Transcript

Rob’s intro [00:00:00]

Rob Wiblin: Hi listeners, this is The 80,000 Hours Podcast, where we have unusually in-depth conversations about the world’s most pressing problems, what you can do to solve them, and what to do if you’re handcuffed to someone on the side of a cliff. I’m Rob Wiblin, Head of Research at 80,000 Hours.

Recent events have had people thinking a lot about interstate conflict, which made it a no-brainer to interview economist Chris Blattman about his new book Why We Fight.

I’ve followed Chris on his blog for many years, and I imagine many of you listeners have read some of his work before. Why We Fight gives a lovely overview of the theory of when and why people choose violence, which is a super important topic we’ve never covered on the show before.

Before that just two quick notices:

First, at 80000hours.org/latest you’ll find a new very informative article titled “My experience with imposter syndrome — and how to (partly) overcome it” written by previous guest of the show Luisa Rodriguez. If that interests you I recommend checking it out.

We’re also hiring a new full-time staff writer to publish well-researched articles that help people use their careers to contribute to solving the world’s most pressing problems. You’d work with Luisa Rodriguez and previous host of the show Arden Koehler, among others. The starting salary for someone with one year of relevant experience would be approximately £61,000 per year. Applications for that close in a few weeks on May 16. You can learn more about the role and apply at 80000hours.org/latest.

Alright, without further ado, I bring you Chris Blattman.

The interview begins [00:01:43]

Rob Wiblin: Today I’m speaking with Christopher Blattman, the Ramalee E. Pearson Professor of Global Conflict Studies at the University of Chicago. Chris is an economist and political scientist who has spent years studying the causes of violence in contexts as diverse as a civil war in northern Uganda, generalized insecurity in post-civil war Liberia, and gang violence in the city of Medellín in Colombia. That led him to write his new book, Why We Fight: The Roots of War and the Paths to Peace, which is going to be the focus of today’s conversation.

Rob Wiblin: We probably won’t get to all of this today, but listeners may also be interested to know that Chris has gone out and coordinated randomized trials looking at the impact of cash transfers, which he is an advocate of. And he’s also studied the impact of so-called sweatshop jobs in Ethiopia, which he became more pessimistic about after conducting a trial to see how they really affected workers’ income and health. Thanks so much for coming on the podcast, Chris.

Chris Blattman: Thanks for having me.

Rob Wiblin: I hope to talk about what causes intergroup violence and how circumstances can be shifted to make the turn to violence less appealing. But first, what are you working on at the moment, and why do you think it’s important?

Chris Blattman: I’ve just started dipping my toe into working in Mexico. One reason is there’s a lot of violence there. I don’t think it’s quite studied enough, and it’s obviously super important: it might be one of the biggest national security issues for the US, and there’s a lot of people harmed. But mostly, from all my international peacebuilding work or research, I just see a lot of techniques that are used — like mediation between armed groups — that just never get deployed, it seems, in US cities. It happens very little in Chicago. It’s kind of a natural thing to try to mediate between warring factions. We don’t do it.

Chris Blattman: Some of that does happen in Mexico. That’s a good example of ways in which I think Mexico is maybe at the frontier of incorporating a lot of these lessons from international peacebuilding into countering violence. And we could learn from that and maybe see that spread.

Rob Wiblin: Is there a reason why this isn’t used in US cities? Is this kind of the classic thing that the US tends to be unusually reluctant to learn from studies overseas?

Chris Blattman: It’s not just the US; it’s in other parts of the Americas as well. It’s done a bit in Colombia, in Medellín where I work. But I think one reason is we just have a blind spot to some degree. There’s an instinct that this is just a more individualized problem, and that policing is the solution. In part, policing is essential, but we can also do other things. I think the more legitimate thing that governments have to take a lot of care with is they have to be careful not to legitimize these groups too much. So I think what you actually need are more civil society actors to step up and try to mediate.

Chris Blattman: And that happens a little bit. I’ve followed a few people who do this in Chicago. But there are very few; they’re hard to find. It’s not very systematic, it’s ad hoc, they do it in their spare time. They could be an ex-gang leader, they could be a social worker, they could be a minister. And they don’t have any particular skill or experience other than whatever they’ve acquired. They haven’t had a chance to learn or share experiences with many other people. So it’s very ad hoc at the moment. I think that should probably change.

What people get wrong about violence [00:04:40]

Rob Wiblin: I think we’ll get to some of the stories about mediated peace between criminal groups later on, but let’s dive right into your book, Why We Fight. In that book — which is probably going to be out by the time this interview airs — you aim to get people to think a bit more clearly about why it is that violence sometimes occurs and how it can be prevented. I think you really succeed on that count. In a sense what you’re saying is really straightforward and just fundamental, but it’s wisdom and modeling that I think is often really missing from discussions of violence and would add a lot of value.

Rob Wiblin: I should note for listeners at the outset that one thing you’re clear that you’re not doing is setting out to analyze individual or heat-of-the-moment violence in the book, because the causes of that are a bit more idiosyncratic and personal, and probably better left to psychologists than political scientists, potentially. But that is fine with me, because my personal interests here are primarily about how to prevent major wars between states, because they’re most likely to be the risk factor for the collapse of civilization or human extinction.

Rob Wiblin: With that out of the way, first off, a lot of people have the perception that humanity is beset by violence, and if you watch the news it might indeed seem that way. But you argue that in fact, doing violence to others absorbs very little direct human effort. Is there a way of quantifying what fraction of all human resources are directed towards physical violence?

Chris Blattman: That’s a good question. Not that I know of, is the short answer. Related to this idea that actually war’s kind of rare, or the exception rather than the rule, is that we tend to overestimate how much conflict there is. And also how much effort is devoted to nonviolent negotiation, bargaining, posturing, all these sorts of things that we do. So we pay attention to the fact that Russia has invaded Ukraine — which we should, just like a doctor should pay attention to the very sick people — but then we sort of overlook the fact that there wasn’t a war for them to almost subsume or cow Belarus, or send peacekeepers into Kazakhstan or Georgia. The list goes on and on and on. So most of the time, these things are settled acrimoniously but without violence.

Rob Wiblin: Yeah. I was thinking about how to answer this question. I think something like 3% or 4% of global GDP is spent on militaries directly. The interesting thing is that we spend all of this money actually training soldiers and buying all of this equipment and so on. And yet 99% of the time, the soldiers are just sitting there, not engaged in any actual war at all.

Chris Blattman: Right.

Rob Wiblin: So the thing that we’re not saying is that people won’t buy armaments. The thing that your modeling implies is that they’ll rarely actually be used.

Chris Blattman: This is I think one of the big insights just from strategic analysis or game theory in general, which is that most of the time fighting doesn’t happen — because it’s really inefficient, and there’s a better way to obtain that. But the big inefficient thing that does happen is arming: that unless we can find a way to really solidly commit one another to disarm, we are just going to invest an enormous amount of social resources in that. And that’s really hard to avoid. I think that when we do have states, when we do have alliances, that partly solves the problem.

Chris Blattman: So you think about the Western Hemisphere: we have the United States, we have all of these countries in the Western Hemisphere, which are all sort of in an alliance with the United States. The US is like the hegemon, and it helps keep the peace. As a result, none of these nations really need to arm against one another. So that hegemonic power and that set of agreements and also their collective willingness to act means that we don’t have to worry about going to war against one another. And so everybody can disarm against those, but then we still have to arm against the possibility of an invasion from some other hegemonic alliance led by Russia or China. So we do have solutions to this, but they’re still partial solutions.

Rob Wiblin: What’s the basic model that people should always keep in mind for why it is that violence should be expected to be really rare?

Chris Blattman: The book is really my attempt to distill insights of my own, but mostly these insights from decades of researchers in psychology and game theory. The core one starts with people like Thomas Schelling and James Fearon. And a whole host of economists and legal scholars as well, who studied strikes and courts — because we’re all talking about the same thing, which is: there are things like warring, court battles, and labor strikes that are inherently costly, that we can negotiate to avoid. So we can either bargain or fight. And war’s just politics by other means, in the words of many, many famous generals, including von Clausewitz.

Chris Blattman: So that’s what we have to remember: that this is a choice, and one of them is incredibly more costly than the other. We can almost always find a negotiated settlement, whether we’re officially negotiating or we’re just posturing from afar. We can always find something that makes us better off than fighting and flipping a coin to see who gets what we’re fighting over. So that’s the starting point. And so the cost of war is the gravitational pull towards peace. It keeps us in that orbit.

Rob Wiblin: Right. Applying this to the current case that is going to be on people’s minds a lot, the Russian Ukraine invasion: it’s been so costly for Russia and it’s been so costly for Ukraine, that it is remarkable that they couldn’t reach a negotiated settlement that would be better for both parties than the absolute shitshow that they’re living through.

Chris Blattman: I mean, that has to be our first instinct: that it is remarkable. Because it shouldn’t happen. And I predicted it wouldn’t happen, only because I think that’s always the best bet. I was wrong. And so in every one of these cases, we then have to say, “OK, well then some other thing must have had such an incredible pull that it overcame those costs, and yanked us out of that peaceful orbit.” And then the search for those causes begins.

Rob Wiblin: Right. So you offer five different explanations for how it is that, despite the enormous costs that are involved in violence, nonetheless states can end up using it, and also smaller groups can end up using violence. I guess if I had to really distill down why is it that violence is rarely used, I would say it’s because when there’s an imbalance of power, the weaker party basically gives up. They don’t surrender completely, they don’t give up everything, but they give up most of what the more powerful party wants. They appease them, effectively. And so they don’t lose a fight, but they’ve lost part of what they wanted because the other side was just more powerful than them.

Chris Blattman: Right. We all get what we deserve, where “deserve” is not maybe some abstract notion of equality and justice in this — I mean, it would be nice — but where what we deserve is what we have the military or other power to demand. And if we’re weak, if we can’t threaten to burn the house down, well, we’re not going to get the house.

Rob Wiblin: Yeah. People often object to making concessions to bullies who want unreasonable things on the basis that it’s appeasement and that’s potentially really bad. And the thing I just always want to remind people of is that appeasement makes the world go round, or appeasement is the thing that has prevented war from being a more constant feature of human life from the very beginning.

Chris Blattman: Right. “Appeasement” is what we call compromise when we find the compromise repugnant. It’s just a synonym.

Medellín gangs [00:11:48]

Rob Wiblin: Yeah. You have this great story of compromise that prevented a conflict that could have broken out between gangs on the streets of this city, Medellín in Colombia. Can you briefly describe how that situation played out?

Chris Blattman: Sure. So we’d been interviewing criminal leaders in prison and increasingly out of prison to understand how this whole apparatus works. And one of the people we interviewed in Bellavista prison was telling us a story about what I later liked to call the Billiards War, because it begins over a game of billiards. In his cell block, there were two rival factions that were gangs on the street, and a lot of their membership was on the cell block. And they were in the game room, they were playing billiards, and he doesn’t remember what happened — maybe it was just a petty dispute or something. One side pulls out their guns and shoots on the other — and why they have guns is a whole other podcast — but when the dust settles, miraculously, no one’s killed.

Chris Blattman: Of course, immediately the other side prepares for retaliatory attacks, everybody arms. In many gangs, there’s a system of alliances. They all activate their alliances. Everybody starts to arm. Is this the next big war in Medellín? Because there have been big gang wars two or three times in the past. And there is a cycle of retaliatory killings and things start to escalate. And then it doesn’t. The Billiards War just never happens, mainly because it’s not in their interests.

Chris Blattman: Also because there are these higher powers in Medellín. And they themselves have a negotiating table, a little bit like the UN Security Council. It’s not that effective, not that equal, but it’ll kind of work some of the time. And they sit these two gangs down, and are like, “No, we’re all going to lose too much if you guys fight, and we’re not going to sell our drugs, and you’re going to get us embroiled in a war. So we’re going to disincentivize you. We’re going to make you pay attention to the costs of war that you will inflict on everyone.”

Chris Blattman: And versions of that have happened many, many times. That’s one reason why, despite being this valley filled with hotheaded young men and lots and lots of guns and armed groups, it has a homicide rate half that of Chicago.

Rob Wiblin: Yeah. An interesting thing is that there’s tiny gangs that occupy individual blocks, and then those are aggregated into these larger gangs that oversee a whole lot of block-level organizations.

Chris Blattman: It’s a super common political form — both for gangs, but also internationally. Remember I described America the hegemon, keeping the peace with its own countries in the Western Hemisphere so that they don’t have to fight with one another. It’s kind of the same system of these whatever, that political scientists would call “hegemonic alliances.” The big guy keeps the peace within their neighboring gangs, and then it’s an easier negotiation problem between two hegemons than between —

Rob Wiblin: 100 small —

Chris Blattman: — 300, 400. Yeah, exactly.

Rob Wiblin: And so the leaders of these higher-level gangs are all together in a cell block, or they’re placed together in a cell block because they’re all in prison.

Chris Blattman: There’s many cell blocks. There’s a leaders’ wing in every prison. So there’s many prisons around, and all of them have a leaders’ wing, so many of them are together. They’re spread out across many leaders’ wings.

Chris Blattman: What’s interesting — we’ve only gotten these accounts from the criminals’ side; we have yet to get someone from the government on the record — but there was a brewing conflict at one point where they were going to war, again. This is actually much more recent than the Billiards War that didn’t happen. The homicide rate had tripled in the city because these factions were actually starting to squabble and skirmish, and maybe it was going to turn into all-out war.

Chris Blattman: And then all of a sudden, one day all of these leaders in all these prisons get transferred to new prisons. They’re just going to move people around. But in order to do that, everybody has to go to the same cell block and actually be in the same place for three days before they all get moved around. And somebody gets arrested, who’s a well-known mediator on the criminal side; he’s a crook. And miraculously, he happens to land in the same cell block, right? So it’s all one grand coincidence. A week later, the homicide rate has gone back down to a third of what it was, back down to its normal level. So the threat of state repression in prison can actually be, and often is, a mechanism for peace building, but also criminal control in a lot of cities in the Americas.

Rob Wiblin: So basically, I imagine that the thing that motivated them all to have these extended discussions in the prison and come up with some sort of peace arrangement is that it would be atrocious for business to go to war. The government will crack down on them. They can’t sell as many drugs when people won’t come and visit them, because they’re worried about getting shot. It’s very financially costly to them, and ultimately this organized crime is a business for them.

Chris Blattman: Right. And a lot of the leaders are in prison, and the government can threaten to make their lives easier or harder.

Rob Wiblin: Yeah. So there is kind of a hegemon?

Chris Blattman: Yeah. I mean, everybody’s trying. In Medellín, this whole system works it much more peacefully than in most big cities. And the state has managed to coordinate in this way without, I think, being co-opted or doing anything that I think is really distasteful. I think they really managed to walk this line incredibly well, of shrinking and reducing the strength of the groups as much as they can, but still acknowledging that they’re pretty powerful. And realizing that they have to negotiate in a way, but negotiating in a really one-sided ultimatum-like way that isn’t legitimizing them as political actors. It’s a really complicated problem to solve if you’re a government, and Colombia’s done it pretty well.

Rob Wiblin: Yeah. I don’t know how these gang leaders get on. They might have some affection for one another in a sense, because they’re in the same industry, but there might also be a lot of bad blood from past events that cause them to personally dislike one another. Obviously, me and my rivals don’t generally engage in shooting wars on the streets, but I think all of us are subject to a similar set of incentives when it comes to interpersonal conflicts in social situations, or when it comes to people who we don’t like in a professional environment.

Rob Wiblin: So I think the closest equivalent to war in my life would be disputes and spats in public, on Twitter and things like that, where there’s people who don’t like me and there’s people who I don’t respect — who I think are really wrong, or maybe I think their projects are bad, and on some level I would like to criticize them.

Rob Wiblin: But if we get into a shooting war, where we’re both talking in public about how much we detest the other one and why, of course it’s a cycle of retaliation in which both of us incredibly lose in terms of reputation. And at the end of the day, what really is likely to be gained? It’s not obvious which party is going to win the PR battle on this, going in to start. And so even people who really detest one another tend to just loathe one another in peace rather than engage in public fights.

Chris Blattman: Right. I mean, if someone insults you on Twitter, you no longer challenge them to a duel to the death, right? So that’s eroded for a good reason, because that’s not really optimal. Now, those used to happen, and so that gets us down the road of why violence does happen. But it’s kind of obvious; most people look at dueling and say, “Yeah, that’s not a great decision.” And they’re right. They’re right. It’s the same logic.

Rob Wiblin: Yeah. To be honest, when people insult me on Twitter these days, usually 90, 95% of the time I have the self-control to not reply because it seems net negative.

Chris Blattman: Right. Absolutely.

Rob Wiblin: I guess that’s the culture of dignity that we have. Can you lay out the analogy between war and human health that you have in the book?

Chris Blattman: So I think of it in a couple ways. One is that we all know that most people are healthy, and we would want doctors to get trained in the healthy functioning of the human body, and by virtue of being surrounded by sick people all the time, not forget that healthy people exist. And unfortunately we’re in a world of war doctors, right? People are trying to figure out and operate in these conflicts, and somehow forget that there’s this normal healthy state, and that’s the default and that actually the disease is rare. So that’s a first mistake we have to get past — that many do, but not all.

