20-8 Export Controls: America’s
Other National Security Threat
Chad P. Bown May
2020
ABSTRACT
While the public was transfixed by the Trump administration’s policies alleging that imports
were a threat to America’s national security during 2017–20, there was a concomitant and more
quiet US policy shift on the export side. Addressing the national security threat presented by
exports posed different economic and institutional challenges from those associated with import
policy, including the acknowledgment that export controls for legitimate national security
reasons can be the first-best policy to confront the problem at its source. Yet, export controls
could also be misused as a beggar-thy-neighbor policy to redistribute economic well-being
across countries, even from one ally to another. This paper describes how US export control
policy evolved over 2017–20, as well as the
international institutions—first the Coordinating Committee for Multilateral Export Controls
(COCOM), then the Wassenaar Arrangement—historically tasked with multilateralizing US
export restrictions used to protect national security. With the potential for US export control
policy to brush up more frequently against WTO rules designed to limit the use of export
restrictions, the paper also highlights new challenges for the WTO’s system of resolving trade
disputes. Overall, a US failure to strike the right balance for its export control policy would result
in it being ineffective at addressing national security risks, costly for the economy, and
problematic for trade and diplomatic relations.
JEL Code: F13
Keywords: national security, export controls, dual-use technologies, ECRA, Wassenaar
Arrangement, uncertainty
Chad P. Bown is
the Reginald Jones
Senior Fellow at the
Peterson Institute for
International
Economics.
1750 Massachusetts Avenue, NW | Washington, DC 20036-1903 USA | + |
Author’s Note: A revised version of this paper is forthcoming in the Duke Journal of
Comparative and International Law. Thanks to Martin Chorzempa, Soumaya Keynes, Nicholas Lardy,
Niall Meagher, and Marcus Noland for extremely useful discussions. All remaining errors are my
own.
WORKING PAPER
WP 20-8 | MAY 2020 2
INTRODUCTION
On January 6, 2020, the Trump administration announced new export controls on artificial
intelligence (AI) software. For the first time, an American company would need to apply for a
special license to sell satellite imagery software And the US government could deny the
application, nixing any revenue from export sales.
That was not the end of it. On February 16, the Wall Street Journal reported that the
administration was contemplating a ban on exports of jet engines
to China for use in civil This threatened to cut off some of General Electric’s jet
engine sales to one of the world’s largest and fastest-growing markets for commercial
aviation.
Then on February 17, the Wall Street Journal reported that the US administration was
considering a new rule to prohibit American companies from supplying equipment to foreign
manufacturers of semiconductors that wanted it to make chips to sell to This could have
curtailed hundreds of millions of dollars of US sales to customers like Taiwan Semiconductor
Manufacturing Company, one of the largest semiconductor foundries in the world.
As the rumors swirled, on February 18 President Donald Trump intervened, tweeting that
America would remain open for business, and that the “United States cannot, & will not, become
such a difficult place to deal with in terms of foreign countries buying our product, including for
the always used National Security excuse.” He added, “I want China to buy our jet engines, the best
in the World,” a sentiment perhaps designed to comfort worried executives at General
The Semiconductor Industry Association responded with temporary relief, applauding “President
Trump’s tweets supporting . companies being able to sell products to China and opposing
proposed regulations that would unduly curtail that ability.”5
Nevertheless, uncertainty persisted. The president’s tweets have not historically been a
guarantee of the path of policy, and in this instance were not enough to assuage those worried
that further restrictions in the name of America’s national security might be forthcoming. The
whiplash around these events did little to clarify whether this was the beginning of a larger
change
As the rumors
swirled, on
February 18
President
Donald Trump
intervened,
tweeting that
America would
remain open for
business...
1 The exception was that the software could be sold in Canada. See Wolf, Emme, and Monjay (2020)
as well as James Politi, “US proposes new export controls on satellite imagery soft- ware,“
Financial Times, January 3, 2020.
2 Ted Mann and Bob Davis, “Trump Administration Considers Halting GE Venture’s Engine Deliv- eries
to China,” Wall Street Journal, February 16, 2020.
3 Asa Fitch and Bob Davis, “. Weighs New Move to Limit China’s Access to Chip Technology,”
Wall Street Journal, February 17, 2020.
4 The series of four Trump Tweets of February 18, 2020, read, “The United States cannot, & will not,
become such a difficult place to deal with in terms of foreign countries buying our prod- uct, including
for the always used National Security excuse, that our companies will be forced to leave in order to
remain competitive. We want to sell product and goods to China and other countries. That’s what trade
is all about. We don’t want to make it impossible to do business with us. That will only mean that orders
will go to someplace else. As an example, I want China to buy our jet engines, the best in the World. I
have seen some of the regulations being circu- lated, including those being contemplated by Congress,
and they are ridiculous. I want to make it EASY to do business with the United States, not difficult.
Everyone in my Administration is being so instructed, with no excuses. THE UNITED STATES IS
OPEN FOR BUSINESS!”
5 Semiconductor Industry Association, “SIA Statement on President Trump’s Tweets Regarding Export
Controls,” February 18, 2020.
WP 20-8 | MAY 2020 3
in US policy toward exports, and one with implications for tens of billions of dollars’ worth of
expected trade. And it was clearly not over when, in late March, additional reports swirled that the
Trump administration was pushing ahead with export restrictions on semiconductor
manufacturing equipment after
Much of the call to action on export controls arose out of growing US government concern
with China. In 2018, with bipartisan support, Congress passed and President Trump signed into
law the Export Control Reform Act of 2018. But its scope and scale remained unclear well after
the statute went into effect. In part, this was because Congress left it to the Trump administration
to interpret and put into practice two key elements of the new law. First, what were the “emerging
and foundational technologies” to be restricted? And second, how to define what was “essential to
the national security of the United States”?
It is not for this paper to assess the national security implications of individual
technologies, nor to define what is in the national security interests of the United States. Rather,
the purpose is to highlight the risks and unintended consequences of controls that might be
poorly designed or badly implemented. Because intelligence-gathering and enforcement
resources are scarce, prioritization matters. If everything is about national security, nothing is
about national security.
There were reasons to be concerned. The Trump administration took an expansive view of
what was “essential to the national security of the United States” when it used that justification
to apply tariffs on imported steel and aluminum from America’s allies in 2018. The president
himself had politicized typically bureaucratic and legal export control decisions by using them
as bargaining chips in trade negotiations with China. Furthermore, he was not known for
careful consideration of economic expertise or of America’s commercial interests when it
came to trade.
And there were large economic interests at stake. Take a suddenly ubiquitous technology like AI;
over a short period, it had been adopted by banks for
fraud detection, retail outlets for online customer support, Netflix for movie recommendations,
and carmakers for autonomous vehicles. What would the new AI export controls announced in
January 2020 portend for its seemingly limitless commercial applications?
Export controls defined too broadly would make the United States a less attractive
place for companies to do their research, development, and production, which they might
shift elsewhere. Equally worrisome was the
continuing cloud of uncertainty about future US export control policy. Would American
companies hold back on R&D and key investment decisions until it was resolved? Squelching
innovation was costly—the competitiveness of the American economy would suffer, as would
American workers and communities losing out from that economic activity no longer taking
place locally.
The final risk was that President Trump’s preference for unilateralism would make
ineffective the US export controls actually needed to address the most worrisome national
security threats. Getting export controls to work at
safeguarding national security would require multilateralism—other governments
Much of the
call to action
on export
controls arose
out of
growing US
government
concern with
China.
