Shortages of essential medical supplies, semiconductors, and other goods and materials during the Covid-19 pandemic have dramatically exposed the vulnerabilities of global supply chains. U.S. government officials have been carrying out President Biden’s executive order to propose remedies to the risks that such networks pose to four specific areas (pharmaceuticals, strategic materials such as rare earth minerals, semiconductors, and large-capacity batteries) and six sectors of the economy (defense, public health, communications technology, energy, transportation, and food production). We think they will find one fundamental cause is the erosion of the U.S. “industrial commons” — the domestic capabilities needed to support the development and production of many goods deemed critical to U.S. interests.
Addressing this decline requires an appreciation of factors related to both the supply side (the industrial capabilities for making these products) and the demand side (how these products are bought and sold). To illustrate both sets of challenges, we describe the efforts of Massachusetts-based Shawmut, a producer of advanced textiles, to help satisfy the soaring need for N95 masks and hospital gowns during the pandemic. Shawmut’s experience offers insight into how the difficulty of rebuilding domestic capabilities may depend on a complex set of factors, including the health of adjacent manufacturing sectors and changes in buying and regulatory practices.
Rebuilding a Weak Commons Must Begin with Imports
Manufacturing any product requires a supply of raw materials as well as the skills and the production tools to convert those inputs into outputs. When a country has a robust industrial commons, a firm can draw on domestic suppliers of materials, makers of production equipment and tooling, and a workforce that has the necessary skills to operate those processes. But American manufacturers offshored a significant amount of production in fields such as electronics as early as the 1970s, and this approach expanded dramatically into information technology (IT) and communications equipment in the 1990s and broad swaths of industry in the early 2000s. This offshoring was accompanied by the movement of suppliers who chose to follow their customers abroad, leading to a widespread loss of U.S. jobs in design and production.
The impact of weakening the commons can be seen in Shawmut’s efforts to respond to an urgent appeal from Massachusetts business leaders and state government officials for N95 masks early in the pandemic. N95s are filtering respirators that use a polypropylene non-woven fabric made by a melt-blowing process. The melt-blown fibers are randomly deposited in a sheet and electrostatically charged to form a highly efficient, fabric-like filter material. The filter material is then ultrasonically welded to a thermo-formed cup to create the face-fitting part of the mask. This is followed by an assembly process that adds head straps, an aluminum nose clamp, nose foam, and the required markings that the National Institute for Occupational Safety and Health (NIOSH) has approved the product.
As it tried to rise to the occasion, Shawmut had to confront two challenges: finding sources of melt-blown filter material and ancillary components and acquiring automated production equipment for the fabrication and assembly of the mask itself. Not surprisingly, China is the major producer of melt-blown fabrics thanks to the scale and scope of its garment industry. As the garment industry moved to Asia in the early 2000s, melt-blown fabric producers moved as well to be close to those customers. Even U.S. domestic manufacturers of masks rely heavily on Chinese sources — including their captive facilities in China — for some of their melt-blown fiber. Though there were a few domestic sources of melt-blown fabric, no manufacturer could offer Shawmut any capacity when it needed it.
Shawmut also struggled to find U.S. sources for mask components such as moldable cup materials, elastic straps, and specially coated aluminum strips for the nose bridge. All of these were available from numerous suppliers in China, but developing a domestic source took months of trial and error.
The void in domestic tool-making capacity was even more challenging. Shawmut was able to acquire a melt-blown fabric line from Reifenhäuser of Germany on an “accelerated” seven-month delivery schedule, but there were no suppliers of N95-cup-mask assembly equipment in the United States or Europe. With most of the world’s mask production capacity in China, equipment suppliers had shifted there as well.
Shawmut was able to identify five Chinese suppliers who offered automated equipment with delivery lead times as short as 30 to 45 days, but adapting the equipment to its line in Massachusetts was challenging due to a dearth of automation engineering expertise in the United States. It took months to reconfigure, debug, and ramp up the new equipment, though Shawmut’s engineers were eventually able to get it running. By November 2020, Shawmut had installed melt-blown capacity to produce 180 million masks per year, started up the first of its 17 mask-assembly lines, and submitted samples to NIOSH for testing and regulatory approval.
It is important to note that Shawmut had to import both materials and production equipment and had to start with relatively little local content, just as China did over the last three decades as it entered new markets. The lesson for U.S. officials and business leaders: It takes many years to build localized capabilities, especially in tool building, and restoring those capabilities once lost can be at least as challenging as building them in the first place.
