Strategies to secure America's supply chains
A guest post by Yann Calvó López and Ben Golub.
This is the second part of a guest post by Yann Calvó López and Ben Golub about the danger facing America’s supply chains. The first part is here:
Calvó López is a Research Associate at Northwestern University. Golub is Professor of Economics at Northwestern whose research focuses on social and economic networks.
In the first part of the post, the authors identified the dangers facing America’s supply chains; in Part 2, they discuss solutions.
1. The Aggregate Advantage: Self-Organizing Robustness
Economies are far more adaptable than individual companies—they have more substitution possibilities.
As Russian tanks rolled into Ukraine in early 2022, Germany faced a stark dilemma. The industrial powerhouse of Europe relied on Russia for 55% of its gas supply and one-third of its total energy needs. Severing ties with the aggressor “would mean the destruction of the entire German economy," warned the CEO of BASF, the world's largest chemical producer, as he predicted "the worst crisis since World War II.”
The last post in this series discussed the ways in which the supply networks underlying modern economies can break down: concentrated dependence on an upstream supplier like Russia; the complexity trap of economy-wide gridlock when most of the nodes and links in a large system are simultaneously hurt by, say, spiking energy prices. During the crisis triggered by Russia’s war, German business leaders were lobbying the government aggressively to take these arguments seriously in forging its foreign policy.
They were successful. But, in August 2022, despite Germany’s forbearance, Moscow turned off the taps, closing the critical Nord Stream 1 pipeline.
The expected economic apocalypse? It never came. Germany's economy actually grew by nearly 2% in 2022. Winter brought only a "technical mini-recession".
How did Germany defy forecasts of economic doom? The answer lies in a textbook macroeconomic idea whose origins go as far as 1955: Economies are far more adaptable than individual companies—they have more substitution possibilities.1 Think of the economy as a forest ecosystem: while a single tree species might struggle with changing conditions, the forest as a whole can adapt and thrive.
By reviving and quantifying this insight, a team of leading economists accurately predicted that the overall impact on Germany of cutting ties with Russia would be small. If some firms could substitute Russian imports in their production, even imperfectly, they could provide a route for the German economy as a whole to go around the missing Russian gas.2
While some heavily gas-dependent firms struggled, others found creative solutions. Munich’s energy supplier, for example, converted two heating plants to run on oil rather than gas. Others, such as the screw manufacturer Würth, switched to electricity. The car manufacturer Audi instead resorted to reducing some of its near-term gas consumption by turning the heating down in its offices. Even the glass industry, where substitution was claimed to be impossible by lobbyists, found a way forward: glass manufacturer Wiegand Glass switched to heating its melting tanks with light fuels instead of natural gas.3
Some firms failed to adapt. However, just as a forest can thrive despite losing individual trees, the German economy remained robust even as some firms faltered. The less adaptable businesses were replaced in other firms’ sourcing by more agile counterparts, or by foreign suppliers. For instance, as domestic production of gas-intensive chemicals such as ammonia plummeted due to energy scarcity, chemical imports soared—Germany relied on the ability of faraway producers to get gas from sources unrelated to Russia. Such substitution via imports allowed sectors such as German fertilizer production—heavily reliant on gas-intensive inputs—to remain surprisingly stable.
Germany's response to the gas shock demonstrates how successful decentralized adaptation can be. Despite the vulnerabilities in modern supply networks, these networks can turn out to be strikingly adaptable in the medium run. On the other hand, the widespread and persistent illnesses of supply chains during the pandemic showed that such resilience is by no means guaranteed, and that disruptions can be surprisingly persistent and painful.
So the question that we should ask is not “Are supply networks adaptable?” Rather, we should wonder what makes them agile or brittle, and how economic policy can affect which outcome we reach.
2. The Economic Fabric
Supply networks do not form in a vacuum. Companies live in a business environment which allows their complex supply networks to rewire in response to disruptions—or not.
This business environment requires infrastructure, investment, and coordination. To make this point, it is useful to introduce one more metaphor for supply networks, building on the ones we have already discussed. Let’s think of the links between firms as forming a fabric or tissue that connects the economy.
