Innovation Files: Where Tech Meets Public Policy

Getting Export Controls Right, With Kevin Wolf

Information Technology and Innovation Foundation (ITIF) — The Leading Think Tank for Science and Tech Policy Episode 82

Export controls exist at the perilous intersection of economic policy and national security. What could go wrong? Rob and Jackie sat down with Kevin Wolf, a partner at Akin Gump, to discuss how to strike the right balance between keeping certain advanced technologies away from adversarial militaries without locking U.S. firms out of global markets. 

Related

Auto-Transcript

Rob Atkinson: Welcome to Innovation Files. I'm Rob Atkinson, founder and president of the Information Technology and Innovation Foundation.

Jackie Whisman: And I'm Jackie Whisman, head development at ITIF, which I'm proud to say is the world's top ranked think tank for science and technology policy.

Rob Atkinson: This podcast is about the kinds of issues we cover at ITIF, from the broad economics of innovation to specific policy and regulatory questions about new technologies. And today we're going to talk about something that when I tell you, you might, your eyes might glaze over and you might yawn, but trust me, pretty interesting and pretty important, and that is export controls.

Jackie Whisman: Our guest today is Kevin Wolf, who's a partner at Akin Gump. He served as Assistant Secretary of Commerce for export administration and has more than 25 years experience providing advice and counseling regarding the laws, regulations, policies, practices and politics pertaining to export controls, sanctions, national security reviews, and other international trade issues. Welcome Kevin.

Kevin Wolf: Nice to be here. Happy to help.

Jackie Whisman: I wonder if you could kick us off with a few seconds on maybe like an export controls for dummies so that we can set the stage for the rest of the conversation.

Kevin Wolf: Sure. My entire 30 year career actually boils down to one sentence, which answers your question. So export controls are the set of regulations that govern three verbs, the export, re-export and transfer, four types of things, commodities, which are physical items, software technology, which is information to do something, and sometimes services which are acts, assistance of people to specific end uses, specific end users or specific destinations in order to accomplish national security or foreign policy objectives. So if you screen any country's export control rules, going back to Roman times, it's all about those three things governing the export, re-export transfer of commodities, software technology and services to specific destinations, end uses and end users to accomplish national security and foreign policy objectives. And so to know what those items are and to know what should be controlled and know what should be granted or denied or requiring government authorization requires a understanding and acceptance of a definition of national security that is not an objective term.

And foreign policy is certainly not an objective term. And during the Cold War, national security in this context for export controls was largely defined at containing the Soviet Union and the East block. And these were broader countrywide strategic controls to accomplish various objectives and constrain the ability of the Soviet Union and the Eastern Bloc for a variety of reasons. At the end of the Cold War, what evolved was a system that was focused on non-proliferation objectives largely. And so there were starting in 1996, the completion of four multilateral arrangements, international political arrangements of the US and partner countries that would identify the lists of commodities, software and technology that had some identifiable relationship to the development production or use of either weapons, advanced destruction, missiles, chemical biological weapons, nuclear items or conventional weapons, and then later anti-terrorism issues. And then the US and the allies would agree to regulate the movement of those commodities, software and technology, whether bespoke for military or a commercial dual use item that had some relevance to the production development or use of one of these items.

And the national security definition was that to achieve, to stem the flow to military end users or terrorist applications or sanctioned destinations, but otherwise to generally approve commercial items for commercial end uses. And that non-proliferation focus system is really largely what governed the thinking and frankly is the structure of the laws of all of the US allies that is regulate the lists of things that are on in those four regimes. Anyway, there's more to it, but that's the essence I'm leading up to. What this rule is, is a change we'll get into, but that's the story for the last 30 years.

Rob Atkinson: Cool. I think that was probably more of an export control for pretty intelligent people as opposed to dummies, but I think I got it. So no, that was good. One thing that you mentioned all the way back to the Roman times, the case I always love to talk about is in the British, in the textile era, revolutionary industrial revolution, they made it a punishable capital offense to take out plans for textile equipment and others. And actually the first American factory really was from a guy named Samuel Slater who worked for Arc Wright in the UK and he came here, didn't have any written plans, but he had 'em all up in his head and he got funding by Moses Brown who made his money in the slave trade and endowed Brown University to build the first textile mill in America. And that's how America got into the textile game. So there's some export controls, it's really hard. You'd have to sort of put that guy in chains and not let him get on a ship. I guess that would've been the way.

