The last two decades have exposed significant conflicts among these traditional influences on policy, largely brought about by three important changes: (1) the end of the cold war, (2) the general stagnation of the Japanese economy, and (3) the globalization of the computer and semiconductor industries into well-established7 and mutually dependent supply chains and markets.
Federal funding of electronic development, from the launch of Sputnik in 1957 almost until the fall of the Berlin Wall, was driven by perceived military requirements, which had significant noneconomic motivation. During this period, significant federal R&D investment was made in innovative semiconductor technology for military application. After some cost reduction and normal technology adoption delay, the same technology and technology roadmap steadily appeared in the commercial market, including advanced compound semiconductors and dramatically new manufacturing equipment, which also found strong commercial adoption, for example, in lithography.
By the mid-1990s this pattern had reversed; that is, the incredible acceleration of the personal computer (PC) and server industries meant that commercial technology was leading rather than lagging behind military technology. This shift led to an increasing focus on the use and adaptation of commercial off-the-shelf technology in federal procurement and contributed to the steady decline in federal funding for R&D,8 given the U.S. preeminence in the area and the already high levels of research investments by the U.S. computer and semiconductor industry.
Today, cutting-edge R&D in semiconductors, the historical engine of computer performance growth, has become unmanageably expensive for the usual U.S. federal agencies. At the same time, it is extremely difficult for industry to invest in long-term R&D, given the near-term expectations of the financial markets. One consequence has been limited commercial R&D investment in hardware and software technologies whose economic return is not realized rather quickly.
In the United States, industrial policy has typically not been viewed as an offensive tool for economic competition or a means to create new industries or accelerate successful ones, but rather as a defensive tool to protect or restore existing industries under competitive economic pressure.9 The perception of favoring certain industries, “picking winners,” by government pressures and incentives, rather than allowing for natural market forces and laissez-faire investment, has been politically toxic. On the other hand, rescuing at least some foundering industries, or attempting to regain lost ground in critical ones, has been generally politically rewarding. The Asian competitor nations, for example, China and Japan, traditionally have both subsidized and protected (by legal and covert subsidies and tariffs) those industries that they choose to target.
In contrast, it is important to recognize that U.S. industries, and information technology in particular, do not tend to receive attention or assistance from federal sources simply because they are slowing down in growth or maturing; there typically must be a specific adversary. For example, once U.S. superiority in electronics and computation (e.g., for guidance systems) over the Soviet Union became assured, the focus of government policy switched to the rising Japanese dominance in electronics, especially including memories.
While countries such as Japan began forming R&D consortia as early as 1956, the practice was illegal in the United States until Congress passed the National Cooperative Research Act in 1984.10 Two years later, concerns of a U.S. decline in semiconductor market share prompted a call by the Semiconductor Research Corporation (SRC) and Semiconductor Industry Association (SIA)11 for increased cooperation to provide the U.S. semiconductor industry with the capability of regaining world leadership in semiconductor manufacturing. As a result of this effort, SEMATECH (Semiconductor Manufacturing Technology) was created in 1987 as a partnership of 14 U.S. semiconductor companies with the Defense Advanced Research Projects Agency (DARPA), which contributed U.S. $500
7While well established and interdependent, these value chains can be highly vulnerable to sudden disruptions from natural disasters, geopolitical conflicts, and so on. Some of these are discussed in greater detail in Chapter 4.
8J. Gansler, 2011, Democracy’s Arsenal: Creating a Twenty-First-Century Defense Industry, The MIT Press, Cambridge, MA.
9U.S. innovation policy can be thought of as “market conforming” in its intent to address problems that economists have deemed weaknesses for technological advancements. In particular, these were externality problems that required collective R&D funding and that funding took specific paths because of appropriation processes in Congress.
10D. V. Gibson, and E. M. Rogers, 1994, R&D Collaborations on Trial, Harvard Business School Press, Cambridge, MA.
11“Founded in 1977 by five microelectronics pioneers, SIA unites over 60 companies that account for 80 percent of the semiconductor production of this country.” (see www.sia-online.org) SIA, along with the European Semiconductor Industry Association (ESIA), the Japan Electronics and Information Technology Industries Association (JEITA), the Korea Semiconductor Industry Association (KSIA) and the Taiwan Semiconductor Industry Association (TSIA), sponsors the International Technology Roadmap for Semiconductors, a15-year assessment of the semiconductor industry’s future technology requirements (see www.public.itrs.net). Last accessed on June 30, 2012.