This fabric is a multifaceted object—much like the tissues in our bodies that contain muscle fibers, nerves, and blood vessels. We’ve been using roads as a loose analogy for supply networks. Actual roads are themselves part of this fabric. More broadly, infrastructure for transportation and logistics, essential for economies to prosper, is a part of the economic fabric.4 And there are a lot of other connections, too: telecommunication technologies, for instance, allow businesses to connect and coordinate cheaply and effectively across great distances. Less visible forms of connection include standard formats for engineering plans and financial data, which permit firms to work together effectively.
There is also a third layer—the least visible, but perhaps the most important, of all. A strong trade link between two firms doesn’t just involve well-oiled logistics, good telecommunications, and interoperable technical standards. It also involves trust and a good understanding of one’s trading partner.
Businesses often have to deal with the unexpected. For instance, consider a firm whose plans to get its goods to market for the holiday season are scrambled by the threat of the shipping strike currently looming over most of the ports in the US. With good foresight, this firm can protect itself to some extent by rerouting shipments through West Coast ports. But to achieve this, the distributors who import the goods will have to call in favors from overwhelmed shipping partners, find room in their warehouses earlier than expected, and make costly arrangements for unusual overland shipping routes. Late-night phone calls will be needed to juggle the many details of the new flow of goods. In the many arrangements they coordinate during this process, the businesses involved will have to trust that they will get the job done together and treat each other fairly—rather than, say, ditching their partner for a more attractive price or shirking on their part of the work needed to adapt. Relationships and understandings among executives, managers, and workers are a vitally important dimension of the tissue of the world economy that makes this kind of adaptation possible.5
3. Institutions and the Wealth of Nations
All the ingredients of the tissue connecting the economy are underpinned by what economists and political scientists have come to call institutions—the rules of the game in a society.
All the dimensions of the tissue connecting the economy are underpinned by what economists and political scientists have come to call institutions. These are “rules of the game” in a society—the laws, norms, and practices facilitating economic transactions and other joint actions. The ways in which government supports commerce are institutions—for example, the formal and informal rules governing how goods go through customs. Good institutions also include things like common understandings in a place or industry about business agreements; courts that hear cases promptly and resolve them consistently; and a political environment in which the powerful do not meddle arbitrarily with exchange.
Institutions are now understood to be a critical factor explaining the divergent outcomes of different societies at a very large scale. Ever since the insights of Douglass North, economists have explored institutions as a deep cause of the differences among societies. Using evidence from the different paths of European colonies, Acemoglu, Johnson, and Robinson (2001) famously argued that differences in the security of property rights—a key economic institution—could explain almost the entire gap in per capita income between countries like Nigeria and Chile.6 In the absence of secure property rights, individuals have poor incentives to invest in land, capital and technology—e.g., due to fears of expropriation. This ultimately results in fewer productive enterprises and lower income levels.
An especially vivid example comes from comparing North and South Korea: two countries that, until their partition, shared the same history, culture, genes, and so on. Their profoundly divergent experiences after the Korean War were undeniably a consequence mainly of different economic and political institutions being imposed in the different countries.7
The importance of institutions also plays out at the level of relationships among businesses. When businesses can trust that their contracts will be honored, that disputes can be resolved fairly, and that the laws governing their trade are predictable, they are more likely to engage in transactions and invest in long-term partnerships. These conditions seem important in creating relationships capable of adapting to shocks.
These are not just theoretical ideas. Researchers have recently been able to zoom in to the level of individual companies to show how one specific type of institution directly affects firms’ networks.
Somewhere in Karnataka, India, the owner of a shirt factory is hiring workers to painfully and slowly hand-embroider designs onto shirts at a wage of about three dollars a day, though machines exist that can do the same work more cheaply and much faster. This choice has its roots in an unexpected place: the crowded hallways of the local courthouse. The shirt company would much rather outsource the work—necessary in only a few of its products—to a specialized supplier that does mechanized embroidery. But in a region where court cases can drag on for years, an outsourcing decision carries hidden risks for the buyer. What if a supplier damages partly finished garments through poor storage practices? What if it demands that the shirtmaker take late delivery, unless it is willing to pay extra? From the supplier’s perspective, there are similar worries: what if the buyer pays for goods much later than is customary, or sends items that jam the machines? With luck, two businesses with an ongoing relationship can deal fairly with each other and resolve disputes. But such relationships benefit profoundly from the backstop of legal enforcement. With courts that never get around to hearing a case, that backstop isn’t available.
