Should American taxpayers provide the microchip industry with a blank check of over $50bn at a time when semiconductor companies are making tens of billions of dollars in profits and paying their executives exorbitant compensation packages? I think the answer to that question should be a resounding NO.
-U.S. Senator Bernie Sanders
Episode Summary
On July 27th, the Senate passed the CHIPS and Science Act of 2022. It is now awaiting President Biden’s signature. Proponents of the bill say it will reduce U.S. dependence on foreign producers of semiconductor chips by building manufacturing capabilities at home. Skeptics – which oddly include both progressive Vermont Senator Bernie Sanders and conservative Republicans – say it amounts to “corporate welfare.” So who is right?
The CHIPS in the CHIPS Act stands for ‘Creating Helpful Incentives to Produce Semiconductors.’ But are the incentives, and the resulting changes, actually “Helpful” for the U.S. semiconductor industry?
In this Dial P for Procurement crossover episode, Kelly Barner digs into the global semiconductor market and talks about some relevant considerations not covered in the language of the legislation:
• What’s in the bill?
• What can and can’t be done with the funding?
• Why has U.S. semiconductor manufacturing capacity has been falling since the 1990s?
• How long will it take to see the benefits?
• And issues like sustainability and intellectual property that still have to be resolved before this bill can achieve the ultimate vision.
Episode Transcript
Intro/Outro (00:01):
Welcome to dial P for procurement, a show focused on today’s biggest spin supplier and contract management related business opportunities. Dial P investigates, the nuanced and constantly evolving boundary of the procurement supply chain divide with a broadcast of engaged executives, providers, and thought leaders give us an hour and we’ll provide you with a new perspective on supply chain value. And now it’s time to dial P for procurement.
Kelly Barner (00:31):
On July 27th, the Senate passed the chips and science act of 2022 proponents of the act say it will reduce the United States dependence on foreign producers of semiconductor chips by building much needed manufacturing capabilities at home skeptics, which oddly include both progressive Vermont Senator Bernie Sanders, and a number of conservative Republicans say the bill amounts to corporate welfare. I guess it is true what they say politics makes for strange bedfellows. Now you might not catch it from audio only, but chips is an acronym that stands for creating helpful incentives to produce semiconductors as cheery as it sounds. It also reminds me of something George Orwell would have come up with, you know, like the ministry of peace being in charge of war are these really helpful incentives in this week’s episode of dial P for procurement, I’m going to cover what’s in the bill. What can, and can’t be done with all of this funding, why us semiconductor manufacturing capacity has been falling when we’ll potentially see the benefits and maybe some environmental and intellectual property considerations that aren’t exactly on the front burner right now.
Kelly Barner (01:56):
But before I go any further, let me pause and introduce myself. My name is Kelly Barner. I’m the co-founder and managing director of buyer’s meeting point. I’m a partner at art of procurement, and I’m your host for dial P here on supply chain. Now I’m constantly scanning the news for complex articles that I think are worth discussing. I look for things that are interesting, but which could easily escape. People’s notice my goal is absolutely never to lead you to think that there’s a simple answer to a topic that I’ve chosen to cover, but instead to provide the background and context, I think are necessary for you to form your own opinion or at least to make you interested enough, to want to learn more. Now DW P releases a new podcast episode or interview every single Thursday. So be on the lookout for future episodes.
Kelly Barner (02:52):
And don’t forget to check out our past episodes as well. Before we get back to today’s topic. Here’s what I’d like, like to ask. If you enjoy what you hear today, take a minute and give us a review on iTunes or offer up some stars on your favorite podcast platform. I’ll take a share or a like on LinkedIn or Twitter. I’m always grateful for your interest and attention. And of course your comments, which I’ll talk about at the end of today’s podcast. So thank you for being part of the listening audience. All right, let’s start with what’s in the chips act. The entire thing is 280 billion in deficit spending. So what that means is that because this is considered a matter of national security, the cost of it won’t be covered by tax income or revenue derived from the bill. 52.7 billion of that is going to semiconductor companies and other related companies in the ecosystem to create incentives for them to develop and manufacture chips in the United States.
