中国汽车行业技术效率及影响因素实证研究

【作者】 李万文；

【导师】 陆文聪；

【作者基本信息】 浙江大学 ， 产业经济学， 2008， 硕士

【摘要】 汽车行业历来是各个工业化国家的主导产业,中国汽车行业经过了50多年的发展历程,基本建立起了比较完整的汽车行业产业链,成为新的拉动宏观经济增长的主导产业。在这个发展过程中,效率是令人感兴趣的研究主题。中国汽车行业上世纪大部分时间,是在缺少足够市场竞争的环境中发展。是否可以认为,在这期间中国汽车行业的效率较低?而自2002年起汽车行业新的一轮发展周期中,汽车行业景气指数直线攀升,这期间汽车行业的效率是否随之提升?哪些因素对中国汽车行业的效率有着重要的影响?基于对以上问题的兴趣,本文对现有中国汽车行业的研究文献进行了回顾,发现目前国内对汽车行业的研究虽然涵盖了很多主题,但研究效率的文献却不多,而且这些研究在数据、实证模型、研究方法上都存在一些缺陷。如使用企业数据估计行业生产函数、简单的使用C-D生产函数而没有进行检验、使用行业时间序列数据而不是面板数据、使用单一的研究方法等等。因此本文使用行业面板统计数据,使用多种模型和方法对中国汽车行业技术效率和影响技术效率的因素进行了深入研究。文献回顾之后,本文对中国汽车行业的现状进行分析。在此基础上选择适当投入产出指标,收集所需统计数据。随后构建模型对中国汽车行业技术效率进行实证研究,模型使用随机前沿生产函数和数据包络分析模型。接下来对技术效率的影响因素提出假设,并通过结构方程模型的实证研究,验证和修改假设。最后得出以下结论:中国汽车行业现状分析部分,本文认为中国汽车行业在成长的过程中,呈现了一些重要的特点:（1）并购重组日益增多,规模经济逐步体现。（2）自主品牌的力量壮大。（3）产业链日渐完善,中国的汽车服务业迅速发展。（4）商用车的国际市场份额扩大,具备较强的国际竞争力。中国汽车行业技术效率实证研究表明,中国汽车行业使用包含技术进步的超越对数生产函数模型是合理假设。中国汽车行业自1992年以来,存在着非中性的技术进步,投入要素之间的边际技术替代率发生显著变化。汽车的生产过程中,存在着随时间变化的显著的非效率项。进一步研究中国汽车行业技术效率的区域分布后,本文得出一些重要的结论。（1）中国汽车行业技术效率和产量的区域分布基本一致。汽车产量较高的省如上海,比产量较低的省如河北的效率均值要高20%。但产量并不是决定技术效率的唯一因素,有些区域虽然产量不高,但在细分市场占据领导地位,同样具有较高的效率。（2）1994年以前,中国汽车行业有着较高的关税壁垒和价格管制。汽车产量虽然不高,但在政策保护下,以产值为产出衡量单位的技术效率较高。90年代后期随着关税的降低和市场竞争的加剧,至1998年效率均值下降了30%。1998年之后技术效率开始上升,2006年的效率均值达到0.95,说明通过改进效率仅能提高5%的产出。（3）技术效率的区域分布在15年间经历了较大变化。有的区域如广东,效率持续提升;有的区域如天津,效率持续下滑。随后本文提出了中国汽车行业技术效率受企业、行业、市场力量、国际因素、产业政策五个方面影响的假设。使用结构方程模型进行实证研究后,得出如下结论:（1）成本费用利润率、研发人员比重与技术效率的关系为正。这说明可以通过加强企业管理和研发人员的投入提高行业技术效率。（2）外商直接投资和汽车产品进出口金额的上升,对于中国汽车行业技术效率的改进有积极的影响。（3）生产集中度CR₅对技术效率的积极影响,显示中国汽车行业的规模经济得以体现。（4）2002年以后关税的进一步降低和其他产业政策共同促进了中国汽车行业效率提高。本文的主要创新之处在于本文采用了更规范的研究方法,对中国汽车行业的效率进行了深入的研究,并得出了一些有意义的结论。本文对中国汽车行业生产函数的设定使用了一般的超越对数模型,并通过相关假设检验,验征假设的正确性。在投入指标选取时,使用因子分析研究投入指标的相关性。在技术效率实证研究中,同时使用随机前沿分析和数据包络分析。并将结果相互验证,这样避免了单一方法估计的误差。行业面板统计数据的使用,也提高了研究的准确性。由于技术效率的影响因素错综复杂,本文考虑了市场结构、外商直接投资、进出口、产业政策等多种影响因素,并使用可以处理潜变量的结构方程模型进行实证分析。这些都是现有中国汽车行业效率研究中很少出现的。本文的工作尚属探索性的实证研究,限于时间和精力,本文在数据和理论支撑方面还有许多不足。更多还原

