Trade wars are damaging, difficult to control and usually best avoided. But what should a country do when economic conflict is imposed upon it?
“The trade wars are coming to us, and we’re going to have to engage.”
Soumaya Keynes, Financial Times economics columnist and co-author with Chad Bown of How to Win a Trade War, joins Ben Yeoh to examine what economic conflict now looks like as the old trading order weakens. The World Trade Organization still exists, but its dispute-settlement system can no longer reliably manage the largest tensions. In that world, simply repeating that nobody wins a trade war is not much of a strategy.
Soumaya argues that defence should come before attack. Governments need to understand their real supply-chain vulnerabilities, maintain strategic stockpiles where the insurance is worthwhile, and identify areas of strength that could deter an adversary. Yet resilience is harder than it sounds: companies often cannot see far enough into their own supply chains, while political pressure can turn sensible preparation into an expensive pursuit of self-sufficiency.
They also discuss tariffs and industrial policy. Tariffs are taxes on imports, generally borne largely at home, and are far too blunt to achieve every objective politicians attach to them. But Soumaya also rejects the idea that tariffs and subsidies are always wrong. Where a country has acquired dominant market power and shown a willingness to weaponise it, governments may need to intervene. The real challenge is doing so without creating wasteful subsidy races, protected national champions or a retreat into autarky.
The conversation then moves beyond trade: to China’s industrial model, the importance of competition and tolerated failure, the relationship between obesity, gender and income, and the wider economic effects of GLP-1 medicines. Soumaya also explains her slow, highly edited writing process, why she keeps AI-generated prose out of her writing process, and how she is turning economics into songs.
Takeaways:
Economic conflict requires preparation, not just retaliation.
Strategic stockpiles can work precisely by never needing to be used.
Tariffs may sometimes be necessary, but they remain costly and unreliable tools.
Industrial policy needs competition, discipline and a tolerance for failure.
GLP-1 medicines may affect employment, relationships and inequality as well as health.
In an age of abundant AI writing, a recognisable human voice may become more valuable.
Summary contents, transcript and podcast links below. Listen on Apple, Spotify or wherever you listen to pods. Video above or on Youtube..
Episode Contents
00:00 — Intro
01:00 — Are trade wars ever worth fighting?
02:09 — What happened to the WTO rules-based system?
04:28 — Can trade wars have any silver linings?
07:59 — Why defence should come before attack
09:30 — Mapping supply-chain vulnerabilities
13:01 — Strategic stockpiles and the politics of insurance
17:29 — What does winning a trade war mean?
20:05 — Market power and economic offence
22:16 — What tariffs actually do
25:15 — The return of industrial policy
32:48 — What the West can learn from China
35:03 — Soumaya’s return to health economics
37:40 — Moving beyond trade
38:28 — GLP-1 medicines and their wider economic effects
39:25 — Patents, prices and access to medicines
41:39 — Soumaya’s writing process
43:50 — AI, writing and preserving your voice
45:31 — Economics songwriting as a creative niche
46:39 — Overrated or underrated
51:37 — Future projects—and no more books (for now!)
54:38 — Soumaya sings her trade-war song
57:10 — Final advice: do less, but better
Transcript (transcribed by LLM and lightly edited, mistakes are possible)
Ben: Hey, everyone. I'm super excited to be speaking to Soumaya Keynes. Soumaya has written with Chad Bown an excellent book, How to Win a Trade War, and she is an all-round brilliant person.
Soumaya, welcome.
Soumaya: Thank you so much for having me and for that lovely introduction.
Ben: Before we talk about how to win, I thought I would ask, is there ever a good time for a trade war, or in an ideal world should we be thinking about avoiding them?
Soumaya: Yeah, clearly trade peace is better than trade wars, right?
Just in the same way that, normal peace is better than military wars. Con- conflict is bad. You heard it here first. I think that the premise of the book is that there are some cases in which you can't avoid the fight, right? Because in some cases the trade wars are going to come to you.
And there is this line that economists like to trot out, which is no one wins a trade war. And oh Lord, the number of times that people have given me that line over the past year and a half. But, that... It's trite. It sounds smug. And it just isn't appropriate anymore for where we are because the trade wars are coming to us, and we're gonna have to engage.
Ben: So is the old rules-based trading system truly over then, or have we given up too early or quickly, or could they come back so this is not the new status quo? It can go back and forth in a kind of equilibrium
Soumaya: Yeah. So it's probably worth being quite precise about what we mean about the rules-based trading system.
So if you think that it is a world of trade deals in general, then there are many of those still in effect, many of those functioning perfectly well. There are some trade relationships that are clearly broken down quite dramatically, primary one being the US and China.
But around much of the world, the trading system is functioning. Now what people often mean when they talk about the rules-based trading system is something quite specific, which is the World Trade Organization and specifically its dispute settlement system, right? Under the old system, if you had a problem with the way that another country was managing its affairs then you would essentially sue them at the World Trade Organization to say I'm filing a dispute."
And then there would be a bunch of independent judges who would say, "Okay, you," "yes, they have broken the rules or not." And if they had, then those judges would essentially say, "Okay this would be a proportionate level of retaliation." And the purpose of this system wasn't to end all trade conflict, right?
It was to manage trade conflict and make sure that it didn't spiral out of control. And it's that system of judges that we don't really have anymore because the Trump administration killed it. We have a sort of interim replacement which the EU tried to set up, but many of the world's biggest players just aren't in it.
And so it can only do so much. The EU and China are in this alternative dispute settlement system, but I think it's notable that you're hearing a lot from Europe right now about its problems with China's economic system. I think it's notable that they haven't filed some major dispute against all of China's economic practices using this alternative dispute settlement system, right?
They're trying to go bilaterally. They're trying to negotiate. They're, they're maybe thinking they might have to impose trade barriers. So it doesn't look like this rules-based system is Europe's first port of call. And so yeah, I... it's clearly not working very well when it comes to the biggest tensions in the global economy today.
Ben: When I speak to some econ people, they say a silver lining of a hot war is that in the medium to long run, maybe you get technological booms or you get a change of paradigm and a change of thinking and things like that. Are there any silver linings to a trade war? I know it's very different, but as we come out of this and assuming we eat somewhere else, is there something which in a creative destruction we could actually end up with something where, oh, we didn't think we could do that, but because of where we got to, we were that?
Or to your earlier point, this is just something we should avoid, we can't avoid it now and it's just gonna be what it is?
