ITIF Search
Podcast: Global Supply Chains Under Pressure, With Willy Shih

Podcast: Global Supply Chains Under Pressure, With Willy Shih

Global supply chains have been under intense pressure during the coronavirus pandemic, particularly when it comes to medical supplies and drugs. What should the U.S. policy response be? Rob and Jackie discuss the issue with Willy Shih, a renowned professor of management practice at Harvard Business School.



Rob Atkinson: Welcome to Innovation Files. I’m Rob Atkinson, president of the Information Technology and Innovation Foundation. We’re a DC-based think tank that works on technology policy.

Jackie Whisman: And I’m Jackie Whisman. I’ve been with ITF for over a decade, and part of my job is making sure Rob doesn’t forget to tell you that we are the world’s top-ranked think tank for science and technology policy. Rob writes the 50-page papers, and I tell everyone how great they are.

Rob Atkinson: Thanks, Jackie. You do that quite well. This podcast is about the kinds of issues we cover at ITIF, from the broad economics of innovation to specific policy and regulatory questions about new technologies. In this episode, we’re focusing on global supply chains.

The pandemic has brought a lot of attention to this question, particularly with regard to medical supplies and drugs. To talk about this, we’re delighted to have Willy Shih with us today. Willy’s a professor of management practice at the Harvard Business School. Prior to coming to HBS in 2007, he spent 28 years in industry at companies like IBM, Digital Equipment, Silicon Graphics, and Eastman Kodak.

He’s also author of one of my all-time favorite books, with his colleague Gary Pisano, a book called Producing Prosperity: Why America Needs a Manufacturing Renaissance. Willy, glad you can join us.

Willy Shih: Hello Rob. Thank you for the invitation to join you today.

Rob Atkinson: So, you recently wrote a piece for the Harvard Business Review, which I believe is online, called “Bringing Manufacturing Back to the U.S. Is Easier Said Than Done.”

In that piece, you wrote quote, “the issue is complex and defies easy solutions. The challenge lies in a combination of how modern supply networks are structured and how the operational metrics apply to manufacturers.” Certainly, the COVID crisis has shined light on this. We’ve heard from Peter Navarro and the White House calling for a Buy America provision, at least for drugs and medical supplies.

Before we get into this question, where are we today and what can we do? You certainly have written about how supply chains have gotten more complex and more specialized and more globalized. You know, if we look back a hundred years at Henry Ford’s giant River Rouge plant, where essentially that factory, uh, largest factory in the world that brought in coal and iron ore on one end and out popped Model T’s on the other end, that’s not what we’re doing today.

How has it evolved? I mean, any sort of shorthand, what’s happened now to make all these supply chains so specialized and so global?

Willy Shih: Well, Rob, in some sense, the world has gotten much more complex. A lot of that is because of sophisticated technologies that require specialists to be able to produce them.

So, one of the examples I cited in that paper is a notebook computer. If you look at a notebook computer, they are assembled by specialist assemblers, and the assemblers, in turn, turn to other specialists—for example, to produce the display screen. The display screen is made by a handful of Asian manufacturers in large factories that costs, the latest generation, costs six-and-a-half billion dollars to put one together.

Okay. And those factories take years of experience and a lot of sophisticated machinery and tooling to produce those displays. But it gets even more complicated than that, because the displays themselves also have components like optical glass, or polarizing filters, or a color filter dies. All of those, in turn, rely on yet another tier of specialists. And then those specialists, in turn, will also turn to more, more suppliers. And that’s true for so many other components in that notebook computer as well: the microprocessor or the batteries, or the keyboard or the solid-state storage devices, or the memories, and so on.

So, what you have is essentially a network of suppliers and tiers, where you have first-tier suppliers, who will supply the primary assembler, and then a second tier, who might supply the first tier, and then the third tier, and so on. And it’s all driven by the high degree of specialization that is necessary to successfully make all those components.