Jackie Whisman: Is that your Canadian accent? Do you mean textiles?

Rob Atkinson: Beauty, eh?

Jackie Whisman: What is he? Is this some new technology? I don't know about the textiles.

Rob Atkinson: Textiles.

Jackie Whisman: Sorry. Okay.

Rob Atkinson: So, we've been doing this for a long, countries have been doing this for a long time.

Kevin Wolf: Yeah, there's lots of good stories like that about how the regulation of movement of things, which is what this is about, or information whether in your head or in documents or has been regulated for different reasons over the year. And what's really good comparison to that very real story and the case over the last 30 years is that economic considerations were deliberately excluded. And that's relevant to what's going on now. Economic considerations were deliberately excluded. Trade considerations, trade policy issues were deliberately excluded from the scope and purpose and use of export controls. And when I was in government, for example, our mantra was we do not pick winners and losers. We look at items based upon their inherent characteristics or how they might be used or misused. And if they had some relationship or potential diversion for military chem, bio nuclear missile, et cetera, or terrorism applications or sanctioned destinations or human rights violations, then we would deny, deny the license.

But we didn't take into account or try to help industry by approving or denying licenses for the sake of economics alone. And absolutely we considered the economic impact, no doubt, but that was in connection with making sure that we regulated no more or no less than what warranted control. And so for example, we wouldn't approve a license that was for billions of dollars because it would really help that US company, but it would be to a recipient where there would be harm to the war fighter or to broader national security interest. And on the flip side, we wouldn't necessarily just deny a license because it was insignificant economically. So there's economics and national security traditionally were not mixed. That is evolving, which we'll get to I'm sure with later questions.

Jackie Whisman: Well, yeah, you probably know where we're going with this, but I want to ask what I was going to ask. What the role of export controls vis-a-vis China is?

Kevin Wolf: Right. So the construct of focusing on nonproliferation and conventional military and anti-terrorism objectives and human rights objectives that I described in a small part, the human rights were not a major part of the regulations was pretty much the structure for really the 1990s until near the end of when I was in the Obama administration. And what started to change was about, I know others will say they go earlier, but from my perspective, really about 2015, which was the maid in China 2025 program and the dramatic increase in technology acquisition efforts by the Chinese government in strategic and foundational technologies for broader objectives of the Chinese government, the extraordinary increase in evidence of the civil military fusion policy of requiring commercial companies and commercial technologies for use with modernizing the Chinese military.

And then there was a report that we presented just near the last few months of our time in government about the $150 billion or so of investment into the Chinese semiconductor industry. And what evolved in 2017 and 18 was a rather thoughtful, completely bipartisan public debate as part of both plans to change the CFIUS rules, the foreign direct investment rules into the us. And then the bookend to that was how to give congressional authority for an export control system, which had been lacking for decades. And what evolved out of that in the summer of 2018 was an agreement and a statutory requirement that the US needed to identify emerging and foundational technologies that were not within the scope of this more classical export control system that I described, but were nonetheless essential to national security. And so these words were used and the mandate was given by Congress to do this, but there was no definition of what national security meant.

There were almost as many definitions as there were people, there were the hawks, the dogs and the owls out there, and each of them had their own definitions. And the Trump administration asked questions and gave lists, but never really articulated a coherent definition of what either emerging or foundational meant or what national security meant outside the classical construct. And then the Biden team was largely quiet on it as well until first the Russia invasion and then a speech that Jake Sullivan gave in September of this year. So if you want, I'll ask your questions and I'll move on to giving sort of what changed and where the coherent definition started coming in.

Rob Atkinson: Yeah, why don't you do that? Because I know that one of the issues with our allies now is, and we're getting a little bit ahead of ourselves, but they're saying, well, wait, are these Biden controls and the ones you're trying to get us, are they really related to military and national security? And I feel like we're kind of living in this world that we do have to clarify what are we really trying to do and what that controls. So Kevin, yeah, why don't you go ahead and just explain to how that changed?