Soumaya: Yeah, that's an interesting question. So historically I don't know that there are many silver linings of trade wars that I can think of. And, I think one of the messages in the book is, you really shouldn't run into these conflicts happily, right?
Our tools are terrible, our weapons are very blunt. They have all sorts of unintended consequences. They're very difficult to wield well. There is a reason why for many years the norm or, the kind of the instruction from economists was, do not- do not deploy tariffs, export restrictions, industrial policy bec- they're just very challenging to use well.
Now, it depends on your worldview, right? So there is a world in which actually, you know, a move in a trade war might be the catalyst to persuade everyone else to act, right? So let's just take an example of China's restrictions on rare earths exports, which happened last year.
And, from the US perspective, that was obviously very stressful at the time. I spoke to people, and the consensus seems to be that the Chinese made a strategic mistake, right? Because they didn't just target the US, they targeted... whether through incompetence or design they hit everyone, right?
Everyone was struggling to get rare earths. And what that did ultimately is galvanize everyone else and make America's trading partners think, "Oh, actually we also need to work to set up alternative sources of supply for these rare earths." And had China not done that, and had China only hit the US, I think there would've been a greater chance that those other countries could have maybe sat on the sidelines and thought, "Nah, this is a US problem.
We're not gonna try that hard to do something." Tactical mistakes in trade wars can inadvertently strengthen the other side. I guess I would, I probably wouldn't go so far as to say, "Oh, trade wars make us richer in the end." I don't know about that. I guess maybe one more point though would be, talking about industrial policy, right?
The, the most powerful argument against industrial policy is that we're just... the government's just really bad at doing it well, right? We just need to... We don't have the state capacity. Need to be realis- realistic about what the government's good at doing. And, managing massive subsidies and protection is not one of them.
I feel like the conversation now has been moving away from the question of whether we should do industrial policy at all and towards the question of how to do it well. So you know, if an outcome of trade wars is that governments build up some of that capacity and learn lessons about how not to create massive white elephants or repeat a kind of enormous waste of Chine- Chinese...
the China subsidy infused model, and then that could be a positive outcome. I haven't seen it yet, but I'm hopeful.
Ben: So hard to find a silver lining, but possibly depending on which side you're on or your worldview, there could be better coordination between some actors, and possibly we're taking industrial policy more seriously, which gives us a chance to do it better, but let's wait and see on that.
I... In the book, you talk about defense and offense in a trade war, which I thought was really smart, and I hadn't really thought about the fact that you've got offensive ideas and defensive ideas, and you also talk about needing objectives, which I hadn't really thought about either because w- what does it mean to win?
And I take it from the book, you lean into establishing defense first as potentially something that a nation ought to think about. Is this still your view, a little bit further on and having written it through? And if so, how do you build a good defense?
Soumaya: Yeah, I... It's absolutely still my view, right?
And it's, it's the basic recommendation that the Trump administration didn't follow, right? So you know, this idea that trade wars are like real wars and logistics win real wars, right? It's that preparation, that planning that you need. You can't just run onto the battlefield waving around a weapon, in their case tariffs, and hope everything will be okay because the other side might also have a big weapon, right?
You need to plan what you would do if that happened. And so yeah, so the structure of the book is we talk about the stakes, the players, the rules, and then the defense and then finally the attack and that, it, it takes a while to get to the attack because, that instant gratification is not gonna come if you want to engage in a trade war well.
Okay, so how do you build a defense? So there are three chapters in the book. One is about identifying your vulnerabilities. Another one is about subsidizing and the third one is actually my favorite which is about stockpiling. And the theme running through the book is, I think, a big problem in the way that count- countries have approached international economic policies.
They've ultimately been too insular, right? So they've assumed "Oh, this is the way we do things and everyone else is a problem," and, we- we're- we're not gonna think about spillovers of our actions to other countries. And, everyone's basically just been far too inward-looking, right?
And so the big lesson is, okay if we are going to engage in trade wars, if we are going to use these tools that we have not actively... they- they've not been the fashionable tools, right? And so we don't have a huge amount of evidence or experience on using them really well. We're gonna have to look to our trading partners, right?
So winning a trade war means learning from the other side. And so just to give ex- some examples, right? So this came out of our reporting. So we spoke to a ton of government officials when we were reporting for the book, and we spoke to officials around the world who'd been engaged in these big vulnerability mapping exercises.
And, the- these came out just after COVID, right? The politicians were completely freaking out. There had been these supply chain shocks and, maybe these shocks could come from everywhere else. Like it was, it was very alarming for everyone.
And what happened is that the civil servants, the economic officials were tasked with doing these, and they yanked back control, right? Because they were like no. We're not gonna let you use these as an excuse to panic about everything. We're gonna look at the data. We're gonna be evidence-based, and we're gonna actually show you that in most areas we are not massively vulnerable to an international trade shock."
Actually, when you look at the data, we do have a relatively diversified set of imports. In the EU, they actually were able to get more of a handle on internal and domestic production data so they could look there. So these exercises were all marketed as vulnerability mapping exercises, but in practice they were actually more like invulnerability mapping exercises with the economists trying to calm everyone down.
So that was how these exercises happened in more than one jurisdiction. But then, I think o- one thing that becomes really obvious is, you go through these exercises, right? And you think about it, okay these economists have reassured us that in most products there are no vulnerabilities really. But the problem is actually these exercises are very difficult because of the nature of supply chains, because these supply chains end up running so far back behind your border often to other countries, right?
And so it can be very challenging to see where the true vulnerabilities lie. And actually when you get really into the weeds of this and we, we do and g- give lots of examples, there are reasons why companies just don't know where their own vulnerabilities are.
Their suppliers won't tell them. And so actually it's really hard. And so slowly one by one we spoke to people who... Th- there was this realization that actually you needed to completely reframe this challenge that actually thinking about your vulnerabilities wasn't the best way to go about it.
Because in cases where you have identified a vulnerability, sure, plug that one. Great. Let's try and fix that. But actually what might be more effective is in international trade conflict, rather than trying to plug every hole, if someone comes to you threatening to exploit a vulnerability, if you can go back to them and say hey, we've got this strength and we're willing to exploit it," that actually might be a more effective strategy than focusing entirely on this kind of defensive approach
Ben: That's the whole best defense is actually a good attack.
Yeah.
