191

Chapter 6

Building Resilient Supply Chains

The year 2021 was when supply chains—the networks of producers,

transportation companies, and distribution centers that develop and move

products and services—entered dinner table conversations. Though this term

has certainly been part of the lexicon going back to the 1980s, and has been

a part of doing business for centuries, COVID-19 highlighted supply chains’

vulnerabilities, which became front-page news. Supply chains have become

more complex, interconnected, and global than they were in decades past.

The share of world trade that crossed at least two borders increased from 37

percent in 1970 to nearly 50 percent in 2014 (World Bank 2020a, 2020b).

This increasing segmentation of the production process has reduced prices in

the United States, while also raising productivity and aggregate incomes in

many of the low-income countries that are integral to global supply chains

(World Bank 2020a). However, the globalization of production has also

made supply chains more vulnerable to disruption. This fragility has been

exacerbated as firms have removed excess capacity (e.g., extra inventory,

or reserves of people with the time and skills to solve problems), making

supply chains less resilient. That is, they have less ability to recover quickly

from unexpected events. Thus, though modern supply chains have driven

down consumer prices for many goods, they can also easily break (Brede

and de Vries 2009; Baldwin and Freeman 2021; Miroudot 2020; de Sá et al.

2019; White House 2021a).

Though it was not inevitable, movement toward this more fragile con-

figuration has been happening for decades, as public and private policies

have undermined firms’ incentives to invest in such capacity to ensure

192 | Chapter 6

resilience. The COVID-19 pandemic is not the first time that supply chains

have been disrupted; the production and distribution of goods have been

regularly snarled by natural disasters, cyberattacks, labor strikes, supplier

bankruptcies, industrial accidents, and climate-induced weather emergencies

(de Sá et al. 2019). The pandemic simply exposed just how complex and

interconnected modern supply chains have become. These highly publicized

disruptions and product shortages made the public painfully aware of the

many steps involved in getting a product produced, transported, and placed

on shelves or doorsteps.

The first section of this chapter describes modern supply chains and explains

their evolution, focusing on manufacturing. Supply chains are shaped by

a complex network of relationships; these relationships affect not just the

movement of supplies from place to place but also the incentives of lead

firms and suppliers to invest in producing new products, in providing good

jobs, and in achieving resilience. The second section describes how increas-

ingly frequent disruptions of the economy suggest that supply chain fragility

will continue to be a problem. The third section outlines the private sector’s

incentives to become more resilient in the face of these challenges. Finally,

the fourth section suggests vital roles for government in helping to shape

supply chains and overcome market failures.

21st-Century Supply Chains

Supply chains are the linkages in the production process that facilitate the transformation of raw materials into finished goods or services. A supply chain is made up of producers and logistics providers that move inputs from one stage to the next, and also of participants in the distribution channels for the finished product, including wholesalers, distributers, and retailers. This chapter primarily focuses on manufacturing supply chains that facilitate the production of physical products from unprocessed materials.1

Figure 6-1 depicts some of the ways supply chains are commonly organized. Even within the same industry, firms have different supply chain

1 In addition to goods, services are also part of supply chains and often face some of the same issues that are discussed in this chapter.

Building Resilient Supply Chains | 193

configurations (Kamalahmadi and Parast 2016; Lund et al. 2020). This figure gives four stylized examples of how supply chain relationships could be formed:

Vertical Integration with Isolated Industries Panel A of figure 6-1 illustrates a three-firm configuration, where each firm (shown by the dots in the figure) is self-sufficient—that is, completely vertically integrated. Thus each firm produces everything, starting from raw materials and ending with the finished product. In this configuration, supply chains are completely internal to a firm. A prototypical example of this is the automaker Ford’s River Rouge Plant, which in the 1930s included a steel mill, glass factory, power plant, rubber factory, foundries, machine shops, stamping plants, assembly lines, a cement plant, a paper mill, a leather

Source: Adapted from Cavalho (2014). Note: From left to right: A, vertical integration with isolated industries; B, outsourcing with isolated industries; C, outsourcing and offshoring with isolated industries; D, outsourcing with a central node (star- shaped). Arrows denote flows of products, information, and the like between companies.