6 Bob Davis and Katy Stech Ferek, “. Moving Forward with Rule to Limit Chips to Huawei,”
Wall Street Journal, March 26, 2020.
WP 20-8 | MAY 2020 4
need to agree to also hold back supplies of comparable technologies. But cooperation was being
challenged by rising levels of distrust that some countries would seek to restrict exports—alleging
a national security threat when there was none—to achieve political-economic gain in a beggar-
thy-neighbor form. Thus, the multilateral institutions tasked with facilitating and policing export
controls were confronted with new pressures that threatened to disrupt the delicate balance among
cooperative rule making, exceptions to protect national security, and effective dispute resolution
when inevitable frictions arose.
THE TRUMP ADMINISTRATION LINKS NATIONAL SECURITY,
ECONOMIC SECURITY, AND TRADE POLICY
Almost from inauguration day, the Trump administration tied together trade and national security
in ways not seen in US policy for decades. White House National Trade Council director Peter
Navarro set the stage in early March 2017 with a high-profile speech implying that the US trade
deficit was a threat to American national The next month, the administration began two
investigations, under Section 232 of the Trade Expansion Act of 1962, into whether imports of
steel and aluminum threatened national security. In August, it opened another inquest, under
Section 301 of the Trade Act of 1974, into whether China’s unfair trade practices worked to
“undermine American manufacturing, services, and innovation.”8 Finally, in December, the
administration released its National Security Strategy blueprint, defining its America First policy
with the mantra that “economic security is national security.”9
What had been mostly rhetoric in 2017 became concrete policy actions in 2018.
Beginning in March, the administration imposed 25 percent tariffs on steel and 10 percent
tariffs on aluminum, affecting nearly $50 billion of imports. Its Section 232 reports alleged
that such imports were a threat to national security, despite the fact that most such imports
were from Canada, NATO-allied countries in Western Europe, Japan, and South Korea. In
May, the administration turned to the Section 232 statute again, launching a new
investigation into whether $350 billion of imported automobiles and parts posed a threat to
American national security. This began a sustained period in which the president repeatedly
threatened the European Union and Japan with additional national security tariffs—threats
suddenly made credible with his restrictions on steel and
An important inflection point for US policy came after the Trump administration released
details of its unfair trade investigation of China. Made public was a long list of American
grievances laying the groundwork for
Almost from
inauguration
day, the Trump
administration
tied together
trade and
national
security in ways
not seen in US
policy for
decades.
7 See CSPAN, National Association for Business Economics Conference, Peter Navarro Remarks, March
6, 2017.
8 White House, Presidential Memorandum for the United States Trade Representative, August 14, 2017.
9 White House, National Security Strategy of the United States, December 2017.
10 Nearly a year later, in May 2019, as the Mexican Senate was meeting in a ceremony to consider passage
of the US-Mexico-Canada Agreement on trade just negotiated with the president,
Trump threatened to invoke the International Emergency Economic Powers Act and impose a 25
percent tariff on all goods imported from Mexico. His motivation was not trade related:
Trump felt Mexico was not doing enough to address the flow of migrants arriving from Central
America. In the end, he backed down without imposing tariffs.
WP 20-8 | MAY 2020 5
subsequent policy Some concerns involved explicit Chinese policies, such as
Beijing’s Made in China 2025 industrial policy, rolled out in 2015, which identified 10
priorities for sectoral advancement that the US administration felt posed a direct threat to
American technological Other complaints alleged covert Chinese policies, such as
state-sponsored cyberhacking, theft of industrial secrets, espionage for commercial (as
opposed to intelligence- gathering) purposes, as well as predatory foreign investment and
purchases
to acquire advanced American technology. The administration essentially alleged that Beijing
was pursuing a Chinese version of “economic security as national security.”
The Trump administration’s subsequent “trade war” with China focused on import-related
policies to Beginning in July 2018, the administration rolled out a series of tariff actions
that ultimately covered $360 billion, or nearly two thirds, of US imports from China by
September Most of those imports faced additional US tariffs of 25 percent and remained
in place despite the administration’s truce with China implemented in February
That the United States would ultimately deploy policies in addition to import tariffs was
made explicit in the March 2018 presidential memorandum kicking off the trade In a
section on potential restrictions on Chinese investment, for example, the president directed his
administration to consider “any available statutory authority, to address concerns about
investment in the United States directed or facilitated by China in industries or technologies
deemed important to the United States.” Thus, it was not surprising when the administration
turned to other US statutes, including those allowing the government to control what American
companies could export to China.
The tariff decisions had signaled a major shift in US policy toward imports. Yet, on export
control policy, some of the president’s subsequent actions still caught many by surprise.
That the
United States
would
ultimately
deploy policies
in addition to
import tariffs
was made
explicit in the
March 2018
presidential
memorandum
kicking off the
trade war.
11 USTR, Findings of the Investigation Into China’s Acts, Policies, and Practices Related to Tech- nology
Transfer, Intellectual Property, and Innovation under Section 301 of the Trade Act of 1974, March 22,
2018. See also USTR, Update Concerning China’s Acts, Policies and Practices Related to Technology
Transfer, Intellectual Property, and Innovation, November 20, 2018.
12 See Made in China 2025, Notice of the State Council, May 8, 2015. The 10 industrial policy pri- orities
concerned the next generation of information technology; robotics, artificial intelligence, and
automation; aerospace equipment; offshore engineering equipment and high-tech ships; advanced rail
transportation equipment; new energy vehicles; power systems; agriculture ma- chinery and equipment;
advanced materials; and biomedicine and high-performance medical devices.
13 In addition to the tariffs, the administration filed a WTO dispute over Chinese laws and regula- tions that
prevent foreign patent holders from enforcing their rights against a Chinese joint venture after a
technology transfer contract ends. See China – Certain Measures Concerning the Protection of
Intellectual Property Rights – Request for consultations by the United States, WTO legal document
WT/DS542/1, March 26, 2018.
14 See Chad P. Bown and Melina Kolb, Trump’s Trade War Timeline: An Up-to-Date Guide, PIIE Trade
and Investment Policy Watch, February 14, 2020 (last updated).
15 See Chad P. Bown, Phase One China Deal: Steep Tariffs Are the New Normal, PIIE Trade and
Investment Policy Watch, December 19, 2019.
16 White House, Presidential Memorandum on the Actions by the United States Related to the Section
301 Investigation, March 22, 2018.
WP 20-8 | MAY 2020 6
US EXPANSION OF EXPORT CONTROLS UNDER
EXISTING STATUTES DURING THE TRUMP-CHINA
TRADE WAR
US export controls were one important example of President Trump’s periodically injecting other
policies into the mix during the first two years of his trade war with China. The primary initial
targets of US export restrictions were Chinese telecommunications giants ZTE and Huawei. These
companies faced the threat of lost access to American-made semiconductors, software, and other
technologies on which they relied for production for sale in China and other foreign markets.
Preceding the trade war, in March 2017, ZTE had agreed to a settlement with the US
government for its failure to abide by US sanctions prohibiting the sale of certain technology
to Iran and North Korea. ZTE’s agreement to plead guilty, pay a fine, and change some of its
internal practices followed a multiyear investigation brought under the Obama
administration; while serious, the settlement was received as relatively routine and
That perception began to change when, on April 15, 2018, the US government
enacted a denial
order against ZTE that would have resulted in export controls for the company’s violation of the
terms set out in the March 2017 But in a surprising move, President Trump brought
the export controls levied on ZTE directly
into his trade war negotiations with President Xi Jinping. Trump overruled his Commerce
Department’s denial order and demanded the department negotiate a settlement to lift the export
controls and restore ZTE’s access to American-made goods and
Huawei faced a separate set of actions during the trade war. In May and August 2019, the US
government claimed that Huawei and a number of its affiliate companies were involved in
“alleged violations of the International Emergency Economic Powers Act (IEEPA), conspiracy to
violate IEEPA by providing prohibited financial services to Iran, and obstruction of justice in
connection with the investigation of those alleged violations of US sanctions.”20 Huawei was thus
added to the “entity list” of companies that were required to receive a license to obtain any
product subject to existing US export controls.
