Three essays on R&D cooperation among firms

Abstract

The dissertation consists of three chapters. The first chapter examines the functional specialization of SMEs' technological competence and its moderating role in the effect of external R&D on their innovative performance. Technological competence consists of many functional dimensions such as basic research, product architecture, process construction, testing, and evaluation, which constitute a sequence of innovation tasks. The specialization of technological competence allows SMEs to utilize economies of specialization in R&D, enhance their bargaining power and appropriability conditions in the process of external R&D, and attract promising R&D partners. However, competence specialization may hamper SMEs' capabilities to coordinate and integrate diverse external R&D projects. Using a sample of SMEs in Korean manufacturing industries, we find the following results. First, competence specialization positively moderates the effect of external R&D on SMEs' innovative performance. Second, the positive moderating effect of competence specialization diminishes as the share of external R&D increases. Third, the moderating effect of competence specialization differs across industries depending on the degree of market dominance by a few large firms (i.e., market concentration) and the novelty of technologies pursued by SMEs in each industry. The second chapter investigates the relationship between the technological distance among joint R&D partner firms and the performance of their research joint ventures (RJVs) by utilizing a unique panel dataset of U.S. RJVs across various industries. Using the number of patents co-produced by an RJV's partner firms as a measure of its performance, we employ zero-inflated negative binomial (ZINB) models with random effects in order to control for RJVs' unobserved heterogeneities in both their performance and propensities to patent. The key findings are as follows. First, the relationship between technological distance and RJV performance is inverted U-shaped, implying that RJV performance first increases up to an optimal level of technological distance and then decreases with technological distance. Second, the degree of market competition among RJV partners decreases the optimal level of technological distance. Third, the optimal level of technological distance increases with RJV partners' average technological competence. The findings show that the beneficial and harmful effects of technological distance on RJV performance are conditioned by the degree of market competition among RJV partners and their average technological competence. Lastly, the third chapter studies asymmetric learning among partner firms in research joint ventures (RJVs) by investigating their differential rate of enhancement in technological competence. We argue that the asymmetric learning among joint R&D partners is primarily attributable to their differences in the potential to access and utilize the knowledge pool created in their RJVs. Using a unique dataset on firms engaging in U.S. RJVs, we show the following. First, the rate of enhancement in a firm's technological competence, as measured by its R&D productivity, during its RJV participation is positively associated with the size of the RJV's knowledge pool. Second, the degree of a firm's learning from the RJV's knowledge pool is largely influenced by the following three factors: its relative technological competence among RJV partners, the technological distance from the partners, and the strength of the partners' R&D appropriability conditions. Specifically, when the technological distance is low (high) or the appropriability conditions are weak (strong), technological laggards (leaders) are better positioned than their counterparts to learn from the RJV's knowledge pool and thereby to enjoy the technological-competence-enhancing effect of joint R&D.