Figure 6-1. Common Types of Supply Chains

B A

C D

C ou

nt ry

A

C ou

nt ry

B

194 | Chapter 6

plant, and a textile mill (Weber 2019). Ford also owned a rubber plantation in Brazil, coal mines in Kentucky and West Virginia, and railway cars to transport raw materials. This allowed Ford to maintain direct control over the entire manufacturing process. However, this complete vertical integra- tion also made it difficult for Ford to cut costs during the sharp decrease in demand for cars during the Great Depression, as the automaker continued to bear the fixed costs of component production. In contrast, Chrysler, which was much less vertically integrated during this time period, did not need to bear these fixed capital and administrative costs; Chrysler’s suppliers did (Chandler 1962, 1992). A firm’s decision to vertically integrate depends in part on whether the costs of transacting in different markets outweighs the cost of managing these activities internally (Coase 1937).

Outsourcing with Isolated Industries Panel B of figure 6-1 represents three industries, each with significant supply-chain relationships. Here, inputs travel “downstream,” where they are transformed into a final good. The lead firm typically designs products and directs production by multiple tiers of suppliers in many locations, but it does not own most of these suppliers. This is called outsourcing. Outsourcing allows the lead firm to contract with firms that may have lower production costs due to lower wages or other competitive advantages (see box 6-3 below).

The chain includes direct suppliers of the lead firm (tier 1 suppliers), as well as suppliers to those suppliers (tier 2 suppliers), and so on—all the way back to the raw materials used to produce the good. A firm can have hundreds of tier 1 suppliers and thousands of tier 2 suppliers, as shown in figure 6-2 (Lund et al. 2020).2 Looking at the publicly disclosed lists of suppliers for 668 companies, the McKinsey Global Institute found that the number of direct suppliers was large and that the network of indirect suppli- ers was even larger, often numbering in the thousands (Lund et al. 2020). As discussed below, the degree of coordination between the firms, represented by the arrows in figure 6-1, can vary between two extremes: arm’s-length transactions and collaborative relationships.

Offshoring and Outsourcing with Isolated Industries If lead firms choose suppliers across national boundaries, this is called offshoring, as shown in panel C of figure 6-1. Offshoring gives companies expanded scope to locate production in areas with lower wages, or that have other competitive advantages not available in their home country, such

2 Note that, due to data limitations, the tier 2 suppliers in figure 6-2 may not be supplying inputs into the lead firms’ products; rather, they are suppliers of the tier 1 suppliers, which usually produce for more than one lead firm.

Building Resilient Supply Chains | 195

as access to natural resources or better technology (Antràs 2020; World Bank 2020b). Competitive advantage may be the result of naturally occur- ring endowments or developed by government or private sector policies (Mazzucato 2016; Lee 1995). In the past, internationally traded goods were largely either raw materials, such as cotton, or finished goods, such as cloth- ing. Since the early 1990s, there has been a large rise in trade of “intermedi- ate goods” or components, such as fabric that has been cut but not sewn.

In both panels B and C of figure 6-1, no connections exist between the blue industry and the parallel orange and black industries. In this diagram, nodes are industries with few overlapping suppliers, such as electronics and autos in the past.

Outsourcing with a Central Node In contrast to the isolated industries depicted in panels B and C of figure 6-1, supplier firms usually sell to more than one lead firm and may sell to several different industries, as shown in panel D (Carvalho and Tahbaz-Salehi 2019; Carvalho 2014). One example is a star-shaped configuration, with one cen- tral node (the green node) that is used in production by all other nodes. Firms in this general-purpose industry supply a wide number of other industries and often also use inputs from the industries they supply (Carvalho 2014).3 These types of supplier relationships allow firms to take advantage of

3 In practice, some suppliers, and even the central node of panel D, may be offshored as well as outsourced; for simplicity, this configuration is not depicted.