Rob Atkinson: So, going back to the River Rouge example, you can imagine a River Rouge for Dell or HP. They’d have a giant factory and sand would come in one side and aluminum ore, and out would pop laptops—obviously not possible, because things you said are so complex.

Willy Shih: Right. That’s the old example of, I actually went earth, wind, and fire coming in one end, and then all these other products.

But you know, if you look at something like semiconductors, that’s the classic example of how you have a division of labor among all these specialists, right? A lot of people talked about the resurgence of the semiconductor industry in the United States in the 1990s and that was really driven by a separation of design from manufacturing, right? So, a lot of the companies that design worked in the United States—and that was companies like Nvidia or Qualcomm—are kind of the poster childs of that. And then a lot of the manufacturing went off to Asia. And I think, for example, you know, right now, in the crisis, we’re focused on things like how come we can’t get enough medical supplies and personal protective equipment?

People haven’t really focused on semiconductors so much. The extent to which we are dependent on many parts of the world for the actual manufacturing of the semiconductors, the packaging, the testing, and then the incorporation into products.

Rob Atkinson: So I want to come back to semiconductors, because certainly people may not be focusing on it at the moment, but certainly before the crisis, there was a lot of focus on that, particularly in the Defense Department and concerns about how vulnerable we might be.

So, when you think about the issue of supply chain vulnerability, it seems to me there’s really three different components. There’s an issue about dependency on a single factory, or a company, where there could be some kind of production interruption. We’ve seen that in the past, where one factory that was making some kind of resin had a fire, and there was a big shortage in the world. And we can see it, for example, maybe there might be a tsunami somewhere, or an earthquake.

And the second kind of dependency would be on nations—clearly, China being the case here, where they have used their leverage to extract concessions from other countries. We saw that with their limitations on rare-earth minerals exports, for example, unless you opened up a factory in China.

And the third is something you alluded to, and that’s about surge capacity. Even if we had a lot of factories in the U.S. to make ventilators or personal protection equipment, we’d still have an issue around surge capacity.

Willy Shih: Well, Rob, I think that’s a good characterization, especially when you talk about the dependency on single factories. So, for example, the world is highly dependent on Taiwan for semiconductor production. That has about half the world’s foundry capacity. TSMC has 70 percent of that foundry capacity in Taiwan.

But, to the first point about diversifying across single factories, they have multiple factories spread across the island, predominantly in three different science parks, right? So they kind of diversify across that. I think your point about depending on a single nation, like China, for rare earth, oftentimes that is driven by resource specificities. So, we also see, for example, the world dependent on South American nations for lithium supplies, for example.

The third one, which is the one on surge capacity, I think that is actually a really important question. And I also talked about this in the HBR piece about why there isn’t that much surge capacity. One thing to remember, a lot of our shortages now are driven by more than a 20X increase in demands. For example, four N-95 facemasks and surgical masks.

And you know, the whole question around surge capacity is how much are you willing to pay the have that, right? Because everything we have taught our operating managers and supply chain managers is about optimizing your production capacity, which means I want to run that thing as close to 100 percent as I can, right? I don’t want to have a lot of spare capacity, because that means I’m paying for capital that I’m not fully utilizing. We have taught generations of students to say, “you know, capital efficiency: I don’t want to have wasted capacity,” right? I don’t want to have a lot of stockpiles—which is another name for inventory, in case I have a sudden need—because somebody has to pay for that. And the point I was trying to make in this article is that until we, as consumers, until we, as a country, are willing to pay for those things, it’s going to be economically hard for producers to justify having that kind of surge capacity.

Rob Atkinson: Yeah. I mean, the whole lean production movement, it started really by Toyota and the in the 60s and 70s and spread here. In fact, a lot of it by MIT and Harvard business school professors back in the 80s. But, hey, that’s a good thing.

Willy Shih: Well, yeah, but let me also embellish that a little bit, okay, because if you look at the whole Toyota concept of lean—and even to this day, they practice it very well—which is, I want to have suppliers clustered around me, or not a great distance. And I will partner with those suppliers so that they understand my needs, and I understand their capabilities, and I’ll have strategic partnerships with those suppliers.