【Abstract】 Automobile industry has always been a leading industry in many industrialized countries. After more than 50 years of development history, China’s automobile industry has established a relatively complete industrial chain, and becomes a new leading industry which stimulates the macroeconomic growth. In the process of development, efficiency is an interesting research topic. For most of the last century, China’s automobile industry developed in an environment which lacked of adequate market competition. Is it true that China’s automobile industry is less efficient in this period? Since 2002, a new round development cycle began, the automotive industry sentiment index rose line, did the efficiency of the automobile industry rise too during the period? Which factors have important impact on the efficiency of China’s automobile industry?Because of the interesting in above questions, we reviewed the existing literatures about China’s automobile industry and found that the research on the efficiency of China’s automobile industry was not enough, although many other topics had been covered. And these studies had some drawbacks in the data, empirical models and study methods, such as only using corporate data to estimate industry production function, simple using of the C-D production function without inspection, just using time-series data rather than panel data and using a single research method and so on.We use industrial statistical panel data and a variety of models and methods to research the technical efficiency of China’s automobile industry in-deepth and factors impacting the efficiency. After reviewing of the literature, we analyze the status of China’s automobile industry, on this basis, selecting the appropriate input and output indicators, collecting the necessary statistical data. Then, we construct the model to do empirical study on technical efficiency of China’s automobile industry, using stochastic frontier production function model and data envelopment analysis model respectively. Next, we promote an assumption of factors impacting the technical efficiency, testing and modifying the assumptions through empirical research by using of structural equation model. Finally we come to the following conclusions:In the section of China’s automobile industry status analysis, we believe that there are a number of important features in the developing process of China’s automobile industry: (1) Merging gradually increased, reflecting economies of scale. (2) The power of China’s brands expanded. (3) Industrial chain increasingly improved and China’s automobile service industry developed rapidly. (4) China’s commercial automobile industry expanded the international market share, with strong international competitiveness.The empirical research on China’s automobile industry technical efficiency shows that it is a reasonable assumption that China’s automobile industry production function is a trans-log function form, including the technological progress. Since 1992, there is a non-neutral technological progress which means the marginal technical elements alternative rate has a significant change in the China’s automobile industry. And there is significant non-efficiency item that changes with time in automobile production process. After further studying the regional distribution of technical efficiency, we come to some important conclusions. (1) China’s automobile industry technical efficiency and yield have the similar regional distribution. The provinces which have higher production, such as Shanghai, have higher efficiency than the lower output province, such as Hebei, by 20 percent. But output is not the sole factor which impacts technical efficiency. Because some regions have high efficiency also, as a leader in some segments market, although their production is not high. (2) Before 1994, China’s automobile industry had a fairly high tariffs and price controls. Vehicle production was not high, but under the protection of the policy, the efficiency which output is valued using money was still high. In the late 1990s, as the tariffs reduced and market competition intensified, there was a 30 percent decline in average efficiency till 1998. Then, it started to rise from 1998, the efficiency in 2006 reached 0.946, which means that it is possible to increase output by 5 percent, only through improves efficiency. (3) The regional distribution of technical efficiency experienced substantial changes in the past 15 years. Some regions, such as Guangdong, steadily improved efficiency while other regions, such as Tianjin, continued decline in efficiency.Then, we give an assumption that the technical efficiency of China’s automobile industry is impacted by five factors: enterprise, industry, market forces, policy, and international factor. After empirical study using structural equation model, we come to the following conclusions: (1) the cost of profit margins and the proportion of R&D personnel has positive relationship with technical efficiency which shows that it is possible to improve technical efficiency through strengthen the management of enterprises and R&D staff input. (2) Foreign direct investment and the amount of import and export for China’s automobile industry have a positive effect on technical efficiency. (3) Production focus degree CR5 shows positive impact on the technical efficiency reflecting the economies of scale of China’s automobile industry. (4) Further reducing tariffs and other industrial policies promulgated jointly promote China’s automobile industry more efficiency after 2002.The main innovation of this paper is that we conducted in-depth research on the efficiency of China’s automobile industry and reached some useful conclusions, using more standardize research method. In this paper, we assume China’s automobile industry production function is a trans-log form, and through hypothesis testing to validate if the assumption is correct. When select the input and output indicators, we use factor analysis to research the relevance of input indicators. In the empirical study on technical efficiency, we use stochastic frontier analysis and data envelopment analysis at the same time. Meanwhile, we contrast the two results to avoid the estimated error when use a single method. The use of industry statistical panel data also improves the accuracy of the study. Because of the complex factors that impact technical efficiency, our thinking cover the market structure, foreign direct investment, import and export, industrial policy, and other factors. Besides, we use of structural equation model which can dealt with latent variables to began empirical analysis.This paper is just an exploration work and has many deficiencies in theoretical supporfc and data, with limited time and energy.更多还原