Kevin Wolf: Yeah, no, you're exactly right. And that's why when I talk about this, I think it's really, really important to give the history of export controls really the last 30 years and the shift from the Cold War strategic mindset. And the reason for that is that the allied export control systems and their laws and regulations and the mindsets frankly, of the individuals who are in charge of these agencies is really sort of a time capsule in many ways of this post Cold War system that is based around items whether bespoke or dual use that have some identifiable relationship to the development or production of specific weapons. So what I was in government for example, and each year we would make our submissions to the four multilateral regimes for changes to be discussed with the allies in the other partner countries. We would generally start with the military item first.

We would start with the weapons system and work our way backwards. Here is a military radar for military aircraft. What are the commercial components in it that are particularly important for its functioning? What are the thresholds? What are for these microwave monolithic integrated circuits, for example, that have great five G applications but are also critical to an aisa radar? We would identify those and we would regulate them because of a clear identifiable relationship to a weapon system. There are a few exceptions obviously over the regs that we can get into including high performance computers. But that essentially was the mindset and is the current mindset. So as the us, however in really publicly in 2017, 18 started rethinking national security, we have gone far beyond in terms of a broader than classical national security threat that is China than most of our allies. Now, there are different people with different opinions of course in every country, but the main point is though that their laws and limitations are, their laws are largely limited by this mindset.

And so what really started to change was in the extraordinary allied response to Russia's invasion of Ukraine. And for the first time since the Cold War, 37 countries started adopting controls that only the US had historically regulated itself, including basic semiconductors and basic civil aircraft parts and basic consumer electronics, things that no country had ever identified before as being a regime worthy, a multilateral control worthy type of item if destined to Russia, and then much broader sanctions and export controls if for military end users in China. And the key psychological and policy turning point with that response is that export controls for the first time since the Cold War were being used in a coordinated fashion to achieve strategic objectives. And what I mean when I say strategic, what I mean by is broader than just items that have some relationship to a weapon system, for example, even if indirect. And so for example, semiconductors are being controlled and prohibited, not because they're necessarily going into or bespoke or modified for a weapon, but they're needed for the ecosystem of Russia to function the commercial ecosystem which is needed to then provide and rebuild its equipment that's being destroyed in the war. And that's a broader strategic objective.

And the allies were finding authorities, they were creating authorities for these plu lateral controls and plural, lateral is a key word here because it means a collection of countries working together using their own individual authorities. But outside this classical multilateral regime process, which by the way, Russia is a member of three of the four regimes and all of the regimes operate by consensus, meaning one country can veto a change to any aspect of the regime and the list of items which I'll get to later. And so that was a first break on the theory and philosophy and practice of export controls since the Cold War. And then the next big change from a US perspective, and what was really the first coherent articulation by a senior government official in the Biden or Trump administrations or Congress, was a speech that Jake Sullivan gave to a conference in September of this year where he laid out a strategic priority for the use of export controls differently.

And he identified three technologies that were what he referred to as force multiplier technologies. And the number one, which is relevant to what we're going to get into here in a moment, are computing related technologies, advanced semiconductors, supercomputer applications, the types of things needed to run artificial intelligence capabilities. And he said explicitly that we need to revisit the premise that was implicit in export controls for the last 30 years of maintaining just a relative advantage over competitors. Now, that wasn't a written rule, but it was basically a practice for licensing decisions for things that met those regime thresholds. And he said that a sliding scale approach is no longer acceptable and we need to maintain in these three force multiplying technologies or enabling technologies as large of lead as possible. So we basically forecast that the very existence of advanced node semiconductors, the very existence of supercomputer and advanced computing applications is a per se national security threat, even if no clear or identifiable or direct relationship to a particular weapon. And this was said in the abstract, China was never referred to, he wasn't announcing specific controls, but he set out a coherent intellectual vision for the definition of national security and export controls that was really much broader than the conventional wisdom for the previous 30 years.

Rob Atkinson: So I guess a couple things. Do we need to operationalize or admit that or formalize it more or put it in statute? Because the way we look at export controls, particularly when you have a country like China that is a systematic manipulator, cheater, mercantilist, IP theft, you name it, thief in my view, there are areas where we should have export controls, even if it doesn't directly help the military because it helps the overall broad techno system in China, which at the same time, the stronger they get, the weaker we get. But it doesn't seem like we're quite there yet. We still have to sort of fall back on this. Well, ultimately it's going to affect their military. Do you think we should do that and could we do that?

