted, is that the presence of a Joint steering committee displays a weakly significant and positive relationship on the first (p = 1.338, p < 0.05) and a positive and significant relationship with the third (p = 1.272, p < 0.01) dispute resolution alterative. Although I conclude that Hypothesis 1 is rejected, I will present the moderation effect of a Joint steering committee. In Hypothesis 2, I propose that the presence of a Joint steering committee moderates the positive relationship of Technological proximity on the contractual Dispute resolution complexity. This would suggest a negative sign for the interaction variable in model 3. In the full-specification model,

Technological Proximity shows a sufficiently significant relationship for the second alterative, and

therefore only this relationship can be strengthened or weakened by the moderating variable. The results indicate a positive interaction effect in the presence of a Joint Steering Committee. Therefore, it does not support my second hypothesis and is therefore rejected.

To clarify this moderating effect, I graphically depict the interaction from Table III (third model, second alterative) in Figure III. I plot the marginal effects of the presence of a Joint steering

committee over the percentiles of technological proximity (std.). I opt for percentiles over scaled values

because the data is skewed due to the overrepresentation of no technological proximity in the contracts – where no patents are assigned to the firms. Additionally, I plot a horizontal line at y = 0 to display the threshold of marginal effects of the moderator. This plot confirms the positive marginal effect of the presence of a Joint steering committee.

FIGURE III Effect of interaction between

J.J. van Heyningen TABLE II

Descriptive statistics and correlation

Variable

Mean

S.D.

Min

Max

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(1)

Dispute Resolution Complexity

1.98

1.15

0

3

|

(2)

Technology Proximity

0.21

0.32

0

1 -0.07

|

(3)

JSC

0.35

0.48

0

1

0.21

0.06

|

(4)

JV Scope

3.36

1.62

0

6

0.15

0.09

0.15

|

(5)

Relative Firm Size

0.53

0.42

0

1 -0.04

0.29

0 0.11

|

(6)

Deal Size

33.37

133.13

0

1500

0.02 -0.05 -0.06 0.07 -0.05

|

(7)

Ownership Asymmetry

0.45

0.5

0

1 -0.02 -0.16 -0.07 0.01 -0.14

0.06

|

(8)

Prior JV Experience

0.66

0.47

0

1

0

0.26 -0.02 0.15

0.24 -0.02 -0.13

|

(9)

Contract Length

17772 14366.12

673 85930

0.26

0.24

0.19 0.15

0.12

0.14 -0.21 0.07

|

J.J. van Heyningen TABLE III

Multinomial logistic regression results

Variables 1 2 3 1 2 3 1 2 3

Bi-lateral mechanism(s) and no arbitration)

Arbitration and no bi-lateral mechanism(s) Both bi-lateral mechanism(s) and arbitration Bi-lateral mechanism(s) and no arbitration) Arbitration and no bi-lateral mechanism(s) Both bi-lateral mechanism(s) and arbitration Bi-lateral mechanism(s) and no arbitration) Arbitration and no bi-lateral mechanism(s) Both bi-lateral mechanism(s) and arbitration Control variables

Joint venture scope 0.860** 0.505 0.375+ 0.884** 0.513 0.369+ 0.956** 0.712* 0.392+

(0.33) (0.33) (0.19) (0.33) (0.33) (0.19) (0.35) (0.36) (0.21)

Relative firm size 0.109 0.112 -0.057 0.107 0.292 -0.007 -0.014 0 -0.144

(0.30) (0.26) (0.19) (0.30) (0.29) (0.20) (0.31) (0.32) (0.22)

Deal size -0.544 -0.173 -0.234 -0.601 -0.19 -0.22 -0.508 -0.079 -0.086

(1.06) (0.16) (0.23) (1.11) (0.17) (0.23) (1.20) (0.32) (0.37)

Ownership asymmetry -0.153 1.048* 0.18 -0.183 1.038* 0.182 -0.559 0.353 -0.18

(0.56) (0.49) (0.37) (0.56) (0.49) (0.37) (0.59) (0.57) (0.42)

Prior joint venture experience 0.153 -0.859 0.006 0.139 -0.711 0.078 0.156 -0.486 0.129

(0.62) (0.55) (0.39) (0.63) (0.56) (0.40) (0.66) (0.64) (0.45)

Contract length 0.647 0.980** 1.176*** 0.649 1.193** 1.259*** 0.654 1.307** 1.382***

(0.40) (0.35) (0.29) (0.41) (0.38) (0.31) (0.45) (0.43) (0.35)