Chris Blattman: But then once we’re there, the funny thing about a lot of violence reduction — I see this in the city of Chicago, I see this in west African civil wars — there’s often a whole bunch of one-size-fits-all solutions, right? “Oh, this city used violence interrupters, so we must need violence interrupters.” Or, “This study used this new hot spots policing, or focused deterrence, or some policing tactic, and that worked there, so we need to import that.” You see that a lot in new cities.

Chris Blattman: The other thing you see is there’s a package: there’s this humanitarian peacekeeping complex in Africa that says, “First we need mediators. Then we need a demobilization and disarmament and reintegration program, DDR. Now we need a peacekeeping mission. Now we need a truth and reconciliation commission. And now we need elections within two years of the conflict.” There’s like this package, and you’re going to get it whether you like it or not.

Chris Blattman: If our doctor did that, if we just showed up at the doctor and said, “I’m kind of sick,” and he’s like, “You know what you need? Radiation therapy.” We’d be like, “Whoa, whoa, whoa. I didn’t even tell you what was wrong with me. You didn’t even spend any time thinking about it. What is going on? There’s this thing called diagnosis.” But of course it happens, because how do you solve the disease if you don’t have the diagnosis? And somehow we forget that. We just think, “Oh, it’s all violence. We need this again, just like South Africa had a truth and reconciliation commission. We need a truth and reconciliation commission.” It’s bonkers, but it’s super common.

Rob Wiblin: I think another analogy to health that I find that I have to constantly remind myself not to miss, is that death is rarely caused by a single malady, a single problem. Someone who’s 85 dies of pneumonia nominally. But in fact, they’ve died of so many different things that weakened the system until it was at a point that a viral infection could actually kill them. And so it is with conflict situations, where there was the fight in the bar that started the war. But actually that’s just the last straw. And what’s really going on is, why was the patient so incredibly sick that that could spread?

Chris Blattman: Exactly. I think I used the example of the First World War in my book, where I talk about how most people think the assassination of Archduke Franz Ferdinand, plus the miscalculations of Europe’s mediocre leaders, slept-walked the whole continent and the world into war. And that’s true in the same sense that pneumonia killed an 85-year-old person. But the only reason that person was so vulnerable — and the only reason we’re so vulnerable to some wacky little idiosyncratic situation or leader miscalculation — was because all the fundamentals were so fragile. But those fundamentals are less apparent.

Rob Wiblin: Yeah.

Chris Blattman: I think at least less readily apparent. You know, it takes ages for people to figure these out after the fact, and it’s often hard to see in real time. I have the luxury of summarizing thousands of people who have thought deeply about it for a century. So of course that’s going to sound wise, but that’s sort of like the slow lesson that I think a whole legion of social scientists have learned.

Rob Wiblin: So in the World War I case, where the temptation is to say, “World War I was, on some level, caused by an assassination.” The way that we instinctively check that is to think, “Well, what fraction of wars were preceded by an assassination?” And that is moderately common. But we also need to think, “What fraction of assassinations are followed by a war?” And that is actually quite uncommon. We need to look for dis-confirming cases as well. And that is something that I, even knowing this for many years, find so counterintuitive.

Chris Blattman: There were several Balkan Wars just in the few years before World War I that did not trigger World War I. There were innumerable crises that were very similar in the 20 years previous that didn’t cause World War I then. So something else was going on.

Overrated causes of violence [00:23:53]

Rob Wiblin: Yeah. What are some things that people think lead to violence, or are underlying causes of violence and wars, that you are more skeptical of?

Chris Blattman: Right away you talked about how I’m not talking about interpersonal violence — even though actually that’s my day job, a lot of it is actually studying and developing programs to counter interpersonal violence. I do think human emotions and passions in these hot, reactive moments that do explain a lot of violence are really central to understanding a lot of interpersonal violence. And not just that: my relationship with my nine-year-old son, and whether we argue with our boss or our wife or whatever is really important.

Chris Blattman: And I think it’s not that it’s irrelevant when it comes to big groups; I just think of all of our human errors and foibles and frailties, it’s probably the least of them. Because it gets mediated by big bureaucracies. It fades over time. Maybe some monarch getting angry in July or August 1914 helped contribute to World War I. But by 1917, presumably hotheaded reactive violence and passions were not what was going on. So that’s a good example of one that I think is maybe overstated. We tend to take our interpersonal violence and then we project that onto nations, and I don’t think that’s a good guide.

Rob Wiblin: Yeah. When I read that part of the book, I was trying to think about why it would be that people are biased towards thinking about these kind of personal emotions as a driver of large-scale conflict, and two things jumped to mind. One is that, the thing that we have experience of all the time is interpersonal interactions — like our interactions with our spouse, our interactions with our boss — where these personal feelings are more key. And so we are building this analogy up to a much bigger scale.

Chris Blattman: And that is to me, a version of the availability bias: we take what’s available, and we think that’s more common. And also our tendency is what I call projection bias: we then project that onto other situations, maybe erroneously. So we actually form and hold strenuously onto erroneous beliefs, despite contrary evidence.

Chris Blattman: And that is the misperception, not the passions: it’s our ability to draw the wrong conclusion from incomplete information and then project it to other people and situations. That is the misperception — that’s one of these causes of war. It’s these biases to have persistently erroneous beliefs that really matter in group behavior and individual behavior, and the passions fade away. So I want to push us towards focusing on the mistakes that bureaucracies and military generals and accountable presidents make all the time, not the ones that get filtered out.

Rob Wiblin: I think the other big contributor for why it’s natural to reach for these personal explanations is that it’s something that you can do without having a lot of very specific detailed knowledge about the strategic situation. So in the case of Russia-Ukraine, in order to think about this in a systematic or strategic way, you would probably want to know: What is the Donbas? Where is the line between Ukraine and Russian control? The Minsk Accords, what was in them, and to what degree were they agreed? What is Zelenskyy’s policy towards Russia? All things that I don’t really know that much about now, and I certainly didn’t know much about in January. But without knowing any context about this specific case, I can always say, “Oh, it’s because of Putin’s personality.” That’s an easy thing to say.

Chris Blattman: Yeah. And it’s easy to understand, that too is a little bit of the availability heuristic going on. But I think it permeates everything we do.

Rob Wiblin: Yeah. Some other drivers of war that I hear people talk about that you’re skeptical of include climate change and water scarcity. Can you talk about why it is that you’re skeptical of this idea of water wars?

Chris Blattman: So I think scarce water, any scarce resource, is something which we’re going to compete over. If there’s a little bit, we’ll compete over it. If there’s a lot of it, we’ll still probably find a way to compete over it. And the competition is still going to be costly. So we’re always going to strenuously compete. It’ll be hostile, it’ll be bitter, but it shouldn’t be violent. And the fact that water becomes more scarce — like any resource that becomes more scarce — doesn’t take away from the fact that it’s still costly to fight over it. There’s always room for that deal. The fact that our water is shrinking in some places, we have to be skeptical. So what is actually causing this? And then empirically, I think when people take a good look at this and they actually look at all these counterfactual cases where there’s water and war didn’t break out, we just don’t see that water scarcity is a persistent driver of war.

Chris Blattman: The same is a little bit true of climate change. The theory is sort of the same. How things getting hotter or colder affects interpersonal violence is pretty clear, but why it should affect sustained yearslong warfare is far less clear. That said, unlike water wars, the empirical evidence is a little bit stronger that something’s going on. But to me, it’s just then a bit of a puzzle that still needs to be sorted out. Because once again, the fact that we’re getting jostled by unexpected temperature shocks, unexpected weather events, it’s not clear why that should lead to sustained political competition through violence, rather than finding some bargain solution.

Rob Wiblin: Yeah. I remember when I first heard this idea of water wars, and some alarm bell went off in my head, and I’ve been skeptical of it from the beginning. I think one trigger that it sets off is: so water will be more scarce in some places in the future than it is now. Fair enough. But lots of resources are scarce now, and water is scarce in some places now. So why aren’t people fighting wars all the time about all of the things that are already scarce — including, in some cases, water? I mean, occasionally they do, but mostly they don’t. And it’s presumably because violence is incredibly costly and an agreement is better. And that will remain true in future.

Chris Blattman: Right. I think you said it better than me.

Rob Wiblin: Another thing that people think drives violence is having men in power rather than women. The evidence on that is a little bit mixed. It’s kind of an interesting picture. Can you explain that?

Chris Blattman: I mean, it’s more complex. So there’s this instinct that we see: men are obviously more aggressive in interpersonal violence. Most of the soldiers are men. Most of the war leaders and generals are men. So it’s totally natural to try to extrapolate from individual aggression to group aggression. But it’s like the passions: I think aggression is just one of these emotions that gets filtered out in larger groups over long periods of time in long fights. So it’s not surprising to me that when people actually compare the likelihood of going to war with male or female leaders, that they’re just as likely to go to war with one another. And when people have found clever natural experiments in current times or history — when it was quasi-randomly a woman leader or male leader — they too are not necessarily much more likely to go to violence, or go to war.

Chris Blattman: So that may not matter very much. Now, people could retort — and they should — to say, “Well, these are just women leaders. This is not actually an integrated government.” And maybe that would be different. We don’t really have 50%, or have women represented to their proportion of the population, almost anywhere in history. So it’s not a fair empirical comparison. But I think the theory is strong that we shouldn’t expect this individual aggression to become dominant.

Chris Blattman: I absolutely think that enfranchising women is going to make people more powerful. Anything that causes war is going to be something that helps you overlook costs. One big category of that is if a leader is an autocrat. In the extreme, the leader as an autocrat who bears very few of the costs. If an unchecked leader bears very few of those costs, why would they consider them in that calculation? They’re going to be far too ready to use violence. They may even have a private incentive, something that benefits them — not others, not their group — to go to war.

Chris Blattman: Now let’s think about a democratic leader where only half of the population gets to vote for them. Maybe it’s only men, which of course has been true in many times and places. Maybe it’s only one ethnic group or only one religious group. Well, half of that population is going to be ignored — their interests, the costs of war that they bear is going to be ignored. And so the leader is going to be far too ready to use violence and far more vulnerable to other failures and breakdowns that lead the group into war. So anything that makes them more accountable to a wider swath of the people — including women, irrespective of whether they are more or less likely to support violence in surveys or something — is just going to make the world more peaceful. So that’s the first-order thing we should be thinking about.

Rob Wiblin: Yeah. In part of the book, you suggested that poverty probably doesn’t increase war initiation. Just for the obvious reason that just because you’re poor, or just because there’s less stuff to go around, doesn’t make it more appealing to fight versus reach an agreement.

Chris Blattman: Right. The same argument with the water scarcity.

Rob Wiblin: Right. Exactly.

Chris Blattman: It’s a scarcity of other things.

Rob Wiblin: But you thought it could lead to prolonging wars or more escalated wars. And the argument was that poverty reduces incomes for nonviolent occupations. So you might be likely to leave the farm because the weather’s terrible and go get a military job as well. But I was thinking, doesn’t that reduce incomes from violent occupations basically by the same amount? So in a poor country, if your side wins the war, then there’s less GDP to steal. And at the same time, if the economy’s doing badly, you grab less by winning and there’s less tax revenue to go around to pay your salary as a soldier. So you might think that this effect kind of cancels out.

Chris Blattman: It could cancel out, and that might be why I don’t think the effect is super strong in the data. But often the people fighting, and the people leading those who are fighting, are able to capture most of these — whatever revenues and resources there are, they are able to capture those, and they make sure that the people with guns get paid first. And so I do think there is this fact that the gains and the assets in society will not be equally distributed in that context, so that’s probably what’s going on in these cases. But I agree. I just think the relationship between poverty — or sudden adverse shocks to income — and violence in general are not particularly strong in any circumstance. It’s only in this one circumstance where they do seem to play some role.

Rob Wiblin: Yeah. I suspect that people really overestimate the causal channel from poverty to conflict, because we observe a strong correlation. But it’s because the causal chain is more often the other way: that conflict leads to poverty. And you can’t easily observe which one came first.

Chris Blattman: Right. Finding natural experiments with changes in income is actually a lot easier than finding changes in a lot of other factors that I think do drive conflict. Because you can measure commodity prices, and you can measure temperature shock, and you can measure whether or not rainfall changed and destroyed your crops — and so we tend to just write a lot of studies about how poverty and economic shocks affect conflict, because we can. And then we don’t write studies about the things that are hard to measure, that the causes are a little harder to pin down. So we tend to obsess, and then we make a lot out of some statistically significant correlation that maybe isn’t actually even that close to important and only applies to some cases, because it’s the thing we can measure. I think that’s been a blind spot in research over the last 20 or 30 years.

Rob Wiblin: Yeah. When I try to think about what is causing a specific conflict and what might be able to help to cause it to stop, I try to put myself in this odd headspace where I mentally set aside which side is right — which side is just, and which side is unjust. Because I think on that positive question, I guess. This is what economists call a positive rather than normative question: a merely descriptive question of what would cause it to stop and what caused it to start. The question of who is right and wrong actually just doesn’t really affect that at all. And at the same time, having those normative facts in mind — who do you agree with and who do you disagree with — can strongly cloud our judgment on these non-normative questions. Is that something that you also try to do?

Chris Blattman: 100%. In fact, I think I do it too naturally now and too well, in the sense that I’ve realized in talking about contemporary conflicts, where passions run high, I find people get very upset at my dispassionate discussion. But I always have my own private views, so it’s this hard thing. I think to be a good social scientist and to actually analyze things clearly, you have to do that. But more importantly, I think you have to do that even if you are the general on one side or the other. They call it “strategic empathy”: you need to try to understand the conflict from the point of view of your enemy in a very clearheaded way, partly to avoid these kinds of availability and projection biases and other misperceptions that will just lead you to make strategic errors.

Cause of war #1: Unchecked interests [00:36:40]

Rob Wiblin: All right. Let’s now go through the five causes of war that you lay out in the book. I think this is the most important thing for listeners to remember. Hopefully many of you will go away and read the book and get all of the details, and hopefully these five causes will stick in your mind. But if you don’t, this is a really important section to keep in mind, so we’ll try to give specific names to these five different phenomena, these five different drivers.

Rob Wiblin: The first one you call “unchecked interests.” That’s basically where there’s a divergence in interests between the people who decide to go to war and the people who experience the consequences of the war. Can you briefly explain that phenomenon?

Chris Blattman: Right. So it’s a little bit like what I described in the context of women in war. Which is to say, think of an autocrat — think of a Putin, for example. If the main thing that is keeping us from going to war are the costs of war, and the person who makes that decision — in this case, a personalized autocrat — doesn’t have to bear the costs of everyday soldiering and death or deprivations from sanctions. Some of them — he’s bearing a lot of pain, let’s make no mistake, and risk — but doesn’t consider all of them. Well, that’s going to really reduce the incentives for peace and make him ready to use violence.

Chris Blattman: And what’s worse is when the circumstances around war give those leaders a private incentive: something that benefits them to go to war, but not their group. In Liberia, certain natural resources — not water, but diamonds, also timber, a few other goods that could be exported in this war economy — give warlords an incentive to go to war and stay at war, because they can use that disorder to enrich themselves. That’s a classic example. While that could be going on in a place like Russia, I think it’s more likely that there’s some reason for invading Ukraine, for example, that gives a private incentive. It may be a belief that this will entrench their power and solidify their political control, and continue to leave them unchecked.

Chris Blattman: So that is that first logic. It’s really simple: costs of war keep us from fighting. Leaders who ignore the costs because they can are more ready to fight.

Rob Wiblin: You have this nice example of how in provoking the French and Indian War, which I guess then blew up (although I don’t know whether it was causal) — but it expanded into the Seven Years’ War between Britain and France, and then also helped to launch the US War of Independence. George Washington actually had a lot to gain personally, and this might have influenced his judgment. Can you explain that?

Chris Blattman: Right. So I don’t think this is the primary explanation for the American Revolution, but I think we can’t ignore it, because it’s a factor. And I wanted to use something that was counterintuitive for people. It’s really easy to think, “Oh, that Liberian warlord, of course he’s selfish. We would never do that.” And I wanted to take one of the most iconic and high-integrity characters in history, and point out that even they were somewhat unchecked.

Chris Blattman: Because certainly after the American Revolution was run, most Americans didn’t vote, couldn’t vote. Certainly they weren’t holding George Washington or the other Founding Fathers fully accountable in the years running up to the revolution. And what’s interesting, and many historians have noted, is that most of the Founding Fathers had a big economic stake in independence, or a very particular trade and tax policy with Britain. It’s no coincidence.

Chris Blattman: And coincidentally, a lot of the people who remained Loyalists had economic stakes in union with Britain and with the British policies that Britain wanted to impose on the United States. And so that surely influenced. Along with all of the other more noble motives, that played a role in their willingness to launch the revolution. Washington was a notorious land speculator. He was a consumer: he wanted the fanciest clothes, the fanciest carriages, the land, trappings of wealth. He pursued that his entire life. And his pursuit of that land wealth is one of the things that set off the French and Indian War. So we just can’t ignore it.

Rob Wiblin: So Washington, among others, was going out West, trying to claim territory as his personal property — which involved taking it from the Indians and the French, potentially. The British weren’t necessarily keen on this, because this could lead to a very expensive conflict. And I think towards the point at which they were claiming independence, the British you said were interested in returning some of the land that Washington had claimed for himself personally.

