Panel On Industrial Policy For National Security

- Okay, well thanks for coming to the first EPWG of the spring. We're delighted to have a four panelists On our industrial policy panel. We have Steve Davis, senior fellow here, Josh Rao, also senior fellow here, Philip Ko, senior fellow here, and also Dan Wang, who will be joining us virtually because he's not in town right now. So the order will be Steve Davis, Josh Rao, Philip Zko, and Dan Wang. And each panelist is going to talk for up to 12 minutes and then we'll have open question and answer and general discussion. So go ahead and take it away, Steve.

- Great, thank you Valerie. It's so wonderful to have an opportunity to address this assemblage. I've got a lot to say. I'm gonna mostly read my remarks. United States practices many forms of industrial policy and has long done so. I'll offer some empirical observations about US industrial policy, touch on a few examples and distill some lessons. I won't confine my remarks to national security matters because the main challenges that arise with industrial policy are generic industrial policy creates special interests that benefit from the policy and lose out when it ends naturally. These interests organize politically to preserve their benefits. As a result, industrial policies are self entrenching within a democratic political process. That brings me to my first lesson. Once implemented, industrial policies are hard to kill. That also holds when they fail to achieve their objectives, when they are not cost effective and when they outlive their usefulness. No matter how carefully the experts design an industrial policy, it gets mangled and distorted by the political process, which brings me to my second lesson. Even sound well conceived, industrial policies usually under form underperform in practice because they are not designed, implemented, and operated as conceived. A former US Secretary of Defense famously remarked you go to war with the army, you have not the army you might want. Likewise, we execute industrial policy with the government. We have not the farsighted, benevolent, highly competent government that many academics and policy mavens imagine another observation. Industrial policies are often rife with unintended harmful consequences. When commercial concerns produce harmful consequences, firms lose money and stop. When public sector initiatives yield costly consequences, they go on and on. See lesson one. Here's an example, are many policy efforts to force March American households and auto firms to move from internal combustion engines to EVs damage to the US economy and undermine our national security. China leads the United States in battery technology and production and in control and has control over much of the battery supply chain. Yet we have pursued costly policies that push the US to greater dependency on batteries and battery technologies, amplifying the leverage of our chief geopolitical adversary. And that brings me to lesson three, new proposals, new proposals for industrial policy warrant rigorous scrutiny, great humility about unintended consequences, and a sober recognition of lessons one and two. The political process is ill-equipped to assess performance against vague open-ended objectives like Beat China, win the future, win the war on this or that. Performance assessments against such objectives is hard even for experts, let alone political leaders and voters. As a result, the political process is slow to recognize policy failures, to verify why they happen and to respond with corrective actions. In contrast, market processes mitigate these challenges or obviate them entirely. The profit and loss system compels hard choices without the need to explain why to politicians and voters. If a consumer dislikes a particular product or a business finds fault with an input supplier, there's no need to verify the dislike to outside parties or explain why that consumer or business simply withholds their spending power. The compulsion to stop spending taxpayer money on failed public sector init initiatives is much weaker and sometimes altogether absent. So that brings me to lesson four. Be especially wary of industrial policy initiatives in pursuit of vague open-ended objectives that lack clear yardsticks for success and failure. Despite the many challenges, US history offers some examples of industrial policies that largely achieve their stated objectives. Here are just three. The Manhattan Project, the goal was to design, build tests and deploy and upon an atomic bomb In theater, success was evident. The objective was easy to explain and understand, afford the fact to those who participated. And after the fact. For political leaders and voters, the benefit cost calculus was extremely favorable for the United States and its allies. A more recent example, operation warp speed. The goal was to quickly develop and manufacture a COVID vaccine. The objective and its urgency were transparent and easy to grasp. For everybody, the extent of success was reasonably evident and the project was essentially self-liquidating. The guaranteed procurement feature was also important, which contains another lesson and one that is highly relevant for munitions. For example, third JFK's pledge to put a man on the moon also had a clear verifiable objective. NASA began to flounder after the Apollo program ended. And in the absence of an equally clear objective, that's probably not coincidental. So lesson five, successful industrial policies often involve clear, easy to grasp objectives for which success or failure is evident. It helps if the, if the policy naturally expires, it's often claimed that some countries accelerated their economic development through well-designed industrial policies. Japan, South Korea, Singapore, and Taiwan are often held up as exemplars in this regard. Let's accept that claim. It does not follow that industrial policy will work for the United States. Why not? Several reasons. I'll mention two in my spoken remarks. Our enabling infrastructure. And here I'm borrowing a phrase from Steve Kotkin is inferior to these other countries in key respects, especially in the lack of high quality, widely available primary and secondary school in much of the US and in the scarcity of high quality job training programs. Second, we must implement industrial policy without government, not theirs. That's lesson two. Lesson six, indu successful industrial policies are not readily portable across time and space. Another observation, other aspects of our enabling in the infrastructure are best in class or nearly so our universities capacity for basic research innovation, ecosystem capacity to allocate capital at scale to promising new technologies and convert commercial ventures. And historically, our ability to attract STEM and entrepreneurial talent from the rest of the world are enormous strengths. That brings me to lesson seven. The United States should leverage its strengths in pursuing its national security objectives. Lesson eight, import tariffs are poorly suited for most national security objectives. They distort multiple production and consumption margin in the pursuit of national security objectives that often pertain to concerns about production capacity, not production itself and resilience in the face of supply disruptions. Now that's a pretty abstract state statement, so let me make it concrete. In the case of steel, steel tariffs, steel tariffs discouraged domestic consumption of steel using products of American imports of steel using exports, excuse me. And they discouraged American exports of steel using products and they discourage overall domestic production of steel using products. None of that is in the US economic or national security interests. Tariffs on US steel imports do encourage US steel production, whether they encourage us. Steel production capacity is a more complex issue that turns on expectations about future tariffs. Among other things, in some tariffs on steel imports are at best a poorly targeted policy tool for achieving what might be a sensible national security objectives. And in the interest of time, I will skip over the other reasons that import tariffs are not well-suited to achieving most national security objectives. The US has many negative, for lack of a better term, industrial policies that damage our economic performance indirectly or indirectly weaken our national security. The lesson seems obvious and seems like the obvious place to start fix or end our many broken and negative industrial policies. Again, I'm gonna just touch very lightly on this in the interest of time, I believe Josh is gonna say something about the first one there, regulations and permitting practices that impede new ener energy infrastructure, factories and housing. At least the first two of those infrastructure and factories are key inputs to not just our economy but to our national security. And there are major problems in the United States, and I could go on here, there may be, we may come back to this topic later. Last lesson and observation. Heavy reliance on industrial policy fosters a rent seeking culture among business executives because their success then becomes dependent on anticipating and shaping government policies and on remaining in the good graces of government officials, this dynamic is on vivid display in the second Trump administration. And if you don't believe me on this point, spare yourself 60 seconds to watch this video clip in which you can watch many of the most successful business executives in in, in the United States kowtowing to President Trump. That brings me to the last lesson, one that might seem distant in the present moment, but it's a path other countries have gone down. It's a path. Puerto Rico has gone down part of the United States and that's a path that I don't want us to go down. So I'll just put it this way. If ring kissing and that's a polite way to put it, becomes the path for commercial success, then over time the ranks of senior business leaders will become populated with world class ring kissers rather than men and women who owe their success to customer service better product sufficient operations on the commercialization of cutting edge technologies. Now, I don't want you to think there's no role for least on what I've said so far for beneficial industrial policies, new ones, and reformed ones in the United States. I do think there are several areas in which there's much room for new or improved industrial policies as I've defined it in the United States. I've listed them here. I'm sure we'll get into some of these later and I will be happy to comment on them as we do. But I don't want to overextend my time and how am I doing? I got 50 seconds with your stuff just enough there. There's a remarkable neglect of failure of industrial policies, especially among proponents and I'm not gonna talk about that here, but you can see my written remarks. I alluded to Puerto Rico, which is a case in point. I do wanna make one last comment and that's about the subtext around this discussion by my reading, enthusiasm for industrial policy to advance US national security correlates positively with what I see as an exaggerated sense of China's relative economic performance in recent decades. So I wanna close with a chart that shows China's GDP per capita relative to the United States in recent decades. For comparison, I show the same type of chart for the Soviet Union before its dissolution. The handout shows some related comparisons, which you are free to look at the point I wanna leave you with. For those of you especially who see China as some remarkable economic success stories to ask yourself why is it 50 years after the end of Mao's craziness us GDP per capita remains five or six times higher in China. I'll stop there.

