There is an old saying: “You can lay all the economists end to end and they will still not reach a conclusion.” While there may be some truth to this, there is widespread agreement among most economists (with Peter Navarro perhaps being an exception) that unrestricted trade is beneficial and protectionism is bad for any country. Trade wars can never be won and they are detrimental to an economy, as the US experience with the Smoot-Hawley tariffs during the Great Depression clearly demonstrated.

To point out the folly of a trade war is not to say that some countries aren’t pursuing unfair practices. We know that they are and action is appropriate. China may be dumping steel on the market at below cost, but that can’t go on forever, and China is not the major supplier of steel to the US. In 2017 China accounted for only 2% of US steel imports. Furthermore, no entity can overcome a negative spread on a product by increasing sales volume. But the proper response to abuses and attempts to corner a market is coordinated pressure and political pressure by allies together with us of the WTO dispute process, not tariffs. Pursuing remedies is a long and torturous endeavor, but so is dealing with retaliatory responses when a country like the US suddenly makes a unilateral move to impose tariffs or engage in other protectionist acts, especially those that harm our allies.

Such actions don’t often – if ever – lead to wins, let alone easy wins, as the President will soon find out. In the past the US has occasionally engaged in protectionism by imposing unilateral tariffs. Interestingly, one of the most recent episodes was carried out in 2002 by the Bush administration. While Republicans are most often supporters of free trade, President Bush imposed a complex set of temporary three-year tariffs on steel imports, mainly from the EU. Given this comparatively recent episode and the fact that it involved steel, one of the two products targeted by the tariffs being rolled out by the Trump administration, the 2002 experience may provide some clues as to how President Trump’s tariffs will fare.

Before we turn to the history of the 2002 effort to impose steel tariffs, it is important to put forth some basic statistics on US steel and aluminum imports. The US is the world’s largest importer of steel, yet it actually accounts for only 8% of total world steel imports, and it imports steel from over 110 countries.[1] The top ten sources of US imports account for 78% of the total, with the largest four being Canada (16%), Brazil (13%), South Korea (10%), and Mexico (9%). Note that 25% of our total imports come from our closet North American neighbors. More importantly, the steel imports by the US from Canada and Mexico account for 89% and 68% respectively of their total steel exports. Thus, tariffs imposed by the US on imports from these two countries that our closest neighbors would be devastating to their steel export business. That belated realization is clearly behind the fact that President Trump backtracked on his initial tariff announcement and exempted these two countries from his tariffs. He has also hinted that some others, like Australia (less than 1% of US imports), may also be exempted, as they were from the Bush tariffs in 2002.

How significant are steel imports relative to US production? In the aggregate imports account for about 1/3 of US steel consumption, with domestic production accounting for the rest. The US steel industry has been in a long, steady decline, lagging behind because of its failure to invest in new technologies to produce steel, the impact of union activity, and competition from low-cost producers abroad. Even if successful, the US steel industry does not have the potential to create hundreds of thousands of new jobs, as is argued. Just as an illustration, a new plant in Austria now produces with 14 employees what it would have taken 1000 to produce with the technology employed in the 1960s.[2]

Aluminum import realities raise similar concerns. Canada supplies approximately 59% of US aluminum imports, which are even more important to the US than Canadian steel imports. Other key suppliers are Russia (6%), United Arab Emirates (6%), China (5%), and other countries (24%). In total, the US imports about 90% of its primary aluminum. The imposition of tariffs on Canadian aluminum would devastate that industry and raise costs in the US for no good reason. Finally, aluminum is made from bauxite, but the US has few meaningful supplies of bauxite, mainly in Arkansas, Alabama, Georgia and Virginia, which make the Trump administrations national security rationale for its imposition of tariffs on aluminum suspect.

While the Constitution grants Congress the authority to regulate international commerce, Congress has, through a series of actions beginning in 1930, delegated to the President the authority to impose tariffs.[3] Section 232 of the Trade Expansion Act of 1962 authorizes the Secretary of Commerce to conduct an investigation of the impacts of imports and possible threats to national security; and if the President agrees, the Secretary can then take actions like raising tariffs. Given the dependency of the US on Canada for both steel and aluminum and Mexico for steel, which are now exempt, the imposition of tariffs under the emergency provisions of applicable US law would likely be accompanied by abrupt responses by many other affected suppliers, similar to those mounted by the EU in response to the 2002 Bush tariffs.