The takeaway is that when courts work efficiently, they don't just resolve disputes—they create an environment where businesses can confidently forge partnerships, specialize, and innovate.
So the shirt company chooses the safer, if costlier, route: in-house production. It doesn’t make sense to buy and maintain a $100,000 machine that will operate only one day of the week, so the embroidery is done by hand. It's not the most efficient solution, but it's the one that keeps things under the company’s control.
As shown in a landmark paper by researchers Johannes Boehm and Ezra Oberfield, this pattern repeats across industries and regions. Using plant-level data from India’s Annual Survey of Industries, the researchers showed that, in Indian states where courts are most congested and slow-moving, companies systematically make decisions that lead to higher production costs. Many do this just like our shirt factory, pulling more production in-house and using inefficient processes, even when they would prefer to outsource; another strategy is to rely on family-owned businesses for supplies, even when those are far from the most efficient options.8
Boehm and Oberfield’s insightful study makes it clear that institutions play a large behind-the-scenes role in supporting reliable, productive relationships. The takeaway is that when courts work efficiently, they don't just resolve disputes—they create an environment where businesses can confidently forge partnerships, specialize, and innovate.
The impact of institutions on business relationships becomes intuitive—at least for those who have lived in both the developed and the developing world—in cross-country comparisons, even those less dramatic than North versus South Korea. In many countries, accomplishing any complex task feels like a minor miracle. In other nations, it can be expected, though rarely guaranteed. Private enterprises can adapt flexibly while being able to count on key public goods—ranging from well-functioning roads to reliable telecommunications to a culture of trust backed by the predictable, sound operation of government agencies. Germany is one of the world’s prime examples of healthy institutions, with a business culture that prides itself on professionalism and reliability. We think that this played a big role in how Germany was able to get around Russian gas—with many firms forming critical new relationships and significantly reconfiguring their operations in a matter of weeks.
4. What is to Be Done: Takeaways for Businesses
An outcome like Germany’s during the gas crisis requires many good decisions, at the individual business level and at the level of the economy—both in anticipating disruptions and reacting to them.
In the rest of this post, we will offer some takeaways for individual businesses as well as governments thinking about strengthening supply networks to foster this kind of resilience.
By mapping their supply networks beyond the first few tiers, businesses can see past the mirage in order to understand and address hidden vulnerabilities in their sourcing.
By themselves, individual businesses can do little to combat aggregate vulnerabilities—including the complexity trap. One thing they can do is to better anticipate the risks that they individually face. For example, recall diamond-shaped networks and the diversification mirage, which can make supply networks at the level of an individual firm appear much more diversified than they are. By mapping their supply networks beyond the first few tiers, businesses can see past the mirage in order to understand and address hidden vulnerabilities in their sourcing.
Concentrated dependence does not arise by chance; as we discussed above, it is often the product of incentives that push upstream firms to source their inputs from a single region, or even just a few suppliers. Recognizing these incentives can empower downstream businesses to counteract them. Buyers can seek out suppliers whose cost structure favors regions different from those that dominate in the buyer’s current inputs. They can also offer a premium to suppliers that add sourcing options having low overlap with the other firms in the buyer’s supply network. In other words, the buyer can pay for what it values: true sourcing diversification.
One might wonder whether firms should worry about their suppliers’ sourcing. Perhaps they should just mind their own business and trust their suppliers to deliver reliably. After all, the market is already a powerful source of incentives for reliability: suppliers make more money when they operate well. Unfortunately, the research on this topic shows that the price mechanism is usually not powerful enough to coordinate sourcing correctly.9 To make up for what the market fails to provide, firms should consider mapping their supply chains for essential and volatile inputs at a granular level—and then use that mapping to identify and respond to business risks via strategies such as those sketched above. The research suggests this can be a better business decision than accepting hidden vulnerabilities.
When exposure to a single point of failure such as TSMC is unavoidable, another strategy for businesses—especially those without very deep pockets—is to hedge using the financial markets. For instance, a company can buy contracts that pay in case of trouble in regions upstream of them. This can create the financial slack to respond to disruptions, rather than leaving a firm caught out seeking credit when many other businesses are, too.
