End Semiconductor Industrial Policy

May 8, 1998 • Commentary
This article appeared in The Journal of Commerce on May 8, 1998

Should taxpayer money be used to develop technologies that will be licensed to foreign firms? That’s the question raised by critics of a new alliance of Intel, Motorola, Advanced Micro Devices, and the Department of Energy to develop extreme ultraviolet light (EUV) chip‐​making technology.

The critics argue that technology funded by U.S. taxpayers should be kept in American hands. Here’s a better idea: instead of putting strings on corporate welfare, let’s end it altogether.

The Cato Institute recently published a briefing paper titled “Silicon Valley versus Corporate Welfare” by Cypress Semiconductor CEO T. J. Rodgers. He argues that the estimated $65 billion a year that the U.S. government now spends on business subsidies harms American industry in general and Silicon Valley in particular. U.S. competitiveness would be enhanced by eliminating those programs and devoting the resulting savings to cuts in corporate income, personal income and capital gains tax rates.

Attached as an appendix to the study is a “Declaration of Independence” from corporate welfare signed by more than 50 Silicon Valley CEOs and leaders. That declaration includes the following pledge: “If an independent commission similar to the military base‐​closing commission identified a fair and substantial government spending cut in the area of so‐​called ‘corporate welfare,’ I would support that cut, even if it meant funding cuts to my own company.”

It’s noteworthy that among the signers is Jerry Sanders, CEO of Advanced Micro Devices — one of the participants in the EUV venture with the Department of Energy.

Of course, supporters of high‐​tech R&D subsidies contend that this kind of government involvement isn’t really corporate welfare. Rather, it’s a necessary investment that ensures U.S. technological leadership over Japan and Europe. That’s why the EUV venture has come in for criticism: it’s feared that the results will be shared with foreign companies.

Fortunately, paranoid technological nationalism has abated somewhat in recent years. Only a few years ago, people like Clyde Prestowitz at the Economic Strategy Institute were convinced that the United States was in decline and “trading places” with Japan. Now, in light of current economic realities, even Prestowitz admits that he misjudged the Japanese “threat.”

Even so, misconceptions still abound. In particular, supporters of high‐​tech corporate welfare cling to the myth that government support was responsible for the U.S. semiconductor industry’s competitive turnaround with respect to Japan. The facts, though, tell a different story.

The U.S. government’s major foray into “saving” the semiconductor industry was Sematech — a consortium of large chipmakers that received half its funding from the Pentagon (Sematech still exists, minus the Pentagon money). Sematech was launched in 1988 during a panic over the mass exit of U.S. producers from the DRAM (dynamic random access memory) market. By 1986 ferocious Japanese production had chased all but two American companies away from making DRAMs for sale; the U.S. share of the world market had fallen to 10 percent from 100 percent a decade earlier.

Sematech’s champions believed that commodity memory chips like DRAMs were the key to overall semiconductor competitiveness. The loss of DRAMs, they feared, would lead to the eventual collapse of the whole U.S. industry. Sematech was promoted as a way to help U.S. producers regain their footing in this “strategic” sector.

Notwithstanding Sematech, U.S. producers didn’t return to DRAMs, and the U.S. share of commodity memory production never rallied significantly. Sematech didn’t accomplish what it set out to accomplish. Yet the U.S. industry thrived anyway, making up the lost DRAM market share in other faster growing and more lucrative segments of the industry — notably microprocessors and design‐​intensive logic chips. Japanese producers still remain strong in DRAMs, but there was no domino effect; in fact, Korean and Taiwanese competition has made the DRAM segment much less attractive than those dominated by American firms.

In short, Sematech’s supporters totally misread the future of the industry. It was by ignoring their advice that American chipmakers prospered. Sematech thus deserves no credit for the U.S. industry’s subsequent good fortunes.

Of course taxpayers shouldn’t subsidize technology developed for foreign firms. We shouldn’t subsidize commercial technology for American firms, either. It’s time to end corporate welfare as we know it — in Silicon Valley and elsewhere.

About the Author