This Is What President Biden's CHIPS Office Actually Did

across the US. It was not just a program. It was like an actual like breaking ground and new things were going up and dollars spent by the private sector partially to get public subsidies, et cetera. You could certainly say that it was like badly designed, or that it was wrong, or there were too many rules whatever, like all you know, that's all of this is fair play.

But the idea that suddenly this is just some new impulse and not like something that's real, existing reindustrialization that's going on, I find it infuriating or at the very least very annoying.

the Chips Program Office. I believe he's officially left the job so he could talk now about what he saw inside Hustin. Thank you so much for coming back on Odd Lots Joe Tracy.

people were losing their jobs. Obviously there's a national security angle as well, because we were reliant on overseas factories for chips that go into military equipment. And as the Director of Economic Security, my role was for it was sort of twofold first, helping sort of set the strategy. What was our vision for what we wanted to accomplish with the Chips Program Office. So before I joined, we published a vision for success paper in February of twenty

twenty three. I'm actually quite astonished. I think very few people who talk about chips actually read that paper, and I still think it's worth reading because it laid out a roadmap for what we wanted to do within the different categories. And then the second job that I had was helping our teams understand the value of each proposed project to US economic security. So why would the factory

that X company is proposing improve our economic security? And obviously there are different ways in which you can do that, whether that's advancing technological capabilities, improving supply chain resilience, plugging various gaps in the supply chain, et cetera, et cetera. And that's why we did sort of a deal by deal analysis on those metrics. But it was really that twofold going not just strategically across the portfolio, but on a deal by deal basis.

or national security. If you're looking at two pitches on your desk or I guess it was an online application like a portal, but you're looking at those applications and one is for building I don't know, in video GPUs and the other is like making an improvement on basic chips that go into MCUs and cars or whatever, lagging versus leading. How do you decide between those different proposals.

very rapidly to AI supremacy. So in real time, you know, if you ask Congressman why are you passing this bill, they would have said, well, we can't have these car shutdowns, or we can't have car factory shutting down. Appliances are too expensive. And then by early twenty twenty three, it was what we have to win the AI race so we sat back as an office and we really said, we don't want to be chasing just one category. And I think again our strategy was, we want to make

sure we're making investments across the entire supply chain. So we said, hey, we're going to functionally target a majority of our funding, the vast majority of our funding towards the leading edge. Why those are the most expensive facilities. So Intel, TSMC, Samsung and Micron between them got nearly twenty eight twenty nine billion dollars and don't you know,

you can check my math afterwards. Knowing that getting those facilities in the United States at the scale that they were investing in has downstream consequences too, because now you're building out the supply chain necessary for the entire industry, and that spills over to some of the other facilities

that are going to come up online. We did have a statutory requirement to invest at least two billion dollars in what we're called legacy node chips, and our office spend a lot of time trying to understand what our strategy could be on shoring up legacy supply. So we made investments, you know, large ones in TI and global foundries that are in the sort of meat of the

legacy node supply chain. But we also made actually dozens of investments that I think at short shrift because they just aren't as headline grabbing, but they plugged up a lot of our capabilities in rf in power, summi conductors, the sort of un sexy types of electronics that are critical not just for infrastructure today but infrastructure in the future.

And how we thought about trade offs. I think the way we tried to think about it was we really tried to bucket our funds and say, hey, for the leading edge, we want to be able to say, preserve x amount of our budget. It was about that twenty eight billion dollars for the leading edge and make sure that we retain sufficient funding on the back end for the legacy nodes, for advanced packaging, for the supply chain, because we knew that we needed to make investments across

the entire supply chain just get to the resilience. That was, you know, the reason that the bill was passed.

it's like, well, do you want to build the chip plant or not? Because if you do, then why did you put all of these other burdens that have nothing to do with building chips per se onto the money? In your experience, what is the role of these other elements in the speed of grant programs or project development in the US?

language or what your commitments are to community investments. If you look at the DFAs that we write, the direct funding agreements, the terms that we had around what you might call everything. Bagel policy basically codified the commitments the firms themselves had made to the communities that they were investing in. It essentially said, hey, you told the community that you were going to be investing in, you know, the schools or water reclamation projects, whatever those community investment

funds could be. All we're doing is memorializing that commitment that you've made. Secondly, on childcare, this is another one that first in terms of the amount of funding that we put towards it, I think it was a total of about ten million dollars across the thirty nine billion, Okay,