Chris Blattman: Right. It was a private, illegal militia who had appointed a junior member of their elitist clique of Virginia families to go off and seize land from the French and Indians. In the British point of view, this like errant, selfish, materialistic, splinter militia was going off and provoking the enemy unnecessarily, and was going to bring everything crashing down in their own self-interest. And that’s, to some approximation, exactly what happened.

Rob Wiblin: Yeah. I’ll stick up a link to a really good resource on that, for those who want to learn more about that story.

Cause of war #2: Intangible incentives [00:41:40]

Rob Wiblin: The second cause of violence you point out is that people often get value out of war, above and beyond any material gains that they might achieve if they win. You call this category “intangible incentives.” I guess I might think of it as violence as a terminal value, or harming the other party as a terminal value. What are some of those intangible incentives for conflict?

Chris Blattman: Yeah. This is a really diverse category, and I never found a label I loved. So “intangible incentives” was the best of a bad bunch that I think captured it. You could think of it as violence as a terminal category. But you could also think of it as the thing that you value is actually not violence, but it’s something that only violence can deliver. Glory is an example. Vengeance is another example. There may be certain nationalist or ethnic or religious ideals about extending the power of your group, exterminating the heretic, exterminating the heretical view, racial purity.

Chris Blattman: So these have been more or less noble — many of them less noble — throughout history. And all of these stories, as different as they are, share this similar logic, which is that there was something ethereal. It wasn’t the material country or territory or economy or policy space that you were battling over — there was some value that you achieved through violence, like glory, or that only violence could deliver you.

Chris Blattman: But maybe the more important one for understanding a lot of conflicts — especially a lot of contemporary conflicts — which is also an intangible incentive, is that there’s some ideal that we hold that we refuse to compromise on. So the Ukrainians refuse to compromise on their liberty. The American revolutionaries refused to compromise on their liberty. And the Taliban refused to compromise, not so much on their liberty, but on a set of principles. The American administration, when dealing with the Taliban, refused to compromise on a certain set of principles of justice and of punishment of a grave crime.

Chris Blattman: And so our intransigence on a principle is a kind of intangible incentive that essentially says, “There’s a bargain to be made that we should accept” — semi-sovereignty as an American revolutionary or as a Ukrainian — “because they’re strong and we are weak. And I refuse.” That is the sensible compromise: “War will be costly, but because I value this ethereal thing so greatly, I will fight for it and I will not regret it afterwards. It’s not a mistake.” The principle of glory.

Rob Wiblin: “I stand by my choice.”

Chris Blattman: Exactly, “I stand by my choice.” And I think that’s very powerful. I think we can understand some fraction of these wars through these instances of noble or ignoble intransigence.

Rob Wiblin: Yeah. There’s this idea that a lot of people have heard, that Enlightenment values should cause us to fight less often. But I think it’s actually very theoretically ambiguous, because Enlightenment values not only say, “Yeah, sure. It’s bad to kill a random person overseas.” It does come with that. However, it also says that there’s these — at least in many interpretations — inviolable principles around human rights.

Rob Wiblin: And if that’s the case, if you’re just not willing to make a dirty bargain with a dictator in order to maintain peace — where you sell some strangers down the river, you allow the dictator to get away with what you regard as human rights violations — if that’s not on the table because you are too concerned with the moral purity based on the values that you have, based on the Enlightenment, then you might be willing to go to war more. Because you have values that are global, universal. You care about what’s happening in these other countries, where a purely self-interested country might not care at all what’s happening in Rwanda or what atrocities are occurring elsewhere.

Chris Blattman: Yeah. I think that’s a nice way to encapsulate something I’ve been writing on, just to sort of understand how the product of the Enlightenment and this humanitarian revolution is both peacemaking and not. It’s peacemaking because it makes us care about the other side. We always care about the cost to ourselves, maybe not to the others. So any regard we have for the wellbeing of others, and any repugnance we have for violence, is going to make us less willing to fight, because it’s more costly.

Chris Blattman: But you’re right. It leads to more idealism and intransigence potentially. I think that’s absolutely the case. That’s why I think we may see — I hate to say it — it has a little echo of this Clash of Civilizations story to it, which is to say different ideologies will fight. But I think it’s different. It’s saying that this triumphant thing that at least some of the world has accomplished — which is this human rights revolution, to the extent that we’re really rights-based and some principles we will never compromise on — is going to make peace more difficult in the future.

Rob Wiblin: Yeah. I mean, you can imagine that this human rights ethics approach could lead to less conflict when you have a really strong hegemon that has those values, and basically cows everyone into going along with them, and everyone else is willing to accept that. But in a multi-polar world, where you have all sorts of different groups with what they regard as pure moral values that they’re not allowed to compromise on, and they’re different ones. Then you should expect just constant conflict, more or less.

Chris Blattman: Right. It’s the tension between realpolitik, which is just the willingness to make these bitter unequal compromises in order to preserve the peace and in the pursuit of your own self-interest and these ideals.

Rob Wiblin: Some of these intangible incentives that you talk about, to me they seem downstream of a sort of game theory of conflict. So for example, having a desire to retaliate against a party that wrongs you. It seems like you could say revenge is an intangible incentive. But you might also think that actually this is a game theoretical strategic thing, because in order that people not attack you in the first place, you have to credibly be able to commit to punish them, even if it’s not in your interests, like narrowly after the fact. Should we put maybe some of these things in a different bucket, or think of some of them as intangible incentives and other ones as strategic considerations?

Chris Blattman: I think you’re absolutely right. And that’s the 201 level of the argument, but this is the podcast. So first of all, I would say there’s nothing about intangible incentives that isn’t strategic, which is in the sense that we value what we value. And so when you’re trying to — through game theory or just basic reasoning — think about how people are going to act, you have to understand what they value.

Chris Blattman: So intangible incentives just introduce a whole bunch of unusual non-standard things, but there’s nothing irrational about them: “the things we’re going to be strategic about” is maybe one way to think about them. So that might be glory or some intangible ideal. But what you correctly point out is that one of these, which is vengeance, seems like it’s the consequence of actions — that it almost can be generated by conflict. And that’s true.

Chris Blattman: I think we are programmed as a species to be vengeant. A lot of our social norms and cultures vary, but they can really augment that as well, and make us even more vengeant when wronged. So if you’re engaged in a game, if you’re engaged in a strategic interaction with an adversary, you actually don’t want to make them vengeful. You don’t want to do something that’s seen as so unjust that they want to attack you, because that’s going to make them ignore some of the costs — and then you’re going to find it harder to get a bargain. And more importantly, because they’re willing to pay some price just to punish you, they’re going to refuse a whole bunch of bargains that are advantageous to you.

Chris Blattman: So every time you make your enemy vengeful, you’re actually changing that range of deals available to things that are just more and more disadvantageous to you, because you just want them to stop. And so we should never do that. We should never start that kind of feud and create those cycles of vengeance. And often we don’t, because we can look to, like, “Well, I can try to repress them, but maybe that’ll backfire and then they’ll want to turf me out.” Or, “Maybe I shouldn’t invade, because then they’ll get really angry and they’ll want to punish me no matter what.”

Chris Blattman: And so they kind of backwards induct, and they decide not to do it. So then you have to explain that first unjust act and subsequent unjust acts. I think you then do that not through an appeal to things that are misperceptions and irrational: that people make mistakes, or that groups make mistakes. And so it’s a complex interaction between two of the five logics.

Rob Wiblin: It’s interesting. This is the part of the book that I love and I hadn’t really thought about before — actually, this is a feature that shows up again and again, and I’d like to return to — which is that you try to make things safer on one margin, and then you make them more dangerous in a different way. Or you make things more dangerous in one way, and then there’s this offsetting behavioral effect elsewhere that actually makes it safer. And it’s ambiguous like what the net sign is.

Rob Wiblin: So inasmuch as people want revenge, inasmuch as they can get into these feuds where they hate the other side and they’re willing to pay costs just in order to harm them, you might think that creates a very unstable situation in which wars are often going to escalate and be exceedingly costly and be common. However, if everyone knows that even if they do a tiny slight to the other side, that’s going to get them to attack back and then your side will be riled up and they’ll attack back and so it’s going to escalate, then you really don’t want to even slight the other side at all. And so plausibly, this desire for vengeance could lead to peace, because everyone’s going to be so cautious about it. It’s kind of like nuclear weapons potentially keeping the peace.

Chris Blattman: Exactly. Sociologists and anthropologists call this “peace in the feud,” which is that the threat of setting off one of these cycles of violence keeps people from engaging in those acts. But then if you do set it off, it’s a real doozy. And that’s the story of any kind of military technology in arming — and, as you say, nuclear weapons. Nuclear weapons will make war really costly, so it makes it less likely. But if it happens because of some other reason, it’s going to be really bad.

Rob Wiblin: I think this kind of offsetting behavior phenomenon shows up elsewhere. Where, let’s say that you elect a president of the United States who’s extremely peace loving and just says, “I’m incredibly reluctant to ever use nuclear weapons or extremely reluctant to go to war,” then the other side can be like, “Oh, great. Well, we’re going to take advantage of that. And we’re going to push up against the boundaries really hard. We’re going to invade Taiwan. We’re going to do all these other things because they’ve declared that they’re not going to do anything unless we do something even more extreme than that.” And so they walk up to the boundary again, because they have a particular desired level of risk of war. And then that president who’s really dovish actually hasn’t reduced the net risk of war; they’ve just changed what fraction of the spoils they receive.

Chris Blattman: Right. Which is why we don’t fight, but we get stuck in this trap of escalating arms at the very least. But it is a strategy. I mean, if you want to improve your bargaining position vis-à-vis an adversarial nation, and you elect a hardliner or someone who’s a little bit crazy and unpredictable to become president, that raises the risk of war, but it also improves your bargaining position. And that’s a classic — it’s very easy to derive from game theory.

Chris Blattman: And then we can think of lots of examples. A lot of people criticize Trump for his, “I’m going to keep people guessing.” It’s not clear that’s the right strategy. It’s certainly a risky strategy. But in some sense, electing a hawkish, somewhat unpredictable, seemingly crazy and somewhat unconstrained leader is a bargaining strategy vis-à-vis allies and enemies. We can disagree whether it’s the best bargaining strategy, but it is one.

Cause of war #3: Uncertainty [00:53:04]

Rob Wiblin: I’ve slightly messed up the order here, because I feel like this number two is slightly blowing into number three, which is “uncertainty.” So the third cause is about this issue that parties don’t know exactly how militarily powerful the other side is, and they also don’t know how much they’re willing to suffer or risk in order to defend their interests. Can you help explain how that uncertainty makes violence more likely?

Chris Blattman: So this is a world where we assume there’s no unchecked leaders: the leader is faithfully representing their group. Let’s cross that off. Let’s assume there’s no intangible incentives. All we live in is a world where two groups are uncertain about the strength and resolve of the other. So surely we won’t have war then, right?

Chris Blattman: And the answer is, actually a lot of the time we won’t, because they might start off thinking, “Oh, I think I’m pretty strong and they’re weak,” but erroneously. Well, I have incentives to actually invest in information. I’m actually going to spend a lot of time with spies and intelligence, and I’m going to spend a lot of time carefully looking at their signals, because I know I don’t want to fight and I know they don’t want to fight. And so we’re going to try to give each other credible information, so we avoid this bad outcome.

Chris Blattman: So uncertainty isn’t an inevitable path to fighting. But what happens in this situation is you’re like, “Wait a second, are they really as strong as they’re telling me? They ran those military parades, they did those border skirmishes, they have these big public investments in art — all these things that they’re signaling to me, is that a bluff?” It’s a little bit like poker. And in that circumstance, you can’t really verify amidst this uncertainty. So just like in poker, when you’re not sure what hand your opponent holds, you might fold and you might decide sometimes it’s worth it to call — or escalate, then call.

Chris Blattman: The same is true in warfare: sometimes it makes sense to escalate and call amidst this uncertainty. And in fact it makes sense to be somewhat unpredictable in that — just like you’re unpredictable in your bluffing and you’re unpredictable in your folding and calling. So it creates an environment where it’s in your strategic interest to sometimes fight.

Rob Wiblin: So I was thinking about this and how to model it. I vaguely remember looking at game theory models of poker back when I was an undergrad. I think actually, maybe rather than think about poker, because poker has various complications, you can instead imagine a game where two parties are rivals. They each draw a number between one and 100, then they do a poker thing, where whichever party has the higher number wins if they both have to end up showing their number, but they don’t know one another’s number. And they could decide whether to up the ante or call or fold.

Rob Wiblin: In that kind of game, sometimes even when you draw a really low number like five or 10, and you’re almost certain to lose if the other party calls you, then you want to bluff. Because if you always raise when your number is above 50 and you always fold when it’s below 50, then you’re completely predictable and the other party can exploit that total predictability. I think the best thing is a mixed strategy, where you raise in some proportion — there’s going to be some function, but you raise proportional to like, whether you drew one or 100. So you might always raise if you draw 100, and occasionally do it if you draw one, and somewhere in between if you draw 50. Is that right?

Chris Blattman: Yeah. It varies a lot depending on how you set up the game. But basically, when you’re in the situation where you’re really, really weak, and you know you’re weak but there’s some uncertainty, you’re much less likely to bluff than when you’re potentially more evenly matched. But you’re right that the incentives are never zero, or seldom zero. Sometimes they’re zero, but there’s whole ranges where they’re not zero, and that is dangerous territory.

Chris Blattman: Now, after the fact, because the uncertainty always gets resolved, when the fighting begins as it has — and a lot of the uncertainty’s been resolved in the current Russian invasion of Ukraine — then we look back at those bets as mistakes. “Oh, we were misinformed.” And you’re like, “Well, maybe we guessed wrong, maybe we had bad information. Maybe we were naive. Maybe we had all sorts of misperceptions. But also it was very tremendously uncertain and we didn’t really know.” So we can’t just say that bets where we got lucky were good ones and bets where we got unlucky were bad ones. That’s a bad way to learn from the past.

Rob Wiblin: You know that this kind of uncertainty, bluffing-calling model predicts frequent but brief skirmishes, but it has a harder time explaining prolonged wars. Why is that?

Chris Blattman: So there’s two things. One is it predicts actually a lot of potential violence, which could be skirmishes just to capably signal to your enemy how strong you are. So a little skirmish in Kashmir, or a little skirmish on the Chinese-Indian border, or a little skirmish here, a little skirmish there. In the book I give the example of gangbanging. This now friend and colleague who’s a social worker and peacemaker in Chicago used to be a gang leader, and he’d go into the neighboring gang’s neighborhood and shoot up buildings. That’s violence, that’s a way of communicating how strong you are, that’s what this colleague of mine was doing.

Chris Blattman: Then, if despite all those signals you’re still not sure about the bluff, you start fighting. Typically, any of the uncertainty gets resolved pretty quickly. So we were really unsure how strong the Russian military was. We were really unsure how plucky and organized the Ukrainian military was. We were really unsure about the unity of Western sanctions. That, we now know.

Rob Wiblin: Weighed it up pretty fast.

Chris Blattman: That we found out. So now a lot of the uncertainty is resolved, and so we would anticipate that if uncertainty was the main source of fighting, or one of the main sources, that it might wrap up pretty quickly. Sadly, “pretty quickly” doesn’t always mean a week. It might mean five months, especially if there’s other things leading us to fight. But very seldom do wars go years and years and years.

Rob Wiblin: An interesting effect of this dynamic in wars, and also in negotiations around wars or indeed any kind of conflict — even just in trade negotiations for different countries — is that both sides have really strong incentives to give extremely misleading impressions about their negotiating position about what they would accept and what they won’t accept. I noticed this during the Brexit negotiations between Britain and the EU a couple of years ago. It constantly seemed like they were never going to make a deal, like it was impossible that they would agree on something.

Rob Wiblin: And I learned over time to just basically completely ignore this. I thought that this news was worthless. It was worse than the paper it was written on, because of course both sides always want to claim that they’re right about to walk away from the table, because it improves their negotiating position. And basically as far as I could tell for someone like me, the situation is inscrutable, there is actually just no way for me to know.

Chris Blattman: It’s true. Now, that choice to bluff is also a strategic choice. So they don’t do it always, but it certainly pays off in a lot of circumstances. And we all do it. If we’re going to buy a used car, that’s exactly the same kind of everyday instinctual bargaining. “Oh, I could never pay more than this.” “Oh, I could never get it off the lot for that much.” It’s the same posturing because there is some uncertainty: I don’t know exactly how much this dealer paid. I don’t know. And he doesn’t know how much I’m willing to pay.

Chris Blattman: The only difference between that situation and conflict is if I don’t buy the car, it’s not like there’s some huge costs that we pay. But if we started bargaining and then some genie appeared and said, “By the way, if you guys don’t sell the car, I’m going to execute one of your family members each,” then we would find a bargain. And we might not find it in our interests to bluff so much, right? So it’s not a foregone conclusion that we’re going to bluff.

Rob Wiblin: This issue showed up also in the leadup to the Russian invasion of Ukraine, where I was just completely agnostic about whether they were going to invade. Because I was like, well, they have every incentive to claim that they’re going to, even if they’re not. So literally I just am not going to know until the tanks roll in or they don’t.

Rob Wiblin: It’s like that dynamic where a government has to lie to the other party, and in so doing, also lie to their people. It’s very difficult for democratic accountability, because it’s not possible to say one thing to Germany and another thing to voters. So voters actually can’t really tell on what basis you’re negotiating, and either keep you or beat you out of office based on your behavior, because you just have to lie to everyone.

