The Function And Performance Of Patent System: An Incomplete Contract Approach

In the thesis, patent system is explained as an exchange-for-secrets contract that social planner supplies for the potential innovators, and its basic function is help firms to measure the costs and benefits of innovations and thus make investment decisions. The thesis consists of four chapters in all.At the beginning of chapter I, the introduction of this dissertation, we articulate three fundamental questions: Why do we need a patent system? How does it function? What happens to social welfare if there is a change of patent system? After a comprehensive but selective literature review, we point out the drawbacks in the existent literature that need to be improved. For example, people often discuss patent system in the absence of trade secret, which is always an alternative choice for R&D appropriability. For optimal patent design, an usual assumption is that social planner knows what are the firm's profits and the project's value. But, according to mechanism design literature, reward system will strictly dominate patent system under this condition and there will be no need for "optimal patent design" at all. As to division of profits in accumulative innovations, existing papers often ignore the incentive for the second generation innovators. At last, we define the basic assumptions in the thesis and give an figural illumination for the analytical framework.In chapter II, we construct a basic model to evaluate the functioning and performance of patent system. First, we introduce the conception of project space, in which every project has three dimensions, i.e. project value, the extent of knowledge tacitness and the extent of spillover effect. Trade secret profit plays as innovator's rationality constraint for whether to accept patent contract. It is shown that the most important contribution of patent system to social welfare is that it changes the acceptance set for innovators, especially makes possible the projects that have very high project value but very low extent of tacitness, and thus will not be invested in the absence of patent system. In a poor patent system, strengthening protection will definitely increase social welfare. However, in a strong patent system, its welfare effect may be positive or negative, which depends on more factors. Then, we prove an equivalence theorem, and justify the simplification of patent contract in the basic model.Furthermore, we point out the intrinsic contradiction in the concept of the effective patent life concept advocated by O'Donoghue et al (1998) and suggest a more reasonable index O, the extent of entire protection. Finally, as a natural and immediate application of the basic model, we revisit the famous Needham's Puzzle, and argue that patent system may be the very key to it. Of course, this is merely an alternative hypothesis that needs to be more confirmed.Chapter III focuses on innovators' patenting decision, and consists of two parts. In part one, by introducing patentability, pecuniary cost and waiting time of patenting application, we reconsider the patenting subgame in the chapter II, and define one pair of discriminate "patent tools". Pecuniary cost is biased against projects with low value while waiting time against those with high value. Without censor capability limit, social optimality requires that both pecuniary cost and waiting time be zero. However, with censor capability limit, an increase of pecuniary cost, by discouraging patenting of the projects with low value and decreasing waiting time, will enhance the patenting incentives of the projects with high value and finally result in higher level of social welfare. In part two, we develop a simple two-stage information disclosing model to evaluate intermediate innovation's patenting in an accumulative context. It shows that the level of patent protection, or the value of licensing rate, will be decisive. The stronger the protection is, the higher the patenting incentive. In this sense, it will be of vital importance to scrutinize the mechanism of licensing between sequential innov...