臺大管理論叢 NTU Management Review VOL.29 NO.1

107 NTU Management Review Vol. 29 No. 1 Apr. 2019 strategy, its purpose of alliance is more likely to discover something new to its current knowledge base; it will focus on the “R” in the research and development process (Rothaermel and Deeds, 2004). The focal firm engaged in an exploratory alliance expects to acquire new knowledge and technologies from the distant parties. From the perspective of strong ties of social network, the internal stakeholders’ knowledge is more likely to be homogeneous and redundant to the focal firm’s existing knowledge base. Such homogeneity and redundancy make the opportunities of radical innovation via combination of new and idiosyncratic knowledge relatively rare. In contrast, cooperating with the external stakeholders (e.g., the local professionals or universities) who mostly do the basic and distant research may let the focal firm gain more new knowledge and technologies. The professionals, research institutes or scholars often patent or publish their research for academic purposes; this may reduce risks of information asymmetry. Additionally, the position of the external stakeholders is relatively interest-neutral so that might mitigate the problem of expropriation. For example, Rothaermel and Deeds (2004) depict the exploratory innovation through the cooperation between a firm and its external stakeholder. In their research, BioGen, a biotech firm, cooperates with the University of Zurich, and this cooperation leads to the discovery of Intron A. Hong et al. (2010) argue that having the R&D interactions with universities is more preferable in the relation- oriented cultures when the knowledge involved is more tacit or personalized. Miotti and Sachwald (2003) claim that the collaboration with public research institutions is the most attractive to firms who undertake R&D at the technological frontier. Therefore, we expect that a firm will prefer to cooperate with its external stakeholders for the exploratory innovation. Hypothesis 2b: When a firm undertakes the exploratory learning, it will be more inclined to select its R&D collaborative partner from its external stakeholders rather than internal stakeholders in the host country. 2.4 Institutional Voids The institutional context is different from country to country. Hitt et al. (2004) suggest that the institutional environment of a host country, specifically its legal aspect, should be considered when studying IJV partner selection. The institutions provide the formal and informal rules that bind and shape behaviors of and interactions among individuals and organizations in societies (North, 1990). The function of institutions in an economy is to lower costs of transactions and information by reducing uncertainty and by