Making semiconductors locally sounds reassuring at a time of global supply chain challenges. But in the modern world of complex and interdependent manufacturing it’s an empty promise.
As the old aphorism goes, governments are terrible at picking winners, but losers are great at picking governments. It’s advice which many governments seem to have forgotten as they start using taxpayers’ money to build industries that are deemed to be vital.
The latest vital manufacturing industry supposedly in need of government help to build “sovereign capability” is semiconductors. Semiconductors are an essential component in electronic devices, from communications, computing and cars through to military systems, healthcare and clean energy.
COVID-19 wreaked havoc on global semiconductor supply chains. And with China as a growing major global supplier of semiconductors, and production concentrated in North-East Asia, the United States has taken the brave step of moving to subsidise its semiconductor industry – thus “Japanising its industrial policy”, the Financial Times has said. Some advise US partners in Asia, like Australia, Japan and other countries to now do the same.
There’s a major problem with this strategy: it won’t work. Sovereign capability in semiconductors is a largely empty promise.
There are four fallacies that undermine the plausibility of the sovereign capability idea.
The first is the “made locally” fallacy. The notion that our supply chain challenges are solely the result of international trade does not stack up. This fallacy was exposed during COVID-19. Plenty of the things people have struggled to buy during the pandemic were made locally. Making things locally does not guarantee a resilient supply chain.
The second fallacy is the input fallacy. Things we think we “produce” locally actually include many inputs that come from overseas.
Take the simple surgical face mask as an example. Face masks have at least four components: the medical grade fabric, the elastic, the thin metal strip that bends over your nose and the stitching that holds it together. Having genuine sovereign capability in the production of face masks means being able to produce all four of these things domestically, and then assemble them domestically, too.
And that’s just a simple face mask. A ventilator has more than 240 components, all of which have their own array of inputs as well. To be sovereignly capable, a country needs to make every single one of them.
The third fallacy is the “just in case” fallacy. A new catchphrase of politicians is that we need to move away from “just in time” business models towards “just in case”. Requiring businesses to hold buffer stocks, they argue, is a great way to make our supply chains more resilient.
It’s also a way to make countries poor. Business owners will tell you that stockpiling inventory costs money. Requiring local businesses to hold buffers in case a once-in-a-100-year pandemic strikes would make those businesses uncompetitive.
The bigger question for “just in case” supply chains is: just in case of what? If local cafes need to prepare for a once in 100-year pandemic, why not the zombie apocalypse, World War III, and world-destroying asteroids? Where do we draw the line?
The fourth fallacy is the “essential business” fallacy. The idea of having sovereign capability in producing essential goods and services assumes we know what an “essential” good or service is. Many governments around the world have had three years to agree domestically on what constitutes an “essential business”. They still don’t have an answer.
What constitutes “essential” depends on the crisis. Hand sanitiser is essential during a pandemic, but pretty low on the list during World War III. And what about inputs? Many governments designated PPE as essential during COVID-19, but not the components used to make it. Governments designated ventilators as essential, but not the machines used to treat sleep apnoea, even though it’s basically the same machine.
“Sovereign capability” is not a serious solution. But supply chain resilience is a serious problem. So, what can be done about it?
First, we need to understand when a supply chain is at risk and when it is not. Whether a supply chain is at risk depends on how many alternative suppliers are available for a particular good or service and how easily businesses in one industry can substitute over and start producing that good or service (think: sleep apnoea machines being switched to ventilators, or gin distilleries making hand sanitiser). It has nothing to do with exposure to international trade.
Second, and related to this, we need to get smarter with data. Understanding all the components needed to make something – and all the components needed to make those components – is a tough job when there are millions of products that are constantly changing. Modern datasets make this easier. Bank transaction data – combined with a proper system of classifying what businesses actually produce – allows us to build a picture of a country’s production capacity. It tells us who makes what, how much they make, and how many alternative suppliers are available.
Third, we need to strengthen our connections as a country. Diversification can involve re-globalisation, as WTO director general Ngozi Okonjo-Iweala calls it, that brings more countries into supply chains. A cost-effective way to reduce risk across supply chains is to form more robust multilateral partnerships.
Robust multilateral partnerships may be easier said than done with strategic competition in full vogue. The US-led Chip 4 alliance with Japan, Taiwan and South Korea could help secure semiconductor supply chains, but US efforts to cut China out of semiconductor supply chains will not be met with the same enthusiasm by all members. South Korea and China recently agreed to their own semiconductor agreement, unsurprising with 40 per cent of South Korea’s chips sold to China. The Indo-Pacific Economic Framework (IPEF) will be tested if it seeks supply chain resilience from China given almost all the 14 IPEF members have China as their largest trading partner.
Many of the world’s supply chain challenges came from inward-looking protectionist policies. Vaccine hoarding during COVID-19 is a significant example. The G20 countries need to recommit to their promise during the global financial crisis to resist trade protectionist measures. And regional arrangements such as RCEP and CPTPP are set to play a role.
The solution to the world’s supply chain challenges will be found in data, reducing barriers to entry for new firms, boosting competition and pushing back trade protectionism. Sovereign capability sounds reassuring but in the modern world of complex and interdependent manufacturing it’s based on empty promises.
Adam Triggs is senior research manager at the e61 Institute and visiting fellow at the Crawford School of Public Policy at the ANU. Shiro Armstrong is director of the East Asian bureau of economic research at the Crawford School, ANU and Visiting Fellow at RIETI.