Apple

856 1,676 638 717

18,000+ 12,000+

7,400+ 5,000+

Airbus Nestlé General Motors

Tier 2 suppliers and belowPublicly disclosed tier 1 suppliers

Source: Adapted from Lund et al. (2020), relying on the Bloomberg Supply Chain Database.

Figure 6-2. Examples of Tier 1 and Tier 2 Supply Relationships

196 | Chapter 6

economies of scale, where per-unit costs decrease as the number of units produced increases and the supplier is able to sell to multiple firms.

Firms’ decisions regarding the design of their supply chains lead to a complex web of connections. Aggregating firm-to-firm supply chain con- nections, industry A has supply chain connections to industry B when firms in industry A purchase inputs from firms in industry B. Though compre- hensive data on firm-to-firm supply relationships are lacking for the United States, the network structure of the U.S. economy can be visualized at an industry level. This industry-level analysis can shed light on which indus- tries supply inputs to many other industries and the structure of network con- nections between industries. These connections can amplify microeconomic disruptions.

The U.S. economy is complex and interconnected, with several central hub industries that have connections to most other sectors. Using the most disaggregated, publicly available sectoral data—the Bureau of Economic Analysis’s (BEA) Input-Output Accounts Data—it is possible to see the supply chain connections between 417 different industry sectors, as depicted in figure 6-3. Each node is a sector, and the connections between them rep- resent flows of inputs from one supplying sector to another. The network is sparsely connected; on average, each narrowly defined industry is connected to only 11 other industries (Carvalho 2014). However, a small number of hub industries are highly connected to many others in the network. Although most industry pairs are not directly linked, they are indirectly connected by a small number of steps through these hub industries (Carvalho 2014). The most-connected input supply sectors (the numbered nodes) in 2002 included real estate, electricity generation and distribution, iron and steel mills, depository and credit intermediation, petroleum refineries, and truck trans- portation (Carvalho 2014). The CEA’s analysis of the 2012 input-output tables shows that semiconductors have become a highly connected industry, while truck transportation has dropped from the top 10 list (Carvalho 2014; Bureau of Economic Analysis 2012). Other countries also have similar pat- terns of central hub industries, though the central industries may be different (Carvalho and Tahbaz-Salehi 2019; Fadinger, Ghiglino, and Teteryatnikova 2015; McNerney, Fath, and Silverberg 2013).

Arm’s-Length and Collaborative Relationships The arrows in figure 6-1 represent connections between the nodes in the sup- ply chain. The nature of these connections can vary between two extremes: arm’s-length transactions and collaborative relationships.

In an arm’s-length transaction, a firm purchases a standard input from an unaffiliated firm, often choosing from a large set of possible sellers. In this case, the connection is very simple: the seller provides an off-the-shelf

Building Resilient Supply Chains | 197

product to the buyer, which sends payment. If there is a problem with a supplier (e.g., the price is too high or a disaster causes it to be unable to produce), the buyer can easily find another supplier. Lead firms may benefit from these relationships because they are able to easily change suppliers, creating competition that requires suppliers to reduce their prices to win business.

In collaborative relationships, firms in a supply chain communi- cate frequently about the product and production process; performance requirements (e.g., price, quality, specifications, and delivery schedule) are customized for a particular product, and are usually set by the lead firm (Gereffi 2020). In some instances, these are transactions between affili- ates of a large company, while others involve a lead firm and financially independent suppliers. For instance, companies such as Nike do not own the facilities in which their products are manufactured; instead, they provide the design, product specifications, advertising, distribution, and coordination of the complex network of contractors that make the shoes (Gereffi and Korzeniewicz 1994).

