Reducing CHIP Supply Risks

When semiconductors where invented in the U.S. in the late 1950 and began to replace vacuum tubes in electronic circuits, the world of electronic circuitry changed dramatically forever (from the computer I am composing this note on, to the mobile phone on which I might discuss it with you or the electronics in the car I might drive to meet you while listening to the radio these chips made possible, not to mention of submarines, planes, and satellites they empower).

American companies continue to lead the world in the design of the most sophisticate chips and semiconductor circuits. However, they have increasingly found it more economical to outsource their manufacture to facilities in foreign countries. Almost all the most sophisticated chips (still designed in the US) are now produced in Taiwan by TSMC, which produces about 56% of world output of semiconductors.  Especially given the increasing suggestion, even by the President of the United States, that the U.S. might renounce its acceptance of China’s claim to Taiwan, such reliance on TSMC for our most advanced chips is an economic and security risk we should reduce.

What is the best way to reduce the risk of our heavy dependance on Taiwan’s supply of such Chips? The rest of this note briefly compares the market approach with the government (socialist/industrial planning) approach to reducing that risk.

American and other firms concentrated the manufacturing of the chips they designed and/or needed where it was cheapest to produce (and deliver) them. Where China violated the WTO rules of fair trade via state subsidies, importing countries are allowed by WTO rules to impose tariffs at levels designed to neutralize such artificial advantages. WTO rules also allow the use of tariffs to diminish the risk to national security of dependance on foreign supplies.

The opposite approach is for a government to subsidize the otherwise uneconomical manufacture of semiconductors (or whatever) in their own country.  In the U.S., the CHIPS and Science Act of 2022 appropriated $280 billion in part to subsidize factories to produce such chips in the U.S. Why was so much needed to get firms to produce chips in the U.S.? “’It’s much cheaper to build the chips and the factories in Taiwan than it is in the United States,’ former Google CEO Eric Schmidt told Semafor. ‘Similarly, the workforce quality is not as good as it is in Taiwan.’” “Chip war-US-Taiwan”

The Biden administration’s industrial policy approach suffers all the well-known disadvantages of industrial policy. First, like China’s subsidies, it violates WTO trading rules, which the U.S. seems all too willing to do when it is the violator rather than someone else. Second, it, rather than market factors, must decide who gets the subsidies (and tax breaks), either by establishing the rules for access or by outright picking winners. Governments’ records at picking winners, especially picking technologies, have historically been poor compared with the search for profit by entrepreneurs, most of whom fail and quietly fade away without further cost or waste. Third, when governments pick winners, they establish an economic incentive for corruption by those seeking to be “picked”. Governments, like everyone else, tend to bend to such temptations.

Rather than paying hundreds of billions of American taxpayer’s money for more costly Made in American products, imposing tariffs on imported chips sufficient to reflect the existing sole source risk would leave it to the market to find the best alternative and more diversified sources (India, Korea, Japan, Viet Nam, etc.). The full cost of lower living standards from industrial policies will only be felt in the longer run. “The slippery slope”

The same economic forces and arguments apply to slowing or preventing further global warming. A carbon tax reflecting the global warming externality of carbon producing activities leaves to the market the search for the best technologies for reducing carbon emissions without loss or with minimal loss of output.