Discover more from Discourse
Does America Need an Industrial Policy for Semiconductors?
George Calhoun and Scott Lincicome speak with David Masci and Christine McDaniel about the CHIPS Act and its effects on both the domestic and the global semiconductor industry
Last year, Discourse editor-in-chief David Masci and Mercatus Center senior research fellow Christine McDaniel spoke with Professor George Calhoun of the Stevens Institute of Technology about China’s semiconductor industry and the risks of Taiwan’s global chip-making dominance.
One year later, on October 2, 2023, they had a follow-up conversation, this time joined by Cato Institute vice president for general economics and trade policy Scott Lincicome, to discuss semiconductors and industrial policy in the U.S. They spoke about the domestic and global effects of the CHIPS Act, the incentives the act created, national security concerns, how industrial policy affects innovation and much more.
Below is a transcript of their conversation that has been slightly edited for clarity. The audio can be found here.
DAVID MASCI: It’s been a little over a year since the passage of the CHIPS Act, which aims to reshore semiconductor manufacturing back to the U.S. by providing subsidies for new factories. While much if not most of high-end chip design takes place in the United States, our country actually makes only about 12% of the world’s semiconductors. Most chip manufacturing now takes place in Asia, something that American policy would say is unacceptable in an age of increasing geostrategic instability, particularly given the growing rivalry with China.
Furthermore, U.S. companies like Intel, which were once leaders in high-end chip manufacturing, are now playing catch-up with a number of Asian firms, notably contract chip manufacturer Taiwan Semiconductor Manufacturing Company (or TSMC, as it’s called) and the South Korean conglomerate Samsung.
The subsidies that have been dangled before companies by the CHIPS Act have had an impact. Big chip firms such as TSMC, Samsung and Intel have all announced huge investments in new domestic chip-making facilities in the U.S. In some cases, these companies have broken ground on these new factories. Will all of this new money make a big difference? If it does, is it important that chips be made in the U.S. rather than, say, Taiwan or Korea? In other words, do we need a national industrial policy for semiconductors?
With me today to discuss this and related questions are three experts: Christine McDaniel, George Calhoun and Scott Lincicome. Let me start by introducing Christine, who is a Mercatus Center colleague and will help me moderate this discussion. Christine is a senior research fellow here at Mercatus, where she specializes in trade and globalization. Before that, she held a number of high government posts, including at the U.S. Trade Representative’s Office and at the Treasury Department, where she was a deputy assistant secretary.
George Calhoun is a professor of business at the Stevens Institute of Technology, where he heads up the quantitative finance program. Before going into academia, George worked in the tech industry for 30 years, running a number of companies, including a superconductor firm.
Scott Lincicome is vice president for general economics and trade policy at the Cato Institute. Before joining Cato, Scott practiced international trade law for 20 years with the firm of White & Case.
Welcome to all three of you, and thanks for joining us. As I just mentioned, Christine is helping me moderate this discussion. We’ll each ask a few questions to George and Scott, and then we’ll open it up to give our audience a chance to also ask questions.
If I can get started, let me ask a very basic scene-setter to our two experts. It’s been a little over a year since the enactment of the CHIPS Act. So far, has it been working as the Biden administration and their allies in Congress planned? If so, is that a good thing?
George, do you want to get us started?
Impact of the CHIPS Act
GEORGE CALHOUN: OK. Well, the CHIPS Act has been on the books for a year, but they haven’t spent any money yet. To ask what the impact has been—we’re talking about the psychological impact on the industry, not the substantive financial impact so far.
I do think the psychological impact has been pretty significant. It has been interpreted as a strong signal by the industry. That’s why you’re seeing the acceleration and expansion of some of the construction programs that have been announced. I think the passage of the act as a signal that the government was going to be doing what it could to be favoring this development was received by the industry loud and clear.
MASCI: That’s shown itself in the fact that so many, or a sizable number of, companies have announced some sort of project, right?
CALHOUN: I was just going to say, I think it’s been widely commented that the most striking aspect of the CHIPS Act is that it represents a decisive move in the direction of industrial policy, which is something new under the sun, at least over here in the United States. I think the private sector has got the message, and they’re trying to be responsive to what they think that’s going to mean going forward.