Maybe before we go onto the attack and winning objectives I am interested in the stockpiling issues you bring up 'cause it's not immediately clear to me, and when you speak to different people you get a little bit of a different view from this as a little bit what you're expressing.
And some of it is also how much insurance do you want to buy because these vulnerabilities might not come up. But I think about it particularly, one, around sort of pandemic preparedness and another really perhaps around food security.
So when you speak to a lot of the food systems experts, because they're within the food domain some of them get some more alarming stats like UK's less self-sufficient and some of these.
And depending on your nation state you are less self-sufficient, but there is also more globalization in your supply chain than the others. And again on pandemics it's because there seems to be a small chance but definitely not zero, and then it comes, but then of potentially quite bad outcomes.
But how much do you wanna stockpile for that? Can you do it? Can you do it across political cycles? But then when you speak to the more generalist econ person or particularly when you're trying to put the political economy thing, you end up with they're saying actually it's not so bad.
It probably isn't necessarily worth this insurance risk. We'll have to live within that." Is that something that you've come across as a reading, and is it slightly different from the specialty? It feels to me a little bit adjacent to what we hear from AI world, like AI people, some of them are like, "Oh, we're gonna lose 10 million jobs of X and 50%," and blah, blah, blah.
And when you speak to the econ person, it's when you look at these other transitions, from an economics point of view it doesn't quite work like that." And this is interesting around various domain knowledge shifts. I- is that something which I've observed as just barking up the wrong tree, or is there some truth to that?
Soumaya: I know, I think there is a lot of truth to that, and obviously it just comes from the different areas of experience of stockpiles, right? In food systems, stockpiles can be a very wasteful consequence of subsidy programs, right? So if you're guaranteeing prices to farmers that means the government might have to be buying up lots of excess supply.
And then, you get just very distorted markets, right? If the government becomes a massive buyer. Obviously if, when the government then tries to sell some of it, you can get accusations that they're driving the price down, right? It's a very challenging thing to manage stockpiles in a food context.
When it comes to medicine, obviously stockpiling is baked in, right? It's just it's not really that controversial. It's just something that is done, right? For, vaccines for cholera and what have you those happen, they are boring. And they can be useful because they essentially can by building up your stockpile gradually, you are keeping companies that can make those drugs in business, right?
You are giving companies an incentive to s- sustain the capacity to make those vaccines, so the stockpile can be quite handy. I think the thing that is different and more difficult about a stockpile in the context of a kind of geoeconomic attack is that there's this tragedy that if it is effective, you will never use it, right?
And so the kind-- The political economy of a geoeconomic safety stockpile is much, much more challenging than other types, right? Because of course, if you've got such a massive stockpile that there's no point in an adversary weaponizing their supplies, then they won't bother, and it'll look like your stockpile was completely useless, right?
And maybe over time, then you allow the stockpile to dwindle, and then, oh, look, we didn't have a stockpile, and we're stuffed. And so, that kind of dynamic is quite important, and it's I think undermined some stockpiles in the past. I think it's really interesting, early this year we experienced-- we've experienced a massive oil shock in the Strait of Hormuz closure, right? And really it looks like what saved us is the fact that we had massive stockpiles of oil on hand, right? So there are clearly cases where actually stockpiles were great, right?
We have, possibly avoided economic calamity because of those large stockpiles. And it's not exactly clear what China did, whether it actually used those, its stockpiles or whether domestic demand was soft.
Ben: So that seems like a classic insurance problem. To some extent you never wanna use it.
It's very successful, but then it disappears and it's the background, and it's a regular cost, and people pay for it. But then when you really need it it comes in to save you. That does lead me to think about what winning looks like, and you mentioned a few things. But I'm intrigued as to the mix or how we should weigh ideas. Do you want to really change another country's behavior?
Are you strengthening your own economy or nation state or maybe your allies, or are you simply suffering less damage, so everyone's going down, but you're going down less or is it something else? And thinking about those, then, what makes a good offense depending on your objective function.
Soumaya: Yeah, so we had to really broaden the definition of winning to make this book work, right? I think if you are in the US actually it's complicated, right? And I think one of the big challenges has been that the US office... is often not clear about what it perceives winning to be, right?
If we're in a world where actually we're in a kind of power struggle with China, maybe the US winning is the US doing better than China, right? Which obviously to China doesn't look very nice, right? And, so there is this kind of question whether the US sees that relationship as zero-sum.
And I think in, in, in I... who knows what President Donald Trump thinks, but there are definitely some, important powerful people in America who think that it is zero-sum. In which case winning just means the US doing better than China. Now elsewhere in the world countries often don't have the agency, right?
They're not, they're just not as big and powerful as either of those two, so they don't really have the luxury of saying, "Oh would we like to be bigger, better than the US or China?" So yeah, we had to massively expand our definition of winning to make this book work.
I think in the US there is sometimes an ambiguity about what winning means, right? Is there a straightforwardly zero sum situation in which actually the US just needs to do better than China? Or is there scope for a positive sum interaction? Elsewhere in the world, obviously, smaller countries don't have the luxury of saying, "Oh would we like to be bigger and better than the US and China?"
It's not where they are. So I think for a country like, say, the UK, winning does mean doing as best as you can, minimizing your wounds in this era of economic conflict. Maintaining sovereignty to the extent that you can and being careful about what that really means.
Ben: Yes. What is a good attack then?
Soumaya: What is a good attack?
Ben: But obviously it slightly depends on what w- winning means for you, so that's why I interlinked them.
Soumaya: Yeah. So I think, having identified your strengths or vulnerabilities, and vulnerabilities, sorry you're gonna have to be much more muscular about intervening because the market isn't gonna, isn't gonna fix this, right? And the kind... the core economic problem that we identify and talk about is this case of market dominance, right?
So we've got China as this massive and growing manufacturing power that is incredibly dominant in certain areas and has shown a willingness to, to weaponize that dominance, rare earths being the classic example. And, you can think of this as a straightforward monopolist problem, right? Within domestic markets, we worry about individual companies having too much market power.
The parallel is, China as a government having too much market power and being willing to abuse that power. And so essentially what you need to do is intervene to break that concentrated power. But obviously it's challenging because the dominant actor isn't going to sit by and allow you to do that, right?
So they're going to, they're going to push back. And of course, the very factors that generated that dominance in the first place are gonna be the same ones that make it hard to escape from, right? So in manufacturing, you've just got massive economies of scale, huge costs to move, huge uncertainty.
