Semiconductor Subsidy Mischief
In the coming weeks, Congress could pass a bill showering massive subsidies on the semiconductor industry. The supposed purpose of the plan is to induce firms to produce more semiconductors domestically with an eye toward outcompeting China. My Cato colleague Scott Lincicome noted that politics, not economics, is the driving factor behind the push for this corporate welfare scheme. Meanwhile, Bryan Riley of the National Taxpayers Union highlighted the enormous costs associated with the bill. The American Action Forum has noted that the private sector is already working to address the global chip shortage – announcing more than $830 billion worth of private investment over the next decade. Largely missing from the current debate is the strong potential for future trade frictions as countries try to protect their domestic semiconductor industries from subsidized import competition.
The subsidies under consideration—grants and tax credits for domestic production—are classic examples of specific subsidies under globally accepted definitions: “financial contribution[s]” by a “government” or a “public body” conferring “benefit[s]” on the recipients. Indeed, the World Trade Organization’s Agreement on Subsidies and Countervailing Measures (SCM) explicitly mentions grants in Article 1.1 (a)(1)(i) and tax credits in Article 1.1 (a)(1)(ii) as prohibited trade‐distorting subsidies.
Likewise, the U.S. is already a major exporter of semiconductors. As Bryan Riley noted, in 2021, the U.S. exported about $11 billion worth of semiconductors to China; $4 billion to Taiwan, $3 billion to South Korea and $2 billion to Germany. Meanwhile, according to the Semiconductor Industry Association, the United States had a 47% share of the global chips market in 2020. Korea had a 20% share of the global market; while the European Union had a 10% share; Taiwan 7% and China 5%. With the exception of China, these other nations are strong allies of the U.S. In other words, U.S. policymakers are going to heavily subsidize the domestic semiconductor industry—likely leading to a glut of chips—and then the subsidized products will be exported to closely‐aligned, major semiconductor‐producing nations—many of the same countries the U.S. should be partnering with to pressure Beijing to raise its international trade and investment standards.
Given the strategic importance placed on semiconductor production by policymakers around the globe, the U.S. subsidies are ripe for a long, drawn out trade battle with longstanding allies. Even non‐trade watchers are probably familiar with the decades‐long skirmish between the U.S. and Canada over softwood lumber and the dispute between the U.S. and E.U. over subsidies for Boeing and Airbus. Semiconductor subsidies have the potential to make those fights look rather quaint.
Simply put, there are better ways to boost the domestic semiconductor industry than heavy subsidies, which Scott Lincicome recently laid out in a Cato blog post, including expanding immigration, improving the tax treatment of capital investments, removing existing tariffs, pursuing more trade agreements, and reforming export control laws. High tech, 21st century industrial policy is still just plain vanilla industrial policy, which has an ignominious history. The semiconductor subsidies under consideration are expensive, will create incentives for rent‐seeking behavior for firms, distort trade flows and engender frictions between allies. Policymakers can do better.