MASCI: Assuming it continues, is this a good thing? Should we be happy that companies like Samsung and TSMC and Intel and others have announced these huge, in some cases $100 billion, projects to build factories here in the United States? Is this something we should be applauding?
CALHOUN: Well, I think that’s a more complicated question to unpack, maybe, as we go forward in the interview a bit. I would say the general answer at the moment probably shades in the direction of yes for me. But it’s complicated, so we’ll get to some of those complications.
MASCI: OK. Scott, is it complicated? Is this a good thing?
SCOTT LINCICOME: Yes. I’ll just reiterate a couple things that George said. I think it’s undeniably true that we’ve seen an uptick in announcements for the construction of new semiconductor manufacturing facilities, or what we call “fabs,” right? I think there’s a couple caveats.
One is, a lot of these investments were actually announced before CHIPS was implemented. I know it’s hard for us to remember things all of 18 months ago. CHIPS was far from assured in, say, the spring of 2022. In fact, a lot of people thought it was dead in the water at the time. The same goes with the IRA [Inflation Reduction Act], in fact. Those existing investments had already indicated that there was a desire by American companies that consumed chips, and by chip-makers themselves, to diversify and expand some in the United States. I think you can say that, yes, CHIPS accelerated this preexisting trend.
The other big caveat, though, is it’s always critically important to distinguish between the political spin, including the spin from certain corporations about the size of their investments, and then the actual investments themselves at the end of the day, right? A great example of this is, Samsung put in paperwork in the state of Texas for hundreds of billions of dollars, I think it was, in new investments, but all of that is far in the future. So yes, there has been an increase, but I think it really remains to be seen as to the actual magnitude of that.
You mentioned that the United States had about 12% of global fab capacity. The other big point is that, even if all of this works well, the Semiconductor Industry Association, which is the main lobbying arm of the chip industry, only projects that we’re going to get a couple of percentage points uptick in global market share, and that’s if everything goes well.
I think, again, to reiterate what George said, is this good? Well, it depends on what we get, and we don’t know yet. Are we going to get productive companies producing valuable items at a reasonable cost, or are we getting something else? Quite frankly, I think there are legitimate concerns that we’re not getting a lot of the former. Again, it remains to be seen.
But first, you have pretty substantial concerns about the politics of all of this, right? Are these the best places? Are these the best investments? Are the subsidy conditions that the Biden administration tied to these programs—are they going to be costly or distortive? The most famous one, I think, is requiring companies to provide child care for their employees.
What’s going to be the response from the global market? Everybody didn’t sit still in response to CHIPS. Everybody else around the world provided their own subsidies. China, and then in response to U.S. export controls, boosted its subsidies and other industrial policy as well. Now you hear concerns—at least me, as a former trade lawyer, I get really worried when you hear Department of Commerce Secretary Raimondo say things like they’re worried now about a global glut of semiconductors. Because normally when the commerce secretary is saying that, the next thing is you have trade wars, right? You have new tariffs and new trade remedy cases.
Then, finally, I’m concerned about economic and policy bottlenecks that still exist to actually getting these fabs up and running again on any reasonably valuable basis. There are very high costs. TSMC has been the loudest complainer in this regard related to labor and training but also construction materials. Of course, we apply tariffs on basically everything you need to build a fab these days, including steel.
We just announced new trade remedy duties on glycine [used in the making of semiconductors] from Japan. We have big immigration issues. Again, TSMC has been out there talking about this: the need for high-scale immigration. These bottlenecks are also going to make what might look good on paper right now, in terms of these announcements, turn out to be far less attractive than they seem right now.
Incentives and Signaling
MASCI: I was interested in what you were saying about labor costs, because I’ve read about that as well, and that they’ve delayed, I think, by a year the opening of the factory in Arizona because of that. Also, as a counter, I remember that the chairman of Intel was saying that he was reconsidering this Ohio investment that he had committed to before the CHIPS Act unless Congress acted.
Is it possible that the CHIPS Act brought some of these projects, at least in terms of final commitment, over the finish line?
