Why Wilbur Ross’s Approach on Trade Will Hurt US Competitiveness

September 25, 2017 1:45 PM
Photo Credit: 
REUTERS/Kevin Lamarque

In an op-ed in the Washington Post, Commerce Secretary Wilbur Ross raises concerns over fewer US parts being used in imports from Mexico and Canada than in the past. He argues that the North American Free Trade Agreement (NAFTA) needs stricter rules of origin and US specific content requirements to halt this decline. Such rules would limit NAFTA preferences to imports from Canada or Mexico with high levels of regional and US content.

His argument misses the main reason that the North American supply chain is valuable to the United States. The purpose of the supply chain is not to import from ourselves, but to lower production costs, making our firms more competitive globally. Supply chains allow countries to specialize in the stage of production in which they are most efficient, increasing productivity and lowering prices for consumers.

NAFTA already has the strictest rules of origin of any major trade agreement, so tougher rules or a US content requirement could reverse some of the existing productivity gains and hurt US manufacturing. Not only would they disrupt supply chains, damaging US industry, but they could also perversely lead to less US content in our imports because more goods will be imported without NAFTA preferences or from other countries.

Ross’s concerns stem from a new paper from the Commerce Department, relying on the Trade in Value Added (TiVA) database compiled by the Organization for Economic Cooperation and Development (OECD). The paper shows that US value added in manufacturing imports from Mexico and Canada has fallen from 1995 to 2011 (the most recent year in the database).[1]

Such statistics on value added are useful to help us understand how supply chains function and how they are changing over time, but they should not be used to make value judgements about trade. For example, according to the TiVa database, among US imports from Mexico, paper and pulp have the highest US content, 21 percent, and mining and quarrying have the lowest US content, 2 percent (see chart).[2] But that doesn’t mean paper imports are “better” than fuel imports. Producing fuel is costlier in the United States than in Mexico, so the United States saves money by importing fuel and using scarce resources for more efficient types of production instead.

The Commerce Department also does not show that tougher rules are needed to expand the US content share. For example, paper and pulp not only have the highest US content among US imports from Mexico, but US content also increased sharply from 16 percent in 1995 to 21 percent in 2011, bucking the downward trend among other manufactures. The increase had nothing to do with rules of origin. In fact, imports of paper and pulp face no tariff in the United States, so they do not even enter through NAFTA. The increase in US content happened under free trade with no rules of origin.

US value added share in US goods imports by exporting country, 2011

To the extent a decline in the US content of imports reflects a drop in the competitiveness of US production, we should be concerned, but content restrictions will, if anything, reduce US competitiveness further. Rules of origin or content restrictions add to the cost of production and trade because they force producers to use more expensive parts than they would otherwise, and there are administrative costs to documenting the origin of those parts.

Tougher rules could also magnify the decline in US content in some products, like cars, where existing rules are binding. If rules are very strict and tariffs are low, companies will be more likely to import products without using NAFTA preferences. Alternatively, companies with subsidiaries around the world might import products from other countries instead of from our NAFTA partners, where US firms are far less integrated into supply chains and the US content in imports is lower. In both cases, US manufacturing will be hurt and the US value added share of total imports will be reduced.

The way to expand North American manufacturing is by making it more competitive, and stricter rules will not achieve that important goal.


1. The database does not take into account processing zones in Mexico, which use US inputs more intensely than aggregate production, and hence has been shown to understate US content in imports from Mexico.

2. Manufacturers not elsewhere specified and recycling are excluded from the chart and have a 22 percent US content share.


Michael Twomey


Dr. Freund,

While I am in agreement with the position you state in the headline of this piece, I wish to make some comments.

