【作者】 赵丹；

【导师】 王宗军；

【作者基本信息】 华中科技大学 ， 工商管理， 2012， 博士

【摘要】 作为当今除自主创新以外最迅速和有效提高企业技术创新能力的方式——技术许可,对其进行研究具有两方面的意义：其一,对于我国企业而言,由于整体研发能力较弱,一味鼓励所有企业进行创新,这不仅导致投入较大,而且效果较差,很可能得不偿失。若借助技术许可则可能缩短与国外研发效率较高企业的差距,提高我国企业自主创新的基础,为以后“蛙跳”做准备。二战后日本在几十年里迅速赶超欧美的经验就是值得借鉴的例子。其二,对于我国创新能力较强、研发效率较高的企业而言,特别在为核心技术申请专利之后,如何更快回收研发投入、获取更高经济利润是一个值得考虑的问题。而技术许可则为研发效率较低企业缩短成本差距、研发效率较高企业提高投资回报率等问题提供了一个可行的途径。一般而言,对技术许可契约的设计主要包括许可机制以及许可证发放数。因此如何设计最优的技术许可契约成为值得深入研究的课题。本文在寡占理论、产业组织理论、博弈论以及技术许可理论的基础上,利用定量和定性相结合的方法,特别在对博弈模型求解时选用了非合作博弈论中求解子博弈完美解所用到的逆推归纳法作为工具,对在位创新企业和外部创新机构在不同阶段(具体指研发成功前、研发成功后但专利有效性确认前、专利有效性确认后的正常产品市场和网络产品市场以及存在研发政策影响时的开放市场)时的技术许可行为进行了详细且深入的研究。在研发成功前,由于此时的研发结局存在不确定性,研发成功的概率可能会影响到最优技术许可的选择。第2章所建立的三阶段双寡头博弈模型的结论表明：许可方企业应根据研发成功的概率大小以及研发溢出程度,进而确定研发投入的高低,来选择不同的许可策略。当研发成功后但专利有效性确认前,研发投入变成了沉没成本。由于存在被潜在受许企业拒绝接受许可后通过法院诉讼来主张专利无效的可能,这时弱专利有效性成为这一阶段关注的重点。本文还证实了Farrell和Shapiro(2008)的部分结论,但弱专利只在某种条件下存在可能导致“柠檬”市场的过度补偿问题。最后还提出——禁止专利持有方拒绝许可技术给曾经不成功挑战其专利有效性的企业以及鼓励潜在受许企业形成战略诉讼联盟对其权威提出挑战,可能是解决过度补偿的对策。在专利有效性确认后的正常产品市场,本文得出产品异质性、研发溢出、企业吸收能力、创新规模、创新后企业间成本差异以及创新企业的讨价还价能力会对在位企业最优许可策略的选择产生重要影响。而在网络产品市场,市场规模、产品替代程度、网络强度、市场集中度以及研发效率在决定在位企业最优许可策略的选择以及许可证数的发放上起关键作用。本文还验证了正常产品市场上在无约束条件下两部制许可为最优许可策略的结论,然而在网络产品市场上并非总是成立。最后则解释了现实中多种许可机制并存的原因。当存在外部干预(政府研发政策)时,政府会根据竞争市场(许可方市场、受许方市场还是第三方市场)以及创新规模等关键因素选择不同的研发补贴政策,进而影更多还原

【Abstract】 There are significance of both for research on technology licensing which is today’s one of most rapid and effective ways to improve ability of enterprises’ technological innovation, besides independent innovation. Firstly, as far as Chinese enterprises are concerned, owing to the overall weak R&D capability, if we encourage all enterprises to innovate blindly, negative results can arise, even outweigh benefits. With technology licensing, the gap between our country firms and foreign high R&D efficient enterprises is possible to shorten, ability of Chinese enterprises’ independent innovation can be improved and prepare for further leapfrog. Such the example worthy learning is that Japan in last decades caught up Europe and paralleled United States experience After World WarⅡ. Secondly, Chinese enterprises with high ability to innovate need to consider how to leap R&D investment and obtain much economic profit. Technology licensing is one of feasible paths to resolve such questions as shorten cost gap between china vs. others developed country and increase return on investment. In general, designs of technology licensing contracts include licensing mechanism and number of licenses mainly. Therefore, how to design the optimal technology licensing contract is becoming a subject worthy of further study.Based on oligopoly theory, theory of industrial organization, game theory, and technology licensing theory, this paper makes a detailed and thorough research upon technology licensing behavior of incumbent innovative enterprise and outsider innovative institution at different stages (including R&D pro-success, R&D success but patent effectiveness pro-validation, patent effectiveness validation on normal-product market and network-product market and open market with R&D policy interference)with the combination of quantitative and qualitative methods, specially making use of backward induction to solve the sub-game perfect solution usually in non-cooperative game theory.Because of R&D outcome is uncertainty before R&D success, possibility of R&D success may affect the choice of optimal technology licensing. Chapter 2 in the paper establishes a three-stage duopoly game model, and makes conclusions that:licensor should base on possibility of R&D success and R&D spillover, and thus determine level of R&D investment in order to choose different licensing strategies.When patents have not validated after R&D success, R&D investment becomes a sunk cost. Weak patent effectiveness is paid much attention at this stage because potential licensees who refused licensor’s offer may assert the patent invalid through litigation. The paper has also confirmed part of conclusions of Farrell, and Shapiro (2008), but overcompensation question which may result in lemon market for weak patent arises only under certain conditions. Finally, counter-measures that prohibiting patent-holder refusing to license technology to unsuccessfully challenger and encouraging potential licensees to establish strategic litigation alliance for challenging the authority, are proposed to solve overcompensation.The paper makes some conclusions that combinations with product heterogeneity, R&D spillover, enterprise absorptive capacity, innovation size, cost difference between enterprises after innovation and bargaining power of innovative enterprise, have crucial effect on choice of optimal licensing strategy of incumbent in normal product market after patent effectiveness validation. In network product market, market size, degree of product substitution, network strength, market concentration and R&D efficiency play a key role in decision on both choices of the optimal licensing strategy of incumbent firms and number of licenses. This paper also verify conclusion that two-part tariff licensing in non-binding constraint is optimal licensing in normal product market,which not always hold in network product market. Finally we explain the coexistence of a variety of licensing mechanisms in reality.When external intervention (government R&D policy) is present, government R&D subsidy policy will from change to change according to competitive market (licensor’s market, licensees’ market or third-party market), as well as such key factors as innovation size, thereby affecting corporate R&D motivation and licensing behaviors.更多还原