- Thanks Steve. Our next speaker is Josh Ra.

- Okay, thanks very much for inviting me to speak about industrial policy for national security. So what I thought I would do here is to try to provide an economic framework for industrial policy in the context of both national security concerns on the one hand and the economic concerns of free trade and freedom from state intervention and private economy. On the other hand, and to be clear about where I'm coming from upfront, I'm, I'm concerned about both. I think there are obviously many cases, and Steve has mentioned some of them where it's obvious that there's been gross overreach in industrial policy and we must fix and end many of these broken industrial policies. When I walked into the room, John Cochran said, I have two words for you, corn ethanol. I said, yes, but I also have two words for us here on the other side, which is rare earths. So there are cases where policy intervention actually should be unambiguous and obvious, and it's been unfortunate to see such a mismatch between the targeted approach that we should be taking and the shotgun approach that both parties have been taking to industrial policy. A shotgun approach for those of you who aren't familiar with munitions is when you spread the shot all over the place instead of targeting something that you should be targeting. I don't know how familiar people are with munitions, but the, the primary argument I wanna make is that the US has a genuine demonstrable supply chain problem with China when it comes to national defense. And the problem is not a lack of capital or necessary political will, but it's actually related to our own regulatory and procurement processes. So I'm not making a protectionist argument, okay? I believe conventional free trade theory is correct broadly speaking, but the problem is that the assumptions on free trade don't really hold when it comes to certain aspects of China in our national defense. So, so, so first of, you know, free under, you know, kind of free trade example, free trade theory, you know, you assume the trading partners are not gonna be weaponizing supply chains. China has basically demonstrated that it's willing to do that, right? It had a 2010 rare earth embargo against Japan. It imposed export controls in 2024 and 2025 on both critical and strategic metals and metallos as well as on rare earths. These controls after the late 2025 negotiations, they're no longer at their most restricted 2025 peak, but they, they have not disappeared. China still retains and uses an export control architecture. And when your trading partner has the capability of cutting you off at your worst possible moment, you have to factor that in to the downside scenario. Bullet point number two, you know, comparative advantage just seems market competition, but China's rare earth dominance was built through state subsidies. I can hear John Cochrane saying, who cares if the Chinese government wants to subsidize their firms to sell me cheap stuff then great, I agree. When we're talking about patio furniture, I even agree if we're talking about passenger EVs to some point, if we can develop a little more battery producing capacity in this country, let's have $5,000 passenger EVs. If we can at least make some batteries and add them to the, get some more of those US made batteries in the, in, in, in the supply chain, you know, maybe requires US software and some deep inspection, maybe that adds a thousand dollars to the price, great. We can have $6,000 Chinese EVs. I have no problem with that. The problem is about our being dependent on China for truly existential material when our economic interests may not be aligned. And it's worth keeping in mind that the free trade framework assumes that disrupted supply chains can be rebuilt us mine to production presently takes 29 years. That's the number from s and p Global that looked at 268 mining projects, development projects across the globe 29 years. We're second only to Zambia in that in the United States. So if we don't, if we don't fix that, then we have a problem or a situation where if we are hit with a national security crisis and we need to ramp up production, we won't be able to do so. Here are some numbers that I find pretty stark. These are from Veni, which is an analytics firm that works directly with DOD contract data. So the numbers here come directly from procurement records, okay? 40% plus over 40% of the semiconductors sustaining DOD weapon systems depend on Chinese suppliers. That there's been an astronomical rise in supply chain dependencies since 2014. In, in electronics industrial equipment and defense materials. 78% of US weapons system contain components with gallium germanium and Tim Tungsten or thorium, all of which, by the way, Chinese imposed export controls on in 2024 and 2025, around 10% of tier one defense contractors. Those are major suppliers. Just below prime contractors are Chinese firms, and the number of Chinese suppliers in the defense industrial base is quadrupled since 2005. Those are pretty surprising numbers, but actually what I think should worry us most is the factoid at the bottom of this slide. The Gini data and reports also managed to not only give some numbers, but to name names. So among the defense systems that are dependent on Chinese semiconductor supplies include the B two spirit, stealth bomber, and the Patriot Missile Defense System Supply chains. For these, these, these defense assets are running straight through China, okay? I don't think Churchill bought parachutes from the RAF for the RAF from Herman Goring, or at least now maybe Neville Chamberlain might have thought about doing it, but Churchill did not. This would be like Alex Navalny asking VLA and rep Putin to go fill up his prescription medicine that the uptaken, okay, this is, this is the kind of thing where we absolutely need industrial policy and as for rare earth's, we can mine it, but we can't refine it. So the US is actually the world's second leading miner of rare earth ore. We have the rare earth for here. And the problem is that we export over 90% of that or to China for processing since we have almost no refining capacity. In 2024, the US had just one operating refinery with 1.3 kilotons of alloy output. These are alloys, they're used in magnets. The same year China produced 240 kilotons of finished magnets that same year. It is actually a problem E even if it's only probabilistic, that we could end up in a serious national security conflict with China. It is actually a problem. There are supply chains for our defense assets run through China, and these types of magnets go into F 35 guidance systems that go into patriot interceptors, and they go into Virginia class submarines graph on the right makes the point visually, China is 92% of global rare earth refining. The rest of the world is basically the, the, the rounding error. And again, the bottleneck is not, is not geological. We have the ore, it's not financial. We have the capital. The problem is surely regulatory, the average US permitting time for a rare earth refinery is seven to 10 years. Our allies in Canada are able to do it in about two to three years and one to two years respectfully in Canada, two to three years in Canada, one to two years in Australia. Okay, so now that I've gotten very worked up about that, of course not everything is a national security issue, and that is, has been the major problem is the assertions for the vast gray area in the middle. The assertions that the, that the many things are national security issues, when in fact they are, they are not. So the left column, I think are things where the national security case is clear and obvious. Defense grade semiconductors, rare earth magnets for guidance systems, gallium and germanium for radar electronics. These are obvious places where the US government should worry about the country's national security and intervention. There is clearly warranted industrial policy for national security. I think the middle column are things that are genuinely contested. Broad semiconductor fabrication, pharmaceutical APIs from China and India, advanced battery supply chain steal for direct DOD use. We can maybe debate these things. Think there are legitimate arguments on both sides. I'm not gonna take a hard stance there. I will say on the right column on fails to test, there are many other things in fails, the test fails. The test is actually, you know, a massive category and it's where we've probably focused most of our industrial policy in the us but things like upholstered furniture, softwood lumber, passenger vehicles, and patio chairs from my outdoor furniture, these are subject to section 2 32 national security tariffs. Currently, I'll be direct, I don't care if China makes cheap patio chairs or even as I mentioned before, passenger EVs, but I do think we should really be losing sleep if our, so much of our, of our defense assets supply chain runs through China. Of course, though, if everything becomes national security, then nothing is. We need to be very, very careful of a broadening of the category that protects all sorts of industries that fundamentally have nothing to do with national security. So I'll end with a few points about things I think we could do to make, get, get the policy right. First two actually don't cost anything. One is permitting reform for critical mineral refining administrative action. Further streamlining of NEPA reviews, consolidating the federal process, creating pre-approved site categories for refining hubs if ca Canada and Australia are able to permit to, if Canada and Australia are able to permit their facilities in one to four years, there's no physical or technical reason that we would need seven to 10. It's just our own government getting in our own way. Second, extend supply chain audits through the tiers I mentioned. Tier one is being the, you know, first tier of first tier of of defense procurement. Country of origin certifications currently require stops at tiers one and two. We, we need to reach more into the supply chain or supply chain audits. Okay, I guess it doesn't cost literally nothing because of course, you know, the, the, the DOD has to actually execute on those audits and you know, the, the procurers are gonna have to procure from other places and it could be, it could be more expensive. Third prescription requires some money, but would be targeted multi-year DOD purchase commitments under the Defense Production Act. Title III would give private rare earth refiners to demand certainty to justify billion dollar capital investments. You can imagine something like offtake agreements. There actually has been some movement on this practice. The DOD has already committed over 400 million under dpa, a title three to rare earth supply chains since 2020, including some examples of offtake commitments. The idea of an offtake commitment, just the DOD is pledging to procure the materials, certain amount of materials under a certain price under a certain conditions. There is a very interesting a hundred percent offtake commitment of 7,000 megatons per year of magnet manufacturing from the company MP materials last year that could serve as, as, as some, as some kind of model. So I guess finally I'll just mention important institutional fix that would make this all coherent, which is that if a tariff is gonna be classified as a national security measure under section 2 32 require the Secretary of Defense to certify a specific national security interest, not just some broadly defined economic security. We could have avoided maybe the patio chair tariffs while preserving all of the important interventions on the earlier slide that actually matter and using a much more targeted approach to industrial policy that would really be for national security instead of doing all kinds of things in the name of national security, but that actually are really just a, a waste of money.

- All right, thanks so much, Josh. All right, our next next speaker is Philip Sal.

