Project interface choice and knowledge creation: Evidence of international science cooperation in Taiwan

Highlights

• The endogenous nature of governance decisions.

• The project interface matters with international science cooperation.

• Team competence and inter-team trust are more critical than project interface.

• The governance misfit hurts knowledge creation.

• Asymmetrical performance decay in the cross-country science collaboration.

Abstract

To heed the recent calls for more inquiries into the endogenous nature of governance decisions, this study aims to develop a comparative perspective on the performance implication of the project interface design in the context of international scientific cooperation. Using Heckman's two-stage model in a sample of 217 international scientific cooperation projects sponsored by the Taiwanese government, this study shows that project performance characterized by the collective knowledge creation is explained more by team competence and inter-team social capital than by the project interface choice per se. Furthermore, the knowledge co-creation performance is impaired in the presence of project interface misfit, either over- or under-guarding the knowledge exchange in science cooperation. The findings not only yield implications for managing international cooperation in scientific research but offer comparability and equivalency with the organization-level studies characterized by knowledge exchange/creation and governance difficulties.

Introduction

Given the high failure rate of cross-country collaboration burgeoning in the academia, a project interface design, as a governance choice of the knowledge exchange in imperfect markets, has long been considered critical to the project outcomes. However, as the endogenous nature of governance decision has not yet been well factored into the project management research, this study aims to explore whether the knowledge co-creation performance of a science collaboration project is explained by the implicit factors underlying the interface design more than the interface choice per se, and if the knowledge governance misfit, in terms of under- or over guarded project interface, causes performance decay in an asymmetrical manner.

On the theoretical front, the high failure rate of cross-border collaboration is attributed not only to the difficulties in cross-cultural adaptation (Duan et al., 2010; Hemmert et al., 2014) but to the transactional hazards caused by external uncertainty and behavioral opportunism (Bosse and Alvarez, 2010; Brahm and Tarziján, 2014). The literature of transaction costs economics (TCE) has prescribed either a structural or relational form of governance design for the project interface which dictating the ways of task division and knowledge exchange (Faems et al., 2008). The former emphasizes the importance of structural contracts or formal agreements in writing between exchange parties, which tasks are well defined and divided beforehand (Reuer and Ariño, 2007), while the latter focuses on relational processes within ongoing bilateral relationships, and stresses the role of trust for binding and coordinating all sides involved in the collaboration (Jiang et al., 2013). The two extreme modes not only reflect the different logics of project interface design but also dictate the degree of contractual formalization and guide the division of tasks. Although the two governance modes are usually viewed as equivalent or substitutes, their complementarity has also been reported in certain circumstances (Poppo and Zenger, 2002), implying the endogeneity of governance decisions. In light of the disparity and interdependence of two governance modes, this study views a choice to take either a formal or flexible interface a critical decision in the beginning of any cross-country scientific cooperation project.

Alliance studies have attempted to identify the antecedents of specific governance decisions that are used to reduce uncertainty and risks arising from knowledge exchange between partners. For example, the seminal works of Teece (1986) suggest that contractual solutions to innovators are determined by the appropriability regime, the status of in the dominant design paradigm and the presence of complementary assets, while Chesbrough (2008) further demonstrates how contracting mechanisms help a firm leverage other firms' knowledge assets. In addition to contractual solutions in the imperfect knowledge exchange market (Bosse and Alvarez, 2010; Choi and Contractor, 2016), Blomqvist et al. (2005) view trust as a complementary mechanism to bridge the differences between partners' nonhomogeneous backgrounds and culture. However, there are still some gaps for our understanding of high failure rates of cross-border collaboration. The first gap refers to the governance choices for non-economic exchanges, such as mutual learning in science collaboration projects. In contrast to the make/buy decision in commercial transactions, a science project governance is concerned more with the efficiency of cross learning than with the risk of losing one's own proprietary knowledge to other side (Foss, 2007). However, the value-driven in lieu of hazard-aversed project interface between research teams has been sidelined in the TCE literature. The second gap is about the normative implication of governance choices (Gulati and Nickerson, 2008); it has been questioned partly because mixed findings on the performance effect of governance decisions may be a consequence of self-selection bias (Leiblein, 2003), particularly when a majority of studies have assessed the performance implication of distinct governance choices without accounting for the fact that firms purposely, instead of randomly, choose certain governance modes (Hamilton and Nickerson, 2003). Moreover, although the TCE theorists postulates the efficiency-maximizing effect of governance fit (Castañer et al., 2014), it remains under-explored how a governance misfit hurts an exchange (Reuer and Ariño, 2002), not to mention a further probe into the asymmetrical performance decay effect of under- and over-governed exchanges.

To heed the recent calls for more inquiries into the complex nature of governance choices in various contexts (Leiblein, 2003; Geyskens et al., 2006), this study aims to develop a comparative perspective of the performance effects of scientific projects. Viewing project interface design as a proxy of governance choices for knowledge exchange in the international science collaboration, this study sets out to investigate the extent to which a project interface choice is shaped by the imperfectness of knowledge markets, and, in turn, affects project performance measured by knowledge co-creation. Furthermore, considering the necessity of a multi-theoretical lens upon the context of international science collaboration, this study examines if the implicit factors at the team- and dyadic levels, namely, team competence and inter-team social capital, also yield influences, along with the project interface choice, on project performance. The potential effect of unobserved heterogeneity among scientific cooperation projects is captured by using the Heckman two-stage model (1979), which simultaneously tests the relative influences of observed and unobserved determinants of a specific project interface choice on project performance. Finally, the study aims to probe into the performance decay problem in the presence of the misfit between knowledge market conditions and interface choices.