The controls—and threats of more—heightened awareness at ZTE and Huawei, and in the
Chinese government, of the companies’ vulnerabilities. Cutting off access to American exports of
semiconductors, software, and related technologies could have been devastating, putting tens of
thousands of Chinese out of work. Thus, one unintended consequence of the Trump
administration’s policy may have been for Beijing to pursue an even more aggressive approach to
industrial policy. The fear of being cut off could have created the incentive
US export
controls were
one important
example of
President
Trump’s
periodically
injecting other
policies into the
mix during the
first two years
of his trade war
with China.
17 See Aruna Viswanatha, Eva Dou, and Kate O’Keeffe, “ZTE to Pay $892 Million to ., Plead Guilty
in Iran Sanctions Probe,” Wall Street Journal, March 7, 2017, and Stephan Haggard, The ZTE Case,
PIIE North Korea: Witness to Transformation blog, March 15, 2017.
18 Paul Mozur and Ana Swanson, “Chinese Tech Company Blocked from Buying American Com-
ponents,” New York Times, April 16, 2018.
19 “President Xi of China, and I, are working together to give massive Chinese phone company, ZTE, a way
to get back into business, fast. Too many jobs in China lost. Commerce Department has been instructed
to get it done!” Tweet of Donald J. Trump, May 13, 2018.
20 See BIS, Department of Commerce Announces the Addition of Huawei Technologies Co. Ltd. to the
Entity List, May 15, 2019, and Department of Commerce Adds Dozens of New Huawei Affiliates to
the Entity List and Maintains Narrow Exemptions through the Temporary General License, August 19,
2019.
WP 20-8 | MAY 2020 7
to speed up diversification of its supplier base, perhaps with the Chinese government
increasing its already considerable state support for domestic
The US administration’s export control actions also revealed some immediate costs to
American businesses of being cut off from Chinese buyers. Google’s Android would lose out
if Huawei chose other operating systems for its Qualcomm, Acacia
Communications, and other American companies suffered hits to their stock prices when
markets discovered that ZTE could be forced to stop purchasing American-made
Nevertheless, the March 2020 revelation that the administration was still considering an
extension of US export controls to semiconductor manufacturing equipment was a major potential
Despite the president’s February 18 tweets, the Trump administration seemed
intent on going ahead with a modification to something called the “direct product rule,” that
would attempt to extend the reach of US controls to other countries’ suppliers seeking to make
sales to Huawei.
The policy had farther-reaching economic implications for American industry, as illustrated
with the example of a non-Chinese semiconductor company, Taiwan Semiconductor
Manufacturing Company (TSMC), that had considerable sales to The proposed US
export controls were reportedly designed to confront a company like TSMC with a choice: To
retain access to US equipment, TSMC would have to give up its sales to Huawei of chips made
using that equipment.
Alternatively, TSMC could keep its Huawei business, but it would need to switch from
American-made to semiconductor manufacturing equipment produced by firms in South Korea,
Japan, or elsewhere.
The proposed export controls were a bet that companies like TSMC would choose
continued access to American equipment over their future sales to Huawei. But what if they
didn’t? TSMC’s decision would also depend, of course, on the availability of substitute
equipment from non-American suppliers. And that availability would also hinge on whether the
US export control would be multilateralized, so that other exporting countries applied it too. In
this example, the South Korean and Japanese governments would need to impose controls on their
equipment suppliers’ sales to firms like TSMC.
A failure to get US-imposed controls multilateralized to other countries would potentially be
devastating for the US semiconductor industry. Its fears were described in a March 2020 study
commissioned by the Semiconductor Industry The study imagined a scenario in
which a US policy of the sort
The March
2020
revelation that
the
administration
was still
considering
an extension of
US export
controls to
semiconductor
manufacturing
equipment was
a major
potential
escalation.
21 See OECD (2019b).
22 Yang Jie and Dan Strumpf, “Who Needs Google’s Android? Huawei Trademarks Its Own Smart- phone
OS,” Wall Street Journal, May 25, 2019.
23 Jay Greene, “In ZTE Battle, . Suppliers Are Collateral Damage,” Wall Street Journal, April 24, 2018.
24 Karen Freifeld, David Shepardson, and Alexandra Alper, “. prepares crackdown on Huawei’s global
chip supply – sources,” Reuters, March 26, 2020. For additional discussion of the direct product rule,
see Whitten and Mays (2019).
25 Asa Fitch and Bob Davis, “. Chip Industry Fears Long-Term Damage from China Trade
Fight,” Wall Street Journal, March 9, 2020.
26 See Semiconductor Industry Association, Report Shows Risks of Excessive Restrictions on Trade
with China, March 9, 2020; Boston Consulting Group, How Restricting Trade with China Could End
US Semiconductor Leadership, March 9, 2020.
WP 20-8 | MAY 2020 8
reported in February 2020 was a bad bet; ., that unilateral US export controls would result in
many foreign companies choosing to engage with China instead of the United States for
equipment and input sourcing. The study estimated the cost to the US industry at tens of billions
of dollars of annual revenue—revenue that was the main source of R&D funding needed for
American companies’ next generation of chips. Because less R&D would make the next round of
American semiconductors less competitive globally, future customers would be even less likely
to choose US equipment and inputs. The study pointed to a vicious cycle of American industrial
decline, with the pain extending to American workers at these companies and their communities
reliant on the jobs and economic activity supported by the industry.
MAJOR CHANGES TO US LAW AND TO EXPORT CONTROL
REGULATIONS BEGAN IN 2018
In the midst of the trade war and the battle with ZTE in 2018, the US government undertook a
separate legislative process to overhaul its export control regime.
The debate between Congress and the Trump administration involved a number of proposals,
some even more far-reaching than the final
Nevertheless, the new statute had the potential to severely curtail US exports for products that
might have “dual use” (both military and commercial application). The Export Control Reform
Act of 2018 (ECRA) came into law on August 13, 2018, as part of the John S. McCain National
Defense Authorization Act. It was passed with bipartisan support of 87 votes in the Senate and
351 votes (including 131 Democrats) in the House of Representatives.
To many, the ECRA was long It codified existing US government practices into
law by replacing executive orders that had been issued annually under the IEEPA since the
statutory authority for the Export Administration Regulations (EARs) set up by the Export
Administration Act of 1979
lapsed in 2001.
But ECRA also had the potential to do much more. Under the law, Congress tasked the
Bureau of Industry and Security (BIS) in the Department of Commerce to update US export
controls on “emerging and foundational technologies” that were “essential to the national security
of the United States.” BIS would lead an ongoing interagency process to identify and add
products to the EARs. To fulfill its very challenging new mandate, BIS eventually requested
additional budgetary resources from
In the implementing regulations to guide the process, BIS came up with 14 new categories of
representative technologies for which it sought public input into whether to implement new
export controls and, if so, how far-reaching they
In the midst of
the trade war
and the battle
with ZTE in
2018, the US
government
undertook
a separate
legislative
process to
overhaul its
export control
regime.