Figure 6-3. The Production Network Corresponding to U.S. Input-Output Data in 2002

Source: Carvalho ( 2014). Copyright the American Economic Association; reproduced with permission of the Journal of Economic Perspectives.

198 | Chapter 6

Suppliers in collaborative relationships provide these highly custom- ized inputs on a repeated basis, usually without complete or easily enforce- able contracts (Hart and Moore 1990). Both the buyer and supplier invest in capital, equipment, or knowledge that is useful only with a particular partner (Antràs 2020). These relationship-specific investments increase the cost of finding and switching to a new supplier, but often pay off in components that better fit the lead firm’s needs and in quicker responses to unexpected situa- tions (Antràs 2020; Helper 1991; Gibbons and Henderson 2011). A large lit- erature describes the potential benefits to lead firms of having collaborative relations with suppliers, such as reduced costs, defect rates, and lead times; and increased investment, responsiveness, innovation, and problem solving (Delbufalo 2012; Gibbons and Henderson 2011; Aoki and Wilhelm 2017).

A key reason for the long-term profitability of firms such as Toyota and Honda is their investment in collaborative relationships with their suppliers (Aoki and Wilhelm 2017; Liker 2004; Lieberman, Helper, and Demeester 1999). The rise of these sticky buyer–seller relationships is a dis- tinctive aspect of the recent rise in global value chains (World Bank 2020a). Understanding why some firms adopt collaborative relationships and others do not is an area of active research in many disciplines, including econom- ics, management, and sociology (Bernstein 2015; Gil and Zanarone 2018; Schrank and Whitford 2009). Box 6-1 provides an example of how one firm currently combines domestic production, offshoring, vertical integration, and offshoring to make its products.

However, there is no single optimal way to organize a supply chain. Even within the same industry, firms often choose different strategies. For example, on average, automakers producing in the United States have 4.7 suppliers for each product category, a financial stake in 22 percent of trans- actions, and relationships with suppliers that last 2.4 years. However, there are substantial differences among automakers in these practices. Japanese vehicle manufacturers engage more in collaborative outsourcing than do their U.S. counterparts; therefore, Japanese relationships with suppliers last 70 percent longer, and they have fewer than half as many suppliers for each part as do U.S. automakers. These differences persist even when automak- ers are selling similar products in the same market, and after controlling for component volume and mix (Helper and Munasib 2022). Automakers differ in their offshoring strategies as well. For example, in 2020, Ford had 24 percent more production offshore than did Stellantis.4

44 These data, from the American Automobile Labeling Act (AALA), do not allow the separation of U.S. and Canadian content (Center for Automotive Research 2020).

Building Resilient Supply Chains | 199

Drivers of Change in Supply-Chain Structures Global supply chains that involve offshoring, and often outsourcing, multi- plied rapidly from 1990 to 2008, though their growth slowed after the 2008

Box 6-1. The Supply Chain of a Hot Tub The M9 hot tub is made by Bullfrog Spas in Utah, where 500 workers assemble almost 1,850 parts from 7 countries and 14 states (see figure 6-i). The hot tub top shell starts as a flat acrylic sheet from Kentucky, which is then combined with a different type of plastic in Nevada and sprayed with an industrial chemical from Georgia. Parts of the frame shell of the hot tub are driven in by trucks from Idaho several times a week. Many of the electric motors come from China and are assembled into water pumps in Mexico and then driven to Utah. Additional material for exterior cabinets is transported from Shanghai on container ships through the ports of Long Beach or Oakland. Water-spraying jets are made in Guangzhou, China; are sent through the Panama Canal and Eastern ports to the supplier’s warehouse in Cleveland, Tennessee; and then are sent on to Utah. Once fully assembled, the finished hot tubs are placed on trucks or trains and delivered to retailer warehouses. This example illustrates both the extent of outsourcing, which increases the number of individual companies involved in the production of a single good, and the geographic distance traveled by each component, estimated to total nearly 900,000 miles, as well as the dependence on transportation and logistics this entails.

