Information Processing In Firms

对企业规模收益的研究传统上侧重于生产过程中的技术规模收益。然而，随着公司生产过程的规模扩大，其管理机构也在扩大。这一行政机构的大部分任务是信息处理，包括沟通和决策。这项任务对公司的业绩很重要，因此与规模回报相关。这项研究使用了随机控制模型和组织中信息处理的并行处理模型，该模型具有显式且昂贵的策略计算。预计将获得各种结果，其回报取决于对要处理的信息的统计性质和损失函数的假设。进一步的研究包括规模的区位模型，企业的显式模型，其规模是由物理投入和产出而不是环境的信息大小衡量的，以及信息处理的分布式计算模型。这项研究跨越了计算机科学和经济学之间的前沿。它涉及信息处理、组织设计和功能，并将增进我们对一个长期存在的、在经济学中具有重要意义的问题的理解。

The study of returns to scale in firms has traditionally focused on technological returns to scale in the production process. However, as the scale of a firm's production process grows, so does its administrative apparatus. The task of much of this administrative apparatus is information processing, including communication and decision-making. This task is important to the performance of the firm, and hence relevant to returns to scale. This research uses a model of stochastic control with explicit and costly computation of policies, coupled with a parallel processing model of information processing in organizations. It is expected that a variety of results will be obtained for which returns depend on assumptions on the statistical properties of the information to be processed and on the loss function. Further studies include a locational model of scale, explicit models of the firm with scale measured by physical inputs and outputs rather than by the informational size of the environment, and a distributed computational model of information processing. This research works across the frontier between computer science and economics. It relates information processing and organization design and function, and will enhance our understanding of a problem of long standing and great importance in economics.