A Nonsequential R & D Search Model
构建了一个非顺序研发搜索模型，识别出当前技术水平、决策期限和基础研究三个影响研发投入的因素，并证明了保留技术水平的存在，为研发项目管理者何时停止搜索提供了决策规则。
Managers of research and development (R&D) programs are often confronted with questions of what determines R&D spending (i.e. R&D search intensity) and when R&D spending is unprofitable (i.e. Reservation Technology Level). Three determinants of R&D spending are identified: the current technology level, the length of decision horizon, and basic research. Comparative statics of the effects of these determinants of R&D spending are discussed. The higher the current technology level, the less intensive is R&D search because, ceterus paribus, the lower the probability of finding a better technology. The longer the decision horizon, the more intensive is R&D search because fruitful R&D results can be applied to more time periods and generate more present value profits. Finally, technological opportunities can be improved via basic research. The improvement of technological opportunities in turn spurs R&D search. Existence of a Reservation Technology Level (RTL) is established. It constitutes a stopping rule for R&D search. Whenever the current technology level is greater than or equal to RTL, R&D search stops; otherwise R&D search continues. One interesting property of RTL is that it is strictly less than the highest possible technology level. Thus the economic limit of R&D (RTL) differs from the technological limit of R&D. Ex ante, an R&D program manager should not press for the technological limit of R&D once the current technology level exceeds or equals RTL. However, ex post, one may end up with a technology level greater than RTL. Moreover, R&D search may be resumed in an abandoned R&D area if (i) interest rate declines, (ii) R&D quasi-fixed cost reduces, or (iii) basic research provides a favorable shape-preserving shift of the distribution of technological opportunities.