Herriman, Utah

Frame Seat backs

Acrylic sheet Urethane

Water jets Touch screens

Electric motors Cabinet pieces

Component

Figure 6-i. Sources of the Components of a Hot Tub

Source: Adapted from Hufford, Kim, and Levinson (2021).

200 | Chapter 6

global financial crisis (World Bank 2020a). Manufacturing firms also out- source services, including logistics, cleaning, and security. That is, workers providing these services are no longer direct employees of manufacturers; instead, they work for financially independent contractors. For example, in food, cleaning, security, and logistics services, the share of those working for such contractors in the United States rose from about 5 percent to about 30 percent between 1950 and 2015 (Dorn, Schmieder, and Spletzer 2018).

Two key changes have increased the attractiveness of outsourcing and offshoring. The first change is increased access to foreign suppliers, making offshoring more cost-effective for firms, largely due to advances in information technology (IT) and reductions in trade barriers since the 1990s. Advances in IT allow firms to convey detailed information about product and process specifications across long distances, while improve- ments in transportation, such as containerization, allow goods to be moved more quickly and consistently (Grossman and Rossi-Hansberg 2006). These developments make it possible to segment the production process, keeping highly skilled functions, such as research and development and management, in more advanced economies, while moving others, such as production of components or assembly, to countries with lower wages (Gereffi 2020).

Major trading nations have signed agreements that reduced barriers to trade, such as the 1994 North American Free Trade Agreement. These trade pacts contain strong protections for property rights of corporations, but far weaker protections for labor rights. This disparity increased the attractive- ness to multinational firms of offshoring production to low-wage countries (Drake 2018). The result has been increased availability of cheaper goods for American consumers, but also significant pressure on wages and ben- efits that have often driven workers from the middle class (Hakobyan and McLaren 2016).

Finally, widespread international government subsidization of manu- facturing industries has lowered prices that lead firms pay for inputs, and has oriented many nations’ domestic industry toward global supply chain participation (Hauge 2020). For instance, in the past few decades, the Taiwan Industrial Technology Research Institute has facilitated relation- ships between young, domestic semiconductor manufacturers and mul- tinational buyers. The institute helped organize two firms—the Taiwan Semiconductor Manufacturing Company and the United Microelectronics Corporation—and gave them intellectual property. By 2020, these two com- panies accounted for 60 percent of global semiconductor revenue (Lee 2021; Breznitz 2005). Taiwan and China have extensively subsidized their semi- conductor industries, with subsidies often approaching nearly 30 percent of a company’s revenue, according to the U.S. Department of Defense (2022, 36). “Made in China 2025,” China’s 10-year plan to transform itself into a world leader in high-tech industries, promotes policies that increase Chinese

Building Resilient Supply Chains | 201

firms’ market share and builds globally competitive industries in key sectors without relying on foreign firms (Congressional Research Service 2020). (See box 6-2.)

The second key change is the growing role of financial criteria and institutions in corporate decisionmaking. This “financialization” of the economy has encouraged outsourcing and offshoring because of savings in costs that are easily measurable. Firms increasingly tie executive compensa- tion to such financial measures as earnings per share, stock prices, and return on equity. Before the 1970s, only 16 percent of the chief executive officers in Standard & Poor’s 500 companies had compensation based on such mea- sures; by the 1990s, 47 percent did, and in 2021, the vast majority employed by large corporations did (Admati 2017). Such incentives have encouraged managers to focus more on these financial statement numbers than on less easily measurable metrics, such as resilience.