27 For a discussion, see Kevin J. Wolf, Steven C. Emme, Thomas J. McCarthy, and Andrew R. Schlossberg,
“The Export Control Reform Act and Possible New Controls on Emerging and Foundational
Technologies,” Akin Gump International Trade Alert, September 12, 2018; and Mar- tin Chorzempa and
Gary C. Hufbauer, Trump Awaits Congress on Investment and Technology Controls, PIIE Trade and
Investment Policy Watch, July 9, 2018.
28 The Obama administration had attempted a reform of US export controls, but it was never enacted
into law. For a discussion, see Congressional Research Service (2020).
29 “Commerce requests 8 percent BIS funding hike to counter China’s tech rise,” Inside US Trade,
February 12, 2020.
WP 20-8 | MAY 2020 9
should New limits would be considered for goods and services such as AI, machine
learning, quantum computing, and 3D printing. In seeking to define the specific emerging and
foundational technologies to control, BIS established technical advisory committees “composed
of representatives from industry,
academia, and the . Government and reflect[ing] diverse points of view on the concerns of the
exporting community.”31
The American business community was nonetheless concerned that the BIS scoping
exercise would be indifferent to its input. Billions of dollars of R&D expenditures had been
premised on access to foreign markets, and more
expansive export restrictions could hamper the expected commercial As Kevin Wolf,
who directed US export control policy as former assistant secretary of commerce for export
administration in the Obama administration, said in testimony before the Senate Banking
Committee almost a year after
the new law went into effect, “many are wondering what the impact on their businesses will
be and how BIS will justify any new controls based on the ECRA standards.”33
Uncertainty over the commercial implications of new US export control policy increased
with the activity and reports of January and February 2020. US government concerns with ZTE
and Huawei were longstanding and predated the Trump administration. But these events signaled
something new.
The January 6 announcement of new export controls on satellite imagery software, for
example, was significant because AI was on the BIS list of “emerging and foundational
technologies” under examination. The first control on AI exports seemed narrow, but was it just a
start?
Autonomous vehicles were another example of a product for which AI was Fears
that the United States would attempt to control AI for self-driving cars were reportedly
impacting joint ventures between American and Chinese startups, as well as their access to
The American
business
community
was
nonetheless
concerned that
the
BIS scoping
exercise would
be indifferent to
its input.
30 See Federal Register, Review of Controls for Certain Emerging Technologies: A Proposed Rule by the
Industry and Security Bureau on 11/19/2018. These 14 categories were biotechnology; artificial
intelligence and machine learning technology; position, navigation, and timing technol- ogy;
microprocessor technology; advanced computing technology; quantum information and sensing
technology; logistics technology; additive manufacturing (., 3D printing); robotics; brain-computer
interfaces; hypersonics; advanced materials; and advanced surveillance tech- nologies.
31 See Federal Register, Technical Advisory Committees; Notice of Recruitment of Members: A Notice
by the Industry and Security Bureau on 04/01/2019.
32 See Martin Chorzempa’s PIIE Trade and Investment Policy Watch blogs, Worst Case Averted on
Foreign Investment Reviews, August 20, 2018; and The Trump Administration’s Rush to Curb
Technology Leakage Is in Danger of Backfiring, January 8, 2019.
33 Kevin Wolf, “Confronting Threats from China: Assessing Controls on Technology and Invest- ment,”
testimony before the Senate Committee on Banking, Housing, and Urban Affairs, June 4, 2019.
34 For early research on AI and international trade, and potential implications for policy, see Gold- farb and
Trefler (2019).
35 Trefor Moss, “.-China Trade Tensions Jeopardize Rollout of Self-Driving Vehicles,” Wall Street
Journal, September 24, 2019.
WP 20-8 | MAY 2020 10
More generally, economists Kyle Handley and Nuno Limão have documented evidence from
a variety of settings showing that uncertainty over foreign market access due to import tariffs can
hamper investment, production, and Changes in US export control policy provided one
more channel by which costs of elevated levels of uncertainty might impact American-based
businesses.
Finally, as one other central part of ECRA, Congress mandated that BIS ensure the
multilateral adoption of any new US controls, noting that unilateral US export controls on
“widely available” goods would be ECRA indicated that if the administration
did not succeed in getting a particular control adopted by other countries within three years,
the US government should drop
This multilateralization requirement mattered for both national security and commercial
reasons. A control would do little to safeguard US national security if the equivalent technology
were available on global markets from other foreign suppliers. And American companies would
be hurt commercially if they were the only ones unable to sell it.
But getting other key national governments to adopt any new US export control required
international cooperation. Understanding how to work the multilateral process was fundamental
if the United States was to considerably expand the scope of its export controls. (The Trump
administration’s poor record of engaging in, as opposed to disrupting, other multilateral
initiatives presents a particular challenge for international )
Previous multilateralization of US export controls, and some of the problems that arose,
provided clues as to where challenges might be expected.
COCOM AND MULTILATERAL EXPORT CONTROLS DURING
THE COLD WAR
How the United States would convince other countries to adopt more stringent export controls
was not, of course, an entirely new question. The origins of the modern approach date to
shortly after the Second World War, as a response
to growing tensions that the United States and its Western European allies had
Getting other
key national
governments to
adopt
any new US export
control required
international
cooperation.
36 See, for example, the discussion in Handley and Limão (2017).
37 Section 1752, Statement of Policy, reads “Export controls applied unilaterally to items widely available
from foreign sources generally are less effective in preventing end-users from acquir- ing those items.
Application of unilateral export controls should be limited for purposes of protecting specific United
States national security and foreign policy interests.”
38 In particular, Section 1758, Requirements to Identify and Control the Export of Emerging and
Foundational Technologies, states
(C) Multilateral Controls.—
(1) In General.—The Secretary of State, in consultation with the Secretary and the Secretary of Defense,
and the heads of other Federal agencies, as appropriate, shall propose that any tech- nology identified
pursuant to subsection (a) be added to the list of technologies controlled by the relevant multilateral
export control regimes.
(2) Items On Commerce Control List Or United States Munitions List.—If the Secretary of State
proposes to a multilateral export control regime under paragraph (1) to add a technology identified
pursuant to subsection (a) to the control list of that regime and that regime does not add that technology
to the control list during the 3-year period beginning on the date of the proposal, the applicable agency
head may determine whether national security concerns war- rant the continuation of unilateral export
controls with respect to that technology.
39 The Trump administration’s destructive actions concerning the multilateral trading system and WTO
(Bown and Keynes 2020) are one example; others include pulling out of the Paris Climate Accord and
the Joint Comprehensive Plan of Action for Iran sanctions, as well as threats to defund the World
Health Organization in light of the COVID-19 pandemic.
WP 20-8 | MAY 2020 11
with the Soviet Union. The fear was that the Soviet Union would improve its military
capabilities through acquisition of western equipment and commercial technologies that might
have dual use, fears that intensified over the following 40 years of the Cold War.
The allies negotiated the Coordinating Committee for Multilateral Export Controls
(COCOM), which went into effect on January 1, 1950. The original members were the United
States, United Kingdom, France, Italy, Belgium, the Netherlands, and Luxembourg;40 with the
exception of Iceland, members of the North Atlantic Treaty Organization (NATO) all joined
over the following years.