LINCICOME: Yes, sure. Yes, I think it’s possible. Look, you’re at Mercatus, and you guys have done amazing work on how the incentives game works, right? It is very, very often the case that companies make these claims that they’re not going to invest if they don’t get the subsidy cash, or it was the subsidy cash that sealed the deal—when in reality they’re being strategic about trying to get even more subsidies down the road.
Again, I’m a critic of industrial policy generally, but I do think that there’s a legitimate argument that this boosted a few investments, accelerated trends. I would push back at least so far on the idea that it’s a game changer.
CALHOUN: Well, I would agree that there is a lot of signaling going on in both directions at this point, and that’s probably the short-term name of the game, is people sending signals to each other to try to influence the next shoe to drop.
If I may, I’d like to just give a little bit of a larger perspective, a more historical perspective on this, as I see it. I think the chip industry evolved a strategic grand concept about 25, 30 years ago to go asset-light, as it’s called: to separate the incredibly capital-intensive fabrication process as a business from the much more lucrative and asset-light model of designing chips, where the technology really—you can make an argument that the value-added of the technology mainly resides in the design, not in the fabrication.
That’s a very oversimplified statement, but that strategic concept has been taken up by just about everybody except Intel. AMD [Advanced Micro Devices] split themselves up a while back. Many other companies—NVIDIA, Qualcomm—never had fabrication in the business plan at all. If I would have looked at that—and I did look at it maybe five years ago—that looked like an incredibly smart grand strategic concept for the industry to have followed.
What that has run into is the Taiwan problem, and the concentration as a natural consequence of the intensifying capital expenditure associated with it, in a place that suddenly has a strategic vulnerability that is at least as important as the advantages of this split business model that the industry evolved. I think that is the big driver behind all of this. I think that you can look at the CHIPS Act as a corollary or a go-along and one of the straws in the wind that another $30, $40 billion is not a decisive financial commitment.
That’s what the industry, I think, has started to recognize, maybe a couple years ago, and the big ship is now turning, and the grand strategic concept is being rethought.
National Security Concerns
CHRISTINE McDANIEL: George talks a lot about the national security argument, and I always like to play devil’s advocate. Although in this conversation it’s fun to play devil’s advocate from both sides. Clearly there is a national security argument to be made. Let’s assume it is legitimate.
My question is, OK, assuming the national security argument is legitimate, then what is the best reallocation of resources moving forward on this, keeping in mind the economic realities of the huge scale and scope economies this industry really thrives off of and really needs, right? What does that mean for the industry if they have 30% of their market cut off? Will we still be able to achieve those goals with a smaller market?
CALHOUN: I think—you’re framing it as national security. I think it also is—from the perspective of these companies, of the firms in the business, it’s their business model security, if you will. It’s a version of the security argument that is much more focused on whether or not they have a survivable business plan in the event of X or Y or Z taking place in the Far East that compromises their exposure in a way that would really be harmful.
As I said, I think this is a big, big change in the direction of the whole industry. It’s going to play out over a decade, and we’re watching the ball go back and forth across the net here, but it’s a long game.
McDANIEL: Scott, what do you think about this? Let’s say, OK, even if it’s not national security per se, there’s the threat that the most advanced chips are being made, not to mention on a fault line, but then also in a geographical location that is at risk.
LINCICOME: Well, I would start with—given Intel’s stumbles in previous iterations of its bleeding-edge chips, there’s a very good chance that the tippy-top edge is still going to not be made in the United States in the future. TSMC is not building their most advanced stuff in Arizona yet. I think that’s an issue—some context we need to consider.
I agree, I should say, entirely that there is a legitimate national security issue related to defense-related technologies, supercomputers, all this type of stuff. I think there is a government role here. But the problem we have is that the CHIPS Act is not really targeting that stuff. Most famously, you have billions of dollars in the CHIPS Act that is specifically earmarked for legacy chips that automakers need—and that was inserted by a bunch of (big shock) Michigan lawmakers, right? There aren’t these guardrails that I think would be essential to ensure that we really are focusing on just the chips that have these really strong national security implications.
