The end of 2016 and the beginning of 2017 saw a dramatic change in the global trade picture. Following on the heels of the British referendum supporting the exit of the United Kingdom from the European Union, the U.S. presidential election resulted in the victory of Donald Trump, who made many protectionist statements during the campaign, in particular threatening China, Mexico, and Germany with import duties. At various points, he said that he would impose tariffs of 35 percent on some Mexican imports and 45 percent on Chinese imports as a way to protect American jobs from unfair foreign competition.
There exists a large economic literature on trade wars. The famous Smoot-Hawley Tariff Act has in particular received extensive attention from economists. In June 1930, eight months after the Wall Street crash of October 1929, the U.S. Congress raised tariffs on 20,000 imported goods. The average tariff on protected imports increased from 39 percent to 53 percent. Many U.S. trading partners retaliated in the following months. When global trade collapsed after 1930, the U.S. share of world trade fell from 16 percent to 11 percent between 1930 and 1935. Irwin (1998) concludes that Smoot-Hawley led to an efficiency loss in the U.S. Gross National Product of between 0.3 percent and 1.9 percent.
Would a new trade war launched by the U.S. government make America lose again? And what could be the economic and trade consequences for U.S. trading partners?
In a new IFPRI discussion paper, we provide an evaluation of potential trade wars launched by the U.S., with the help of a world model of international trade that looks at 18 scenarios of trade wars. In these trade wars, we vary the trading partners involved (either China, or Mexico, or both China and Mexico), and also the form that U.S. protection and trade retaliations could take (the U.S. implements either a 35 percentage point increase on import duties on all imported goods except energy coming from these partners or an approximately 10 percentage point increase on these import duties). We then study different types of trade retaliation implemented by the U.S. trading partner(s) on imports from the U.S.; these result in import duties also varying from about 10 percent to 35 percent.
Our first policy conclusion is that there is no scenario under which the U.S. can benefit from either unilateral protectionism applied on these two trading partners or a trade war with these partners. In all scenarios, the impact on U.S. welfare or U.S. GDP is either zero or negative (welfare is defined as the monetary amount that a representative consumer would accept in place of the implementation of a specific protectionist policy). It is true that some U.S. sectors may benefit in these scenarios in terms of value added (textiles, wearing and leather products, electronic equipment, etc.), but this gain would be to the detriment of other sectors (chemical, rubber and plastic products, crops, meat and dairy products, motor vehicles and parts, and transport equipment), as well as to the detriment of workers, both skilled and unskilled, and in both agricultural and non-agricultural activities.
A second policy conclusion comes from the following finding: A 35-percentage point increase of U.S. protection against China and/or Mexico, followed by a similar retaliation, clearly constitutes an overreaction in terms of trade policy. Welfare and GDP losses are significantly larger when the US increase in protectionism and trade retaliation is large (35 percent) compared to moderate increases (10 percent).
A third policy conclusion concerns U.S. trading partners. While China is significantly affected by trade wars (in terms of welfare, China experiences losses between -0.2 percent and -1 percent), the welfare losses for Mexico are potentially huge (between -0.5 percent and -3.2 percent for welfare, and similar losses for GDP under the same scenario). Furthermore, value added may drastically decrease in some sectors. For example, the Mexican motor vehicles and parts sector and the transport equipment sector currently represent 3.3 percent of total Mexican value added. In the scenario of a relatively large augmentation of protectionism by the US and a retaliatory response by Mexico, value added in these sectors diminishes by about 22 percent!
A fourth policy conclusion concerns the emergence of free-riders—that is, countries or regions that benefit from a bilateral or trilateral trade war between the U.S. and its trading partners. This is the case of the Central America Free Trade Agreement (CAFTA) region, which obtains welfare gains ranging from 0.3 percent to 0.8 percent under the studied scenarios. This impact is of course associated with increased exports from CAFTA to the U.S. to replace either Mexican or Chinese goods.
Does this study exaggerate the potential magnitude of a trade war? It is true that the recently elected president of the U.S. has made fewer protectionist threats against China and Mexico since he took office compared to the statements made during his presidential campaign. It is also true that even if these threats are executed, we do not have much information about the extent to which trading partners will react (if they will react at all).
Beyond the good markets, China has potentially large retaliatory capacities either through capital and currency markets (for example, through the amount of U.S. dollar and U.S. Treasury bonds detained by the Chinese central bank) or through export restrictions (for example, using its high market share in global production of rare earths). These are clearly strong mechanisms with which the Chinese government can exert pressure on the U.S. government. However, the negative outcomes for unskilled workers in Mexico and the deterioration of the Mexican labor market following an increase in U.S. tariffs will also trigger additional incentives for legal and illegal migration, leading to additional challenges to U.S.-Mexico relations and additional costs for the U.S.
As a concluding remark, it is important to restate that trade wars are potentially harmful for the world economy. Protectionism is not the right way to reduce current account deficits, as these are mainly the consequence of insufficient net savings. It is also important to state the importance of the multilateral trade system which provides mediation for the litigation of trade disputes. If the U.S. government believes that some of its trading partners implement unfair trade practices at the expense of U.S. producers, the World Trade Organization (WTO) is the right arena in which to complain and obtain a removal or a change of these practices.