When Adjacent Sectors Share a Commons, Rebuilding is Easier
Shawmut’s experience in manufacturing hospital gowns during the pandemic was quite different from its N95 effort and illustrates how capabilities in a commons can cut across multiple sectors. In the early days of the pandemic, the company received a call to help a supplier to the Federal Emergency Management Agency (FEMA), which requested enough fabric to make tens of millions of urgently needed isolation gowns. As one of the relatively few textile producers remaining in the United States, Shawmut had moved into advanced textiles used by automakers for interior components such as seats and headliners. Because auto manufacturers rely on just-in-time production and interior components are bulky and costly to ship, suppliers of these components have located themselves close to vehicle-assembly facilities.
Though isolation gowns have typically been made of woven textile fabrics, there were not enough looms in the United States to meet the surge in demand at the beginning of the pandemic. Shawmut devised an alternative product made of a lightweight knitted fabric. With demand surging, it reached out to a number of smaller specialty-textile producers to supplement its internal capacity in knitting, finishing, and lamination. It also worked with a producer of polyurethane foam to create a variant of the gown fabric that could be processed on the company’s seven high-output automotive lamination lines. Shawmut’s coalition totaled 25 firms.
The U.S. Defense Logistics Agency (DLA) then took over from FEMA and issued a request for proposals for finished gowns. Because of the pressing need, Shawmut decided to extend its efforts into manufacturing finished gowns. It expanded its coalition to include both a high-fashion dressmaker in New York and a maker of high-end climbing suits in Oregon to assist with pattern designs and garment production as well as an upholstery manufacturer, military dress uniform supplier, and mattress company (among others) to supplement sewing capacity.
In parallel, Shawmut devised a novel welding process to assemble gowns in-house using presses designed and built by its engineering and mechanical teams, which normally worked on textile processing lines. Shawmut’s internal gown operation went from a standing start to an output rate of 350,000 gowns per month in just over 90 days, with the capacity to double or triple output again on short notice. Between fabric and finished garments, Shawmut and its coalition partners supported the production of 11 million urgently needed isolation gowns, with most of it occurring during the early stages of the pandemic to address acute shortages.
In contrast to its efforts to manufacture N95 masks, Shawmut’s gown experience highlights an important point: An industrial commons generally spans sectors. Even though most of the textile industry was offshored from the United States by the mid 2000s, residual domestic capabilities still existed because of producers that served specialty markets. Additionally, the Berry Amendment, which gives preference to domestically produced, manufactured, or grown fiber and fabrics for U.S. Department of Defense (DoD) procurements, helped preserve some local manufacturing capabilities.
Because the shutdown of businesses amid the pandemic had idled capacity across the board, Shawmut was able to piece together a domestic supply chain of firms with the requisite capabilities. This responsiveness illustrates the impact of a shared supply base across product sectors — an important feature of a commons. To ensure that such a response becomes possible in other contexts, U.S. officials should look beyond the six areas that were named in President Biden’s executive order and identify important overlaps between sectors and shared skill bases. Such an assessment would help identify capabilities that are in decline and need to be supported across a wide swath of the economy.
A More Resilient Supply Chain is More Expensive
Supply chains that rely on a diverse array of sources are more resilient, but they are also likely to be more expensive if one expects to include higher-cost domestic manufacturers. As a case in point, India’s decision to ban the export of 26 drugs and drug ingredients in February 2020 led to an outcry in the United States; at the same time, U.S. pharmaceutical purchasers were largely unwilling to pay more for the assurance of some domestic supply.
Such outcomes are partially the result of how these goods are bought. Many medical goods that were in short supply during the pandemic flow through group purchasing organizations (GPOs), and consolidation has increased those firms’ purchasing power. Three pharmaceutical benefit managers (PBMs), GPOs for pharmaceutical purchasing, control roughly 80% of the U.S. generics purchased. The raison d’être of these organizations is to drive prices down, not ensure domestic supplies. Their emphasis on price drives purchases to offshore manufacturers, primarily in India and China, that either have lower-cost positions or benefit from government subsidies.
Shawmut faced a similar challenge. Procurement of personal protective equipment (PPE) typically flows through GPOs such as Premier, Vizient, Intalere (which Vizient is in the process of acquiring), and HealthTrust’s Healthcare Group Purchasing Industry Initiative (HGPII). Premier controls more than $67 billion in purchasing, HPGII accounts for $45 billion, and Vizient claims to serve more than 50% of U.S. acute care providers, 95% of academic medical centers, and more than 20% of ambulatory providers that collectively represent $100 billion in annual purchasing.
As is the case with PBMs, they exist to drive costs down, and they historically have been unwilling to pay a premium for domestic sourcing. Until their customers instruct them to do otherwise, their mission will stay focused on delivering the lowest cost. Buyers also favored suppliers with whom they already had relationships and that might offer urgently needed deliveries in exchange for committing to long-term supply contracts. Though this may be a reasonable course of action for both sides, it makes life much harder for new domestic suppliers.