Year dummies yes yes yes yes yes yes yes yes yes

Inustry dummies yes yes yes yes yes yes yes yes yes

Independent variables

Technology Proximity 0.054 -0.646+ -0.222 0.001 -1.030** -0.585*

(0.27) (0.34) (0.20) (0.34) (0.38) (0.25)

Moderating variable

Joint steeering committee 1.338* 0.74 1.272**

(0.66) (0.70) (0.49)

Technoligical proximity x Joint steering committee 0.145 0.883 0.761+

(0.57) (0.79) (0.46)

Pseudo R2 0.303 0.303 0.303 0.310 0.310 0.310 0.303 0.303 0.303

Log-likelihood -233.612 -233.612 -233.612 -231.213 -231.213 -231.213 -205.857 -205.857 -205.857

Notes: n = 257. *** p<0.001, ** p<0.01, * p<0.05, + p<0.1. Clusterd robust standard error in parentheses

6

Discussion

The ensuing section discusses the empirical findings of this study and the subsequent interpretations I derived. Furthermore, I present the theoretical contributions and implications in light of prior research as well as potential managerial implications. Finally, I elaborate on the limitations of this study and subsequent avenues for future research.

6.1 Theoretical contributions and implications

This study leveraged a unique database consisting of 257 analysed joint venture contracts and the USPTO patent database in order to examine a previously unexamined antecedent of contractual dispute resolution in the form of technological proximity.

In this thesis, I firstly investigated whether partners, who possess similar technological portfolios, prior to establishing a joint venture, contractually impose more complex dispute resolution mechanisms. I sought to clarify proprietary technological knowledge antecedents of contractual dispute resolution in joint ventures. This was done incorporating ideas from transaction cost economics that suggests that complex inter-firm exchanges, such as joint ventures present substantial risks of opportunism (Williamson, 1985). Appropriate contractual arrangements and the subsequent governance it imposes on the relationship, underlies the process to facilitate cooperation, prohibit opportunism and protect priority knowledge of the partners involved (Hackett, 1993). I focussed on approaches to contractual dispute resolution (Brett, Goldberg & Ury, 1990) and its ability to hinder the misuse of sensitive technological knowledge (Lerner & Malmendier, 2010). Whereby approaches involving tri-lateral, legalistic mechanisms may be more appropriate when partners face opportunistic behaviour of their counterparts (Lin & Germain, 1998; Brett, Goldberg & Ury, 1990).

Secondly, I pursued whether the presence of a contractually established joint steering committee mitigated partners from including more complex dispute resolution alternatives. I attributed this argument to the notion that partners may seek adaptive contractual features which may diminish disturbances in the foresight of unforeseen contingencies (Williamson, 1975, 1981). This is because joint steering committees present functions which help partners in their interactions and address potential hazards, which might otherwise go unnoticed and mitigate incipient conflicts (Reuer & Devarakonda, 2016).

contractual complexity in alliances in safeguarding valuable knowledge (e.g., Colombo & Piva, 2018, Devoranka & Reuer, 2018). Thus far, prior research in this field has not focussed on technological proximity as an antecedent of dispute resolutions per se.

The empirical findings in this study revealed no evidence for the proposed hypothesis that a more similar technological portfolio incites partners to opt for more complex dispute resolution mechanisms. Conversely, the results indicate the opposite relationship for the inclusion of arbitration provisions in the contract. Specifically, this study implies that a higher extent of technological similarities indicate that partners are less likely to include a tri-lateral form of dispute resolution in the form of arbitration. As such, a possible explanation for the absence of proof for the proposed relationship may lie in the inherent complexity in the construction of the inter-firm contract. Implying a myriad of other factors may be at play beyond technological proximity. Although the nature of this thesis extends research on the antecedents of contractual dispute resolution by recognizing the role of technological proximity, this study and this dataset may not offer confiding conclusions of this relationship.

Secondly, this study enhances alliance research on the methodological approaches to investigate contractual dispute resolution by presenting arrangements among bi-lateral and tri-lateral approaches. Existent research on dispute resolution recognized the different approaches and mechanisms to handling interfirm disputes (Ury et al., 1988; Brett, Goldberg & Ury, 1990; Lumineau & Malhotra, 2010), as well as that these approaches may not be mutually exclusive (Ury et al., 1988). The methodological approach proposed in this study, through the means of different dispute resolution mechanisms, presented distinctive combinations of previously acknowledged contractual mechanisms to address disputes. In that, these combinations of alternatives range in complexity due the levels of hostility and cost involved (Koolwijk, 2006).