5. The Data Frontier
Supply chain intelligence is essential in such efforts. Some providers offer firm-level supply chain visibility, helping firms map the indirect sources of their inputs, identify bottlenecks, and learn about hidden vulnerabilities. Currently, deep insight into supply networks is mainly available to giants such as Toyota, but the plummeting cost of data and analytics promises to change this. If firms coordinate on demanding such information, economies of scale and scope can make it cost-effective for them to get the data they need.
One obstacle to rich quantitative analysis of supply chain robustness is that much of the relevant data is text-based—emails and Slack messages complaining about delays, earnings calls estimating the financial stakes of supply chain risks, etc. This has just changed from a curse into a blessing
Better data is also indispensable as governments think about how to strengthen supply networks. Given the importance of infrastructure and institutions for good interfirm relationships, governments can help supply networks by strengthening these. Unfortunately, this statement is almost worthless at that level of generality: if governments could improve infrastructure and institutions on demand and had the incentives to do so, all countries would probably be rich. Even when benevolent governments can and want to do something useful in these dimensions, they must pick and choose where to focus their efforts.
A key difficulty in wisely targeting the use of limited resources lies in identifying the strained and troubled areas of a supply network—i.e., in figuring out which links are in a bad state. With physical infrastructure, assessing whether a road is in a good or a bad state is relatively easy: one would look for potholes, test the functionality of traffic lights, perform traffic measurements, and the like. Assessing the state of a supply chain link is not so straightforward. Relevant factors include the quality of the firms’ logistics,10 how well businesses can navigate the regulations surrounding their relationships, and even the amount of trust between the firms. We currently lack a framework to systematically measure and aggregate all these factors—and likely need to uncover hidden ones beyond the ones just mentioned.
Nevertheless, we have reasons to hope that the problem of measuring supply chains is going to get technologically easier both for firms and public authorities. One obstacle to rich quantitative analysis of supply chain robustness is that much of the relevant data is text-based—emails and Slack messages complaining about delays, earnings calls estimating the financial stakes of supply chain risks, etc. This has just changed from a curse into a blessing. Large language models and other AI tools have made miraculous leaps in their ability to process text data—ranging from news articles to the content of a company’s Slack—into quantitative forms and integrating it into predictive models of risk. These tools can allow for more proactive decision-making by tracking relevant news and public information about important suppliers in real time, for instance.11
Unfortunately, a big challenge remains despite these technological advances. A lot of information on supplier relationships remains largely confidential. Businesses carefully safeguard information about their supply practices. One simple reason is that their competitors can leverage it to mimic hard-won tactics for improving reliability. And a company’s customers can sometimes use such information to cut a company out of a deal altogether, sourcing directly from that company’s suppliers instead. Nevertheless, if the public sector is to help improve supply networks, some information of this form must make its way outside firms.
There is an interesting echo here of the aftermath of the 2008 global financial crisis, which highlighted the need for increased disclosure for financial stability assessments. The proliferation of low-quality assets and the lack of transparency about their true roles on the balance sheets of various firms were key factors in the global financial turmoil. That crisis, therefore, made it clear that the government needed to know a lot more about the assets firms were trading, and the details of those trades. Without that knowledge, authorities could not monitor, let alone guard against, the emergence of systemic fragilities. Consequently, policymakers pushed banks to disclose detailed information on financial risks, and the subsequent reporting and research led to better systemic risk regulation.12
Something similar is now needed for supply chains. COVID-19 demonstrated the stakes involved in their good functioning and the fact that they could become macroeconomically important. Greater disclosure of supply relationships, including from private firms, is a first step in being in a position to predict and react to future crises, and to help the market to price and react to risk better. The solution to this problem—paralleling the aftermath of the global financial crisis—will have to involve using a mix of government power, technology, and enlightened self-interest to coax data about supply networks out of businesses.13
This is not an easy problem. But recent breakthroughs in modern cryptography offer an assist, addressing some of the misgivings that firms have about disclosure. There has been immense recent progress in developing systems for secure data sharing. These technologies allow very complex calculations to be run on firms’ private data, without revealing that private data to anybody, not even the researchers running the calculations. This can give firms more security in sharing information about some of their most important business processes, and correspondingly reduce the amount of coercion needed in order to get useful aggregate measures of supply chain health out into the open.