Chris Blattman: And there’s layers of uncertainty. So if you’re the president of this country, actually, you’re not quite sure of the level of support for standing up to an invasion. And you’re not quite sure of your population’s level of support for you committing an invasion. So there’s layers to this game. And what that might mean is actually, what does negotiation look like as well?

Chris Blattman: One “negotiated outcome” would be for Russia to have merely unilaterally sent in tanks and planes — maybe shocking and aweing, maybe not — and then people just standing down. It’s kind of what happened in Crimea. They say half the Ukrainian military in Crimea defected over to the Russians, half of them quietly retired, very few shots were shot. And some version of that could have played out. A lot of people in retrospect say, “Oh, that’s foolish to even think that’s true.” And I don’t think people knew that beforehand.

Cause of war #4: Commitment problems [01:02:24]

Rob Wiblin: The fourth underlying driver of violence you list, you call “commitment problems.” It’s the inability to commit not to abuse your power in the future. Basically, if you have two actors and one is quickly becoming more powerful than the other, then the rising party would probably like to commit to not take too much of the pie, to not exploit the other party in the future, because that would stop them from being attacked now.

Rob Wiblin: But no matter what they say, the declining power knows that it’s going to be in the rising power’s interest to take advantage of them in the future. So the rising power could basically try to directly pay off the declining power now not to attack them — which does apparently often happen — but if their rising power is too steep, then they may just not be able to afford a sufficiently large payout to motivate the declining power not to engage in a preemptive strike in order to stop their rise. Can you explain how this kind of commitment problem might help explain why World War I occurred?

Chris Blattman: Sure. I should say, it sounds like an esoteric concept, and you’re like, “Oh, does that really happen?” And I think it’s actually so fundamental to understanding so many things. Actually, political economists also use these commitment problems to understand a lot of issues of domestic politics and dictatorships and low economic growth. So it’s worth thinking a lot about and learning, because I think it ends up being really useful. Some people say that every long war has a commitment problem at its root. I think that’s a slight exaggeration, but it’s true often enough that it’s worth pointing out.

Chris Blattman: And World War I is one of the classic examples. It’s a great example of where a lot of historians are very quick, as I mentioned, to point out the unchecked leaders in Europe, to the misperceptions that they had about this idea of a “jolly little war; home by Christmas.” Which is actually not what most leaders believed, but people like to tell that story. And they talk about nationalists and other ideologies or intangible incentives that led to war.

Chris Blattman: So we end up explaining World War I with unchecked leaders, misperceptions, and intangible incentives. Then a good number of other people said, “Yes, that’s true. But why were we so vulnerable to that?” And for this situation you just talked about and summarized so nicely — this launching a preventative attack and a preventative war — one of the keywords you’ll hear in the vernacular is a “closing window of opportunity.” As soon as you hear “closing window of opportunity,” it’s your signal for a potential commitment problem.

Chris Blattman: What was going on briefly in World War I was Germany’s fear that Russia would soon rise to be so powerful that they would be at a complete strategic disadvantage. And there was an opportunity to prevent that from happening, and that window of opportunity was closing very soon — it was probably going to close by 1917 or 1918, according to some of the German generals who were arguing this.

Chris Blattman: And the historians who don’t buy this fully are totally right to say, “Well, there’s ways to get out of that; it wasn’t a total commitment problem.” That’s why I sort of buy both stories. It was mostly a commitment problem. It was really hard to find a deal that Russia would be able to commit not to overpower Germany in the future and assure Germany of some stability. But it was really hard. And that’s the point at which our unchecked leaders and our misperceptions and our intangible incentives keep Europe from making a deal. And so we need both — it’s not one or the other.

Rob Wiblin: I think some people have argued that commitment problems are also behind the Russian invasion of Ukraine. I suppose we’re constantly courting trouble by talking about a contemporary case. But the argument there goes that Russia wants influence, power over Ukraine, whatever their goals within Ukraine are, that they want to have power over what’s going on there. They notice that the Ukrainian military is getting way more competent, way more powerful — as evidently they have over the last eight years. And apparently also Ukraine was investing in more equipment, investing in longer range missiles, stuff that would make it easier to defend themselves, these drones that seem to have really packed a punch.

Rob Wiblin: And so the Russians were like, “Well, Ukraine’s not actually going to be more powerful than us in five years, but they’re going to be more than capable of defending themselves in five years. So if we’re ever going to engage in a kind of preemptive attack in order to prevent them from getting to that stage, then we have to attack now.” And that can explain the kind of otherwise quite peculiar and arbitrary-seeming timing.

Chris Blattman: I think to an extent, absolutely, I think you definitely need more than that to explain this invasion, because there was a way out. The way out was for Ukraine to go the way of Belarus and submit to some degree, and that didn’t happen. So I think at minimum, you need an intangible incentive on the Ukraine side, like we said earlier. Like the American revolutionaries: this noble refusal to accept the cruel bargain on offer. And those two things combined, and now you have pretty limited space to avoid an invasion, I think. So that’s the strategic story.

Rob Wiblin: Yeah. I maybe should have said more at the outset. I think you and I are probably picturing this diagram in the book, which is this pie chart where it’s like this is the pie that we have to split between the various different parties. We’ve got 100 different points of resources. And violence is going to destroy say 20% of the pie, because we’re destroying all of the capital, we’re destroying all of this output, all of this stuff that we want.

Rob Wiblin: So we’ve got this 20% wedge there, which is the negotiating space where both parties would prefer anything inside that wedge. And each of these different five reasons just shrinks the size of that wedge, because it gives people the reason to go to war — in order to learn information about how strong the other side is, gives people intrinsic motivation to go to war, it means that if the leaders have different interests then that slice of the pie is narrower for them than it is for the population as a whole. Because violence is so destructive, it’s very rare for one factor to be sufficient to shrink that pie to zero.

Chris Blattman: Sometimes there are certainly instances, I think, and certainly conceivable. But these events have a lot of terrible roots, and there’s always something going on. So I agree, but it’s conceivable just one could do it.

Rob Wiblin: So balance of power politics is one way that these commitment problem wars can potentially be avoided. Can you explain how that would work?

Chris Blattman: There’s one sense in which we think of a balance of power as potentially being stabilizing — this is a common approach to understanding long pieces for long stretches of European history. So for the 100 years before World War I, there were, of course, wars in Europe, but it was a reasonably peaceful time between the Napoleonic Wars. And what this suggested is there was a shifting pattern of alliances that would essentially make sure that no one power could have that incredible rise, create commitment problems, and thus lead to a world war — until, of course, that happened in World War I.

Chris Blattman: But the reason that that’s not always particularly stabilizing, I think, is that these shifting alliances and the unpredictability of these alliances can also create the potential for power shifts. The example I gave in the book actually is this ancient Greek war, the Peloponnesian War, which has not been analyzed very formally by political scientists through this lens, to my surprise. I was hoping to find the ideas and the explanations served up on a platter, but surprisingly, they’re not.

Chris Blattman: Let’s be clear: I am not an ancient Greek historian nor political expert. But nonetheless, my take is that the chief thing that led to this famous long war between Athens and Sparta was the fact that there was a previously neutral party that was sitting by the side — another Greek state — that all of a sudden started to look like it might tip towards one side rather than the other. So it created this sudden potential shift in power that would forever change the balance of power and led one party to preventatively strike — in this case, Sparta — as this neutral party crept towards Athens.

Chris Blattman: And so it goes to a larger lesson of game theory, which is that everything I’ve told you so far — this natural gravitational pull of peace — makes a lot of sense when we’re talking about two players trying to find some alternative, and we use that pie that you talked about. Not necessarily true when you have many players and you have these shifting patterns of alliances and all sorts of possible factions. In that multiplayer setting, some of the predictability breaks down and it’s a little bit harder to maintain peace. And we do live in a multipolar world, so it isn’t quite so simple as peace, peace, all the time, peace, peace, and very rarely war. There’s more room for this to happen with all these actors.

Rob Wiblin: In that ancient Greek example, I guess the stylized description is that Athens was this rising power, and Sparta was kind of declining relative to Athens. And so they were getting very skittish, they were getting very nervous about what Athens might do in future when they’re more powerful.

Rob Wiblin: The interesting thing is because there was this third party, this third balancing power, Athens could be in favor of that third party allying with Sparta so that Sparta is less concerned about the rise of Athens, because they’ll still be balanced in the future and so won’t feel the need to launch a preemptive attack on Athens. It’s like this third party makes a commitment credible, or is a way of trading off something now, so that Sparta doesn’t have to worry.

Chris Blattman: The neutral power had the second-largest navy after Athens, so that would not be their preferred solution. Their preferred solution was actually for this third power, Korkyra, to remain neutral. And they tried to maintain that. So as Korkyra tried to pull them into some of their disputes and things, the Athenians resisted and they tried to be neutral observers, and it was just a very difficult path to walk.

Cause of war #5: Misperceptions [01:12:18]

Rob Wiblin: The fifth cause of violence you call “misperceptions.” I think this audience might think of it as systematic poor judgment, or poor judgment that occurs in a systematic way. Basically these are the thinking traps that people classically fall into in conflict situations: failing to put yourself in the mindset of someone else; not thinking up all of the possible underlying reasons that might explain an adversary’s actions, and jumping to the worst conclusion rather than thinking about how their actions could be more innocently explained; overestimating our capabilities relative to other parties.

Rob Wiblin: Maybe the main thing that jumped out at me about this chapter is that you cite a lot of psychology research that I personally would be pretty wary of relying on. Because it’s old enough — I suppose maybe this could apply to all psychology research, even up to now — but at least the older stuff I’d really worry that there was widespread enough bad research practices within the social sciences that you really can’t trust that the results are legitimate. How did you try to handle that issue?

Chris Blattman: I think a couple things happened. First of all, we have an endless list of behavioral biases that people have proposed and which there’s both good and bad research for. Mostly I was searching for the ones that could actually affect strategic behavior. So I think I approached it theoretically, in the sense that I think we have a really good theoretical sense of what kinds of biases are going to be worse for this particular situation.

Chris Blattman: A lot of the behavioral science and psychology and economics that people are so familiar with right now are actually about things that have to do with maximizing behavior, about optimizing a choice — not in a strategic interaction with somebody else, but just either with my future self or a mistake I make now in the moment, ignoring the actions of anybody else.

Chris Blattman: And a lot of those just aren’t that important, or just recede in importance in this particular kind of strategic interaction when the stakes are so high — because a lot of behavioral biases that we talk about go away when the stakes are super high. Not all of them, but many do. We have to focus on the limited number of ones that persist in big groups and bureaucracies that are working really hard to avoid a really, really bad outcome — and then can keep working three years into the whole mess. That’s what we need. So it limits the scope of possible things, as the first step.

Chris Blattman: Then you have to dive into the evidence on which ones actually seem to be true. I was guided a lot by the history and case studies and the sense of what came up when I was reading about Israel, and Palestine, and the Northern Irish, and the Greeks, and this, and that, were a set of persistent biases. So in some sense, it narrowed down where to look. And then you have to go to the harder psychological evidence, which as you note can be spotty, whether it’s in psychology or in economics or whatever.

Chris Blattman: Yeah, I try to be clear about when things are stronger than not. I think the preponderance of evidence in places, plus the tie-ins to the cases, plus the theory, they buttress one another. But there have been lots of efforts — I’m standing on the shoulders of people like Rose McDermott and Bob Jervis, who’ve been trying to categorize and really systematize our understanding of these misperceptions in international relations. So we’re getting closer, but there’s really a dearth of attention in research to these specific problems. I think I just tried to be transparent about the strongest case: the strongest case is the conceptual one, and the evidence is suggestively saying this is true.

Rob Wiblin: Yeah. Despite that question, I think it’s very likely that all of these misperceptions play a role. Firstly, if there’s one study, that’s a really bad sign. If there’s a whole bunch of studies, that’s a whole lot better. But even then, a whole literature can be confused. But a big thing is that all of these make a ton of sense to me; none of these are counterintuitive. All of these are things we could see in our own minds, I think. And just that prior probability, the fact that it makes so much sense, feels right, and isn’t surprising, I think it makes it a lot more believable.

Rob Wiblin: One interesting one is an example that you gave for how we fail to put ourselves in the mindset of others, that we could even fail to put ourselves in the mindset of ourselves at other times. There’s an example where people who are hungry buy more food at the supermarket. People who are cold buy warmer clothes, rather than anticipating how cold they’re going to be in future.

Rob Wiblin: This isn’t that important, but I was thinking about that, and I was thinking that actually, this could be rational. Because it’s so easy to tell what your preferences are now: you can just inspect them. It immediately feeds back without any energy. Actually analyzing it and trying to figure out how you would have felt in the past or how you will feel in the future is more effort demanding. And so if a decision’s not very important, like how much food to buy at the supermarket, then maybe it just actually isn’t worth the effort. That wouldn’t really explain the war case, but…

Chris Blattman: Well, it might for a very specific reason. I mean, you’re saying there’s like a transaction cost to thinking through something.

Rob Wiblin: Exactly.

Chris Blattman: And a lot of behavioral biases are about transaction costs and our brains working efficiently to avoid them — especially for lots of everyday decisions, which makes total sense. In some sense, I could have a sixth bucket of explanations called “transaction costs”: that it’s actually expensive and difficult to negotiate, and to think about your war and your opponent’s war interests and things, and those transaction costs are a cause of war.

Chris Blattman: And technically that’s true, but as a matter of fact, because war is so costly, the transaction costs mostly pale. So it’s more of like a conceptually true thing that I just think isn’t worth elevating to the level of a sixth cause. But conceptually, it is a sixth cause, and it’s the source of a lot of the behavioral biases that I think don’t matter in the case of war.

Rob Wiblin: Yeah. If a country ended up in a war with another country and you asked them why, and they said, “Oh, we couldn’t afford the diplomat’s salary,” it would strain credibility.

Chris Blattman: Right. I mean, there are limits on our spy budgets and our diplomatic budgets, and you might say we invest too little in some of these things. But yeah, I think we put a lot of effort into discerning what’s going on.

Rob Wiblin: Yeah. I think of the misperceptions that you listed, the one that I’m most nervous about is overestimating your capabilities relative to others. I think people have this strong perception that people just always overestimate themselves. They think they’re better drivers than everyone else. They think they’re better at maths than everyone else. They think they’re more beautiful than everyone else.

Rob Wiblin: But if you look more closely, you see that there’s quite a lot of situations in which people under-place their ability relative to others. Some of those, from what I’ve read, is where a task is difficult in absolute terms, where they’ve never done it before; where measurement of success is objective rather than subjective; and where it’s not central to their identity. A case that we talked about with a previous guest, Spencer Greenberg, is marathons, where people actually under-place themselves. Because for most people, including me, having to run a marathon isn’t key to our identity. I’ve never done it before. It’s pretty clear whether I did it or didn’t. And it’s also something that I just think of as difficult.

Rob Wiblin: Arguably, fighting in a war has some of these characteristics. You could imagine a leader underestimating their ability to fight on the field because they don’t have experience and it’s clear whether you lose. What do you think?

Chris Blattman: I think this is just something we don’t know the answer to and is just hugely important. Speculatively, I think that there might be a lot of interpersonal variation. So I don’t think all CEOs are overconfident. I don’t think all mutual fund managers are overconfident. But there might be selection into CEOs and selection into mutual fund managers that draws overconfident types. And then if you happen to get one, they might get you into trouble, right? And they clearly exist.

Chris Blattman: But it’s not every CEO, so something’s going on. And it’s not in your interests as a firm to hire consistently overconfident people, so why it exists is a bit of a puzzle. Then I think we have to understand why it persists even in a somewhat checked-and-balanced organization, like a shareholder-held, publicly traded corporation. And it does. So there’s a bureaucratic answer rather than maybe a purely psychological answer. I struggled to find much on that. I looked exhaustively. I’m sure there’s things I missed, but I don’t think we really have a good answer to why our mutual fund managers are just so ridiculously overconfident in their performance.

Rob Wiblin: Yeah. I think it could be the case that on average in these studies of over-placement and under-placement, a typical person isn’t overconfident in general. There’s just some things where they’re overconfident and some things where they’re underconfident. But it wouldn’t shock me if politicians by nature tend to be overconfident. Because just to begin with, most people who run for office, most people who try to pursue a career in politics fail. So maybe it selects for people who are willing to go for long shots, or people who are willing to accept that.

Chris Blattman: Oh, I think we know there’s selection, but then there’s counterincentives against those selections. Now in politics, it might be hard to discern and get those people out of power. But if you run a trading firm, you should be aware of the selection problem — that I don’t just want a bunch of overconfident young yahoos, because that’s going to lose money. And over time, the bigger and better firms should emerge that don’t have that problem.

Chris Blattman: So there’s some inefficient persistence of overconfidence in something as high stakes as our financial markets and our largest corporations. If it persists there — where there’s information and explicit measurement and accountability and instantaneous results — it has to persist in political situations. And wouldn’t it be nice if we understood more about that.