COCOM was very different from other agreements affecting international commerce
developed in parallel, most notably the General Agreement on Tariffs and Trade (discussed
below). COCOM was not treaty-based; it was an informal agreement that was established in
secret and did not create binding legal obligations on the countries involved. Consensus drove
decision making, which also meant that any country had a veto. An export control arose by
agreeing to place a product on one of three lists. The first two lists concerned international
munitions and atomic energy. For export control purposes, products on these lists faced an
The third list, involving dual-use technologies, was referred to as the International List, and
its products were subject to export control review as opposed to bans. Products on this list could
be exported subject to a licensing requirement. Frequent debate emerged between COCOM
countries over whether to add a new product to the International List, and the United States was
often more keen than the European members. The geographic proximity of the European allies to
Eastern Europe and the Soviet Union—and their Warsaw Pact alliance—meant that the average
export control had a greater impact on European foreign commercial interests.
On occasion, COCOM debates did escalate into conflict. Most notable was a 1980s dustup
involving exports of submarine technology from Japan and Norway to the Soviet Union. In what
became known as the Toshiba-Kongsberg Incident,
a Japanese and Norwegian firm were charged with falsifying documents and evading government
controls to export quiet submarine propellers to the Soviet fleet to help it evade Western sonar
Each company was sanctioned by its government, but a new debate emerged within
the US government as
to the appropriate American policy response. The Congress was furious that it might have to
allocate hundreds of millions of dollars of additional military
funding to address the new national security threat caused by Soviet acquisition of the sonar-
evading propellers. It proposed legislation, the Garn Amendment, that would have slapped US
sanctions on Toshiba and Kongsberg.
COCOM was
very different
from other
agreements
affecting
international
commerce
developed
in parallel,
most notably
the General
Agreement on
Tariffs and
Trade.
40 For more on COCOM, see Whang (2019). When the Korean War broke out in 1950 and Chinese troops
became involved, the United States pushed for COCOM to apply its export controls to China. They
eventually did, and for the next four decades the relative restrictiveness of the ex- port control regime
with respect to the Soviet Union and China shifted back and forth, based on political developments. For
a discussion, see Meijer (2016, pp. 33–54).
41 Over the second half of the 20th century, other agreements emerged that banned exports of specific
types of military goods, including chemical and biological weapons (Australia Group), nuclear
weapons (Nuclear Suppliers Group), and missiles (Missile Technology Control Regime).
42 See Wrubel (1989) and Morehead (1988-1989) for historical accounts of the Toshiba-Kongsberg incident,
as well as David E. Sanger, “. Changes Its Stance On Damage by Toshiba,” New York Times, March
14, 1988.
WP 20-8 | MAY 2020 12
The Reagan administration pushed back against congressional demands for additional US
penalties beyond those imposed by the Norwegian and Japanese governments. One
administration concern was that the sanctions might cripple Toshiba’s ability to fulfill its
contracts, hurting American consumers and suppliers to the companies. But another worry was
that additional US punishment could have systemic implications for COCOM itself: COCOM
was voluntary and even allied governments might choose to pull out if penalties for violating
rules on export controls became excessive.
This incident highlighted the trade-offs that any multilateral export control regime must
navigate. The lack of multilateral controls meant that one country’s (the United States’)
unilateral export restriction was insufficient to protect national security (., the target country
acquired the technology). But allowing for excessive punishment of parties that violated the
agreement could have resulted in reduced engagement and cooperation overall.
THE WASSENAAR ARRANGEMENT AND
MULTILATERAL EXPORT CONTROLS SINCE 1995
COCOM was dissolved with the end of the Cold War, in 1994, and a new form of multilateral
cooperation emerged to take its place. In 1995, the Wassenaar Arrangement was established
with 33 members as the new, multilateral vehicle for export controls. Importantly, it included
Russia, which, as part of the Soviet
Union, had obviously been a main target of the COCOM export control While
participation in the Wassenaar Arrangement expanded to 42 countries by 2020, China remained
a major The European Union was also not a formal participant, despite
engagement by EU member states in their national capacities.
The Wassenaar Arrangement followed the COCOM model in only some respects. It
continued to be based on voluntary submissions of products that countries wanted to control.
However, it prioritized transparency and shifted its focus to nonproliferation. To do so, it
maintained two lists: The Munitions List covered conventional arms and included rifles, handguns,
machine, submachine guns, bombs, torpedoes, rockets, and missiles; the list of Dual-Use Goods
and Technologies, where more commercial concerns arose, covered nine categories: special
materials and related equipment, materials processing, electronics, computers,
telecommunications and information security, sensors and lasers, navigation and avionics,
marine, and aerospace and
A good or service must satisfy four criteria to be subject to the dual-use list for export
control. First, there was little utility in subjecting an export to
restraints if it was already available from countries that did not participate in the Wassenaar
Arrangement. Second, countries were discouraged from proposing
The Wassenaar
Arrangement
followed
the COCOM
model in only
some respects.
43 See Wassenaar Arrangement on Export Controls for Conventional Arms and Dual-Use Goods and
Technologies, Public Documents, Volume I, Founding Documents, WA-DOC (17) PUB 001.
44 The Wassenaar Arrangement has periodically engaged in “outreach activities” to nonpartici- pating
states, including China. See, for example, Wassenaar Arrangement on Export Controls for
Conventional Arms and Dual-Use Goods and Technologies, Public Documents, Volume IV,
Background Documents and Plenary-related and Other Statements, WA-DOC (19) PUB 006.
45 For the lists as of December 2019, see Wassenaar Arrangement on Export Controls for Con-
ventional Arms and Dual-Use Goods and Technologies Public Documents, Volume II List of Dual-
Use Goods and Technologies and Munitions List, WA-DOC (19) PUB 002.
WP 20-8 | MAY 2020 13
limits on products where the restriction would knowingly not work. Third, product definitions
should include a “clear and objective specification”; broad descriptions would catch commercial
items for which controls were unnecessary. And finally, the product should not be already
controlled by some other regime (., the Munitions List or the Nuclear Suppliers Group).46
Nevertheless, a potentially more expansive US export control policy suggested
at least three areas of tension with the Wassenaar Arrangement framework.
First, the modern pace of technological change posed an immense challenge to any control
regime. Consider an attempt to control semiconductors. According to Moore’s Law, the number of
transistors on a chip doubles about every two years. A controlled technology today could become a
potential commodity item tomorrow. Such a fast pace of innovation raised the concern that BIS
couldn’t possibly evaluate and multilateralize controls for new products quickly enough— .,
before the technology became widely distributed.
A second concern involved the breadth and precision of new export controls. Some emerging
and foundational technologies under the US review process did not naturally fit under the nine
Wassenaar Arrangement categories. Furthermore, the Wassenaar criterion of an item having a
“clear and objective specification” could run afoul of US attempts to introduce more general
classifications of products it wanted to control.
Finally, an asymmetry remained between the potential target of US controls— which might
include China—and those agreed multilaterally. The traditional
focus of the Wassenaar Arrangement had been on nonproliferation and keeping controlled items
away from rogue states. Thus, any US attempt to get partners to control a technology with
respect to a certain country (., China) would require additional bilateral engagement outside
of Wassenaar with the other key suppliers.
For example, recall the Trump administration proposal to stop granting GE a license to
sell commercial jet engines to For this control to protect national security, BIS would
have needed to convince the governments of competing engine makers—., Rolls Royce—to
similarly deny such export licenses from their jurisdictions. Failure to convince the UK
government would have left GE at a commercial disadvantage relative to its global
competitors.
Overall, the consensus nature of the Wassenaar Arrangement, as well as potentially divergent
economic interests among its participating countries, posed a challenge for US unilateralism and
thus an effective export control policy. It would turn out to be much more difficult to
multilateralize US export control priorities than even during the Cold War—the last time such
controls for dual-use items were put to such a stringent multilateral test.