First, I believe that conventional wisdom places US content in US auto imports from Canada and Mexico  at closer to a half, and not the data Ross cites, nor that which you mention in your blog “A US Content Requirement in NAFTA..”). Responding to Secretary Ross’s op-ed, Toronto’s Globe and Mail published on October 4, 2017 an op-ed by Flavio Volpe, the president of Canada’s Automotive Parts Manufacturers Association, who challenged the OECD data on US content in its imports from Canada, asserting that “[A] recent Scotiabank Economics report put that figure at almost 60 per cent,” – without, however, providing a specific citation, nor have I found it. On October 11, 2017, an article in Automotive News cited Eduardo Solís, President of the Mexican Association of the Automotive Industry (AMIA), that “[A] Mexican-made auto currently has about 40 percent of its parts and components from U.S. sources.” In an article in the Mexico City newspaper El Universal of October 10, 2017 by Sara Cantera “AMIA rechaza estudio de la OCDE,”  that same Mr. Solís is cited as affirming that the US content in Mexican exports to the US is between 37% and 39.5%.  The New York Times published “What Would Happen if the U.S. Withdrew from NAFTA,” by Ana Swanson and Kevin Granville, October 13, 2017, page B6 that asserted that US content in cars imported from Mexico is about 40% [this may simply refer to the Solís result]. The Center for Automotive Research (2017) in its “NAFTA Briefing: Trade benefits to the automotive industry and potential consequences of withdrawal from the agreement,” on page 8 cited a 2014 Wharton School report also asserting that the US content of cars imported from Mexico was 40%, up from 5% before NAFTA. The referenced Wharton report provides no information on the methodology used for this conclusion. Finally, INEGI, the Mexican statistical agency, published “Statistics in Occasion of… The Automotive Industry”, (2016) [translation into English of Estadísticas a propósito de… la Industria automotríz 2016], which on p. 16 states that the fraction of the final value of Mexican produced cars and trucks attributable to imported inputs was 44.6%, while that for automotive parts was 41.4%. They used an input-output table, but their methodology was quite distinct from that of the OECD.

Secondly, and still referring to the Ross – Flatness/Rasmussen Department of Commerce work, I should think that a more telling criticism of Ross is that effectively he only considers jobs created via our exports, that are re-imported as components of motor vehicles. But it would seem that much more is at stake in our exports of finished cars, or exports of parts that will be consumed elsewhere. The TiVA data has estimates of nominal values of these series, but of course converting that into jobs is a significant task.

Thirdly, in your iie blog “Why Wilbur Ross’s Approach Will Hurt …” the two links in your first end note both go to the same source: the Koopman et al. NBER paper #14626.  The point you raise that they supposedly refer to is important, that the OECD data does not consider trade through Mexico’s processing zones. If Koopman et al. refer to this limitation on OECD/Mexican trade data, I missed it. Or was the discussion in the other paper whose link got lost?

Finally, and getting back to your on-the-spot assertion that the administration’s proposed policies will ruin these auto value chains, I have a question as to whether you know of any empirical work on that line? While reading on this subject, I recently ran across a reference from a consulting group in Boulder called ImpactECON, “Reversing NAFTA: A Supply Chain Perspective.” Any other references would be welcome.  They talk about employment declines in the US and Canada of over 100,000, which sounds credible to me. Moreover, their toll for Mexico is over 900,000, which while both alarming and large, does reflect the large number of low productivity (and income) workers in auto parts.

Your response to these comments would be most welcome.

Michael Twomey

Prof. Emeritus, Economics

Univ. Michigan, Dearborn



Caroline Freund

Thanks for your comment. There are a number of different estimates out there for the US content. The TiVA data are only one estimate and I do not think they are the best. In fact, since I wrote this blog, a better estimate has come out which puts the number at close to 40 percent. The new paper (https://scholar.harvard.edu/files/alonsodegortari/files/degortari_dgvcs_october2017.pdf) uses firm-level data, allowing the author to control for the unique input mix of firms that export to the United States.

I agree that counting the US content of imports is the not the best way to evaluate the gains from NAFTA.  But, Ross was responding to comments from others about supply chains and US content.

A host of studies are being done on NAFTA supply chains now. In addition to the one you mention, some summaries of others are available at these links:

Mary Amiti and I offer some statistics on supply chain trade here: https://piie.com/blogs/trade-investment-policy-watch/why-renegotiating-nafta-could-disrupt-supply-chains

Brookings discusses state effects: https://www.brookings.edu/blog/the-avenue/2017/03/30/how-u-s-states-rely-on-the-nafta-supply-chain/

Emily Blanchard summarizes the existing NAFTA work in this context: http://voxeu.org/article/renegotiating-nafta-role-global-supply-chains

Add new comment