Rob Wiblin: Yeah. So some of these misperceptions you might expect would be very prominent for individuals, but they would be tempered in groups. Because you’d think, OK, an individual might fail to put themselves in another country’s shoes. But surely we have a diplomatic corps. This is kind of their whole job to think about it from the other side’s perspective, and they would help to correct this mistake. But I think interpersonal interactions can in some cases make this even worse. Or at least it doesn’t fix it as much as you might think, because interactions between people within a society can increase the pressure to indulge in misperceptions.

Rob Wiblin: A classic one that really stands out to me is, imagine that Joe Biden got up and gave a speech where he was like, “Well, you might be very mad with Putin, but you’ve got to understand the fourth cause of war is commitment problems, and Ukraine was improving its military. And so there was a closing window of opportunity in which Putin could act.” “Boo!” People are going to hate this kind of explanation of his behavior. That’s a crazy example, but you could get this in a bureaucracy even, where people don’t really want to go to the meeting and be like, “I’m the person who deeply understands Putin’s mind and can get myself inside that,” because you look like a jerk.

Chris Blattman: 100%. The late Bob Jervis and his analysis of intelligence failures in the US and in the invasion of Iraq is a great study in some particular bureaucratic failures to process information and draw the right conclusion and insulation. The thing is, most of the time I don’t think that happens so seriously that this terrible action, which is war, happens. It only happens some of the time, or maybe it happens all the time, but not to a degree that it’s decisive.

Chris Blattman: But then sometimes our decision making gets better through bureaucratic decisions. Everyone thinks “bureaucratic” sounds like a bad word, but bureaucratic means systematized, rule-based. A lot of our decisions are better when we make them in groups. And so there’s a lot of difference of opinion. I get some criticisms from colleagues saying, “This is nonsense, this group decision making, a lot of this goes away.” And then other people saying, “This is nonsense, the fact that you’re even acknowledging that group decision making could ever get better than individual decision making ignores all this other…”

Chris Blattman: The fact is, this is one of those additional instances — I’ve pointed at one or two previously — where I think we don’t know a lot. And it has a lot in common with us not knowing why mutual fund managers are overconfident. We’ve had all this investment. All this behavioral science that is so intuitive now to so many people is super individualistic. We can all quote 100 studies, right? Even the average person off the street, probably they could tell you about one. But nobody can tell you about how this breaks down in organizations, and some intuition about it. I thought it must be out there, and some of it is. But that’s our big failing as a social science right now.

Rob Wiblin: OK, we’ve come to the end of laying out these five different drivers of violence in order. Just to recap for people, they were unchecked interests, intangible incentives, uncertainty, commitment problems, and then misperceptions. What’s the best argument for a cause of violence that doesn’t fit into the above five buckets? I guess you’ve maybe already answered that with the transaction costs thing, but is there a seventh potential bucket?

Chris Blattman: I think I also mentioned this multiplayer problem, and I talked about it as maybe a kind of commitment problem, because it creates these shifting and hard to predict alliances. A big problem if you’re trying to bargain with an enemy is the fragmentation of your group, and splinter factions going off and doing something. Like George Washington was a splinter faction the British had to worry about. Yasser Arafat spent a lot of his time worrying about splinter factions. Things of this nature.

Chris Blattman: So is that a separate bucket or not? Again, I think this is where I don’t think we have the equipment to understand for sure. Kind of in both of these cases I think we push against the frontier of what we know. So it could just be that it fits into the buckets, or it could be its own bucket. I haven’t decided yet.

Weaknesses of the model [01:26:08]

Rob Wiblin: Yeah. That makes sense. What’s the thing that you’ve said so far that would be most controversial among political scientists or economists?

Chris Blattman: Well, the way I’ve tried to skirt controversy is not by trying to say, “This is the way the Iraq War started, and not this way.” I just am trying to classify people’s explanations. It’s a way of actually trying to keep everybody happy, which is maybe my inner Canadianness. So partly I just assiduously skirt all that, and then maybe I haven’t heard it yet, because the book hasn’t come out. So you should probably ask me in a couple weeks. It comes out April 19th. April 21st in the UK. Two weeks after that, I will know the answer to the question.

Rob Wiblin: I expect you’ll get some people who are maybe mad about the case studies because your description might seem like it’s justifying really bad behavior.

Chris Blattman: Yeah. The dispassionateness bothers a lot of people. Not the political scientists — the political scientists are empathetic to that sort of strategic empathy, that we’re trying to understand these people. At the same time, I avoided contemporary conflicts in the book for a couple reasons. One is I just didn’t want it to be dated the day it came out, because things change. I wanted this to be more timeless. But more importantly, I didn’t want people to get angry about the cases so much, because I wanted to sort of provide that dispassionate toolkit.

Chris Blattman: For people in the UK, Northern Ireland is still very fresh. For people in the US and the UK, the war in Iraq is still very fresh. So I’m not totally skirting things, but people are finding it very hard to think analytically about the Russian invasion of Ukraine right now. And that’s actually a problem for peace.

Rob Wiblin: One dynamic that I wasn’t sure where it fit into these five categories, is that people argue that if it’s easier to win a war if you start it first. And that’s true of both parties. Then that’s a reason to attack someone else before they attack you. And it’s a reason for them to attack you really quickly, so you want to get in even sooner. So you get in this kind of race to attack. Where does that fit?

Chris Blattman: That’s another classic kind of commitment problem, because both of you would like to avoid this costly outcome, which is war. And because you have to make this simultaneous instantaneous choice and there’s no third party to restrain you, you can make it. There’s lots of varieties of the commitment problem, actually. I just happened to emphasize the one that I think is most intuitive to people, which is the preventative war. It was a pretty comprehensive book already without being super comprehensive. But I think you’ve picked up on something that I slunk past in the service of not trying people’s patience.

Rob Wiblin: Brevity, yeah.

Chris Blattman: So very quickly, you’ve brought us to the 201 class again.

Rob Wiblin: Yeah. That’s what subscribers to the show want.

Chris Blattman: Perfect.

Dancing on the edge of a cliff [01:29:06]

Rob Wiblin: Maybe the answer’s just going to be the same, but another kind of game theory model that people talk about, especially in the context of nuclear war, is this kind of beautiful little setup I think created by Thomas Schelling, who wrote this famous game theory book, The Strategy of Conflict.

Rob Wiblin: The setup is you’re standing at the edge of a cliff, chained by the ankle to someone else. You’ll be released and one of you will get a large prize as soon as the other gives in. How do you persuade the other guy to give in when the only method at your disposal — threatening to push the other party off the cliff — would doom both of you because you’re tied together? The answer is, you start dancing closer and closer to the edge of the cliff. That way you don’t have to convince them that you would do something totally irrational, like plunge both of you off the edge of the cliff. You just have to convince them that you are prepared to take a higher risk than they are of accidentally falling off the cliff. If you can convince them that you’re more of a risk taker than them, then you can win. Where does that fit?

Chris Blattman: It kind of depends how you set it up. The way it’s set up there it’s mostly a story about uncertainty. Fundamentally, the problem is, “I don’t know whether you’re crazy or how resolved you are.” So it goes back to that whole discussion we had about it’s a bluff, but it might not be a bluff. You’re creeping closer to the edge as a way of signaling your willingness to just burn the whole place down if you don’t get what you want. And creeping too close to the edge is a good signal only if nobody but the craziest person, as you slowly get there, would actually do that. So now you’re credibly signaling. Now the fact is, standing far from the edge or close to the edge, I don’t think it’s a perfect example. Because that’s actually not a very credible signal at all, because I’m not actually going over the edge.

Rob Wiblin: It’s not that dangerous.

Chris Blattman: So that’s one way to think about it: it’s just uncertainty and bluffing, and it can result in somebody potentially going off the cliff. And it can be strategic. I think what subsequent game theorists did was refine that logic. Schelling sort of created all of these puzzles for people to work out, which then a lot of people just filled in the blanks.

Chris Blattman: The other way it could work though, is again, related to something we mentioned earlier. Let’s say each group gets to elect who gets chained. You may have an incentive to elect the crazy person to be chained up, because that’ll actually improve your bargaining position. But it’s a difficult calculated decision, so you might play a mixed strategy in whether or not you elect a crazy person.

Rob Wiblin: Yeah. All these strategic cases, they suggest that there is some minimal risk of war that we can get to that we can’t get below, because it’s just not a stable, competitive situation. Because one party then has a reason to increase the risk of war.

Chris Blattman: The central contribution of game theory to the study is to say that actually the fact that we have shifting power, the fact that we live in an uncertain world, and the fact that there are unchecked leaders — which game theorists would call “agency problems” — means that there can be lots of rational war. And that’s a huge contribution.

Chris Blattman: I think the problem is, for a long time, that discipline has totally ignored the irrational war, and not tried to do this sort of mind meld that I’ve tried to do in the book, and try to make some of these things work in our models and explore the implications. And I’m not a game theorist, but I do know enough that you never really know how the model and the prediction is going to work out until you work it out. And there’s a lot of counterintuitive insights that might seem obvious when I explain it now, but weren’t obvious beforehand.

Chris Blattman: There’s a lot of things we think about irrational behavior and this story you’ve just told, that if you actually work it out, you might not get the conclusion you expect. And the sad thing is there’s maybe only a handful of people who actually work on conflict-related game theory these days.

Rob Wiblin: Huh. It’s just out of fashion?

Chris Blattman: I don’t know. It’s also not that big a field and it’s also a hard-to-acquire skill. Maybe it’s out of fashion. It hasn’t been favored by the technology of research, right? There’s so many empirical researchers and a lot of the empirical research has boomed. Let’s think about it. When I went to college, we didn’t have computers. There was no software to run statistical analysis — and if there was, the computer wasn’t fast enough to do it and nobody had created the data. So we’ve had an empirical revolution because of the sort of shifting technology of research. So it’s not so much made it unfashionable; it’s just people chasing the frontier are not doing as much theory.

Rob Wiblin: Now you can run too many regressions. We’ve got the opposite problem!

Chris Blattman: True. Exactly.

Rob Wiblin: I guess we had to pause in 1992 or something. The model that you present in the book is kind of this dividing pie thing. And for simplicity in the book, you conceptualize things in terms of: different groups have different levels of power, and that changes their probability of winning the war and getting the whole pie — or at least the whole pie that’s remaining after the destruction of the war.

Rob Wiblin: Which is a nice simplification, but it seems more often in conflict, a small group would almost certainly lose a war because they’re just so much weaker. And instead what’s going on is that they have some ability to harm a much more powerful party, even if they can’t actually win a pitched battle. Does thinking about dividing up the pie more in proportion to how much you can harm someone else, rather than your chance of winning an outright war, can it change anything important about the conclusions that spit out?

Chris Blattman: Generally, no. I think that’s why I could make the simplification, is most of the things we get are intact. I think what you end up getting is you then start to understand things like terrorism better. Terrorism is the weapon of the weakest: it’s the way that you exercise bargaining power versus an incredibly strong foe. Not because you think the terrorists are going to manage to take over the country, which very seldom happens, but they leverage violence against civilians, with the fear of violence to obtain concessions or obtain some other aim.

Chris Blattman: So I think it helps you understand a greater variety of violence. But a lot of the same principles, the idea that peace is optimal, and a lot of the five sources of breakdown still apply.

Confusion around escalation [01:35:26]

Rob Wiblin: I want to talk a little bit about this issue that I’m kind of confused about. I’ve been reading this book Only the Dead by Bear Braumoeller, where he does a whole bunch of statistical analysis to try to figure out whether wars are becoming less common — like, the initiation of wars is becoming less common, or their escalation is becoming less frequent.

Rob Wiblin: He has this really interesting quote that stood out at me when I was reading it this week, which is:

The fact that both intensity and severity are not just thick tailed, but also fit a power law distribution well, is especially ominous. That fact means that the outcomes of wars are consistent with a process in which the only difference between a small war and a very large one is random chance. Even people who can manage to wrap their heads around the incredible escalatory potential of war often find this point hard to believe. Don’t human beings have control over the outcomes of wars? Can’t we just put a stop to them if they’re getting too bloody? There’s a short answer to this objection and a longer, more detailed one. The short answer is that, yes, we certainly can put a stop to a war that we’re fighting if we’re willing to lose. But if we were willing to lose, we probably wouldn’t be fighting the war in the first place, would we?

Rob Wiblin: So something I still don’t quite know how to model in my head is the choice that both parties have in some sort of auction kind of scenario, where they’re both fighting one another over something that they care about. And they have a choice in each period to either fold or raise. They can fight harder or not. And they both end up paying the price of the war, so it’s sort of a double-pay thing, where it’s like they both either have to pay with effort or with the violence that’s inflicted on them. How can one think about that more structurally, the choice to escalate or not?

Chris Blattman: Let’s see. I like that book a lot. I read it a long time ago. The first caution I would just say is that there is a difference between statistical noise and real randomness, which is to say that the things that aren’t in our model that we can’t observe might be what’s driving the ones that escalate from the ones that don’t. We actually don’t have that many wars to study, and so our statistical power to discern what’s noise or not really deteriorates. We can’t even measure most of the things I just talked about, like: is there a commitment problem? Is there an intangible incentive? Is there some ideological issue at stake? It’s hard to observe. How do you measure that in the data? So those don’t go into the models.

Chris Blattman: So we see a lot of escalation, we see a lot of non-escalation, and then there’s a whole bunch of conflicts we never observed — they’re not in the dataset because they weren’t conflicts. So it’s just hard to do. Then I think we have to think a little bit more, and we can run regressions and do careful studies about what escalates. Once again, I think the most important things we can’t measure. So that’s why in spite of being, in my day job, this applied statistician in some ways, when it comes to understanding why we fight or why we escalate, I end up going back to the cases and the theory a lot for the guidance, and then thinking about how we’d build better evidence on that kind of thing.

Chris Blattman: So for example, what does it mean to refuse to lose? I think one interpretation of that is to say losing entirely just isn’t in our bargaining space — that we actually just want to find something. Getting zero isn’t an option because we can always do better by fighting than zero. So there’s a rational story there. Then we have to say, “Well, why didn’t we get something in our bargaining range?”

Chris Blattman: But there’s another interpretation that would say that sometimes there’s something in that bargain that we should accept, and we refuse to. Like Ukrainians or American revolutionaries compromising on their liberty. And that’s hard to predict, right? Why did Ukrainians resist, and the Belarusians did not? Why have there been color revolutions in some Eastern European countries, and there hasn’t been a major color revolution in Russia? Why didn’t the 2018 protests escalate? Why did repression against protestors in one place work, and media control in one place work, and why did it unravel in the other? That’s just hard to answer.

Chris Blattman: We have some answers I think, but that to me is like, what’s going on. There’s a certain sort of unpredictability to it that is not because it’s just a power law or something mechanistic. But because we don’t observe, it’s just hard to predict.

Rob Wiblin: So a kind of model that I had in mind is an extension of the uncertainty poker thing that we were talking about. As we noted, a slightly perverse implication of the uncertainty thing — needing to bluff, needing to seem tough, needing to signal that you have resolve — is that a weaker party can choose to escalate a war or start a war, or accept that a war is going to happen, even when they suspect that they’re weaker and that they’re going to lose in a direct sense. Because they have to seem fierce. They want to seem like they’re not bluffing, like they’re willing to go to war if it comes to it.

Rob Wiblin: And the optimal strategy involves randomization here. So, the party might say, “We’ve drawn a random number. We’ve decided to escalate in this case in order to signal our fierceness.” And the other party could basically do the same thing, and you could just get extremely unlucky, where all sides decide to escalate again and again, and proceed to a really full-scale war that they never really wanted in the first place. Just because they feel like it’s necessary in order to signal the kind of tough guys that they are. Does that make sense?

Chris Blattman: Yeah, I think it does make sense. I think it’s hard to get to a long war from that. I think the escalation, escalation, escalation, and both being willing to do that is improbable enough. Except in a world where — it goes back to this multiplayer games —

Rob Wiblin: Everyone’s watching.

Chris Blattman: Everyone’s watching. So if I’m Saddam Hussein facing the US, as I point out in the book, I was really struck by — and now I don’t remember which historian pointed this out — that in Saddam’s mind, the US wasn’t even enemy number one or enemy number two. Wasn’t even enemy number three. Saddam Hussein was chiefly worried about Iran. He was worried about the Shiites. He was worried about Israel. And they were all watching what Saddam would do.

Chris Blattman: And likewise when the United States was trying to convince the Taliban government of Afghanistan that we’ll invade if you don’t hand over bin Laden: both the Taliban were thinking about their reputation, but more importantly, I think the United States was thinking about every other potential enemy, conventional or terrorist, that would be watching to see what they do. So I don’t think we can understand America’s 20-year war — and I don’t think we can understand Saddam Hussein’s brinkmanship, or the Bush decision to invade Iraq — without thinking about how they were trying to, in an uncertain world, signal resolve to a wider audience in order to move every other bargaining game a little bit in their favor.

Rob Wiblin: That’s the thing that can make a lot of sense, I suppose, of how you’d be willing to bear such huge costs. Because why would you go into a total war just to show that you’re fearless? You’re losing everything that you were trying to protect. So you need to have some broader considerations that are beyond the scope of that war in order to make sense of the cost.

Chris Blattman: I think so.

Applying the model to the war between Russia and Ukraine [01:42:34]

Rob Wiblin: Maybe let’s come back to that, because I want to talk a bit about the war in Ukraine and what we’ve gotten right and wrong, and maybe how these models can apply to that beyond what we’ve already talked about. I think this might be a good case to bring up later.

