While the U.S. reaped the benefits of cheap imports from its new normalized trade partner, the American manufacturing sector took a beating: China’s production explosion accounted for 59.3% of all U.S. manufacturing job losses between 2001 until 2019—about 4 million jobs. Economists David Autor, David Dorn, and Gordon Hanson coined this phenomenon the “China shock.”

A quarter of a century later, some economists have compared this industrial transformation and today’s rise of AI. Like the China shock, AI growth has been associated with a shift in labor: Despite many economists seeing little evidence, so far, of mass job displacement as a result of AI, tech companies have used the technology to justify laying off thousands of workers. Last month, Snap CEO Evan Spiegel announced the reduction of about 1,000 roles at the company, 16% of its staff. Klarna CEO Sebastian Siemiatkowski anticipates AI shrinking the company’s white-collar workforce by one-third by 2030.

“The AI shock is following the same playbook,” Apollo chief economist Torsten Slok said in a blog post this past week. “The displacement force is different this time, impacting cognitive and white-collar work rather than factory floors. But every other element of the structure is remarkably familiar.”

China shock vs. AI shock

According to Slok, the shared themes of labor market upheaval between the AI and the China shocks may not be a bad thing. Following China’s WTO entry, overall U.S. unemployment remained low. As for manufacturing, the sector’s share of the labor market was already in decline well before the China shock as the U.S. transitioned to a service-oriented economy.

Meanwhile, cheaper intermediary goods from China helped boost manufacturing productivity, resulting in a 50% increase in real manufacturing value added from 2001 to 2024.

Slok sees similar trends in productivity and labor in the AI future.

“If history is any guide, the gains will be substantial,” he said. “Just as cheaper Chinese inputs helped U.S. businesses grow and hire, AI is already accelerating business formation and productivity gains across the economy.”

Slok has previously invoked Jevons paradox to explain why AI will actually create more jobs overall, despite some employers attributing layoffs to the technology. In 1965, economist Willian Stanley Jevons observed that following the invention of the Watt steam engine that improved the efficiency of the coal fired engine, coal consumption actually increased significantly as coal-powered energy became cheaper, incentivizing greater coal use.

Similarly, Slok argued in a blog post last month, as AI makes some white-collar work more efficient, the market for those positions expands, creating more jobs. This paradox can be seen in real time in radiology: AI has been able to automate parts of the imaging process, but the number of active radiologists in the U.S. over the past decade has grown by about 10%.

In in latest post, Slok said AI has the potential to shift concentrations of jobs in different areas or create new jobs altogether, similar to how the rise of Chinese manufacturing cemented the U.S.’s service economy while growing its productivity in manufacturing.

“The bottom line is that we have seen this before,” he concluded. “Just as the China shock gave way to new industries and stronger businesses, AI will drive productivity gains and create opportunities that will more than replace jobs lost today.”

The case against a China shock redux

Autor, the economist who helped coin the term “China shock,” isn’t as convinced of the parallels. In an episode of the Possible podcast hosted by LinkedIn co-founder Reid Hoffman, he said AI “will not be, in any sense, a repeat of the China trade shock.”

Unlike Slok, Autor argued AI will displace jobs, and will do so differently than the China shock did. He suggested AI will target job functions, not specific industries or regional geographies, increasing the likelihood of AI creating even greater labor changes—though not wiping out any one profession.

How these labor changes are perceived will also be different than the early 2000s, Autor said. 

“The China trade shock was experienced by U.S. firms as a pure negative competitive shock,” he said. “All of a sudden, they couldn’t charge the prices they were charging. Someone else was charging much less. And so from a firm perspective, this was all bad.”

AI, however, has the potential to drive up productivity and lower prices, Autor argued, making it appealing from a business perspective, but potentially more disruptive to labor.

“AI will be experienced by many firms as productivity increases, so it may still lead to displacement of workers,” Autor said. “In fact it will, I don’t want to suggest it will not. But it will have a very different texture.”