Over the years I’ve learned that working on export controls is like being in quicksand—once in, it’s very hard to get out. So, once again, I will use my 900 words on that topic. I leave the detailed description of what the Bureau of Industry and Security (BIS) recently announced to the practitioners and other experts. (CSIS’s Greg Allen has put out an excellent detailed summary of what the new rules mean for artificial intelligence, or AI).
In brief, the new rules impose licensing requirements—in some cases with a presumption of denial—for exports of high-end semiconductors and chip design software that would enable China to enhance its AI or supercomputing capabilities. Similar controls are placed on advanced semiconductor manufacturing equipment in order to make it more difficult for China to make its own advanced chips—both logic and memory. In addition, the process for maintaining and adding companies and individuals to the BIS Entity List is revamped, in large part to help BIS deal with the front company problem—a company is listed and then promptly sets up a shell company with a different name that, being new, is not listed. (This is a fitting reminder that export controls amounts to an eternal cat and mouse game where each side always has another move it can make.) The new rules denote a significant change of policy, which National Security Adviser Jake Sullivan articulated in two recent speeches, one at the Special Competitive Studies Project and one at Georgetown, and I suspect there is more to come.
Sullivan acknowledged that U.S. policy for at least the last 25 years has been to keep our adversaries one or two generations behind us technologically, and he said that the Biden administration was moving beyond that and would now seek to actively degrade China’s military capabilities, which in practice means recontrolling some items we had previously allowed to be exported.
Our past policy has been effective because it enabled U.S. high-tech companies to expand their sales and profits, which they plowed back into development of next-generation products, while at the same time keeping China behind us. Because the United States did not cut off all high-tech exports to China, we also did not increase incentives for them to go it alone and develop their own technology.
The change, however, is not a surprise. For more than a year, I have been suggesting that our policy may have run out of gas. Xi Jinping has chosen to pursue a path that focuses on independent development of Chinese technology and products (albeit with stolen U.S. intellectual property), and the deteriorating bilateral relationship, accompanied by expanding U.S. sanctions, has only accelerated that effort. As a result, a policy to simply keep them behind may no longer be enough. (Although, staying ahead by running faster is an important corollary of our policy—see the CHIPS and Science Act—that the Biden administration is pursuing.)
The next shoe to drop will be due to China’s civil-military fusion doctrine, which makes clear that Chinese companies in the civilian sector are expected, and may be called on, to provide support for the military. The consequence is that there are no longer any reliable end-users in China, and our more than 25-year-old policy of distinguishing between civilian end users and end uses and military ones is no longer viable. That portends a return to the Cold War days of broad, blanket controls. Thus far, the administration has not explicitly abandoned that policy and has instead chosen to add companies to the Entity List, making a license mandatory for any export to those listed. But stay tuned.
The policy question is whether the benefit of these changes exceeds their cost. The benefit is denial to China of critical high-end technology that would enhance its AI and supercomputing capabilities that have both military and human rights implications. Note that the benefit is significantly reduced if the United States cannot persuade other manufacturing countries, primarily Japan, South Korea, the Netherlands, and Taiwan to go along. That appears to be a work in progress, and it is surprising that the administration did not nail it down before making its announcement.
The cost is the risk of overcontrolling. One part of that is the problem of kneecapping oneself—taking away from companies a revenue stream that has been an important part of their profitability and whose absence may retard the development of next generation technologies and get in the way of our running faster strategy. The other part is the incentive the new rules give China to accelerate its independent technology development. The administration’s counter will be that the controls are airtight. It seems to have done a better job of eliminating loopholes than the Trump administration, but there is always leakage, particularly if other countries only cooperate reluctantly. Another thing to stay tuned for is Chinese retaliation. They always do, though not always right away.
Unfortunately, we won’t know who is right for some time, and much will depend on how the new rules are implemented and enforced. (There are already exceptions for some foreign-owned companies manufacturing in China.) That means the political debate will continue. China hawks will demand a near embargo, and business will complain the rules go too far. Neither will put forward a better plan. The hawks cannot explain away the damage their proposals would do to our economy, and business cannot claim our old policy still has gas in the tank. The administration’s challenge is to continue walking the fine line between over- and under- controlling without tilting too far in either direction.
William Reinsch holds the Scholl Chair in International Business at the Center for Strategic and International Studies in Washington, D.C.
SUBSCRIBE TO WILLIAM REINSCH'S WEEKLY COLUMN
Commentary is produced by the Center for Strategic and International Studies (CSIS), a private, tax-exempt institution focusing on international public policy issues. Its research is nonpartisan and nonproprietary. CSIS does not take specific policy positions. Accordingly, all views, positions, and conclusions expressed in this publication should be understood to be solely those of the author(s).
© 2022 by the Center for Strategic and International Studies. All rights reserved.
