This article is part of a new blog series called INFER Insights. Each post will explore a question we're monitoring on INFER and compare the crowd consensus with expert perspectives surrounding the possible outcome.
In October of last year, the U.S. announced new export controls preventing the sale of certain advanced semiconductors to China—along with the software and machinery required to make them. The new controls are a key part of the Biden administration’s stated goal of maximizing the U.S. technological advantage over China.
But will the controls work? From their announcement, U.S. officials conceded that they would not be effective without cooperation from allies.
"We recognize that the unilateral controls we're putting into place will lose effectiveness over time if other countries don't join us," Reuters quoted one U.S. official as saying. "And we risk harming U.S. technology leadership if foreign competitors are not subject to similar controls."
Forecasters on INFER have been tracking the likelihood that allies announce they're joining the U.S. effort.
The INFER crowd forecast indicates that it’s more probable than not that both Japan and the Netherlands will announce that they’re joining the U.S. chip ban by the end of March (57% and 58% chance respectively). On the other hand, they think it’s improbable that South Korea will make such an announcement (25%) and highly improbable that any other country does (16% chance).
Those forecasts are in line with reporting from Reuters and Bloomberg that a deal has been reached—or will soon be reached—between the U.S., Japan, and the Netherlands. And they reflect the current U.S. focus on the machinery necessary to manufacture certain advanced semiconductors. The U.S., Japan, and the Netherlands are the only makers of such machinery, so a deal between them would make it very difficult for China to access those machines. South Korea is a different case: It produces two-thirds of the world’s memory semiconductors and is a buyer of some of the advanced machinery that the U.S. is seeking to control. There is concern that manufacturing equipment could find its way to South Korea and then to China, according to one export-control expert who was not authorized to speak on the record.
The reported deal with Japan and the Netherlands is likely the start of a new U.S. export control effort, not the end of it. And the forecasts available on INFER, along with analysis by experts in the field of export controls, point to two key barriers that the U.S. will have to overcome to win more converts to its agenda.
First, allies need to have the political will: They must believe that the benefits in terms of security exceed the commercial costs. And second, they need the legal ability to actually participate—or the ability to secure new legal authority. The seriousness of these barriers varies by country, by technology, and based on the details of a given policy. And INFER forecasters’ discussion of the chip ban highlights the importance of both.
“In making policy decisions, allies will consider many factors including their own pragmatic interests,” says Jeanette Chu, a non-resident senior associate at the Center for Strategic & International Studies.
That means allies’ participation will hinge in part on the extent to which they worry about China’s technological development. It will also depend on how they perceive the tradeoff between economic and national security concerns.
An export control prevents a country’s companies from accessing a given market or from selling to a given customer. The more important that business is, and the more influential an industry in a country’s political system, the harder the sell.
The commercial concerns are real: U.S. unilateral controls on commercial satellites in the 1990s “seriously hurt [the] U.S. satellite industry and helped Japanese and European competitors,” according to Kevin Wolf, a non-resident senior fellow at Georgetown’s Center for Security and Emerging Technology. “Plurilateral” deals between several countries can help blunt that effect, but export controls can still harm an industry’s competitiveness and invites competition from companies based in countries not party to the deals.
These twin concerns—over security and competitiveness—shape the political will of a given ally to participate. And not all U.S. allies are on board with the breadth of the new controls. Historically, deals on export controls have tried to limit themselves to specific goods linked to specific harms, like munitions. Semiconductors, by contrast, are a foundational technology for commercial as well as military uses.
“Most allies are simply not convinced that regulating commercial semiconductor inputs is necessary for national security,” says Wolf.
Even many allies that support the new U.S. export controls may find it difficult to sign on because they lack the legal authority to do so.
Since the end of the Cold War, international export controls have been dominated by the Wassenaar Arrangement
, a multilateral agreement involving 42 states and mostly focused on issues of non-proliferation and non-state actors. Wassenaar does not allow the targeting of individual countries. Designed in the wake of the Soviet Union’s fall, it counts Russia as a participant.
The U.S. has long had its own, varied approach to export controls, above and beyond the Wassenaar Arrangement. It includes lists of banned items (like munitions), banned uses (like making a weapon of mass destruction) and banned customers (like Huawei).
But few other countries have such a history. Almost none use the approach of banning end-users, for example.
“The main issue is the allies don’t have the legal authority in their systems to impose controls on software and technology and equipment outside the multilateral regime system,” says Wolf, adding that Japan and the Netherlands have had to stretch to find the authority to join the U.S.
Whereas the U.S. delegates broad authority for export controls to the president, most allies need to go back to their legislatures for new authority. In some cases, countries’ legal authority is tied to multilateral agreements: Japan, for example, is legally prohibited from undertaking export controls unilaterally. To join in the current U.S. effort, many allies will have to pass new laws—and that could take time.
“It could take several months for Japan and the Netherlands to implement the restrictions as they need to prepare domestic legislation,” one INFER forecaster noted
, citing reporting from Bloomberg.
And there’s at least some possibility of an ally agreeing to a deal but not announcing it. Wolf believes it’s highly probable that Japan and the Netherlands will publicly agree to a deal with the U.S. on semiconductor equipment exports. But he notes that it’s at least conceivable, if unlikely, that things could proceed less publicly. “Things can be done much more informally,” he says, with governments quietly asking a couple of companies to just not sell certain machines to China.
In a September 2022 speech
, Biden’s national security advisor Jake Sullivan laid out several areas of technology that would be crucial for U.S. policy going forward. That list included quantum computing, AI, biotechnology, and clean energy. It’s possible that the U.S. export control strategy against China will expand beyond chips to include some of these areas. If it does, success will once again depend in part on allies. And those allies’ willingness to participate will hinge on the same factors that experts and forecasters have highlighted here: the combination of legal authority, economic competitiveness, and political will.
By: Walter Frick
Walter Frick is the founder of Nonrival, a newsletter that lets readers make predictions about tech, business, and the economy. He is a contributing editor (former senior editor) at Harvard Business Review, former executive editor at Quartz and has written for publications including The Atlantic, the BBC, and MIT Technology Review.