Restricted from buying some semiconductors from the outside by the U.S. and unable to purchase critical semiconductor equipment as well, self-sufficiency has taken on a new urgency for China, a country that imports about 70 percent of the semiconductors produced in the world and mainly re-exports these as finished electronic products.

To combat the problem, the Chinese government has engaged in subsidies and tax incentives in partnership with domestic Chinese semiconductor companies.  The Big Fund, which totalled about $22 B in Phase I in 2014 by some estimates, now has a second phase estimated at around $29 billion of investment in semiconductors.  The big fund consists of subsidies and direct equity investments by the government.  The fund does not include local subsidies and investments by municipalities, estimated to exceed even the Big Fund.  

Most economists in the U.S. remain skeptical about top-down government planning to promote industrial development.  Indeed, China’s has had several corporate semiconductor failures in the past.  Fraud by the founders may have been a factor, but lack of expertise and making unrealistic promises to investors played a significant role.

Economists in the U.S. mention these failures, but mainly point to Japan’s stagnant economy and France’s failure to create a vibrant technology sector as examples of poor outcomes from top-down industrial development in semiconductors.  To make matters worse for top-down development in the U.S, the Government often shapes policy with cronyism and favoratism in mind.

Instead of government planning, the U.S. has relied mainly on a free-market approach in semiconductors favoring start-ups and innovations led by companies with big R&D budgets.  As an example of start-ups, in 2003, Strategy Analytics listed more than 100 Wi-Fi chip start-ups.  Wi-Fi was pretty new and only about 43 million Wi-Fi devices shipped that year globally.  Most of the start-ups had similar business plans focused on the design and marketing of CMOS radio SoCs for Wi-Fi.  At the time, these SoCs did not yet exist outside of the laboratory.

Atheros and Airgo succeeded, and Qualcomm acquired both.  Broadcom acquired two companies, which helped Broadcom take the leading share position from Intersil.  About 95 Wi-Fi chip companies quietly went out of business, for the most part selling their IP for pennies on the dollar, and as far as I know nobody got accused of trying to scam investors.  This is typical in the U.S. and the EU, where something like 19 out of 20 semiconductor start-ups fail, with the successes making up for the failures for investors.

In retrospect, the free market approach has encouraged specialized semiconductor companies to form and move production to lower-labor cost countries, while encouraging mergers and acquisitions.  This has led to concentration of the semiconductor supply chain into relatively few players located in a handful of different countries.  Trade disruptions among these countries have contributed to the present global semiconductor shortage.

While central planning is notoriously difficult to implement and often impractical, a completely laissez-faire approach such as traditionally favored by the U.S. is tantamount to having no industrial policy whatsoever and doing nothing to promote economic growth.   And even if a country’s industrial policy does not inhibit its competitiveness, more than just industrial policy determines competitiveness; a country’s education and social systems also play key roles in determining the productivity of the labor supply.

In support of more U.S. Government support for technology businesses including semiconductors, the newly-appointed U.S. Trade Representative Katherine Tai said at her confirmation hearings “. . . recent years [have] taught that we need to revisit how we conduct our economic activity . . . not to become China, but to be true to ourselves and our traditions and be more strategic, knowing the quantity and the strategy and ambition that we are up against.”

A report by the Boston Consulting Group (BCG) and the Semiconductor Industry Association (SIA) came to the conclusion that fully “self-sufficient” local supply chains in each nation able to meet their current levels of semiconductor consumption would require at least US$1 trillion in upfront investment, raising the cost of finished ICs by as much as 65 per cent, and ultimately leading to higher costs for electronic products for consumers.

China seems to have hit upon a good mix of targeted industrial policy and free market incentives to raise their own self-sufficiency.  Full self-sufficiency may prove impossible, but China has a good start on reducing imports of semiconductors as discussed in the complimentary report China Well Positioned to Attain Semiconductor Self-Sufficiency.

You can watch a video interview featuring Sravan Kundojalla and me talking about this topic by clicking here.