In this episode of Escape Forward, Cristina Caffarra speaks with Shahin Vallée about Europe’s economic and geopolitical choices at a moment of growing pressure. They discuss productivity, AI, industrial policy, China, the transatlantic relationship and global trade imbalances — and ask whether Europe can turn shifting ideas into decisive action before events move faster than policy.
ESCAPE FORWARD Ep. 12, 15th September 2026
EUROPE’S SLOWLY SHIFTING, BUT TOO LATE FOR THE FIGHT? Industrial policy, trade, AI, China with Shahin Vallée German Council on Foreign Relations, Senior Fellow, and Catalyse Europe
Cristina Caffarra (CC)
Hello everyone. I’m Cristina Caffarra, and this is Escape Forward, the space where I have the privilege of having conversations with people I like, people I learn a lot from, and go beyond my original haunt of antitrust and competition to join the dots with areas of policy that are salient for Europe today. Where are we now? Third/fourth quarter of 2026. We had a big talk about competition reform needed to support competitiveness from the president of the Commission, but antitrust remains pretty much in suspended animation: I’m not really inspired by the so-called reforms, we are paralyzed by terror of U.S. reprisals when it comes to digital, so enforcement is pretty much going nowhere. Industrial policy should be a major driver of where we’re going in terms of policy initiative, and yet there is little action that we can see and much radio silence from national governments. Trade policy is another area which is red hot, but we are in the dark at this point about what anyone is thinking of doing. And then there is the whole space of digital and AI, there’s a lot of action globally, but in Europe we don’t have very much that is coherent and big underway and can put Europe back on the map. Everyone seems to be keeping their head down below the parapet, difficult to read the tea leaves as to where the Commission and the national governments are actually going and what they’re doing.-
I’m delighted to be able to welcome for this discussion Shahin Vallée, who is well known to many, Senior Fellow at the German Council on Foreign Relations and Catalyse Europe, among many other things, writes a lot of interesting commentary and papers. Every week, I religiously read your Notes on Europe Substack, which is one of the sharpest commentaries out there on the salient facts in Europe in the previous week. Shahin, you’re very welcome.
Shahin Vallée (SH)
Thank you.
CC 03:25
I want to use as a jumping off points the two big economic debates of the summer. It is hard to think that nerdy economic debates can capture anyone’s enthusiasm and attention, but there were lively discussions not just in academic papers, but also on social media, on Substacks and X of these themes.
The first is what has been described as “the Great PPP War of 2026”, the “purchasing power parity war”, and we’ll explain briefly what that is. The second one is to do with trade imbalances and the way they are shaping the discussion around the Second China Shock and where Europe is and what we should be doing. You’ve been involved, of course, as commentator and debater and opinion maker in these discussion. I find them useful to go in two directions: one is industrial policy and one is trade.
Let’s start with the first, which is the debate around productivity growth in Europe. This as we know was started by the famous Draghi report that pointed to a big decoupling in productivity growth, at least on his analysis, between Europe and the U.S. in the last 20 years, and that is a concern. We need to catch up. To be brief, the debate that took place over the summer was between two camps. One camp was Luis Garicano, Philippe Aghion, Bergeaud, essentially saying this is serious, and Europe really needs to get its act together, we are falling behind. And then there was another camp led by Paul Krugman, another Nobel Prize, saying you are overreacting, you’re exaggerating. If you measure it in different ways, the productivity growth decoupling is actually not there, or it it is much smaller. That was a big discussion that occupied a lot of economic space over over the summer. Tell us where you stand on that, and why you think it matters, if it does indeed matter?
SV 06:01
I think it does matter because this is where a lot of the Draghi diagnosis starts: there is a large and growing divergence between the U.S. and and Europe and productivity growth, and that leads to essentially big differences eventually in living standards. And I think this is maybe the translation between a measure of GDP per capita and conclusions on living standards: moving from one to the other is not always obvious. I think Luis Garicano really played on that Draghi alarm, stressed that this divergence in GDP per capita, measured in PPP terms, so adjusted for purchasing power, was a very big issue. I tend to be of the view that it’s not as big an issue, and in that sense I’m probably closer to Krugman. In part because when measuring things in PPP terms, we do these price adjustment, and it’s important to look at the composition of the consumer’s basket. Basically we tend to overstate GDP in the U.S. which has grown a lot through services that are typically free or relatively cheap in Europe and whose prices have grown massively in the U.S.: education, healthcare, and so on.
