【作者】 钱宏菲；

【导师】 赵银德；

【作者基本信息】 江苏大学 ， 国际贸易学， 2010， 硕士

【摘要】 中美农产品贸易一直是中美经贸关系的重要组成部分。随着中美农产品贸易往来的不断增多,双方在农产品贸易中的争端也层出不穷,美国对我国出口的农产品提起反倾销调查是中美农产品贸易争端中最重要的一种形式。鉴于农业对于我国的重要战略地位以及国内在出口反倾销预警方面研究的不足,本文旨在建立美国对华农产品反倾销预警模型,以此来丰富和完善该领域的研究。同时希望能够为政府相关部门完善倾销与反倾销立法,建立健全外贸管理秩序提供一些参考性意见,尽量减少美国对华农产品反倾销调查的数量和由此带来的损失,促进中美农产品贸易的健康发展。倾销与反倾销受经济、政治、文化、双边贸易关系等多重因素的影响。本文为了简化分析、突出重点,笔者假设经济原因是引起美国提起反倾销调查的主要原因。本文在阐述反倾销预警理论和研究意义的基础上,结合国内外研究现状,进行预警模型的选择,最终选定使用二元逻辑斯蒂回归这一方法来构建反倾销预警模型。首先按照指标选取的原则,选出15个有代表性且经济意义明确且能够反映美国对华农产品提起反倾销动机的指标。在此基础上分两步对这15个指标进行择优筛选,构建预警指标体系。第一步,进行因子分析简化指标,使余下的指标成为能反映美国对华农产品反倾销预警的最小完备集。第二步,进行时差相关分析,遴选出有预警作用的指标,保留提前指标和同步指标,删除滞后指标。接着,利用预警指标体系进行二元逻辑斯蒂回归,构建出美国对华农产品反倾销的预警模型。代入原始数据,根据最终计算出来的概率值判断美国是否会对我国出口的农产品提起反倾销,发出预警警报。最后,为了验证模型的有效性,笔者利用历史数据对预警模型进行回判测试,计算出二元逻辑斯蒂模型的总体预警准确率和误判率。结果表明本文所建立的模型能够起到美国对华农产品反倾销的预警作用。基于上述分析结果和预警信号,笔者有针对性地提出了防范美国对华农产品反倾销的对策。更多还原

【Abstract】 Sino-US agricultural trade has been an important part of bilateral economic and trade relations. With the growing Sino-US agricultural trade volume, the disputes about agricultural product trade is also increasing and the United States’antidumping toward china’s agricultural products is the most important form in the Sino-US agricultural trade disputes. In view of the important strategic position of agriculture and anti-dumping warning of inadequate research, this paper aims to establish U.S. anti-dumping early warning model toward China’s agricultural products in order to enrich and improve the research in the field. At the same time hope to help relevant government departments to improve the dumping and anti-dumping legislation, establish and improve the foreign trade management in order to provide some reference observations, minimize the number of anti-dumping investigations and the resulting losses, and promote the healthy development of Sino-US agricultural trade.Economic, political, cultural, bilateral trade relations and other factors will affect a country’s dumping and anti-dumping actions. In this paper, in order to simplify the analysis and give prominence to key areas, I assume that the economic reasons are the major cause of the U.S. antidumping investigations. This paper describes the early warning of anti-dumping theory and research implications, based on the situation of research at home and abroad, I compared the early warning models, and ultimately I selected using the method of binary logistic regression to build the anti-dumping early warning model. First, in accordance with the principles of index selection, I selected 15 indicators of representative and clear and reflect the economic significance of U.S. antidumping motivation. On this basis, into a two-step selection of these 15 indicators, aims to build early warning indicator system. The first step, factor analysis simplified indicators, so that the remaining indicators become the best to reflect the minimum anti-dumping complete set of early warning. The second step, the time difference correlation analysis, I selected the early warning indicators, preserved early indicators and coincident indicators in advance, removed the lagging indicators. Then, I used the early warning indicators and their data into the binary logistic regression to build the model of The United State’s antidumping toward China’s agricultural products. Substituted into the original data, calculated according to the probability of the final value will determine whether the United States will filed anti-dumping investigations toward China’s agricultural products, show the early warning signals. Finally, to verify the validity of the model, I used historical data back to the early warning model for sub-test, binary logistic model to calculate the overall accuracy rate and false alert rate. The results have shown that the model can play a U.S. early-warning role in anti-dumping of China’s agricultural products. Based on the above analysis, I have put forward some measures against U.S. anti-dumping measures toward China’s agricultural products.更多还原