【作者】 陈小毅；

【导师】 周德群；

【作者基本信息】 南京航空航天大学 ， 管理科学与工程， 2013， 博士

【摘要】 产业集中问题是产业组织研究的核心内容之一。通篇论文结合煤炭产业的特殊性，从市场和地理两个维度，侧重于定量、系统分析，务求客观描述中国煤炭产业的集中程度、变化趋势和调整速度，分析市场结构特征及其形成原因，探讨煤炭产业集中与利润率、买方市场势力之间的关系，从而提出相应的政策建议。论文通过理论研究和实证分析，得到一些重要结论：第一，首次测算了1979-2010年中国煤炭产业市场集中度指标，这是论文的重要基础之一。研究发现，2009年以后中国煤炭市场结构出现类似寡头竞争的特征。第二，构建了研究我国煤炭产业市场集中度变化的静态和动态模型。首次分析了集中度影响因素，结论是在我国煤炭工业集中度水平受国家产业政策影响最为显著：产业市场化程度与其负相关；国有重点煤矿产量占比与其正相关。其他重要因素依次是行业利润水平、地理因素和需求因素。首次运用1986-2010年时间序列数据，计算出煤炭市场集中度偏离长期均衡水平时的年均调整速度为8.57%，说明我国目前煤炭市场结构调整速度较快。第三，提出决定我国煤炭产业绩效的理论模型，包含市场集中度、供需基本面、安全研发投入等关键要素。实证结果表明：产业利润率与市场集中度显著正相关；公司利润率与市场份额正相关，而同研发投入负相关。运用CDW方法分析煤炭产业利润率与市场集中度的正向关系，结论是效率因素较市场势力更具解释力。第四，首次量度了我国煤炭产业的买方市场势力，得到趋势向上的结论。目前买方市场势力与煤炭产业集中度高度正相关、与利润率显著正相关，后者与大多数实证结论相悖，原因是目前煤炭需求的特殊性和产业“边际”集中度增速较快。在煤炭细分市场，国有企业寡占的电力产业集中度与动力煤价格高度正相关；低集中竞争的钢铁行业集中度的提升推动了寡占型炼焦煤市场价格，我国经济增长模式和焦煤赋存有限的背景下，焦煤龙头企业间不存在竞争。运用2003-2009年炼焦煤上市公司数据，选择变截距确定效应模型，结论支持Galbraith论断，但是目前钢铁产业集中度的提高对炼焦煤企业利润的削弱作用非常微弱。第五，首次测度了中国煤炭产业地理集中的五个指标：CR10和Herfindahl指数，与美国煤炭产业进行对比，表明我国产业集中程度呈向上趋势，但总体水平低于美国，而且周期性特征更明显；空间Gini系数、E-G指数和M-S指数，发现煤炭产业地理上高度集中，但是迄今为止尚未形成产业集聚，外部性不明显。通过建立确定效应变截距模型考察影响地理集中因素，结论是资源禀赋、规模经济性和市场潜力与产业地理集中存在正相关关系，影响程度依次减弱；产业政策与地理集中程度存在负相关关系，而且目前影响程度远远超过了普遍认为最关键的资源因素。同时讨论了煤电产业地理布局重构的可行性，改变煤炭富集地能源输出结构以确保国家能源安全。更多还原

【Abstract】 Industrial concentration is one of the most important research fields of industrial organization. Thepaper uses a large number of data about market and geography due to the coal industry particularity toshow the industrial concentration and the annual rate of structural adjustment, and analyzes the marketstructure and the causes, and explores the relations between the concentration and the profit rate andthe countervailing power. On the basis of that, I advance some suggestions.After theory and empirical study, I draw some important conclusions as follows:First, this study calculates the concentration ratio (CR) of Chinese coal industry from1979to2010.It is one of the most important bases, which shows the market structure is para-oligarchy competitionsince2009.Second, a steady-state model and a dynamic model are developed to study the market concentrationadjustment of Chinese coal industry. The paper is a first attempt to analyze the industry-specificdeterminants of changes in industry concentration, and it shows that the industry policy plays the mostsignificant role: marketization is negative to the concentration, and the share of state-ownedlarge-sized coal enterprises’ output is positively related to the concentration. The other importantfactors include industrial profit, geography and demand. Using a partial adjustment model, atime-series analysis has been carried out from1986to2010for the first time, and it shows that theannual rate of structural adjustment or8.57%is very fast.Third, this paper presents an academic model about coal industry performance including industryconcentration ratio, supply and demand, safety cost due to the industrial particularity. The empiricalresults are as follows:(1) that the industry profit rate is positively related to the concentration ratio,which cost efficiency plays more important role on than market power by CDW (1984) and (2) thatwithin coal industry, those firms with the larger market shares have higher rates of profit, but anegative correlation between company profit rate and R&D.Fourth, measures of buyer market structure of China’s coal industry, e.g. DSPH, BCR, have beencomputed for the first time. The upward buyer concentration was positively correlated with sellerconcentration highly and seller profitability significantly due to the demand conditions and thedifference between “marginal” concentration’s growth of upstream and downstream industries. Theconcentration ratio of electrical industry controlled by five state-owned enterprises was positivelycorrelated with the steam coal prices, and the oligopolistic coking coal market experienced an inclinein prices while the iron and steel industry that was competitive became more concentrated. There is noexplicit competition among those leading coking coal enterprises because of economic growth mainly depending on the input of resources and its limitation. I employ a fixed-effects model with variableintercept based on four listed coking coal mining companies from2003to2009, and the finding isconsistent with Galbraith’s theory, but the iron and steel industry’s upward concentration adverseeffect on the coking coal companies’ performance was very weak.Fifth, the research calculates the geographical concentration indices of Chinese coal industry for thefirst time. The CR10and Herfindahl index from1993to2008with U.S. comparison show that theoverall degree of industrial concentration has risen up, but the level of China is lower than that of U.S.,and is affected more easily by the economic cyclical fluctuation. I also measure the spatial Ginicoefficient, E-G and M-S index, and find that the coal industrial concentration is high, but the sector islocalised at locations with little industrialization. I employ a fixed-effects model with variableintercept to account for the coal industrial geographic concentration, and find that resourceendowments, scale economies and market demand are positively related to the concentration, and theinfluence extent decreased, but a negative correlation between industrial policy and the concentration,and the extent is far beyond the reserves that is regarded as the most important factor generally. Then Idiscuses the feasibility of those provinces with rich coal will export more power instead of coal toensure energy supply.更多还原