Kevin Wolf: Well, a couple of things. So the US is largely alone in having a really broad statutory authority to define national security. However it sees fits, I mean, at a minimum, at a floor, the statute, the Export Control Reform Act codifies the practice and the expectations of the last 30 years of export control policy, both for military and dual use items and adherence to the regimes and human rights objectives. But it left open some really broad phrasing for what national security meant outside of that. And there's been in congress and the media and think tanks, lots of debate as to what that means with respect to China, everywhere from complete decoupling to business as usual, which is obviously a minority, but still, nonetheless they're in some respects and only regulating items and have a clear relationship to a military item. And that's part of the problem with the public debate, which is a great public debate that's happened the last couple of years, is that it hadn't really been codified or articulated.

And the reason that that's relevant is for any regulatory system or any legislative objective, the first rule is to define the problem to be solved. And so the codification, if you will, of the Jake Sullivan speech in actual practical terms, was the rule that the commerce Department published on October 7th, which laid out a really coherent national security objective and why items involving advanced compute were directly relevant to US national security objectives. And then Secretary Raimondo after this rule gave probably the most coherent articulation of US-China policy in this space and every other space. Remember, export controls do not operate alone. They are not the solution to all the world's problems. And generally the dynamic when speaking of national security and economic security, which gets rolled in and equivocated on a lot, is a run faster and a keep away strategy. And the run faster national security strategy is industrial support. And the prime example of that, of course, is the CHIPS act to subsidize US industry and to otherwise support them to be more competitive economically. And the export controls are focused on the keep away regulating commodity software and technology for that which would support issues.

Jackie Whisman: The other big news in this space is the Biden order on China and semiconductors and semi equipment. How effective do you think this will be?

Kevin Wolf: Very effective for a variety of reasons. And the summary of the rule is that it is X four controls that are designed to limit the development of production in China, specifically in China of four things, advanced node semiconductors, which are certain types of memory and logic chips, second semiconductor production equipment of any kind. Third advanced computing items largely through controls on graphic processing units that are important for artificial intelligence applications. And then fourth supercomputers. And the reason I believe that the rule is really quite effective in this regard in the short term, and I'll get to what I mean by the short term here in a minute, is that the US government is imposing these controls in a traditional way, which is regulating the movement of things from the United States, but in an expanded way in applying the rules extra territorially that is for foreign made items that are made with US equipment or US software, and also to the activities of US persons, whether employed by foreign companies or US companies and their actors overseas.

And through those three techniques of exports from the US, foreign made items being affected and activities by US persons, it pretty much is stopping the inputs for advanced nodes, semiconductors production in China, semiconductor production equipment, advanced computing and supercomputers. And US obviously has great dominance in a lot of the key choke point technologies in this space. And so it's quite effective. All of the news media lately has been however, whether the US will or will not get support from its key allies who also have capabilities in these technologies, largely Japan and the Netherlands. And the reason I say very effective in the short and the medium term is that the US has dominance in these sectors, but there are lots of clever people outside the United States, including in Japan and the Netherlands, and none of these things are prohibited by their companies or their citizens to perform or to provide for the same end uses in China.

And the history of export controls has always shown that unless countries work together in controls or plural, lateral or multilateral over time at different levels, depending upon the technology, they become less or ineffective. That means the items still make it to the end user and end user of concern, and they become counterproductive, meaning that the work and the money and the flow goes to the competitors for the US companies, not the US companies themselves. And so that's really the story to watch. And this is not new. The administration has recognized it itself and has made, so this is not a controversial statement that I'm making, that the effectiveness will be enhanced over time if other countries start imposing these controls themselves. And this all goes back to where we started, which is it depends on if the US can convince allies that these types of controls, which are far more indirect than classical controls to the national security, the weapons related application are part of a common national security interest. And that remains to be seen. I'm optimistic, and this is too big to fail, I believe, for the administration. I think the majority view is that the allies won't come on board, but that's the topic of the day.

