Semiconductors and Modern Defense Spending
September 8, 2020
A New Competition
The United States must again confront authoritarian powers, but the nature of conflict has changed. Now, technological leadership is one of the most important areas for competition. U.S. opponents learned from the United States that technological leadership provides countries with influence, power, and authority in the international environment and that it is crucial for military strength. They seek the leadership that we have held. The United States is in a new kind of conflict where technology and ideas will play as big a role as militaries in creating national power. We need to adjust to this new and different kind of conflict.
China is our chief competitor. It has one clear advantage over us in this conflict. It is willing to spend money. China spends on its military, but it is not just spending on weapons. It spends on research and on its companies. This is the new space for competition. China, unfortunately, is not the sluggish Soviet Union with its turgid economic planning. It has found a way to introduce a degree of market dynamism into its state-controlled economy. But even if it was the Soviet Union, we would still be at a disadvantage. The United States no longer has the federal investments in technology needed for competition, having slashed technology budgets since the end of the Cold War.
China is investing heavily in semiconductor fabrication plants (fabs) and manufacturing equipment, hiring engineers away from companies like Taiwan Semiconductor Manufacturing Company and, of course, engaging in espionage to bolster its chip-making efforts. China’s goals are leadership in semiconductor production and ending its reliance on U.S. suppliers. These efforts are supported by more than $58 billion in government semiconductor investment funds, accompanied by pledges of another $60 billion in semiconductor funds created by local governments and include a 10-year corporate tax exemption for chipmakers producing advanced chips.
Other policies aim to strengthen the chip workforce, expand R&D, and incentivize foreign chip companies to relocate to China. China has become the world’s largest purchaser of semiconductor manufacturing equipment (according to SEMI), and China plans to double its capacity in the next five years. Realistically, it will take China more than a decade for success in memory chips and wafers, and perhaps longer if China no longer has the same levels of access to foreign technology. The issue, however, is not only preventing China from catching up but keeping the United States in the lead.
There is growing recognition of the competition with China in key technologies, but for semiconductors, the measures the United States has taken so far are defensive. Measures to block exports of semiconductors manufacturing equipment, restrict sales of chips made using U.S. equipment, and block Chinese acquisitions of U.S. chipmakers work to preserve the status quo, for now. They slow China’s growth, but they do not provide for a strong U.S. semiconductor industry. Doing this requires increased federal investment to accompany policies oriented toward ensuring access to the global market.
Investment for National Defense
Federal investment in strategic industries is no longer a taboo concept. Let’s start by noting the United States used industrial policy to win in each of its global conflicts. Not only was industrial policy the key to helping win those conflicts, but the technology base built in World War II and later expanded tremendously for the Cold War is the foundation of our national wealth today. Silicon Valley was built on that federal investment, but those investments were made decades ago. We are coasting on spending in science, research, and technology that date to the 1980s. To defend ourselves against hostile powers, federal investment is again necessary.
The original language of the Creating Helpful Incentives to Produce Semiconductors (CHIPS) for America Act (now merged in the National Defense Authorization Act with the American Foundries Act) provided strong support for the semiconductor industry, including $3 billion of R&D funding for the National Science Foundation (NSF), $2 billion for the Department of Energy, and $2 billion for the Defense Advanced Research Projects Agency. It created a National Semiconductor Technology Center to conduct research with the private sector, with a recommended budget of $3 billion over 10 years. It proposed a $10 billion trust fund to match state and local incentives, created an Advanced Packaging National Manufacturing Institute at the Department of Commerce, and authorized a semiconductor research program at the National Institute of Standards and Technology. Finally, it called for tax credits for qualified semiconductor equipment or manufacturing facility expenditures through 2027. Looking at the levels of spending in this industry, fully funding the CHIPS Act will create incentives, competitive advantage, and support national security.
This sounds like a lot of money, but money should not be a problem. The United States is much richer than China and has always been willing to spend on defense. And investment in technology is a new requirement for defense. The United States was able to spend (by some estimates) $6 trillion over the last 18 years in Iraq and Afghanistan, or over $330 billion annually. In contrast, China spends on building its technology base. It devotes huge sums to the pursuit of technological supremacy. China learned from the United States about the importance of technology for national power. The United States needs to now spend similar amounts in the strategic competition with China. Even investing 10 percent of the amount we spent in Iraq and Afghanistan on R&D and strategic industries would strengthen our ability to compete with China.
Instead, Congress has pared back funding significantly. A good way to think about spending on semiconductors is to compare it to other defense expenditures. A fab costs almost as much as a new aircraft carrier. Of course, there are crucial differences. Aircraft carriers do not provide direct commercial benefit by enabling a range of other key industries. Nor is the cost of a fab borne solely by the government. The United States spends more than other countries on weapons, but if it does not learn to spend more on technology, this kind of defense spending will be a declining source of advantage.