- So I'm going to focus on industrial policy for national security. I wanna start with this important preparatory understanding. If you don't make a product, you will soon lose the ability to innovate in that product or even fix it. Most of the d and r and D is learning by doing in the private sector. So making a product is goes beyond the prime contractor and the end producer. It involves a whole ecosystem of making that's an ecosystem of materials of the kind that Josh was just talking about. Tools, sub assemblies, component suppliers, and production knowhow. You may think of US Steel and the Ohio Valley. The Ohio Valley was an entire ecosystem that supported steel production, for example, with dozens and dozens of companies. There's similar stories for Silicon Valley and so on. When you outsource or lose too much of the production capacity and you can no longer innovate or adapt, you can actually end up losing your ability to be in that product category at all. A number of American companies have learned this lesson. The hard way. I cite here, for example, Boeing almost collapsed as a company in the early two thousands. It had so outsourced its supply chain for the 7 87 aircraft that it came very close to bankrupt, the bankrupting the entire company. It could not make adaptations or fixes in this advanced aircraft that use these new carbon fiber materials. And Boeing just barely survived that crisis circa 2002, 2003, 2004. It partly was able to revive 'cause of better design of the Triple seven project and eventually learning its lesson. Boeing began insourcing more of some of its key production, even though it might buy components from elsewhere in the free world so that it moved more of its aircraft production back into the United States, including in places like South Carolina. Another kind of lesson you'll see was learned by a man named Elon Musk. Elon Musk, when he decided to go into the automotive business, went to an extreme degree of vertical integration, including the whole manufacturing technique called giga casting. He then adopted that approach to extreme vertical integration of the, almost the entire component supply inside his company when he moved into SpaceX and SpaceX's ability to innovate in rocket motors and in the whole design of its spacecraft has, I think, just been a historic achievement, but partly accomplished then actually through not adopting the trend of widespread dispersion of value chains. I wanna call it a very different example. In the case of a company like Corning. Corning makes the gorilla glass that's on every one of your iPhones. It makes the fiber optics that the United States relies on, including in the semiconductor plants in Arizona and makes another a number of other very optics. Corning actually has quite a number of its fabs overseas, partly because the displays that Corning designed have to be made overseas close to where the televisions are assembled. But Corning has been very careful to keep key parts of its fabs in the United States. Forget about wages and labor issues. Corning keeps absolutely critical fabs in New York, in North Carolina and in Michigan in part 'cause Corning actually wants to stay in the frontier of the business. And it knows is if it puts all its fabs overseas, however it might please its stockholders in the short run, it'll ultimately destroy the competitive advantage of the company and that area of innovation. Also, notice the story for GE Aerospace. GE Aerospace is the gem of what's left of the formerly famous GE conglomerate. GE Aerospace is absolutely vital to the United States. It's the source, it's the source of our ability to do jet engines. GE Aerospace actually also has dispersed fabs, but keeps key fabs in the United States, close to the center of its r and d. GE Aerospace in many respects is a company that is 50 50 American and French American and French. About half of the production in GE Aerospace is done by Safran Aviation based in France, and the avionics actually are done by a British company, but which GE Aerospace actually owns. But notice the way they've designed their production in order to maintain a balance that allows them to main be a frontier innovator by keeping a lot of this stuff very close to the center of the company, despite the sh short term vicissitudes of pressures they get from their shareholders in the markets. However, in general, us, these are exceptional cases. US ecosystems in the innovation economy broadly have withered. They reached a high point, I would argue between, and I think the economic historians and Valerie is one of them, including Robert Gordon and Alexander Field reached their high point between about 1920 and 1950. Since 1950, the ability to produce at the frontier of innovation and manufacturing has gradually migrated away from the United States in many products. It began migrating slowly into Europe and Japan, and then began migrating further beyond Japan into South Korea, Taiwan, and then China with results that people can now track in many sectors. And that were alluded to in the earlier presentations. So with without, with that in mind. Now let me shift gears. The problem of industrial policy for national security is not the same as the problem of industrial policy for competitiveness. What Steve's presentation mainly talked about are industrial policies that to borrow the shotgun metaphor is the shotgun blast at a thing called competitiveness with many of the results already described. What I do by concentrating on national security is try to posit a concrete goal, which I would state this way in one sentence, which I don't think I need to read aloud. Read aloud. Obviously the question of what's essential is very difficult. This is the central trade-off category in Josh RA's slide. Hard to define yet. You absolutely have to do it. You have to engage that exercise, come up with your definitions, be clear about them, and then de and then defend them and have the debate. I'll note that the United States was dominated by a software driven economy of innovation in the period between say, oh 1980 and maybe even up to 2020, where you got these companies with very, with relatively small labor forces, producing very high margin products that could be deployed on platforms and that occasionally produced unicorns for the VCs. What's happened in recent years is a renewed attention to the saliency of hardware and the examples I have up here, you can see, I think these are, in my view, these are indisputably in the national security category and there are various examples and data one can give under all