Rob Atkinson: So Kevin, just to wrap up because we're running short on time, two things. One I just read yesterday or the day before, I thought the Japanese and the Netherlands were in consideration of joining,

Kevin Wolf: Right? There was a Bloomberg article, there were several articles the last couple of weeks saying the opposite by multiple media outlets. And then there was a Bloomberg story that came out and said, the Dutch and the Japanese are considering something. Now I don't know what the something is. It could be too limited. It could just be one type of tool, not all the types of tools. It might and probably won't include the prohibitions on activities of Japanese or Dutch citizens that US citizens are prohibited in. So it'll still be an unlevel playing field even, but at least it would be progress at expanding the aperture for what national security means and leveling the playing field somewhat for US companies. And at the same time increasing the effectiveness of the controls because China still is very much dependent for now on non-Chinese inputs into advanced node semiconductor production and supercomputer and advanced computing applications. And remember, export controls are never perfect. They're only one tool as part of a larger suite of policy tools. And they're generally at this level more about degrading and delaying because eventually other countries develop the same capability. So when I mean effective, that's what I mean, it's going to delay a lot the innovation and the development of production in each of these key sectors and the length of the delay and the degrade will expand the more countries come on board with similar controls.

Rob Atkinson: Great. Just to wrap up maybe with just maybe a one minute answer, because we are running, we want to keep the time you've spoken articulately. In fact, we're going to be on a panel together for a second time about the need to come up with a new, you mentioned there's four of these export control regimes. You've really talked articulate about coming up with a new export control regime that's a little bit more multilateral, multinational with our allies. A minute on what you envision that to be. Yeah,

Kevin Wolf: Absolutely. So again, this is the reason for telling the history with these stories because everybody is new to export controls, and it's something that only nerds like me have been paying close attention to for several decades. And the four regimes, again, are based upon consensus, meaning that all members must agree, including Russia, which is a member of all three ahead of being invaded, one of the other members, Ukraine, which is unprecedented in the regime history. And then second, the scope and the mandate of the four regimes is focused on these conventional military and non-proliferation WMD applications. And the thing I've been advocating since right after the invasion is that we need a new regime not to replace the four regimes. They still have their purpose and the laws of the allies are still dependent upon them, and there still can be arms control discussions of merit in the existing system, but to address non-traditional national security issues, non-classical rather, and largely China and Russia, specific issues that are broader than just items for weapons as well as human rights objectives, which is not within the mandate of any of the regimes.

And also to deal with the fact that progress in the other regimes is and is going to be blocked by Russia as a member not wanting to agree to controls that are imposed against it and its friends. And so I've been advocating strongly that first informally, which is largely happening already through the response to the Russia invasion, but to codify the goodwill and the hard work and the connective tissue that's been developed by that effort, that a new regime of a smaller group of close allies to deal with these non-classical issues in a smaller group of technologies in addition to human rights issues, in addition to strategic economic dominance objectives, in addition to potentially even supply chain issues, that a smaller group of 10 or fewer countries come together and create a new regime with a new mandate and a new articulated coherent scope of export controls beyond the classical.

And the reason I say an articulated scope is otherwise all you have is chaos because the first rule of regulation, again, is you have to define the problem to be solved, to know what the regulation is, to have plural, lateral, and allied support. You need to have an articulated description of what you're trying to accomplish together, so you're all on the same page. And I think that discussion is happening informally, but it needs to be codified in ways that can result in actual controls that have actual effect going forward to achieve those non-classical objectives without Russia.

Rob Atkinson: That's great. Yeah, and I would argue the second rule is due as little damage to yourself as possible when you are imposing these controls and your regime is critical to allowing that to happen.

Kevin Wolf: Absolutely.

Rob Atkinson:

Hey, we could go on. This is fascinating. At least to me it's fascinating and it's certainly critical to the nation and to the world. So Kevin, this was really great. Thanks so much.

Kevin Wolf: Happy to help.

Jackie Whisman: And that’s it for this week If you liked it, please be sure to rate us and subscribe. Feel free to email show ideas or questions to podcast@itif.org. You can find the show notes and sign up for our weekly email newsletter on our website itif.org. And follow us on Twitter, Facebook, and LinkedIn @ITIFdc.

Rob Atkinson: And we have more episodes of great guests lined up. We hope you'll continue to tune in.


People on this episode