So it was never a focus. It never became in any of the negotiations that I set in a discussion where the company came back and said, hey, we really want to build this plant, but the million dollars that you're giving us for childcare and the requirements that you

have simply aren't enough. Many of these firms are investing in child care facilities for their workers anyways, and you see it like there's Wall Street Journal report a few months ago about a company that wanted to expand and with new workers that were mostly coming from like Hispanic background, and they found that the biggest thing they could do to help bring them on was have childcare on site, because their workers were like, I can't come to the

office because I don't have a place to leave my kids. So it's just it's actually what the private sector is doing anyway, And oftentimes the funding that we brought to those initiatives actually helped them think outside the box and think across firms to come up with regional solutions that scaled better than they would on a firm by firm basis. I will say, however, where I think critics of sort of the everything bagel approach do have a point is whereas a lot of the terms that I just described

were not deal stoppers. They didn't slow down negotiations, they weren't the points of contention where there are points of contention between different stakeholders, I think you need top leadership to be able to come and say our number one goal is to get the factory built, and various stakeholders

have to get in line. And where I'm talking about stakeholders is where I think the abundance folks also speak to them, groups like labor and environment, right, you have to come and say, hey, do we want this project to happen? There will inevitably be trade offs. There is no world in which you can build a massive factory

and have zero environmental impact. Right. You have to also even in the context of labor, we have to understand that this is a globally competitive industry, and so the demands that labor is making have to be viewed from the context of like what does it take for the factories in the US to be globally competitive? And I think you need top leadership to come and say we're not going to allow concerns that are being raised by

the community to sort of halt negotiation. So there is a balance to be struck in terms of what's our number one goal. Is it to get the factory done or is it to make sure that no one's upset at the fact that the factory is getting done.

really trying to build. Then we had what we called a pre app process, and we can come back to that in just a moment my thoughts on the process, but it essentially said, hey, submit a simplified version of your final application and we'll give you some preliminary a feedback kind of like a draft application, and will help identify where we think on our scoring rubric you need

to make adjustments in order to score better. By the way, our response to the pre app was non binding, so if we basically said hey we don't like your pre app, you could still apply and submit a full application. But you know, we did take into consideration whether or not you responded to the feedback from the pre app. The

full application was sort of your final submission. We started to receive for our first full applications in the late summer early fall of twenty twenty three, and so you saw we got to a first preliminary announcement by the end of twenty twenty three with BAE. So it took us, you know, on the order of about a little more than a quarter to get through a first full announcement.

The exact longest negotiations that it took, I have to think back for a moment, but we had one final step after the preliminary announcement following sort of exactly how you do it in the private equity world, you know, have a preliminary announcement saying, hey, we intend to make this investment, we intend to go forward with this, but it's subject to due diligence and i'd be the direct

funding agreement. Those negotiations did drag on through twenty twenty four, and I think a lot of it came down to sort of dotting the eyes and crossing the t's on what did it mean for the government and semiconductor firms to make a commitment to each other on these facilities. Right, There was a lot of not on the sort of everything bagel terms, but there was a lot of negotiation and what does it mean if your company is sold?

What does it mean if you violate guardrails statutory requirements? Right? We really had to work through that because we'd never worked through it as a country before with firms at

this scale. But what you saw routinely was as we reached a milestone, So as we reached the first preliminary memorandum of terms and we reached the first direct funding agreement, the second, third, fourth agreements would happen much faster because we at that point had worked out a template and could say, hey, here's how other firms are thinking about doing it. There's already comfort with this format. Let's try and you work off that, and you saw them happen in rapid succession.

by organized labor. I don't know if the organized labor, especially in the private sector, is a big Democratic constituents anymore. But Democratic Party certainly wants to be liked by organized labor. They certainly want people who concern about the environment talk to us about the reality of how these different impulses can collide with each other.

these projects, right. So, I think one of the consequences of this tension of there's an urgency to move fast, an urgency to catch up to our geopolitical competitors, but not really a readiness to sort of tear down the frameworks that we had. And I think for good reason, you have to come back and say, well, what is it going to take for us to catch up and so on. A lot of these projects, you saw environmental

groups raising concerns on you know, pollution impacts. You saw labor groups sort of saying, hey, unions are being left out in the cold. And I would come back and say, I think a lot of that really was noise because there was like an open negotiation going on sometimes through the media where these various groups were trying to say, hey, make sure you don't forget about us. But I do also think for policy, you have to be able to come out and say what is the most important thing.

