How robust are supply chains?

That is the topic of my latest Bloomberg column, here is one excerpt:

Consider the supply chain of the Apple iPhone, which stretches across dozens of companies and several continents. Such complex cross-national supply chains generate relatively high profits, giving them a kind of immunity to small disruptions. If there is an unexpected tax, tariff or exchange movement, the supply chain can generally swallow the costs and move on. Profits will be lower within the supply chain, but production will continue, as it is too lucrative to simply shut everything down.

Do not be deceived, however: Supply chains are not indestructible. If the new costs or risks are high enough, the entire structure will be dismantled. By their nature, supply chains do not fall apart slowly, because each part of the chain relies upon other parts to add its value. It does not help much to have the circuit components of the iPhone lined up, for instance, if you cannot also produce the glass screens. In this way, these supply chains are less robust under extreme conditions.

Global supply chains have yet to come apart mostly because trade and prosperity generally have been rising. But now, for the first time since World War II, the global economy faces the possibility of a true decoupling of many trade connections.

It is not sufficiently well understood how rapid that process could be. A complex international supply chain is fragile precisely for the same reasons it is valuable — namely, it is hard to construct and maintain because it involves so many interdependencies.

The nature of the cross-national supply chain makes it especially vulnerable to shocks coming from the coronavirus. These supply chains do not adapt so well to complete cutoffs in materials or labor, as may happen if Chinese coronavirus casualties continue and workplaces find it hard to operate effectively.

Imagine that closed Chinese factories cannot produce the components of many American medicines. It is not a question of the supply chain simply losing some profits; rather, some critical pieces of the production process are missing. The medicines won’t work without these inputs. The U.S. medical establishment might try to source those components elsewhere, but it isn’t easy for other suppliers to produce enough of them at sufficient scale and quality.

U.S. medical producers might try to bid more for the Chinese medicine components, but if the workers are prohibited from even showing up at the factory, no feasible market clearing price can make this arrangement work. Production just won’t be possible. Fashionable practices of near-zero inventories can make these shortages appear all the more rapidly. About 80% of the active pharmaceutical ingredients in U.S. medicines rely on Chinese or Indian components, so this does represent a very real public health risk for the U.S., even if the coronavirus itself does not.

You will note that when it comes to ex ante planning, companies do not in general internalize the costs of a supply chain cut-off to their customers, since consumer surplus for the infra-marginal buyers exceeds market price.  Supply chain are thus too fragile relative to an optimum, though that matters only under very limited circumstances, as we may be seeing right now.