After starting out as an importer of American semiconductors in the 1970s, Japan rapidly rose to dominate the chip industry, surpassing the U.S. in global sales of memory chips by 1982 and accounting for 51 percent of the global semiconductor market by 1988.
The meteoric rise was made possible by massive government investments: Semiconductor R&D accounted for 26 percent of Japan’s total R&D spending in 1977. Further, private-public partnerships allowed Japan’s leading tech companies to collaborate, while largely keeping foreign chips out of the country.
But it also made the Japanese semiconductor industry rigid. When market forces started to change in the late 1980s and early 1990s, Japanese companies struggled to adjust and the industry fell almost as quickly as it rose.
Now, as President Trump works to boost the American semiconductor industry, he seems to be falling into the same trap that led to Japan’s collapse in semiconductor relevance.
In August, Trump and Intel announced a “historic agreement” that saw the U.S. government purchase 10 percent of the struggling semiconductor manufacturer for $8.9 billion. To buy the stock, Trump used $5.7 billion of CHIPS and Science Act grants that had already been awarded to Intel, with another $3.2 billion coming from a Defense Department contract that was also previously awarded to Intel.
Shortly after the deal, National Economic Council Director Kevin Hassett said the administration was interested in buying similar stakes in other companies, so as to boost the domestic companies leading America’s AI race against China, but the move has left even some Trump allies questioning the decision to directly influence the economy.
“I don't think the federal government should be taking equity stakes in private companies,” Senate Commerce Committee Chair Ted Cruz told National Journal in early September. “I think that is a mistake and a slippery slope.”
The Intel press release said the government would only have “passive ownership” over the company, with no board seats or other governance rights.
But Trump has shown in the past that he is more than willing to use the power of the executive branch to bend any agreement, and force companies to comply with his will and whims.
On Thursday, Nvidia announced a $5 billion investment in Intel that will allow the two companies to collaborate on R&D. The company previously agreed to give the government 15 percent of certain chip sales to China, further bending the divide between government and private industry.
If Trump uses CHIPS and Science funds, or any other money the executive branch may have access to, to buy stakes in more chip-manufacturing companies, it would place the American semiconductor industry in the same position as Japan in the 1980s, potentially hurting domestic innovation and flexibility, and undercutting the biggest advantage the U.S. has in the race to beat China in the AI market: innovation.
The Chinese government already has ownership stakes in its domestic chip-manufacturing companies, allowing Beijing to tip the economic scales to help its favored companies, while China's willingness and ability to invest in chip R&D will likely always outpace America’s. If America tries to beat the Chinese in the AI race by playing by their rules, it will lose.
America’s ability to lead the global economy has always relied on the flexibility of the American economy and willingness to innovate, even if innovation means the death of former national champions in industry like Intel.
—Philip Athey