Is it for the factory to get done on time, or is it that we leverage union labor or is it that we make no impact to the environment. And there will be times in every complex project in the public or private sector, you have to make trade offs between different objective functions, and I think for what we saw was there was like an unwillingness sometimes to really say to stakeholders, hey, we hear your needs, but they're gonna be second priority in order to get the project

done right. And that complicates the discussion on how are we going to get these things done quickly? And I think there was a tension between the urgency that firms and folks within the Chips Program Office felt, and outside stakeholders who really were saying, well, don't forget about us.

basically stopped going to the gym. Do you know before the TSMC fab came online in late twenty twenty four, when the last leading edge fab in the United States was built, came online, No it's good. What is it?

understand the environmental impacts of these facilities? And so when you talk about the delays in construction, oftentimes those delays are from the permitting processes that are handled at the state and local level. Well, in a lot of the places that we're building these facilities, state and local regulators hadn't seen a facility like this before because we hadn't been building them in over a decade, and so they didn't know what the impacts were, and you know, it

required an education process. And I think a lot of the noise that we heard in the last two years was because we were kind of starting this again after not going to the gym for over a decade. And you know what happens when you don't go to the gym, you go back one time you're really really sore the next day, but if you keep going, your body kind

of gets used to it. And that's why, for example, take TSMC's FABS in Arizona, you don't hear the same noise about labor unions or permitting concerns over Fab two because the entire system sort of got into shape, right.

And I think if the chips program off is going to be looked at as a success, it's going to be because the second, third, fourth, fifth facilities that are being built at these sites are showing rates of learning in how long it takes for them to get brought online, brought up to speed, brought up to a similar capacity to what they have and their overseas benchmarks. So I would look at it's like the first fabs are like

a proof of concept, can we do this? And it's really in the second, third, fourth fabs at these local projects that you'll start to see the ecosystems of the church.

an issue of capitalism? And investors in semiconductor companies don't want US manufacturing because that's lower margins. You move that overseas, et cetera. You don't want design and fabrication in house together because then you mix a high margin company with a low margin company, et cetera. Obviously, to some extent, the idea of using public money is to solve this problem.

But just when you're look in general at the questions of US manufacturing and high tech areas or whatever, how much is it about capitalist incentives.

go use it in other use cases. Yeah, I mean famously, look at the case of Intel, which for a long time was returning a lot of cash to shareholders through dividends and stock buybacks and fell behind the leading edge curve. So that tension is absolutely real. But I do think firms and investors understand the value of having these facilities. I think the challenges creating a structure where the government can help equalize the returns so that the private value

is similar to the public value. Let me put it in another way, the government highly values these manufacturing facilities being in the United States for the economic and national security reasons I laid out above, right, But private shareholders don't value them as much. But there are levers that we can pull to help make them look more attractive. And I think the biggest under discuss lever was the Investment

tax credit. Right. The investment tax credit is a twenty five percent tax credit for firms that invest in manufacturing

in the United States. I think there's a world where the future of industrial policy I'm putting quotes around that really comes and says, hey, the focus should be on tax credits that give firms certainty on what their cost structure and return structure is going to look like for capital investments made in the United States, and maybe the disbursement funding that's subject to review by bureaucrats is a smaller pot that is really geared towards firms that have

capital shortcomings or capital concerns. Right, So, I think you could plausibly make the claim that the intels and tsmcs of the world don't necessarily need cash from the government because what they're optimizing for is like an NPV function on their capital investments, and tax credits can solve all of that, in fact, can happen with less government intervention.

But there are smaller firms who really do need cash infusions in order to bring, you know, to bridge the value of death that we've been talking about for decades in this country but never really had an approach to SAULT.

