America's Advanced Manufacturing Problem and How to Fix It

Link

https://americanaffairsjournal.org/2023/08/americas-advanced-manufacturing-problem-and-how-to-fix-it/

For OSE

• Impressive critique - by people from MIT Open Learning, industrial policy authors
• Hire the boys

Solution

• Invent Integrated Education and Integrated Enterprise to make manufacturing interesting
• Create upward mobility by open source enterprise distribution - DE
• Reinvent tech and college edu (same as first point)
• Cut the hubris of 'clean jobs' - integrated jobs are multifaceted and more integral. Specialization is for insects.
• And of course start with Collaborative Literacy and open innovation
• Create nonprofit defense industries, without offensive capability, relegating offense to other thugs

Article - see below

America’s Advanced Manufacturing Problem—and How to Fix It

by David Adler and William B. Bonvillian

Industrial policy is no longer taboo in the United States.1 In the last two years, the federal government has undertaken multiple industrial‑innovation policy initiatives. The chips and Science Act of 2022 is designed to revitalize domestic production of semiconductors as well as to add an applied science directorate to the National Science Foundation (NSF) focused on advanced technologies. The Executive Order on Bio­tech and Biomanufacturing hopes to ensure that the next generation of medicines are manufactured in the United States. And the Infrastructure Act and the Inflation Reduction Act include massive investments in clean energy technologies.2 These industrial and innovation strategies are underpinned by a common vision, one centered on nurturing advanced technologies. The belief is that these policies will result in the United States once again being able to manufacture critical technologies at home.

But can the United States still make things domestically? What if America lacks capabilities for advanced manufacturing which the most recent round of industrial policies have not fully addressed? These go beyond the lack of resilient supply chains, or the unsurprising fact that a country which has aggressively deindustrialized during the last several decades currently lacks a work­force with skills in advanced manufacturing. What if there are more profound weaknesses inhibiting advanced manufacturing in the United States, as is indicated by the trade deficit in advanced technology products? This deficit is accelerating, growing from $128 billion in 2019, to $195 billion in 2021, to $244 billion in 2022.3 In 2021, the United States accounted for 78 percent of the total trade deficit among industrial nations, while China accounted for 45 percent of the total trade surplus.4 Productivity in manufacturing in the United States, across most measures, has actually been declining.

The United States was once the global leader in manufacturing, ushering in the mass production era from the end of the nineteenth through the mid-twentieth century. It is not a global leader in the advanced manufacturing of the twenty-first century. (Advanced manu­facturing can be defined as the application of innovative technologies to improve manufacturing processes and products, adding significant value through productivity advances and innovation. These would include digital technologies, robotics, 3-D printing, advanced materials, bio-fabrication, artificial intelligence, and nanofabrication.5)

The United States does not currently have the correct institutional infrastructure and accompanying operational mechanisms to support ad­vanced manufacturing. Industry, government, and academia are largely unlinked when it comes to advanced production technology and processes, and there is a similar lack of interagency coordination within the government. Pathways necessary for diffusing new technologies and getting them to market are missing, including a lack of scale-up financing mechanisms. The vocational education system has withered as has the corporate lab system. The Department of Defense’s (DoD) mission has traditionally been one of military security rather than economic security and assuring a strong American industrial base. Yet economic security and military security are now inseparable, and by failing to pursue innovation in production, the DoD is putting U.S. economic and therefore national security at risk. Financial markets do not reward advanced manufacturing. They favor outsourcing and the disaggregation of integrated firms. Corporations are not rewarded for pursuing production as opposed to, say, stock buybacks. What is sometimes called the U.S. developmental state has many strengths—in basic research as well as applications in the areas of defense technology, software, and biopharma development—but advanced manufacturing is not one of them.

Robert J. Gordon, in his widely read book The Rise and Fall of American Growth, argues that recent generations of technologies are inherently less conducive to job creation, compared to earlier breakthroughs, leading to lower growth.6 But the real culprit may be the way in which our innovation system was designed, leaving a manufacturing focus out of the innovation equation. The result of all this has been the decline of U.S. manufacturing and the corresponding weakening of the American working class, growing economic inequality, and protracted political confrontation.

How did we arrive at this state of affairs, and how can we fix it? The design of the existing system has been informed by many factors: which sectors in the United States are seen as prestigious and which are not; false beliefs about the strengths of U.S. manufacturing; beliefs that a postindustrial future is both inevitable and desirable; policy decisions that promoted diffusing technologies globally and not domestic indus­trial strength; the DoD’s assumption that it would always have a robust industrial base to rely upon; hubris and insularity dating from the period when the United States was the unchallenged leader in mass manufacturing; financial market pressures; and the ideological and methodological limitations of mainstream economics.

A pervasive source of weakness is the disconnect between the U.S. production system and the U.S. innovation system. Unlike Germany, Japan, Korea, Taiwan, and now China, the United States failed to put manufacturing at the center of its innovation system. In fact, it didn’t even consider manufacturing as part of innovation. When the United States was assembling its innovation system after World War II, it assumed its lead in manufacturing would be eternal. It built its innovation system in the area it needed to strengthen: R&D. Thus America’s innovation system is in many ways a historical artifact, maladapted to the global economic competition of today.

The challenges America faces in adopting advanced manufacturing practices are, however, surmountable. Overcoming these challenges would allow current and future industrial-innovation policies to reach their full potential and revive domestic manufacturing. Yet these problems are not amenable to an updating of a single process or a management-consultant fix. Solving them will require a major system redesign, relinking innovation to production, building on efforts already underway as well as new initiatives.

America as the Leader in Mass Production In the 1930s, Detroit was the technological and commercial leader of the world’s auto industry. Detroit used mass production techniques con­sisting of assembly lines and standardized products to produce an ever-increasing number of cars at reduced cost. Detroit had nearly double the wages of other industrial centers like Chicago. At the time, nearly five out of six of the world’s cars and trucks were manufactured in the USA.

The most colossal and advanced factory in the world’s most advanced center for car manufacturing was Ford’s River Rouge plant. Here, all production used the assembly line method with conveyor belts bringing together different components of the car in synchronized harmony. Mass production technologies and the division of labor created a workforce consisting of relatively unskilled employees on the line combined with a smaller group of highly skilled workers and the engineering teams that designed it. Visitors from around the world came to view this modern technological marvel, including delegations from Nazi Germany and Soviet Russia. Stefan Link, in his book Forging Global Fordism, writes, “Ford’s new factory became the destination of engineering delegations. Italian, German, Russian, and Japanese special­ists traveled to Detroit and spent weeks, months, even years at River Rouge to learn the American secret of mass production.”7

Mass production was furthermore a dual-use technology. Industrialist Henry Kaiser, who had no background in shipbuilding (he built the Hoover dam and the Oakland–San Francisco bridge) applied mass pro­duction techniques to shipbuilding in World War II. Kaiser shipyards could produce Liberty Ships in a matter of weeks and sometimes even days as opposed to the two-thirds of a year it took to build a comparable ship in the past. The mass production method dramatically raised manufacturing productivity. The United States produced so much war materiel so quickly that it had to scrap more planes after World War II than Nazi Germany was able to produce during the last four years of the war.8

Mass production itself has military origins. The system of interchangeable machine-made parts was implemented in the early 1800s at the federal government’s Harper’s Ferry and Springfield, Massachusetts, arsenals, for the production of firearms. These production techniques then spread to the civilian sector. This precursor to mass production, which used specialized machines for the production of interchangeable parts, was called the American system of manufacturing or sometimes, more specifically, “armory practice.”9 Ford added the moving assembly lines which marked the starting point of true mass production.

Postwar Innovation Policy and Its Assumptions about U.S. Manufacturing The U.S. victory in World War II was supported by rapid advances in science and technology and the scaling of mass production, with central planning combining the two, and the military promoting improved production techniques for private firms. Before the war was even over, President Roosevelt considered the implications of this organized approach to scientific development for peacetime. On November 17, 1944, he wrote a letter to Vannevar Bush, his science adviser and the director of the Office of Scientific Research and Development, stating that Bush’s office “represents a unique experiment coordinating scientific research and in applying scientific knowledge to the solution of the technical problems paramount in war. There is, however, no reason why the lessons to be found in this experiment cannot be profitably employed in times of peace.”10 Roosevelt asked for Bush’s recommendations, adding, “New frontiers of the mind are before us.”

Bush’s report, written in response, Science—the Endless Frontier, is widely credited as the basis for postwar U.S. science and innovation policy.11 It conceived of innovation as a linear pipeline. The federal government would fund basic research and the development of scientific talent. The knowledge would be available to industry which, it was assumed, would develop it further, moving it along the pipeline until it was ready for mass production, eventually reaching the marketplace. The report called for a new central science agency to oversee the research funding.

There were other influences on postwar innovation policy, however, which would prove to be as important as Bush’s report. John Steelman, assistant to President Truman (today’s equivalent of a chief of staff) proposed instead a more decentralized model of research no longer overseen by a central agency, which is how federal research would come to be organized, overturning Bush. Nevertheless, Bush’s emphasis on basic research remained the focus of the U.S. innovation system, at the NSF, other government agencies, and U.S. universities, as did his linear innovation model.12

The Department of Defense was a holdout from the Bush model in that it continued to support prototyping, testing, demonstrating, and implementing technologies in addition to basic research. Particularly after Sputnik, the DoD pursued the more integrated wartime approach, and was also the customer for the new technologies it developed. In contrast, the civilian R&D agencies and universities supported research only through early-stage development. The civilian sector therefore fol­lowed Bush’s basic research architecture as well as Steelman’s decentralized one. The result, effectively, was two innovation systems in the United States, an integrated defense one, and a disjointed civilian one.13

Missing from both new R&D-led innovation systems was any focus on manufacturing or the understanding that manufacturing-led innova­tions could occur along the pipeline rather than just as an afterthought.14 The Endless Frontier report doesn’t contemplate that new technologies could transform manufacturing processes themselves—the word manufacturing only occurs once. The Pentagon likewise didn’t pay that much attention to broad manufacturing capabilities in the United States, nor did it need to, given the country’s recent performance during the war.

Instead, the U.S. postwar approach to innovation rested on the assumption that the United States was—and would remain—the pre­eminent manufacturing power, deploying the most advanced production processes from a position of unchallenged supremacy. After all, the industrial bases of the Axis powers had just been destroyed. The foreign policy establishment, assuming U.S. industrial strength would be endur­ing, made trade deals aimed at building up allies at the expense of domestic manufacturing, a position that has only recently changed.15

The private sector similarly took America’s supremacy in manufacturing for granted. The corporate lab system of AT&T, GE, GM, Ford, Xerox, and IBM, much like the government lab system, focused on technology research, not incremental advances or breakthroughs in manufacturing production. Henry Kressel, who led the central research lab at RCA, explains why: “The returns are deemed faster and higher in focusing on product development and marketing. And there was a lack of trained people—manufacturing carries a stigma for college trained elites. Regarding the old labs, they rarely focused on production technology—with the exception of Corning.”16 The corporate labs declined with the pressures of global competition, with most cutting back their role on potential breakthroughs in favor of safer, shorter-term incremental advances. There has been no replacement for the old lab system when they phased down. And manufacturing technology was never an empha­sis.

Overall, the U.S. federal innovation system of government labs and government-funded academic research was disconnected from manufacturing. On the nondefense side, it emphasized basic rather than applied research. The focus was on scientific breakthroughs, and not on manu­facturing technology. The government did not coordinate the innovative activities it funded across market actors. There were no features spe­cifically designed to move a new technology along the commercial pipeline.

The United States further lacked collaborative government, academic, and industry institutions centered on improving manufacturing tech­niques and sharing related new technologies. The system was therefore very unlike that of agricultural colleges, cited in Endless Frontier as a model, through which government-supported research is transmitted into practice. (The United States remains highly competitive in agriculture to this day.17)

The government planners and businessmen of the postwar era may have presumed that America’s leadership in the prior manufacturing innovation wave—mass production—would be repeated in subsequent waves, but this was not to be. The disconnect in the United States between innovation and production would lead to profound weaknesses. America often missed out on newer production approaches starting with quality manufacturing. As the mass production era was succeeded by new waves of manufacturing advances, America would no longer be king of manufacturing.

The Japan Shock Ford was building cars in Yokohama, Japan, as early as the 1920s. By the 1930s, Ford, together with GM, completely dominated the Japanese auto market. The Japanese government, however, sought to change this, and pursued interventionist policies to foster a locally owned industry. In 1936, Japan enacted the Automobile Manufacturing Enterprise Law favoring domestic carmakers. The American firms were forced to leave Japan. Toyota and Nissan took over their manufacturing equipment.18

In the 1970s and 1980s, Japan’s automakers, with support of the Ministry of International Trade and Industry (MITI), pioneered the “quality production paradigm.” Japanese producers pursued precision production, just-in-time production, quality built into all production stages (not just at the end), continuous improvements by the entire workforce, and collaborative labor practices, all of which led to dramatically increased efficiencies and quality. MITI funded industry rather than basic research to support this new approach. It was Japan’s turn to lead technological innovation in car manufacturing, with Detroit a follower.

Currency manipulation by the Japanese government further aided its exports, and the oil shock of 1973 was helpful to Japanese automakers given that they specialized in smaller, more efficient cars while the United States produced gas-guzzlers. Japan’s exports to the United States soared. U.S. automakers, in contrast, took too long to understand and then adjust their manufacturing technologies and techniques to the quality revolution. Rather than making the necessary changes or investments, the standard reaction of Detroit was to offshore lower-end models to lower-wage countries. Japan, bolstered by the quality revolu­tion and the modernization of its industrial production process, and taking advantage of open U.S. markets while keeping its own closed, leapt ahead in economic competitiveness with the United States in consumer electronics as well as autos.

The U.S. government eventually countered in the 1980s with industrial policies of its own: the Bayh-Dole Act (which encouraged universities to commercialize federally funded research); the Manufacturing Extension Partnership (modeled on the U.S. agriculture extension program to bring advances to small manufacturers); the SBIR program (R&D grant funding to small and start-up companies); sematech (the consortium to restore U.S. semiconductor technology leadership though improved production efficiency and quality); and the R&D tax credit (which benefited firms that increased their R&D investments). Finally, a joint currency intervention between the United States and its allies, the Plaza Accord, resulted in the depreciation of the dollar.

These industrial policies were modestly funded and piecemeal in comparison to Japan’s efforts. Sematech, for instance, only targeted one sector, semiconductors. Generally, these programs targeted early stages of innovation, rather than actual production technologies. The quality revolution in manufacturing of the 1970s and ’80s, pioneered by Japan, largely passed America by—it was only adopted slowly. The country faced a widespread economic malaise and the emergence of the “Rust Belt” in the Midwest.19 The manufacturing competition with Japan was a harbinger of the far graver China shock to come.

The Future Will Be Postindustrial There are reasons why the United States was slow to adopt the quality revolution in manufacturing and the production innovation waves which followed. The handful of industrial policies launched by the U.S. government in response to Japan’s industrial successes were overwhelmed by larger trends, including beliefs, incentives, and policies pushing in the opposite direction, against reviving domestic manufacturing. Simply put, manufacturing was no longer a prestigious sector in the United States.

America’s success in the IT revolution in the 1990s completely diverted U.S. attention from following through on the piecemeal industrial policy reforms of the 1980s. By this point, the United States was moving too slowly, if at all, on new production techniques, and adopted policies that often hastened deindustrialization altogether.

The widespread belief—and prophecy—that the most advanced economy was a postindustrial one made concerns about America’s manufacturing malaise seem irrelevant. Daniel Bell’s influential 1973 book The Coming of Post-Industrial Society: A Venture in Social Forecasting is illustrative of American intellectuals’ thinking about manufacturing during this era.20 The book took the transition from a goods to services economy as a given. Indeed, it is more of a meditation about the features of this transformed society.