So, if you look at Toyota and Georgetown, Kentucky, which is a huge site now, they have a lot of their suppliers within the state of Kentucky, or nearby, as well as around the United States. In fact, Toyota-produced vehicles in the United States have a higher domestic content than many of the Detroit Big Three vehicles.

Now, I think one of the assumptions people have made has been, well, as long as I have efficient global logistics in the form of reliable container shipping, reliable air cargo, then I can distribute that to the supply chain globally and go to kind of, I refer to it as a global sequential production system, where I’m going to add value in different parts of the world, and I’m going to count on the fact that I can ship stuff around.

Well, one of the things that the current pandemic has exposed to us is that a supply chain is only as strong as its weakest links. Okay. And with the collapse of air cargo, and first a supply shock followed by a demand shock, and the impact on ocean-based container shipping, all of a sudden that has become a critical link in supply chains as well. And a lot of what we’re seeing is an inability to get materials to the right place when they are needed. So that assumption may not have held very well.

Rob Atkinson: Well, you know, it’s interesting, Willy, because the Japanese, as a broad generalization, and Toyota in specific, they have done lean production, but they’ve done it as, you know, in a geographically concentrated manner. There was a book that came out recently, which I enjoyed a lot, and sorry, I don’t remember the title of it, but we’ll put it on the podcast link. And it was an argument that one of the reasons why American Big Three auto companies didn’t do as well is that they dispersed their regionsalized production chains largely to fight unionization, and that this hurt their capacity to innovate. I’d be interested in your comment on that. But also, are there things the U.S. government should do to reshore more U.S. manufacturing, if you will? I looked recently at U.S. government data, and if you leave out the computer sector, which is historically mis-measured and overstated in its output, we’re producing less manufacturing output in inflation-adjusted terms last year than we were in 2008.

So, should we be trying to do more here? And if so, try to regionsally concentrate supply chains?

Willy Shih: Okay. Well, let me talk about your first point. I’ve always thought about, you know, what I admire about Toyota in particular is they take a very strategic view of their supply network and their suppliers as partners, right?

So, they will invest in their partners in terms of improving their productivity, improving their quality, improving their operations. And so, the question I often ask manufacturers is, do you take a strategic view of your suppliers, or do you take a transactional view? And by transactional view, I mean, it’s all about price, and every year are you going to give me productivity? And I’m afraid a lot of manufacturers have taken a very transactional view, which is this all about price every year, give me more productivity. And then, when we get into a crunch and you wonder why they don’t have the capacity to help you when you need them to, I would say that would be my explanation.

Now, what can we do in terms of regionsalization? I think we tend to talk a lot about the supply side of the equation. I think we also need to think about the demand side. I see this huge push right now to make masks, PPE, ventilators, and so on in the U.S., okay, and we have seen this movie before, going back to the H1N1 swine flu, where, you know, domestic manufacturers ramped up and then demand collapsed. And then they had to lay off workers, and so on. So, my advice is, really, we have to think about the economics side of the equation. We have to make the economics of domestic manufacturing work. And there are various approaches you can take to that, but don’t focus only on the supply side. You have to focus on the demand side because in the end, that is what’s going to pull through the supply domestically.

Rob Atkinson: So on that point, I take your point, because the question would be, if Peter Navarro was with us today, President Trump’s advisor principally on China, he would argue that having a Buy America provision or requirement for these products is a way to do demand-side policy.It is a way to do demand. Is that what you were thinking of, or were you thinking more of sort of insured, or guaranteed markets for companies, so that they don’t go out on a limb and end up losing a lot of money?

Willy Shih: Well, I mean, there are various approaches you could take on that. I think manufacturers really are looking for a stable demand. We’re about to embark on a lot of infrastructure spending. I think as a way of kind of reviving the economy. I think there are opportunities in the U.S. where we could generate demand for new technologies, advanced technologies, that would foster domestic production and build domestic capabilities.

