Comparative advantage, the idea put forward by economist David Ricardo in the early 19th Century, suggests each country should make what they are most efficient at. If all countries do this and trade with each other, it maximises welfare for everyone. This theory is the cornerstone argument for globalisation. Some countries have climates most suited to growing certain foods, an abundance of natural resources or a culture geared towards producing a specific good or service.
For the past 30 years, Taiwan has focused on producing semiconductors, with the help of government subsidies. Students are taught electrical engineering as early as high school and the industry is a source of national pride. Last year, semiconductors accounted for up to 20 per cent of Taiwan’s total GDP. As a similarly small island, the UK has also pursued a specialisation strategy, with 12 per cent of GDP coming from financial services.
For Taiwan, this made sense. It has few natural resources and imports almost all its oil. It needed to find a comparative advantage and chose one that was culture-based and could be developed regardless of geographical restrictions, much like the UK.
As ‘culture’ is difficult to measure, it is often dismissed in importance. However, a recent article from the technology publication Rest of World has shown the problems the US has had trying to adopt Taiwanese manufacturing culture. The detailed report refers to many specific tensions between TSMC and its new American employees as it tries to build its US government-backed plant in Arizona. The Taiwanese engineers are reportedly complaining that the Americans are lazier, have less technical ability and aren’t loyal to the company. Meanwhile, the Americans have complained the Taiwanese workers create a false sense of urgency, are unwilling to delegate work and mostly communicate in Mandarin.
The article hints that the Taiwanese working culture is backwards from an American perspective. There is an anecdote about a Taiwanese worker having a bikini-clad model as their screen saver and being asked to take it down by an American employee. Then it refers to how relieved some of the American employees feel after quitting their jobs at TSMC, with many saying they would like to join US rivals Intel.
However, although this article implies that TSMC is doing something wrong, the financial performance of the company over the past decade suggests otherwise. Since 2019, TSMC’s revenue has more than doubled with an operating margin consistently above 40 per cent. In the same period, Intel’s revenue dropped 25 per cent while last year it barely made any operating profit at all.
Culture might seem unimportant compared with natural resources, however, the reason why it is unmeasurable is the same reason why it is hard to replicate. The US is trying to reproduce decades of accumulated cultural values and, unlike buying natural resources, it is not obvious that more money can fix this problem. This is good news for TSMC shareholders, given that its main rival Intel has the majority of its manufacturing based in the US.
In the recent earnings call, TSMC chief executive C.C.Wei said customers would have to pay more if they wanted their semiconductors manufactured in the US. The company is happy to accept the billions of cash subsidies because ultimately it is not the one having to pay the higher price. It is customers, such as Apple, and ultimately US taxpayers and consumers. Public debt is expected to reach $45.7 trillion, or 114 per cent of GDP by 2033, up from 97 per cent at the end of 2023, according to the Congressional Budget Office.
A better strategy for the US would be to subsidise semiconductor manufacturing in Japan and South Korea, countries with a long history in electrical engineering, while also providing defence capabilities, something the US is culturally very good at, to deter China from invading.
Governments hardly have a good track record of picking industrial winners. But if the US insists it wants to boost its economy with fiscal stimulus it should spend on areas it has a comparative advantage in, such as AI, chip designing and weapons development. At least, that's what David Ricardo's 200-year-old theory would recommend.