As early as the 1960s, Bell argued, “the leadership of the new society will rest, not with businessmen or corporations as we know them, for a good deal of production will have been routinized . . . but rather with intellectual institutions. Social prestige and social status will be rooted in the intellectual and scientific communities.”21 It is quite clear which sectors and activities Bell valued. The idea that production could move beyond the routine is not considered. Instead, manufacturing seems to almost vanish entirely in his conception of the future; it is a relic.

Bell also claimed, “what has now become decisive for society is the new centrality of theoretical knowledge, the primacy of theory over empiricism.”22 This statement contains a great deal of self-insight, about Bell, but also American intellectuals and social scientists. Bell’s theory-driven approach lacked any empirical understanding of what was going on in American manufacturing in the 1970s. The routinized production in the United States he refers to was being globally displaced by the newer type of manufacturing of Japan.

Bell’s book and others like it, such as those from futurist Herman Kahn, were prescient about the future of the United States—but not about the future of East Asian countries. These nations did not embrace the postindustrial paradigm theorized by American intellectuals and academics. More damagingly, the prophecies about a postindustrial American future might have been self-fulfilling.

Workforce Education The state of American vocational education is emblematic of the low prestige associated with production. The U.S. vocational education system of the 1950s, ’60s, and ’70s was not quality education. It shunted off too many students into what became a dead-end track. It was representative of the low status in the United States of factory jobs or jobs which did not require a college degree.

The college-for-all movement, which arose in response, was a well-intentioned reform which created its own problems. It largely dismantled vocational education. It did not, however, replace it with a better system. Instead, academic courses crowded out vocational courses, but most students who were previously on the vocational track did not end up going to college.

Some states and regions maintained and improved their vocational schools. Today, the United States has a small number of effective vocational-ed high schools, often with long waiting lists for admissions. These typically put students into programs that combine work and learning. But these remaining programs are the exceptions.

Instead, the education system for America’s technical workforce is largely broken.23 Yet without a workforce trained in the skills required for new technologies, those technologies cannot be implemented. The overall problems include: a history (until recently) of disinvestment by government and employers in workforce education; Labor Department training programs that do not reach either higher technical skills or incumbent workers that need to upskill; Education Department programs that are focused on college not workforce needs and not linked to the Labor Department programs; underfunded community colleges that lack the resources to provide advanced training in new fields and have too-low completion rates; colleges and universities that are disconnected from workforce education; no system for lifelong learning; underfunded advanced technical education programs (at the NSF and at advanced manufacturing institutes); and a broken labor market information system. These problems are compounded by the reality that the existing actors are in a “legacy” sector—the long-established education sector is hard to change.

There is a burgeoning literature on innovation ecosystems as essential to technology-based growth. However, the role of workforce education as a factor in innovation ecosystems has largely been missed by these analysts.24 It is time to put it into the equation. Overall, the United States has never created a widespread or effective work-learn model supporting skilled manufacturing jobs, akin to the vocational and apprenticeship programs of other industrial countries like Germany or Japan. But the issues go beyond just the collapse of vocational education in the United States. In addition, the American higher educational sys­tem is largely disconnected from manufacturing technology and pro­cesses—exactly as Vannevar Bush designed.

Economists’ Perspectives In the 1970s, when American heavy industry was decimated by the combination of the oil shock and the rise of East Asian exporters, neoclassical economists could offer little advice about how to stem the decline. The Keynesian demand regime which had successfully managed the U.S. economic cycle over the previous thirty years now only pro­duced stagflation and was no longer tenable. The neoliberal approach which replaced it, with its emphasis on privatization, deregulation, monetarism, and free trade, also was unable to revive manufacturing—and was not particularly interested in doing so. These inadequacies stem from both the methodological limitations and ideological biases of mainstream economics.

Macroeconomic theory has been unable to adequately model innova­tion, nor does it deeply grapple with innovation systems. Nevertheless, there have been some attempts to include innovation in growth models: Robert Solow’s models identify growth as coming predominantly from technological progress in addition to labor and capital, but this is treated as an exogenous input, a black box with the model offering no further insight. Other models try to endogenize innovation, incorporating learning by doing.25 But the details of America’s innovation system and its weaknesses are essentially alien to macroeconomics and not part of this field.

In addition to these methodological flaws, ideological preferences creep into modeling. A standard assumption is sector agnosticism. Manufacturing is treated just like any other sector, despite the evidence of the importance of manufacturing innovation to long-term growth. Although manufacturing is only 11 percent of U.S. GDP and 8 percent of direct employment, it drives 20 percent of capital investment, 30 percent of productivity growth, 60 percent of exports, and 70 percent of business R&D.26 Yet economists have often held manufacturing in disdain, with deindustrialization a sign of progress akin to sociologists’ prophecies. Alan Blinder, former member of Clinton’s Council of Eco­nomic Advisers and vice-chair of the Fed, wrote in 2005, “The shift to services is still viewed with alarm in America and many other rich countries, where people bemoan rather than welcome the resulting losses of manufacturing jobs.”27

There are parallel beliefs, misconceptions, and dogmas in trade theory, though its recommendations were largely ignored until the neoliberal era. The United States was historically a high tariff nation, only reducing tariffs once it reached economic supremacy after World War II. It turned a blind eye to the mercantilist trade practices of South Korea, Germany, and Japan (until the imbalances with Japan became too outsized and the Plaza Accord was enacted) because these were allied nations during the Cold War. The United States offered these countries access to its markets for geopolitical reasons. The result too often was one-sided free trade.

Then, during the neoliberal era, ideology suffused into policy. The United States extended the same courtesy of domestic market access to China, not an ally, that it had previously offered Germany and Japan. It did so for many reasons: the desire to change China; hubris after winning the Cold War; pressure from U.S. corporations; but also because standard trade models showed this to be net-welfare improving for the United States. N. Gregory Mankiw, chairman of George W. Bush’s Council of Economic Advisers, in his 2004 Economic Report to the President, stated “when a good or service is produced more cheaply abroad, it makes more sense to import it than make or provide it domestically.”28 Mankiw told reporters, “offshoring is the latest mani­festation of the gains from trade that economists have talked about for centuries.”29

Since then, there have been changes in trade policy attempting to reverse this neoliberal approach. Theory, however, has generally lagged policy with some exceptions. Economists’ apparent disdain for manufacturing is still a constant. Mankiw’s indifference to America’s manufacturing sector remains representative of economists’ thinking. Although manufacturing jobs result in significantly higher net employment than services jobs when indirect jobs are included, the report treats manufacturing as just another sector. It is oblivious to the underlying weaknesses in U.S. manufacturing and its many causes, which are not addressed. The 2004 Economic Report’s advice: “The best policy response to recent developments in manufacturing is to focus on stimulating the overall economy.”

Mis-Incentives and Misconceptions Though economics is conspicuous for its lack of insights about what exactly is wrong with American manufacturing, this doesn’t mean economic incentives are not key. Financialization and globalization combined to reduce incentives for manufacturers to adopt advanced production techniques in the United States.

American financial markets favor a capital-light production model, or one of no production at all. Jack Welch was well aware of this in his transformation of GE from an engineering company—one cofounded by Thomas Edison—to a financial engineering company. In 1985, during his early years as CEO, he canceled the company’s “factory of the future” initiative consisting of automation systems, robotics, and ad­vanced machine tools, which would have allowed GE to catch up with Siemens and Japanese competitors in its offerings.30 Instead, that same year he acquired RCA, parent company of the television broadcaster NBC. Eventually, 60 percent of company profits came from GE capital. Welch’s strategy worked, at least for a while: in 1993, GE became the world’s most valuable company, before the stock price collapsed during the great financial crisis and GE was broken up.

The assessment of author David Gelles, in his critical biography of Welch, The Man Who Broke Capitalism, is that “instead of trying to fix American manufacturing, he effectively abandoned it, and would soon start shuttering factories around the country and shipping jobs overseas.”31 Welch once stated, “Ideally, you’d have every plant you own on a barge to move with currencies and changes in the economy.”

This dream might have been unthinkable in a previous age, when American elites had more of a national commitment. But it would have been unfeasible without the disaggregation of production and the rise of the internet conveying information. Political economist Suzanne Berger has explained that IT-based precise specifications and designs enabled manufacturing, which previously had to be vertically integrated, to be distributed across many sites. Manufacturing could take place in coun­tries which subsidized it through direct grants, cheap electricity,32 and manipulated currencies. None of these factors favored investing in technological improvements in U.S. production. Ultimately, the IT revolution made the whole manufacturing issue moot in the United States. The future was the Information Age, and platform companies which manufactured nothing. Manufacturing didn’t matter. It was a legacy sector at best. Daniel Bell’s prophecy about a postindustrial future seemed true.

Paradoxically, buried in this disdain for manufacturing were long-held assumptions about U.S. strengths, updated with beliefs that IT was transforming American production. The unprecedented decline in U.S. manufacturing employment, falling by 28 percent between 2000 and 2016, was often interpreted as resulting from automation rather than trade with China—what would later be called the China shock—and other Asian nations. This faulty analysis, which still remains widespread, rests on the mistaken belief that U.S. manufacturing output remained solid. In fact, statistical anomalies involving how growth in computer processing capabilities is credited to growth in manufacturing output gave a misleadingly positive picture of the overall state of U.S. manufacturing.33 There is no need to turn to explanations about automation to explain the seeming disconnect between manufacturing employment and output, because there is no actual discrepancy: both declined from the China shock.

In fact, the United States has been slow in adopting factory automation compared to leading competitor nations. All evidence points toward weakness. According to the International Federation of Robotics, the world’s top five most automated countries in manufacturing in 2021 were South Korea, Singapore, Japan, Germany, and China. The United States ranks a distant ninth, just slightly ahead of Slovenia in robot density per worker, a key indicator of automation.34 The United States is not close to being as dominant in advanced manufacturing as it was in mass manufacturing during the first half of the twentieth century.

Nonetheless, even today, there is a burgeoning genre of white papers from management consultants which insists that U.S. manufacturing is undergoing a technological renaissance. Often published as “native advertising” (sponsored content) in the financial press, they discuss the Third Industrial Revolution—using IT to automate production—and now the Fourth (4IR), which involves biological processes, blockchain, virtual reality, and more. The most prominent of these white papers are published in partnership with the World Economic Forum,35 and the recommendations are distinctly neoliberal, meaning they are CEO-led rather than government led. There is no analysis of the design flaws of the overall U.S. innovation system, including its delinking from produc­tion, nor of the successes of the Chinese system, including the heavy use of subsidies. Instead, resilience, sustainability, and inclusivity are the key words of this corporate literature. They can appear in any order. Words that do not make an appearance include industrial policy, subsidies, mercantilism, and the CCP. Reading about this imminent Fourth Industrial Revolution, one would never imagine that during the Covid pandemic the United States struggled to produce the most basic PPE gear and doctors resorted to wearing plastic bags.

Case Study: Clean Energy Often for the United States, the choice hasn’t been between mass production and advanced production, but whether to pursue any domestic production at all. The federal government continues to fund basic research as Vannevar Bush planned, but the production chain beyond that is broken. Most manufacturing—and products developed based on innovations in manufacturing processes—takes place outside the United States.

The clean energy sector is just one example of many. Despite the initial invention of key technologies, the United States has ceded manufacturing leadership of one energy sector after another, from solar to offshore wind to batteries to nuclear power. Jonas Nahm in his book Collaborative Advantage36 analyzes these differences, comparing the United States, Germany, and China’s experiences in developing clean-energy industries.37

“In the United States firms have typically taken the form of start-ups with skills in the invention of new technologies, but with far fewer capabilities in the commercialization and production of these inventions,” Nahm writes. In contrast, he found Germany’s highly skilled small manufacturers had prowess in the customization of small batch production of equipment and components. China pursued what Nahm terms “innovative manufacturing”: Chinese manufacturers “focused on the R&D required for scaling and commercializing novel technologies. Far fewer firms prioritized the production of manufacturing equipment or the invention of new technologies.”

As a result of China’s strength in innovative manufacturing processes and scale-up, supported by its financial system, demand policies, and workforce training—as well as massive production subsidies, local content requirements, and forced technology transfer—only China cur­rently manufactures new energy technologies like solar and increasingly batteries and soon EVs at worldwide scale. Although Nahm sees the new global system as a positive because it presents new opportunities for firms and countries to specialize and collaborate, the United States misses out on the large potential base of employment from, and control of, the technologies it invented.

The Innovation System Revisited—by China Today, it is China which is the pacesetter for auto manufacturing, not Germany or Japan, and certainly not Detroit. China dominates the production of full electric vehicles and, in 2023, surpassed Japan as the world’s largest auto exporter. As soon as 2025, Europe could become a net importer of cars, given the rapid growth of electric vehicles made in China. Some of these are Chinese brands but others are European (and American) brands that have shifted their EV manufacturing to China. China leads in battery production and offers well-developed supply chains, a vast domestic market, and technological expertise.

The Wall Street Journal observed, “Thanks to competition and [China’s] focus on execution, the EV industry went from a niche industrial policy project to a sprawling ecosystem of predominantly private companies. Much of this happened below the Western radar.”38 The Journal also referenced a statement from a Chinese entrepreneur who said, “Ideas are not important in China—execution is.” This attention to execution in production is very different from the basic research ambitions and goals of America’s innovation system. But China’s seemingly overnight success in electric vehicles derives from something else, too: the industrial and frequently outright mercantilist policies39 it has successfully used in industry after industry. These include market access and consumer financing contingent upon local production, forced technology transfer, exclusion of foreign battery producers, and massive subsidies through government guidance funds.40

China’s emphasis on manufacturing is the exact opposite of the postindustrial discourse which has been dominant in the United States from World War II until very recently. “Industry is the main engine of economic growth. Industry is the main battlefield of technological innovation,” writes Party Secretary Jin Zhuanglong, head of the Minis­try of Industry and Information Technology. His article, “Accelerate the Promotion of New Industrialization,” published in an official CCP theory journal, further stated, “keep the proportion of manufacturing industry to GDP basically stable, and prevent the economy from ‘moving away from the real to the virtual.’ Promoting new industrialization is urgent to build a large country’s competitive advantage. It is the foundation for winning the initiative in international economic competi­tion.”

The United States, however, is no longer sitting still. It has put in place significant industrial policies under the Biden administration, some of which passed with bipartisan support in 2021 and 2022. China is not sitting still, either.41 Its industrial policies include the equivalent of up to $400 billion in annual scale-up financing for growing industries—the United States has no comparable program.42 It is rapidly reforming its own innovation system, including addressing its needs in basic research. This is the last missing link in its “innovation chain,” one required for Chinese technological self-sufficiency. Funding for basic research increased 24 percent in 2021 compared to a year earlier.

“China’s policy makers are seeking to systematically address and integrate every step of the innovation process,” notes a 2023 report by the German think tank merics.43 “These include launching large projects, promoting industry-university-research synergy, and integrating the innovation, industrial, capital, and talent chains.”

Basic as well as applied scientific research is directed towards national strategic goals such as technological self-sufficiency according to the merics report. “China’s government has brought the work of State Key Laboratories closer to commercialization by embedding labs in compa­nies including Huawei and ZTE.” With government backing, China is reproducing in its way the vanished U.S. corporate lab system. Even basic research conducted at universities shares these strategic and commercial imperatives. Under the National Key R&D Projects pro­gram, which provides research funding, “there is no strict division between funding for basic research and support for developing practical applications,” according to the merics report. China is now second in the world in R&D spending and on a path to pass the United States.44 Unlike in the United States, in China there are combined thrusts for invention, innovation, and production, integrated with trade policy.