Over time, the real question you want to ask is, what kind of capabilities do we have domestically to respond to a sudden crisis as we are experiencing now?

Rob Atkinson: So that gets to, I think another question. In your book with Gary Paisano, you highlighted a number of different technologies that the U.S. no longer can produce. I was talking to the CEO of one company a few years ago, and he was telling me about this new product they were coming out with, and they were making one of the core components, I think in Taiwan, and I asked him why he didn’t make it in the U.S., and he said, “Well, we looked, we looked and looked and looked, and we could not find a single supplier who had that capability anymore in the U.S.” So, I guess the first question there would be, how much do you think the federal government actually knows about what our capacities are, both at the first tier or OEM level, and then all the way through the supply chain?

DOD knows little bits here and there about defense products. But do you think the government should be able to have a better sense of intelligence of what’s happening in their area and where we’re weak?

Willy Shih: I think there are some people in the government, I would say particularly in DOD and some of the intelligence organizations, who have a relatively good understanding of some of this. I have spent my career in manufacturing and global manufacturing. I’ve had factories around the world reporting to me, and now that I’ve been at HBS I’ve spent a lot of time looking at these global supply chains, and even for me every time I want to explore a new product area, it takes a fair amount of research and talking to people to really understand some of those areas. So, for example, recently, in the last week, I’ve been exploring this question of, you look at the vaccine candidates for the COVID-19, okay, which is a very exciting area—which we have companies in the U.S., like Moderna Therapeutics, who are pushing forward with these vaccines, and we have these DNA-based vaccines, and we have viral vector methods as well—and underneath all of them, there’s a requirement that you are able to feed these systems with oligonucleotides. So, I’ve said, well, let’s go look at the supply chain of that, right? I still haven’t gotten to the bottom of who makes each part of that.

So, the world is a very complex place. It would be hard, I think, for a lot of people, especially congressional staff or Commerce, to understand each one of these areas, right? I would hope they would tap into the base of expertise in this country, where people know where these things come from and are able to trace it. But it’s hard work to really trace all of these things and understand it.

Rob Atkinson: There’s a lot of interest in Congress about doing something to bring more manufacturing back here. I think you saw that Prime Minister Abe of Japan recently announced a big incentive package for Japanese companies to move production from China back to Japan or, or frankly just out of China. Do you think that’s something we should be looking at, and if so, any thoughts on what are some of the best tools we could be using to get more manufacturing back home? And obviously we’re not going to bring back Happy Meal toys, nor should we, but maybe some other things?

Willy Shih: So, Rob, I think what we as a country really need to do is we need to think strategically about how we are going to get some of these capabilities back, or develop some of the leading-edge capabilities for the future that are important.

I often think about like platform technologies. What are things upon which a lot of other things are built? And therefore, I would want to have those capabilities within the country, or closer, or at least guaranteed accessible. I worry about short-term incentives, because I worry about the economic viability, because at the end of the day the companies who might be lured back, they still have to be competitive on a global stage.

So, the problem I always have with tariffs or things like that is that doesn’t help me maintain competitiveness, because what you’re doing is you’re saying, I’m going to allow you to charge a higher price effectively and not be globally competitive. I think there are examples that the U.S. has done in the past, which really demonstrate how we can help develop global leadership in this country. Those examples tend to be what I call pre-competitive, fostering pre-competitive collaboration, or pre-competitive research. The Human Genome Project in the 1990s, with heavy investments from NIH and the federal government, really put the U.S. on the map relative to others because of how much work—and that was really a pretty competitive nature—how much work, and practical know-how we developed in the organization.

I often point to NASA and the E3 program of the late 70s. E3 was energy-efficient engine program, where they funded pre-competitive research at both Pratt & Whitney, which is now part of Raytheon Technologies, and GE Aviation: more energy-efficient, clean combustors, higher pressure ratios, and the high-pressure turbines and things like that. And if you look at both of those companies, especially GE Aviation, it’s at the core of their roadmap, which led to their global leadership—large, high-bypass turbo fans—and they’re beyond comparing that segment, okay? And so, but those were long-term investments that go beyond one election cycle, okay? And I think we need to do those.

