Specification and Estimation of Technologies

技术规范和评估

• 批准号: 8720761
• 负责人: Robert Pollak
• 金额: $ 2.7万
• 依托单位: University of Washington
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1988
• 资助国家: 美国
• 起止时间: 1988-09-15 至 1990-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=8720761&HistoricalAwards=false
• 关键词: Specification; Estimation; Technologies

This project examines three topics in production analysis. These are: returns-to-scale and technical progress, the effects of imposing regularity conditions, and direct estimation of production functions from data on inputs and output. Empirical estimation of the returns-to-scale and changes in technical proficiency have long been central questions in the analysis of production. Past work has typically specified constant returns and technical progress due to the inherent econometric tractability. This project extends that work by incorporating production functions which are non-homothetic in returns and technology, and are more realistic, general, and flexible. Investigation of the imposition of regularity conditions at every point in the sample data addresses an important but often neglected issue in empirical production analysis. The theory underlying the economics of production assumes that the conditions of monotonicity and convexity apply to every combination of inputs and output. Empirical data often do not meet those conditions, and this project systematically examines the effects of violating the regularity assumptions. The comparison of direct estimation of production functions with the indirect method makes a fundamental contribution in empirical econometrics. Estimating functions directly has the advantage of not requiring data on factor prices nor does it assume cost minimization. Estimating input demand systems with factor price data is often econometrically intractable. This project addresses some important problems in the economic analysis of production. Economic theory in regard to production is quite well developed; indeed analyzing the interaction of production and consumption has been a fundamental endeavor for many years. From an empirical standpoint, however, the analytical techniques are much less straightforward. Transferring the conditions imposed by economic theory into empirical econometric analysis has proven to be quite difficult, and often restricting assumptions about the data must be made to render empirical analysis tractable. These distinctions between the theoretical models and the more tractable but simpler empirical models are very important in that most decisions made by production managers and policy makers are based on empirical analysis. This work promises to provide insight into the importance of some often-used empirical assumptions. By using various empirical specifications this project examines the effects of some fundamental assumptions in production theory. These are: 1) returns-to-scale, which is the manner in which product output changes with respect to changes in inputs. Of particular importance is the assumption that output will expand or contract in exact proportion to changes in input, 2) regularity conditions, which imply that the theoretical properties of the production function are met by all sample data values, 3) direct versus indirect specification of the production function, the former specification incorporating data on input and output quantities, the latter using only data on the prices of inputs.

本项目考察生产分析中的三个主题。它们是：规模回报和技术进步，施加规则条件的影响，以及根据投入和产出数据直接估计生产函数。长期以来，规模收益和技术熟练程度变化的实证估计一直是生产分析中的核心问题。由于固有的计量经济可追溯性，过去的工作通常指定恒定的回报和技术进步。本项目扩展了这项工作，纳入了在回报和技术上不一致的、更现实、更普遍和更灵活的生产功能。在样本数据的每一点的规则条件的强制调查解决了一个重要的，但往往被忽视的问题，在经验生产分析。作为生产经济学基础的理论假定单调性和凸性条件适用于每一种投入和产出组合。经验数据往往不满足这些条件，本项目系统地检查违反规则假设的影响。生产函数的直接估计与间接估计的比较在实证计量经济学中作出了根本性的贡献。直接估算函数的优点是不需要要素价格的数据，也不需要成本最小化。用要素价格数据估计投入需求系统在计量经济学上往往是棘手的。本项目解决了生产经济分析中的一些重要问题。关于生产的经济理论相当发达；事实上，多年来，分析生产和消费的相互作用一直是一项基本工作。然而，从经验的角度来看，分析技术远没有那么直截了当。事实证明，将经济理论所施加的条件转化为经验性计量经济学分析是相当困难的，而且为了使经验性分析易于处理，通常必须对数据作出限制性假设。理论模型和更容易处理但更简单的经验模型之间的这些区别非常重要，因为生产经理和政策制定者做出的大多数决策都是基于经验分析。这项工作有望提供对一些经常使用的经验假设的重要性的见解。通过使用各种经验规范，本项目检验了生产理论中一些基本假设的影响。它们是：1)规模回报，即产品产出相对于投入变化的方式。特别重要的是假设产出将与投入的变化成正比地扩大或缩小，2)规律性条件，这意味着所有样本数据值都满足生产函数的理论性质，3)生产函数的直接或间接规范，前者包含投入和产出数量的数据，后者只使用投入价格的数据。