Implementing Advanced Manufacturing Policies As of 2021, China’s manufacturing sector accounted for a 28 percent share of the value add of its GDP. In the United States, manufacturing makes up only 11 percent of the value add of its GDP. Roughly 40 to 45 percent of China’s manufacturing output is directed to exports.45

Furthermore, the United States ran a trade deficit in advanced technology goods, largely with Asian nations (predominantly China), as noted above, that grew to $244 billion in 2022. This is not a deficit in commodity goods; it is a deficit in the goods the United States needs for technology and innovation leadership. And this is an accelerating prob­lem for the United States, not a declining or stabilizing one.

Although it’s long been in denial, the United States is now recognizing that it has a deep manufacturing problem. If it is going to tackle this problem, it must shift to advanced manufacturing methods to get the kind of productivity and efficiency gains needed to compete in these goods. The issue now is how to do this. A systematic approach is needed; trying to tackle the problem by relying on just R&D capabilities will not be enough. Trade policy, for instance, will be part of any successful solution and is necessary to protect manufacturing from predatory pricing by mercantilist competitors. We focus here, however, on pragmatic steps the United States could take to strengthen its innovation chain. The general theme, usually overlooked in standard policy analysis, is relinking the innovation system to production.

Policy discussions in recent years increasingly focus on “direction­ality,” that is, the idea that policies should set the direction for technol­ogy development, as opposed to simply letting markets work their will. But what role should government play in setting that direction for manufacturing? And how can “directional” policy decisions be pursued given the complex mix of actors (from industry in numerous sectors to universities to government) and the highly federalized political system of the United States?

Given this complexity, it is important for manufacturing policy design to bring together the critical actors in the system. These include industry, including small manufacturing firms and large; government, including federal, state, and local; and educators, including universities and community colleges. The tool kits of each set of actors will all be needed to form effective policies. Below are ten steps that the federal government, working with the other actors, could undertake.

(1) Improve the Manufacturing Institutes. Although policymakers began to see that American manufacturing was facing challenges when Japan’s quality model began making inroads into U.S. auto and consumer electronics sectors in the 1970s and 80s, only limited policy steps were taken. Industry, while trying to copy Japan on quality, continued to believe that the overall solutions to its problems lay in tax and trade policies. It was not until the Obama administration that an innovation-based set of policies started to evolve, with the creation of sixteen manufacturing innovation institutes during the period of 2012 to 2017.46 The motivating concept was that U.S. manufacturers could only compete with lower-cost Asian producers, particularly from China, if they became much more efficient and productive, to offset their cost and wage disadvantages. The institutes were collaborations, with federal core funding cost-shared with industry, universities, and state governments. Each institute was organized around a particular key manufacturing technology strand—robotics, digital production, bio-fabrication, ad­vanced composites, etc.—and they were loosely modeled on Germany’s Fraunhofer Institutes, which play a comparable role. Congress has recently approved three more institutes. While the institute model involves the critical actors required for manufacturing innovation listed above, program details need attention.47

End the term limits. The institutes have never had the budget or long-term commitment that a manufacturing technology transformation requires. At the outset of the program, the institutes were placed on term limits, with federal support set to end after five years. This was a political decision but was never realistic: the structural problems in manufacturing that the institutes were designed to tackle cannot be solved in five years; they represent long-term challenges. The term limits forced the institutes to organize around the needs of large companies—the only source that could sustain them over time when the government pulled out—which meant that they had to neglect the smaller firms that were furthest behind in adopting new technologies and limit their workforce education investments. The term limits should be ended, with institutes subject to rigorous review at the end of their current terms; if they are successful they should be extended with comparable funding. The Commerce and Defense Departments have now adopted this approach, although the Energy Department has not, and many of its institutes have become shells with very limited programs. While Commerce has reviewed and fully funded its first institute, DoD has reviewed its institutes but renewed them at significantly lower funding levels. These problems require correction.

Network the institutes. In addition, efforts need to be networked across the institutes. Companies don’t want to adopt advanced manufacturing for one technology at a time just because they were developed at separate, stove-piped institutes. Instead, institutes need to cooperate to offer packages of their technology advances to firms. For example, firms don’t want to adopt just robots; they want robotics integrated with digital production technologies and other technologies such as 3-D printing or advanced composites. We need an institute network function to pull together advances from across institutes and package them for easier adaption.

Collaborate closely with state and local governments. Although the institutes must have a national mission, manufacturing is highly regional, and the states and local governments play the leading roles in economic development and workforce education. Collaborations between them and the institutes are ongoing but deepening them will be key to institute success. New manufacturing technologies need to be available for national adoption but tested and piloted regionally, which means institutes must operate at both national and regional levels.

Undertake testing, demonstration, and certification. The Defense Department has long supported the testing and demonstration stages, key steps for technology acceptance and scaling. Most manufacturing institutes have created technology centers and should receive additional funding to undertake rigorous testing and demonstration of new tech­nologies developed by participating companies in the areas the institutes support, such as robotics or 3-D Printing. Part of these demonstrations should be to identify the cost and efficiency savings to manufacturers these technologies can provide. After effective demonstration, the insti­tutes should certify and validate the feasability of the new technologies, which can help with prompt market acceptance.

(2) Back R&D for manufacturing technologies. We also need to enlist our federal R&D agencies in the cause. Manufacturing is not a subject of their research, but it needs to be, so technology advances move into the manufacturing institutes for experimentation and adop­tion. The federal basic research agencies work at what DoD and NASA have characterized as “Technology Readiness Levels” (TRL) 1–3, meaning they support early-stage research up to proof-of-concept experimentation. Since industry typically devotes most of its resources to the scale-up and initial production stages—TRL levels 7–9—there is a major gap (the infamous “valley of death”) in between. The manufacturing institutes are designed for these in-between levels, from development and prototyping to technology demonstration (TRL levels 4–6).

But if the institutes focus at these post-research stages, over time they will need new input from earlier stage research. Our R&D agencies have historically shunned research on manufacturing technologies—that’s been viewed as industry’s purview, although industry emphasizes later-stage development not research. We need to see manufacturing as a system, with connections throughout, from research to production.

Although the research agencies never developed significant portfolios around manufacturing, now is the time for them to do so if we are to have a connected innovation system to spur advanced manufacturing. This means, too, that they will need to include applied work that adds manufacturing solutions to their earlier-stage research. We need a cross-agency agenda for R&D on manufacturing technologies and processes at our R&D agencies, linked to the development initiatives at the institutes. Otherwise, over time, tech development at the institutes will become stranded and thin out.

(3) Provide scale-up financing. While venture capital has been a major force for financing innovation in the United States, it is now focused overwhelmingly on software, biotech, and various services sectors. It is largely out of “hard tech”—innovations that must be manufactured. This means there are few mechanisms to scale up manufacturing production in the United States. In recent decades, firms have been shifting prototyping and production of hard-tech goods to China.

A significant manufacturing gap in the United States is seen in its lack of entrepreneurial start-up firms in this sector, which are needed to advance innovation. The share of young manufacturing firms (less than five years old) among U.S. manufacturers has steadily declined over the past three decades, falling far below the levels in sectors supported by VC, such as biotech, software, or IT.48 Yet newer firms tend to adopt and develop new manufacturing technologies and processes at a higher rate than established firms. Manufacturing is missing among new entrants, and scale-up financing could help encourage their entry into manufacturing and corresponding production innovation.

If the United States is to shift to advanced manufacturing at both old firms and new, it will need scale-up financing to support it.49 In contrast to the United States, China has long provided massive manufacturing scale-up assistance to its firms, which accounts for much of its dominance of world manufacturing output. What form could scale-up fi­nancing take in the United States? Operation Warp Speed in 2020 used guaranteed contracts to Covid-19 vaccine makers to reduce their risks and assure production. Senator Chris Coons and colleagues have pro­posed an industrial finance corporation for innovative manufacturing.50 The DoD in 2022 created a new “Office of Strategic Capital” for tech­nology scale-up, although it appears focused more on national security rather than broader economic security goals.51 The Biden administration has proposed repurposing an established federal bank, the Ex-Im Bank, to provide manufacturing scale-up support alongside its long-standing export financing role.52 The Depart­ment of Energy’s Loan Programs Office, since 2005, has provided financing for scale-up of new energy technologies. While the program suffered a political backlash for a failed loan to a thin-film solar company, Solyndra, in 2011, it has had numerous successes. The most notable was a $465 million loan to Tesla in 2009–13 for expansion of its Fremont, California, production site, which enabled the company to avoid bankruptcy and accelerated the scale-up of electric vehicles in the United States. In addition, states in their economic development role have often provided financing to firms to support regional job creation. In contrast to the United States,53 China has long provided massive manufacturing scale-up assistance to its firms, which accounts for much of its dominance of world manufacturing output.

There are thus a range of possible mechanisms available, from forming a new industrial bank to expanding the role of existing institutions. Applying a number of these mechanisms with expanded lending authority for advanced manufacturing scale-up could be a solution to this glaring manufacturing system gap.54

(4) Use government procurement power to promote new manufacturing technologies. The DoD is by far the largest procurement agency, and it needs to strengthen its industrial base and build more resilient supply chains for increasingly pressing national security needs. The productivity and efficiency gains possible from advanced manufacturing therefore should be a priority. DoD historically played a role in fostering new manufacturing technologies, such as the interchangeable machine-made parts developed at Army arsenals in the 1840s. A more recent example is computer numerically controlled (CNC) equipment, pervasive now in all manufacturing sites. It was first developed through DoD-supported research at MIT. When DoD saw the new level of precision manufacturing it enabled, it required its contractors to implement CNC machining for its missile programs, which spread this advance throughout U.S. industry.

DoD, working with manufacturing innovation institutes, could identify productivity savings available from new technologies such as digital production advances, robotics, or 3-D printing, and require its defense contractors, through specifications and contracts, to implement them. It also has authority under the Defense Production Act to acquire these new technologies and lease them back to its contractors. While DoD procurement does not dominate U.S. manufacturing, it has a significant market share. DoD could adopt the approach of mandating improvements, along with funding support for new equipment, that it used with CNC machining, enabling larger-scale adoption of the new manufacturing technologies.

(5) Direct production support. In the case of some critical technologies, we may need more direct production support to build factories. We need a more sustained chips Act, and for other critical sectors, not just semiconductors.

In the 2022 chips Act, the federal government provided $39 billion in grants as well as loan guarantees and investment tax credits to semiconductor makers to fund new advanced production facilities. Similarly, the federal government, through the Commerce or Defense Departments, could provide support for firms to acquire and install new advanced manufacturing technologies. This could be applied particularly toward production of critical or national-security-related technologies.55

(6) Provide both “top-down” and “bottom-up” support. Many support tools are going to be needed for a manufacturing revival and will involve a series of organizations and agencies, from the advanced manufacturing institutes, to the R&D agencies noted above, and the Defense and Commerce Departments. Each has existing policy imple­mentation mechanisms and could acquire new ones based on the recommendations here. These tools can be offered to manufacturing firms in two basic forms. Top-down tools have long been employed by darpa and were also employed by Operation Warp Speed. Here, the government identifies a technology challenge to be addressed, and picks a small portfolio of leading companies to pursue it. The DOE’s funding of eight companies for fusion development is yet another example. Bottom-up tools involve incentives widely available to interested com­panies to meet a technology challenge—it’s up to the firms to pursue them. For example, the Inflation Reduction Act provides tax credits for purchases of EVs manufactured in America if the vehicles meet sourcing and processing requirements for critical materials used in their batteries. Tesla notably took advantage of many other bottom-up tools in rising to EV technology leadership.56 The manufacturing chal­lenge is complex so both top-down and bottom-up approaches will be needed.

(7) Build a manufacturing focus into existing industrial policy programs. The United States has recently passed a major package of industrial policy programs (including the energy demonstrations in the Infrastructure Act, the chips and Science Act, and the Inflation Reduction Act) with over $500 billion in new funding. Yet programs specifically focused on manufacturing technologies are largely missing from these initiatives. While the new legislation aims to rebuild U.S. supply chains, if the bulk of this manufacturing does not occur in the United States, supply-chain resilience will remain elusive. In effect, we need to put the “industrial” into these industrial policy programs. Each requires a policy focus on manufacturing. And future industrial policy approaches require more of a specific manufacturing focus. Rather than just incentivizing the growth of a particular manufacturing sector through subsidies or tax credits, this would also mean incentivizing adoption of advanced manufacturing processes such as digital technologies, robotics, artificial intelligence, 3-D printing, or bio-fabrication, both in general and in those sectors.

(8) Map and fill gaps in supply chains. The Covid pandemic and the resulting economic fallout exposed the weaknesses of U.S. supply chains not only in protective equipment, pharmaceuticals, and pharmaceutical materials but in a host of areas, from semiconductors to critical materials and supplier manufacturing capabilities. A fundamental lesson from Operation Warp Speed in 2020 was the need to better map supply chains in order to assure production capability, and the Defense Production Act enabled the Defense Department, working closely with industry, to intervene and fill gaps to assure sufficient vaccine production to meet the emergency. Similarly, in 2020, the executive branch, through a series of participating agencies and the White House, began a major effort to identify and fill supply-chain gaps in advanced batteries, pharmaceuticals and pharmaceutical ingredients, semiconductors, and critical miner­als.57 Particularly for the production of critical and nation­al-securi­ty‑related technologies, relevant federal agencies could work with industry to support the mapping of supply chains for implementing advanced manufacturing technologies and processes.

(9) Fix workforce education. Germany has long gained productivity improvements from a famously well-trained manufacturing workforce based on an apprenticeship system. In contrast, U.S. companies have generally tried to get productivity gains from capital, plant, and equip­ment investments and ignored the workforce side. German firms under­stand, however, that productivity gains from new equipment can soon spread worldwide, while a gain from a high-quality workforce will be enduring and provides a long-term competitive edge.

The United States has a broken workforce-education system, with a deep disconnect between the education system and workplaces. If the United States wants to adopt advanced manufacturing, its workforce must be ready for it. This requires rebuilding much of workforce education at all levels.58 Community colleges must introduce advanced manufacturing curricula, create short programs more adapted to upskilling workers already in the workforce, establish certificates around particular skills that stack toward degrees, and turn around low completion rates. At the federal level, disconnected Labor and Education Department programs need to be integrated and efforts to expand registered apprenticeship programs accelerated. At the industry level, firms must collaborate with each other and with community colleges to build new training and apprenticeship programs, including youth apprenticeship starting in high school. Solutions will have to be pursued throughout the U.S. federal system, since education is largely state- and local-government led. These governments need to improve community college funding and support, integrate advanced manufacturing pro­grams across community colleges, promote apprenticeships, and work closely with area firms on reforms and curricula.

(10) Put someone in charge. The above steps require a series of agencies to act in concert, not an easy task in our stove-piped government, and ways must be found to pull these varied manufacturing pieces together. An interagency committee will be inadequate for the task. The Obama administration in 2008, faced with the bankruptcy of most of the U.S. auto industry, named a manufacturing czar to supervise federal support and put the industry back together, who then began to attack manufacturing challenges more broadly.59 A comprehensive program to spur and implement new manufacturing technologies and processes, with varied mechanisms of support, as well as new trade approaches, calls for such a position in a permanent White House office. Operation Warp Speed provides guidance on how to organize it. OWS was a task force with a specific mission that included key representatives from relevant agencies led by a highly experienced industry leader and senior Army logistics expert, as well as an independent and expert supporting staff. That model could work here, assuring White House authority, links to agency powers, and independent leadership. Since the mission will require industry collaboration, an advisory group of industry, engi­neering, university, and labor leaders could be formed.60

This is not an exclusive list for what needs to be done through government policy for the United States to shift to advanced manufacturing, but it marks a beginning. Most are relatively manageable initial steps, many within reach of the existing policymaking process and not requiring new laws. Nearly all require collaboration between the actors in the U.S. manufacturing system—industry, government, and education—and there will be no substitute for committed companies. The remedies here are largely federal but some are shared across federal, state, and local governments. But more “directionality” from new government policies will be required for this shift to advanced manufacturing.