Jackie Whisman: So, drilling down just a little bit, if I’m a member of Congress or congressional staffer looking to help solve this, and I’m able to legislate on a federal level, what can I do?

Willy Shih: Well, I think invest more in those kind of long-horizon programs. This is what is the U.S. strengths. So, I would say a combination of those things. I would invest in some of those long-horizon programs. I would direct infrastructure programs, infrastructure spending, to what I would call known weaknesses in our infrastructure. This time we were hit by a pandemic. We all know that the grid in the U.S. needs modernization and our grid is obsolete, okay? And one of the reasons countries like China and Germany are ahead of us is because they’ve invested, in going to distributed generation and distributed consumption from a centralized generation, centralized consumption. But if you look at like high-voltage DC power transmission lines, the Chinese and the Europeans are way ahead of us. We haven’t built any of that infrastructure. In doing that, what you would do is you would drive development simultaneously on group III-V semiconductors for power semiconductor devices. We all know that’s an important technology for the future as well.

So, I think, you know, a thoughtful program that balances kind of short-term, medium, and longer-term investments with a real focus on developing capabilities within the country would benefit us enormously over the long term.

Jackie Whisman: So, final question, since we are almost out of time, what is the technology that you’re most excited about for the future?

Willy Shih: I am most excited about what’s going on in biotech. Specifically, if you think about cell therapies, we’ve gone from drugs being small molecules to large molecules, to cell therapies, really being living cells as therapies. I think if you look at some of the vaccine approaches that are out there, the mRNA and DNA-based approaches and viral vector approaches, we’re actually heading to a point where what we’re going to do is we’re going to give the body information in the form of a DNA sequence or an RNA sequence and then have the body develop antigens to fight and so on. So, I think there is so much going on in that area. It’s really an exciting area that U.S. is still really a leader in that area, and I think we should really strengthen that, particularly in the manufacturing of the precursors we need to support that industry. You would be surprised how much of that has been offshored to this point, but that’s an industry for the future. We all talk about the 21st century being the century of life sciences, you know? So, I think we should look very carefully at that and strengthen our capabilities.

Rob Atkinson: So, Willy, to that point, we hear a lot of talk about bringing back APIs—in other words, the ingredients, the active pharmaceutical ingredients, a lot of them are in China. And, and to your point, we don’t invest enough in that. We have a great program in the U.S. that ITIF actually helped launch, or get off the ground, called the Manufacturing USA networks, and we have one really good center, I believe in Delaware, that’s focused on part of that. But it seems to me that’s an area where Congress could do more: really expand that program focused on the process. So, and that gets to the other point about supply chain vulnerability. It would be better processes. It allows us to tool, scale things up more quickly.

Willy Shih: That’s right, Rob. And I think the opportunity there is in process innovation, like continuous manufacturing for APIs. But then we also have to make sure that we understand all the precursors that we need as, as we shift that. We don’t want to just shift the supply chain vulnerability from one set of materials to another.

Rob Atkinson: Willy, this has been great. Thank you so much. Can you tell our listeners what’s the best way to look you up? Is it your website at Harvard Business School?

Willy Shih: Website at Harvard Business School. I try to keep that current. I’m also on LinkedIn, and I’m also on Twitter, and we can provide you the links later as well.

Rob Atkinson: Great, thank you.

Jackie Whisman: Thanks for listening. If you liked it, please be sure to rate us wherever you get your podcasts. Then go to and sign up for our weekly email newsletter. Also follow us on Twitter, @ITIFdc. And we’re on Facebook and LinkedIn, too.

Rob Atkinson: That’s it for now, but we have more episodes of the podcast coming up. We have got great guests lined up to talk about China and U.S. economic relations; the impact of COVID-19 on the future of the U.S. economy and technology; and the implications of COVID for privacy policy, among other topics. So please come back. Thank you—and thank you, Willy.

Willy Shih: Thank you, Rob.

Back to Top