Rob Wiblin: So basically, I’ve been following what both of us have been saying about the war in Ukraine, and over time, I’ve started to worry that I’ve gotten some stuff wrong, or at least my recommendations in some areas would’ve been harmful if they’d been followed. And I think you have a similar way of analyzing these issues, so maybe you’ve done the same. I guess you thought that Russia probably wouldn’t invade Ukraine. Basically you just thought that on priors, they probably wouldn’t invade because it was a bad idea. Is there anything more systematic that we can learn from that? Or is it potentially the case that you just got unlucky?

Chris Blattman: To be determined. We don’t really know what’s going on. A good example is I have a Russian colleague, who’s an eminent game theorist and was once the head of one of the most eminent universities in Moscow (he’s now here), who has a completely different view than mine. He obviously has more information. But despite being a strategic thinker and someone who studies strategic thinkers, he thinks this is very irrational and it’s all misperceptions and intangible incentives in an insular, degenerated, inner circle around Putin.

Chris Blattman: I don’t know if that’s right. I’m not totally persuaded by him. I certainly think it’s a role. I can’t quite convince myself he’s fully right. Not everyone agrees with him, even other experts, but it’s a good example. It’s a totally plausible state of the world.

Chris Blattman: So what we learn from this is going to depend on what was actually going on. Because I can articulate, and I have here, the strategic logic of the invasion. Especially Russia’s strategic logic to invade makes even more sense in a cold, calculated, brutal way. But in a realpolitik kind of sense it makes sense, given if the Ukrainians are these nobly intransigent defenders of liberty. Then you’re just going to have to seize it, is the rationale. Just like that was maybe the “correct decision” for the UK in the American Revolution.

Chris Blattman: But if I’m wrong and my colleague is right that they’re just not thinking strategically at all, that means that any solution that follows my advice would be a terrible one. Likewise, if he’s wrong, any solution that follows his advice would be a terrible one. It’s a little bit like going back to our doctor analogy: the treatment suits the disease, and we don’t have a clean diagnosis right now because there’s so many mysteries.

Rob Wiblin: So I suppose you might disagree about which of the five drivers is doing most of the work, and there’s kind of a question of how personal is it versus how strategic is it?

Chris Blattman: One thing I would say is never believe anyone who’s super confident. Because that to me is what’s misleading. That to me is what’s misleading about so many people: an unwillingness to admit uncertainty over which of the five is operative when they’re all plausible is suspicious.

Rob Wiblin: Yeah. I suppose part of the reason to think that war was unlikely was that it would be bad for Russia. And it seems like probably, ex post, it is going to turn out to be bad for Russia. And yet they did it anyway. I guess I need to maybe look into this, but people who have looked into the cases of countries that initiated wars of choice over the last 100 years, how often does it work out positively for them, in people’s general assessment? It seems like it’s very rare that it works out in the interests of the country as a whole. Things usually go worse than they expect; the costs tend to be higher. If that’s the case, then maybe this is slightly the wrong framework, because we’re just going to get what’s happening surprisingly often, when it’s not in the interests even of the party choosing to do it.

Chris Blattman: I think the right way to ask that question would actually be to say: how many times did it make sense to credibly threaten violence, and then potentially roll the dice and decide to use it? The credible use of violence is extremely important, not just in achieving your interests, but in maintaining peace as well. So I think that would be the fair comparison. If we think that violence is, by definition, the inefficient thing to do, then we’re kind of stacking the deck. Of course it was always the wrong thing to do. It typically is not going to work out in your interest because we know the negotiated solution. So we have to have the fair comparison.

Chris Blattman: One example I like to point out is that everybody knows about the US invasion of Afghanistan. Most people don’t know about the US invasion of Haiti — and that’s because they won before they even landed the planes. All it was was the credible threat of actually the invasion and actually having the planes in the air with the troops, and a non-democratic coup leader said, “OK, OK, OK.” And more often than not, it doesn’t even come to putting the troops on the planes and having them over the Gulf of Mexico. Those instruments of power are in many countries’ interests, so that’s why the US arms. That’s why the US, and any imperial power, or any power, gets its way in the world: because it can credibly threaten violence.

Rob Wiblin: I suppose going to war in a case where it’s not in your interests might be an even stronger signal, because it shows you’re crazy enough to do it, even if it’s too costly to you in a specific case.

Chris Blattman: That is one explanation for the 20 years that the US spent in Afghanistan: that it’s a costly signal exercise. And from one perspective, it’s not clear. It was costly to the Afghans. It’s not clear how costly it was to the United States. You could say it cost about 0.5% of GDP, and in bad years, dozens of deaths, and in other years, many, many fewer. And that was the price paid for no terror attacks on US soil and intimidating potential future al-Qaeda. Ex ante, is that a price that any president would willingly pay? Maybe.

Rob Wiblin: Maybe, yeah. So the obvious thing that I think both of us got wrong — along with a lot of other people, to be fair — is that at least my thought was Ukraine should probably be willing to give up more than it seemed like it was willing to give up in order to prevent the war from happening. Because I assumed, as I think many people did, that their military position was very weak. My prediction was that Russia would have a hell of a time trying to occupy Ukraine. But I thought they’d be at least able to sweep in to begin with and try to oust the government. But basically it seems like the things that people predicted would happen with the occupation, where they would struggle to do it, kicked in on day two.

Chris Blattman: Yeah.

Rob Wiblin: They just attacked on all sides, and it’s kind of this failed occupation even just of the border areas. Now, one way of making sense of this is just that it was extremely uncertain: this is just a case of uncertainty on this list. We, along with Russia, along with lots of other parties, thought Ukraine was weaker than it was. But maybe the Ukrainians got lucky, or maybe Zelenskyy had private information that he couldn’t credibly communicate about just how resolved they were and how strong their military was. But in a sense, we just got unlucky. What’s the probability that we got something else more structurally wrong than that, that we misunderstood the situation?

Chris Blattman: What I’ve heard, not so reliably, but what I’ve heard from some people who are studying this more intently than I am is that, for example, what you just said about your expectation of how strong Russia was relative to Ukraine was not inconsistent with US intelligence reports. Maybe those were systematically wrong and structurally wrong, and maybe we’ll find out. I’m not sure that’ll ever get resolved. No one’s ever going to open up the Ukrainian or Russian archives to the extent that, say, the CIA let scholars like Bob Jervis peer in.

Chris Blattman: So I think probably it’s mostly uncertainty. What’s structurally wrong? I think the Americans have not invested a great deal in understanding Ukraine. I don’t know that; I’m guessing we don’t have huge diplomatic missions there, didn’t before the war. And certainly we’re training their military. But I’m not sure how much of the foreign policy apparatus was that experienced, so maybe it was more uncertain than it could have been if people were taking this threat more seriously.

Chris Blattman: It’s also possible, this idea that in order to preserve power, and by becoming a more personalized and repressive leader systematically over the last 10 years, Putin has manufactured an inner circle that is insular, that is full of second-rate people. A defense minister with zero military experience. That he has, in order to preserve his power, built a rickety structure that fed him bad information. That’s a really plausible story. It’s 100% true to some extent, whether or not that’s it, with no uncertainty — I think it’s some combination.

Rob Wiblin: Following the conversation about this, there’s a bunch of folks who are somewhat jubilant now. I would say people who wanted to engage in appeasement of Russia, people who wanted to reach some kind of negotiated settlement ahead of time, kind of like me, we had it wrong. We shouldn’t have done that because Russia was asking far too much, relative to their military strength. I think there’s a lot to be said for that.

Rob Wiblin: On the other hand, I worry that because we’ve been kind of happily surprised by how the war has gone from a military sense, we’re not paying attention to the fact that a bunch of cities have basically just been flattened, that tens of thousands of civilians are dead, that the economy of Ukraine is doing much worse than it would have otherwise. And that even now, even knowing that the military has over-performed, the war has been so costly — not just for Ukraine, but for the world, for the global system. That even a bad agreement, even after all of this, even after you’ve done unusually well, could still have been better than the outcome that is now going to be on the table.

Chris Blattman: Yeah. Quite honestly, let’s think about how this might resolve. It’s easy to imagine a world where in six months, Russia de facto controls and has incorporated much of eastern Ukraine and Crimea into itself. Maybe Ukraine doesn’t acknowledge that officially, maybe the West doesn’t acknowledge that in law. It seems implausible. They might, but let’s say they don’t. But they kind of recognize that that’s what it is. And Ukraine, in order to attain a peace agreement, declares neutrality, does not militarize, does not acquire these long-range missiles for example. Everyone renounces Ukraine’s eligibility for NATO, et cetera, et cetera.

Chris Blattman: Let’s say all of these things happen. You’re like, “Well, that kind of was one of the deals on offer beforehand.” Now, you can say that was one of the less attractive options for Putin, and given that he thought he would do better, he might not have accepted that. Maybe. But yeah, I’m not sure about the way things shake out, because Ukraine is relatively weak, no matter how well they’ve been doing. I’m not sure that the final outcome is going to look that different from what was on the table in the beginning.

Rob Wiblin: So let’s come back to the reputation and escalation thing, because I think some listeners will be shouting into their podcasting app about this. Basically, a key reason why it would have been bad for Ukraine to make such a deal — where it gave up part of its territory to Russia in order to see off a war — is that this is the way in which a powerful country like Russia can just bully countries one by one and take advantage of everyone around them. As you can imagine, now they don’t have to fight a war in Ukraine, they’ve stolen a bunch of its territory. Now the lines are renegotiated. But now they can just go and do that again to someone else. They can go and bully them.

Rob Wiblin: So the argument goes that Ukraine — and indeed all of the countries that face a threat from Russia, including the West as a whole — need to draw a line in the sand. And even though what Russia might be asking for wouldn’t be worth a total war, we have to be willing to go to war, to an extreme, to suffer enormous cost, in order to prevent ourselves from being picked off one by one and ultimately, in the long run, paying a huge price.

Chris Blattman: So first of all, I think Russia’s divide-and-conquer strategy with its neighbors has been working, and they’ve not had any mechanism by which they can band together to collectively resist this. Tim Frye, who’s a Russia expert at Columbia University, has a fabulous book, published last year, where he also talks about that’s how many dictators, including Putin, grab power and personalize power. There’s lots of oligarchs, there’s lots of military generals, there’s lots of powerful governors as well, but they were never banded together to collectively resist him, and so he was able to sort of divide and conquer. That’s a strategy that works well for him.

Chris Blattman: And now there is an alliance, the NATO alliance, plus Ukraine, that is drawing a line. Sort of, right? NATO isn’t really drawing that firm a line. They’re supporting Ukraine; they’re not actually allies. They’re not militarily getting involved. So you can understand maybe the West’s willingness to fuel this war on those terms. And that’s a kind of unchecked leaders or unchecked interests argument, that says, “Listen, Ukrainians are paying the costs, not us.” It’s really easy to draw a line in the sand if it’s not our sand. It’s the story behind every proxy war.

Chris Blattman: So we’re fighting a proxy war in Ukraine. It’s a nobler proxy war than most proxy wars, but every proxy war is in some ways the product of unchecked interests. Because there’s some superpower or some neighboring power that’s fueling one side of the conflict, even when it makes sense to deal, from the perspective of those two very local warring parties.

Rob Wiblin: Yeah. So one narrative is: we have to arm the Ukrainians in order to protect Ukraine. A different narrative would be: we have to fight an incredibly destructive proxy war in Ukraine that is bad for Ukrainians, in order to protect Taiwan.

Chris Blattman: Yeah.

Rob Wiblin: And these are very different stories for what is going on.

Chris Blattman: Yeah. Both of these aren’t decisions about peace. These are decisions about bargaining power. In some sense, there’s lots of efforts each side can make to alter their bargaining power in this conflict and alter their actual bargaining power in future conflicts, both through actual investments and also through signaling. And that’s how to explain a lot of what’s going on: we’re fighting a war now because we’re thinking about all the future wars we want to avoid. Or more importantly, the future wars we know will be avoided, but we’ll be in a stronger bargaining position. We’ll get a better deal without violence than before.

Rob Wiblin: Yeah. So I laid out just a second ago the argument for escalation on the basis of trying to show how fierce you are. From NATO’s point of view, an interesting thing is that Russia internally could have a similar argument, where they’re like, “Well, we deployed all of our troops to the border and threatened to go to war. And if we back down now, then people are just never going to take our threats seriously again. So we have to go to war, even if we think it’s a mistake in this case, just in order to, again, convince people that we’re willing to go to war in future cases, so people don’t laugh in our faces.”

Chris Blattman: Yeah. That kind of face saving, if it’s about signaling reputation, can happen. I think it’s maybe a little exaggerated and all sides recognize this, and so finding ways that Russia can save face is sort of a natural peacebuilding strategy. But a special case of this is something political scientists call “audience costs.” Which is that if we think that Putin would be deposed if he lost, then on the one hand, you might think, “Great. Glad to see you go.” On the other hand, that is an incentive for him to just keep fighting beyond where it makes any sense whatsoever. And so there’s a certain logic to letting someone like Putin not lose completely and letting him save face, and paying the price of knowing that the regime will not change but at least it will end the war.

Rob Wiblin: Yeah. OK. I’m not sure exactly what to take away from all of that, but I suppose that’s natural at this stage of the situation. I think if we tried to tie it up more neatly than that, we would still be proven wrong by events in just a couple of weeks.

Chris Blattman: I think it’s a good example of how I find it really useful to have this analytical frame and the five forces to analyze the situation. But despondent because, even with all the information available, and even in retrospect looking afterwards, it’ll be difficult to really say, “It was this.” And so you have to be comfortable with a degree of uncertainty and ambiguity in what we know. The good thing is, at least it’s a little bit more organized than the messy way I was thinking about it before.

Rob Wiblin: One thing I just take away — which you already know, but it’s always good to be chastened by events — is just how hard it is to predict the future. I talked earlier about my estimation of the military situation as if I had put thought into it, as if I had actually analyzed or collected any information. But it’s not even at that level. I just kind of assumed that Russia was in a far stronger military position.

Chris Blattman: As if either of us could have found Ukraine on an unlabeled map five months ago. Right? So I hope I’ve stressed that in some way, I know enough to be dangerous, as most of us do.

Rob Wiblin: Yeah.

Chris Blattman: But I guess the way I justify this is, I think there’s a lot of use in bringing these lessons from dozens and dozens and dozens of other wars, to try to organize what people talk about. So rather than say, “This is what I think is going on,” again, I try to do this thing of just classifying and organizing the way people are talking about it. And then of course I shoot my mouth off about what I think will happen and I often mispredict. But hopefully not too terribly.

Rob Wiblin: Yeah. I wish my performance could rise to the level of me being wrong because of something I thought about a lot, rather than something I didn’t even think about at all. This is a part of the book, actually, that we won’t get to talk about very much, but you have this observation that we tend to be very cautious and circumspect about our opinions about things that are closer to us, because we understand the details and the difficulties. And then we’re more confident about things the further away they are, precisely because we don’t know what we don’t know.

Chris Blattman: Yeah. A lot of ill is done in the world by having people go and be in charge of policy in a place they’re not familiar with. Which could be in their own city.

Rob Wiblin: Yeah. So before being too humble, do you think that if you were combined with someone who knew a ton about the specific Ukraine-Russia situation, that you would be able to add value because of the frameworks that you bring to thinking about these things? Like you might be able to structure their thinking somewhat better than they might, if they’re just someone who follows politics in that region?

Chris Blattman: Well, I think yes. Only because just not knowing much about Ukraine but having a lot of practice at helping people analytically organize arguments and think and remind people of strategic incentives, I do add value. I blog and tweet and write articles a lot, and I avoid things when I’m just too unknowledgeable. But actually, I feel totally comfortable doing this — not because I am a specialist, but just because there’s not enough people calling attention to these psychological and strategic logics. Now, it would be even better if I knew something about it, but here we are.

Rob Wiblin: Yeah. We can only know so much.

Great power wars [02:01:46]

Rob Wiblin: Let’s turn to what practical lessons we can potentially take away from all of the above political science thinking. I’m involved in the effective altruism and longtermist research communities. In those groups, we have this special desire to reduce the probability of a great power war, because we think that’s the kind of war that could have really big and persistent effects that could make the future worse for many generations, even lead to extinction. And for more or less the same reason, we’re particularly concerned about the use of nuclear weapons or other unconventional arms, such as bioweapons or cyberattacks, and also future technologies that could be particularly destructive, such as weaponized AI.

Rob Wiblin: Actually, something we haven’t talked about much on the show, but is interesting and hopefully will come up in the future, is that there’s this fear that an AI that’s programmed to engage in retaliation or to use blackmail or to engage in random escalation for strategic reasons — for all of the game theory things that we’ve been talking about — that that could be disastrous one way or another in its interactions with people or other AIs. I guess both, if it’s programmed to do that really well. And also if it’s programmed to do that very poorly, or at least it does it in a way that’s inconsistent with the way that other actors on Earth behave.

Rob Wiblin: Anyway, setting that aside. The problem that we face is that having identified that the above is really important, we find it really hard to find projects that one can engage in that might actually move the needle on any of them. It’s like, “War with China is bad. Yeah, yeah, yeah. OK, so now what does that imply?” It’s kind of not obvious. Now, your expertise is maybe more on the civil war side rather than great power wars, but I’m particularly keen to ask you a whole lot about this, because I expect other podcast hosts won’t be as aggressive about talking about great power conflicts.