If you account for that, what Krugman says at the end of the day is the Europeans yes have a smaller GDP and GDP per capita, but they’re actually able to buy with the same basket services that are a lot cheaper in Europe than they are in the U.S. and goods that are essentially traded globally. And so if you’re a European you can buy an iPhone roughly for the price of an American, but you have healthcare services and education and other services that are free when an American doesn’t, is it so clear that the living standards in Europe are so much worse? I think that’s a fair question to ask. And so I tend to think that we have a productivity problem that we should not deny, but I think concluding from that that there are profound divergences in living standards between the U.S. and Europe to me is a bit of an exaggeration, and it’s an exaggeration that I think can be useful for momentum or policy purposes, but in rigorous terms may not necessarily be be true.
CC 08:48
So you also think that the analysis of Mario Draghi somehow overstated the problem?
SV 08:58
Yeah, I have two issues with the Draghi report and its foundations. The first is this one: I think it overstates the divergence with the U.S.. In addition I think it overstates another problem, which is painted as central for Europe: the problem of internal trade frictions that would supposedly create the equivalent of very substantial tariffs between European countries. So the motto of the Draghi report is we need to bring down these internal non-tariff barriers, and they are super large. And when we do that, we’re going to have a massive productivity boost, and I doubt that. I’m not denying that there are internal market frictions, and we don’t have a fully functioning Single Market. I do believe that there are some economic gains to having a more integrated Single Market. But I don’t believe we’re going to have a leap in productivity from that. And I think we overstate these non-tariff barriers. A lot of these frictions are calculated assuming people would be agnostic and neutral about their baskets of goods and services if these frictions did not exist. So, for instance, these measures would say if Italians buy Italian wines it is because there are market frictions. No, Italians buy Italian wine because they have social and cultural preferences that lead them to buy domestic products. So the belief that we’re going to erase these social local preferences just by some kind of act of magic deregulation is a fantasy. That’s not going to happen. We overstate these frictions, and if we were to reduce some of them, and I think we can and should, I don’t think this is going to lead to such a massive leap in productivity. So I am a bit circumspect with the Draghi push. I’m all in favour of Single Market integration but I don’t want to mislead people into thinking that if we do that, we’re suddenly going to move our potential growth from 0.5 to 1.5%, we’re not going to gain a percentage point of potential growth by doing these things.
CC 11:19
This is fascinating because it is the absolute lament that you hear across the corridors of Brussels, across the capitals everywhere: the reason we are where we are is because we haven’t got a deep enough internal market and we haven’t achieved the Single Market. This is what I hear endlessly and it becomes an excuse for people’s lack of agency because people are then saying, “What can I do? I’m not going to do very much because the Single Market isn’t there”. It becomes chicken and egg, which I very much resist because whatever the barriers, they’re not in any event going to be eroded in my lifetime.
SV 12:02
These barriers exist to an extent also in the U.S. in other forms. We lament the lack of a Capital Markets Union. But also, let’s not fantasise what the American capital market is. There are profound regulatory and supervisory issues in the U.S. that create deep frictions, and so they have issues of their own of their own, which we should not entirely forget and minimise. Where I would agree with the Luis / Draghi / Rhine Group conclusion is that we have potentially an innovation problem in Europe. I think that’s real and profound and it’s connected to many different things, bigger than Single Market integration. We have a human capital issue, which is connected to our educational system and our immigration policy. The reason the U.S. today is leading in terms of innovation it’s because it has a much better higher education system than ours. It puts a lot more money into it than than we do, and it attracts talent globally that either we can’t or we don’t want to. More often than not, we don’t want to.
CC 13:33
I want to push back on the innovation focus that you have. What do you mean exactly by that? I think that Europe has tremendous capabilities and talent in tech. We are certainly no less good than U.S. engineers or technicians in any way. We have great research. What matters is we are not funding that, from startup into a scale up. That is the endemic problem. We don’t scale up. I wouldn’t call it an innovation problem. It’s a productization problem. We innovate, but we don’t fund the scale up. Is that where you are?
SV 14:23
Yes we have very talented engineers and scientists in in Europe, but the reality is that we have less than in the U.S. Simply because the U.S. imports some of the best scientists and engineers in the world, and so there’s this joke about the AI race as being a race between the U.S. and China, and people say no no no no the AI race is a race between Chinese engineers in Silicon Valley and Chinese engineers in China. I think that’s the reality. If we want to be at the leading edge of frontier labs in AI, there is no way we will be just with the stock of human capital that’s available in Europe. In order to be there, you have to have the best engineers coming of India, from China, and we don’t have them. Why? Because these people, and maybe this is something that’s changing, preferred to get education in the U.S., had better work opportunities in the U.S. after getting their education, and we’ve done very little to compete with that. I also think we have all we need to compete with that, probably more so today than ever, given the way immigrants are being treated in in the U.S. If Europe had an open door policy to every PhD in the world, things could look very different on the innovation front.
