Industrial Productivity, Urban Growth, and Air Quality Regulation

该项目有三个相互关联的部分，一个统一的主题是受集聚经济、地理和监管影响的工厂、行业和人口的区位和集聚格局的动态变化。另一个统一的元素是所利用的数据。第一部分考察了环境监管对主要污染行业的影响，其中除了研究监管对地点和产业结构的影响外，重点还涉及不同地点、不同时间以及工厂规模和年限的工厂运营效率。各工厂的资本强度按年龄比较，在严格的环境监管出现之前和之后，以及在受管制的未达到(‘处理’)和较少或不受管制的(‘控制’)地区之间。估计准固定要素成本函数，以便能够计算资本的影子边际产品和有效经营规模。需要分析的问题包括环境引起的投资在多大程度上有助于提高生产率，以及监管后和未达标地区的工厂相对资本过剩和规模相对较小的程度和成本。该提案的第二部分考察了高科技和传统产业的空间演变，以及当地静态和动态外部性对植物生产力的影响。城市间高技术就业分布的空间演化模型为平稳的一阶马尔可夫过程。除了研究“趋同”的问题外，目标是确定哪些城市的选址特征定义了哪些行业的良好选址。城市高科技“奇迹”在多大程度上是可预测的？场地特征包括预定的城市人口和就业特征、外生地理特征和邻近城市的特征。例如，如果基期大都市区人口是一个关键特征，那么还有一个问题是为什么。为了回答这个问题，我们将考察城市外部性，以确定它们是静态的还是动态的，以及它们是否是本地化与城市化的。第三部分主要论述城市的演变。该理论结合了在一个经济体中产生多个有限规模城市的地方规模经济与不经济的城市要素，以及通过涉及本地化知识溢出的人力资本投资实现的外生国家人口增长和内生经济增长。除了对城市演变进行建模外，这个想法是为了研究城市之间不断演变的收入不平等。在任何时刻，不同的商品和城市类型通常会有不同的人均人力资本均衡水平和实际可观测收入。此外，工人可能会根据能力水平在不同城市之间进行排序。一个关键问题涉及本地化知识溢出、父母教育、能力代际传递、城市生产模式和收入不平等之间的相互作用。实证部分考察了美国从1900年到现在的城市演变。这项研究使用了三个主要的数据集。首先是1963年至1994年的纵向研究数据库，研究工厂生产率、大都市区工业环境和环境法规(还结合人口普查局的污染治理成本和支出数据)。第二个是美国19001990个县的县级数据集，这些县随着时间的推移一直被定义，涵盖了人口普查和行业特征，并纳入了地理特征。最后是1977-1995年的年度县商业模式数据，给出了所有私人就业的行业细节，并结合了关于政府就业和城市特征的人口普查数据。

This project has three interrelated parts, with a unifying theme being the dynamics of patterns of location and agglomeration of plants, industries, and populations as affected by agglomeration economies, geography, and regulation. Another unifying element is the data utilized. The first part examines the effects of environmental regulation on major polluting industries, where, apart from studying regulatory effects on location and industrial structure, a focus concerns plant operating efficiencies in different locations, over time, and by plant size and age. Capital intensities across plants are compared by age before and after the advent of strong environmental regulation and between regulated non attainment (`treatment`) and less or unregulated attainment (`control`) areas. Quasifixed factor cost functions are estimated so as to be able to calculate shadow marginal products of capital and efficient operating sizes. Issues to be analyzed include to what extent environmentally induced investments contribute to productivity and to what extent and at what cost plants after regulation and in non attainment areas are relatively overcapitalized and relatively under sized. The second part of the proposal examines the spatial evolution of high tech and traditional industries and the effects on plant productivity of local static and dynamic externalities. Spatial evolution of distributions of high tech employment across cities are modeled as a stationary first order Markov process. Apart from studying issues of `convergence,` the object is to determine what urban site characteristics define good sites for what industries. To what extent are urban high tech `miracles` predictable? Site characteristics include predetermined city population and employment characteristics, exogenous geographic characteristics, and characteristics of neighboring cities. If, for example, base period metro area population is a key characteristic, there remains a question of why. To answer this question, urban externalities will be examined to determine if they are static versus dynamic and if they are ones of localization versus urbanization. The third part focuses on urban evolution. The theory combines the urban elements of local scale economies versus diseconomies yielding multiple finite size cities in an economy, with exogenous national population growth and endogenous economic growth through human capital investment involving localized knowledge spillovers. Apart from modeling urban evolution, the idea is to study evolving income inequality across cities. At any instant, different goods and city types will generally have different equilibrium per person human capital levels and real observed incomes. Moreover workers may sort across cities by ability levels. A key issue concerns the interaction among localized knowledge spillovers, parental education, inter generational transmission of ability, urban production patterns and income inequality. The empirical part examines urban evolution in the USA from 1900 to the present. The research uses three major data sets. First is the Longitudinal Research Database from 1963 1994 to study plant productivity, metro area industrial environments, and environmental regulation (combined also with Pollution Abatement Costs and Expenditures data of the Census Bureau). Second is county level data set for 19001990 for all USA counties consistently defined over time, covering Census population and industrial characteristics and incorporating geographic characteristics. Finally there is annual County Business Patterns data for 1977 1995, giving industry detail on all private employment, combined with Census data on government employment and urban characteristics.