The Research On The Early Warning Of The United States' Anti-dumping Toward China's Agricultural Products

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.