A Rare Opportunity Only about twenty countries have managed to move from developing to developed-nation status since World War II. Fourteen of them applied industrial policies based on governmental interventions.61 This approach has been most famously expressed in the East Asian development model. Directed and targeted interventionist policies were key to Japan’s recovery after World War II and were also applied by Korea, Taiwan, Singapore, and in recent decades by China. All these countries relied on strong manufacturing through industrial policy approaches to drive economic growth. Since these examples suggest that the restoration of manufacturing prowess requires interventionist government policies, there are lessons here for the United States about the kinds of policies it needs to consider.

U.S. global power paralleled its rise in manufacturing power. It was the first nation to introduce interchangeable machine-made parts before the mid-nineteenth century. It grew this breakthrough in technology and process into its system of mass production, which led the world in manufacturing by the early twentieth century. By the end of World War II, it had achieved production supremacy—no other nation was close to U.S. manufacturing capacity and output. Through this manufacturing advance it forged a deep connection between national security and economic security with manufacturing as the enabler of both. But this link, never explicit, has been neglected since the end of the Cold War. America’s problems with, and indeed indifference to, manufacturing have historical and societal roots, which predate the rise of China and the offshoring of manufacturing. In some ways, offshoring is the culmi­nation of the trend, not the cause. But by missing a focus on manufacturing, the United States has missed an innovation capability step that many other leading nations have grasped.

The recent industrial policies of the Biden administration mark a sea change in America’s approach. They are attempts to address both the technological competition with China and the need for new energy technologies.62 These initiatives are important and necessary—but they are not sufficient. No nation can maintain a world-power position while walking away from manufacturing. Major gaps remain in advanced manufacturing and scale-up financing, which are not a significant focus of the new programs compared to the level of effort needed. The United States is now pursuing a series of industrial policies, although so far they are light on the “industrial.” For the United States to successfully adopt advanced manufacturing at scale, its industrial innovation policies need to address production directly. China is facing its own challenges, including adverse demographic trends, rising wages, regulatory uncer­tainties, and wasted resources, but a negative shift in China’s growth model does not mean the United States will succeed in manufacturing.

The way forward is for the United States to redesign its innovation system, so that innovation is more closely linked to production. It can’t just focus on basic R&D as has been the case since World War II. The United States needs to retain these basic research capabilities, but at the same time sharpen its focus on execution. Overall, the United States needs to more fully integrate government policies, trade, finance, education, innovation, and production.

There is a rare opportunity to do this. Economic security, including the technological competition with China, as well as the quest to renew American economic leadership, provide a strategic imperative.63 But there is something more. The United States created breakthroughs in technologies which can transform advanced manufacturing: 3-D print­ing, advanced composites, new materials, biomanufacturing, photonics, power electronics, and other emerging fields including AI manufacturing.64 We are at a once-in-a-century moment where we could fundamentally change the way we undertake production.

Whereas EU nations define advanced manufacturing primarily in terms of current digital technologies, the federal government’s modest manufacturing programs are researching and pursuing a broader host of technologies—a more revolutionary approach. The United States may have fallen behind in implementing today’s advanced production tech­niques, but with the right political commitment, new scale-up financing and other tools, and tighter linking of innovation and production, America may be able to leapfrog to the frontier of an industrial trans­formation.

This article originally appeared in American Affairs Volume VII, Number 3 (Fall 2023): 3–30. Notes 1 Reda Cherif, Marc Engher, and Fuad Hasanov, “Crouching Beliefs, Hidden Biases: The Rise and Fall of Growth Narratives,” Working Paper No. 20/228, International Monetary Fund, November 2020; Reda Cherif and Fuad Hasanov, “The Return of the Policy That Shall Not Be Named: Principles of Industrial Policy,” Working Paper No. 2019/074, International Monetary Fund, March 2019, 23–24; Reka Juhasz et al., “The Who, What, When and How of Industrial Policy: A Text-Based Approach,” Structural Transformation and Economic Growth (STEG) Working Paper 050, January 12, 2023, 22. 2 For details on each, see William B. Bonvillian, “Industrial Innovation Policy in the United States,” Annals of Science and Technology Policy 6, no. 4 (2022): 315–411.

3 “Trade in Goods with Advanced Technology Products,” U.S. Census Bureau, 2023.

4 Ian Clay, “China Exporter—United States Importer,” Information Technology and Innovation Foundation, February 28, 2023.

5 Manufacturing.gov, the federal government’s advanced manufacturing portal, defines advanced manufacturing as “Use of innovative technologies to create existing products and the creation of new products. Advanced manufacturing can include production activities that depend on information, automation, computation, software, sensing, and networking.”

6 Robert J. Gordon, The Rise and Fall of American Growth (Princeton: Princeton University Press, 2016).

7 Stefan J. Link, Forging Global Fordism: Nazi Germany, Soviet Russia, and the Contest over the Industrial Order (Princeton: Princeton University Press, 2020), 3.

8 Dwight Jon Zimmerman, “Henry J. Kaiser and the Liberty Ships,” Defense Media Network, June 24, 2021.

9 David Hounshell, From the American System to Mass Production, 1800–1932: The Development of Manufacturing Technology in the United States (Baltimore: Johns Hopkins University Press, 1984), 3.

10 Vannevar Bush, Science—the Endless Frontier: A Report to the President (Washington, D.C.: United States Government Printing Office, 1945), reprints Roosevelt letter of November 17, 1944.

11 Bush, Science—the Endless Frontier.

12 William B. Bonvillian, “Encompassing the Innovation Panoply,” Issues in Science and Technology 38, no. 2 (Winter 2022): 37–43.

13 Bonvillian, “Encompassing the Innovation Panoply.”

14 William B. Bonvillian and Peter L. Singer, Advanced Manufacturing: The New American Innovation Policies (Cambridge: MIT Press, 2018), ch. 2.

15 Jake Sullivan, “Remarks on Renewing American Economic Leadership,” White House, April 27, 2023; Robert Lighthizer, No Trade is Free (New York: Broadside Books, 2023).

16 Henry Kressel, email interview, 2023. See also Kressel, “Edison’s Legacy: Industrial Laboratories and Innovation,” American Affairs 1, no. 4 (Winter 2017): 115–29.

17 Bonvillian and Singer, Advanced Manufacturing, n. 12.

18 Jeffrey A. Hart, “A Comparative Analysis of the Sources of America’s Relative Economic Decline,” Understanding American Economic Decline, eds. Michael A. Bernstein and David E. Adler (Cambridge: Cambridge University Press, 1994), 207.

19 Bonvillian and Singer, Advanced Manufacturing, ch. 3.

20 Daniel Bell, The Coming of Post-Industrial Society: A Venture in Social Forecasting (New York: Basic Books, 1973).

21 Daniel Bell, “Notes on the Post-Industrial Society,” Public Interest (Winter 1967): 27.

22 Bell, “Notes on the Post-Industrial Society,” 28.

23 See William B. Bonvillian and Sanjay E. Sarma, Workforce Education: A New Roadmap (Cambridge: MIT Press, 2021).

24 Paul Lewis, “Innovation, Technician Skills, and Vocational Education and Training: Connecting Innovation Systems and Vocational Education and Training,” Journal of Vocational Education and Training, May 27, 2023.

25 For more, see Bonvillian and Singer, “Advanced Manufacturing,” ch. 4.

26 Rana Foroohar, “Why Manufacturing Matters to Economic Superpowers,” Financial Times, April 14, 2021.

27 Alan S. Blinder, “Fear of Offshoring,” CEPS Working Paper No. 119, December 2005.

28 U.S. Council of Economic Advisers, Annual Report of the Council of Economic Advisers: Economic Report of the President (Washington, D.C.: United States Government Printing Office, 2004).

29 Jonathan Weisman, “Bush Adviser Assailed for Stance on ‘Offshoring’ Jobs,” Washington Post, February 11, 2004.

30 Peter Petre, “How GE Bobbled the Factory of the Future,” Fortune, November 11, 1985.

31 David Gelles, The Man Who Broke Capitalism: How Jack Welch Gutted the Heartland and Crushed the Soul of Corporate America―and How to Undo His Legacy (New York: Simon and Schuster, 2022).

32 Suzanne Berger, How We Compete (New York: Crown Business, 2005).

33 David Adler, “The Real Challenge for U.S. Industry,” City Journal, March 29, 2017.

34 “China Overtakes USA in Robot Density,” International Federation of Robotics, December 5, 2022.

35 See Klaus Schwab, “The Fourth Industrial Revolution: What It Means, How to Respond,” World Economic Forum, January 14, 2016; “Fourth Industrial Revolution,” World Economic Forum, 2023.

36 Jonas Nahm, Collaborative Advantage: Forging Green Industries in the New Global Economy (Oxford: Oxford University Press, 2021).

37 Although Nahm’s book predates the major energy investments in the 2022 Inflation Reduction Act, that act focused on consumer and industry tax and other incentives for spreading green technologies, not on manufacturing capability. Although the chips and Science Act calls for regional innovation hubs to try to spread innovation mastery to new areas, potentially including production capability, it is not yet funded. So Nahm’s analysis remains relevant for understanding comparative approaches to innovation and production.

38 Greg Ip, “China’s EV Juggernaut Is a Warning for the West,” Wall Street Journal, June 7, 2023; Zeyi Yang, “How Did China Come to Dominate Electric Cars?,” MIT Technology Review, February 21, 2023.

39 Gregor Sebastian and Francois Chimits, “‘Made in China’ Electric Vehicles Could turn Sino-EU Trade on its Head,” merics, May 30, 2022.

40 David Adler, “Guiding Finance, China’s Strategy for Funding Advanced Manufacturing,” American Affairs 6, no. 2 (Summer 2022): 17–40.

41 Barry Naughton, Siwen Xiao, and Yaosheng Xu, “The Trajectory of China’s Industrial Policies,” Working Paper, UC Institute on Global Conflict and Cooperation, June 2, 2023.

42 Gerard DiPippo et al., Red Ink: Estimating Chinese Industrial Policy Spending in Comparative Perspective (Washington, D.C.: Centre for Strategic and International Studies, 2022).

43 Jeroen Groenewegen-Lau and Michael Laha, “Controlling the Innovation Chain: China’s Strategy to Become a Science and Technology Superpower,” merics, Feb 2, 2023.

44 Bailey Crane, “China’s Drive for Leadership in Global Research and Development,” Center for Strategic and International Studies, June 30, 2023.

45 Andrew Batson, “Breaking Down China’s Manufacturing,” The Tangled Woof (blog), June 13, 2023.

46 See, generally, Bonvillian and Singer, Advanced Manufacturing.

47 For additional discussion of needed improvements, see William B. Bonvillian, “Ensuring Manufacturing USA Reaches Its Potential,” Federation of American Scientists, August 10, 2021.

48 Ben Armstrong, Suzanne Berger, and Bill Bonvillian, Advanced Technology, Advanced Training (Cambridge: MIT Initiative for Knowledge and Innovation in Manufacturing, 2021): 23–24.

49 Although investments in manufacturing plant construction have accelerated this year (see Javier David, “1 Big Thing: A Manufacturing Supercycle is Starting,” Axios, June 16, 2023) this is largely spending in the semiconductor sector driven by the chips Act, and for several EV and battery plants mostly in the South by foreign investors; it is not broad-based.

50 “Sen. Coons, Colleagues Seek to Create New Domestic Manufacturing Investment Corporation,” Office of Senator Chris Coons, news release, August 12, 2021.

51 C. Todd Lopez, “New Defense Office Connects with Next-Gen Tech Developers with Much-Needed Capital,” U.S. Department of Defense, December 1, 2022.

52 “Statement From the Export Import Bank of the U.S. President and Chair Reta Jo Lewis on the One Year Anniversary of President Biden’s Supply Chain Executive Order,” Export-Import Bank of the United States, February 24, 2022.

53 Tesla’s loan from the Department of Energy (DOE) was approved in 2009 and was available in January of 2010. Tesla repaid the loan in 2013, ahead of schedule; see “Tesla: Loan Programs Office,” U.S. Department of Energy, accessed July 22, 2023; Tim Higgins, Power Play: Tesla, Elon Musk and the Bet of the Century (New York: Doubleday, 2021): 109–12, 159, 169, 194.

54 Gerard DiPippo et al., Red Ink; David Adler, “Guiding Finance.”

55 The chips and Science Act identified ten critical technology areas for government programs to focus on, with a process to periodically update this list (See, HR 4346, 117th Cong., 2nd Sess., Subtitle G, 2021; Tim Clancy, “Chips and Science Act Enshrines Policy for New NSF Technology Directorate,” American Institute of Physics, November 23, 2022.

56 Bonvillian, “Industrial Innovation Policy in the United States.”

57 “Building Resilient Supply Chains, Revitalizing American Manufacturing, and Fostering Broad-Based Growth,” White House, June 2021.

58 See detailed recommendations in William B. Bonvillian and Sanjay E. Sarma, Workforce Education: A New Roadmap (Cambridge: MIT Press, 2021).

59 Noam Scheiber, “Manufacturing Bloom,” New Republic, December 7, 2009.

60 The Advanced Manufacturing Partnership, an Obama administration panel of industry, university, and labor representatives that developed advanced manufacturing policies in 2011–14, provides a model (Executive Office of the President, “President Obama Launches Advanced Manufacturing Partnership,” news release, June 24, 2011).

61 Reda Cherif and Fuad Hasanov, “The Return of the Policy That Shall Not Be Named.” See also David Oks and Henry Williams, “The Long, Slow Death of Global Development,” American Affairs 6, no. 4 (Winter 2022): 122–50.

62 Bonvillian, “Industrial Innovation Policy in the United States.”

63 Tai Ming Cheung and Thomas G. Mahnken, The Decisive Decade: United States–China Competition in Defense Innovation and Defense Industrial Policy in and beyond the 2020s (Washington, D.C.: Center for Strategic and Budget Assessments, 2023); Sullivan, “Remarks on Renewing American Economic Leadership.”

64 The Georgia AI Manufacturing (ga-aim) coalition, led by the Georgia Tech Research Corporation, will receive approximately $65 million from the Commerce Department’s Economic Development Administration to accelerate the adoption of artificial intelligence across the state’s legacy industrial sectors. See U.S. Economic Development Corporation, “U.S. Department of Commerce Invests Approximately $65 Million to Accelerate Integration of Artificial Intelligence Technologies in Industry in Georgia through American Rescue Plan Regional Challenge,” news release, September 2, 2022.

Embed

America’s Advanced Manufacturing Problem—and How to Fix It

by David Adler and William B. Bonvillian

Industrial policy is no longer taboo in the United States.¹ In the last two years, the federal government has undertaken multiple industrial‑innovation policy initiatives. The chips and Science Act of 2022 is designed to revitalize domestic production of semiconductors as well as to add an applied science directorate to the National Science Foundation (NSF) focused on advanced technologies. The Executive Order on Bio­tech and Biomanufacturing hopes to ensure that the next generation of medicines are manufactured in the United States. And the Infrastructure Act and the Inflation Reduction Act include massive investments in clean energy technologies.² These industrial and innovation strategies are underpinned by a common vision, one centered on nurturing advanced technologies. The belief is that these policies will result in the United States once again being able to manufacture critical technologies at home.