Because, quite frankly, if there’s an invasion of Taiwan, you’re still going to have a massive disruption in global semiconductor supply. And, quite frankly, we’re going to have much bigger fish to fry than just global semiconductor supply, right? Certainly I’ve said, ballpark, there’s a bill that exists in an ideal national security world that’s much, much smaller than the CHIPS Act and targets those issues, but that’s just not what we’re getting. Unfortunately, in certain aspects, like the tax credits, it’s just an open-ended subsidy for anything and everything. And that is a big problem.
MASCI: Obviously, we’ve witnessed the power of disruptive change in whole industries, but particularly the technology industry, over the last three or four decades. In the 1980s, for example, IBM and Digital Equipment Corporation were among the big players in the technology sector. Now one of them doesn’t exist, and the other is obviously not one of the top firms anymore.
I’m curious about this phenomenon in relation to the semiconductor industry and especially in the context of the role that (from what I’ve read, anyway) chemical and material science is playing—a greater role, I gather, in the development of new chips, the manufacture of new chips. As well as something I think I mentioned the last time George and I and Christine spoke about this: three-dimensional chip, which seemed to be the bleeding edge of the industry and where it’s going.
My question is, are we—and, again, maybe the answer is “No, it’s not a problem.” But when we’re trying to get some of these factories up and running and trying to duplicate what they’re doing in Asia, are we ultimately investing hundreds of billions of public and private dollars in something that we’ll look back in 10 years and say, “Oh, we were building yesterday’s industry”? Again, I don’t know enough to know what the answer to that is, but do either of you gentlemen have an opinion about that?
CALHOUN: I think it’s quite possible we will look back that way. Let’s look at the history of the industry. It’s been a steady cycle of disruptions for several decades. If you go back and look at the top 10 firms in the business year by year by year, there’s an awful lot of triage done.
I will say that I think Intel is kind of unique here. I’m more optimistic about Intel with the new management and their plans. I think they have a pretty good shot to recover and be in the mix, particularly if there is greater pressure in the Far East. Because the customers for the fabrication are going to look for the alternative strategy, as they’re increasingly calling it. I think if Intel can stand up a credible option for fabrication, it’s going to be a very interesting play.
But yes, it’s hard to project where the next disruptive technology is going to come from. Moore’s law is done, but logical progress is not done. It’s going to come through software. It’s going to come through material science. It’s going to come through chip architecture and advanced packaging. I think the point of your question is well taken.
MASCI: Scott, anything you want to add to that?
LINCICOME: Sure. Well, first, I would note, I defer to George on the tech side. In fact, I’ve cited George’s work on some of this stuff in some of the things I’ve written. I would note on the history, though, yes, I think there’s a pretty big concern. Quite frankly, there’s a historical parallel in the United States and our previous semiconductor industrial policies, in fact, going back to the 1980s and ’90s, back then.
MASCI: Sematech and all that.
LINCICOME: Yes. Well, and the DRAMs, memory chips, right? Back in the 1980s, the Department of Defense wrote this very urgent report saying that memory chips were the future, that we were losing ground to Japan and that we needed industrial policy. We needed Sematech; we needed antidumping duties on Japanese semiconductors—but memory chips.
At the time, it turns out that everybody was actually moving to logic chips, which are now so much more important and what we think about when we think of the chips race. (Though of course memory chips remain an integral part of all of this.) We put all these duties on DRAMs and other products. In fact, my first case ever as a new trade lawyer was on DRAMs. From then, it was Korea, because the duties on Japan actually helped boost the Korean industry. It’s very messy. We picked the wrong horse.
The Technology Pork Barrel
LINCICOME: Not only that: The government remained invested for political reasons, for bureaucratic reasons, at continuing these measures even after the memory move had happened towards logic. And that actually cost downstream industries in the United States. We saw offshoring of the computer industry in a lot of ways that was cited based on the high prices that those duties created.
Politicians remain heavily invested in these things long after failure has been demonstrated or at least the market has moved on. In fact, there have been entire books written on the technology pork barrel that happens as investments from the government don’t pan out, and yet the projects continue for years, if not decades.
It is certainly a concern. Again, I defer to George on the tech and some of the things you mentioned, but the historical concerns are quite, quite valid because we have, unfortunately, a lot of experience.
MASCI: George, do you want to follow up with that?