Federal investment is crucial. It is unrealistic to expect to compete with China without spending money. In some technology areas, such as semiconductors, China’s announced plans to outspend the United States by almost 50 to 1. China’s government does not have to worry about voters and elections, but a strong case can be made to U.S. voters that new investments will provide both security and economic benefits that outweigh the costs. Investment will also partially compensate for the revenue lost because of new 5G-related sanctions on China. This is not the time to balk at supporting the semiconductor sector, as companies say the lower revenue damages their ability to spend on R&D
Defense Production Act (DPA) funds are not a realistic substitute. Currently, there is about $228 million available for DPA spending. If all of this went to semiconductor incentives (and some will go to Covid-19 responses), that would mean China would be outspending the United States 20 to 1 on this strategic industry.
Semiconductors should be the focal point for this new kind of defense spending. They are a “foundational technology.” We do not want to lose this advantage. Semiconductors are the backbone of economic and military performance in the digital age. The Chinese know this, which is why they have been willing to spend billions of dollars for decades to build their own industry. Semiconductors power our economy and our weapons, but there is now a real risk that the United States will lose its leadership in this foundational technology. Funding the CHIPS Act will compensate for revenues lost to new export sanctions on China and help maintain strength in a foundational technology.
One reason for this is that we invest far less now than during the golden age of U.S. technological leadership. According to statistics compiled by the NSF, the United States spent $83.4 billion on basic and applied science and technology research in 2019. This is about 1.9 percent of all federal spending and 0.4 percent of U.S. GDP. NSF, which funds “hard” science and engineering, received total research funding of $6.74 billion, 0.1 percent of the federal budget and less than 20 percent of the budget of the National Institutes of Health. In 2018, U.S. companies spent $28 billion on basic research and $314 billion for development. In contrast, China’s research and development spending reached $554 billion in 2018—more than that of Japan, Germany, South Korea, France, and the United Kingdom combined.
There is also the competition among nations to attract semiconductor fabs to their countries. This affects both “reshoring” (e.g., moving supply chains back to the United States) and tech leadership. China spends the most, but it is not the only government to do so. Few fabs are built today without some form of public support, and there is fierce competition among a dozen nations to attract the next fab. The Organisation for Economic Co-operation and Development estimates that semiconductor firms have received more than $50 billion in government support between 2014 and 2018. Countries see chip production as a priority and provide incentives to attract semiconductor manufacturers, including grants, equity investments, infrastructure support, and tax credits. These subsidies mean it is more expensive to build a fab in the United States than in other countries, creating an incentive structure that disadvantages it. Some argue that the U.S. chip industry does not need federal support. This would be true in a perfect world where no other country used subsidies. Nor does China have any intention to end its subsidies.
The CHIPS Act and U.S. Security
This leads us to the CHIPS Act. CHIPS could be a significant step forward for national security if it is adequately funded. It would provide incentives for investment in semiconductors in the United States and better position it to compete. While the bill contains a number of provisions to support the industry, the simplest to implement would be to provide tax credits for investments by semiconductor companies. Some companies would like these tax credits to be “refundable,” but Congress has often been unwilling to approve refundable tax credits in the past.
We have already learned the cost of not intervening to protect a strategic industry. U.S. companies once led in building telecommunications infrastructure. Now, there are none left. While the situation is not as dark for semiconductors, U.S. companies will lose market share and revenue as a result of sanctions. This could provide an advantage to foreign suppliers and will further incentivize China to build its own chips. The end result will be an increased reliance on foreign suppliers for a critical technology. A fully funded CHIPS Act can avert this.
Incorporating language from CHIPS, the National Defense Authorization Act (NDAA) would bolster domestic semiconductor manufacturing using Commerce Department-administered grants of up to $3 billion each for facilities construction and modernization. The bill creates a public-private National Semiconductor Technology Center for research, with $914 million in proposed funding. Additionally, the National Science Foundation would receive $300 million and the National Institute of Standards and Technology $50 million for semiconductor research. The Department of Defense would create a consortium of U.S. companies to develop “advanced, measurably secure” microelectronics. Treasury would administer a “Multilateral Microelectronics Security Fund” to finance international partnerships for developing chips. This adds up to about one-quarter of the cost of a new aircraft carrier.
Spending for Twenty-first Century Security
The United States has always been willing to spend what is needed for defense. Investment in technology is now a requirement for national defense and security in the conflict in which we now find ourselves. Federal investment in the semiconductor industry provides both economic and military benefits. The CHIPS Bill offers several funding mechanisms. The easiest to implement would be tax incentives to support a strategic industry. Passing the CHIPS bill as part of the NDAA was a milestone in recognizing the requirements of this new conflict, but failing to fully fund it yields the advantage back to China.
James Andrew Lewis is a senior vice president and director of the Technology Policy Program at the Center for Strategic and International Studies in Washington, D.C.
Commentary is produced by the Center for Strategic and International Studies (CSIS), a private, tax-exempt institution focusing on international public policy issues. Its research is nonpartisan and nonproprietary. CSIS does not take specific policy positions. Accordingly, all views, positions, and conclusions expressed in this publication should be understood to be solely those of the author(s).
© 2020 by the Center for Strategic and International Studies. All rights reserved.