December 2-3, 2022, hosted by the Center for International Finance and Economics Research (CIFER) of the National Institute of Financial Research, Tsinghua University.Academic Symposium on "2022 International Trade Disputes and the Restructuring of Globalization"The event was successfully held.Doctoral student at the School of Statistics, Beijing Normal UniversityXia JunAttendees included a CIFER researcher and a lecturer at the School of Statistics, Beijing Normal University.Lu BingHostTrade Structure Sub-Forum 7Sharing with collaborators (as shown below) on the topic ofClarifying the Misleading Statistics in China-US Bilateral Trade Due to Additional Tariffs (Clarification of misleading bilateral trade statistics between China and the United States due to tariffs)The article.
Xia Jun, PhD student in Economic Statistics, Class of 2021, School of Statistics, Beijing Normal UniversityMy research interests include Sino-US trade friction and trade benefits. During my doctoral studies, I received the Doctoral Freshman Scholarship and the Third-Class Academic Scholarship from Beijing Normal University. I have been a major contributor to three academic papers, one of which is a CSSCI (Statistical Research) paper and two are currently under review for SSCI (Q2) publication.
Xia Jun
Since the start of the US-China trade war in 2018, the statistical discrepancies in bilateral trade between the two countries have exhibited characteristics different from historical trends. In 2020, the absolute value of the US trade deficit with China fell below the US trade surplus with China by $6.6 billion for the first time, and this difference further widened to $41.3 billion in 2021. Clarifying the misleading impact of tariffs on US-China bilateral trade statistics and avoiding statistical misjudgments of the tariff effects between the two countries is crucial. This paper employs a difference-in-differences model, based on the HS8-bit code, to empirically examine US imports from China that are net affected by Section 301 tariffs, Chinese exports to the US that are affected by the reduction in the net export VAT rate, and goods affected by both. The conclusions show that the imposition of Section 301 tariffs led to US importers underreporting imports, by approximately 10.4% of the pre-2017 US imports from China; the reduction in the net export VAT rate encouraged Chinese exporters to overreport exports, by approximately 2.3% of China's 2017 exports to the US; in contrast, the significant decrease in the US trade deficit with China under the Section 301 tariffs is the main contradiction causing the discrepancy in trade statistics between the two countries; the revised US trade deficit with China was revised upward by an average of 14.4% over the three years from 2019 to 2021, while China's trade surplus with the US was revised downward by an average of 3.4%.Regarding the impact of the US-China trade war on the trade imbalance between the two countries, existing literature shows that imposing tariffs does not help improve the current trade imbalance. Both countries arrive at conclusions favorable to themselves based on official data published by their own statistical departments. This paper starts with the mirror error phenomenon prevalent in international trade. Building upon an analysis of the reasons for the significant mirror error in bilateral trade before the mutual imposition of tariffs, it studies the impact of the tariffs on bilateral trade statistics, leading to a reversal of the mirror error. This paper first outlines typical examples of how tariffs have reversed the statistical differences in bilateral goods trade between the US and China, proposing two possible explanations: 1. Chinese exporters overreport exports due to export tax rebates; 2. High tariffs between the US and China create a tax avoidance motive for US importers to underreport imports. This paper empirically tests these two reasons, quantitatively assessing the impact of these two factors on the reversal of the statistical differences in Sino-US goods trade, clarifying the misleading effect of tariffs on bilateral trade statistics, and avoiding statistical misjudgments regarding the impact of tariffs on both countries.Difference-in-differences model:
The dependent variable is the logarithmic difference between imports from China to the United States as statistically reported by the United States and exports to the United States as statistically reported by China. The core explanatory variable is the interaction term between three different scenarios affected by the Section 301 investigation and the policy's effective date (2019).data:Data on US tariffs on Chinese imports of goods with HS8-digit codes in eastward trade comes from the USTR's 2018 report on goods totaling approximately $550 billion in the first to fourth phases of the Section 301 investigation. Data on Chinese exports to the US comes from the China Customs database. Export tax rebate information from 2018-2020 comes from the Ministry of Finance and the State Taxation Administration of China. This paper selects a list of goods whose export VAT rates have changed by more than 1% as the research object.1. Eastbound trade remains the primary source of discrepancies in US-China trade statistics. The underreporting of imports by US importers to avoid taxes and the falsification of exports by Chinese exporters to obtain export tax rebates have led to a reversal in the statistical discrepancies between bilateral goods trade between the US and China since the imposition of tariffs, deviating from historical trends. Among these discrepancies, the underreporting of imports from China by US importers due to the US Section 301 tariff policy is a major aspect of the statistical anomaly.2. Despite the mutual imposition of tariffs by China and the US, China's trade surplus with the US has continued to expand, according to Chinese statistics. This is partly due to data distortion caused by overstating, but also due to factors such as the pandemic and tax rebate policies affecting the trade balance between China and the US.
Zhu Kunfu
Associate Professor Zhu Kunfu of the School of Economics at Renmin University of China commented on the article. Professor Zhu first summarized the article, concluding that its topic was significant, its analysis logically rigorous, and that it was an excellent academic paper in economic statistics. Professor Zhu's main suggestions for improvement are as follows:1. The mirror error problem in trade statistics, especially bilateral trade statistics, has a long history. As early as Fung and Lau (1998), they began studying the mirror error in Sino-US bilateral trade, and Koopman et al. (2008) also explored this issue. At the same time, there is also a great deal of research on the fictitious trade component in trade statistics; for example, Professor Wei Shangjin has also conducted relevant research. This paper can appropriately expand upon the literature review.2. Mirror error in trade statistics is a common problem. Comparing the similarities and differences between China-US bilateral trade and other major bilateral trade (China-EU, China-Japan, Japan-US, and Europe-US) can enhance the significance of this paper's topic.3. The article's main conclusion is that after the US imposed high tariffs on Chinese goods, US importers underreported the value of their imports, which is a major reason for the reversal and widening of the mirror image error in Sino-US bilateral trade in recent years. The article further needs to analyze why, after the imposition of high tariffs on Chinese imports from the US, the phenomenon of underreporting import values did not occur.4. The article sums the US HS10-digit code data into HS8-digit code data and compares it with the Chinese Customs HS8-digit code data. This work is debatable. In fact, the HS6-digit code is internationally accepted. Both the US HS10-digit code and the Chinese HS8-digit code are extensions of the international HS6-digit code. It would be better to simply sum them all into the HS6-digit code before making a comparison.5. The descriptions of the three scenarios in the article processing group, especially Scenario 1, need further specification. It should actually be "trade goods only included in the US Section 301 investigation Phase 1-4 tariff list (US$550 billion) and the list of goods whose net VAT rate in China has decreased by more than 1%."