CALHOUN: Just a real quick follow-up in agreement. One of the worst parts of the CHIPS Act, I thought, was the attempt to specify what the technology objective should be: talking about going to a low three-nanometer or whatever. It was very strange. In a document that was basically pretty technology agnostic most of the way through, and then they throw a couple of these—those are almost certainly going to be looked back on as stale objectives that were already stale at the time they were written into the law. I think your point is well taken, Scott.
McDANIEL: Yes. In fact, the technology pork barrel—I think you’re referring to perhaps Roger Noll’s book, Scott?
LINCICOME: The Brookings book, yes.
McDANIEL: Yes, Roger Noll. He is amazing. I was honored to sit next to him at an Aspen forum in August.
His point—and I know you know this point too, Scott, and we all know it inherently, but fundamentally: OK, we have to think of all of the economic forces that led to where this industry is today, right? Will these resources that we’re reallocating, taking away from one sector of the economy and putting them toward another sector, will those counteract sufficiently all of those economic forces that have been there for decades?
For the economists out there, of course, we’re talking about relative factor input prices, right? But for everyone else, we’re talking about labor, capital, politics—all of the prices. Roger Noll’s point—his book is a couple of decades old now, but I read his book, and it rings just as true today as it did when I was looking at Sematech back in the—well, I don’t know; was that ’80s, ’90s?
I think—we have a very special guest here, Weifeng Zhong, my dear colleague here at the Mercatus Center. He’s been looking at different aspects of this. I wanted to bring him in so he could share an insight. Weifeng?
Innovation and Industrial Policy
WEIFENG ZHONG: Oh, thank you. Thank you. I have, I guess, a provocative point I want to lay out here on the table, and I welcome everybody’s critique on this: which is that I think it’s inconsistent for someone to be for the export control of advanced chips to China—if you support this policy, you can’t be also supporting the CHIPS Act as an instrument for industrial policy to support domestic firms.
The reason is this. There’s a notion that—so China is very good at reverse engineering. So they often take a product somewhere else and then take it apart, look inside, and then they figure out how to make something very similar. But somehow this just doesn’t work very well for semiconductor chips, right? Because even though with a microscope you can see everything inside, if you’re not part of the innovation ecosystem, you don’t have the right chip-making machines and all the talents. You just can’t reverse engineer.
I think that reasoning is what underpins the export control on advanced chips, with the reason being that if we cut it off, cut China off from this ecosystem, they can’t make anything close to that. But if that’s the case, by the same reasoning, industrial policy in the United States would not work either, because innovation doesn’t come from government command. One can only support one of the two, but you can’t reasonably support both of them.
CALHOUN: Well, I guess I would just—I agree with the analysis in the immediate moment, but I go back to the fact that I view this as a large shift coming mainly from the private sector itself. The government has jumped in as a cheerleader and so far hasn’t spent any money but will spend some money, but the money they’re going to spend is not that decisive.
I think the broad picture here is that the industry has already made its call that they’re going to have to be in a different place in 10 years than they are today. It’s going to take a lot of time to get there.
MASCI: Before Scott jumps in, I’m just curious, George: When you look ahead, and if you say to yourself, “If all of the things that we think may happen do happen”—all these plants that are being talked about are actually built—how much different are things going to be?
Will the United States now control, let’s say, 25% or 30% of global chip manufacturing? Or, as Scott says, are we talking about just a few percentage points? Are we going to make a huge difference in the global picture?
CALHOUN: Well, I don’t think the scorecard is about a percentage of manufacturing: That’s the first thing I would say. I think it’s about the security of the supply chain through the whole industry. By security, yes, it’s national security, but from the perspective of the customer, it’s the end users. From General Motors’ perspective, national security is one thing, but they want to see a secure supply chain for the chips that they need to build automobiles.
We all know how many tens of billions, or maybe a couple hundred billion, dollars’ worth of lost sales the auto industry suffered in the pandemic due to some elemental blockages in the semiconductor supply chain. That’s the way I view it.
Maybe I’m a little bit too private-sector-oriented. I don’t have the experience with interacting with the government and the legislative process that Scott has or that you guys have. I think this is, by and large, an initiative that’s coming out of the private sector anyway, and it’s going to continue.
