On April 9, Peter Navarro, director of the White House National Trade Council, published an article in the Financial Times claiming that the trade deficit with China has cost the United States millions of manufacturing jobs during the Bush and Obama administrations since 2001. This conclusion has been widely circulated and quoted by American officials. The basis of this criticism is research reports written by some US think tanks, especially one published by Robert E. Scott, a researcher at the Economic Policy Institute. However, Scott's research method is obviously wrong, and hence the research finding fails to withstand close scrutiny.

The method he uses to assess the impact of the trade deficit on employment is the input-output model. Specifically, he calculates the coefficient that reflects the correlation between the output and employment indicators in different industries, according to the input-output tables and related data prepared by the US Department of Commerce and the Bureau of Labor Statistics. After that, he links the trade and employment indicators and analyzes the impact of changes in trade on employment through the input-output table. But Scott's research method has three material flaws.

First, Scott's research method only considers the impact of the trade balance on total demand from the perspective of national economic accounting, and ignores the role played by trade in improving the efficiency of resource allocation, especially the role of imports in improving supply and increasing employment.

No country in the world, including the US, its largest economy, has all the resources it needs, nor can it produce all the products and services it requires. For example, the US imports large amounts of resource-related products and other intermediate inputs from other countries every year. Without these products, the US would not be able to fully utilize the productive capacity of many industries, total output would be significantly lower, and demand for labor would inevitably be reduced, leading to a rise in unemployment. Trade enables one country to fully realize its comparative advantages and compensate for its resource shortages, so that every trading partner can benefit. Unfortunately, Scott only sees that imports have displaced employment in certain industries of the US, which only tells us part of the truth.

Second, Scott's research method assumes that trade flow has no impact on employment in such sectors as wholesale trade, retail, and advertising. He argues that advertising campaigns and sales are always required for products produced in the US or imported from other countries. Obviously, he ignores the large number of people employed during the follow-up distribution of imported goods. In particular, employment in the wholesale and retail sector accounts for about an eighth of total employment in the private sector and employment in retail alone exceeds the total number employed in manufacturing. Scott assumes trade has no impact at all on employment in these industries, despite their close relationship with both trade and employment. He did so either for political reasons or because he intentionally chose to ignore such an impact. Taking Walmart as an example, the US retail giant purchases large quantities of goods from different developing countries, including China, for re-sale in the US and consequently has to employ hundreds of thousands of people in the US. Scott only sees the negative impact of Walmart's imports of Chinese goods on employment in the US manufacturing sector, and turns a blind eye to the employment created by Walmart.

Last but not least, Scott's approach allocates the so-called job losses to different regions, but fails to consider the different industrial structures of these regions. In his research report, Scott breaks down the impact of the estimated trade flow on employment in regions down to the level of every congressional district. Generally speaking, the US industrial system is relatively developed and complete, but there are clear differences between the industrial structures of various regions. Some industries do not exist or have very low production capacity in certain regions. If the industrial system in one region is far from that of the US as a whole, then industrial shortcomings can easily emerge. If the imported products are not produced at all in certain regions, importing such products will not have any impact on local employment. Therefore, the method of breaking down the so-called job losses into different regions and congressional districts is absolutely unreasonable.

The three major flaws in Scott's analytical method mean he is bound to reach the wrong conclusion.

According to a paper published by Scott in early 2017, 3.4 million jobs in the US were lost or displaced by China due to the trade deficit from 2001 to 2015. This is the conclusion frequently used by Navarro and other people to criticize China. Nevertheless, such a conclusion is invalid because of the major flaws in Scott's analytical method and is also inconsistent with US' economic performance.

On the one hand, there is no strong correlation between the unemployment rate and the US trade deficit, and in many cases the reality is the opposite. According to Scott's research method and conclusion, when the US trade deficit is high, the unemployment rate should be relatively high as well because many jobs are lost or displaced by other countries, and vice versa. Therefore, the unemployment rate should be going in the same direction as the trade deficit. In reality, however, the unemployment rate in the US is not closely correlated to the trade deficit, and sometimes they go in opposite directions. For instance, the US had a current account deficit of $805.9 billion in 2006, accounting for 5.8 percent of its GDP, the largest ratio in US history, but the unemployment rate was only 4.6 percent, significantly lower than the long-term average of 5.6 percent. From 2003 to 2006, the ratio of the current account deficit to GDP increased year by year but the unemployment rate decreased. And when the ratio of the current account deficit to the GDP decreased every year from 2007 to 2009, the unemployment rate increased.

On the other hand, we will reach an unbelievable conclusion if we extend Scott's research method. By following the research method and parameters used by Scott to assess the impact of the US trade deficit on employment, we estimated the impact of the total trade deficit on employment from 2001 to 2015. We found that the jobs lost or displaced because of the trade deficit exceeded the total number of unemployed in the US in some years. In other words, if the US had achieved trade balance during this period, it would not only have eliminated unemployment but would have discovered its labor force was insufficient. If that is the case, the US can realize full employment when it achieves trade balance. This conclusion, however, is clearly unbelievable because unemployment always exists, irrespective of whether the US has balanced trade or a trade surplus.

As can be seen from the analysis above, Scott's analytical method is obviously flawed and the resulting conclusion is biased and is misleading the US Congress, the government and the general public. This is also why Scott's previous reports were criticized by some institutions such as the Cato Institute in the US. Regretfully, the US government ignores the flaws in his research method and those of other researchers, and has even formulated trade policies and initiated protectionist trade measures against China according to such papers. This will have a negative impact on the development of US-China trade and an adverse affect on US economic development.

The author is deputy chief economist for the China Center for International Economic Exchanges. The views expressed do not necessarily reflect those of China Watch.

All rights reserved. Copying or sharing of any content for other than personal use is prohibited without prior written permission.