CC 15:53
But we have the problem of funding. We pay people less, and that’s also why they don’t come, and that’s why they emigrate over there. Isn’t this putting too much emphasis, though, on frontier labs as the magic sauce? We have been certainly radicalised into fetishizing American frontier labs. There are the three leading guys, and we need to be on our knees as Europeans to get access to their models. In exchange for which we can give them access to build data centres for us. But for most users, aside for frontier research and people really requiring something very special, the majority of European businesses can do very well with models that are not frontier. Plus, the Chinese have reached almost parity, so do we really need to? My position in all of this is that on in AI what we need to focus on is not whether we can produce the next frontier model because we can’t, Mistral itself essentially abandoned that ambition. They wouldn’t say they abandoned it, but they are not prioritising it, and there is a much different stance when it comes to AI that Europe can still do something about-not frontier, but close to frontier, investing in the stack and being a platform. How do you see it?
SV 17:27
I’m completely with you on that. This is something indeed which has been largely engineered by the Frontier Labs: this belief that we are in a in a race, and that if you’re not running the race to win it, then there’s no point. I agree with you. Europe knows that it’s not going to win that race for a number of reasons. One being our resources when it comes to engineering, and the other even more important is the financial resources we’re prepared to commit to that. The U.S. is going to commit 700 billion this year on data centre capacity, we haven’t committed that in the last 10 years. But I’m not sure this is actually such an issue, provided and this is a big if, and that speaks to the geopolitics of AI, provided AI remains fairly open. This is where this friction between open weights versus closed weights model is so central. The U.S. has decided, or at least the hyperscalers, except Meta, have decided that this is a race that needs to be close, and they want to be at the frontier with closed models, and that’s the game that Altman and Anthropic are playing. Meta and interestingly now Nvidia are playing a different game. They’re saying that the race is to keep open weights model accessible because actually the real value in AI in the future will be in the application layer, which is where I think Europe has a role to play, and where Europe could actually be as good, if not better, than the US and others by actually deploying a very thick application layer across our economic space, so that could be powerful.
The caveat for the European strategy that that you describe and that both you and I seem to agree on is that there is still between China and the U.S. a desire to keep open weights and to allow others to build an AI stack on their foundational models, and this is where the danger is: that we’re reliant on the idea that the U.S. and/or China, or today actually China and/or the U.S. is going to build a significant amount of capacity in open weights model for the rest of the world to build its application layers. I tend to think that they will do it because it makes economic sense for them. But you could imagine a world that becomes extremely adversarial between the U.S. and China, where none of the current discussions we’re having around AI security architecture lead anywhere, and where we turn into what the hyperscalers, in particular the frontier labs, really want, which is a war of attrition between China and the U.S. for the best closed model and for AI dominance. I don’t think this is the world in which we are or necessarily will be, but I think we have to accept that this is one possible scenario. And in that scenario, the European strategy that both you and I describe and agree on is a very dangerous one because then we are very vulnerable. To some extent, it’s a strategy that is consistent with the old European tech strategy, which is let the U.S. build the expensive stuff, and we’ll build our own things on top. It’s created profound dependencies that we regret today. The strategy is consistent with what we’ve done, and it’s consistent with our endowment, both in terms of human and financial capital. But it carries some risks, and I think we have to be aware of these risks.
CC 21:30
I’m very animated by this issue because I do believe that we’ve been leading ourselves towards this dependency; but more than that, when you are building on somebody else’s infrastructure, you don’t capture value in full, and that entrenches your dependencies even more because investing on somebody else’s infrastructure is not the same as investing on your own. But I agree with you. There is so much uncertainty at the moment. What you see is the U.S. lab somehow flipping as Nvidia buys Hugging Face, and everyone is hedging towards open and closed weights. And then you’ve got the attraction of the Chinese models that for now are super cheap. They’re up to 1/200 of some of the cost of frontier lab tokens, so hugely attractive for Europeans. I know that many are experimenting. But the issue is this could be a “bait and switch”. At some point, they could turn into extracting value and getting much more from it. On the other hand, we don’t have an option. We need to build the industrial applications on top. This is dangerous either way. Building them on American closed models and infrastructure is a bad idea because it just completely gives up every form of agency and autonomy that we have. How much does this worry you?