However, financial metrics can be misleading. Although an outside or offshore supplier may offer a lower unit price, these savings may be eaten away by hidden costs, such as longer lead times, increased vulnerability to disruption, and reduced access to ideas for innovation due to linguistic and geographic distance (Gray, Helper, and Osborn 2020). Such hard-to- estimate costs are often ignored, even though they may negate the estimated savings from outsourcing (Barthelemy 2001). These less easily measurable metrics are often characterized as “soft” information, which, in contrast to “hard” information, may require knowledge of the environment and/or personal relationships to collect and understand.

Soft information includes operational measures that use physical units, such as defect rates or downtime, and involve such intangibles as the value of research and development or of employee training (Liberti and Peterson 2019; Edmans, Heinle, and Huang 2016). It is often difficult to convert soft information into dollars. For example, it is not easy to measure how much an investment in training will improve quality, and how much this improve- ment in quality will flow to the bottom line. Such investments are also hard for outsiders or those without experience with a given product to verify. Thus, the pursuit of favorable performance as measured by financial indica- tors may induce firms to act in ways that could be detrimental to long-term performance, essentially trading longer-term resilience and sustainability for nearer-term profitability (Edmans, Heinle, and Huang 2016).

For firms increasingly driven by short-term investors’ demands, the temptation to ignore these costs has often been great. A survey of senior U.S. financial executives found a willingness to sacrifice long-term shareholder value to meet Wall Street earnings targets or smooth reported earnings. For example, when managers were asked if they would “accept a sacrifice in value . . . to avoid volatile earnings,” 78 percent said yes; 55 percent would “delay starting a new project, even if this entails a small sacrifice in value”

202 | Chapter 6

Box 6-2. The Role of China in U.S. Supply Chains A significant factor in the recent evolution of global supply chains has been the rise of China, which is now the largest source of U.S. imports. China’s manufacturing began exploding in the 1990s, and its share of world manufacturing exports rose from 3.1 percent in 1991 to 17.6 percent in 2015, before dipping to 14.2 percent in 2018 (Autor, Dorn, and Hansen 2021).

Initially, China specialized in simply assembling products from imported components and designs. For example, it is estimated that in 2010 China provided less than 2 percent of the value added of the Apple iPhone 4; the product was designed in the United States, and the com- ponents were made in places like Japan and South Korea; no Chinese suppliers contributed components (Linden, Kraemer, and Dedrick 2011). However, China learned quickly, and for the iPhone X in 2018, it contributed more than 25 percent of the value added, including assembly and high-value components such as the battery pack and touch screen (Linden, Kraemer, and Dedrick 2007; Xing 2019).

China’s entry into global supply chains was facilitated not only by technological advance in transportation and communication but also by changes in institutions. Particularly important were the United States’ granting of Permanent Normal Trading Relations (PNTR) to China in 2000 and the admission of China to the World Trade Organization in 2001, steps that gave imports from China permanent access to the relatively low tariff rates reserved for members of the World Trade Organization. These steps did not require China to change its labor policies, which banned workers from joining independent trade unions, involved reprisals against workers who sought higher wages, and involved forced labor. These policies suppressed wages in China, increasing the competitiveness of firms, including multinational firms, that produced there.

China’s competitiveness was also facilitated by large subsidies, and requirements that multinationals transfer technology to Chinese firms. As the Congressional Research Service concluded, China aims to advance its national development goals and future global economic position through industrial policies that seek global civilian and military leadership in advanced and emerging technologies. China’s policies fea- ture a heavy government role in directing and funding Chinese firms to obtain foreign expertise and intellectual property in strategic industries, including aerospace, semiconductors, microelectronics, pharmaceuticals, and electric vehicles (Congressional Research Service 2020). Through these policies, and aided by U.S. companies pursuing asset-light strate- gies, China gained large degrees of market power in a variety of critical supply chains. For example, China has 97 percent global market share of the ingots and wafers used to make solar panels (U.S. Department of

Building Resilient Supply Chains | 203

to avoid missing an earnings target (Graham, Harvey, and Rajgopal 2019, 8). Underlying this willingness is a view that stock market investors lack the informa

Related Tags