Overall, the
consensus
nature of the
Wassenaar
Arrangement, as
well as
potentially
divergent
economic
interests among
its participating
countries,
posed a
challenge
for US
unilateralism
and thus an
effective export
control policy.
46 “Wassenaar Arrangement “Criteria for the Selection of Dual-Use Items” (adopted in 1994 and
amended by the plenary in 2004 and 2005).
47 Ted Mann and Bob Davis, “Trump Administration Considers Halting GE Venture’s Engine Deliv- eries
to China,” Wall Street Journal, February 16, 2020.
WP 20-8 | MAY 2020 14
GOVERNMENTS SOMETIMES LIMIT EXPORTS FOR
POLITICAL OR ECONOMIC REASONS UNRELATED TO
NATIONAL SECURITY
Any US export control would be effective only if adopted multilaterally and if the technology
were not already widely available from noncontrolled sources. One challenge for rule making
was thus already clear: allied trading partners may not have commercial incentives to sign on to
America’s additional controls. American-made jet engines vacating the Chinese market would
pave the way for additional sales from Rolls Royce, for example. This posed one challenge
to cooperation.
But there was a separate problem. Even allied governments suspected that the true motive
behind US efforts to restrict exports was economic or political (redistributive) gain at their
expense. A quick tour through a simple, hypothetical economic model clarifies why a country can
have incentives to abuse the national security threat justification for its export restrictions and why
partners are wise to be skeptical.
Take the example of the United States being a “large” global supplier of semiconductor-
making equipment. Being large simply means that a change in the level of US exports affects
the world price of the equipment. If it increases supply, the price in the rest of the world falls;
if the United States limits supply, the world price increases.
Now assume that a legitimate national security threat exists and a “negative externality”
arises: The United States and other countries experience social costs not taken into
consideration by commercial actors (similar to the impacts of cross-border pollution or climate
change)—but only if the equipment is traded. No extra social costs result from local US
production or consumption, but something bad happens if the equipment is sold abroad. In this
instance, the standard Pigouvian economic logic for policymakers holds: the first-best
government policy is to attack the externality at its source and limit equipment
exports. If the negative externality of the national security threat is large enough, the first-best
policy could even be a complete export
But now suppose there is no legitimate national security threat to trading the equipment. The
concern is that the United States may sometimes benefit from restricting exports of the equipment
anyway. The wariness of trading partners arises because this benefit occurs at their expense.
But why might the United States impose an export restriction when there is no national
security threat? Limiting foreign sales means more is kept locally, with American consumers
(., the US semiconductor chip industry) enjoying lower prices and increased equipment
availability. However, the reduction in how much the US equipment industry can sell globally
and the lower prices
for domestic sales result in a loss to the industry’s economic well-being. Overall, losses to
American equipment makers are larger than the gains to US
Any US export
control would
be effective
only if adopted
multilaterally
and if the
technology were
not already
widely available
from
noncontrolled
sources.
48 This would likely be the argument of those proposing the restriction of exports to China—., that
China’s access to technology allows its industry to develop and this in itself poses a na- tional security
threat to the United States. Again, it is not for this paper to assess the national security implications of
any individual technology or industry.
WP 20-8 | MAY 2020 15
Even so, the US government might implement an export restriction for political
reasons—for example, because it values the well-being of the US semiconductor chip industry
(consumers) more than that of the equipment makers (producers).
Nevertheless, any benefit to the United States arising through the export restriction comes at
the expense of its trading partners. Limiting US equipment exports increases the price to foreign
consumers (semiconductor companies like TSMC), hurting them more than the benefits to the
rest of the world’s equipment makers (in Japan or South Korea). Similar to the more familiar
example of a large consuming country imposing a small import tariff, the US export restriction
here is therefore a beggar-thy-neighbor
Beyond political motives, there may be additional economic incentives pushing the US
government to impose such restrictions. As one last tweak, suppose the consumers in the
model—the semiconductor industry—are themselves producers and their industry benefits from
increasing returns to scale. Because of learning by doing, each additional unit that the domestic
industry produces allows it to lower its average costs. In that case, the US export restriction on
equipment generates a separate channel through which the American consumer (the
semiconductor industry) benefits at the expense of its competitors in the rest of the world. The
export restriction means firms, like TSMC, in other countries face higher costs for their
equipment inputs relative to the US semiconductor industry. TSMC’s having to reduce its output
increases its costs while the US industry enjoys a reduction to its costs by producing more with
inputs made cheap only by the export restriction on equipment.
The conundrum confronting national export control policy and international cooperation is
now clear. Export controls for legitimate national security reasons can be the first-best policy to
attack the problem at exactly its source and provide benefits to allies. But export controls can
also be misused as a beggar-thy-neighbor policy to redistribute economic well-being, even from
one ally to another.
Trading partners may be suspicious that the real motive for the policy is economic if there is
an informational asymmetry as to whether the national security threat is legitimate. Information
asymmetries may be difficult to overcome if it is hard to foresee all of a technology’s potential
(nefarious) uses at the time the good is traded, or if the full details of the adversarial threat can’t
be
The conundrum
confronting
national export
control policy
and
international
cooperation is
now clear.
49 For the country as a whole, a standard economic result is that a small export restriction im- posed as a
tax can allow overall American economic well-being (of consumers, producers, and the government
combined) to be higher than it is in free trade. But this result is contingent
on the receipt of tax revenue. An export control that prevents a product from being exported altogether is a
quantitative restriction—or quota—with the volume limit set at zero. More gener- ally, the licensing
procedures associated with export controls (even if all applications for export are accepted) are a nontariff
barrier: They impose additional compliance costs to firms that are similar to a tax, but in which the
government collects no tax revenue.
50 See, for example, Bagwell and Staiger (2002). Note that the simplest example—that a country can be
slightly better off with a unilateral export tax relative to free trade—relies on the col- lection of tax
revenue. An export control such as a ban or a nontariff barrier with compliance costs will not generally
make a large supplying country better off. Nevertheless, a large country that imposes an export control
for political or distributional reasons will still pass some of the cost of that policy on to trading partners.
WP 20-8 | MAY 2020 16
revealed to protect the source of the information. Suspicions are also heightened after the excuse
has been abused, as when the Trump administration imposed national security tariffs on steel in
2018.
Thus, just as there is a need to multilateralize legitimate export controls, there are economic
efficiency gains to agreeing to international rules so that governments cooperate and do not
impose excessive and reciprocal export restrictions when national security threats are not present.
Without such rules to guide policy, the noncooperative outcome could prevail: one country
limits its exports (and imposes costs on partners), and other countries do the same
(imposing reciprocal costs). This is the classic prisoner’s dilemma in which all are made worse
off relative to cooperation.
WTO RULES AND EXPORT CONTROLS
In addition to unilateral and multilateral export controls, governments have developed
international rules that help limit the imposition of beggar-thy- neighbor export restrictions for
economic or redistributive gain. A number of rules in this vein, as well as exceptions for
permissible export restrictions, are set out under the World Trade Organization (WTO). The
WTO has also provided a forum for dispute resolution when inevitable trade frictions
However, the WTO has only begun to face the challenges of interactions between trade,
national security exceptions, and export controls.
The WTO came into effect in 1995, building on its predecessor, the General Agreement on
Tariffs and Trade (GATT). A contemporary of COCOM, the GATT established the initial
multilateral rulebook for national commercial policies affecting exports and imports following
the Second World War.