Shawmut fulfilled all orders issued under its emergency contract with the Defense Logistics Agency. Though the contract duration was one year and Shawmut expected to receive new orders, the agency said it had no more demand for gowns. Shawmut’s production stopped on February 15, 2021 — just 140 days after the original award — resulting in layoffs for its workforce and those of other members of its coalition.
This rapid turn of events highlights another problem: Manufacturers in low-cost countries can sell into export markets, while a higher-cost U.S. manufacturer will have difficulty competing and will have to rely mainly on premium-priced domestic demand. Shawmut fell into the role of being a swing producer that was asked to produce during an emergency. But once existing suppliers were able to catch up, it had difficulty winning follow-on business.
The lesson: If the U.S. wants domestic capacity, it will have to find a way to make the business model of domestic production sustainable. Legislation that extended the Berry Amendment to PPE is one approach. It would follow an approach taken by other countries. For example, Medicom, a Canadian producer that imports masks and gowns from China, was recently awarded a $382 million, 10-year fixed volume contract by the Canadian Government to produce 50 million masks a year in Canada as an incentive to build a domestic N95 factory. Medicom received a similar deal from the government of the United Kingdom.
Government purchases, for both ongoing use and routine replenishment of a strategic national stockpile, could be targeted at domestic production to ensure stable and predictable demand for U.S. production despite any price premiums. Such stability would allow manufacturers to plan, amortize their investments, and hire and retain key staff. Tax credits, reimbursement, or other incentives for private-sector hospitals and GPOs could be used to foster additional domestic PPE demand and open up distribution channels.
Regulatory Agencies Need to Work in Partnership
Another challenge facing Shawmut, and new suppliers more generally, relates to the regulatory environment. Hospital gowns are rated from Level 1 to Level 4, according to the ANSI/AAMI PB70 standard for PPE, and must be tested for their liquid-barrier performance. Levels 3 and 4 are used for surgical applications, and these are categorized as Class II medical devices. Manufacturers must file a 510(k) premarket notification with the U.S. Food and Drug Administration (FDA) at least 90 days in advance of marketing such products. The submission process is complex, takes about six months for approval, and the application can cost $100,000 or more to prepare.
This 510(k) filing allows the FDA to determine whether a device is equivalent to one that is already on the market and is therefore acceptable for use. But when the Defense Logistics Agency requested bids for disposable Level 3 and 4 gowns in late summer 2020, it received no proposals from domestic manufacturers that already held the necessary 501(k) approvals, and domestic firms like Shawmut that did not already have a 510(k) approval at the time could not bid. The agency thus went back to Chinese suppliers to purchase millions of Level 3 and Level 4 gowns.
The regulatory approval process for N95 masks was frustrating as well. After Shawmut produced its first mask, it submitted samples to the NIOSH for testing and approval. The approval process is summarized in an 81-page document, which intended to clarify the approval process under the Federal Code 42 CFR Part 84. Even though Shawmut was a global manufacturer with nine ISO-registered manufacturing facilities — two of which are also FDA registered — NIOSH required a detailed audit of the company’s quality systems and documentation. Shawmut’s masks tested well above the NIOSH standard, but the company did not receive approval until February 19, 2021, nearly three months after submission. Meanwhile, Shawmut’s investment in equipment stood idle. Unlike the case of Covid-19 vaccines, there was no apparent urgency around establishing new domestic sources of N95 masks.
Looking forward, policymakers should consider how regulatory processes favor existing producers, most of whom are located outside of the United States. One can imagine that for urgently needed products, testing could be accelerated, as was the case for the FDA’s rapid review and granting of emergency use authorization for three Covid-19 vaccines. As with other steps required to support the development of domestic suppliers, efforts to streamline the regulatory process will require the investment of additional resources.
A Multi-faceted Problem
The pandemic triggered countless efforts at universities, hospitals, and firms across the United States to make products, such as the PPE manufactured by Shawmut, on an ad hoc basis. Yet as was the case for Shawmut, many innovative designs failed to receive widespread support from purchasers and regulators. Even when manufacturers are able to overcome supply-side challenges to rebuild domestic capability, the hurdles created by buyers that continue to rely on pursuing the lowest price and regulators that are unwilling to reconsider bureaucratic processes often prove to be too much to overcome. Absent change and additional support, those innovators that pivoted to help during the Covid-19 pandemic may be less inclined to do the same the next time they are called upon.
We don’t recommend that the United States or any country try to be self-sufficient in everything; there are too many benefits to be derived from global sourcing. But in preparing for the next crisis, it is worth understanding why a country as well-resourced as the United States was slow and, in many cases, unable to respond to basic and critical needs. The goal should be to nurture domestic capabilities that will allow a country to cope during a prolonged crisis such as a pandemic.