And there’s another dimension in all of this AI discussion, which is the growing concern around the financial bubble that’s built around it. There are all sorts of magic tricks – Nvidia is lending to its own customers, there is all this circular finance, but more than that, you quoted a big number in terms of this year’s investment alone, much of it is debt, and I’m hearing voices in the U.S. on the financial side – I’m talking about insurance companies, institutional investors who are beginning to fear this is all just about holding it together ahead of these IPOs, but then there will be a considerable shakeup, and they are the ones who will be left holding the problem. The final burden is going to be ultimately on pension funds and insurance, and all of this securitized on student debt and so on. How do you see it?
SV 24:21
I am concerned as well about the financing of this infrastructure capex boom. It’s important to put it in perspective. It’s really historical. I mean we’re we’re talking about sums in terms of percentage of GDP that are higher than we’ve ever seen, higher than the tech bubble of the late 90s/early 2000, but also higher than the electrification investments, higher than the railway investments of the late 19th century. Now it’s not because it’s enormous that it’s necessarily a bubble. What makes it a bubble to me is that the revenues in front of these investments are, for the time being, very small, they need to grow, and for that you need to have a very substantial jump in productivity. I think there will be some, but it could also be that it’s a lot smaller than what people make it to be. We see AI everywhere, except that we don’t see yet this boost in productivity, is not nearly as large as what will be required in front of these large investments.
And then, even if there is eventually a big jump in productivity, who is going to capture the value? It’s not certain that the value is going to be captured by those who have built the infrastructure, and so you could be in a world where indeed AI pays off, but it doesn’t pay off for those who have actually invested. And this is where I think the Europeans have maybe a way into this. It might pay off for the application layer that’s actually used and actually generates revenues, but maybe not for the big data centre that has buried billions to actually build the foundational model underneath. So I worry about that, for a long time, essentially until the middle of this year, this was all financed by the cash flows of the hyperscalers, and so you could say in the worst case they’ve burned their cash flow, their shareholders will lose money, but nobody’s going to get burned. Now, as you said, we’re entering a new phase. Essentially, since this year, cash flows are negative across the board amongst the hyperscalers, and this is no longer financed by equity. This is essentially financed by debt, and that changes fundamentally the moment in the cycle in which we are. It’s financed by debt. It is sometimes hidden debt, so it doesn’t sit on the balance sheet of the hyperscalers. It’s financed by debt securities that are not listed, and the price discovery mechanism is therefore a bit blurred. It’s financed a lot by private credit, so we don’t have to mark this to market. This is where I worry. The circular financing, the debt financing, the private credit financing, all coming together into feeding a beast that may deliver a return, but maybe much later and maybe not as large as what investors have put their money for.
CC 27:55
Agree. Does it surprise you we are in a narrative in the last week that seems to say, well, there’s no job apocalypse, and in fact, AI has created a million jobs in the US. Where is that surprising to you?
SV 28:09
No, because I think in the short term, the extent of the capex build out, the need to build these data centres to create the fibres, the cables, the construction- will create jobs. The short-term boost to GDP is there. But It’s interesting to see the distribution of jobs. You have a lot of construction jobs. But where you see struggling is parts of the labour market for the young graduates. A young graduate in the U.S. struggles to get into the job market a lot more than an unskilled labourer who is actually working on a construction site. So that’s an interesting rotation, and in aggregate, it looks like employment numbers are okay.
CC 29:04
At least temporarily…
SH 29:06
You can have both for a while, so long as the capex build out continues to fuel the machine. The moment the capex build out slows, then it’s going to be a different picture.
CC 29:19
We could talk about this for a long time. But I want to pivot back to one of the places I wanted to dig into, let’s move away from AI for a moment and and think about where Europe is in terms of the other sectors. Industrial policy is one of the tools that Draghi pushed, for better or worse, he is very much saying we have passed the time globally in which there was suspicion about industrial policy. Industrial policy is an essential way of progressing. What is Europe doing in this space? Because it is hard to decipher. Of course, there is the usual problem you have DG Grow that has no real mandate and does not hold the purse strings. They are held in Berlin. They are held in Paris. They really don’t have any spending ability, and so what the Member States are doing matters as much or more than what Brussels is doing. That said, we we’ve come from the experience of the U.S. where IRA was an extremely effective instrument that created both on the supply side and the demand side great stimulus to many sectors. Here we don’t. It’s hard to discern what Europe is doing to boost itself in a situation where our predicament is so grave.