The WTO treats the two main export policies—taxes and quotas—quite differently,
albeit in parallel to its treatment of import-restricting policies. Countries are generally
prevented from exercising export restrictions in the form of quotas or bans. Just as the WTO
frowns on quantitative import limits, export quotas are discouraged under GATT Article XI.
Export taxes, on the other hand, are broadly permissible under the WTO, similar to the
preference for import tariffs over import quotas. Export taxes must also be implemented on a
nondiscriminatory basis and are thus subject to the WTO’s most favored nation (MFN)
There are important exceptions. One is GATT Article XXI, which allowed a broad
carve-out for “security exceptions.” When imposing export controls, as long as countries
respected the spirit of Article XXI, matters addressed by
COCOM did not generally come up under the GATT. A second involves Article XI(1), which
allows for “Export prohibitions or restrictions temporarily applied to prevent or relieve critical
shortages of foodstuffs or other products essential to the exporting contracting party.” This would
presumably serve as the justification
In addition to
unilateral and
multilateral
export
controls,
governments
have
developed
international
rules that help
limit the
imposition
of beggar-
thy-neighbor
export
restrictions for
economic or
redistributive gain.
51 See Bown and Keynes (2020) for a discussion of the WTO’s current dispute settlement status.
52 See Mavroidis (2016, pp. 87–89) on export taxes under the GATT and WTO. Wu (forthcoming)
provides a discussion of export tax commitments that countries have undertaken as part of preferential
trade agreements.
WP 20-8 | MAY 2020 17
adopted by a country challenged to explain its export restriction on food staples imposed during
the commodity price spike of 2008–11, or on medical supplies during the COVID-19
One crucial difference between export taxes and import tariffs is that few countries have
taken on legal commitments at the WTO to constrain how high their export taxes might go. For
decades starting in 1947, countries negotiated rounds of reductions to their import tariffs and
then agreed to schedule (bind) them, legally promising not to raise them above a certain level.
For the most part, governments have not made similar promises about their export taxes under
the GATT or There are a few exceptions, with China being the most significant. China
committed to schedule and bind its export taxes when it acceded to the WTO in 2001.
WTO DISPUTE SETTLEMENT, NATIONAL
SECURITY, AND EXPORT RESTRICTIONS
In addition to its rules, a second main function of the WTO historically has been to provide a
forum to resolve commercial disputes between Because countries have political and
economic incentives to impose export restrictions unilaterally, some frictions over such policies
were inevitable.
Nevertheless, there have been very few documented cases in which a government imposed
an export control, faced a WTO dispute, and used national security as a defense. One example is
the WTO dispute over Japan’s more stringent controls on exports to South Korea in 2019. Japan
suddenly made Wassenaar Arrangement–controlled fluorinated polyimide, resist polymers,
and hydrogen fluoride—inputs used to make products like smartphones, television displays, and
semiconductors—subject to license requirements for sale to South
South Korea quickly filed a WTO dispute; in its view, Japan’s action did not arise from
an increased national security threat but was simply retribution for a diplomatic flare-up
involving reparations for Japanese mistreatment of
South Koreans during the Second World War. At the timing of writing, the WTO
There have been
very few
documented
cases in which a
government
imposed an
export control,
faced a WTO
dispute, and
used national
security as a
defense.
53 See Giordani, Rocha, and Ruta (2016) and Martin and Anderson (2012) for export restrictions on food; for
export restrictions related to COVID-19, see Soumaya Keynes, New trade barri- ers could hamper the
supply of masks and medicines, The Economist, March 11, 2020; Chad P. Bown, EU limits on medical gear
exports put poor countries and Europeans at risk, PIIE Trade and Investment Policy Watch, March 19,
2020; and Chad P. Bown, COVID-19: Trump’s curbs on exports of medical gear put Americans and others at
risk, PIIE Trade and Investment Policy Watch, April 9, 2020. The US export controls of April 2020 on
medical gear were imposed under the Defense Production Act and not ECRA.
54 The United States does not generally implement export taxes as they are banned under US
Constitution Article I, Section 9, Clause 5.
55 WTO dispute settlement is facing a separate challenge due to the Trump administration’s refusal to
allow the appointment of new members to its Appellate Body (Bown and Keynes 2020). While
critical, these additional institutional challenges are not a focus here.
56 See, for example, , , and in Wassenaar Arrangement on Export Controls for
Conventional Arms and Dual-Use Goods and Technologies Public Documents, Volume II List of Dual-
Use Goods and Technologies and Munitions List, WA-DOC (19) PUB 002. Prior to Japan’s action,
South Korea was on a white list of countries where such shipments didn’t require an individual license.
WP 20-8 | MAY 2020 18
dispute was still in process. If it moves forward, Japan may claim that its export controls arose
after a threat to its national security, justifying its actions under an Article XXI
There have even been few WTO disputes in which countries adopted the national security
defense for challenges to their import-restricting policies.
The only case to have reached a legal decision involved Russia, which used the justification when
Ukraine challenged trade barriers imposed during the military conflict between the two
There are other such defenses in the pipeline, however, including a number of
challenges to the Trump administration’s tariffs on steel and aluminum imposed in
There have been three main reasons behind the limited number of formal WTO disputes
involving a country’s national security: Countries were hesitant to impose trade restrictions in
the name of national security, trading partners were hesitant to file disputes in which that was
the likely defense, and countries were hesitant to invoke the defense if challenged. These three
hesitations arose out of recognition that the WTO would be put in a lose-lose position if forced
to rule on any country’s national security defense. Striking down the measure would jeopardize
the legitimacy of the WTO from one side—the WTO would
be accused of threatening a member country’s sovereignty. But upholding the measure meant
attacks from the other side—countries would be free to invoke the defense over seemingly
anything, rendering meaningless even the most basic WTO
Unrelated to national security, countries have made some WTO challenges to the beggar-
thy-neighbor effects of trading partners’ export restrictions.
China has faced the most disputes, in part because it has taken on the most commitments over
limiting its export restrictions. Japan, the United States, and the European Union, for
example, felt the brunt of China’s export restraints on rare earth elements and brought a
dispute in 2012. At the time, China provided 97 percent of the world’s supply of elements of
critical importance for both renewable energies and the defense But countries have
been concerned about the negative impact on their industries of other Chinese export restrictions,
including some on raw materials and primary aluminum. By
Hesitations
arose out of
recognition that
the WTO would
be
put in a lose-
lose position
if forced to
rule on any
country’s
national
security
defense.
57 See Japan – Measures Related to the Exportation of Products and Technology to Korea, WT/
DS590/3, September 30, 2019.
58 See Russia – Measures Concerning Traffic in Transit - Report of the Panel, WT/DS512/R, May 4, 2019.
59 See, for example, United States – Certain Measures on Steel and Aluminium Products – Request for
consultations by the European Union, WT/DS548/1, June 6, 2018.
60 Pinchis-Paulsen (2020) notes these and related concerns that came up during the original GATT
negotiations in the 1940s that resulted in Article XXI.
61 See Morrison and Tang (2012) as well as Bond and Trachtman (2016) for a discussion of the WTO
dispute.
WP 20-8 | MAY 2020 19
limiting exports strategically, Beijing was providing unfair advantages to Chinese manufacturing—
which relied on inputs made cheap locally because of the restrictions—that caused harm to foreign
Because most other WTO members have fewer legal obligations, there was less constraint on
their use of export restrictions and thus there were fewer disputes. Nevertheless, some cases have
arisen. In the late 1990s, out of concern for its manufacturers of footwear, automotive seating, and
other leather- consuming industries, the European Union challenged Argentina, Pakistan, and India
for their limits on cowhide and leather
In a few instances, disputes arose when countries were too aggressive at countering the
beggar-thy-neighbor effects of the export restrictions unilaterally. Consider Indonesia’s export
tax for palm oil and Argentina’s export tax for soybeans. 64 The economic effect was that each
country provided an implicit subsidy to its downstream biodiesel industry, which the European
Union targeted with countervailing and antidumping duties. Indonesia and Argentina challenged
the EU trade remedies targeting the export restrictions, but the restrictions went unaddressed in
Geneva.