6. What Can Governments Do?
The ultimate goal of supply network intelligence at the level of national policy is to know where macroeconomically significant fragility may be building. Once we know that, policy can begin to address the fragility. To some economists and business leaders, that step feels too close for comfort to old-school industrial policy: governments supporting some companies, necessarily at the expense of others.
But the perspective we have developed in this post shows that the government can help supply networks without aggressive intervention to prop up individual firms. Governments can instead think of themselves as being akin to city planners directing repair crews to support infrastructure—making sure the roads and sewage systems run well.
What would this involve, practically?
There are some obvious avenues that involve large investments: upgrading seaports and airports; making border crossings more efficient for both people and goods; maintaining roads; and beefing up the protection of international shipping lanes. As Larry Summers has emphasized, in recessions—when real interest rates have recently been low or negative—it seems like a no-brainer to improve infrastructure in the long run while creating much-needed jobs and economic activity.
But even when government funds are not as plentiful, there are high-leverage strategies that should be on the table in policy discussions. We will not give an exhaustive analysis, but point in a few directions.
First, supply chains are often encumbered not by physical or deep economic constraints, but by policy and regulation. Brexit gumming up the works of UK-Europe commerce is a prime example: the difficulties Britain experienced after exiting the customs union were based only on a regulatory shock, not any changes in the material facts of production. Around the same time, the process of Covid vaccine development and deployment taught the world that governments are capable of removing constraints on progress when that is critical. So in cases where supply networks are on the brink of gridlock, a remedy that governments can implement is to suspend cumbersome rules. To take one small example, some ports have limits on what hours loading and unloading can take place. Governments can coordinate the suspension of such rules in the national interest. This is analogous to addressing a traffic jam on one side of a highway by allowing some lanes on the other side to drive in the “wrong” direction. Here, the key scarce ingredients are urgency and coordination: a clear mandate from high levels of government to cut through red tape.
Queues of ships at ports have no mechanism for priority service: a ship holding up ten million dollars of delayed production has the same priority as one delivering low-value souvenirs that will sit in a warehouse for months.
Second, governments and other public and semi-public organizations provide key inputs supporting business activity, but do not always allocate these efficiently. For instance, queues of ships at ports have no mechanism for priority service: a ship holding up ten million dollars of delayed production has the same priority as one delivering low-value souvenirs that will sit in a warehouse for months. Economists love congestion pricing for roads. They should also lobby to resolve traffic jams in infrastructure that is equally critical, but less visible.
Moving to less visible institutions, we saw in the research on Indian supply chains that courts can play a large role in helping businesses operate efficiently. In cases where access to enforcement is a constraint on the formation of important relationships, governments can allocate the scarce public resource of court time better. They can, for example, let firms pay higher taxes (i.e., user fees) when it’s important for them to get a quick hearing. Similar considerations apply to firms’ interactions with regulators, when a judgment from a regulatory authority is needed in order to facilitate a new sourcing relationship.
The broader point is that many of the markets and institutions that support commerce could be designed better. An idea that has reshaped economic theory over the past 50 years is that the mechanics of markets do not have to be taken as natural facts. Markets can, instead, be carefully designed to meet society’s needs. Examples extensively studied in the economics literature on market design include organ exchange, the matching of students to schools, the assignment of refugees to host locales, and the allocation of donated food items to needy families. When economists redesign such systems, the markets often end up working more smoothly.
What is striking is that the markets mentioned above—which economists have spent a lot of effort on—are tiny in importance relative to the world supply network. Meanwhile, as we have seen, that network is supported by many systems that are not in any way optimized.
If we manage to quantify key bottlenecks and estimate (even roughly) the value of addressing them, we can target government action much better. Sometimes this will require adding capacity, but a lot of value can be obtained simply by creating and redesigning markets to better allocate the capacity that we already have.