The- Yeah. Skills of your labor force aren't necessarily easily found elsewhere. It's gonna, it's gonna take a lot of effort. It's gonna take a lot of work. Probably gonna have to do it in coordination. This is one of the big fights between me and Chad. Chad's answers to everything in the book was, "We should just work with our friends and hug each other and look at some rainbows."
And I was the cynic saying, "Oh, people are rubbish. It never happens." And, the disagreement wasn't actually that big. I think we both agreed that yes, it would be better if countries coordinated. I think I was just a bit more skeptical that it would actually happen and therefore skeptical of the utility of just that being the recommendation for everything.
That's all
Ben: that matters ... but
Soumaya: yeah, I think j- in certain areas acknowledging that, yeah, the market isn't gonna fix this, the government does need to intervene. But, try not to go too far because autarky is not great.
Ben: Yes. Okay. So an issue of market power.
So what should we think about tariffs? When they first came around there was a lot of chat about, even about tariffs and all of this, and perhaps a little bit of confusion in the ordinary person. But how should we think about what tariffs actually are and how they are used in the trade wars we are in at the moment?
Soumaya: Tariffs are a tax on imports. They are technically paid by the importer. Now economists know very well that who physically pays the tax isn't necessarily the party that actually bears the cost of the tax because prices might change. The evidence so far suggests that in the US when President Trump has applied his tariffs, most of the incidence was borne by the importer.
So it was Americans who were paying those higher prices. There are a couple ways to think about tariffs. You could think of them as a kind of Trumpian Making America Great Again tool to revive manufacturing, bully your allies, extract concessions, raise tax revenue, and all the good things.
Reduce your trade deficit being another one. And, I think the message there is to try not to be too confident in what these very blunt weapons can achieve. I think that the relationship between tariffs and trade deficits is one that has generated a lot of yelling over the past few years.
And maybe this is controversial, but I think there is plausibly a relationship between tariffs and trade deficits. The problem is it's just incredibly unreliable, right? So in an extreme situation, if you imposed a 1,000,000% tariff on everything, then yeah, you probably would affect the trade deficit, probably generate a recession as well.
But the problem is, there isn't a straightforward linear relationship there. And President Trump applied very broad tariffs last year and the trade deficit did, diddly squat overall. These tools aren't necessarily all they're cracked up to be.
So yeah, tariffs as a kind of a wonder weapon is not what I would describe them as. That said, I do think that blanket screeching against all tariffs at any time ever is probably not right either. You can think of tariffs as a demand signal, right?
You're sending companies a sign that actually if they source from one place rather than another, it's gonna be more expensive. And that could be a trigger to persuade them to start sourcing from somewhere else. Now they're very difficult to do correctly or, as, for many reasons I go into in the chapter.
But I guess the headline is y- there are cases where actually you might want to use them. You might need to use them. Maybe they're the least bad option of many bad options. Yeah.
Ben: So be very careful of your blunt tools and blunt weapons, of which broad tariffs are probably a blunt weapon and a little bit hard to know how to do that.
But there is a case for selective use of it. So again, the purest on the other side may not be exactly right. So I guess we had the tariff part, and then another part which comes up in the book, and we've mentioned it earlier, was industrial policy, which has risen up nation states' agendas as well.
And I guess it does depend on which nation state we're talking about, but how should we think about industrial policy? Maybe if we wanna talk about three big blobs, we've got the US, China as two major players, and then as we're sitting in the UK, we could think about UK industrial policy if maybe we're trying to advise a UK government or prime minister on it.
But what should we think about industrial policy? And if you wanna think about it across a couple of countries, that's one way of looking at it.
Soumaya: Yeah. So industrial policy is an incredibly broad term, right? In a sense it includes subsidies, tariffs, export restrictions, any kind of government intervention to distort or shape the structure of your economy, presumably to try to generate some industry.
In the chapter on subsidies, the big analogy that we use likens subsidies to cooking, right? And we talk about the lessons that you can draw from there. And the reason we do that is because Chad and I are both absolutely horrendous at cooking, right?
And so the best place to... the best way to come at this question is from a place of deep humility. I think I mentioned this in the book, but Chad once baked gluten-free apple pie, and it was truly the worst thing I've ever eaten in my life. Really astonishing. It can go really badly wrong.
No, I think from the US' perspective they have the fiscal firepower to, they can spend big. The Chinese obviously have been spending big. Now, the big danger of a big turn toward industrial policy is that you end up spending just because the other side is spending.
You're in a wasteful subsidy race. Who wins in a market has nothing to do with who's most innovative or providing the best service to consumers, but it's actually just 'cause, who's got the deepest pockets whose backer has the deepest pockets. And I think that, that kind of warning has been one of the, yeah, that, that's been pretty loud, I think, over the past few years. But I do think there's increasing recognition that in some areas where there are these big distortions, where you have essentially a kind of non-market player then you do need to erect some defenses. Just thinking about China and how it has these local government authorities that are providing cheap credit to Chinese companies that lack domestic demand and so end up exporting loads and loads to the rest of the world.
And then when prices collapse, in most other countries, those businesses would just fold. But in China they're not allowed to fold. The- those credit lifelines just carry on going, and so they're able to sustain their exports, right? So when you're confronted with that, when you're faced with that- It's, y- the purest response would be to say look these foreign governments, these Chinese, Chinese government, that they're subsidizing our consumers.
We should just welcome the cheap stuff. Sounds great." I think there's a recognition that we're not in that world anymore, and actually allowing the de-industrialization that comes when that happens across the board is not very wise strategically. And so you do need to act.
I think the big debate now is not whether there needs to be action, say, in Europe, it's how broad would it need to be and can we do anything without eliciting China's retaliation, right? Because the other difficulty is that obviously in the short term, the problem is there, right?
China has dominance, and so it's actually very challenging to introduce your industrial policy without the Chinese retaliating and triggering the very thing that you're trying to protect yourself against.
Ben: I can't believe you are a bad cook. You had such an entertaining conversation with Tyler Cowen about food and economics and things that I don't truly believe your recipes will be as bad as that.
But I can-
Soumaya: I can cook one thing. I make really good za'atar salmon. Okay ... and then it goes to salads. Yeah. And that's where I am.
Ben: Maybe that's the thing you need. As a country, you just need to be so super great at one thing-
Soumaya: Yeah ... that's enough. No, totally, right? Just be in Taiwan with TSMC.