But can the United States still make things domestically? What if America lacks capabilities for advanced manufacturing which the most recent round of industrial policies have not fully addressed? These go beyond the lack of resilient supply chains, or the unsurprising fact that a country which has aggressively deindustrialized during the last several decades currently lacks a work­force with skills in advanced manufacturing. What if there are more profound weaknesses inhibiting advanced manufacturing in the United States, as is indicated by the trade deficit in advanced technology products? This deficit is accelerating, growing from $128 billion in 2019, to $195 billion in 2021, to $244 billion in 2022.³ In 2021, the United States accounted for 78 percent of the total trade deficit among industrial nations, while China accounted for 45 percent of the total trade surplus.⁴ Productivity in manufacturing in the United States, across most measures, has actually been declining.

The United States was once the global leader in manufacturing, ushering in the mass production era from the end of the nineteenth through the mid-twentieth century. It is not a global leader in the advanced manufacturing of the twenty-first century. (Advanced manu­facturing can be defined as the application of innovative technologies to improve manufacturing processes and products, adding significant value through productivity advances and innovation. These would include digital technologies, robotics, 3-D printing, advanced materials, bio-fabrication, artificial intelligence, and nanofabrication.⁵)

The United States does not currently have the correct institutional infrastructure and accompanying operational mechanisms to support ad­vanced manufacturing. Industry, government, and academia are largely unlinked when it comes to advanced production technology and processes, and there is a similar lack of interagency coordination within the government. Pathways necessary for diffusing new technologies and getting them to market are missing, including a lack of scale-up financing mechanisms. The vocational education system has withered as has the corporate lab system. The Department of Defense’s (DoD) mission has traditionally been one of military security rather than economic security and assuring a strong American industrial base. Yet economic security and military security are now inseparable, and by failing to pursue innovation in production, the DoD is putting U.S. economic and therefore national security at risk. Financial markets do not reward advanced manufacturing. They favor outsourcing and the disaggregation of integrated firms. Corporations are not rewarded for pursuing production as opposed to, say, stock buybacks. What is sometimes called the U.S. developmental state has many strengths—in basic research as well as applications in the areas of defense technology, software, and biopharma development—but advanced manufacturing is not one of them.

Robert J. Gordon, in his widely read book The Rise and Fall of American Growth, argues that recent generations of technologies are inherently less conducive to job creation, compared to earlier breakthroughs, leading to lower growth.⁶ But the real culprit may be the way in which our innovation system was designed, leaving a manufacturing focus out of the innovation equation. The result of all this has been the decline of U.S. manufacturing and the corresponding weakening of the American working class, growing economic inequality, and protracted political confrontation.

How did we arrive at this state of affairs, and how can we fix it? The design of the existing system has been informed by many factors: which sectors in the United States are seen as prestigious and which are not; false beliefs about the strengths of U.S. manufacturing; beliefs that a postindustrial future is both inevitable and desirable; policy decisions that promoted diffusing technologies globally and not domestic indus­trial strength; the DoD’s assumption that it would always have a robust industrial base to rely upon; hubris and insularity dating from the period when the United States was the unchallenged leader in mass manufacturing; financial market pressures; and the ideological and methodological limitations of mainstream economics.

A pervasive source of weakness is the disconnect between the U.S. production system and the U.S. innovation system. Unlike Germany, Japan, Korea, Taiwan, and now China, the United States failed to put manufacturing at the center of its innovation system. In fact, it didn’t even consider manufacturing as part of innovation. When the United States was assembling its innovation system after World War II, it assumed its lead in manufacturing would be eternal. It built its innovation system in the area it needed to strengthen: R&D. Thus America’s innovation system is in many ways a historical artifact, maladapted to the global economic competition of today.

The challenges America faces in adopting advanced manufacturing practices are, however, surmountable. Overcoming these challenges would allow current and future industrial-innovation policies to reach their full potential and revive domestic manufacturing. Yet these problems are not amenable to an updating of a single process or a management-consultant fix. Solving them will require a major system redesign, relinking innovation to production, building on efforts already underway as well as new initiatives.

America as the Leader in Mass Production

In the 1930s, Detroit was the technological and commercial leader of the world’s auto industry. Detroit used mass production techniques con­sisting of assembly lines and standardized products to produce an ever-increasing number of cars at reduced cost. Detroit had nearly double the wages of other industrial centers like Chicago. At the time, nearly five out of six of the world’s cars and trucks were manufactured in the USA.

The most colossal and advanced factory in the world’s most advanced center for car manufacturing was Ford’s River Rouge plant. Here, all production used the assembly line method with conveyor belts bringing together different components of the car in synchronized harmony. Mass production technologies and the division of labor created a workforce consisting of relatively unskilled employees on the line combined with a smaller group of highly skilled workers and the engineering teams that designed it. Visitors from around the world came to view this modern technological marvel, including delegations from Nazi Germany and Soviet Russia. Stefan Link, in his book Forging Global Fordism, writes, “Ford’s new factory became the destination of engineering delegations. Italian, German, Russian, and Japanese special­ists traveled to Detroit and spent weeks, months, even years at River Rouge to learn the American secret of mass production.”⁷

Mass production was furthermore a dual-use technology. Industrialist Henry Kaiser, who had no background in shipbuilding (he built the Hoover dam and the Oakland–San Francisco bridge) applied mass pro­duction techniques to shipbuilding in World War II. Kaiser shipyards could produce Liberty Ships in a matter of weeks and sometimes even days as opposed to the two-thirds of a year it took to build a comparable ship in the past. The mass production method dramatically raised manufacturing productivity. The United States produced so much war materiel so quickly that it had to scrap more planes after World War II than Nazi Germany was able to produce during the last four years of the war.⁸

Mass production itself has military origins. The system of interchangeable machine-made parts was implemented in the early 1800s at the federal government’s Harper’s Ferry and Springfield, Massachusetts, arsenals, for the production of firearms. These production techniques then spread to the civilian sector. This precursor to mass production, which used specialized machines for the production of interchangeable parts, was called the American system of manufacturing or sometimes, more specifically, “armory practice.”⁹ Ford added the moving assembly lines which marked the starting point of true mass production.

Postwar Innovation Policy and Its Assumptions about U.S. Manufacturing

The U.S. victory in World War II was supported by rapid advances in science and technology and the scaling of mass production, with central planning combining the two, and the military promoting improved production techniques for private firms. Before the war was even over, President Roosevelt considered the implications of this organized approach to scientific development for peacetime. On November 17, 1944, he wrote a letter to Vannevar Bush, his science adviser and the director of the Office of Scientific Research and Development, stating that Bush’s office “represents a unique experiment coordinating scientific research and in applying scientific knowledge to the solution of the technical problems paramount in war. There is, however, no reason why the lessons to be found in this experiment cannot be profitably employed in times of peace.”¹⁰ Roosevelt asked for Bush’s recommendations, adding, “New frontiers of the mind are before us.”

Bush’s report, written in response, Science—the Endless Frontier, is widely credited as the basis for postwar U.S. science and innovation policy.¹¹ It conceived of innovation as a linear pipeline. The federal government would fund basic research and the development of scientific talent. The knowledge would be available to industry which, it was assumed, would develop it further, moving it along the pipeline until it was ready for mass production, eventually reaching the marketplace. The report called for a new central science agency to oversee the research funding.

There were other influences on postwar innovation policy, however, which would prove to be as important as Bush’s report. John Steelman, assistant to President Truman (today’s equivalent of a chief of staff) proposed instead a more decentralized model of research no longer overseen by a central agency, which is how federal research would come to be organized, overturning Bush. Nevertheless, Bush’s emphasis on basic research remained the focus of the U.S. innovation system, at the NSF, other government agencies, and U.S. universities, as did his linear innovation model.¹²

The Department of Defense was a holdout from the Bush model in that it continued to support prototyping, testing, demonstrating, and implementing technologies in addition to basic research. Particularly after Sputnik, the DoD pursued the more integrated wartime approach, and was also the customer for the new technologies it developed. In contrast, the civilian R&D agencies and universities supported research only through early-stage development. The civilian sector therefore fol­lowed Bush’s basic research architecture as well as Steelman’s decentralized one. The result, effectively, was two innovation systems in the United States, an integrated defense one, and a disjointed civilian one.¹³

Missing from both new R&D-led innovation systems was any focus on manufacturing or the understanding that manufacturing-led innova­tions could occur along the pipeline rather than just as an afterthought.¹⁴ The Endless Frontier report doesn’t contemplate that new technologies could transform manufacturing processes themselves—the word manufacturing only occurs once. The Pentagon likewise didn’t pay that much attention to broad manufacturing capabilities in the United States, nor did it need to, given the country’s recent performance during the war.

Instead, the U.S. postwar approach to innovation rested on the assumption that the United States was—and would remain—the pre­eminent manufacturing power, deploying the most advanced production processes from a position of unchallenged supremacy. After all, the industrial bases of the Axis powers had just been destroyed. The foreign policy establishment, assuming U.S. industrial strength would be endur­ing, made trade deals aimed at building up allies at the expense of domestic manufacturing, a position that has only recently changed.¹⁵

The private sector similarly took America’s supremacy in manufacturing for granted. The corporate lab system of AT&T, GE, GM, Ford, Xerox, and IBM, much like the government lab system, focused on technology research, not incremental advances or breakthroughs in manufacturing production. Henry Kressel, who led the central research lab at RCA, explains why: “The returns are deemed faster and higher in focusing on product development and marketing. And there was a lack of trained people—manufacturing carries a stigma for college trained elites. Regarding the old labs, they rarely focused on production technology—with the exception of Corning.”¹⁶ The corporate labs declined with the pressures of global competition, with most cutting back their role on potential breakthroughs in favor of safer, shorter-term incremental advances. There has been no replacement for the old lab system when they phased down. And manufacturing technology was never an empha­sis.

Overall, the U.S. federal innovation system of government labs and government-funded academic research was disconnected from manufacturing. On the nondefense side, it emphasized basic rather than applied research. The focus was on scientific breakthroughs, and not on manu­facturing technology. The government did not coordinate the innovative activities it funded across market actors. There were no features spe­cifically designed to move a new technology along the commercial pipeline.

The United States further lacked collaborative government, academic, and industry institutions centered on improving manufacturing tech­niques and sharing related new technologies. The system was therefore very unlike that of agricultural colleges, cited in Endless Frontier as a model, through which government-supported research is transmitted into practice. (The United States remains highly competitive in agriculture to this day.¹⁷)

The government planners and businessmen of the postwar era may have presumed that America’s leadership in the prior manufacturing innovation wave—mass production—would be repeated in subsequent waves, but this was not to be. The disconnect in the United States between innovation and production would lead to profound weaknesses. America often missed out on newer production approaches starting with quality manufacturing. As the mass production era was succeeded by new waves of manufacturing advances, America would no longer be king of manufacturing.

The Japan Shock

Ford was building cars in Yokohama, Japan, as early as the 1920s. By the 1930s, Ford, together with GM, completely dominated the Japanese auto market. The Japanese government, however, sought to change this, and pursued interventionist policies to foster a locally owned industry. In 1936, Japan enacted the Automobile Manufacturing Enterprise Law favoring domestic carmakers. The American firms were forced to leave Japan. Toyota and Nissan took over their manufacturing equipment.¹⁸

In the 1970s and 1980s, Japan’s automakers, with support of the Ministry of International Trade and Industry (MITI), pioneered the “quality production paradigm.” Japanese producers pursued precision production, just-in-time production, quality built into all production stages (not just at the end), continuous improvements by the entire workforce, and collaborative labor practices, all of which led to dramatically increased efficiencies and quality. MITI funded industry rather than basic research to support this new approach. It was Japan’s turn to lead technological innovation in car manufacturing, with Detroit a follower.

Currency manipulation by the Japanese government further aided its exports, and the oil shock of 1973 was helpful to Japanese automakers given that they specialized in smaller, more efficient cars while the United States produced gas-guzzlers. Japan’s exports to the United States soared. U.S. automakers, in contrast, took too long to understand and then adjust their manufacturing technologies and techniques to the quality revolution. Rather than making the necessary changes or investments, the standard reaction of Detroit was to offshore lower-end models to lower-wage countries. Japan, bolstered by the quality revolu­tion and the modernization of its industrial production process, and taking advantage of open U.S. markets while keeping its own closed, leapt ahead in economic competitiveness with the United States in consumer electronics as well as autos.

The U.S. government eventually countered in the 1980s with industrial policies of its own: the Bayh-Dole Act (which encouraged universities to commercialize federally funded research); the Manufacturing Extension Partnership (modeled on the U.S. agriculture extension program to bring advances to small manufacturers); the SBIR program (R&D grant funding to small and start-up companies); sematech (the consortium to restore U.S. semiconductor technology leadership though improved production efficiency and quality); and the R&D tax credit (which benefited firms that increased their R&D investments). Finally, a joint currency intervention between the United States and its allies, the Plaza Accord, resulted in the depreciation of the dollar.

These industrial policies were modestly funded and piecemeal in comparison to Japan’s efforts. Sematech, for instance, only targeted one sector, semiconductors. Generally, these programs targeted early stages of innovation, rather than actual production technologies. The quality revolution in manufacturing of the 1970s and ’80s, pioneered by Japan, largely passed America by—it was only adopted slowly. The country faced a widespread economic malaise and the emergence of the “Rust Belt” in the Midwest.¹⁹ The manufacturing competition with Japan was a harbinger of the far graver China shock to come.

The Future Will Be Postindustrial

There are reasons why the United States was slow to adopt the quality revolution in manufacturing and the production innovation waves which followed. The handful of industrial policies launched by the U.S. government in response to Japan’s industrial successes were overwhelmed by larger trends, including beliefs, incentives, and policies pushing in the opposite direction, against reviving domestic manufacturing. Simply put, manufacturing was no longer a prestigious sector in the United States.

America’s success in the IT revolution in the 1990s completely diverted U.S. attention from following through on the piecemeal industrial policy reforms of the 1980s. By this point, the United States was moving too slowly, if at all, on new production techniques, and adopted policies that often hastened deindustrialization altogether.

The widespread belief—and prophecy—that the most advanced economy was a postindustrial one made concerns about America’s manufacturing malaise seem irrelevant. Daniel Bell’s influential 1973 book The Coming of Post-Industrial Society: A Venture in Social Forecasting is illustrative of American intellectuals’ thinking about manufacturing during this era.²⁰ The book took the transition from a goods to services economy as a given. Indeed, it is more of a meditation about the features of this transformed society.

As early as the 1960s, Bell argued, “the leadership of the new society will rest, not with businessmen or corporations as we know them, for a good deal of production will have been routinized . . . but rather with intellectual institutions. Social prestige and social status will be rooted in the intellectual and scientific communities.”²¹ It is quite clear which sectors and activities Bell valued. The idea that production could move beyond the routine is not considered. Instead, manufacturing seems to almost vanish entirely in his conception of the future; it is a relic.

Bell also claimed, “what has now become decisive for society is the new centrality of theoretical knowledge, the primacy of theory over empiricism.”²² This statement contains a great deal of self-insight, about Bell, but also American intellectuals and social scientists. Bell’s theory-driven approach lacked any empirical understanding of what was going on in American manufacturing in the 1970s. The routinized production in the United States he refers to was being globally displaced by the newer type of manufacturing of Japan.

Bell’s book and others like it, such as those from futurist Herman Kahn, were prescient about the future of the United States—but not about the future of East Asian countries. These nations did not embrace the postindustrial paradigm theorized by American intellectuals and academics. More damagingly, the prophecies about a postindustrial American future might have been self-fulfilling.