Finally, research co-authored by one of us14 suggests that quick mobilization of government resources to directly relieve pressure in a crisis can be a surprisingly wise antidote to the gridlock caused by a complexity trap. For example, the first few stages of the post-pandemic snarl involved misplaced shipping containers and a shortage of labor in trucking. We have discussed how governments can often help simply by suspending rules stopping businesses from resolving the problems themselves. But when that doesn’t work, governments can help more directly by mobilizing military manpower and vehicles, as they do for natural disasters. The overall economic impact of supply chain collapses is likely much larger, investments to unblock infrastructure have enormous returns, and governments can recoup the costs from affected businesses with a claim on future profits (as in financial bailouts). Another approach is to spend political capital—e.g., the Biden Administration invoking the Taft–Hartley Act to prevent the crippling U.S. port strike currently looming. All these interventions are drastic, but when they are needed, they solve drastic problems.15
7. Supply Networks and Geopolitics
Much of our analysis above focuses on peacetime concerns. Our final policy insight, however, is that the fabric of supply networks should be a first-order consideration in geopolitical strategy—and, in particular, in thinking about a potential US-China conflict.
There has been a lot of recent policy discussion about decoupling from Chinese inputs. The most basic reason for the US to seek such decoupling is the need for military production that does not depend on a potential adversary. But policymakers’ interest in decoupling seems to reach well beyond military concerns. A recent study presented at the influential Brookings policy conference documented in detail that the US is highly reliant on Chinese inputs in many non-military goods.16
Why was this research done, and why does it matter? At a first pass, Western policymakers might reasonably wonder why they should care about how much their economies rely on China indirectly. For instance, while a severe conflict might prompt a halt to some or even most direct Chinese imports to the US, these imports account for a small fraction of US GDP.17 And when it comes to indirect inputs, it seems unlikely that China would or could stop the indirect use of its inputs in US goods. To take a relevant analogy, despite the heavy sanctions that were imposed on Russia after it invaded Ukraine, Russia continues to indirectly source US and European inputs—and such inputs even constitute a significant contribution of its war machine. The world economy is so interconnected that even major conflicts are unlikely to cut off indirect sourcing.
Thinking about supply network fragility helps rationalize policymakers’ concerns. A large geopolitical shock to US-China relations would not necessarily cut off the supply of any indirect input, but it would damage the commercial fabric linking many nations across the Pacific. This damage would come through trade wars, threats to commercial navigation, worries about commercial espionage, and regulatory obstacles—all of which are not hypothetical, but already playing out to some extent. Recall that even a fairly genteel disruption—Brexit—reduced EU-UK trade and very likely harmed Britain’s economic performance. Trans-Pacific tensions nearing the brink of war would likely cause much more damage to much higher-value supply networks in much larger nations.
The economic damage of war in the Pacific would likely be drastically amplified through damage to the fabric of world commerce. The harm would go beyond what happened after COVID, and would pose corresponding challenges to economic governance—with inflation being a key threat.
From this perspective, the reason to fear a sharp US-China decoupling is not that the US would entirely lose access to specific indirect inputs. Instead, the fear is that a level of generalized strain akin to what followed the COVID-19 shock would lead to gridlock in a huge swath of complex networks.
There are two large geopolitical implications. First, the US has a strong strategic interest in building supply networks for high-value goods that are as insured against systemic trans-Pacific shocks as possible. The way to create these is not by fiat; instead, governments (and industry groups, when possible) should systematically support businesses in forming diverse sourcing networks. The key deliverables include understanding the key bottlenecks and providing good infrastructure—both physical and institutional—to relieve them. (These investments are likely to have large, positive returns even putting fragility aside.) Direct subsidies to foster diverse sourcing in the interest of national security are also likely to be important, when the money is available—but excellent targeting is especially important for these, and we do not yet have the science to do this well except in the most obvious cases.
Second, leaders should recognize that such efforts will achieve, at best, only partial success. In any realistic scenario, the economic damage of war in the Pacific would likely be drastically amplified through damage to the fabric of world commerce. The harm would go beyond what happened after COVID, and would pose corresponding challenges to economic governance—with inflation being a key threat. The recognition that a large trans-Pacific conflict would create systemic shocks to supply networks unlike any seen in previous conflicts should make policymakers even more eager to deter and avoid war.
There is huge value in protecting the fragile fabric on which much of the world’s wealth relies.
Houthakker showed that, at the aggregate level, substitutability between inputs can be high even if it is essentially zero at the firm level. More recently, Oberfield and Raval showed how substitutability at the aggregate level (at the level of the manufacturing sector) is higher than at the plant (an in industrial plant) level; focussing on substitutability at the plant level misses important mechanisms of adjustment.