Yeah. Great. Everyone- Be South Korea with memory. Brilliant ...
Ben: Everyone needs you. I have to say, I did probably lean slightly towards this older view initially that industrial policy doesn't seem to work, it's all really tricky, let's not think about it 'cause we had subsequent things. Which I would've said, I don't know, maybe 10 or 20 years ago was the sort of mainstream median econ view.
Soumaya: But can I just, can I just- Go on ... so industrial policy doesn't work is a familiar phrase to me. But I think what's happened is we've had to change our definition of working, right? Yeah. So in the past, industrial policy was supposed to generate profitable, self-sustaining, efficient industries, right?
Generate dynamic economic growth. And sadly that's not the only thing that we care about anymore, right? Because essentially China has been using industrial policy to get market share, right? And actually, if you think about whether industrial policy works, if you just wanna buy market share, it probably does a bet- have a better track record of doing that, Yes
than it does in getting efficient, self-sustaining-
Ben: Yeah. So I was gonna come to that. So-
Soumaya: Sorry, yeah ...
Ben: Yeah, that, the consensus maybe 10 or 20 years ago, or sometime in the past, and maybe I was slightly persuaded by those who talk about it, although I'm slightly suspicious about macro overall. We could talk about the macro and micro.
And then you see these individual companies sometimes do this. So you argue like a ride sharing company produces, gets all the market share, and then when they are dominant market share, they can then do something with their prices. So within that, as a long-term strategy, that's potentially it.
But then I looked at China and its healthcare policy, particularly around biopharmaceuticals, so this is biotech and the other, where initially we were like they're spending a lot of money. The state doesn't necessarily know how to choose. They're sponsoring a lot of these. You're gonna get a lum- lot of zombie biotechs."
Okay, fine. But you fast-forward to today, and there's now an acknowledgement that the Chinese biopharmaceutical sector is really quite great. They've got, Everything is coming out. Yes, a lot of them failed, but actually that's what biopharmaceuticals do. So by enabling them to survive for longer than maybe they would because of the fact that your probability of success is quite low, but you need lots of shots on goal, and that's been really intriguing.
It did also seem to go hand-in-hand, and this is why it's a little bit more complex than just saying, "Oh, it's a subsidy or it's a tariff," by also fostering essentially clusters. So this is the ag- agglomeration effect, but you have places in China where there's very good biotech and there's very good biopharmaceuticals, and that seems to have come around.
And you can argue about this with some of the supply chain into factories and the like. And if you then say what is this success?" You now actually do have thriving Chinese biopharmaceutical companies and also R&D tools, and they're arguably now from nowhere right up there as global. I don't know.
I don't know. We'll never know quite how much money they had to spend to get there, but it probably is a win for that sort of policy. So I have had to rethink about, yeah what does it mean? There's also a slight spillover because we'll get more medicines, which is probably net-net good for the world, but that's also a kind of other interesting thing.
But yeah, I don't think it's obvious which way industrial policy goes in terms of it, it can be bad and it can be good, and it depends. Do you think that we might see more of that, and that's essentially what's happening in the world when nation states think about this?
Soumaya: Yeah. I think one of the lessons from the Chinese model is that you need competition, right? I think in the past when industrial policy was just national champions picking winners, allowing big countries to get big behind protective walls, I think that that has been discarded.
I think one of the elements of China's model is that you see it with these state-owned enterprises. And I'm not... Biopharma is your area of expertise, not mine. But, one thing that you can see with, say 'cause the railways, SOEs you had competition within them.
So different branches would compete against each other for contracts. They were, looking at the two biggest ones, they weren't... One of them was never allowed to get too much bigger than the other, right? They were both supposed to be keeping the other honest essentially.
And I think what China's benefited from is obviously being massive, and so you can have an element of domestic competition. So yeah. So I think hopefully the rest of the world will take that lesson. And then, I th- and I completely agree, right? This kind of...
The sense that actually when it came to industrial policy, that the West was too intolerant of failure. That has been taken on board. In the US, Solyndra is the main example, right? So there are stories about when the Biden administration was setting up the CHIPS Act office to do industrial policy looking into chips and, there was this fear that, we can't do another Solyndra, right?
And that actually isn't particularly helpful because actually maybe you do need some failures because only If you've had the big failures, that shows that you're taking enough risk, right? If you can never have a big failure, then what's the point?
Ben: Yeah. What did you try to do?
I think you make a really good point about the competition. A- and potentially this is some parts where China learned in some of their sectors, because on biopharmaceuticals they didn't take big stakes and try to do state-owned backing. They seeded a lot of private companies- ... essentially, which were profit-seeking.
But they seeded much more and allowed them to live longer than probably would've been. So they were still competing, but they didn't really bet. They just said, "Okay, we're gonna essentially over-fund this," in terms of that. So that was maybe an interesting contrast. Maybe sticking on the same theme of health, I have here in my notes that you began working on health economics at the IFS a while back.
Do you still follow the health economics field, and are there any health questions that you think are of interest and importance to you now? Or is it all trade and health and all of these other things are not of so much interest to you at the moment?
Soumaya: Yeah. So that was a very long time ago.
Yeah. So that was when I was first at the IFS and was doing some academic research looking into the interaction between health and socioeconomic status, right? W- what causes what, right? And actually, funnily enough, just this week, so yesterday I published a column looking at the relationship between obesity and income.
So I actually very recently have returned to this in quite a r- relatively direct way. So my column this week was about gender and obesity because if you look at the data everyone knows that poorer people tend to be fatter than richer ones. But if you look at the data, that difference is only really there for women in the US.
The- there isn't really a gap between the poorest and richest third of men in terms of their obesity rates in the US which is really astonishing. I think, elsewhere in the world and, maybe depending on the data source there is still a gradient for men, but it's always much more extreme for women.
And that's super interesting because it suggests that it's not as straightforward as, oh poverty means that you have a worse diet and therefore you're obese, right? Because why would that only be the case for women and not men? And then, the reason why I wrote that study, that column was 'cause of a new study looking at the effect of GLP-1s on women's economic outcomes, right?
So if you think that weight is bad for their economic outcomes, does weight loss actually improve their economic outcomes? And there's this new paper by Rebecca Diamond, who's a professor at Harvard University, who finds these extraordinarily large effects on the employment rates of women who start off unemployed and start taking these drugs.