of them. Josh, for instance, focused very hard in his good presentation on critical materials. Let me drill down on the issue of semiconductors and give you an example of how hard it's for the United States to attain self-sufficiency despite a, a really quite heroic effort made by the Biden administration. You know, given the circumstances the, they had to define operational objectives for that program, which boiled down to, according to the chief strategy officer of the CHIPS program, Dan Kim, who discussed this with me, basically trying to get resilience down to about two years if there was a serious supply interruption. Even so today, despite the tens of billion doll, billions of dollars put against this, a little of which has gone to Corning, by the way, Arizona still can't do printed circuit boards, it can't do the wet chemical industry that's needed to make all this work. It doesn't do high bandwidth memory, which mainly comes from South Korea, two companies, it can't do the advanced packaging that puts all these things together. In many cases it can make chips, and the chips then have to be transship back again to Taiwan to do the advanced packaging from which they can then be shipped back again across the Pacific to Nvidia. And of course, the workforce, not just the production workforce, but the engineering workforce. And increasingly in these sectors, the engineering workforce is quite large. In the, in aerospace, by the 1950s, the engineering workforce might be considered 40% with the production workforce, about 60%. The amount of engineering man hours needed to design the B 52 before it flew in 1952. In 1952 was 4 million engineering work hours or the B 52 design. So this just gets you some sense of what's then involved in recovering ecosystems and managing this. And of course, at this stage, you have to think of this in a free world context because the United States loan cannot possibly solve these problems. So now let's shift gears a little bit now and drill down a little harder on that resilience problem. That's the objective of your industrial policy for national security. I'll say that in, in trying to meet the goal that I postulated a few slides ago in against the possibility of a war with China, which is not a zero possibility that would present a resilience challenge. We have not faced on this scale since the World War II era. So then it really is important to go back and revisit the economic history of mobilization in that era and the economic warfare of that era. By the way, very few people understand this story. There are a variety of celebratory narratives about, you know, America overcoming everything. The Soviet Union had a celebratory narrative about how the communist Sparta could overcome everything am in the United States. The mobilization story is a giant mess with some outstanding examples of successes and examples of big snafus, huge failures, trial and error, the reinvention and proliferation of numerous government entities, et cetera. This is a very complicated and messy story in which it's not clear the United States performed way above potential. Fortunately, enormous potential had been established in the 1920s and 1930s on which we could build and did build during the war. Valerie has worked on this as have others in the resilience problem. You'll find if you look at that experience there, some absolute limits simply cannot be overcome. You cannot get the substitutability, they're just, they are absolute limits. In World War ii, the saline example for the United States was natural rubber. We never solved that problem and we never eliminated the need for natural rubber. You could blend natural rubber with synthetic rubber in certain applications, but you always needed some natural, and we were actually just about to run out of it as the war ended in 45 and it would've been an an acute crisis. However, some resilience limits are simply become functions of time and planning to do this substitutability, the synthetic rubber cases, very interesting. We adopted the German style synthetic rubber process. We, we, however, we probably went to that process a year, two a year later than we should have. Boy did that loss year cost us. And then when we went to the German process, instead of using alcohol as the feedstock, we decided we would use petroleum as the feedstock, which was not a brilliant idea. There were reasons we did petroleum. This is some of the reasons aren't very good. It created some very serious problems for the availability of high octane aviation fuel and other things that the government had to manage heroically vaccines in war speed, by the way, which is a story I think I, the group I led have studied this in more detail than anybody else and published in our lessons from the COVID War book a few years ago. Fortunately, we had already through NIH developed the MNA delivery vehicle and, and fortunately we're able to acquire the necessary genetic codes in January. It took us, when we first got to get warp speed going in a brilliant piece of improvisation, it took us about six months. Six months was okay for a disease that had one to 2% mortality rates. If the disease had had 10 to 20% mortality rates, which is not a farfetched scenario. Six months would've seemed like a really long time actually after this we developed a plan in which how you could, how can you get six months down to 60 days. There are actually ways to do that, but we've done nothing to do any of that. So, but some of this is time and planning. Some limits become a function of geography and transport. The great example in the second World War is the German U-boat warfare against tanker traffic on the east coast and in the Caribbean completely disrupted that tanker traffic. And we could not get Texas oil to northeast industry. That was turned out to be a big problem to solve that problem during the war. We built the big inch and little big inch pipelines between 42 and 44 that barely solved that problem, but at great cost to the production that we had on the East coast. And again, through these heroic efforts in mid war with years of time delay to solve. But some of this stuff is geography and transport, even if there's theoretical availability. And I conclude the problem now, when you think about the free world and what our geographies are and where the key reliances are, this is a way harder problem now than it was for America and World War ii, which at that time was geologically nearly entirely self-sufficient and had all this fantastic manufacturing based located in the country. In this light, I wanna stress

