On the margins of the G-20 gathering of world leaders on November 14, President Joe Biden affirmed that the United States would “compete vigorously” with China. America has its work cut out for it in the tech sector. China has spent years mobilizing national resources to control the world’s tech supply chains, and Washington is playing catch-up in the form of the recently-passed CHIPS and Science Act (hereafter the CHIPS Act). That legislation may financially incentivize a few large companies to establish more semiconductor factories at home, but winning the long-term battle for technology superiority requires catalyzing U.S. entrepreneurial energy through special manufacturing innovation zones (SMIZs).
Subsidies are not a panacea for winning the semiconductor race: they create an unlevel playing field that reduces innovation and competition. Nor do bureaucrats know how to funnel money to the right activities, and the CHIPS Act risks doing just that. Intel, which will likely be one of the recipients of the subsidies, could use the money to build fabs in Ohio for older-generation chips while it pays Taiwan Semiconductor Manufacturing Company (TSMC) to do higher-end work in Taiwan. TSMC, another likely recipient, could also start chips in its subsidized fab in Arizona that still have to be finished in Taiwan.
While such investments may reduce some risks, the United States would still be dependent on an island that manufactures 90 percent of the world’s most advanced chips, and that China has repeatedly threatened to take by force. And the U.S. would need to legislate additional subsidies if Washington hopes to attract future semiconductor investments or reshore additional hi-tech sectors, which are just as vulnerable to Chinese domination.
Facility construction, labor, and utilities are all more expensive inside the United States than in the places abroad that dominate production. The CHIPS Act aims to offset these cost disadvantages, but that prospect is unlikely. Indeed, clauses in the legislation such as those requiring firms to pay prevailing union wages for construction jobs will actually raise costs, undermining the bill’s intent. In April, Morris Chang, the founder of TSMC, called efforts to re-shore production “a very expensive exercise in futility,” given that that chips made in the United States are 50 percent more expensive than those made in Taiwan. The bill also does nothing to address another issue Chang raised: the difficulty of hiring high-tech manufacturing talent in America.
Across the Asia-Pacific
Get the Newsletter
A better approach than mammoth amounts of direct subsidies – or at least a complementary one, now that the CHIPS Act is law – would be to build on the lessons from one of China’s best devices for catalyzing business development: special economic zones. China initially established SEZs in 1980 to bypass the country’s own rigid, statist governance system, which could not attract foreign investment. SEZs boast favorable geography (near ports across from Hong Kong and Taiwan), excellent physical infrastructure, business-friendly legal and regulatory frameworks, and procedures to fast-track permitting and construction. Ample tax incentives have helped lure companies. The most successful SEZs, such as Shenzhen, have also actively recruited the talent companies need.
Enjoying this article? Click here to subscribe for full access. Just $5 a month.
Consequently, SEZs have become one of the key engines of China’s economic growth. Today there are hundreds of such zones in China promoting innovation, commercialization, and technological development, and they have generated roughly one-fifth of China’s GDP, three-fifths of its exports, and more than 30 million jobs.
The closest U.S. approximations to SEZs are Foreign-Trade Zones (FTZs), which are approved by a federal board and run by public or private corporations. By sheltering investors from federal customs duties and excise taxes, they incentivize greater private investment. There are 195 active FTZs operating either within a defined area (typically near a port of entry) or a specific facility, with 3,400 companies using them to export close to $100 billion worth of goods in 2020. Instead of doling out subsidies directly to firms, the U.S. government should take a different page from the Chinese playbook and adapt the FTZ program to create Special Manufacturing Innovation Zones (SMIZs).
SMIZs would focus squarely on developing industrial hubs in the key tech sectors where the United States currently lags. They would function much like Chinese SEZs by offering tax credits, rebates against construction and labor costs, exemptions from certain regulations (and fast-tracking permitting processes for the rest), and special visas for high-tech workers.
Offering these benefits to any qualifying firm, regardless of size, is a more efficient means of inducing reductions in cost than government handouts for behemoth chip manufacturers such as Intel. And by being strategically located to leverage existing infrastructure, talent pools, and livable neighborhoods – think of places such as Madison, Wisconsin – they would have wind at their back from Day 1. A separate program could offer competitive grants to small companies launching new technological innovations into high-scale production, something the U.S. has notoriously struggled with.
The presence of these zones would catalyze states to both compete for them and then work hard to ensure that each zone succeeded. It would also provide the type of level playing field that enables entrepreneurs and small companies to challenge tech giants. If success were achieved in a few sectors, the program could be expanded to additional technological fields deemed important for national security.
While industrial policy can work in some cases, the United States government is not well equipped to deliver in the way the CHIPS Act conceives of. Creating the conditions for markets and customers to drive innovation is a far better way to cultivate domestic production of key technologies such as chips than one-off investments from Washington.