Diffusion and Multiple Equilibria

拟议的研究议程由三个主要部分组成。第一部分主要是对技术创新扩散的实证研究，以及这种扩散如何影响产出的总体行为。更具体地说，它关注于从创新的时间路径和扩散速度来解释国民生产总值的时间序列行为。这被表示为一个随机的简化形式的方程，将GNP中当前的创新与过去具有各种滞后的创新联系起来。然后，使用关于国民生产总值增长率和创新活动的横截面国家数据，例如通过专利申请衡量的数据，来估计方程的参数。研究的第二部分包括关于企业采用技术创新和技术知识传播过程的三个模型的理论发展。这些模型中的第一个试图解释为什么大公司通常最先采用一项新的创新。另外两种模式处理的是信息传播中的摩擦，重点是个人接触。这项研究的最后部分涉及对存在多重均衡所引起的问题的计量经济学调查。在这种情况下，偏好和技术的参数会产生一组均衡，然后从这一组中选择一个结果。争论的问题是，如何估计马尔可夫过程以供选择，因为平衡中的一部分运动将是由于外部冲击的存在。这项基础性研究对技术冲击如何影响长期增长和周期性波动这一总的主题产生了非常重要的影响。这项研究所关注的技术冲击，是那些导致我们的知识存量永久增加的冲击。由于存量太大，任何人都不熟悉，其扩散将决定其对产出、生产率、投资和其他总量的影响。这位研究人员的论点是，这种创新主要发生在公司层面，然后以一种尽管其中心地位仍未得到完美理解的方式在整个经济中传播。已经完成的工作和这一提议计划的进一步工作构成了对这一专题的前沿研究，这在现代“内生增长”理论和现代商业周期理论中具有至关重要的意义。

The proposed research agenda consists of three broad parts. The first part involves a largely empirical investigation of the diffusion of technological innovations, and how this diffusion affects the aggregate behavior of output. More specifically, it is concerned with explaining the time-series behavior of GNP in terms of the time path of innovations and the speed with which they are diffused. This is expressed as a stochastic reduced- form equation relating current innovations in GNP to past innovations with various lags. Cross-section country data on GNP growth rates and innovation activity, for example, as measured by patent applications, are then used to estimate the parameters of the equation. The second part of the study consists of a theoretical development of three models regarding the adoption of technological innovations by firms, and of the process of diffusion of technical knowledge. The first of these models seeks to explain why bigger firms are usually the first to adopt a new innovation. The other two models deal with frictions in the spread of information, with an emphasis on personal contact. The final part of the research involves an econometric investigation of issues arising from the presence of multiple equilibria. In this setting, the parameters of preferences and technology generate a set of equilibria and then an outcome is chosen from this set. The question at issue is how to estimate the Markov process for selection given that part of the movement in equilibrium will be due to the presence of exogenous shocks. This fundamental research bears in a very important way on the general topic of how shocks to technology impact on both long- term growth and on cyclical fluctuations. The technological shocks, on which the research is focussed, are those that result in permanent additions to our stock of knowledge. Since the stock is much too large for anyone to be familiar with, its diffusion will determine its effect on output, productivity, investment, and other aggregates. It is the thesis of the investigator that such innovations occur largely at the firm level, and then spread throughout the economy in ways that despite their centrality are still imperfectly understood. Work already completed and the further work projected by this proposal, constitutes frontier research on the topic, which is of critical importance in the modern theory of "endogenous growth", and in modern business cycle theory.