- Phil.

- Yeah,

- Yeah, yeah. We wanna give Dan time and time for the general.

- Yep, I'm almost, almost done. We have forgotten why we created the National Security Council. The National Security Council was driven in part by an entrepreneur, main principle part named Ferdinand Everstat, who was Ferdinand Everstat. He was running the oil effort for the United States on the civilian side and the war, the reason we created the National Security Council was mainly to solve the problems of economic management that everyone knew about during the war. And everyone had been shocked about during the war. I think this has been entirely forgotten, almost all the histories of the NSC actually neglect this point almost completely except for Douglas Stewart. The success story they had in mind was not the American success. They incredibly admired the British and they sought to create a system that mimicked the British War cabinet system for economic management. And its elaborate committee systems and here is why they admired it. Is this chart a look at these comparisons? This is, people forget how much Britain punched above its weight during the Second World War. It's, it's maximizing its potential. I want you especially notice the column for aircraft with heavy bombers and brackets, aircraft and heavy bombers are fantastically capital intensive to produce. Another example of this can be the surface ships and submarines, but you, you get the point is that Britain was producing the German empire in 1942 and it was doing this in part because the quality of its management of the economy, the gr the most important industrial policy success story of the United States of America in the entire Cold War is represented by this one slide. We had no ballistic missile program worth the name in 1954. The Soviets had started seven years earlier and were way ahead of us, which is why they have ballistic missiles in 57 and we don't, and their ballistic missiles in 58 are still better than ours. And by 61 we had completely trumped them because of a story that begins really in 54 and is done and successful by 61. And it's an amazing story. And credit to American leadership involving names of people you've never heard of, like Trevor Gardner and Bernard Reever. This is the last slide. The last slide basically says, what is it that you want the management to do in, in this kind of industrial policy? You wanted to do these six things and so the quality of your ability to do industrial policy and national security terms on your government's ability to do these six things. And that is in a way why we originally created the National Security Council and which we have forgotten. So I'll stop there.

- Thanks

- Phil. Sorry for going on so long.

- All right, so our fourth speaker is Dan Wang, who will be joining us virtually. And I don't, I you don't have slides right Dan?