SV 30:54
When I take the long view, I feel that actually very profound changes taking place in Europe on some of these issues, on trade policy, on competition policy, on industrial policy. If you look back to where we were 10 years ago, we’ve travelled quite a lot of distance intellectually. But I think it’s important to distinguish how our thinking has changed, and how we are implementing these changes. I think intellectually the consensus has shifted a lot. I think the IRA that you’ve mentioned has played a big part in making us realise that the U.S. is doing aggressive industrial trade and competition policy together. This was a big shock, realising that the U.S. was moving there very aggressively. We’ve ignored for a long time the industrial policy efforts of China, even though what was announced in 2015 with Made in China 2025 was a superb and very well-designed industrial policy that we should have psid attention to already back then. We did not take enough notice of what was going on. But I think now and the discussions we’re starting to have in Europe with the Industry Accelerator Act, to me are philosophically very profound. It’s the first time we’re discussing introducing local content rules. It’s the first time we’re discussing introducing investment FDI screenings and mandatory joint ventures and mandatory technology transfer. So, I think intellectually we’re pushing a lot the boundaries of what was deemed possible just a few years ago. That takes the form of an internal trench war inside the European Commission between DG Trade, DG Comp, DG Grow, and so on that needs to percolate, and also takes the form of intense fights between Member States who have different views on that. What is very interesting to me, and I think that’s maybe the last shoe to drop, is that Germany, which had historically been very conservative on these issues, is finally realising that something has to be done, and that’s an important shift. It has not fully occurred yet, Germany the biggest automaker in Europe has not voted on whether we should have EV tariffs against China. It has not voted. It’s not even that it voted yes or no. It abstained. It’s the single most important policy decision that the EU has taken for Germany, and because the German coalition government could not agree, it abstained. But that’s changing. If there were a vote today as to whether we should have EV tariffs on Chinese vehicle, I think the German government could not get away with abstaining. It would have to take a decision. So I think things are moving, but very slowly in a time where economic changes are very rapid and profound, and my worry is that the intellectual battle that I’ve described and that I think is taking place will only produce results too late for the fight. It will produce results once the economic impact has been so grave that it becomes almost irredeemable.
CC 34:43
But isn’t that the whole problem with our history? I don’t disagree with you. You call it an intellectual discussion which is progressing. I think you’re being generous because I still detect a huge amount of resistance and entrenchment. At the level of the individual competences, is still manifests itself very much. Yeah, in the long term, there may be an improvement of vision, but seen from the trenches, I don’t see that. The Commission’s President says competition policy needs to be more sensitive to the European predicament and no one in DG Comp is taking any notice other than paying lip service. Oh yes, we’re going to have something about dynamics and something about benefits of mergers but it’s the same old stuff with a wrapper. Take industrial policy. I agree with you that there is an interesting effort on can we use demand tools? It’s never been done before, But the resistance to that at the level of Member States and DG Trade too seems huge. This debate is yet to happen this autumn. Where will it go?
SH 36:10
My fear is that it’s not going to be conclusive. In principle, a lot of these issues ought to be settled by the end of the year if we want to have the IAA finalised by then but I think that it’s not going to be legislated by the end of the year, and that will take more time, and that eventually the IAA will be a very important intellectual leap for the European Commission and for Europe, but it’s not going to produce massive economic effects because we’re going to create a machinery that is extremely complex and bureaucratic, and actually doesn’t bite as much as it should. In part because of disagreements and compromises that we have to find. The first discussion was on local content, but then there was this big fight about how do we define “local”, and our friends from Bruegel have been saying well maybe we should stop talking about “made in Europe” and talk about “made with Europe”, which basically meant anybody that has a trade agreement with Europe and anybody that is a party to the WTO procurement arrangement, which essentially meant the entire world, other than China. And that to me is is is really a problem. We cannot define local content in that way and assume that there is reciprocity from these countries because they are party to WTO or of the public procurement agreement.
What I would be a lot more comfortable with is a narrow definition of “made in Europe”, which can be opened up for countries that prove to us that they are actually applying reciprocity. So let’s take South Korea or Japan, if European firms are equally eligible to public procurement in South Korea, I think Korean firms should be equally eligible in Europe. But that’s not the case today. If European vehicles are equally eligible for EV subsidies in Korea as they are in in Europe, then we should reciprocate. But that’s not the case today, these are the things that we need to change, and I am afraid things are moving in that direction, but very slowly.