Workforce Education

The state of American vocational education is emblematic of the low prestige associated with production. The U.S. vocational education system of the 1950s, ’60s, and ’70s was not quality education. It shunted off too many students into what became a dead-end track. It was representative of the low status in the United States of factory jobs or jobs which did not require a college degree.

The college-for-all movement, which arose in response, was a well-intentioned reform which created its own problems. It largely dismantled vocational education. It did not, however, replace it with a better system. Instead, academic courses crowded out vocational courses, but most students who were previously on the vocational track did not end up going to college.

Some states and regions maintained and improved their vocational schools. Today, the United States has a small number of effective vocational-ed high schools, often with long waiting lists for admissions. These typically put students into programs that combine work and learning. But these remaining programs are the exceptions.

Instead, the education system for America’s technical workforce is largely broken.²³ Yet without a workforce trained in the skills required for new technologies, those technologies cannot be implemented. The overall problems include: a history (until recently) of disinvestment by government and employers in workforce education; Labor Department training programs that do not reach either higher technical skills or incumbent workers that need to upskill; Education Department programs that are focused on college not workforce needs and not linked to the Labor Department programs; underfunded community colleges that lack the resources to provide advanced training in new fields and have too-low completion rates; colleges and universities that are disconnected from workforce education; no system for lifelong learning; underfunded advanced technical education programs (at the NSF and at advanced manufacturing institutes); and a broken labor market information system. These problems are compounded by the reality that the existing actors are in a “legacy” sector—the long-established education sector is hard to change.

There is a burgeoning literature on innovation ecosystems as essential to technology-based growth. However, the role of workforce education as a factor in innovation ecosystems has largely been missed by these analysts.²⁴ It is time to put it into the equation. Overall, the United States has never created a widespread or effective work-learn model supporting skilled manufacturing jobs, akin to the vocational and apprenticeship programs of other industrial countries like Germany or Japan. But the issues go beyond just the collapse of vocational education in the United States. In addition, the American higher educational sys­tem is largely disconnected from manufacturing technology and pro­cesses—exactly as Vannevar Bush designed.

Economists’ Perspectives

In the 1970s, when American heavy industry was decimated by the combination of the oil shock and the rise of East Asian exporters, neoclassical economists could offer little advice about how to stem the decline. The Keynesian demand regime which had successfully managed the U.S. economic cycle over the previous thirty years now only pro­duced stagflation and was no longer tenable. The neoliberal approach which replaced it, with its emphasis on privatization, deregulation, monetarism, and free trade, also was unable to revive manufacturing—and was not particularly interested in doing so. These inadequacies stem from both the methodological limitations and ideological biases of mainstream economics.

Macroeconomic theory has been unable to adequately model innova­tion, nor does it deeply grapple with innovation systems. Nevertheless, there have been some attempts to include innovation in growth models: Robert Solow’s models identify growth as coming predominantly from technological progress in addition to labor and capital, but this is treated as an exogenous input, a black box with the model offering no further insight. Other models try to endogenize innovation, incorporating learning by doing.²⁵ But the details of America’s innovation system and its weaknesses are essentially alien to macroeconomics and not part of this field.

In addition to these methodological flaws, ideological preferences creep into modeling. A standard assumption is sector agnosticism. Manufacturing is treated just like any other sector, despite the evidence of the importance of manufacturing innovation to long-term growth. Although manufacturing is only 11 percent of U.S. GDP and 8 percent of direct employment, it drives 20 percent of capital investment, 30 percent of productivity growth, 60 percent of exports, and 70 percent of business R&D.²⁶ Yet economists have often held manufacturing in disdain, with deindustrialization a sign of progress akin to sociologists’ prophecies. Alan Blinder, former member of Clinton’s Council of Eco­nomic Advisers and vice-chair of the Fed, wrote in 2005, “The shift to services is still viewed with alarm in America and many other rich countries, where people bemoan rather than welcome the resulting losses of manufacturing jobs.”²⁷

There are parallel beliefs, misconceptions, and dogmas in trade theory, though its recommendations were largely ignored until the neoliberal era. The United States was historically a high tariff nation, only reducing tariffs once it reached economic supremacy after World War II. It turned a blind eye to the mercantilist trade practices of South Korea, Germany, and Japan (until the imbalances with Japan became too outsized and the Plaza Accord was enacted) because these were allied nations during the Cold War. The United States offered these countries access to its markets for geopolitical reasons. The result too often was one-sided free trade.

Then, during the neoliberal era, ideology suffused into policy. The United States extended the same courtesy of domestic market access to China, not an ally, that it had previously offered Germany and Japan. It did so for many reasons: the desire to change China; hubris after winning the Cold War; pressure from U.S. corporations; but also because standard trade models showed this to be net-welfare improving for the United States. N. Gregory Mankiw, chairman of George W. Bush’s Council of Economic Advisers, in his 2004 Economic Report to the President, stated “when a good or service is produced more cheaply abroad, it makes more sense to import it than make or provide it domestically.”²⁸ Mankiw told reporters, “offshoring is the latest mani­festation of the gains from trade that economists have talked about for centuries.”²⁹

Since then, there have been changes in trade policy attempting to reverse this neoliberal approach. Theory, however, has generally lagged policy with some exceptions. Economists’ apparent disdain for manufacturing is still a constant. Mankiw’s indifference to America’s manufacturing sector remains representative of economists’ thinking. Although manufacturing jobs result in significantly higher net employment than services jobs when indirect jobs are included, the report treats manufacturing as just another sector. It is oblivious to the underlying weaknesses in U.S. manufacturing and its many causes, which are not addressed. The 2004 Economic Report’s advice: “The best policy response to recent developments in manufacturing is to focus on stimulating the overall economy.”

Mis-Incentives and Misconceptions

Though economics is conspicuous for its lack of insights about what exactly is wrong with American manufacturing, this doesn’t mean economic incentives are not key. Financialization and globalization combined to reduce incentives for manufacturers to adopt advanced production techniques in the United States.

American financial markets favor a capital-light production model, or one of no production at all. Jack Welch was well aware of this in his transformation of GE from an engineering company—one cofounded by Thomas Edison—to a financial engineering company. In 1985, during his early years as CEO, he canceled the company’s “factory of the future” initiative consisting of automation systems, robotics, and ad­vanced machine tools, which would have allowed GE to catch up with Siemens and Japanese competitors in its offerings.³⁰ Instead, that same year he acquired RCA, parent company of the television broadcaster NBC. Eventually, 60 percent of company profits came from GE capital. Welch’s strategy worked, at least for a while: in 1993, GE became the world’s most valuable company, before the stock price collapsed during the great financial crisis and GE was broken up.

The assessment of author David Gelles, in his critical biography of Welch, The Man Who Broke Capitalism, is that “instead of trying to fix American manufacturing, he effectively abandoned it, and would soon start shuttering factories around the country and shipping jobs overseas.”³¹ Welch once stated, “Ideally, you’d have every plant you own on a barge to move with currencies and changes in the economy.”

This dream might have been unthinkable in a previous age, when American elites had more of a national commitment. But it would have been unfeasible without the disaggregation of production and the rise of the internet conveying information. Political economist Suzanne Berger has explained that IT-based precise specifications and designs enabled manufacturing, which previously had to be vertically integrated, to be distributed across many sites. Manufacturing could take place in coun­tries which subsidized it through direct grants, cheap electricity,³² and manipulated currencies. None of these factors favored investing in technological improvements in U.S. production. Ultimately, the IT revolution made the whole manufacturing issue moot in the United States. The future was the Information Age, and platform companies which manufactured nothing. Manufacturing didn’t matter. It was a legacy sector at best. Daniel Bell’s prophecy about a postindustrial future seemed true.

Paradoxically, buried in this disdain for manufacturing were long-held assumptions about U.S. strengths, updated with beliefs that IT was transforming American production. The unprecedented decline in U.S. manufacturing employment, falling by 28 percent between 2000 and 2016, was often interpreted as resulting from automation rather than trade with China—what would later be called the China shock—and other Asian nations. This faulty analysis, which still remains widespread, rests on the mistaken belief that U.S. manufacturing output remained solid. In fact, statistical anomalies involving how growth in computer processing capabilities is credited to growth in manufacturing output gave a misleadingly positive picture of the overall state of U.S. manufacturing.³³ There is no need to turn to explanations about automation to explain the seeming disconnect between manufacturing employment and output, because there is no actual discrepancy: both declined from the China shock.

In fact, the United States has been slow in adopting factory automation compared to leading competitor nations. All evidence points toward weakness. According to the International Federation of Robotics, the world’s top five most automated countries in manufacturing in 2021 were South Korea, Singapore, Japan, Germany, and China. The United States ranks a distant ninth, just slightly ahead of Slovenia in robot density per worker, a key indicator of automation.³⁴ The United States is not close to being as dominant in advanced manufacturing as it was in mass manufacturing during the first half of the twentieth century.

Nonetheless, even today, there is a burgeoning genre of white papers from management consultants which insists that U.S. manufacturing is undergoing a technological renaissance. Often published as “native advertising” (sponsored content) in the financial press, they discuss the Third Industrial Revolution—using IT to automate production—and now the Fourth (4IR), which involves biological processes, blockchain, virtual reality, and more. The most prominent of these white papers are published in partnership with the World Economic Forum,³⁵ and the recommendations are distinctly neoliberal, meaning they are CEO-led rather than government led. There is no analysis of the design flaws of the overall U.S. innovation system, including its delinking from produc­tion, nor of the successes of the Chinese system, including the heavy use of subsidies. Instead, resilience, sustainability, and inclusivity are the key words of this corporate literature. They can appear in any order. Words that do not make an appearance include industrial policy, subsidies, mercantilism, and the CCP. Reading about this imminent Fourth Industrial Revolution, one would never imagine that during the Covid pandemic the United States struggled to produce the most basic PPE gear and doctors resorted to wearing plastic bags.

Case Study: Clean Energy

Often for the United States, the choice hasn’t been between mass production and advanced production, but whether to pursue any domestic production at all. The federal government continues to fund basic research as Vannevar Bush planned, but the production chain beyond that is broken. Most manufacturing—and products developed based on innovations in manufacturing processes—takes place outside the United States.

The clean energy sector is just one example of many. Despite the initial invention of key technologies, the United States has ceded manufacturing leadership of one energy sector after another, from solar to offshore wind to batteries to nuclear power. Jonas Nahm in his book Collaborative Advantage³⁶ analyzes these differences, comparing the United States, Germany, and China’s experiences in developing clean-energy industries.³⁷

“In the United States firms have typically taken the form of start-ups with skills in the invention of new technologies, but with far fewer capabilities in the commercialization and production of these inventions,” Nahm writes. In contrast, he found Germany’s highly skilled small manufacturers had prowess in the customization of small batch production of equipment and components. China pursued what Nahm terms “innovative manufacturing”: Chinese manufacturers “focused on the R&D required for scaling and commercializing novel technologies. Far fewer firms prioritized the production of manufacturing equipment or the invention of new technologies.”

As a result of China’s strength in innovative manufacturing processes and scale-up, supported by its financial system, demand policies, and workforce training—as well as massive production subsidies, local content requirements, and forced technology transfer—only China cur­rently manufactures new energy technologies like solar and increasingly batteries and soon EVs at worldwide scale. Although Nahm sees the new global system as a positive because it presents new opportunities for firms and countries to specialize and collaborate, the United States misses out on the large potential base of employment from, and control of, the technologies it invented.

The Innovation System Revisited—by China

Today, it is China which is the pacesetter for auto manufacturing, not Germany or Japan, and certainly not Detroit. China dominates the production of full electric vehicles and, in 2023, surpassed Japan as the world’s largest auto exporter. As soon as 2025, Europe could become a net importer of cars, given the rapid growth of electric vehicles made in China. Some of these are Chinese brands but others are European (and American) brands that have shifted their EV manufacturing to China. China leads in battery production and offers well-developed supply chains, a vast domestic market, and technological expertise.

The Wall Street Journal observed, “Thanks to competition and [China’s] focus on execution, the EV industry went from a niche industrial policy project to a sprawling ecosystem of predominantly private companies. Much of this happened below the Western radar.”³⁸ The Journal also referenced a statement from a Chinese entrepreneur who said, “Ideas are not important in China—execution is.” This attention to execution in production is very different from the basic research ambitions and goals of America’s innovation system. But China’s seemingly overnight success in electric vehicles derives from something else, too: the industrial and frequently outright mercantilist policies³⁹ it has successfully used in industry after industry. These include market access and consumer financing contingent upon local production, forced technology transfer, exclusion of foreign battery producers, and massive subsidies through government guidance funds.⁴⁰

China’s emphasis on manufacturing is the exact opposite of the postindustrial discourse which has been dominant in the United States from World War II until very recently. “Industry is the main engine of economic growth. Industry is the main battlefield of technological innovation,” writes Party Secretary Jin Zhuanglong, head of the Minis­try of Industry and Information Technology. His article, “Accelerate the Promotion of New Industrialization,” published in an official CCP theory journal, further stated, “keep the proportion of manufacturing industry to GDP basically stable, and prevent the economy from ‘moving away from the real to the virtual.’ Promoting new industrialization is urgent to build a large country’s competitive advantage. It is the foundation for winning the initiative in international economic competi­tion.”

The United States, however, is no longer sitting still. It has put in place significant industrial policies under the Biden administration, some of which passed with bipartisan support in 2021 and 2022. China is not sitting still, either.⁴¹ Its industrial policies include the equivalent of up to $400 billion in annual scale-up financing for growing industries—the United States has no comparable program.⁴² It is rapidly reforming its own innovation system, including addressing its needs in basic research. This is the last missing link in its “innovation chain,” one required for Chinese technological self-sufficiency. Funding for basic research increased 24 percent in 2021 compared to a year earlier.

“China’s policy makers are seeking to systematically address and integrate every step of the innovation process,” notes a 2023 report by the German think tank merics.⁴³ “These include launching large projects, promoting industry-university-research synergy, and integrating the innovation, industrial, capital, and talent chains.”

Basic as well as applied scientific research is directed towards national strategic goals such as technological self-sufficiency according to the merics report. “China’s government has brought the work of State Key Laboratories closer to commercialization by embedding labs in compa­nies including Huawei and ZTE.” With government backing, China is reproducing in its way the vanished U.S. corporate lab system. Even basic research conducted at universities shares these strategic and commercial imperatives. Under the National Key R&D Projects pro­gram, which provides research funding, “there is no strict division between funding for basic research and support for developing practical applications,” according to the merics report. China is now second in the world in R&D spending and on a path to pass the United States.⁴⁴ Unlike in the United States, in China there are combined thrusts for invention, innovation, and production, integrated with trade policy.

Implementing Advanced Manufacturing Policies

As of 2021, China’s manufacturing sector accounted for a 28 percent share of the value add of its GDP. In the United States, manufacturing makes up only 11 percent of the value add of its GDP. Roughly 40 to 45 percent of China’s manufacturing output is directed to exports.⁴⁵

Furthermore, the United States ran a trade deficit in advanced technology goods, largely with Asian nations (predominantly China), as noted above, that grew to $244 billion in 2022. This is not a deficit in commodity goods; it is a deficit in the goods the United States needs for technology and innovation leadership. And this is an accelerating prob­lem for the United States, not a declining or stabilizing one.