All told, the relatively limited frequency of trade disputes over export controls, export-
restricting policies, and invocations of the national security justification could change.
Certainly a US policy decision to impose additional export controls for dual-use technologies
would likely bump up against other US commitments in international agreements, including
those at the WTO, and lead to more of such frictions. Given the extreme political sensitivity of
such cases and those who would point to this as an erosion of national sovereignty,
the multilateral trading system may need to contemplate new means of resolving such frictions.
CONCLUSION
Much of 2017–20 found the Trump administration debating and then imposing tariffs under the
justification that imports, as well as China itself, posed a threat to America’s national security.
Somewhat less public was a concomitant, but perhaps more politically bipartisan, potential shift
of US export policy. Though the exact direction of US export control policy remains uncertain,
farther- reaching government restrictions on foreign sales of American-made goods and services
seem likely.
Exports pose a distinct national security threat than imports. And how to effectively restrain
American exports presented a number of different policy challenges—domestically, in the
Wassenaar Arrangement, and even at the WTO. Furthermore, any sudden policy shift would lead
to both short- and long-run
The relatively
limited
frequency of
trade disputes
over export
controls,
export-
restricting
policies, and
invocations of
the national
security
justification
could change.
62 See the WTO disputes China – Raw Materials (DS394) and China – Raw Materials II (DS508).
According to OECD (2019a), China’s downstream aluminum manufacturing has benefited from implicit
subsidies resulting from its export restrictions on primary aluminum. Thus this would have also likely
been an issue in the China – Subsidies to Producers of Primary Aluminium (DS519) that the Obama
administration filed at the very end of its administration but that was not pursued by the Trump
administration.
63 See Argentina – Hides and Leather (DS155), Pakistan – Export Measures Affecting Hides and Skins
(DS107), and India – Measures Affecting Export of Certain Commodities (DS120).
64 For a discussion, see Fischer and Meyer (2020) as well as EU – Biodiesel (DS473) and EU –
Biodiesel (Indonesia) (DS480).
WP 20-8 | MAY 2020 20
costs for the US economy. In the interim, lingering policy uncertainty over future access to
foreign markets may have crimped US investment in R&D and imposed separate costs of its own.
New conflicts between the United States and its allies also seemed likely to emerge in the
struggle to align export controls, because of divergent commercial interests. Cooperation was
hindered by increased international skepticism—fed by American abuse of the national security
justification—that countries were acting without their national security actually being under
threat. With the existing multilateral framework for adopting export controls the legacy of an
earlier era, a result could be more trade frictions sent to the WTO—a multilateral institution both
without much experience resolving these types of disputes and already under attack by the US
administration.
As of the time of writing, US export policy was still a work in progress. Major
unknowns included how America’s own list-review regulatory process would evolve, as well
as whether and how successful the United States would be at getting allies to adopt similar
controls multilaterally. What is clear is that
failure to strike the balance between protecting national security and minimizing negative
commercial consequences would be costly for the US economy, ineffective at addressing
national security risks, and problematic for trade and diplomatic relations.
REFERENCES
Bagwell, Kyle, and Robert W. Staiger. 2002. The Economics of the World Trading System.
Cambridge, MA: MIT Press.
Bond, Eric, and Joel Trachtman. 2016. China–Rare Earths: Export Restrictions and the Limits of
Textual Interpretation. World Trade Review 15, no. 2: 189–209.
Bown, Chad P., and Soumaya Keynes. 2020. Why Trump Shot the Sheriffs: The End of WTO Dispute
Settlement . Journal of Policy Modeling, forthcoming.
Congressional Research Service. 2020. The US Export Control System and the Export Control Reform
Initiative. CRS Report R41916, updated January 28. Washington.
Fischer, Carolyn, and Timothy Meyer. 2020. Baptists and Bootleggers in the Biodiesel Trade: EU-
Biodiesel (Indonesia). World Trade Review, forthcoming.
Giordani, Paolo E., Nadia Rocha, and Michele Ruta. 2016. Food prices and the multiplier effect of trade
policy. Journal of International Economics 101: 102–22.
Goldfarb, Avi, and Daniel Trefler. 2019. Artificial Intelligence and International Trade. In The Economics of
Artificial Intelligence: An Agenda, ed. Ajay Agrawal, Joshua Gans, and Avi Goldfarb. Chicago:
University of Chicago Press.
Handley, Kyle, and Nuno Limão. 2017. Trade under . policies. In Economics and Policy in
the Age of Trump, ed. Chad P. Bown. London: Centre for Economic Policy Research.
Martin, Will, and Kym Anderson. 2012. Export Restrictions and Price Insulation During
Commodity Price Booms. American Journal of Agricultural Economics 94, no. 2: 422–27.
Mavroidis, Petros C. 2016. The Regulation of International Trade, Volume I: GATT.
Cambridge, MA: MIT Press.
Meijer, Hugo. 2016. Trading with the Enemy: The Making of US Export Control Policy toward the People’s
Republic of China. Oxford: Oxford University Press.
New conflicts
between the
United States
and its allies
also seemed
likely to emerge
in the struggle
to align export
controls,
because of
divergent
commercial
interests.
WP 20-8 | MAY 2020 21
Morehead, Jere W. 1988-1989. Controlling Diversion: How Can We Convert the Toshiba- Kongsberg
Controversy into a Victory for the West. Northwestern Journal of International Law and Business
9: 277–95.
Morrison, Wayne M., and Rachel Tang. 2012. China’s Rare Earth Industry and Export Regime: Economic and
Trade Implications for the United States. Washington: Congressional Research Service.
OECD (Organization for Economic Cooperation and Development). 2019a. Measuring distortions in
international markets: The aluminium value chain. OECD Trade Policy Papers no. 218. Paris.
OECD (Organization for Economic Cooperation and Development). 2019b. Measuring distortions in
international markets: The semiconductor value chain. OECD Trade Policy Papers no. 234. Paris.
Pinchis-Paulsen, Mona. 2020. Trade Multilateralism and US National Security: The Making of the GATT
Security Exceptions. Michigan Journal of International Law 41: 109–93.
Whang, Cindy. 2019. Undermining the Consensus-Building and List-Based Standard in Export
Controls: What the United States Export Controls Act Means to the Global Export Control
Regime. Journal of International Economic Law 22, no. 4 (December): 579–99.
Whitten, Reid, and Rachel C. Mays. 2019. Chasing Huawei: BIS May Change the Rules of the Game to Target
One Player. National Law Review, December 18.
Wolf, Kevin J., Steven C. Emme, and Robert Monjay. 2020. A Look at New Limits on Geospatial
Imagery Software Exports, Law360, January 7.
Wrubel, Wende A. 1989. The Toshiba-Kongsberg Incident: Shortcomings of Cocom, and
Recommendations for Increased Effectiveness of Export Controls to the East Bloc. American
University International Law Review 4, no. 1: Article 8.
Wu, Mark. Forthcoming. Export Taxes. In Handbook of Deep Trade Agreements, ed. Aaditya Mattoo, Nadia
Rocha, and Michele Ruta. Washington: World Bank.
