Competition vs. collaboration in the generation and adoption of a sequence of new technology

Although there is quite a rich literature relating to competitive innovation there is \nrelatively little relating to technological collaboration. However, ignoring \ncollaborative possibilities may result in overestimation of the importance of selfinnovation. \nThis thesis is therefore mainly concerned with the determinants of \ncollaboration in innovation, taking both a theoretical and an empirical approach. \nThe empirics relate to the manufacturing industry in a Chinese region. The \nthesis is particularly innovative in emphasising how collaboration costs will be \nshared when collaboration occurs. \nWe provide a game theoretic exploration of the decisions of firms on \nwhether to compete or collaborate in the generation and adoption of a \nsequence of new technologies. Different from the models proposed by Vickers, \nwho concentrates upon process innovation and a two-strategy (innovation or do \nnothing) set, our game theory model emphasises product innovation and either \na three-strategy set (innovation, collaboration, and do nothing), or a fourstrategy \nset (innovation, collaboration, imitation and do nothing). In particular, \nMATLAB programming is employed for generating the equilibrium solution for \neach strategy set. We found that the relationship between imitation and \ncollaboration and collaboration cost is not univariate. It depends upon the \nmarket type and various market characteristics, such as technology gap, \ntechnology level, the product substitution index, transaction costs and the \ndiscount rate of price sensitiveness. The results also show that the elasticity of \ncollaboration opportunity with respect to transaction costs in a persistent \ndominance market is much greater than in an action reaction market. \nBy using data on manufacturing in a Chinese region from 2005 to 2007, \nderived from the China Innovation Survey and the Annual Corporate Financial \nSurvey, we empirically explored innovation and collaboration patterns. Three \nfactors, innovative ability, absorptive capacity, and catching up capacity were \nproposed to positively affect both innovation and collaboration. This led to six \nhypotheses, which were tested using a number of econometric models \nencompassing selection bias, timing, and dynamics issues. The major finding \nfrom the empirical models suggests that innovative ability, absorptive capacity \nand catching up capacity all impact significantly and positively on collaboration, \nwhilst innovation is positively related only to absorptive capacity. Also, we found \nthat collaboration cost may increase with R&D, employees‘ education, the \ntechnology gap and collaboration cost in previous periods, but decrease with \ntransaction cost, patents held, the technology level and perceived price. \nThe thesis makes three contributions. Theoretically, our game theory \nmodel not only extends the understanding of the impacts of collaboration \npossibilities and collaboration cost in dynamic game theory, but also clarifies the \nimpacts of transaction costs and imitation (and thus intellectual property rights \n(IPR)) on the outcome. Empirically, by introducing new data our work is the first \nto investigate collaboration patterns and collaboration cost sharing strategies in \na mid-income level developing country. Last but not least, using MATLAB \nanimation programming to simplify the calculation process of the game theory \nequilibrium may be considered as a methodological contribution.