- Correct. No slides. Yeah, I have three points that I will make. I'll, I pretty briefly, mostly related to China. The first point is that I'll, I'll, I'll echo Josh a little bit here. That the adversary has a lot of choke points over the United States. So it was about a year ago that President Trump announced the tariffs on, on China, which reached up to pretty high levels. It was about 150%. And then China Xi Jinping mostly managed to get Donald Trump to, to back down pretty substantially to have tariff levels that are much, much lower. And there's a variety of explanations here. And a big part of that was because China denied the United States rare earth minerals, especially the magnets. And there were press reports that a Ford production plant had to shut down because the US couldn't bake the, couldn't get the magnets anymore. And one of the things that still strikes me is that Beijing exercised relative restraint in terms of its exercise of these choke points. Because we can take a look at segment by segment. There are plenty of other technologies, which Beijing has a pretty strong choke points. And so there's a lot of pharma dependences on China, a lot of antibiotics, a lot of fermented antibiotics are made in China. A lot of ibuprofen is made in China. I think something like 40% of cardiovascular drugs are made in China. And so imagine if, you know, we have some more serious crisis and the American elderly cannot get cardiovascular drugs, you know, that feels like a pretty big leverage point that the Chinese can have over the United States. Now in this case, Beijing exercised these restraints to, you know, make tariffs go down. So perhaps that's on net positive, but probably this is not, I think a, a good thing that any other country can have pretty substantial restraints on American political power to do what it likes to do. Okay, so that's one broad point. Another broad point that I want to make is that it is absolutely the case that China remains relatively poor. It's per capita GDP, it's about one sixth American levels. China's economy is performing significantly below potential. And so over the last three or four years, essentially since the end of COVID China has not recovered its economy very well. There's all sorts of problems with property sector that is still on the slow burn down used unemployment. It's very, very high. There's all sorts of consumer confidence issues and one can actually kind of trace a pretty straight line between China's economic under performance right now with its focus on industrial policy. But you know, the focus on industrial policy has, I think, achieved quite significant capabilities that directly threaten Amer a lot of American interests. And so if we take a look at a lot of Chinese capabilities in let's say electric vehicle production, a lot of their batteries, a lot of the, you know, pharmaceuticals, there's a growing sense that China will also do very well in biotechnology. And so, you know, if if true were, and you were in the room, I think he would us all that, there's a lot of excitement about China's development of pharmaceuticals. And so this is where I think that, you know, yes, the China does not have a thriving economy. I think that in fact the industrial policy, we may even be able to again, trace a pretty straight line between China's, you know, economic weakness. It's, it's broad, it's it's inability to create broader human flourishing with this very intense focus on technology production. But they have the technology production and they have the, you know, deepening dependencies that America has on China between the supply chains with, you know, defense goods as well as with pharma as well as everything else. And I think this is kind of the, you know, one of these defining characteristics of the communist party as I understand it, is that they really, really wanna be a technological power. Part of the way that China has understood its history is that, you know, the foreigners had technology, the, the, the Europeans had technology, the Japanese had technology, China did not. Therefore it lost various wars and had various humiliations. Now it really wants to have the technology and it by hook or by crook, it is really going to try to pursue a lot of these different technological capabilities. And you know, even if they waste a lot of money, I think what they've sort of decided is that waste is a political problem. They are going to be able to tolerate quite a lot of waste. They're also going to jail. A lot of executives as they have in the semiconductor industry that have been too wasteful and they're gonna try to tamp down, you know, all of the fraud issues. But what they really want is semiconductor sovereignty so that the Americans cannot threaten them. And so, and I think that that is the balance of that, that is where they have ended up. And you know, I wondering, you know, what is the American counter strategy here, given that that is what the, what seems like the principle adversary has primarily decided, when we think about something like project warp speed, I think that is a clear success of, you know, America's ability to produce mRNA vaccines. But at the time, you know, the US has also not produced enough of cotton swabs, cotton masks, and that's where a lot of these basics were. It did not do terribly well. It seems like the United States has a munitions problem. I think John Deutsche on this call earlier this week, we, we had a little bit of a back and forth whether the United States really does have a munitions problem. Now, according to a lot of press reports, a lot of the, you know, between Ukraine and Iran, the US is kind of running low on a lot of critical munitions. And again, munitions are not a terribly sophisticated technology. And I'm wondering why the US can't do some of these basic things if US has poor and enabling infrastructure between power infrastructure, you know, all sorts of ports, railroads, you know, why aren't we doing a little bit more industrial policy to try to fix these sort of things. They don't seem like they have to be fixed stocks. Maybe they ought to do quite a lot better. Okay. I think the, the final point that I want to make is to echo a little bit more of what Philip Ko has said, that the challenge is not just, you know, any sort of static comparison between China as well as the us. We have to adopt a future oriented view that, you know, present technolo, you know, future technological capabilities derive to some extent on our current technological foundations. And if the United States keeps losing a lot of these abilities to do not just simple things like masks, cotton swabs, you know, artillery shells, but also somewhat more complex things like aviation products. You know, the woes of Boeing that Phil pointed out earlier, the woes of intel, which are still kind of insight right now. You know, if the United States has a much worse performing aviation industry but has some much worse performing semiconductor industry is going to be able to maintain all sorts of leadership in the future. Is the future really just mostly significantly software if, is it mostly, you know, all of these more, you know, software source of technologies or does it maintain an active and thriving industrial base that is able to pursue both high end goods as well as lower end goods? It seems like the Chinese have ended up on the fact that they want to do much more rather than less, given that is how they've chosen to have a lot of their emphasis. You know, shouldn't the American strategy accommodate that? And the debate ought to be, I think, you know, whether the United States tries to match quite a lot more of these loss capabilities or not. Because if I had to summarize the Chinese strategy in one sentence, I think that Chinese strategy is actually a relatively clear, it was encapsulated by Xi Jinping in the year 2020 in this slogan that he called dual circulation. Dual circulation I think can be understood very simply as an effort to make China more independent from the rest of the world. And at the same time make the rest of the world more dependent on China. We've seen a little bit more of this. We've seen that they have been able to excel and keep doing better on all sorts of new technologies, electric vehicles, the batteries, new drones, all sorts of things. And I'm wondering why the United States cannot at least match some of these capabilities and develop, patch up a lot of these weaknesses. With that, I will conclude.

- Thank you Dan. So, but oh, do you want, I see you've made notes.

- Well, I, I,

- We, we we'll take, we'll take organizer's prerogative you first, and then I'll go and,

- And then Yeah. And maybe we should just collect a lot

- And then we'll collect a lot. Yeah. More.

- Yeah, I, I'll make some just brief comments. Thank you all. These were spectacular presentations and I, I really appreciate as an organizer, I appreciate the work you put into it. It's funny that we didn't really start with the definition of industrial policy. Should I think of, I gave one you didn't want you paying attention. Can you talk any the microphone? That's my first slide. Yeah, let, let me try to redefine it 'cause I'm gonna disagree with you. Okay. We usually think of it as a system of, of protections. Subsidies and government run industries with economic points. So the Manhattan Project and the Man on the Moon were shows the government's ability to do a, something an an big engineering challenge at monstrous cost. Just whatever cost you want, you know, and SpaceX now is showing you, well, we can do this one 10th of cost in the private sector. And it wasn't, it was a direct government program. Warp speed was a great example, but warp speed was largely getting rid of regulations that were in the way of doing things At Josh's example, the rarest 27 years to get a permit. You know, well, the problem there isn't that we need more protections and more subsidies in the government running things. So if you wanna get deregulation as industrial policy, I am a hundred percent all in our shipbuilding industry. Why did we lose shipbuilding? It's because we hobbled it with the Jones Act, with unions and so forth. It's not that we had lost the technological ability to make shit. And again, if deregulation is industrial policy, count me in. And the last comment I'd make here to, to, you know, there is this defense, you know, defense matters, Josh, even I'll agree with that. But we ought to also put it on the defense budget. So it's easy to say we want protections or whatever, but because the go military doesn't have to say how much it costs. So, you know, you can have a hundred billion dollar chip factory. Sure. Or you can have 10 aircraft carriers. Sure, tell me which one you really want, would be helpful.