CC 38:36
But how much time do we have? Because the reality is that even within the Member States you have the French who are the most favourable to these kinds of Buy European rules, historically and traditionally. Then you have the Scandinavians who think that we are all going to be doing terrible things to WTO. In addition to that, there is transatlantic pressure. In digital there was a draft law, the Cloud and Data Act, which was supposed to contain in theory mandates to the public administration to direct procurement towards Europe. And it’s not a secret, everybody knows that just ahead of the issuing of this draft, there was so much pressure from hyperscalers and the American administration along the lines of “nice car industry you have there, wonder what would happen if tariffs were even higher”, that everything to do with mandates was removed. The text is all “the Commission will favour, the Commission will promote, the Commission will encourage”, but that doesn’t mean anything. The IAA contained something about chips and AI but it was removed because everything digital is DG Connect and not DG Grow. How is this going to be composed into something coherent? That is my concern. I don’t see this improvement. Maybe intellectually. But we are beyond that, it’s like oh, the problem is at the door. No, it’s not at the door. The problem has broken the window and is rampaging through the living room.
SV 40:22
Yeah, so here I have a view that’s different from a lot of the colleagues I write on these things about, like Brad Setser or Sander Tordoir, I am less optimistic than they are about trade defence instruments against China, introducing higher trade barriers. I tend to think, and it’s a leap of faith that the only way out of this bind for Europe is essentially to embrace the idea of Chinese investments into Europe. But I think embracing Chinese investments into Europe, and in fact this is probably the least discussed part of the IAA but in my view the most interesting and important: how do we frame an investment framework that is sufficiently conducive for Chinese firms to invest in Europe and produce in Europe rather than export from China, but also sufficiently tight to maximise technology transfers, local employment, local content?
I think that’s possible. And it’s possible because actually the Chinese have done it. The Chinese for the past 20 years have imposed on European companies investing in China a whole host of restrictions. You have to have a joint venture. You have to employ Chinese people. You have to build locally. You have to have your data locally. The result of that has been a massive technology and intellectual property transfer to China, of which we are seeing the results today. The play for Europe now is basically to do the same in reverse. To basically say we are in a reverse technology transfer moment. We are going to be nowhere in batteries. We’re going to be nowhere in EVs. We’re going to be nowhere in all the green tech if we’re not doing this with China in Europe, and so that requires, in my view, not to impose massive trade tariffs on China, which is only going to increase the price of these goods on which China has effectively a global monopoly. We need to basically tell the Chinese we want you in Europe, and these are the conditions for you to invest in Europe. And you cannot tell us that this is very unfair because this is essentially what you’ve done. And so that, to me, is the more promising way out of this bind, it’s not the one that’s talked about the most because that assumes embracing a bilateral relation, economic relationship with China, which is not exactly the one we have now, and actually creates a degree of friction in our transatlantic relationship.
CC 43:21
That was where I was going to go because every American I talk to is saying, “Well, you guys, we’re okay if you guys want to be more independent and decouple from us somewhat, but what you’re really doing is getting in bed with China, and this is not okay”. So there would be a freaking out moment in the U.S. about that if we ever did.
SV 43:42
Yeah, there is, and at the same time, the U.S. has proven to be such an unreliable partner to Europe that I think it’s only fair if Europe diversifies its risks away from the U.S. towards China. So I think that’s a reasonable thing for us to do, but I agree with you. It comes with a risk of retaliation from the U.S. on that basis.
CC 44:10
So we are in between a rock and a hard place because there is also talk that the traditional trade defences don’t work. We need to think of a new tool, and what is it? And there is a big concern about escalation. We’re going to escalate, but what if that doesn’t work. And we are going to upset the U.S. relationship, whatever is left of it.
SV 44:35
No, I think there are two ways to view this. We’re either stuck between a rock and a hard place. That’s the “glass half empty” view of it. The positive way of seeing it is we can be in a position to create leverage both from the U.S. and from China if we basically manage to leverage the fact that we are in the middle and both of them need us. Either as a customer or as a partner, and so that takes us back to the conversation we had on on on AI earlier and open models. We’re potentially very vulnerable, but so long as the U.S. and China want to compete for this, there is a venue for us to say if the Chinese model are not performing X, Y, and Z, we’ll go for American models, and we’ll tell the Americans if you don’t want to do open weights model that have these and that characteristics, and that respect our GDPR and that respect this, then we’ll go China. And that, if you can exploit it, gives you a lot of leverage. I don’t think we’ve been good at exploiting that position so far, but I think that’s an opportunity for Europe.
CC 45:40
But Shahin, where is this leadership and vision going to come from? Because you have clarity of thought, and other commentators are people I could listen to forever. And then look at the reality, and who is going to design this plan in which we can play one against the other? Because it isn’t remotely where we are, and the worry as you said yourself, not only we got the Member States, we got the various functions (fighting). So at the moment, we have an aspiration that we could play this, we could arbitrage between these two great blocks. It’s painful that we are just arbitraging rather than wielding our own power, but where is it going to end?