And also on single women and their chances of finding a partner, which then obviously massively increases their household income. So yes, my, my interest has been sustained. And yeah it's a weird thing. So I, So I used to write for The Economist, and I was the trade correspondent there, and that became my specialty, and I had a trade podcast called Trade Talks, and that was great.
And I k- I became known as this trade person. And then I basically stopped being a trade person for a few years, and then, went to the FT where I was really covering everything. I'm an economics columnist. The FT has some really excellent trade commentators. They don't need another one.
And then I had a baby in 2025, and a few weeks in was like, I need a project. What's my project?" And decided to return to trade. So it's a weird thing of specializing, de-specializing, re-specializing in trade, and then having this book out, so I have to take on my old trade person persona.
And now I'm gonna de-specialize again 'cause I'm not the trade columnist at the FT, right? I'm a general economics columnist. So actually... And honestly, I love trade, and I'll always be a trade person, and will always be a very important thing to me. But I need a break.
Yeah. I need a bit of a detox. You kind of- 'cause it's been a lot.
Ben: Some other things. Yeah.
Soumaya: Yeah. Some other things. GLPs- Yeah ... obesity, gender inequality. I'm gonna do wealth taxes next week. I'm gonna... all sorts of fun.
Ben: Great. Yeah, I thought the paper's really interesting and your column was really good.
I guess my slight suspicion, not suspicion, was that the effect sizes were really huge. Yeah. Having glimpsed that econ papers, it's oh, God that's too large. If you go give that to an average econ, it's really? We don't normally see things as large as that, but who knows?
Soumaya: GLP-1s they're pretty extraordinary, right?
Yeah, they do. And the history of them, the history... like a colleague was saying to me the other day, the history of GLP-1s is people are like, "Really? That much?" And then it replicates. The, I, I... but I do agree, right? They are very large effects. Some of these sample sizes are quite small.
And it's just very difficult to do. You can never be absolutely sure in this kind of setting that the only difference between the women who did and didn't take the GLPs is just the GLPs, right? There might be some other difference that the author wasn't able to control for. Okay. I think she's done as best as she can.
Yeah,
Ben: She tried really hard. It was a, it was a- Yeah ... good paper. And I think- ... even if I might go the effect size I feel directionally it feels quite robust to me and with all of these other things. You left on the column a slight note of oh, but what about the price of GLP-1s and is that gonna be an issue going forward?
But I have a silver lining for this because patents and because patents expire. It's already going generic in Brazil, India and Canada- this year, semaglutide is. And by the 2030s, this generation will be generic overall. So although we might have to wait 10 or so years, this is like the consumer surplus from biopharmaceuticals because patents are enormous.
I always joke that we're still seeing benefits from aspirin today, and that's 100 years, and we'll have 100... for as long as humans manage to keep going, we'll probably have benefits from aspirin. So that- Yeah ... so I feel very optimistic about GLP-1s, and we never... We have s- had some of these estimates about the cost of o- obesity 'cause of these second-order impacts, which I think are probably underestimated, which means that actually in, as we reverse them, we might see unexpected benefits going this other way, which again, because we weren't fully aware of them are potentially quite positive
Soumaya: Yeah, I'm very optimistic, right?
And sorry, just to for people watching this who haven't read the column, the point I'm making at the end is that there is concern obviously that right now only richer people have access to these drugs, right? And so there's this danger that if only the richer obese people can reduce their weight, maybe being obese becomes even more of a signal of socioeconomic status, and that could heighten discrimination against them.
But that was very ... I very much agree that is potentially only a temporary effect while these, the prices are high and while access is limited. I'm very hopeful that as they become more, more widely available, there could just be these amazing health effects. This is so anecdotal that I almost feel embarrassed to say it, but, knowing people who have taken these things, they're doing really good things for them.
Ben: Their life is better across all the domains you might expect. So-
Soumaya: Yeah. And I don't wanna say that, they're happier or more content in themselves 'cause I don't know. I wouldn't go that far.
I don't know. But they seem so much better within them- that's what... That's how they're presenting, right?
Ben: And that's what the survey data also suggests when people have been surveyed with this, with actually potentially a slightly stronger effect in women as well. Although- ... again, slightly hard to know, and also more women use the drug, so all of these types of things.
Yeah. I was gonna turn maybe for a moment to personal creativity and ask this as a two way. So one is I'm interested in your writing process. Are you taking notes longhand and then onto the computer? Are you a morning person, evening person, just have to get it out whenever you can, or do you have a particular process?
And then I also happen to know because I've heard your YouTube and in person that you're a great singer. So I was interested in how singing or other creativity kind of folds into your life and your work, and whether that's also a great outlet and comes in, and how you think about personal creativity.
Soumaya: Yeah. So my writing process is that I'm a very slow writer. I start off making notes, structured notes, and then I write out an outline kind of draft, and then I have to redraft after that. And then often there's a final redrafting process, and I need a night of sleep. Ideally, two nights of sleep in between the various different stages.
So extremely slow. People laugh at me for being that slow. The kind of more experienced columnists are like, "Really? It takes you three days to write your column?" "Yeah. Yeah. No, it takes... It's a full-time job." And then, I do the podcast and research the other days.
So yeah, very slow. I type, I don't write things out handwritten. I do really suffer from just distraction. I'm, like, I'm a human, and so I have a-- I use a software called Freedom that basically has to shut down various websites from my computer. I never, ever write with my phone in the room.
I find writing really hard, and my brain wants to do something that is not writing when I'm trying to make it write, and so I, I need to shut off all of the tempting things. And yeah, and then I just edit and edit. And, cut out the bad jokes and get it more concise and and then so on and so forth.
Ben: Do you edit as you go? Are you like, "Oh, I've written one or two sentences, I gotta go back and edit," or do you try and get it all out-
Soumaya: No ...
Ben: And then edit?
Soumaya: You gotta get it all out- Yeah ... 'cause you need that time to think, right? And to, 'cause often just the structure. So actually if you're getting hung up on two sentences, that's always a sign that you just need to finish the paragraph and move on.
And then the next time you come round to it, you'll see what you're trying to say. You need distance, I think to fix- Yeah ... those things. I agree ... I don't use any AI in writing. I'm actually a bit scared- But helpful in research? Yeah, so I use it. I use it as Google basically.
So I use it to find 10 links to studies exploring this issue. But as much as I can I try to avoid reading the words it gives me, partly because the hallucination rate is non-zero, and that's too high for me. But also I feel I'm a columnist. I have my voice and my personality and my, myself, and the more I read AI, the more I worry that I'm absorbing it- Yes
and I might end up sounding like it, and I d- I need to sound like me, basically.