Rob Wiblin: To start off with, are the drivers of great power conflicts different, on average, than the drivers of smaller wars, as far as you can tell?

Chris Blattman: A lot of the things that occurred to me are questions of scale and difficulty of addressing in a short timeframe, rather than fundamental theoretical differences and what the diagnosis is. Partly because every war is its own little unique flower or snowflake, and so you have to diagnose each time.

Chris Blattman: I think if anything, there’s a common set of very difficult-to-resolve roots to great power wars. Which is that in a city, we can imagine a criminal superstructure organizing the gangs and being that third-party guarantor of peace and security and commitment, and punishing unchecked leaders and solving the five problems. And at a city or even a national level, states perform those functions pretty well. We’ve proven pretty good at constructing those institutions. And as an individual, if I decided I wanted to really make a contribution to those, I could probably make a difference on the margin.

Chris Blattman: On the other hand, it’s very hard for anybody, even the US president, to think, “How could we change the fundamental structure of the UN Security Council?” So there’s a scale of human institutions where we don’t yet have the solutions, and they’re so big and slow and subject to lots of forces and competing actors that they’re hard to manipulate. There’s this fundamental problem of anarchy, meaning there’s nobody above to settle these feuds and resolve these five problems, and there’s a difficulty of tackling that. We still have tools, but we just have to be —

Rob Wiblin: Realistic.

Chris Blattman: Yeah. We have to be willing to work on these low-return margins.

Rob Wiblin: One benefit that we have with great power conflicts, at least compared to civil wars, is that very often when you’re ending a civil war, you’re asking one or both sides to disarm, and that creates this very thorny commitment problem.

Chris Blattman: Absolutely true.

Rob Wiblin: Where if you’ve been fighting a war, and now you’re going to give away all of your weapons, you’re leaving yourself extremely vulnerable to the other side then retaliating against you once you’ve given up basically your power within the bargaining situation. And that doesn’t happen within great power conflicts these days. No one really thinks that the US is trying to colonize all of China or vice versa.

Chris Blattman: That’s true. And that makes great power wars easier to end than civil wars.

Rob Wiblin: I guess you might think that the higher the scale, the more it’s the case that personal idiosyncrasies would be canceled out, or that a huge bureaucracy in a country as large as the US or China or Russia wouldn’t allow one person’s tastes and emotions to guide things.

Chris Blattman: If only. Right? It is an amazing and unfortunate feat just how much Putin was able to personalize power, especially in the last 10 years. And it’s something that Xi Jinping is apparently attempting to do. This is another country I’m not by any means an expert on, but an attempt to further centralize and personalize power in China seems to be underway, and that too would make us, the world, a less stable place. In China right now, power is very institutionalized, very widely shared amongst many actors. So it is a very checked leadership, by the standards of an autocracy. But that worries me the most of all, of all the current global shifts.

Rob Wiblin: That that’s decreasing over time?

Chris Blattman: Seems to be. Or there’s an attempt, and he could be quite successful.

Rob Wiblin: Yeah. We really worry about centralization of power within China and Russia. But something that really stands out to me is that — maybe more, we’re talking about countries in which listeners actually live, because they actually have some potential to shift things — it used to be the case that in the US, you needed Congress’s approval to declare war on another country. That has just kind of disappeared, because seemingly Congress doesn’t want the responsibility of declaring war. They don’t like intervening in foreign policy, because then they can be blamed when things go wrong.

Rob Wiblin: So they’ve just kind of delegated power to the president, with almost unlimited authority to do things militarily and in terms of foreign policy. And I really do wonder whether it wouldn’t be better if we could somehow find a way to bring Congress back into the picture of deciding whether the United States goes to war. At least wars of discretion.

Chris Blattman: I think some more leeway has been given to the executive in the United States. I wouldn’t go so far as to say they’ve completely given up that power. I agree that’s worrisome. The US is the one country that has probably taken decentralization a little bit too far. Maybe not from a conflict perspective, but just from a capacity to get anything done. I have colleagues who have thought a lot about this and think the US needs to be a little bit more executive, a little more centralized. That would be fine, from my perspective. That’s not the thing most countries in the world, even advanced democracies, have to worry about.

Chris Blattman: Even Canada, where I grew up, is a surprisingly unchecked place, at least by formal political institutions. You have a very weak upper house that’s appointed, and so you have really just a single legislature. And if you get a majority in Parliament, then as a prime minister you have an enormous amount of power, and there’s no filibusters. It’s still a pretty federal place, so it’s not totally unchecked. But actually, it’s a little bit surprising to me how stable Canada is, given that. The only way I explain it to myself, so it’s a hypothesis, is that I think the informal institutions — informal checks and balances, what people are willing to tolerate — are really, really powerful in Canada. So it’s all these hidden institutions that check the power of the executive, or in this case the prime minister and the majoritarian party.

Chris Blattman: So there’s lots of ways to check power, is one lesson. But I do think we’d be a safer world if even the more advanced democracies were a little bit more checked.

Rob Wiblin: Yeah. So another of the five factors that I can make an argument that it might not loom as large in the great power war or nuclear war situations is uncertainty. Because the US, China, Russia, they have such advanced espionage capabilities, such impressive satellites. Plus just lots of things are public about the military power of countries that are not, say, about the power of a gang, so that arguably there’s less need to engage, or bluffing is less plausible, because the other side can just see whether you have the aircraft carriers or not. And maybe that helps us.

Chris Blattman: There’s an element of that. At the same time, I think we and our policymakers and our presidents manage to have really oversimplified views of what’s going on in these other countries. I think we can all imagine what those are, the oversimplified views Americans hold of Russia and China.

Chris Blattman: One of the things I’ve heard about Russia is the extent to which the political class and this inner circle around Putin really bought into the rhetoric that the US is polarized, going to have a civil war. You heard a lot of American press in the last year on the likelihood of an American civil war. And I wrote a little bit about how the danger of overestimating that risk, and the paranoid and I think alarmist stuff that was going around. The danger was that other world leaders might start to believe it even though it’s false, because it’s kind of in their interests. And so they’d think the US is weak, ununified. Time to make a gamble for Ukraine or something else. Or time to make a gamble for Taiwan.

Rob Wiblin: Right. Right. Yeah. I suppose Taiwan’s a good example where the US has kind of decided to take the bluffing approach, which may or may not be the right idea. But China doesn’t know, and I guess maybe even the US doesn’t know, whether it would defend Taiwan. It seems like we’re kind of going to roll the dice if the time ever comes.

Chris Blattman: Yeah. I don’t know how to think about what the right thing to do there is. Again, from this perspective of strategic empathy, I try to look at this from China’s perspective and I say, well, what if after the end of the US Civil War, the Confederacy had gone down and occupied the Florida Keys, and it just wasn’t worth it or possible to exterminate them. And so now in southern Florida, there was still [this group]. Maybe it’s not even a slave state any longer, but maybe it’s a repugnant system to us, or a threatening system.

Chris Blattman: But even if that’s not true, there’s no world in which Americans are like, “Yeah, that’s a legitimate political entity.” You’re like, “No, they lost. They’re totally illegitimate and they need to go, and the Florida Keys are ours.” And I’d be super sympathetic to that. So I can see how the Chinese regard this, and so it’s a dangerous game. But at the same time, the thing that we’re standing up for in Taiwan is, I think, very noble, which is democracy and liberty. So it’s really hard.

Rob Wiblin: Yeah. One of the interventions you discuss in the book that doesn’t seem that applicable to preventing great power wars is sending in peacekeeping forces.

Chris Blattman: Yeah. Totally inapplicable. Yeah, yeah.

Rob Wiblin: Another one that’s difficult to use is having overarching rules to police state behavior.

Chris Blattman: That could work.

Rob Wiblin: Oh, really? OK.

Chris Blattman: We do have rules. The International Criminal Court is a great example of an attempt to build in a set of impartial rules that no one nation controls, and it’s actually potentially strong enough and capable enough and impartial enough. That’s the main reason the US is one of the only countries that’s not a signatory.

Rob Wiblin: I see.

Chris Blattman: So it is a little bit rich right now for Biden to be recommending International Criminal Court proceedings against Putin, when the US has refused to sign on. So I think it’s possible to create impartial rules.

Rob Wiblin: But I guess the fact that the US doesn’t sign up is a sign of how it’s challenging, once you’re a powerful enough country.

Chris Blattman: Yeah, although it’s been working reasonably well in spite of that. And I can imagine a world where they do sign on at some point. Likewise, you can imagine more like right now: the sanctions reaction, the sanctions unity and regime towards Russia is super selective. I don’t want to call it hypocritical, but it’s in the West’s interests. We’re using it to punish or deter one invasion and one set of actions. There’s all sorts of worse things that have happened in the last 20 years where we didn’t do that. And there have been times when the US has gone and invaded other countries and evaded [sanctions]. Any set of sanctions rules that were more impartial, that punished any power, great or small, for such a thing I think would deter war. And I think that’s not inconceivable.

Rob Wiblin: Yeah. OK, so I suppose it is possible to create structures that partially constrain great powers, or at least that raise the costs of bad behavior for them, even if it doesn’t actually prevent them from doing those things.

Chris Blattman: Yeah.

Rob Wiblin: That’s the basic idea. And maybe you have this issue that you could sign treaties that you could then get bound to, and even if later on you don’t like it, you’re still kind of stuck with it.

Chris Blattman: Absolutely. Elinor Ostrom, who I talked with in the book, and Amos Sawyer, a Liberian political scientist I talked to in the book, talk about something called “polycentric governance.” Which is one way of saying decentralized governance, but not just decentralized across branches of government or decentralized downwards, but supernational governance. So international treaties. Every time we create a set of international treaties and agreements that do bind us and do come with penalties attached if we abrogate them, then that does bind the hands. That’s another way to check rulers.

Rob Wiblin: Are there any particular interventions to reduce the risk of great power wars that jump out as more relevant than the others?

Chris Blattman: Great power wars are the thorniest, because it’s very difficult to distinguish an intervention or institution that’s done to pacify the situation from one that’s just trying to elevate the bargaining power of one side. So the only way to think about it is: if you have two great powers in conflict, what can the third great power, if neutral, do to mitigate the risk of conflict? There, I think, there are some tools. But otherwise, everything you do is really just a weapon of war or a weapon of competition.

Rob Wiblin: Right. So is this this kind of offsetting issue that I was talking about earlier?

Chris Blattman: Yeah. What would I think? I haven’t thought this through, so I don’t know. Imagine there’s a great power war brewing between China and India. What can you do, as a third party, to deter that? Well, I think you could go and you could tell them that any aggressive actions, any first movers, we are going to punish with sanctions. You could try to use what tools you have.

Chris Blattman: You could also try to mediate, which they would. This is what Israel and Turkey are trying to do in the context of Russia and Ukraine, is take a very costly stance, which is to maintain impartiality and try to help these two sides bargain. And a third great power could do this. But unless there’s some third powerful party that can do these things, even if it’s an over-powerful party, then there aren’t any options.

Rob Wiblin: Because just no one has the resources to meaningfully shift the incentives.

Chris Blattman: Right. Or you need to construct this international architecture that constrains everybody, or at least constrains the great powers.

Rob Wiblin: So you have a bunch of very powerful countries, but you could have an alliance of smaller countries that then might be able to act as if it was a larger country. But I suppose doing that is kind of just adding another great power to the mix, and now you’ve got an even more multipolar world, which sounds worse, potentially.

Chris Blattman: We talked about how it cuts both ways, but I do think that ideas of universal human rights and brotherhood and sisterhood is pacifying, in the sense that we have sympathy for others and so we’re less likely to go to war with them. I think the economic entanglements we have through trade are very protective. So I think untethering ourselves from the Chinese economy too much or too rapidly is not stabilizing, even if it achieves other national goals. That’s something Americans can vote on and decide on.

Chris Blattman: Entangling China in our lives as much as possible through immigration and trade and social exchange and watching their television shows and making sure they get some Netflix or something, are all going to, on the margin, I think help a bit. They’re not magic solutions, but that strikes me as very important.

Rob Wiblin: Entangling economies and cultures and people is one of the interventions that you talk about a bunch in the book. That’s one that I was very optimistic about, back in 2015. It seems it’s one where people have noticed that the entanglement maybe does make war less likely, but it also makes you vulnerable to the other side. It also potentially weakens your position against them, at least if you’re buying military equipment from them or something like that. There’s some that have a greater problem there than others.

Chris Blattman: Yeah. It’s complicated. There’s some countervailing effects. Like why is India not united with the West on this? Well, there’s lots of reasons, but one reason is they depend on Russian military equipment for their entire military apparatus, et cetera, et cetera. But I think on average, it’s pacifying. But people also have had a tendency to exaggerate this in history. There’s lots of Enlightenment thinkers saying, “Oh, the trade and liberty and human rights are going to extirpate war forever.” And that didn’t happen. But it’s not like an on/off switch. It can still be helpful.

Preventing nuclear war [02:18:57]

Rob Wiblin: Are there any interventions that stand out to you in the context of preventing a major nuclear war?

Chris Blattman: I think these arms reduction treaties and any effort and the invention of these international atomic energy associations to monitor these things is maybe one of the most powerful things we can do to systematically disarm. I think anything that improves the quality and rapidity of communications between world leaders are all things that we’ve been doing. What we would do further, I’m not sure I have a good idea. It’s not something I know very well.

Rob Wiblin: It seems like in the past, conflicts between or within smaller countries have sometimes prompted larger countries to end up fighting, because they start a proxy war there and then that potentially escalates. I guess one thing you could take away from that is that actually, preventing smaller-scale conflicts is also potentially useful in terms of preventing great power wars because they can escalate. I guess your model could also suggest that it’s an illusion that the war in the smaller country was actually causing the war between great powers. That just was the proximal cause, but not the deep underlying cause. Do you have any thoughts on that?

Chris Blattman: I’m trying to think of a concrete example. You might think of World War I as an example where there’s this Balkan conflict. But there, I think it’s more of an example of what you said, which is that there’s actually a fundamental strategic challenge there presented by Russia’s rise and Germany merely needed an excuse. I guess in most cases, it would be my suspicion that these proxy wars escalate to larger wars partly because there were fundamental disagreements, rather than just some natural escalatory dynamic that can mechanically happen. I think I’m more skeptical about that.

Rob Wiblin: Another case that is in the book is the Seven Years’ War, where you describe the provocations by the Americans as potentially prompting that. But I imagine if there was another reason, that they would’ve found a way to avoid that.

Chris Blattman: I think so. I actually don’t know enough about what was happening on the European side. I suspect the flareup in these little interior forts between the French and the American colonists was maybe a trigger in the same sense that the assassination of Archduke Ferdinand was a trigger, but I doubt it was the whole thing.

Rob Wiblin: Yeah. I mentioned earlier a way in which, perversely, electing a leader who’s extremely oriented towards peace could potentially not make things safer, but it seems like if every country has leaders who are deeply committed to peace, to whom war is repugnant, that does lower the risk of war — because both sides would be willing to concede more rather than fight, so you’re expanding the range of deals that they think are better than war.

Chris Blattman: Any generalized notion of human rights and the value of human life and the value of the other stranger is going to be pacifying so long as everybody possesses it. I’d have to think about it more. It might even be pacifying if only some people do it. It would reduce your bargaining power in a sense, because you’re not going to be willing to use war. Anything that makes war more costly to you, but not to your enemy, is going to expand the bargaining range in ways that are disadvantageous to you.

Chris Blattman: And that’s a real key. Whenever we think of things, you have to think about: is this something that’s elevating the risk of war, or is this something that’s just changing relative bargaining power? They’re distinct things. And a lot of things that we think, a lot of people say, “Well, don’t counterinsurgency tactics cause war? And new technologies cause war?” And I’m like, no: that shifts relative bargaining power, and then you have to re-bargain.

Rob Wiblin: Well, new technologies create uncertainty, so there’s that channel.

Chris Blattman: There could be that, I agree, but they’re principally shifting bargaining power.

Why undirected approaches won’t work [02:22:51]

Rob Wiblin: OK, yeah. Do you have any other examples of things that people think might help in this context, but it’s ambiguous, or they don’t help?

Chris Blattman: Let me use an example from cities who want to reduce violence — say, intergroup violence between gangs, or interpersonal violence between individuals. In Chicago, they’re like, “We need to economically develop the South Side.” And I’ve been on radio programs in Chicago and people are saying, “Well, you’re doing this, but what we really need are jobs. Jobs, jobs, jobs, jobs, jobs.” And that doesn’t make sense to me for a few reasons.

Chris Blattman: Number one, for reasons we talked about earlier, making the pie larger or smaller doesn’t change the fundamental calculus. The fact that these gang members have jobs available might not change the fact that there’s whatever they’re fundamentally competing against. And we could debate that. What’s more important is just that it’s totally untargeted, right?

Chris Blattman: Chicago’s violence problem is 2,000 or 3,000 people. Any strategy directed towards 500,000 people — especially a job creation program that is probably never going to reach any of these extreme people — is a bad idea. What’s the analogy in the international arena? It’s any generic foreign aid that we give to poor countries that’s not actually specifically targeting the roots of violence and the diagnosis of what’s going on there, but it has this generic, poorly thought-through feel of, “Let’s just help Africa get richer and then there’ll be less war.” And it’s probably not true. It might be good to do that for other reasons, but it’s probably not going to reduce violence. You really have to be targeted. I think that’s the lesson from cities to nations: that you have to really think about who are the violent actors and decision-makers, and focus on them.