SV 46:24
I hear you. I think today it doesn’t look like we are really capable of doing that, but I I’m hopeful that we can. What makes me hopeful? The fact that I think the intellectual consensus is shifting, so we’re still in the middle of the road, with a big part of Europe, not to name it Germany, that is still very focused on restoring or repairing the transatlantic relationship. I don’t think they have completely moved on from that, for lots of reasons that are more than 50 years old and that we can understand very well. But I think that will eventually come. So to me, it’s probably the last time that we have a President of the European Commission that is so transatlanticist, she’s German from a certain generation, and historically this is, in my view a tipping point. I think people have realised across Europe and generationally that the relationship Europe had with the U.S. over the last 50 years cannot be the relationship we will have over the next 50 years. In part because the U.S. actually doesn’t want that relationship anymore, either. So this is why I’m hopeful because I think these slow intellectual battles that we’ve described are slowly bearing fruit. It’s of course very frustrating when you and me, who are kind of in the thick of it, would want to see immediate action. But if you, I think, take the long view. You’ll say, “Oh these were the five years where things moved very slowly, but where they tipped over that edge and they moved decisively in that direction”. And I think this is the moment we’re living.
CC 48:13
You’re more patient than I am. Last question: the midterms are on us. It’s a matter of a few weeks. The general expectation is that there will be a shift, as is frequently the case. How do you see it playing out, and how will it play with Europe? There is even a view that the Dems will retake not just Congress, but also possibly the Senate, and that would make Trump a lame duck, and then it’s over, and we we don’t have to worry about that anymore. We don’t know. We simply don’t know. In two years, will this momentum that you see for changing things in Europe, slow as it is, reverse towards a more transatlanticist kind of thinking?
SV 49:11
The worry I have is that people would read too much into the midterms and think if the Dems win the House and maybe even the Senate, that’s it. This is the end of MAGA. This is the end of Trump, and we can go back to the status quo ante. I think that would be dangerous because there is no going back to the status quo ante. And what’s very important is to realise that some of the movements that are embodied in a very kind of brash way with Trump are actually deeper movements that would follow whoever is the successor, including a Democrat. And I think that’s important to realise.
CC 49:47
My sense is also that Democrats really don’t have a plan. It is “we will win” but then what will you do? Unknown.
SV 49:57
Yeah. What I’m more worried about in the U.S. Is not so much the midterms, but something very important that has happened post Liberation Day, which is the realisation by Trump that the confrontation with China was not possible. To me the more salient thing of the last year is basically the U.S. chickening out and stopping its direct confrontation with China, and I think that has given Europe some space. If we return to a world of very direct confrontation with China, the space that we have today and the leverage that we have to play one vs the other, trade with one and not the other, that leverage goes away very quickly. So I think we should use that window. And the worry I have is that at some point, this structured dialogue that has taken place for the past year, since Liberation Day, the de-escalation, that’s the window that I worry would close, and that to me is more important.
CC
There’s another question I wanted to ask you around the trade debate, a big debate that’s taking place over the summer. So let’s get into it.
CC 53:34
One of the other debates that has been live in the summer was around the Second China Shock and trade imbalances, the fact that China’s trade surplus is growing at great speed, and simultaneously you have a big deficit in the U.S. while Europe is somewhat in surplus. These things, as has been noted, are never about trade alone. It’s not about trade. It’s about the underlying savings, investment, production decisions that are being made. But there is discomfort about these imbalances because they animate President Trump for sure, and they worry others. And there is a view that if one was able to somewhat reduce these imbalances, then there would be a positive effect on growth. This is IMF analysis and so on.
The discussion around the causes of these imbalances has has been animating people over the summer, and one focus has been the role that exchange rate plays and currency valuation. There’s been a debate between yourself, Brad Setser, Sander Tordoir, also Michael Pettis on one side, and on the other side Helene Rey and Gita Gopinath [and Gourinchas] about the role that the possible undervaluation of the Chinese currency plays in all of this. Why don’t you orientate us a little bit on on that and where you stand?