Although it’s long been in denial, the United States is now recognizing that it has a deep manufacturing problem. If it is going to tackle this problem, it must shift to advanced manufacturing methods to get the kind of productivity and efficiency gains needed to compete in these goods. The issue now is how to do this. A systematic approach is needed; trying to tackle the problem by relying on just R&D capabilities will not be enough. Trade policy, for instance, will be part of any successful solution and is necessary to protect manufacturing from predatory pricing by mercantilist competitors. We focus here, however, on pragmatic steps the United States could take to strengthen its innovation chain. The general theme, usually overlooked in standard policy analysis, is relinking the innovation system to production.

Policy discussions in recent years increasingly focus on “direction­ality,” that is, the idea that policies should set the direction for technol­ogy development, as opposed to simply letting markets work their will. But what role should government play in setting that direction for manufacturing? And how can “directional” policy decisions be pursued given the complex mix of actors (from industry in numerous sectors to universities to government) and the highly federalized political system of the United States?

Given this complexity, it is important for manufacturing policy design to bring together the critical actors in the system. These include industry, including small manufacturing firms and large; government, including federal, state, and local; and educators, including universities and community colleges. The tool kits of each set of actors will all be needed to form effective policies. Below are ten steps that the federal government, working with the other actors, could undertake.

(1) Improve the Manufacturing Institutes. Although policymakers began to see that American manufacturing was facing challenges when Japan’s quality model began making inroads into U.S. auto and consumer electronics sectors in the 1970s and 80s, only limited policy steps were taken. Industry, while trying to copy Japan on quality, continued to believe that the overall solutions to its problems lay in tax and trade policies. It was not until the Obama administration that an innovation-based set of policies started to evolve, with the creation of sixteen manufacturing innovation institutes during the period of 2012 to 2017.⁴⁶ The motivating concept was that U.S. manufacturers could only compete with lower-cost Asian producers, particularly from China, if they became much more efficient and productive, to offset their cost and wage disadvantages. The institutes were collaborations, with federal core funding cost-shared with industry, universities, and state governments. Each institute was organized around a particular key manufacturing technology strand—robotics, digital production, bio-fabrication, ad­vanced composites, etc.—and they were loosely modeled on Germany’s Fraunhofer Institutes, which play a comparable role. Congress has recently approved three more institutes. While the institute model involves the critical actors required for manufacturing innovation listed above, program details need attention.⁴⁷

End the term limits. The institutes have never had the budget or long-term commitment that a manufacturing technology transformation requires. At the outset of the program, the institutes were placed on term limits, with federal support set to end after five years. This was a political decision but was never realistic: the structural problems in manufacturing that the institutes were designed to tackle cannot be solved in five years; they represent long-term challenges. The term limits forced the institutes to organize around the needs of large companies—the only source that could sustain them over time when the government pulled out—which meant that they had to neglect the smaller firms that were furthest behind in adopting new technologies and limit their workforce education investments. The term limits should be ended, with institutes subject to rigorous review at the end of their current terms; if they are successful they should be extended with comparable funding. The Commerce and Defense Departments have now adopted this approach, although the Energy Department has not, and many of its institutes have become shells with very limited programs. While Commerce has reviewed and fully funded its first institute, DoD has reviewed its institutes but renewed them at significantly lower funding levels. These problems require correction.

Network the institutes. In addition, efforts need to be networked across the institutes. Companies don’t want to adopt advanced manufacturing for one technology at a time just because they were developed at separate, stove-piped institutes. Instead, institutes need to cooperate to offer packages of their technology advances to firms. For example, firms don’t want to adopt just robots; they want robotics integrated with digital production technologies and other technologies such as 3-D printing or advanced composites. We need an institute network function to pull together advances from across institutes and package them for easier adaption.

Collaborate closely with state and local governments. Although the institutes must have a national mission, manufacturing is highly regional, and the states and local governments play the leading roles in economic development and workforce education. Collaborations between them and the institutes are ongoing but deepening them will be key to institute success. New manufacturing technologies need to be available for national adoption but tested and piloted regionally, which means institutes must operate at both national and regional levels.

Undertake testing, demonstration, and certification. The Defense Department has long supported the testing and demonstration stages, key steps for technology acceptance and scaling. Most manufacturing institutes have created technology centers and should receive additional funding to undertake rigorous testing and demonstration of new tech­nologies developed by participating companies in the areas the institutes support, such as robotics or 3-D Printing. Part of these demonstrations should be to identify the cost and efficiency savings to manufacturers these technologies can provide. After effective demonstration, the insti­tutes should certify and validate the feasability of the new technologies, which can help with prompt market acceptance.

(2) Back R&D for manufacturing technologies. We also need to enlist our federal R&D agencies in the cause. Manufacturing is not a subject of their research, but it needs to be, so technology advances move into the manufacturing institutes for experimentation and adop­tion. The federal basic research agencies work at what DoD and NASA have characterized as “Technology Readiness Levels” (TRL) 1–3, meaning they support early-stage research up to proof-of-concept experimentation. Since industry typically devotes most of its resources to the scale-up and initial production stages—TRL levels 7–9—there is a major gap (the infamous “valley of death”) in between. The manufacturing institutes are designed for these in-between levels, from development and prototyping to technology demonstration (TRL levels 4–6).

But if the institutes focus at these post-research stages, over time they will need new input from earlier stage research. Our R&D agencies have historically shunned research on manufacturing technologies—that’s been viewed as industry’s purview, although industry emphasizes later-stage development not research. We need to see manufacturing as a system, with connections throughout, from research to production.

Although the research agencies never developed significant portfolios around manufacturing, now is the time for them to do so if we are to have a connected innovation system to spur advanced manufacturing. This means, too, that they will need to include applied work that adds manufacturing solutions to their earlier-stage research. We need a cross-agency agenda for R&D on manufacturing technologies and processes at our R&D agencies, linked to the development initiatives at the institutes. Otherwise, over time, tech development at the institutes will become stranded and thin out.

(3) Provide scale-up financing. While venture capital has been a major force for financing innovation in the United States, it is now focused overwhelmingly on software, biotech, and various services sectors. It is largely out of “hard tech”—innovations that must be manufactured. This means there are few mechanisms to scale up manufacturing production in the United States. In recent decades, firms have been shifting prototyping and production of hard-tech goods to China.

A significant manufacturing gap in the United States is seen in its lack of entrepreneurial start-up firms in this sector, which are needed to advance innovation. The share of young manufacturing firms (less than five years old) among U.S. manufacturers has steadily declined over the past three decades, falling far below the levels in sectors supported by VC, such as biotech, software, or IT.⁴⁸ Yet newer firms tend to adopt and develop new manufacturing technologies and processes at a higher rate than established firms. Manufacturing is missing among new entrants, and scale-up financing could help encourage their entry into manufacturing and corresponding production innovation.

If the United States is to shift to advanced manufacturing at both old firms and new, it will need scale-up financing to support it.⁴⁹ In contrast to the United States, China has long provided massive manufacturing scale-up assistance to its firms, which accounts for much of its dominance of world manufacturing output. What form could scale-up fi­nancing take in the United States? Operation Warp Speed in 2020 used guaranteed contracts to Covid-19 vaccine makers to reduce their risks and assure production. Senator Chris Coons and colleagues have pro­posed an industrial finance corporation for innovative manufacturing.⁵⁰ The DoD in 2022 created a new “Office of Strategic Capital” for tech­nology scale-up, although it appears focused more on national security rather than broader economic security goals.⁵¹ The Biden administration has proposed repurposing an established federal bank, the Ex-Im Bank, to provide manufacturing scale-up support alongside its long-standing export financing role.⁵² The Depart­ment of Energy’s Loan Programs Office, since 2005, has provided financing for scale-up of new energy technologies. While the program suffered a political backlash for a failed loan to a thin-film solar company, Solyndra, in 2011, it has had numerous successes. The most notable was a $465 million loan to Tesla in 2009–13 for expansion of its Fremont, California, production site, which enabled the company to avoid bankruptcy and accelerated the scale-up of electric vehicles in the United States. In addition, states in their economic development role have often provided financing to firms to support regional job creation. In contrast to the United States,⁵³ China has long provided massive manufacturing scale-up assistance to its firms, which accounts for much of its dominance of world manufacturing output.

There are thus a range of possible mechanisms available, from forming a new industrial bank to expanding the role of existing institutions. Applying a number of these mechanisms with expanded lending authority for advanced manufacturing scale-up could be a solution to this glaring manufacturing system gap.⁵⁴

(4) Use government procurement power to promote new manufacturing technologies. The DoD is by far the largest procurement agency, and it needs to strengthen its industrial base and build more resilient supply chains for increasingly pressing national security needs. The productivity and efficiency gains possible from advanced manufacturing therefore should be a priority. DoD historically played a role in fostering new manufacturing technologies, such as the interchangeable machine-made parts developed at Army arsenals in the 1840s. A more recent example is computer numerically controlled (CNC) equipment, pervasive now in all manufacturing sites. It was first developed through DoD-supported research at MIT. When DoD saw the new level of precision manufacturing it enabled, it required its contractors to implement CNC machining for its missile programs, which spread this advance throughout U.S. industry.

DoD, working with manufacturing innovation institutes, could identify productivity savings available from new technologies such as digital production advances, robotics, or 3-D printing, and require its defense contractors, through specifications and contracts, to implement them. It also has authority under the Defense Production Act to acquire these new technologies and lease them back to its contractors. While DoD procurement does not dominate U.S. manufacturing, it has a significant market share. DoD could adopt the approach of mandating improvements, along with funding support for new equipment, that it used with CNC machining, enabling larger-scale adoption of the new manufacturing technologies.

(5) Direct production support. In the case of some critical technologies, we may need more direct production support to build factories. We need a more sustained chips Act, and for other critical sectors, not just semiconductors.

In the 2022 chips Act, the federal government provided $39 billion in grants as well as loan guarantees and investment tax credits to semiconductor makers to fund new advanced production facilities. Similarly, the federal government, through the Commerce or Defense Departments, could provide support for firms to acquire and install new advanced manufacturing technologies. This could be applied particularly toward production of critical or national-security-related technologies.⁵⁵

(6) Provide both “top-down” and “bottom-up” support. Many support tools are going to be needed for a manufacturing revival and will involve a series of organizations and agencies, from the advanced manufacturing institutes, to the R&D agencies noted above, and the Defense and Commerce Departments. Each has existing policy imple­mentation mechanisms and could acquire new ones based on the recommendations here. These tools can be offered to manufacturing firms in two basic forms. Top-down tools have long been employed by darpa and were also employed by Operation Warp Speed. Here, the government identifies a technology challenge to be addressed, and picks a small portfolio of leading companies to pursue it. The DOE’s funding of eight companies for fusion development is yet another example. Bottom-up tools involve incentives widely available to interested com­panies to meet a technology challenge—it’s up to the firms to pursue them. For example, the Inflation Reduction Act provides tax credits for purchases of EVs manufactured in America if the vehicles meet sourcing and processing requirements for critical materials used in their batteries. Tesla notably took advantage of many other bottom-up tools in rising to EV technology leadership.⁵⁶ The manufacturing chal­lenge is complex so both top-down and bottom-up approaches will be needed.

(7) Build a manufacturing focus into existing industrial policy programs. The United States has recently passed a major package of industrial policy programs (including the energy demonstrations in the Infrastructure Act, the chips and Science Act, and the Inflation Reduction Act) with over $500 billion in new funding. Yet programs specifically focused on manufacturing technologies are largely missing from these initiatives. While the new legislation aims to rebuild U.S. supply chains, if the bulk of this manufacturing does not occur in the United States, supply-chain resilience will remain elusive. In effect, we need to put the “industrial” into these industrial policy programs. Each requires a policy focus on manufacturing. And future industrial policy approaches require more of a specific manufacturing focus. Rather than just incentivizing the growth of a particular manufacturing sector through subsidies or tax credits, this would also mean incentivizing adoption of advanced manufacturing processes such as digital technologies, robotics, artificial intelligence, 3-D printing, or bio-fabrication, both in general and in those sectors.

(8) Map and fill gaps in supply chains. The Covid pandemic and the resulting economic fallout exposed the weaknesses of U.S. supply chains not only in protective equipment, pharmaceuticals, and pharmaceutical materials but in a host of areas, from semiconductors to critical materials and supplier manufacturing capabilities. A fundamental lesson from Operation Warp Speed in 2020 was the need to better map supply chains in order to assure production capability, and the Defense Production Act enabled the Defense Department, working closely with industry, to intervene and fill gaps to assure sufficient vaccine production to meet the emergency. Similarly, in 2020, the executive branch, through a series of participating agencies and the White House, began a major effort to identify and fill supply-chain gaps in advanced batteries, pharmaceuticals and pharmaceutical ingredients, semiconductors, and critical miner­als.⁵⁷ Particularly for the production of critical and nation­al-securi­ty‑related technologies, relevant federal agencies could work with industry to support the mapping of supply chains for implementing advanced manufacturing technologies and processes.

(9) Fix workforce education. Germany has long gained productivity improvements from a famously well-trained manufacturing workforce based on an apprenticeship system. In contrast, U.S. companies have generally tried to get productivity gains from capital, plant, and equip­ment investments and ignored the workforce side. German firms under­stand, however, that productivity gains from new equipment can soon spread worldwide, while a gain from a high-quality workforce will be enduring and provides a long-term competitive edge.

The United States has a broken workforce-education system, with a deep disconnect between the education system and workplaces. If the United States wants to adopt advanced manufacturing, its workforce must be ready for it. This requires rebuilding much of workforce education at all levels.⁵⁸ Community colleges must introduce advanced manufacturing curricula, create short programs more adapted to upskilling workers already in the workforce, establish certificates around particular skills that stack toward degrees, and turn around low completion rates. At the federal level, disconnected Labor and Education Department programs need to be integrated and efforts to expand registered apprenticeship programs accelerated. At the industry level, firms must collaborate with each other and with community colleges to build new training and apprenticeship programs, including youth apprenticeship starting in high school. Solutions will have to be pursued throughout the U.S. federal system, since education is largely state- and local-government led. These governments need to improve community college funding and support, integrate advanced manufacturing pro­grams across community colleges, promote apprenticeships, and work closely with area firms on reforms and curricula.

(10) Put someone in charge. The above steps require a series of agencies to act in concert, not an easy task in our stove-piped government, and ways must be found to pull these varied manufacturing pieces together. An interagency committee will be inadequate for the task. The Obama administration in 2008, faced with the bankruptcy of most of the U.S. auto industry, named a manufacturing czar to supervise federal support and put the industry back together, who then began to attack manufacturing challenges more broadly.⁵⁹ A comprehensive program to spur and implement new manufacturing technologies and processes, with varied mechanisms of support, as well as new trade approaches, calls for such a position in a permanent White House office. Operation Warp Speed provides guidance on how to organize it. OWS was a task force with a specific mission that included key representatives from relevant agencies led by a highly experienced industry leader and senior Army logistics expert, as well as an independent and expert supporting staff. That model could work here, assuring White House authority, links to agency powers, and independent leadership. Since the mission will require industry collaboration, an advisory group of industry, engi­neering, university, and labor leaders could be formed.⁶⁰

This is not an exclusive list for what needs to be done through government policy for the United States to shift to advanced manufacturing, but it marks a beginning. Most are relatively manageable initial steps, many within reach of the existing policymaking process and not requiring new laws. Nearly all require collaboration between the actors in the U.S. manufacturing system—industry, government, and education—and there will be no substitute for committed companies. The remedies here are largely federal but some are shared across federal, state, and local governments. But more “directionality” from new government policies will be required for this shift to advanced manufacturing.

