The US Innovation and Competition Act, which sailed through the Senate on Tuesday, seeks to strengthen cutting-edge American technologies, manufacturing, and — the reason for this rare bipartisan enthusiasm — competitiveness against China. The bill calls for spending a quarter of a trillion dollars on these problems, a huge increase over paltry earlier efforts.
But from observing technology industries, I know that if the United States is really going to rejuvenate its high-tech manufacturing, Congress will have to ensure this bill strengthens countless small and medium-sized firms you’ve never heard of and doesn’t just throw money at major companies. To better understand why, I reached out to Suzanne Berger, an MIT political scientist and Institute professor I studied with in the 1990s. Berger has led several efforts to examine the decline of American manufacturing, including MIT’s Production in the Innovation Economy project and part of the Work of the Future task force.
Advertisement
Our conversation has been condensed and edited.
You’ve been identifying key problems this innovation act needs to address.
Right. The focus can’t just be on large firms at the top of the manufacturing chain because their ability to produce a range of advanced goods depends on their base of suppliers. And today those suppliers lack the technology and skills to make the parts that would allow the top of the chain to take off.
When we did interviews in 2017, everyone was worried there weren’t going to be any jobs because robots would eat them. We visited strong small and medium-sized enterprises, with up to 500 employees. SMEs represent 98 percent of the firms in manufacturing, employ 43 percent of the workers, and supply major companies. It turned out to be very difficult to find any robots anywhere. One Ohio firm that makes precision screws for companies like Tesla was still using 1940s Davenport milling machines. It had some computer-controlled machines but the new equipment just sat alongside the Davenports. When these firms bought new technology, they just layered it on top of old technology.
Advertisement

How does weakness in small and medium-sized suppliers hurt big companies like Tesla?
Apple said it couldn’t manufacture Mac Pros in Texas because it couldn’t find qualified makers of advanced screws. My colleague Ben Armstrong found many cases where suppliers didn’t have the equipment to produce high-quality components prime defense contractors sought. One Massachusetts supplier to a defense contractor we visited lacked an electron microscope to detect defects in a material it was making. It acquired the microscope through the state’s M2I2 innovation initiative, but such programs are few and far between. The Senate bill proposes 10 regional innovation centers. I hope programs like M2I2 might be set up across the country.
How did US manufacturing sink into this state?
The sector once comprised large, vertically integrated firms. There was a flow from almost basic R&D through manufacturing. DuPont not only invented nylon but scaled up production over a decade. By the 1980s these firms tended to have low market valuations, so financial markets — hedge funds, private equity — descended on them, bought them, and shed factories and workers to drive up stock values. New digital technology facilitated this process. Beginning in the 1990s, a US semiconductor firm could send digital files to Taiwan to actually make the chips. And companies in lower-wage countries — Hong Kong, Taiwan, China — emerged to do the production.
Advertisement
So what happened to SMEs in manufacturing?
The ecosystem around them turned into a desert. Kodak used to train hundreds of apprentices and support community-college classes on optics in Rochester. When it went down, smaller optics firms, some of which are really world leaders in their areas, lacked the resources to support classes. In Germany, a company that wants to develop a new line can go to a local bank it’s worked with for generations. We have no local banks anymore.
How do we even start to rebuild a whole ecosystem?
I think there are ways. We can build institutions like community colleges and technology demonstration centers that many firms can use. Defense procurement is an important lever. When the Defense Department wanted manufacturers to use early computer-controlled machines, it bought and leased them to firms. We discovered, almost by accident, that about half the companies we interviewed were suppliers to defense contractors. For example, when a third-generation family metalworking firm in Ohio, which used to make industrial boilers, started work on Navy contracts, the Navy said it wanted robotic welding. The company saw the prospect of a long-term contract, so it bought a welding robot. And then it started thinking about how to train its workers on the robot.
The government can’t easily tell Apple or Tesla who to source from. Can their incentives be strengthened to support local suppliers rather than ones overseas?
Advertisement
I think they have more incentive now because they have to worry about vulnerability as China moves in a far more authoritarian direction. And many companies that outsourced learned the bitter lesson that it’s very difficult to control quality at a distance.
Jonathan Schlefer is a senior researcher in political economy at Harvard Business School and the author of “The Assumptions Economists Make.”