SV 55:18
I think there is a very interesting discussion taking place, which the French presidency of the G7 played an important role putting on the agenda, which is we have large and growing global imbalances. This is not what the IMF has forecasted over the last 10 years. Over the last 10 years, the IMF was telling us no, no, global imbalances are peaking and they’re going down. Don’t worry. And in fact, the opposite has happened. Global imbalances are large and growing. And what the IMF has said repeatedly is, when you have such levels of global imbalances, they either come down through policy coordination or they come down with a bang through a financial crisis, and so it’s in our interest to opt for the first option and try to create the international policy coordination framework that helps bring down these global imbalances. And this is where understanding the policy levers becomes so critical. Brad Setser and I argued that not only exchange rate policy played a big role in fueling global imbalances and, in particular, exchange rate manipulation in China and more broadly in Asia, so that the currency undervaluation in China was an integral part of China’s economic strategy, and that had helped fuel large Chinese external surpluses. And I think on that, the IMF does not really disagree. They don’t write it in such firm terms, but they broadly agree.
Where there is disagreement is our conclusion that exchange rate manipulation plays a big role in building global imbalances, and the exchange rate can also play a big role in bringing down global imbalances. And that requires exchange rate policy coordination, or at the very least, China, maybe by by extension the rest of Asia, to drive its currency to to appreciate. And this is where, for reasons that are a bit mysterious to me, there is a divergence between Brad and I on on one side, and Pierre Olivier Gita and Helene Rey, who are like all economists that I admire and whose papers have been really trained by, who say, well, actually, the nominal exchange rate doesn’t really matter. What matters is the real exchange rate, and the real exchange rate is going to adjust. And so anything we do on nominal is irrelevant. That to me is very weird because that rubs against decades of IMF advice who would say you need to float your currency or you need to do that exchange. So I find it very counterintuitive that the IMF would conclude that nominal exchange rate doesn’t matter. And to me, what I read in their argument is in fact a political argument, which I think is more interesting. Which is, do we have the leverage to push China to change its exchange rate policy? And it’s probably true that we don’t have much leverage. But then I think it’s a different argument: they could say you’re right, China should move the exchange rate. It would have an effect. But the reality is that we have no leverage to push China to do that.
That’s a different argument than the one they are using, which is… it doesn’t matter that much because in any case, if China revalues the exchange rate, domestic prices are going to fall. There’s going to be a deflationary shock…
CC 58:53
You describe this stance as a bit mysterious. Do you rationalise to yourself why they take this position?
SH 59:01
I think they have internalised that we cannot push China on the exchange rate.
CC 59:08
That’s a different argument.
SV 59:08
Yeah, but I think that’s a wrong conclusion because there are historical precedents where basically faced with the alternative, meaning high tariffs, for instance, China has chosen to let the currency appreciate. For instance, it’s not random that the renminbi has appreciated since Liberation Day. It’s because the Chinese have understood that in their bilateral relationship with the U.S. it’s actually better to improve that relationship to let the currency appreciate slowly. So there is leverage there. I think there is more leverage, and what I think is more important is that the policy advice of the IMF, which is China needs to change its domestic macro industrial policy, that advice is very true, but it’s not going to happen so long as the exchange rate is structurally devalued. If you create pressure through the exchange rate, you have a higher chance of seeing the policy prescriptions of the IMF taking place, and I think we have, even though it’s a long time ago, we have one example of exchange rate policy coordination actually working, and that’s the Plaza Accord of of 1985. Now, is the moment ripe for something like that? Maybe not. Is that what we should have on the horizon as an objective? I think so, and I think this is where there is more discussion to be had between the Europeans, the U.S. in the G7, and potentially eventually in the G20, to see what kind of policy coordination framework we can get to get Asian currencies to appreciate, and in particular the RMB. I find it quite ironic that there is now joint intervention by the MOF in Japan and the U.S. Treasury to appreciate the renminbi, so it must mean that the exchange rate matters to someone, and at the same time, an argument that somehow the Chinese exchange rate doesn’t matter and cannot be touched. And my sense is that, and of course not openly, but I think through the right channels, it’s a conversation that can be had with China. It’s a difficult conversation to have.
CC 1:01:28
Who can have it?
SV 1:01:29
To me, this is both a discussion that takes place between the U.S. Treasury and and China, and when it has taken place in the past, it was at that level, but where I think the Europeans can play a role arguing that revaluation is in the global interest, and I think they’ve done that a bit too timidly. And what I find very interesting is that the Germans have started to talk about exchange revaluation at the G7 in France, and what needs to be happening now is the Europeans beyond that, carrying that torch, at the G7, at the G20, at the IMF to basically push the envelope.
CC
Fantastic! We’re at time. We spoke for an hour. I love your energy, and I love how you move all the things that I consider super interesting. So to me, this is a treat. Thank you so much, Shahin, and till next time.
SV
Thank you for inviting me.
Escape Forward 
Leave a comment