A Rare Opportunity

Only about twenty countries have managed to move from developing to developed-nation status since World War II. Fourteen of them applied industrial policies based on governmental interventions.⁶¹ This approach has been most famously expressed in the East Asian development model. Directed and targeted interventionist policies were key to Japan’s recovery after World War II and were also applied by Korea, Taiwan, Singapore, and in recent decades by China. All these countries relied on strong manufacturing through industrial policy approaches to drive economic growth. Since these examples suggest that the restoration of manufacturing prowess requires interventionist government policies, there are lessons here for the United States about the kinds of policies it needs to consider.

U.S. global power paralleled its rise in manufacturing power. It was the first nation to introduce interchangeable machine-made parts before the mid-nineteenth century. It grew this breakthrough in technology and process into its system of mass production, which led the world in manufacturing by the early twentieth century. By the end of World War II, it had achieved production supremacy—no other nation was close to U.S. manufacturing capacity and output. Through this manufacturing advance it forged a deep connection between national security and economic security with manufacturing as the enabler of both. But this link, never explicit, has been neglected since the end of the Cold War. America’s problems with, and indeed indifference to, manufacturing have historical and societal roots, which predate the rise of China and the offshoring of manufacturing. In some ways, offshoring is the culmi­nation of the trend, not the cause. But by missing a focus on manufacturing, the United States has missed an innovation capability step that many other leading nations have grasped.

The recent industrial policies of the Biden administration mark a sea change in America’s approach. They are attempts to address both the technological competition with China and the need for new energy technologies.⁶² These initiatives are important and necessary—but they are not sufficient. No nation can maintain a world-power position while walking away from manufacturing. Major gaps remain in advanced manufacturing and scale-up financing, which are not a significant focus of the new programs compared to the level of effort needed. The United States is now pursuing a series of industrial policies, although so far they are light on the “industrial.” For the United States to successfully adopt advanced manufacturing at scale, its industrial innovation policies need to address production directly. China is facing its own challenges, including adverse demographic trends, rising wages, regulatory uncer­tainties, and wasted resources, but a negative shift in China’s growth model does not mean the United States will succeed in manufacturing.

The way forward is for the United States to redesign its innovation system, so that innovation is more closely linked to production. It can’t just focus on basic R&D as has been the case since World War II. The United States needs to retain these basic research capabilities, but at the same time sharpen its focus on execution. Overall, the United States needs to more fully integrate government policies, trade, finance, education, innovation, and production.

There is a rare opportunity to do this. Economic security, including the technological competition with China, as well as the quest to renew American economic leadership, provide a strategic imperative.⁶³ But there is something more. The United States created breakthroughs in technologies which can transform advanced manufacturing: 3-D print­ing, advanced composites, new materials, biomanufacturing, photonics, power electronics, and other emerging fields including AI manufacturing.⁶⁴ We are at a once-in-a-century moment where we could fundamentally change the way we undertake production.

Whereas EU nations define advanced manufacturing primarily in terms of current digital technologies, the federal government’s modest manufacturing programs are researching and pursuing a broader host of technologies—a more revolutionary approach. The United States may have fallen behind in implementing today’s advanced production tech­niques, but with the right political commitment, new scale-up financing and other tools, and tighter linking of innovation and production, America may be able to leapfrog to the frontier of an industrial trans­formation.

This article originally appeared in American Affairs Volume VII, Number 3 (Fall 2023): 3–30.

---

Notes

¹ Reda Cherif, Marc Engher, and Fuad Hasanov, “Crouching Beliefs, Hidden Biases: The Rise and Fall of Growth Narratives,” Working Paper No. 20/228, International Monetary Fund, November 2020; Reda Cherif and Fuad Hasanov, “The Return of the Policy That Shall Not Be Named: Principles of Industrial Policy,” Working Paper No. 2019/074, International Monetary Fund, March 2019, 23–24; Reka Juhasz et al., “The Who, What, When and How of Industrial Policy: A Text-Based Approach,” Structural Transformation and Economic Growth (STEG) Working Paper 050, January 12, 2023, 22.

² For details on each, see William B. Bonvillian, “Industrial Innovation Policy in the United States,” Annals of Science and Technology Policy 6, no. 4 (2022): 315–411.

³ “Trade in Goods with Advanced Technology Products,” U.S. Census Bureau, 2023.

⁴ Ian Clay, “China Exporter—United States Importer,” Information Technology and Innovation Foundation, February 28, 2023.

⁵ Manufacturing.gov, the federal government’s advanced manufacturing portal, defines advanced manufacturing as “Use of innovative technologies to create existing products and the creation of new products. Advanced manufacturing can include production activities that depend on information, automation, computation, software, sensing, and networking.”

⁶ Robert J. Gordon, The Rise and Fall of American Growth (Princeton: Princeton University Press, 2016).

⁷ Stefan J. Link, Forging Global Fordism: Nazi Germany, Soviet Russia, and the Contest over the Industrial Order (Princeton: Princeton University Press, 2020), 3.

⁸ Dwight Jon Zimmerman, “Henry J. Kaiser and the Liberty Ships,” Defense Media Network, June 24, 2021.

⁹ David Hounshell, From the American System to Mass Production, 1800–1932: The Development of Manufacturing Technology in the United States (Baltimore: Johns Hopkins University Press, 1984), 3.

¹⁰ Vannevar Bush, Science—the Endless Frontier: A Report to the President (Washington, D.C.: United States Government Printing Office, 1945), reprints Roosevelt letter of November 17, 1944.

¹¹ Bush, Science—the Endless Frontier.

¹² William B. Bonvillian, “Encompassing the Innovation Panoply,” Issues in Science and Technology 38, no. 2 (Winter 2022): 37–43.

¹³ Bonvillian, “Encompassing the Innovation Panoply.”

¹⁴ William B. Bonvillian and Peter L. Singer, Advanced Manufacturing: The New American Innovation Policies (Cambridge: MIT Press, 2018), ch. 2.

¹⁵ Jake Sullivan, “Remarks on Renewing American Economic Leadership,” White House, April 27, 2023; Robert Lighthizer, No Trade is Free (New York: Broadside Books, 2023).

¹⁶ Henry Kressel, email interview, 2023. See also Kressel, “Edison’s Legacy: Industrial Laboratories and Innovation,” American Affairs 1, no. 4 (Winter 2017): 115–29.

¹⁷ Bonvillian and Singer, Advanced Manufacturing, n. 12.

¹⁸ Jeffrey A. Hart, “A Comparative Analysis of the Sources of America’s Relative Economic Decline,” Understanding American Economic Decline, eds. Michael A. Bernstein and David E. Adler (Cambridge: Cambridge University Press, 1994), 207.

¹⁹ Bonvillian and Singer, Advanced Manufacturing, ch. 3.

²⁰ Daniel Bell, The Coming of Post-Industrial Society: A Venture in Social Forecasting (New York: Basic Books, 1973).

²¹ Daniel Bell, “Notes on the Post-Industrial Society,” Public Interest (Winter 1967): 27.

²² Bell, “Notes on the Post-Industrial Society,” 28.

²³ See William B. Bonvillian and Sanjay E. Sarma, Workforce Education: A New Roadmap (Cambridge: MIT Press, 2021).

²⁴ Paul Lewis, “Innovation, Technician Skills, and Vocational Education and Training: Connecting Innovation Systems and Vocational Education and Training,” Journal of Vocational Education and Training, May 27, 2023.

²⁵ For more, see Bonvillian and Singer, “Advanced Manufacturing,” ch. 4.

²⁶ Rana Foroohar, “Why Manufacturing Matters to Economic Superpowers,” Financial Times, April 14, 2021.

²⁷ Alan S. Blinder, “Fear of Offshoring,” CEPS Working Paper No. 119, December 2005.

²⁸ U.S. Council of Economic Advisers, Annual Report of the Council of Economic Advisers: Economic Report of the President (Washington, D.C.: United States Government Printing Office, 2004).

²⁹ Jonathan Weisman, “Bush Adviser Assailed for Stance on ‘Offshoring’ Jobs,” Washington Post, February 11, 2004.

³⁰ Peter Petre, “How GE Bobbled the Factory of the Future,” Fortune, November 11, 1985.

³¹ David Gelles, The Man Who Broke Capitalism: How Jack Welch Gutted the Heartland and Crushed the Soul of Corporate America―and How to Undo His Legacy (New York: Simon and Schuster, 2022).

³²  Suzanne Berger, How We Compete (New York: Crown Business, 2005).

³³ David Adler, “The Real Challenge for U.S. Industry,” City Journal, March 29, 2017.

³⁴ “China Overtakes USA in Robot Density,” International Federation of Robotics, December 5, 2022.

³⁵ See Klaus Schwab, “The Fourth Industrial Revolution: What It Means, How to Respond,” World Economic Forum, January 14, 2016; “Fourth Industrial Revolution,” World Economic Forum, 2023.

³⁶ Jonas Nahm, Collaborative Advantage: Forging Green Industries in the New Global Economy (Oxford: Oxford University Press, 2021).

³⁷ Although Nahm’s book predates the major energy investments in the 2022 Inflation Reduction Act, that act focused on consumer and industry tax and other incentives for spreading green technologies, not on manufacturing capability. Although the chips and Science Act calls for regional innovation hubs to try to spread innovation mastery to new areas, potentially including production capability, it is not yet funded. So Nahm’s analysis remains relevant for understanding comparative approaches to innovation and production.

³⁸ Greg Ip, “China’s EV Juggernaut Is a Warning for the West,” Wall Street Journal, June 7, 2023; Zeyi Yang, “How Did China Come to Dominate Electric Cars?,” MIT Technology Review, February 21, 2023.

³⁹ Gregor Sebastian and Francois Chimits, “‘Made in China’ Electric Vehicles Could turn Sino-EU Trade on its Head,” merics, May 30, 2022.

⁴⁰ David Adler, “Guiding Finance, China’s Strategy for Funding Advanced Manufacturing,” American Affairs 6, no. 2 (Summer 2022): 17–40.

⁴¹ Barry Naughton, Siwen Xiao, and Yaosheng Xu, “The Trajectory of China’s Industrial Policies,” Working Paper, UC Institute on Global Conflict and Cooperation, June 2, 2023.

⁴² Gerard DiPippo et al., Red Ink: Estimating Chinese Industrial Policy Spending in Comparative Perspective (Washington, D.C.: Centre for Strategic and International Studies, 2022).

⁴³ Jeroen Groenewegen-Lau and Michael Laha, “Controlling the Innovation Chain: China’s Strategy to Become a Science and Technology Superpower,” merics, Feb 2, 2023.

⁴⁴ Bailey Crane, “China’s Drive for Leadership in Global Research and Development,” Center for Strategic and International Studies, June 30, 2023.

⁴⁵ Andrew Batson, “Breaking Down China’s Manufacturing,” The Tangled Woof (blog), June 13, 2023.

⁴⁶ See, generally, Bonvillian and Singer, Advanced Manufacturing.

⁴⁷ For additional discussion of needed improvements, see William B. Bonvillian, “Ensuring Manufacturing USA Reaches Its Potential,” Federation of American Scientists, August 10, 2021.

⁴⁸ Ben Armstrong, Suzanne Berger, and Bill Bonvillian, Advanced Technology, Advanced Training (Cambridge: MIT Initiative for Knowledge and Innovation in Manufacturing, 2021): 23–24.

⁴⁹ Although investments in manufacturing plant construction have accelerated this year (see Javier David, “1 Big Thing: A Manufacturing Supercycle is Starting,” Axios, June 16, 2023) this is largely spending in the semiconductor sector driven by the chips Act, and for several EV and battery plants mostly in the South by foreign investors; it is not broad-based.

⁵⁰ “Sen. Coons, Colleagues Seek to Create New Domestic Manufacturing Investment Corporation,” Office of Senator Chris Coons, news release, August 12, 2021.

⁵¹ C. Todd Lopez, “New Defense Office Connects with Next-Gen Tech Developers with Much-Needed Capital,” U.S. Department of Defense, December 1, 2022.

⁵² “Statement From the Export Import Bank of the U.S. President and Chair Reta Jo Lewis on the One Year Anniversary of President Biden’s Supply Chain Executive Order,” Export-Import Bank of the United States, February 24, 2022.

⁵³ Tesla’s loan from the Department of Energy (DOE) was approved in 2009 and was available in January of 2010. Tesla repaid the loan in 2013, ahead of schedule; see “Tesla: Loan Programs Office,” U.S. Department of Energy, accessed July 22, 2023; Tim Higgins, Power Play: Tesla, Elon Musk and the Bet of the Century (New York: Doubleday, 2021): 109–12, 159, 169, 194.

⁵⁴ Gerard DiPippo et al., Red Ink; David Adler, “Guiding Finance.”

⁵⁵ The chips and Science Act identified ten critical technology areas for government programs to focus on, with a process to periodically update this list (See, HR 4346, 117th Cong., 2nd Sess., Subtitle G, 2021; Tim Clancy, “Chips and Science Act Enshrines Policy for New NSF Technology Directorate,” American Institute of Physics, November 23, 2022.

⁵⁶ Bonvillian, “Industrial Innovation Policy in the United States.”

⁵⁷ “Building Resilient Supply Chains, Revitalizing American Manufacturing, and Fostering Broad-Based Growth,” White House, June 2021.

⁵⁸ See detailed recommendations in William B. Bonvillian and Sanjay E. Sarma, Workforce Education: A New Roadmap (Cambridge: MIT Press, 2021).

⁵⁹ Noam Scheiber, “Manufacturing Bloom,” New Republic, December 7, 2009.

⁶⁰ The Advanced Manufacturing Partnership, an Obama administration panel of industry, university, and labor representatives that developed advanced manufacturing policies in 2011–14, provides a model (Executive Office of the President, “President Obama Launches Advanced Manufacturing Partnership,” news release, June 24, 2011).

⁶¹ Reda Cherif and Fuad Hasanov, “The Return of the Policy That Shall Not Be Named.” See also David Oks and Henry Williams, “The Long, Slow Death of Global Development,” American Affairs 6, no. 4 (Winter 2022): 122–50.

⁶² Bonvillian, “Industrial Innovation Policy in the United States.”

⁶³ Tai Ming Cheung and Thomas G. Mahnken, The Decisive Decade: United States–China Competition in Defense Innovation and Defense Industrial Policy in and beyond the 2020s (Washington, D.C.: Center for Strategic and Budget Assessments, 2023); Sullivan, “Remarks on Renewing American Economic Leadership.”

⁶⁴ The Georgia AI Manufacturing (ga-aim) coalition, led by the Georgia Tech Research Corporation, will receive approximately $65 million from the Commerce Department’s Economic Development Administration to accelerate the adoption of artificial intelligence across the state’s legacy industrial sectors. See U.S. Economic Development Corporation, “U.S. Department of Commerce Invests Approximately $65 Million to Accelerate Integration of Artificial Intelligence Technologies in Industry in Georgia through American Rescue Plan Regional Challenge,” news release, September 2, 2022.

About the Author

David Adler is author of the monograph The New Economics of Liquidity and Financial Frictions and coeditor of the forthcoming anthology The Productivity Puzzle, both published by the CFA Institute Research Foundation. He is also a contributing editor of American Affairs.

ALSO BY DAVID ADLER

• The Ghosts of Mont Pelerin: Visiting the Birthplace of Neoliberalism
• Guiding Finance: China’s Strategy for Funding Advanced Manufacturing
• Inside Operation Warp Speed: A New Model for Industrial Policy
• Reshoring Production and Restoring American Prosperity: A Practical Policy Agenda
• Reshoring Supply Chains: A Practical Policy Agenda
• Korean Industrial Policy: From the Arrest of the Millionaires to Hallyu

William B. Bonvillian is a lecturer at MIT teaching science and technology policy, a senior director for special projects at MIT Open Learning, a former senior adviser in the U.S. Senate, and coauthor of five books on technology policy and workforce development.