Ben: Yeah. We read enough AI that it's already gonna seep into you.
Soumaya: Yeah. That's all, there's so much of it everywhere, and so I just really try to avoid it.
Ben: Yes. And we pay to hear your voice…
Soumaya: Yeah ... that's good. But there's a writer at The Economist who I know, who wrote a great piece about AI writing recently.
And then a different friend of mine wrote a newsletter that was making the point that people who are professional writers are very poorly placed to judge whether this is good. Whether AI writing is good. Because, I obviously just have such a strong incentive to hate it.
So I'm gonna, I'm gonna really struggle basically to- Yeah. You can't
Ben: take any other position really. That's-
Soumaya: Yeah ... yeah. And I have lots of reasons why I hate it, but actually I need to acknowledge that I'm not, I'm no neutral party to this. And, And
Ben: I also agree with switching off the internet.
So when I write plays for theatre … and I actually have no internet and no phone. If there's- No ... anything which can connect you, that spells doom. So that's- Yeah ... a way. And then is it the same for your singing practice, or is that completely different?
Soumaya: Yeah. So with the singing stuff, I have the opposite approach in that I'm very much like- I'll write the song and do it, and then it's out.
So I'm actually doing a thing where every week I'm writing a song based on one of my columns. And so I wrote a song this morning. They're songlets, they're not full songs. But there I write the words really quickly, write the chorus, write the melody, and just go. Because I don't...
I don't know. It feels weird with my songwriting, I often, with some exceptions, I often don't feel that they're really improved by agonizing over them. I m- I'm just much less patient when it comes to, when it comes to songwriting. So yeah, they're very f- they're very fast. I'm almost doing it for me.
But also it is my vague attempt to AI-proof myself, right? There's not that many economics columnists who are also singer-songwriters. That's exactly
Ben: it ... so I'm trying to- Cross two- Trying to occupy that niche ... primary domains and you become unique. That is- Yeah ... that is definitely the way.
Maybe I'll try and tempt you at the end in a few minutes to sing a couple of lines of song. I don't know- Oh, yeah ... how we'll feel about that, so we'll do that.
Soumaya: Absolutely.
Ben: Okay. Let's do a little quick round of overrated, underrated, and then we'll do current and future projects, and then a song, and any advice.
Great. So these are quick hits. You can pass, neutral, or agree or not. But- Yeah ... overrated, underrated: universal basic income, UBI.
Soumaya: Ooh.
Ben: I didn't know it's gonna be that tricky, so there'll obviously be that- It's
Soumaya: tricky 'cause by who?
Ben: Oh,
Soumaya: Okay. Some people massively overrate it. Yeah. Okay, I'll go with overrated.
Ben: Overrated. You're already thinking second and third order then. It's by who? Overrated because it doesn't do as much as the advocates would think.
Soumaya: I just, the empirical evidence we have so far has very much been a disappointment, I think, relative to what the most, the strongest ad- advocates were saying. I think it is plausible that there is some future world in which actually it might be necessary- Yeah ... but we're quite far away from that right now.
Ben: Very fair. Fears of falling fertility rates. You did something on fertility rates recently, but- ... fears of falling fertility rates, underrated-
Soumaya: Really underrated ... overrated? Yeah.
Ben: Underrated. It's happening faster than maybe we initially suspected.
Soumaya: Yeah. I... Yeah, and I'm very much like a... Obviously, if women, if people don't wanna have kids, great.
That's on you. But I think the thing I'm worried about is people who's, are saying they wanna have kids and who d- who don't who y- there seems to be a gap between- Yeah ... aspiration and reality, and that, that's the thing that worries me Yes ... with all the, associated domino effects.
Ben: The silver lining is that we might have overestimated some of the climate impact because it relied on people. But the bad thing to that is basically we then have overestimated long run growth, which needs people. So actually yeah, it probably means it is more of a problem than- ... than people think ...
Soumaya: AI
Ben: will
Soumaya: generate the singularity, and we won't need people anymore.
We'll just be... We'll Just- ... veg out playing video games or something.
Ben: All great. So underrated, overrated carbon taxes.
Soumaya: Oh probably, I'm a card-carrying economist, I'll say underrated.
Ben: Underrated 'cause that's the way of solving it. Do you think net zero itself is an underrated, overrated thing?
Soumaya: Temperatures are really high. I... Let's go with underrated
Ben: Underrated, yes. So underrated both carbon taxes and that, and we should bring them in and solve it. I th- it's a political economy problem, isn't it? Rather than a financial economy problem because, yeah, the econ people just say carbon taxes will solve it, and then no one seems to be able to implement them, so tricky.
Soumaya: Yeah. No, ev- we just need to do what we can, so they're underrated by the population at large- Yeah ... therefore- But, yeah ... they're difficult
Ben: to implement. Yeah. Not by economists. AI sovereignty, underrated, overrated?
Soumaya: That's a big question. Probably underrated, right? But you can get sovereignty in many different ways.
Ben: Yeah. And it's becoming potentially more important, although hard to know exactly.
Soumaya: Yeah. I think the debate is between people who are like, "Just give up," right? "It's too late" versus people who say, who think that there is, there's still something we can do.
And I'm a sappy optimist, but I'm not ready to give up quite yet.
Ben: Very fair. And last one on this British meal deal. Do you think the meal deal is underrated or overrated?
Soumaya: Yeah, overrated 'cause they're ultimately often very unhealthy, right? It's like ultra-processed bread plus not enough filling.
Crisps, really bad for you. If the meal deal was, like, a nice salad that was balanced- Salad, yeah ... and, was good for your biome, then I would say great. But meal deal is an odd collection of food with often a sugary drink, also bad for you.
Ben: Yeah. So
Soumaya: calories- I
Ben: mean, if you wanna-
Soumaya: are not good ... if you wanna consume these things then go for it. But I'm, I will not be celebrating them.
Ben: Yeah. And I hadn't realized I feel really late to this, that sandwich makers do this thing where they push the filling into the middle and bulk it up so it looks better. But when you actually open it- You mean push the filling
Soumaya: on the edges
Ben: so that it looks- you, yeah, you sort of shape it- Oh, sorry.