And I think there's a slim down version of industrial policy in the future that really focuses on ay tax credits can equalize our cost structure and make investments attractive where you know, we take target smaller amounts of funding to critical technologies that we want to make sure happen in the United States.

the part of private investors. Did you ever talk to them about, you know, what their concerns were or what they wanted to see from this program. And were you successful, I guess in making chips manufacturing cool again.

the balance sheet. That being said, I do think it's an ongoing discussion with the investment community on how do

we build certainty in these you know, government programs. Right. So, the the biggest thing that the Chips Program Office did was it gave firms confidence in what their returns would look like if they invested in the US because they had a tax credit and award dollars that would come to them, and they could go to their investors and say, hey, look, there is a cost disadvantage, but we feel confident that we'll be able to reduce it with the public dollars

that are coming in. And I really think the biggest lever that the government can pull is giving firms certainty when they're making twenty to hundred billion dollar investments, because they don't want to be caught on the wrong side by a policy change that now gets them underwater on a facility that's half done, And the sunk costs of building a facility and half getting it half equipped are really large, and so I think that is the challenge

for a lot of these firms and for investors. They want to be able to say, hey, is what we're modeling really going to hold in the long term from a cost structure basis for us to feel comfortable in what the returns are going to be.

as an example. TSMC, by the end of the Biden administration has started pumping chips out for Apple and AMD out of its facility in Arizona. And as it continues to move forward with the second and third facilities, that ecosystem is going to mature to the point where the cost differentials versus Taiwan are going to be reduced. The scale is going to bring more suppliers on shore, so they're going to have more of their coems and their gases and their you know, consumable materials sourced from the

United States. And as that happens, you start to you know, build out a broader ecosystem because you know, we heard all the time from suppliers who were saying, Hey, we're building a facility in the United States to service all the fabs that are coming online. Because we can now justify the investment based off of the number of downstream investments that have been made. And as they build out their facilities, then their suppliers are going to come here.

So I think there's going to be an ecosystem maturation that's going to continue, hopefully through the rest of the decade. That's going to bring not just you know, front end fabrication facilities, but their suppliers facilities and then their suppliers suppliers facilities, And now you talk about getting the sort of industrial ecosystems that really had atrophy in the United States. And when you start to go you know, N plus two in terms of the supplier level, you're no longer

just serving the semiconductor industry. You're building fabricated machine parts that go into semiconductor manufacturing equipment and also going to say airplanes or automobiles, and you now you know, buttress the entire industrial ecosystem, even though you're starting from just building semiconductor manufacturing plants. Right. So I think that is what the long term is going to look like. But we have to you know, in this where I have to come back to the point on we have to

also be honest about where we were right. We weren't building these fabs. There's a reason it took so long, and it's not going to happen overnight. And if we're not willing to maintain the investments and the programs that we have, I think a lot of firms are going to say, Hey, the uncertainty isn't worth it for me to continue to invest. Because I can't go to my

shareholders and say this investment has a solid return. They're going to look at it and they're gonna discount it with all that uncertainty and pressure me to not make these investments, or to reduce the investments I make and focus on places where the returns are much more solid. I think that's the situation that we absolutely should avoid.

And you even hear that from the Trump administration, where, for example, Jade Vance at a speech at the American Dynamism Conference talked about making adjustments to tax credits for firms in terms of bonus appreciation and the R and D tax credit, those are very much in line with making these investments less risky for firms the United States.

So if the Trump administration continues down that vein, I think you'll see firms feel confident that they can expand these investments and build out these ecosystems to a a size and scale that's globally competitive. Right. And then you tap into the broader tech stack that we have here, where now the smartest engineers from Nvidia, Apple, and am

D don't have to fly to Taiwan. They can fly to Arizona to make sure that they're getting their designs taped out correctly, and they're working with universities all across the United States on future designs and technologies. And then you get an industrial ecosystem that really leverages our capabilities. Right.

One last point on this one. I think a lot of people made this criticism when the Chips Act was passed that the United States should have just continued to invest in R and D and that's what we should leverage. We should leverage our R and D capabilities. But here's another trivia question for the two of you. When was the quiz?

fact that the R and D first approach was empirically failing. US. We invented EUV technology through our DD National Labs and in partnership with ASML, we installed the first alpha tools in both Europe and the United States. And then, I mean the United States was a half decade behind East Asia in bringing EUV manufacturing to scale, right, so that formula wasn't working. And one of the shortages that TSMC talked about was that they didn't have enough workers who

knew how to install and bring up EUV machines. So you can see how this sort of scades the ecosystem atrophies.