Kelly Barner (04:03):
Now, part of the justification for this is around protecting intellectual property. And we’ll talk about global sources of conductor, semiconductor chips, but right up front in the bill, there are limitations on accepting donations from China. If you happen to be a university, there are limitations on working with Chinese companies and the military. If you’re a manufacturer, even with your own money. And finally there’s a 25% tax incentive for any company that invests in semiconductor manufacturing. So it’s not just companies like Intel that are already in this business that can benefit from the bill. It’s anyone that makes an investment in the talent and facility and knowledge needs required for the us to compete as a manufacturer of semiconductor chips on the global stage. So a few more details about what can and can’t be done with the funding in president Biden’s own words. This is quote, not a blank check, so no dividends can be issued using the money companies can’t do stock buybacks.
Kelly Barner (05:09):
And interestingly, it also requires any company’s accepting funding to build these semiconductor fabs, to pay prevailing union wages, to build the fabs. So this is attempting to reconcile the fact that the United States has fallen behind the rest of the world in terms of manufacturing capabilities. According to data from the SIA and BCG Taiwan is the world leader in fab capacity. They have 22%. They’re followed closely by South Korea with 21% and Japan with 15%. Now, China has been increasing over the last years, and they’re now tied with Japan at 15%, which puts them ahead of the us with 12 and ahead of Europe who have nine, but that’s capacity not production. When it actually comes to production, things are far more constrained. Taiwan currently produces 63% of the world’s semiconductors and just one company TSMC holds 53% of the global market share. Now I mentioned the fact that China’s capacity has been growing in 2000.
Kelly Barner (06:30):
They only represented 3% of worldwide capacity and they have aggressively invested to grow those capabilities. When we look back a little further than that in the 1990s, the United States was a much bigger player in this industry. 37% of chips were manufactured in the us compared to today’s 12%. And of course with all statistics, 12% is not 12%. It tends to be the lower performing, easier to design and manufacture chips that make up most of that 12% that are still currently being produced at home. By the end of this decade, the us commerce department estimates that if things go on changed, the us will represent 14% of global production, but 24% of global demand as with all other business things cost and risk are huge factors in this and building these fabs is an expensive business, semi engineer.com estimates that it costs between 10 and 20 billion to build a leading edge fab.
Kelly Barner (07:46):
And of course after that, they have to be kept up to date. Now we know that the entire funding associated with this bill, at least the part that’s going to semiconductor manufacturers is 52.7 billion. So even if you could efficiently build plants for just 10 billion, you still only would get five fabs. So this bill is definitely not going to represent all of the costs associated with getting this going, making things more complicated. This is a very quickly moving target. And the semiconductor industry is sort of in the middle of a process war. So chips are measured in nanometers and they’re getting smaller and smaller and smaller in early 20, 20 TSMC and Samsung became shipping five nanometer chips while Intel struggled to reach seven nanometer leading firms are working themselves towards being able to produce three nano animator chips. But that number one, worldwide producer TSMC has said, they actually won’t go beyond three nano animator.
Kelly Barner (08:58):
Of course never say never. So these fabs are expensive to build. They quickly become dated. And because it takes time to build them, you’re trying to hit a future spot in terms of manufacturing capabilities, given those complexities and with an eye on profitability, which we understand most us firms in the semiconductor industry have gone either fabulous or fab light in order to save money and focus their research resources on research and development companies like China, Taiwan, Japan, South Korea, and the European union have subsidized semiconductor production for a long time. So the chips act puts the United States on par with how other countries in the world are handling this industry. The question that this raises for me is whether this is no longer a profitable private industry, if it can’t exist worldwide without government subsidies. And of course that also raises a question about ROI simply having capacity does not guarantee that it will be profitable.
Kelly Barner (10:11):
So even if we have the capability to produce even three nanometer chips at the volume desired in the United States, how much are those chips going to cost? They’re likely going to be more expensive than chips produced in Taiwan. Now, worldwide sensibilities around sustainability play an important role here. We’ve talked about the cost and performance reasons for producing chips outside of the us and for our capacity at home to be lagging. But environmental reasons are important to this as well. According to the national institutes of health, semiconductor production leads to groundwater and air pollution, and it generates toxic waste as a byproduct of the semiconductor manufacturing process. So understandably us companies have been only too happy not to deal with that at home, in addition to saving money by manufacturing, these chips overseas. And that brings us to one of the things that’s an environmental risk and cost concern that the chips act does not address.
