【作者】 覃琼霞；

【导师】 陆文聪；

【作者基本信息】 浙江大学 ， 农业经济管理， 2013， 博士

【摘要】 世界贸易组织(WTO)的成立本应成为规范国际贸易纠纷与摩擦的有效制度安排,但是事实却总不尽如人意。在21世纪的前十年中伴随着关税壁垒的逐渐消退,以技术性贸易壁垒(TBT)为代表的非关税壁垒正成为诸多产品国际贸易争端的导火线。由于农产品贸易本身涉及农药残留(MRs)、重金属残留等问题而更易遭受TBT的冲击,而对TBT贸易效应的度量研究是合理解决相关贸易争端的先决条件。当前有关TBT贸易效应的度量研究逐步突破了原先的简单粗糙分析框架,不断引入国际贸易新理论和计量经济学的新方法。在TBT贸易效应度量研究中,主要涉及加总数据模型和细分数据模型的运用,然而加总数据模型与细分数据模型得到的结论大多情况下却存在明显的差异性,特别是细分数据所代表的个体异质性特征在数据加总过程中往往被掩盖掉了,导致加总估计偏误。这类估计偏误对TBT贸易效应度量会带来严重误判,进而可能干扰和误导对外贸易政策的准确实施。本文从加总偏误的角度,从细分与加总数据模型入手,运用加总偏误理论来研究农产品TBT贸易效应的测度偏误。研究涵盖国与国,国家与地区以及地区之间的贸易层次,特别选取了中国与日本、中国与欧盟(EU)以及欧盟与北美自由贸易区(NAFTA)之间的茶叶、花生和大豆等具有代表性的农产品贸易为研究对象,通过不同层次的贸易加总数据来研究农产品技术性贸易壁垒的贸易效应度量偏误问题。对偏误的产生原因、偏误的测度与检验进行深入的研究。本文首先在回顾TBT贸易效应度量方法和加总偏误理论研究基础上,将原有技术方法纳入加总偏误统一分析框架,针对加总偏误问题提出了三类加总形式、三种偏误类型和两类偏误原因与三种检验方法。深入研究发现细分数据模型的内生性和加总过程的内生性是产生加总偏误的根本原因。接着为充分证明上述研究结论特别进行了详细的数值模拟,分析不同情形下的加总偏误特征,考察内生性与加总偏误之间的内在规律。最后运用加总偏误理论对现实代表性农产品的TBT贸易效应测度进行了实证分析,进而为我国农产品精准差异化应对外国技术性贸易壁垒提供了有效政策建议。通过上述研究本文主要得到了以下结论：(1)农产品贸易中的TBT贸易效应在细分数据层面以及细分与加总数据层面上呈现差异性特征。这种差异性来自个体异质性特征。(2)农产品TBT贸易效应度量中的加总偏误是由细分数据模型的内生性和加总过程的内生性因素造成的。(3)通过对不同加总形式下不同加总路径的偏误测度与检验,进一步对代表性农产品TBT贸易效应度量中加总偏误产生的原因和节点进行了准确定位。本文揭示了加总偏误的根本原因,提出了判断和剖析加总模型参数估计偏误的新方法和分析框架,为提高经济数据分析准确性提供技术支持。另外,数值模拟方法通过系统展现内生性与加总偏误之间的关系为加总偏误研究提供了一个理想而有效的实验手段。通过本文的研究还搭建了一个连接细分与加总模型参数估计的理论桥梁,有助于进一步丰富和巩固TBT贸易效应研究的理论基础,同时也拓宽了技术性贸易壁垒研究的视角。更多还原

【Abstract】 The World Trade Organization (WTO) established should become the effective institutional arrangement regulating international trade disputes and frictions, but the fact is not just as wishes. In the first decade of the21st century with the tariff barriers gradually faded, technical barriers to trade (TBT) as the representative of the non-tariff barrier is becoming the disputes fuse of many products in international trade Since agricultural trade itself involves pesticide residues (MRs), heavy metals and other issues are more vulnerable to the impact of TBT. And research on the measurement of TBT trade effects is the a prerequisite of reasonable to solve trade disputes. Current studies on the trade effect of TBT gradually broke through the original simple rough analysis framework against the new field of international trade and econometrics. The measurement of the trade effect of TBT mainly involved the use of aggregated data model and subdivision data model. However, there are obvious differences between them in most cases. Especially the individual heterogeneity represented in disaggregated data often be masked out in the data aggregation process, resulting in a aggregation estimation bias.Such estimation bias will bring serious misjudgment on the effect of TBT which may distract and mislead the exact implementation of foreign trade policy.Therefore this paper will study the effect of TBT in agricultural trade by the aggregation and disaggregated data model from the perspective of aggregation bias and fully research on the causes of bias, bias of measurement and inspection. Firstly, this paper begin by reviewing the original technical approach of the TBT trade effects measurement and aggregation bias theory, and then deduce an unified analysis framework. For the aggregation bias theory, we put forward three kinds of aggregation form, three types of bias, two causes of aggregation bias and three kinds of test methods. Further study found that endogeneity of the disaggregated data model and aggregation process are the root causes of the aggregation bias. Then to fully prove the above research conclusion and find the aggregation bias characteristics in the different situations, we specially carried on the detailed numerical simulation, and study the inherent relationship between endogeneity and aggregation bias. Finally, we use the theory of aggregation bias to carried out the empirical analysis on the real effect of TBT in agricultural trade.and then provide an effective policy recommendations on China’s agricultural trade how to respond accurately and differently to foreign technical trade barriers.Through the study in this paper, we get the following conclusion.(1)The effect of TBT in agricultural products trade showed in the aggregated and disaggregated data levels are different This kind of difference result from the individual heterogeneity.(2)The aggregation bias of the measurement on the agricultural trade effects in TBT is determined by the endogeneity of the disaggregated data models themselves and the endogenous factors in the process of aggregation.(3)Based on the aggregation bias measurement and test in different aggregation form and different aggregation path,we have found the really true causes and given an accurate positioning in the process of the measurement on TBT trade effects of the representative agricultural products trade.This dissertation revealed the root cause of the aggregation bias, and provides new methods and analytical framework to analyze the aggregation model parameter estimation. And also it could be a technical support for improving economic data analysis accuracy. In addition, the numerical simulation method provides an ideal and effective means of experiment for aggregation bias research through systematically to show the relationship between the endogeneity and aggregation bias. Through this study also provides a theory bridges connection with aggregation and disaggregation model parameter estimation, further enriched and strengthened the theoretical basis of TBT trade effect study, but also to broaden the perspective of the research of technical barriers to trade.更多还原