International R&D alliances: The role of governance in realizing innovative potential.

In this dissertation, I examine the determinants of performance in international R&D alliances. A transaction cost framework is developed to address two fundamental questions: (1) How do firms select governance structures for their collaborative R&D? and (2) What are the implications of this choice for R&D alliance performance? Firms entering alliances face considerable moral hazard problems, as partner firm behavior is often unobservable and the costs of opportunism are potentially high. I suggest that effective cooperation in alliances depends on selection of a governance structure that reduces opportunism in a cost-effective way. Cooperation, in turn, determines the extent to which benefits from pooled firm capabilities are realized. In this sense, I suggest a direct link between alliance governance and alliance performance. Opportunism hazards are predicted to be particularly severe in R&D alliances involving a broad scope of activities and a large number of diverse partners, since specification and monitoring of alliance activities is most difficult in these cases. Using transaction cost logic, I argue that an equity joint venture is more effective in such circumstances, since this alliance structure facilitates coordinated adaptation among the partner firms. Bilateral contracts are preferable when opportunism hazards are lower, however, since a joint venture imposes unnecessary bureaucratic costs in this case. In the empirical portion of this study, I test these propositions using two sources of data: the Securities Database Corporation (SDC) alliance database and US patent data from Micropatent. Using a sample of 479 domestic and international R&D alliances in the telecommunications equipment industry, I find that governance is selected based on alliance hazards; higher predicted alliance hazards are associated with the choice of an equity joint venture. Using two alternative empirical methods, mixed support is found for the hypothesis linking alliance governance with performance. Results from the first method suggest that governance, appropriately matched with alliance hazards, elevates performance by over 25%. However, results using the second method are less conclusive: while some support is found for the primary hypothesis, inconsistencies between the two methods make solid conclusions elusive.Ph.D.Commerce-BusinessManagementSocial SciencesUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/132235/2/9959851.pd