And then when firms come and try to do foreign direct investment, they come and say, well, you don't have the skills that we need, even though we can point to all the R and D investments, and I think the problem was we were sort of making these R and D investments in a vacuum and kind of hoping that they'd get sucked into an industrial ecosystem that, you know, despite what a lot of economists say about America still having a very high value add for manufacturing, you look

on the ground and there are tons of anecdotes that the manufacturing ecosystem has atrophied and we have to make investments in order to bring it back up to be globally competitive. And I think an anecdote exactly like the delay in bringing EUV manufacturing to scale in the US,

exemplifies why the old approach wasn't working. And we can debate like what the right ways are and how industrial policy should be structured, and what tax credits, etc. Need to be done, and trade reforms need to be done, but I don't think you can debate whether or not the old you know, let's call it pre twenty twenty approach was actually maintaining America's industrial competitiveness, because it wasn't.

and stuff like that. And then finally, the other thing happening which we should definitely ask about, is tariffs. Right, and maybe tariffs end up being good for domestic manufacturing like semiconductors, but I can imagine that there are also still either components or materials that chips manufacturing actually needs to import. So I guess my question is, how are you weighing all these different things that are going on right now?

We don't know what our cost structure is going to look like, and you know, our project is undergoing change constantly because our cost structure is undergoing change. So, you know, for a lot of these firms, before they're willing to make bets that you know, in some of these smaller firms can be like, you know, life or death size bets for the firm, they really want to have an understanding of what the policy framework is going to look like.

And I think we have to sort of get through the period of you know, a new headline rocking markets every day or it to shake out to understand how it'll affect the long term decisions a lot for a lot of these firms. You know, the sense I got from talking to that firm and from other firms was that they're going to kind of wait it out and see They're going to try and buy as much time as they can to see where things reach a steady

state before reevaluating their investment plans. I think on the flip side, However, there is a bipartisan agreement on the need to bring industrial manufacturing back to the United States, right, So I think the question is going to be on the methods by which we do it. So you know, I go back and you say, is it through tax incentives, is it through trade policy? Is it through industrial policy?

I think all of those tools interact with each other. Obviously, different administrations have different approaches, so I don't know where we'll end up with that. The last thing I'll say is the methods that we developed in the Chips Program Office for trying to get firms comfortable with making investments in the United States and working to accelerate their investments by working with stakeholders across you know, environment, workforce, and

other policy objectives. I think that is actually going to continue. If you look at the Investment Accelerator Executive Order that was announced by President Trump a few weeks back, the sorts of activities that he's saying, the White Glove Service, I think that was pioneered in the Chips Program Office, where we worked with firms to get through the labor issues, to get through the environmental issues, and permitting questions to

make sure that these projects could move forward. It's why I've said repeatedly that there were no Chips Act construction projects that were held up by Neeper review. And I think they're going to take that recipe that was developed and try to scale it across multiple sectors. You know,

certainly going to be different contextual challenges. But if they do that, I think it's going to be a vote in favor of the work that we were doing at the policy level to make sure in manufacturing investments in the United States are viable for firms, and that's going to have to be complemented with a you know, approach to make them financially viable. I don't know, and I don't know that any of us can say what the Trump administration is going to finalize its policy mix on.

And I think firms are going to wait to see what that policy mix looks like from the Trump administration before you know, placing further large bets. So if there's a lot of policy uncertainty, you may see some companies come out and say, hey, we're gonna do price increases and we're gonna maybe pause equipment purchases until we really know what the fiscal impact of tariffs or other. You know, new trade negotiations are going to be for our project.

is actually building factories to produce these things. And so you really need some sort of catalysts to get stuff going, to get people excited about it. Yeah, maybe change the calculation in terms of profit margins and the result is the Chips Act.

that you saw with the first one. That is encouraging. Maybe you have a sort of template to quickly navigate the state and local issues. Some of the questions around you know, the quote stakeholders, et cetera. Which every system has stakeholders. If this isn't not unique in that, I mean, obviously,

you know, every system has to have a way. I think what's important, you know some of the I remember we did a conversation about nuclear construction in China, and it's like they have their own you know, it's not like there aren't environmentalists in China, et cetera. What they have is like a system for allocating like who wins and what the priorities are, et cetera.