Household’s adaptability was, of course, important too: households reduced their gas consumption by 17%.
Germany also adapted by sourcing gas from other countries to partially replace Russian supplies.
Ports, for instance, channel 80% of trade, acting as the starting point and destination of about $20 trillion worth of merchandise.
Sociologists have documented that economic relationships across firms are embedded in social networks of interpersonal connections, and could not function without personal familiarity and trust. This manifests in a variety of informal interactions—ranging from phone calls to double-check orders that don’t seem right, to gift-giving to thank counterparts for going the extra mile to make a transaction work.
This story, which deserves its own long article, starts from the observation that European colonies that were relatively rich in 1500 became relatively poor by 1995, while areas that were poor in 1500 became rich. Many hypotheses had been advanced to explain this, including that hot weather made people lazier. Acemoglu, Johnson, and Robinson set out to examine the hypothesis that the quality of institutions was what mattered; they used arguably exogenous sources of variation in the quality of early institutions and showed that this variation can account for differences in long-run trajectories.
This example led a friend of one of the authors to call the generals and politicians who implemented the partition of Korea along the 38th parallel “the most influential experimental economists of all time.”
The increase of inefficient in-house production is just one of the ways in which congested courts hurt productivity. Other mechanisms include firms tilting their input mix towards standardized inputs, using less intermediate inputs, and being less likely to switch suppliers.
If you’re a buyer, your suppliers do care about their reliability, but in a different way from the way you care about it. You care a lot about their availability when all your other sources are unavailable. To them, it doesn’t make as much of a difference whether you have alternative sources of supply when they are not operational. If you want to make your suppliers care about that, you have to pay them extra. They might care about being your only viable supplier if they can charge a premium at the times when you really need them. But in practice, in times of shortage, price hikes do not seem to be a big part of interfirm relationships. While we think well-targeted price premiums could help, it seems likely that other negotiated rewards for diversified networks have an important role to play.
This includes both actual technologies of shipping, scheduling, etc., as well as potential alternatives in case the “main” ones don’t work, along with the costs of implementing those alternatives. This is just one example of the hidden aspects of the health of a relationship.
Sometimes these tools reveal that superficial claims of fragility are wrong: for example, at the 2023 Brookings conference, Tarek Hassan presented a method for analyzing earnings calls that showed that German executives were mentioning supply chain risks to their investors barely more than usual, even as they expressed panic their media efforts to lobby German politicians. One possible reason for the contrast, as Hassan put it, is that it’s a crime to lie on an earnings call, but not a crime to lie to journalists.
E.g., see the Dodd-Frank Wall Street Reform and Consumer Protection Act and the Basel III framework.
Similarly, the Scale Tool, a computational system developed by the US commerce department with the goal of creating a very granular picture of vulnerability and resilience, might be a significant first step towards a more resilient supply chain fabric. It reportedly includes data “from the entire American goods economy.” It is too early to tell how useful these innovations will be, but they clearly represent a significant step away from “nobody’s in charge.”
See especially Section VII.B.
And the theoretical research we have cited in this paragraph suggests that there is little risk that such emergency measures crowd out business incentives to invest in their own supply networks. By the time supply networks collapse due to the complexity trap, it is due to a systemic problem that cannot be blamed on any single business’s underinvestment, so what happens in these rare states has little effect on the investment incentives of any business. This in contrast to the calm and conventional interventions discussed above to support supply network health, which must be carefully engineered to avoid crowding out businesses’ own logistical investments.
Indeed, once we account properly for indirect dependencies, China was the main foreign supplier to over 90% of US manufacturing sectors in 2018.
US direct imports from China accounted for $562.9 billion in 2022—or around 2.2% of US GDP that year. In the words of Baldwin, Freeman and Theodorakopoulos, “if one looks at the direct source of imports, China is important but not dominant.”
Is there any research on whether vertically integrated companies tend to perform better in general? The article points out that diversified supply chains may not be as resilient as expected, plus it stands to reason that a vertically integrated firm would benefit from detailed information about its supply chain. I know that Tesla and BYD has had success with that strategy.
This sounds a lot like <that whose name cannot be said>, "neoliberalism," strengthening rule of law, regulation according to cost benefit analysis public deficits = Σ(expenditures with NPV>0)