Yeah ... so it looks- yeah ... bigger than it is. And then when you cut it in half And when you open it up- ... it's actually not that much. Yeah. And I felt really tricked. They said, "Oh, yeah, they've been doing this trick for ages," and that, and so that's definitely a meal deal thing.
I guess on this externality thing then a sh- the sugar tax type of thing, are you, do you think that's underrated or overrated?
Soumaya: Oh, yeah, I'm up for sugar tax.
Ben: Sugar tax, yeah. So tax all externalities, right? Straight, straight econ.
Soumaya: Yeah. Is that an external- Yeah. I mean- Ish ... it's ish, sugar's bad for you- Sugar's bad ... if you're addicted. I'm addicted to sugar. Good for you. I would... I think there's a question of whether sugar taxes are enough, right? Yeah. Or even the best way, but- whether you just need regulation to force companies to put less sugar in their stuff.
Yeah. And then there's also, the... I think the bigger question is, I think on some metrics some of the kinds of sweeteners, some of the replacements for sugar that we have probably are also pretty bad for you. And we- we're now straying into kind of Sumaya's unevidenced-
food hunches
Ben: So probably- Anecdotal ... should stop. Yeah. Yeah. The strongest data. Great. Okay. Current and future projects. So we're gonna leave trade behind us slightly because you're interested in so many other things, but yeah. Is there another book for you? Oh, I should have asked also what it's like to be a co-writer versus, versus a not, 'cause that's probably a different sort of experience.
But i- in any event, current projects, future projects, will you do writing together projects or your own, or what's up?
Soumaya: I'm... this experience is still fresh enough that I'm still solidly in the, "I'm never doing this again."
Ben: But is it like running a marathon? So all of these marathon runners, they run it and then they go, "Oh, I'm not doing that again."
Yeah. And then next year they're running it again. This is what seems to happen.
Soumaya: I ran a half marathon and then- And did you say never again? ... I was like, "Never again." Yeah. And I haven't run another half marathon. Okay. So I'm... So this is quite good. Yeah, no, I don't think that- I d- I don't want to do it again.
Don't make me do it again.
Ben: Okay.
Soumaya: There were many things that were really miserable about the experience, right? The main one was the s- the time pressure, right? When I delivered, I wrote and delivered the book incredibly quickly. I still really wish I'd had another round of editing, right?
There are some jokes that should've just been cut, and had I had another round, had I had three and a half more weeks, I would've just, everything would've been better. But that isn't the time that I had. So there's that. And so Chad is, he's my best friend.
He's the godfather of my son. We're still, we're still very close. That's all great. We're not gonna be writing another book together. He's been absolutely brilliant particularly- But not in a fight, yeah ... going around and talking about the book and telling everyone about the book.
It's been really great. And yeah, I think my, I think I want to make stuff and I want to write stuff and produce content that people really wanna really wanna read. And, maybe in 10 years I'll decide that the best format for that is a book. I think for now I have the best job in the world, right?
I write a weekly column for the Financial Times, and I have a weekly podcast, right? It's amazing. And I can write these songs. So I'm gonna be... I just wanna, I wanna do the singing, I wanna develop that. One thing I'm actually working on, one of my- my next kind of big project is I'm gonna try and develop a kind of set.
So I'm gonna try and so I'm hoping, and this isn't formally announced yet, I'm only discussing it, but there's a festival where I'm gonna speak. And the idea is that I'll talk about my columns and also sing songs to go with them. So it'll be a one-woman show combining economics and singing.
So if I can make that really good, that would be pretty great. That'd
Ben: Be excellent ... so that's
Soumaya: my next project. I've
Ben: seen some econ in standup, which can work as well. So more columns-
Soumaya: Yeah, I don't think I'm gonna- ... more podcasts,
Ben: more singing ...
Soumaya: I think I'm done trying to be funny-
for a little while, right? I think- You are funny ... I think with the book- ... there are lots of jokes in the book and some of them are good. But I think I'm, I think I'm done with that. So just straight singing columns, and if people like it, that's great.
Ben: Great. So do you want to tempt us with a little, a few lines from a song?
We could either... I heard your- Yeah ... trade war rules one, which was great, or you could preview something from your fledgling set.
Soumaya: I'll do my trade one.
Ben: Let's do the trade one.
Do you dream about them? No. Do you dream some songs and have to write them
Soumaya: out, or you don't? I don't really dream. Yeah, okay. I just have recurring nightmares about-
having an exam I haven't revised for. My dreams are really tragically obvious. Okay. I could just sing a random song. It doesn't need to be about economics- No, let's do- No, I should probably do one about economics ...
Ben: your choice.
Soumaya: I'll do an economics song. Okay
All so how do you win a trade war? By blowing up the world.
By terrifying with tariffizing. By stopping what you sell. Oh, why not try it? Battle cry singing hi, oh hi, away
Why not try this battle cry singing hi oh hey
So how do you win a trade war? By engaging with the fight. Don't pretend that we're all friends. Subsidize and strategize. Look at your ties. Work with allies. Economize as you build supplies. You've got to do the work. Oh,
Why not try this battle cry? Singing hi, oh hi, oh hey.
'Cause we're all trade warriors now, whether we like it or not. Why not try this battle cry? Singing hi, oh hi, oh hey. Singing hi, oh hi, oh hey
Ben: Ooh. Ooh, that was excellent ... That's what we wanna hear on Radio 4 Thank you. …that's amazing ...
Soumaya: yeah. I've- Yeah ... I've told Radio 4 that I'm available, but they haven't- Yeah.
They've apparently they're a serious radio program- Yeah ... so
Ben: it's
Soumaya: Fine.
Ben: I'll definitely come to your set. Great. So last question: do you have any advice or thoughts for listeners? This could be life advice as to how to combine writing econ, singing, or any observations you've had on writing your book, or anything you'd like to share with listeners.
Soumaya: Don't write a book while on maternity leave. Advice number one.
Ben: That seems to be excellent advice.
Soumaya: Don't over-commit yourself. I feel so over-committed right now. Do less but better-
Ben: Yes ...
Soumaya: This is my advice.
Ben: Don't over-commit. So I think that's also learning the power of saying no.
Soumaya: Yeah.
Yeah, I guess so.
Ben: 'Cause you said yes slightly too many times.
Soumaya: Yeah.
Ben: Yeah. Great. With that, thank you very much. I will flag How to Win a Trade War. Go out and buy it. With that, thank you very much.
Soumaya: Thank you. This was fun.