Kelly Barner (11:22):
That’s rare earths. Now this is something that’s absolutely core to semiconductor chip production and my supply chain now colleague Greg white recently shared excellent thoughts on this point when the bill passed, as he pointed out, even if we magically get fab capacity in the us to where we want it to be in terms of innovation and in terms of volume, we still won’t have resolved the potential for disruption in the semiconductor supply chain, because we haven’t addressed raw materials, which are rare earths. China is responsible for about 80% of all rare earth elements used in semiconductor production. So they can still control worldwide production because they control raw material processing and distribution extracting these rare earths is really bad for the planet. So there are a number of different minerals unless you’re in science. Like I’m not, uh, you’ve probably never heard of their names, but they all end in em, they’re distributed in the Earth’s crust.
Kelly Barner (12:33):
They’re not in chunks or flakes like something like gold would be. And given the destructive nature of the mining and extraction process, most other countries have limited rare Earth’s mining, but not China. Now Brazil and Australia are potential sources of rare earths. But the question is still open as to whether they’re going to allow this destructive mining process. It seems to go against most modern sentiments about sustainability. Now in his comments, Greg suggested the idea of lab grown alternatives. And that’s something that all of us should want to learn more about, but until then all semiconductor supply chains, regardless of where the design work is done, regardless of where the chips are stamped out, still start in China. And the delays are not just in the processing of the stages of the supply chain. They’re over time. One of the things that I really do like from the chips act is their ultimate goal to build a thriving semiconductor ecosystem.
Kelly Barner (13:43):
That includes research, design, fabrication, talent, manufacturing education. We talk more and more about ecosystems within procurement and supply chain. And I love the fact that this evolution of thought made its way into something that the government served up for us. So I think that there’s something really positive. There, there are even incentives for startups to accelerate innovation. There’s a recognition that while we need the big companies like Intel, like AMD, we also need small startups that are willing to attack smaller problems and can think more quickly. And out of the box, global partnerships will be an important part of this. Japan has the world’s leading high tech workforce and have already mentioned Brazil and Australia as a potential source of rare earths if they are willing. But even if all of this falls into place, it’s still going to take years for these new fabrication plants to be built and to be up and running.
Kelly Barner (14:46):
Now, time is of the essence, the tensions between China and Taiwan have been on most people’s radar screen for a long time, even in advance of when Russia invaded Ukraine, we were watching that dynamic worrying that what happened between China and Hong Kong might also happen between China and Taiwan. And there were thoughts that once Russia invaded Ukraine, it might accelerate China’s desire to do a similar thing. Now, as I record this president Biden has yet to sign the chips act into law, but even once he does it, isn’t like that fab capacity will be instantly available. It’s going to take years for that. And in the meantime, the United States semiconductor industry is in a very precarious position for proof of that. We can look no further than the global static over Nancy Pelosi’s trip to Taiwan, which is the home to TSMC. For fact, I actually delayed the recording of this podcast to see what would happen with her, getting into the country, spending time there.
Kelly Barner (16:00):
And fortunately, as we now know safely getting out despite threats from China, if China takes full control of Taiwan, which they already consider to be their sovereign territory before the United States can get production capacity online, we have a serious problem. We need years to get caught up. And even then we won’t be where we are today. But with production in the us, the rest of the world will keep moving forward. In the meantime, and our chips may be more expensive. Now I mentioned at the outset that there are some IP considerations here, certainly part of that is that we wanna have walls between companies that do business with China. We wanna make sure that if the government is investing in these capabilities, that our knowledge, our technology doesn’t make it into the wrong hands. But I also think we have carryover here from another recent episode of I P the one in which I spoke to patent attorney, when she about what the world trade organization did with IP waivers around the COVID 19 vaccine, they waved those patents.
Kelly Barner (17:14):
We have to ask, could they do the same thing here? And we don’t know everything today. This bill is going to take years to play out. In reality, once these semiconductor companies take money from the federal government, will the government stop thinking of the associated IP as belonging to those companies? Remember the agreement that’s already in writing that they can’t issue dividends or do stock buybacks using the money from the government? Well, at the end of the day, where did the money come from? Regardless of which pocket it came out of, we’re kind of talking about the same pair of pants. What questions will be raised for these companies about how they operate, how they invest, how they reward shareholders. If they have taken funds from the government, will they come under greater scrutiny? We don’t know, will the government try to exert greater leverage over them after they’ve accepted the money?