- So great presentations, they, they made me think about so many things. So, so just following up on a couple of the points, you know, this worry about industrial policy, what I've seen from my research, you know, I look back from 1889 to the present during war periods. Something that's very effective and doesn't get into a lot of these problems is if the government takes the approach, if you build it, we will buy it. So that's what was very important for operation warp speed. If you develop those vaccinations, we will buy them. And there are countless examples from World War ii, such as the B 24 liberator. They said, if you do, and, and this government facilitated by getting Henry Ford and his team to come out to San Diego who, who had developed this and figure out how to do it quickly and then build a mile long factory at Willow Run in Michigan. So, so that's one approach to do it. Think of the things that we need. One example is actually rubber. So this, this isn't just a World War II thing, there's a recent report by the Homeland Security Engineering and or other that says we still have a rubber issue in World War II is because the Japanese controlled the source of natural rubber and now it's China controlling the source of natural rubber. It takes seven to nine years to throw a rubber tree. And even with reducing regulations now, that's not gonna help with those delays. So what you can think about, what they're calling for is stockpiling natural rubber. All right? So you, if the government says, if you get it for us, we will buy it. Something else to think about is, you know, you're a finance person thinking about co variances and this marginal utilities in the bad state. Just to give you an idea. 'cause things I'm seeing now are eerily reminiscent of the past. So as of April 20th, the maximum enlistment age for the military is going up from 35 to 42. Those of you who think that you, you know, you have sons and daughters who are beyond the age, not necessarily, and you know, conscription happens. All right? So remember that when you're thinking about whether you're worried about supply chains and whether those bottlenecks could actually result in a lot more American casualties.

- Just three incredibly quick comments. First in my analysis of this issue, the fundamental problem with defense supply chains is a lack of credible forward guidance about government's future demand signal. And for drones, long range precision munitions. Many of these things we have to make in quantity. It's about persuading business, who, which serves one customer that there will be demand for X quantity in 20 30, 20 35. Second, any economic weaponization happens in the context of an entire world market, not just us versus them. So the elasticities or adaptability, the role of the neutral community in helping us adapt has to be part of any calculation. And the critical minerals case is, I think an example. And third, fundamentally we're talking about a trade with allies. We want them to give us back some of that process knowledge to train our workforce and to essentially do that via investment. What do they get in return? They get market access. That's the only way the trade works. And I think we should be more upfront about how the trade should be designed. That's the basis on which we can make this allied coordination work.

- Two very short comments. I ran, sold a lot of oil to China. China sent them air defense systems, which were totally worthless, absolutely failed, couldn't shoot down a single American plane. Total complete utter failure. There's never been a failure like this. I think we should stop worrying a little bit about China's defense capabilities. Second thing is, I'm sitting here waiting patiently. I never heard two words mentioned. Would you care to guess what they were? John, you should know competition, artificial intelligence. How can you people talk about this stuff without talking about the US lead in artificial intelligence, the resources going into it, it's capacity to make this country way stronger as long as it sustains that lead. And I would've hoped to hear one sentence somewhere in an hour of presentations in the middle of Silicon Valley. So I guess I have to live chatty by myself. And that's it. Sorry, I gave up.

- I would like to make a few obvious remarks that should be true, but self evidently, but put together maybe be somewhat uncomfortable. And so one is that if we talk about knowhow and the, the, the engineering capabilities of the world there in China, and if, if, if we think about how China rose, they have a very good shirad of quoting foreigners, of quoting American businessmen, investor corporations, managers of all levels to, to go on to and and teach them how to, how to do things. I mean the, the chairman of the Silicon Valley bank can welcome this, wrote a new memoir about his time in China where the, the counterparts in China was so diligently learning from him, him that by the time he started conducting business, they were able to copy him exactly and, and basically put him out of business. And the same story kept happening for 30 years or so through joint ventures, through tech transfers, through investments, through all kinds of things. And so if we're thinking about how do we revitalize America and America now beyond the back foot, trying to learn from the engineering capacities of China and the rest of the world. Is America ready to learn? Is America ready to import Chinese entrepreneurs the way the Chinese imported American entrepreneurs, American capital and how, what is America's rad of quoing Chinese investors, entrepreneurs, companies, and and engineers? And as much as it is painful to learn from your one students, are the Americans willing to do that? And I think there are different drills, different formula to do it. You could have 50 50 pro ventures at a, at a at, at a locality with expedited permits with the head of the factory being a Chinese. But then the rest of all the engineers trained locally at a university. And there are all kinds of possibilities if we, if we try to go there. Obviously there are political risks just as China worries about foreign infiltration through foreign capital, through foreign culture ideas, America may worry about the same. And so then there are a set of guardrails that could be invented, invented to guard liberal values while importing Chinese technologies. And so I'm just highlighting that realm of possibility is, which seems to be missing from today.

- Yes. By the way, sending them the head of the Silicon Valley bank sounds like a Trojan horse idea that may explain their financial trouble process,

- But it's not a Silicon Valley. No, I know. That was just a joke. Our boys is Elon Musk is everybody

- We get Russ, and then we get,

- So I get, thank you very much. It was, it was fantastic. I wanted to point out one issue, sort of two issues, maybe sort of subtle. One is, I, I don't think it's accurate to talk about the US withering in the, in the past along a dimension. If we look at real per capita GDP growth over the last 50 years, it's about the same as real per capita GD prerow over the 50 years before that, which is about the same as real per capita GDP growth in the 50 years before that. It's not to say what's gonna happen going forward, it's just to talk about withering us. I remember this discussion with Japan as well, is that it was going to take over the world, et cetera. The other thing is I'm not, I, I think we have to be very careful about distinguishing between national security and competitiveness. I don't think that these things can be separated, that one of the components of national security is being at the frontiers tech technologically. And we don't know how all of those technologies that arise in the infinite myriad of industries are going to be put together in terms of national defense. And so it is the competitiveness of the US economy that creates, or at least is a huge contributor to national defense. And therefore things like trade, which allows us to be competitive and specialize things means that immigration, which means that you get entrepreneurs and innovation, all of these things are part of what is going to be what has been a very competitive environment that has produced technological innovations that has indeed created the national defense that we have. So these things can't be separated AI largely private, huge investments by a financial system that can mobilize immense amount of capital for those types of things. So yes, I think, I think Josh's division, there are some things getting the resources necessary. Absolutely. I think your, your de your that division was extremely helpful to me. But these, these two things can't be you, you can't separate the, you can't separate the economy from national security and, and it's economy based ultimately on competition and markets that's been successful.