LINCICOME: I’ll start with—just to go off of what George said, and I agree almost entirely that yes, there’s a lot of private-sector initiative, both on the supply and the demand side here. My only issue is that I think the private sector should pay for it. I don’t know why I as a taxpayer and why the government needs to get involved, particularly given all of the political economy problems that can arise and the economic problems that can arise when we start putting government thumbs on the scales, right? I think it’s great that Apple and GM and others want an onshore supply of chips. I just want them to pay the price and not have taxpayers footing the bill.
LINCICOME: On the export controls point, I think it’s a really interesting one. I would go a slightly different direction, though. And I think one of the issues with the tension between the export controls and the subsidies is that American chip makers—and they’ve said this quite publicly—is they make a lot of money in China. And the export controls, to the extent they’re not very narrowly targeted (and, quite frankly, right now they sure don’t seem to be; at least they’ve expanded a lot in recent months), then that’s denying the very revenue that chip-makers in the United States might need to invest in innovation, to invest in new fabs and you name it, right?
That, I think, is a tension that isn’t really discussed much. Again, the industry has complained about this: that it’s going to cost them billions of dollars in sales. Also, you’re going to get things like Micron’s transaction denied by Chinese regulators and all that stuff that comes up from these open hostilities.
I think that’s where we’ve lost the script a bit on the export control side. These are some of the most innovative companies in the world, and some of them are American companies, and yet export controls—to the extent they’re not narrowly targeted, or to the extent they’re not followed by competitors outside of the United States—they can do a lot of harm for a little good.
McDANIEL: Yes, that’s exactly right. I mean I’m looking at these numbers: 20% for NVIDIA, 60% for Qualcomm, 20% to 30% for Intel. These are big superstars, and what are they going to look like when their Chinese market is cut off?
LINCICOME: Right. You certainly—again, I think certainly from a national security perspective—need to be concerned about the very tippy, tippy-top stuff. If you believe the industry—and that’s, of course, look, big asterisk, believing the industry! But if you believe them, they say that a lot of the chips that are being now targeted, a lot of the technology that export controls are targeting is nowhere near the tip of the spear. That, in fact, it’s things that might go into a smart refrigerator or something, right?
That’s a big concern. You can accept some revenue loss, but if you’re not doing a very good job of targeting it and making it narrowly tailored and ensuring that other countries are doing the same, you end up, again, just with all harm.
Effects on the Private Sector
McDANIEL: I see Mercatus Center senior research fellow Veronique de Rugy is on. Vero, what do you think?
VERONIQUE DE RUGY: It’s a great conversation, but the way I think about it—obviously, I agree with everything that Scott has said. But the way I think about it is, even when we look at the amount of activity that’s been announced, the thing that actually matters is, what kind of additionality is it going to create? Basically, what is it that the government actually is really, truly creating?
One scenario I can think about is that these companies weren’t going to do this, and it’s just the government funding that is actually committing them or creating the incentive for them to come and invest in the U.S. as opposed to somewhere else. That said, as I’ve learned from Scott and from George, the share of the investment created by the government intervention relative to how much the private sector is actually committing is really, really, really small.
You have to wonder whether it’s—do we really believe that it’s the tail that wags the dog, or actually that it’s other incentives and forces that are actually creating all these investments?
Now, talking with George over lunch a few weeks ago, one of the things that he said that I thought was very interesting is that, in a way, what the CHIPS Act is doing is sending a strong signal to the private sector that there’s going to be a favorable environment for investment for this particular industry and that, basically, it is worth pursuing those goals because the environment is going to be friendly.
I thought, “Well, actually, yes, I can see in that sense as a signal that it could be the thing that convinced—that could be tilting the decision towards more domestic production.” At the same time, in the last few weeks, we’ve seen the administration send a lot of signals that is signaling that yes, they want people there, but it’s going to be on their condition.
They’re not—yes, sure, they may be lifting NEPA [the National Environmental Policy Act] for some of the construction of the fabs. I’ve heard that the Biden administration has actually made some noise in that direction. But they’re not talking of removing all the “Buy American” requirements. They’re adding requirements with the child care, as Scott has mentioned—the minority preferences, the union preferences, all of this.