Kelly Barner (18:12):
We don’t know. And in fact, there are a lot of things we do not know that we’re going to have to wait and see play out. There’s the issue of rare earths and the delay of building capacity. There’s the talent lag. And while the bill does allocate funds for education, everything takes time. A recent Bloomberg editorial piece says that about 40% of high skilled semiconductor workers in the United States were born abroad. So they’re also immigration considerations. And remember, this is not just about talent for the fabs. Once they’re open, you have to use union labor to build the fabs. If you take this money from the federal government. In fact, right now, TSMC is trying to build a fab in Arizona and the project is hung up because they don’t have enough engineers and technicians to finish the job. Yet. Another source of uncertainty, the chips probably are going to cost more, even with the automation capabilities that the United States has.
Kelly Barner (19:17):
So will there be enough demand at the ultimate effective price point to justify the kind of capacity we want to build out? This whole industry may end up being subsidized forever, and that’s not supposed to be the point of a private industry. So what does this all mean? You know, we think about some of the wording in the bill, the suppliers that are going to be included, it refers to upstream suppliers. So there are a lot of different kinds of companies that are gonna be involved, packaging and distribution, this whole semiconductor ecosystem. One of the other lesser covered components of the bill is money. That’s being set aside for the public wireless supply chain innovation fund. This is really smart, in my opinion, it’s another way of leveraging small, innovative talent in the us. What they’re actually doing is breaking down all of what Chinese based Huawei does into smaller components, making it easier for smaller companies to compete and allowing them to focus on developing capabilities and specific areas, as opposed to attempting to compete with everything Huawei does.
Kelly Barner (20:30):
And to end my advice is always to keep an eye on the government, especially when they come bearing helpful gifts, as they say. And we always need to remember that there has to be an incentive for companies to be profitable and to be, be innovative in that same Bloomberg article that I cited earlier, there was a terrific quote. It said by shielding companies, from the salutary effects of market competition, they induce complacency, inhibit productivity and dampen the incentive to innovate. It is all too easy to imagine. Bloated wasteful, government dependent, chip makers demanding yet more handouts, a decade down the line end quote. That’s certainly not where we wanna be in an industry that thrives on innovation and competition. Now, semiconductor makers are currently profitable with about 25% return on assets, according to a report from McKenzie and yet not everybody agrees that this is a good idea.
Kelly Barner (21:35):
We know Bernie Sanders and some Republican senators don’t think it’s so great, but the Cato Institute made a very compelling case in an article called the top seven reasons to oppose new semiconductor subsidies. Now, this was written before the house voted to approve the bill, but it’s a position piece. So I actually think it’s still relevant here. It talks about pricing systemic risk. And this reminds me of the last episode of dial P, where we talked about the skeptical eye that the world is looking at profitable refineries right now, profitable refineries, that in some cases are horribly unprofitable all based on the barrel of crude oil price. I think the thing when it comes to systemic change, that concerns me the most here is that once you change the decision making basis and the risk tolerance and the cost implications for an entire industry, it’s very hard to change it back.
Kelly Barner (22:39):
And it’s also difficult to know what the secondary and tertiary ripple effects are going to be. Now we’re waiting for president Biden to sign the bill, but it’s passed the house and Senate. So we’re going to have to wait and read and learn and discuss because this is my point of view. And as I always encourage you, please do not just listen, join the conversation and let me know what you think. Put a comment where you find this on LinkedIn, put a comment where you find it on Twitter or on Facebook dial P supply chain now, and I all share this content. And you’re also welcome to direct message me on LinkedIn. If you have something that you’d like to say maybe a little bit less on the record, I always appreciate the time and enthusiasm that I hear from our listeners. And I appreciate that. So many of you have told me, you’re sharing these episodes with your network. Let’s work together to figure all this out and come up with the best solution until next time. I’m Kelly Barner on behalf of dial P and supply chain. Now have a great rest of your day.
Intro/Outro (23:47):
Thank you for joining us for this episode of dial P four procurement and for being an active part of the supply chain now community, please check out all of our shows and events@supplychainnow.com. Make sure you follow dial P four procurement on LinkedIn, Twitter, and Facebook to catch all the latest programming details. We’ll see you soon for the next episode of dial P for procurement.
Hosts
Check Out Our Sponsors