- I was going back to the, the basic question of what is industrial policy? I thought I knew it coming in here, but I'm not sure I do anymore. And, and sort of thinking like, Josh, I thought your presentation was great and the division was great, but you came down to regulatory barriers and comparing Canada and Australia to us. But when you show the chart of rare earth production, Canada wasn't on that chart, or Australia wasn't on that chart of, they can certainly de get people to through the regulatory process quickly, but it's not at the same time producing outputs that you want. So somehow you have to think a little bit about that. I

- Thought we got a question online. John Deutch, please.

- I'm very sorry that I have not been at your session. I left a couple of weeks ago from Hoover, and I think it's been an absolutely outstanding session with many, many important comments. I'm gonna say one thing. First of all, I feel like I should apologize since I actually was under Secretary of Defense for acquisition and ran the procurement for the DOD, and I know it's from grounds, the ground level. I know the, all of the problems. I'm gonna make two comments. I could make more, but I'm just gonna make two. The first is I have not heard the word of a very big important actor, and that is Congress. A lot of the, the things that happen happen because of congress and Congress, the executive branch having to live by the rules and the pre prejudices you have from Congress. So more should be said about Congress. And the second thing has to do with testing. If you're gonna use defense systems in possibly conflict situations, you have to be sure that those defense systems will, that will perform properly. That means the development and operational testing that goes into a defense product that is going to be used in the field is much different and more stressful than you have in the commercial sector. So you testing is a great difference, but the main difficulty is that the defense system is relying much too much on military determined products and therefore on a defense industry of five firms that has something like 60 or 70% of all the procurement of the federal government, not just of DOD. So there has to be a, a move to use commercial firms, young, innovative commercial firms much more dramatically on a quantitative basis. And I say to you, one optimistic thing. I believe there is a big innovation revolution going on in this country, in our country. Somebody just mentioned why haven't we spoken about ai? But there are lots of different things. There's quantum computing, there's all kinds of communication things. So the real issue is to open up the system to greater co greater participation of commercial commercial firms. That's what I believe will be a useful, and I'm optimistic about the fact that that's going to happen. Thank you.

- Great. Well, how about you guys each take one minute. Sorry. We got,

- I'll touch on a few things that came up multiple times. Government procurement guarantees. Valerie mentioned this, Ike mentioned it in a different way. Maybe John mentioned it. It was, it was definitely a feature of operation warp speed. And it is sent to what John Deutsch was just saying. One of the, one of the problems that is emerged from Congress is their refusal to offer essentially long-term commitment purchases. So all of these things are tied. I agree with all of that very much. I'm sympathetic to Ross's comments about the relative strengths and the sources of them for the United States. Al I didn't mention ai, but I did talk about the capacity of the United States to scale up innovations quickly and that that AI was implied there. AI was implied there, so sorry, we only had 12 minutes. And to, to, to John's point about more, more use of commercial firms opening up the system to competition. I agree that's really moving away from a more centrally planned industrial policy to one that as much as possible tries to rely on market forces, which are the great strength of the US economy and to rely on those to propel our national security. I'll stop there.

- I guess I'll just quickly say, you know, of, of our, it's true that of our three recommendations at the end, one of them was not industrial policy under I think any definition we were thinking about. And that's, that's deregulation and streamlining of reviews. That's, that's not industrial policy. That's just, that's just getting out of our, getting out of our own way. The the other two were though, I mean supply chain audits where you, where you say, okay, you know, department of Defense procuring defense assets, we're gonna go deep into the supply chain. We're gonna say you can't buy from the cheapest supplier if it happens to be China. Your inputs you gotta buy from, from, from someone else. So that's, that's industrial policy and purchase commitments, which Valerie talked about very well as well. That's also in, in industrial policy. So, you know, I I I think I, John, I agree with your definition of industrial policy. I think of our three recommendations, you know, 1, 1, 1 of them was how we can avoid having to do more industrial policy. And the other two were actual types of industrial policy. And I, I think as for like Australian, Canada, I mean CH China wiped them out too, you know, by subsidizing so much. So I think, you know, by subsidizing so much of the production of the, of the refining of, of rare earth. So I think what that demonstrates is that deregulation probably won't be enough. It's gonna have to be paired with some of these other things.

- One minute didn't mention ai. When I spent all that time on semiconductors and Nvidia, I wasn't talking in Arizona. I wasn't talking about transistor radius. Yeah. The, the big point I wanna stress is, is that both Ike and Valerie really called out the significance of advanced mark commitments and predictable demand in order to make a market. They are absolutely right. Therefore, that is what puts the premium on what I call the definitions of success in my very last slide, is you have to then have, why we created the National Security Council is you have to make a variety of decisions in order to make those advanced market commitments. If we had made advanced market commitments to Con Bear, which was the prime on ballistic missiles in 1954, the outcome would've been a disaster. We ended up having to make large scale commitments to a company named Boeing, which made the minute man on a quite different set of design principles that required us to get engineering and scientific input to decide which kinds of designs really showed the most promise. And that is what helps to define an intelligence system of industrial policy in the more narrow definition of national security that I defended.

- Dan, you lack closing words.

- I am skeptical of the ability of AI to fix too much. At least I am skeptical of invoking AI to fix. I, I wanna be very cautious about what AI can do for, you know, how will AI fix the rare earth magnet shortage? How will AI fix the manufacturing base? I think a lot of what that requires is actual real capability and investment in traditional infrastructure, enabling infrastructure in the source of power, the source of workforce education, really to try to do this very well. Because, you know, again, thinking about it from the China context, you know, if the United States, if automation and AI can boost the American manufacturing base, well, it can really boost the Chinese Amer the Chinese manufacturing base because they have just a lot more data, they have a lot more workers that they can put GoPros on to really try to, you know, figure out all of this process knowledge. And they have better robots as well. They're much more roboticized than the American manufacturing sector. And so the therefore, you know, it's like ai. We can't let AI be an excuse for, you know, too much thinking about, you know, how this could be the great salvation. Thank you.

- Thank you everybody. Thank.