In the end, it’s just very hard for me to actually think that if there is any movement in the private sector that is the result of the CHIPS Act, it’s on net going to be productive.
The final point I’ll mention—and Christine has alluded to this—is, in my experience studying the impact of such—I want to call it cronyism, because this is what private-sector subsidies or tax breaks or whatever to a company who mostly would have done it anyway, that’s effectively what it is. As alluded to, we can’t discount, based on experience, the fact that so much of the capital that is shifting away from other investments or other locations may actually be nothing more than malinvestment, especially during all the incredible requirements that are put on those investments and with the specification of how much of what is produced, if anything is produced, is going to be outdated when it finally comes out?
This is why, just overall, I can’t see—I also am very skeptical. National security, there’s absolutely a national security way of doing this. It just should be done in a national security context.
MASCI: George, do you want to comment on some of the things that Vero has been talking about?
An Investible Industry
CALHOUN: I think all of those points go to suggest why the CHIPS Act is not going to ultimately prove to be decisive in moving this. I think it’s going to be a corollary phenomenon. The money will get dribbled out over a number of years. It will be, as you say, encumbered with various irrelevant or semi-relevant social kinds of constraints and issues.
I just don’t think it’s going to have a decisive impact in substance, but it does send the signal to the industry that they are going to get a lot of green lights, a lot of waving on through for things that they do come and ask for. I’m not talking about necessarily even money, but let’s say smoother regulations or eliminating some of the roadblocks to creating these new facilities. And it sends a signal to the financial markets as well that this is an investible—highly investible industry, to pick up on Raimondo’s comment about China not being an investible industry, an investible place, anymore.
May I just offer one—we were talking a minute ago about the exposure of U.S. companies like Qualcomm and Apple to losing revenue in China. I think that’s almost certainly going to happen anyway. We’re seeing it with Apple.
MASCI: Because the Chinese ultimately just won’t want the products.
CALHOUN: The Chinese are beginning to tighten down. There was an article, I think in The Wall Street Journal, maybe six or eight months ago that talked about the incredible embeddedness of Apple’s business model with the Chinese supply networks and came to the conclusion that it just couldn’t be changed. Well, now we’re seeing how it’s going to be changed and altered because the Chinese side of the equation is intensifying its pressure.
I think that’s something that I’m sure Apple and Qualcomm and the others that have significant exposure are having to evaluate strategically—again, from their private-sector perspective. Depending on how fast it takes place, it could be a rather brutal adjustment.
LINCICOME: No, I agree with everything that was just said, honestly. The only thing I would add is that sending signals for investment works well when you have long-term policy that is relatively unencumbered by the political process. Whether you enact a, say, 10-year or permanent full expensing for investment in structures in the tax code or whether you significantly reform high-skill immigration.
I wonder, though, if that’s going to really work very well when you really do have bureaucrats in the Department of Commerce really picking and choosing, or you have to go to the White House to get your visas that you need. That, I think, will act as a regulator—no pun intended—on the signals, right? It will really mute those signals going forward.
The other thing is that we again have to wonder, “What are these green lights doing to the rest of investment in related tech or outside of tech? Is the government preference causing malinvestments?” So causing investment that might go to something that’s an amazing breakthrough, a revolutionary change in the technology inside or outside of semiconductors, to not happen because investors in Wall Street and elsewhere are like, “Well, now I’m going to make good money on chips. That’s where I’m going, because that’s what the government wants.” Those are some of those unseen costs from industrial policy that truly exist but are very difficult to quantify.
MASCI: I think it’s a nice place to wrap things up. First, I want to thank our guests, George Calhoun and Scott Lincicome: Thank you both.
CALHOUN: Thank you, thank you.
MASCI: Thank you both, gentlemen, for joining us.
LINCICOME: My pleasure. Thank you.
MASCI: It was a fascinating discussion. I want to thank Christine for helping to moderate the conversation and asking such good questions.
McDANIEL: Thanks for doing this.
MASCI: Of course, I want to ask the folks who also—Weifeng and Vero asked some excellent questions as well. Thanks again, and take care, everyone.