Exactly what did happen in 2002, and what were the consequences that might reveal the likely outcome today if tariffs were unilaterally increased? The Bush administration imposed tariffs ranging from 8% to 30% on a variety of steel products, in part to make good on a campaign promise designed to appeal to voters in the Rust Belt states of Ohio and Pennsylvania.[4]  However, unlike relying on the Section 232 national security threat authorization employed by the Trump administration, Bush relied upon Section 202 of the Trade Act of 1974, as amended in 1994, authorizing the President to investigate and impose temporary tariffs upon imports that might adversely affect domestic industries.[5] This action was not taken in a vacuum: It followed a series of earlier US complaints to the WTO against the EU and UK and ensuing counterclaims aimed at conserving duties on steel.[6]

The WTO Agreement on Safeguards governs the circumstances and requires studies and related supporting documentation for a country to impose tariffs. The US supplied the relevant documentation, which also contained recommendations for remedial tariffs. Then, on March 5, 2002, President Bush imposed a series of tariffs, which were to remain in place for three years, on a range of steel products. Interestingly, these tariffs also included exceptions for members of NAFTA, namely Mexico and Canada, as well as for Israel, Jordan, and South Korea, along with relaxed levies against steel from Brazil and Russia. Note that these exceptions were essentially for the major suppliers of steel to the US other than the EU.

While tariff changes were being contemplated in the US, the potential affected countries, mainly in the EU, indicated their intention to retaliate, just as countries now are already lining up to oppose the Trump tariffs. After the Bush steel tariffs were imposed, the foreign response was quick, as some seven countries filed a class action suit against the US with the WTO. Read (2005) provides a detailed chronology of the sanctions and tariffs imposed by the EU and the decision by the WTO that rejected the basis and justification for the US tariffs. On December 10, 2003, faced with the adverse decision and the growing threat of a trade war, President Bush rescinded the tariffs before the EU implemented its full list of sanctions. The tariffs lasted a much shorter period than was envisioned because of the impending trade war, the flimsy rationale employed by the US to employ them, and risks to the US that were involved.

What were the economic consequences of the tariffs while they were in place?[7] There were both direct and indirect impacts on both the US economy and our trading partners. Most impacted were those companies that had an inelastic demand for steel inputs. In the US these tended to be among the nearly 200,000 small firms with less than 500 employees.[8] Researchers have studied this episode, and here is a brief summary of those findings.

Preliminary studies estimated that the tariffs would increase steel prices by about 9% and save about 8900 steel jobs, reflecting a cost of about $450,000 per job. A second ex ante study estimated that the tariffs would cost the country about $2 billion but save only 3500 jobs at an estimated cost of $584,000 per job. An ex-post study of the event suggested that the main benefits were a forced restructuring of the steel industry, mergers, abandonment of obsolete plants and equipment, and new investment. At the same time, steel prices went up 39%. One negative consequence of the disappearance of several producers was the shifting of some $8 billion in pension liabilities to the Pension Benefit Guaranty Corporation. In terms of the indirect effects, a study indicated that 50,000 jobs were lost in the fabricated metals industry segment, and overall some 197,000 jobs were lost, which is more jobs than were involved in the making of steel. The impact on aggregate welfare was estimated to be relatively small, in the range of $65 to $111 billion. The impact as a percentage of GDP was infinitesimal.

Clearly, the 2002 experience offers lessons for the present administration. History may be about to repeat itself, for two reasons. First, the choice of the Section 232 national security rationale (and reliance on studies justifying its use) is even more suspect than the rationale used by the Bush administration to justify its tariffs. Second, given the speed at which the affected parties, especially the EU, are already preparing retaliatory tariffs, a case is sure to be brought swiftly before the WTO. These facts suggest that those who oppose the imposition of tariffs are correct. Third, tariffs on steel are not a way to create jobs, on net.

What should we demand as the proponents of the tariffs, and especially the President, as they proceed with their protectionist policies?  They must be pressed to quantify the gains they are attempting to achieve, both in terms of employment and impacts on prices, welfare, and GDP. Unless these ex-ante objectives are clearly articulated, we will be left with false claims of an undocumented victory, when this misguided set of policy actions unwinds.

[1] See Global Steel Trade Monitor, International Trade Administration, December 2017. https://www.trade.gov/steel/countries/pdfs/imports-us.pdf. We are overgeneralizing by using the term steel. Steel actually comprises many sub-products, such as pipe and tube, flat, long stainless, semi-finished, etc.
[2] https://www.bloomberg.com/news/articles/2017-06-21/how-just-14-people-make-500-000-tons-of-steel-a-year-in-austria
[3] For a detailed discussion see Congressional Research Service, “Presidential Authority of Trade: Imposing Tariffs and Duties,” Caitlain Devereaux Lewis, December 9, 2016.
[4] The rest of this commentary relies upon the very thorough and comprehensive study of the entire 2002 tariff episode by Robert Read, “The EU-US WTO Steel Dispute: the Political Economy of Protection and the Efficacy of the WTO Dispute Settlement Understanding,” Chapter 7, in The WTO and the Regulation of International Trade:Recent Trade Disputes Between the EU & US, edited with Nick Perdikis, Edgal Elgar, 2005.
[5] See Kevin Ho, “Trading Rights and Wrongs: The 2002 Bush Steel Tariffs,” Berkeley Journal of International Law,Vol. 21, Issue 3, 2003 fn 2.
[6] See Read fn 4 for a detailed discussion of the WTO Agreement on Safeguards and related international agreements governing the imposition of tariffs.
[7] For more detail see Read (2005) and references therein, fn 4 above.
[8] See Read (2005)