It wasn’t supposed to happen. In January 2025, Chinese startup DeepSeek grabbed headlines with the release of a generative AI program to rival ChatGPT—using half the computing power and developed at a fraction of the cost.
“China shocked the world,” says Harvard Business School Assistant Professor Jaya Wen, especially since for years, the US had banned the export of the most advanced computer chips to China in hopes of preventing just that kind of breakthrough. “It turned out, China didn’t need the most advanced chips.”
There’s a big policy debate about export controls and sanctions. First, do they work? And secondly, do they backfire?
The announcement took Silicon Valley by surprise, triggering a sharp sell-off that erased almost $600 billion from Nvidia’s market value overnight, the largest single-day market-cap loss ever recorded for a US company, in what venture capitalist Marc Andreessen called “AI’s Sputnik moment.”
Wen’s working paper, “Export Controls and Innovation in Sanctioned Countries,” may help explain why China surged ahead. Her research, updated in September, shows that export controls can unintentionally accelerate innovation in other parts of the world, since sanctions often spur companies in other countries to compensate for restricted access to key materials.
“There’s a big policy debate about export controls and sanctions,” says Wen. “First, do they work? And secondly, do they backfire?” By studying a cluster of sanctions placed on China nearly 20 years ago, the researchers found the answer to both questions was yes: While sanctions may work to reduce the supply of goods to other countries in the short run, they often backfire in the long run.
“Our paper is the first to show clear causal evidence that there’s a huge innovation surge in the targeted country in response.”
US policymakers, as well as other government officials around the world, have increasingly used export controls to restrict the flow of sensitive goods and technologies in the name of national security and economic competitiveness. Yet Wen’s findings suggest that officials should exercise more caution in implementing restrictions, lest they spur the very innovation they’re trying to prevent.
The US imposes sanctions on China
It was 2007. George W. Bush was president at the time of an inflection point in US–China relations. On the verge of hosting the Olympics, Beijing had gone from being an inexpensive source of foreign manufacturing to a burgeoning global powerhouse. Contrary to the expectations of some US politicians, the country was showing no signs of liberalizing its political system to become more democratic.
“China was no longer our junior partner whom we wanted to integrate into the world economy. It was becoming a strong and powerful nation,” says Wen, who wrote the paper along with Xueyue Liu and Yu Liu of Fudan University; and Alexey Makarin of MIT Sloan School of Management.
In response, the US imposed two sanctions on China: one preventing trade in military equipment and another prohibiting materials and components that could be used to manufacture technology with national security implications. “These are two of the most common motives for export controls,” says Wen. “One [question officials ask] is: How strong is your military? And the other is: How advanced is your technology?”
The controls covered a wide array of products—from jet engines to hydraulic fluids and even certain ball bearings. In reading the Federal Register announcing the controls, however, Wen and her colleagues realized that the government removed several items from the list after comments from the public, including US manufacturers, and detailed review by policymakers at the Bureau of Industry and Security. That quirk gave them two groups of items they could easily compare—items sanctioned versus similar items almost sanctioned.
Sanctions lead to a surge in patents
Looking at both research and development (R&D) spending by Chinese firms and filings in China’s patent system, the researchers found:
Sanctions led to increased research and patents in China. Directly affected companies that were cut off from controlled items increased R&D spending by 49% and boosted their patent output by 41% compared to firms that bought allowed items.
Directly affected firms increased innovation broadly. Among directly affected firms, innovation increased in both controlled and other technologies. Patenting rose by 65% in controlled technologies and by 42% in other technologies. As Wen explains, firms increased patents for replacements for restricted inputs, but they also redesigned products to avoid those inputs altogether.
Upstream suppliers increased innovation in a targeted way. Upstream suppliers responded in a targeted manner. Among Chinese firms producing goods similar to the restricted items, those positioned to develop substitutes increased patenting in controlled technologies by 361%, whereas their patenting in other topics did not respond to the policy.
The policy succeeded in shutting off imports. But as for slowing down China’s technological development, it basically backfired.
Taken together, the findings show a strong impact on manufacturing in the targeted country, along with a boomerang effect that juiced domestic development. “The policy succeeded in shutting off imports,” says Wen. “But as for slowing down China’s technological development, it basically backfired. It gave China the incentive to figure out how to do without these inputs and make them themselves.”
Should governments impose export controls?
The US uses unilateral controls and multilateral agreements to block dozens of exports, including weapons, tools, chemicals, and software. While those controls could succeed in the short run, says Wen, they might also lead to loss of leverage long-term. Based on the findings, Wen makes the following recommendations:
Analyze intended and unintended effects of export controls
“We have to think really carefully about these tradeoffs,” she says. “If you are on the cusp of entering a hot war and these short-run returns are important, maybe it’s worth it. But if it’s just business-as-usual, you may want to consider other tools.”
Reserve export controls for certain scarce materials
Wen says she’s not suggesting export controls are never effective. They could still work well for items that are difficult for a country to go without or manufacture itself. For example, she points to the rare earth elements that China has imposed export controls on to prevent acquisition by the US and other countries. “If you don’t have your own molybdenum in the ground, it’s going to be hard to go and find some,” she says.
Consider other ways to gain geopolitical influence
Before automatically implementing economic sanctions, a country would do well to consider other tools, such as diplomacy or intelligence.
In the end, Wen says, “there’s no silver bullet” to resolving competitive geopolitical tensions. “We’re in a world of imperfect tools, each with its own costs and benefits. And we can have the best tools in the world, but if we don’t know what we are trying to do with them, we’re not going to succeed.”
Illustration by Ariana Cohen-Halberstam with photo by Mike Mareen/AdobeStock.
Have feedback for us?
Export Controls and Innovation in Sanctioned Countries
Liu, Xueyue, Yu Liu, Alexey Makarin, and Jaya Y. Wen. "Export Controls and Innovation in Sanctioned Countries." Harvard Business School Working Paper, No. 25-004, August 2024. (Revised September 2025.)
