What to do about Intel Corp, the once-storied, now-spiraling US chipmaker?
After years of poor decision-making, Intel’s semiconductors lag competitors and its foundry business has no major customers. It has mostly missed the boat on artificial intelligence. Last year, amid a restructuring, it fired its chief executive, slashed its dividend, undertook mass layoffs and lost about US$19 billion ($24.5 billion).
One thing not to do: Give the US government a 10% equity stake in the company, as the White House has just done. Although the administration has given few details about how this arrangement will work, the risks are all too clear.
A government stake of this size in a private company will almost certainly lead to conflicting objectives and politicised decision-making. It could impede productivity, innovation and growth. It will surely distort competition, invite corruption and set a bad precedent. (Those free-market Republicans newly open to the idea might ponder what a Gavin Newsom administration would do with such powers.)
In fact, nearly every problem Intel faces could be worsened by this idea. Hard decisions will likely be delayed, perhaps indefinitely. Shareholder concerns may be subordinated to partisan ones. A federal backstop may temporarily shore up the company’s stock price but will also erode its competitiveness over time. Has such an arrangement — anytime, anywhere — ever yielded the sort of inventiveness or strategic insight that Intel’s turnaround will require?
Making matters worse, according to some reports, the administration may pressure other technology companies to use Intel’s foundry or buy its chips, thereby forcing more productive actors to subsidize their competitor’s failures — against their will and to their detriment. Although Treasury Secretary Scott Bessent has denied that such a step is under consideration, one has to wonder.
After all, the administration seems partial to this kind of thing. It has previously proposed taking a stake in the social-media platform TikTok, offered export-control exemptions to chip companies that agree to cough up 15% of their revenue, and extracted a “golden share” in United States Steel Corp as part of a deal to approve its sale to a Japanese buyer. Last month, the Pentagon took a US$400 million stake in a rare-earth mining company. Other “deals” are reportedly on the table.
Added up, these ideas start to look like patronage-as-policy. What should happen instead?
Where semiconductors are concerned, Congress should undertake some overdue reforms: Offer a chipmaker’s visa and otherwise expand skilled immigration; do more to rein in the costly legal wrangling induced by permitting laws; slash tariffs on relevant parts and components; invest in transmission lines and grid expansion; eliminate labor and procurement terms that inflate costs; boost semiconductor research and development (R&D) funding; improve the tax treatment of investment spending; and more. None of this would fix Intel overnight, but it would make US chip manufacturing more competitive, which is supposedly the objective.
To the extent the administration is concerned about national security, a better approach is to bolster efforts like the Pentagon’s Trusted Foundry and quantifiable-assurance programs, which aim to create secure areas and processes in commercial plants to meet the military’s needs. A separate effort, called Secure Enclave, which has extended some US$3 billion to Intel to produce chips at a specialized facility, is unlikely to be sustainable unless it’s opened to more companies; the goal should be cost-effective subsides tightly focused on defense requirements, not an open-ended bailout.
As for Intel itself? In the normal course of things, the free market would determine its fate and creative destruction would work its magic as needed. If the government wants to impede this process, it should explain its rationale, offer a plausible vision for the future, and try to mitigate the predictable drawbacks. So far, it’s offering nothing of the sort.