Rob Wiblin: Yeah, you have a section about how it seems like offering cognitive behavioral therapy to at-risk youth in that sort of context can significantly reduce the probability that they engage in murder, which is very cool. I was thinking in the great power context, can we pay Biden to go and get CBT therapy? Or how much would you have to pay Xi Jinping to go to a course? I actually think, even if you could do that, I’m not sure that it would help, because there’s the context where these decisions are not made in the heat of the moment in the same way that gang shootings are.

Chris Blattman: I think you’re right. I think a lot of CBT around violence is doing a few things. One of the things is that it’s trying to help you cope with that hot reactive violence, and that’s not the thing that we think is driving a lot of wars. Maybe personalized dictators. But the other thing it’s doing, is a lot of CBT — and this includes CBT like basic marriage counseling — [targets] all of these misperceptions; these availability bias; these projection bias; these hard, rigid, poisonous views that we have of our adversary; our tendency to interpret their actions in the worst light possible almost automatically without really having an ability to see it from their shoes or have any strategic empathy. Those are really, really, really common problems of interpersonal conflict as well, so it’s not just passions. So that is what CBT is really good at helping people get over.

Rob Wiblin: Right. We’re actually recapitulating a bunch of stuff that I discussed with Robert Wright last year. He has this movement where he is trying to promote peace and reduce conflict, which involves getting everyone to slow down and meditate and do CBT and so on. And I was like, “It seems very scattershot. Most people aren’t going to war anyway. Encouraging someone random to go and do CBT doesn’t seem like it helps.” I was like, what about targeting it on the State Department? Get everyone in the State Department on a CBT course or perspective-taking. That seems more targeted and more useful to me.

Chris Blattman: That’s what we call effective bureaucratic design. Especially in a diplomatic organization, you want people in the State Department and National Security Council to be really, really, really good at perspective-taking, strategic empathy. Really, really good at recognizing their biases. That’s creating an institutional culture. And you want them to question bad information rather than punish them for it. So there’s some principles of organizational design that I think are really important that are less about CBT.

Rob Wiblin: And more about rational analysis, improving the analysis?

Chris Blattman: Yeah, and just avoiding some of the problems, some of the group dysfunctions.

Rob Wiblin: Yeah. I think it will be hard to pitch that to, for example, the State Department, on the basis that it’s pacifying. But it could be possible to pitch it on the basis that it improves the quality of their recommendations because you understand the other side better. It allows you to actually get more of the pie.

Chris Blattman: And I actually go further. I think many of our national security leaders are very aware of this and they deliberately design practices. They have things like red teams whose job is to take down the analysis. I think President Obama in particular was known for, and even George W. Bush as well, going around the room and, “What do you think? What do you think? What do you think? What do you think?” — consciously not telling people what they think ahead of time to not bias. They wanted to get less varnished opinions. There are lots of things as a leader, as an organization, you can do and institutionalize to reduce this. And I think actually a lot of our national security apparatuses in every country are pretty good at that.

Rob Wiblin: Yeah.

Chris Blattman: That’s maybe what my colleague thinks has fallen apart in this insulated, personalized dictatorship in Russia.

Rob Wiblin: Over the last few months, I’ve had some people get quite angry with me when I’ve tried to explain Putin’s motivations using this kind of rationalist strategic analysis. But my goal is to say, “No, this is good. It’s good to do this. You have to set your passions aside. You have to set your moral convictions aside and aim for understanding.” But I may risk losing my moral compass, or both of us are at risk of losing our moral compass sometimes and becoming too compromising. And indeed, I guess, it could weaken your strategic bargaining position if you become too sympathetic to the other side through this method of trying to understand them better.

Chris Blattman: I guess if your intangible incentive is we have to punish wrongdoers.

Rob Wiblin: And people who don’t share that, and putting themselves in a mindset of not sharing that.

Chris Blattman: Yeah. Punishing wrongdoers no matter what is a little bit of this intransigence that will probably continue the current fight, and then maybe increase our bargaining power in the future. It would be fine if people were doing that in a considered way. I do think that if those same people were getting just as mad over Zhan Zhang, and getting just as mad over 12 other things, if they could tell me about the horrible things that are going on in Ethiopia, if they were internally consistent, that would be a lot more credible.

Chris Blattman: But when it’s an ally, and if it’s an ally in Europe, and it’s an ally who has the same skin color as me, and all of these other things that make us a little bit more prone to get mad about this human rights abuse and not others, I think we need to be very suspicious of it. And sadly, I’ve yet to talk to one of these people who feel so strongly that they can even tell me the first thing about what’s been going on in Ethiopia for a year and a half, so I think I am right to be suspicious of that.

Rob Wiblin: Yeah. I have somewhat tracked what’s going on in Ethiopia. It’s an interesting case where it might actually be much easier for us to lean on Ethiopia, because they’re more of an ally and much less powerful than Putin, so you could actually do a lot more good in terms of human rights, potentially, by creating strong incentives for groups that want to be aligned with you.

Chris Blattman: Yeah. I know a little bit, and I know some people who know a lot, and the alternative is that the US government could actually have fumbled and done the opposite by giving too much support to the Abiy regime and demonizing the Tigrayans too much. Neither side is acting particularly well, but I do think there’s a way in which we look back on this and where one of the things that led to this conflict was the US government actually not making a concerted effort to help these parties find agreement, and maybe even doing things that are counterproductive.

Democratic peace theory [02:31:10]

Rob Wiblin: Yeah. An audience member wanted to ask your views on democratic peace theory. I haven’t spent much time on that because it seems like our ability to change whether great powers are democracies or not is somewhat limited, so it’s not clear what the intervention is. But how large do you think the effect is of being a democracy on reducing your warlikeness?

Chris Blattman: This is an area where I think that the theory and a lot of the casual empirics are really strong, and then the hard tests are pretty rare or difficult to believe fully. But I’m partly convinced on the basis of the idea in this case. It goes back to this broader point about overcentralized power being the root of accentuating all five kinds of conflict. Mechanically, it’s going to make leaders more checked and less likely internalized costs, but it means we’re at the mercy of their misperceptions, we’re at the mercy of their intangible incentives, there’s more uncertainty, they can’t commit — so it just aggravates all these sources of war. And that is, to me, a very plausible rationale for why highly checked-and-balanced societies very seldom go to war against one another, and when they do go to war, it’s usually against a relatively unchecked society where they can’t find a mutual bargaining range.

Chris Blattman: But I’m careful to talk in the book about these checked-and-balanced societies rather than democratic — because democratic, most people just leap to thinking, is the president elected or not? But then, there’s a lot of countries, like Liberia, where you’re electing a dictator every five years. That’s a slight exaggeration, but they’re not that checked and balanced; they have a lot of autonomy.

Chris Blattman: So an election for the executive is actually not the most important check, from my perspective, for peacebuilding. It’s actually the division of power within the country, across branches of government, across levels of government in the country, and to these supernational organizations. Anything that does that is going to make things more peaceable. So I totally buy democratic peace theory, but the idea that it’s just some simple, “If we all had elections for president, we’d have a peaceful world” — that’s not true.

Rob Wiblin: Yeah. I think people talk about thick and narrow conceptions of democracy, and I used to think that the people who were trying to use a thicker definition of democracy were slightly cheating; they just define everything that they like as democracy and call it that.

Chris Blattman: That happens too.

Rob Wiblin: I think there’s something to that, where it’s just like, anything that is good is democracy, anything that isn’t isn’t. But I’ve come to think that the thick conception of democracy, where it’s spread-out power, people have more information, is actually the more useful concept than the narrow one, where it’s just, do people vote at some point?

Chris Blattman: Right. That’s why I think China’s a relatively stable place. It’s not democratic, but it’s a relatively checked-and-balanced place. It could be more, of course, and that would make the world a safer place.

Rob Wiblin: I’ve heard a decent amount of skepticism about democratic peace theory. The argument goes that democracies like the US, it’s not that they don’t engage in wars; it’s that they engage in wars that they can win because they have better decision-making apparatus. They’re less likely to make massive mistakes, but nonetheless, they have interests, they’re in conflict with other countries, and sometimes they end up fighting them and winning them and getting the thing that they can get.

Rob Wiblin: Also, the US isn’t at war with Venezuela, but it’s engaging in awfully crushing sanctions that kill a lot of people, and it seems kind of like war, right? It seems like borderline war. And they’re doing that because they can get away with it and it’s in their interests. So there could be a whole lot of quasi-war behavior that democracies engage in, or brief wars or intimidation, where the other side is stood down on the basis of threats of violence that are going on, that are not getting counted here.

Chris Blattman: I would say, yeah, strenuous hostile competition and the wielding of hard and soft power in order to obtain objectives using every means necessary other than armed invasion is totally normal. And what distinguishes that from most war, in one sense — other than the explicit use of armed violence — is the fact that the whole thing that made most kinds of war violence not happen was that it was super costly, and there was a better way. It’s not clear that there’s a better alternative to these other tools that democracies and other countries wield.

Chris Blattman: I’m not saying that democracies don’t pursue their own interests, that they’re not self-interested. I’m suggesting that mostly, because their power is checked and balanced, they don’t engage in activities to obtain what they want that are really, really costly to their own people. So that doesn’t rule out bad behavior. If we want to say, well, is there democratic harmony? No, I don’t believe in democratic harmony. I believe in democratic peace, in the sense that they’re far less likely to actually go to have this highly inefficient thing called warfare.

Rob Wiblin: Yeah. This raises an interesting conceptual point, that the reason we expect countries not to go to war is that it does a lot of harm, and a lot of people suffer enormously. But now, let’s look at the US sanctions against Venezuela. Potentially, many more Venezuelans are dying due to the sanctions than might die if there was a full-on invasion.

Chris Blattman: Yeah. In this strategic story we’re telling, you don’t care about the costs to the other side of your actions.

Rob Wiblin: But it is costly to the US because they can’t buy Venezuelan oil, right?

Chris Blattman: I’d say it’s not particularly costly. If Venezuela had a monopoly on oil, that would be true and we would not be having sanctions on them.

Rob Wiblin: Right.

Chris Blattman: And the elasticity of the supply of oil is pretty high. So it’s true, there’s some people paying. There’s all these refineries in the southern US that were built to process Venezuelan oil, and they’re running silent or not running. So there’s costs to pay, but Americans are not bearing the burden of these sanctions.

Rob Wiblin: I see. So in the accountability story, the point is that they haven’t gone to war in a way that would harm the United States because it’s a democracy and leaders would lose if that happened. But it’s totally consistent with the story for them to harm people in other countries because Venezuelans don’t get to vote.

Chris Blattman: Yeah, and we’re quite willing to support conflicts, like many aspects of the drug war, if we can export some of the costs as well. That’s true of any group. We might have a very different drug policy if we had to face some of the consequences that Mexico and Colombia have to face.

Exchanging hostages [02:37:21]

Rob Wiblin: Slightly facetious question: back in history, during the 14th, 15th, 16th centuries, my understanding is that countries would sometimes effectively exchange hostages — they’d exchange members of the nobility in order to make it more costly to declare war on the other country, because then the hostages would be killed or at least would be at risk.

Chris Blattman: Yeah. They would also marry each other’s children to one another, which is another kind of hostage taking, probably, in the eyes of some of these princesses.

Rob Wiblin: Of the people who were married off. Yeah. I’m not sure that they were getting a vote on that. It doesn’t seem like we’ve come up with an amazing list of things that I feel like we can fund in order to stop the risk of war between China and the US. But given that both sides in that conflict would regard it as catastrophic to have a war, I imagine, is there —

Chris Blattman: They’re building all our iPhones. If you said, “Do we want five Chinese hostages or all of our iPhones to get built in China?” then I’d take the iPhones.

Rob Wiblin: So that’s the interdependence argument?

Chris Blattman: Right. That’s what the marriages were trying to do.

Rob Wiblin: So we wouldn’t actually take hostages now, but you could have an agreement where many very important people from both countries go and live in the capital city of the other country, so effectively, they’re in the firing line if things go wrong.

Chris Blattman: Right. I don’t know how exactly this works, but the fact that the US holds the gold reserves of many countries is maybe a little bit like hostages as well. I’m not sure how that works legally, but we’re all a little bit more interested in this now that they seem to hold all of Afghanistan’s money. So that is a tool of leverage.

Rob Wiblin: Yeah. The US is trying to disentangle itself from the Chinese economy to some degree. And I guess it seems like China is disentangling itself, in some ways, from the rest of the world. Do you have any comments on trade policy here? Are you against that on balance because it could make the risk of war higher?

Chris Blattman: I don’t actually know how much of it is actually happening or if it’s just decreasing a trend. I don’t know the numbers. I’m sympathetic to the idea that I think the US has a lot of domestic political problems and inequality because globalist elites like me rushed very quickly to seek a more integrated world economy — and there were winners and losers, and then we forgot to compensate the losers. So that’s the source of a lot of political ills in the country right now to some degree. Addressing that is just smart, stable domestic politics. I have to think that can be balanced against the desire to remain entangled with China. And frankly, I’m not that worried. I feel like the incentives to remain entangled are so high, I’m not super worried about a huge rollback.

Rob Wiblin: I see. Yeah.

Chris Blattman: Maybe I should be, but …

Rob Wiblin: I guess to wrap up, there is this controversy among people who I read about how optimistic we should be about war having been reduced over the last 70 years, and what are our prospects of major war over the next 100 years. Steven Pinker is famously fairly optimistic about this, takes a fairly positive tone. Bear Braumoeller, who we were talking about, wrote this book arguing back, saying the dynamics that cause wars to start and escalate really haven’t changed so much. They’re both really smart people I respect, who have some statistics on their side. Where do you come down on this?

Chris Blattman: Of course, in the middle. That’s where I’ve been coming down on a lot of these things. Maybe I lean a little bit. I do think our architecture, institutionally, and a lot of things that Pinker talks about — states and norms and humanitarian revolution — I do think those have moved the needle a lot, and I think the destructive power of our weapons have contributed to that. But I don’t share an unadulterated optimism. I think the warnings from someone like Bear and many others are absolutely correct: that we can be a little bit safer than we were in the 20th century, but the 20th century had some pretty bad moments, so we shouldn’t be complacent.

What you can actually do to help [02:41:25]

Rob Wiblin: Yeah, it makes sense. All right. We’ve already gone over time, so I should let you go. To wrap up, I’m left a little bit unsure here exactly what I would want to fund, or what the career implications of all of this might be, although I think it’s really going to help the audience think a lot more clearly about these war and conflict issues going forward. Do you have any more concrete suggestions for what listeners could take away, in terms of actions that they could take?

Chris Blattman: Sure. This is certainly a thing I grapple with myself. I guess I get up every day and work on this because I think it’s possible to chip away at the margin. And I wrote a whole chapter about that, because that’s the way I wanted to end the book in a concrete way. So rather than just have, “Everything’s getting better,” I was like, “OK, here’s how to operate better.”

Chris Blattman: One thing that I’m trying to do, but I think a lot of people can do, is I think it’s really important to change the conversation. I think it’s possible for individuals to shape how people think about and talk about conflict — and that can be in their city, very easily. That could be internationally. I do think a lone voice in Congress, even if it’s an aide who reminds people that this is costly and not the natural thing, or reminds people of some strategic dynamic or some psychological factor — or somebody who makes the State Department just be a little bit more accountable, a little bit more aware of biases — that would be a huge contribution.

Chris Blattman: I think individuals can do that kind of thing all the time. You can change conversations. And you can do that in any field: you can change the conversation in poverty alleviation, you can change the conversation in public health. It’s actually worrisome how easy it is to change the conversation, because I think it’s unsettled.

Chris Blattman: Then the question is, how do you get there? That’s the thing that happens in a lot of my office hours. People are like, “I want to do this. What are the concrete steps that I take?” Most people don’t really think about knocking on the doors that nobody else knocks on. If I wanted to actually change the conversation or change the mind of somebody, honestly, if I were 22 right now, I would just start emailing deputy ministers in 24 middle- and low-income countries where I could operate in the languages I know — just to be their aide, because they don’t have a lot of human capital and no nobody ever emails them. Everybody’s applying to whatever UN internship or something like that, but nobody’s reaching out to these people.

Chris Blattman: There’s dozens of cities on the planet where no one knocks on those doors, and I think you could make a real difference. If you’re a Western person, I think you’re just bringing a certain sort of style of thinking and a level of education, a level of human capital that is partly there, but not fully there. And that’s the theme through a lot of what I do. Why don’t I work in Russia? Everybody else is working in Russia. Nobody is figuring out organized crime in Colombia, as far as I can tell, so I’m going to work there. I’m going to be one of a very small group of people who are chipping away at these problems to change that conversation. And I think that’s been hugely effective.

Rob Wiblin: My guest today has been Christopher Blattman. Thanks so much for coming on The 80,000 Hours Podcast, Christopher.

Chris Blattman: Thank you.

Rob’s outro [02:44:41]

Rob Wiblin: I also wanted to plug another new book out this week by a former guest and friend of the show — A.J. Jacobs.

It’s called The Puzzler, and in it he goes on a journey through the world of puzzles, exploring why we love them, what they do to our brains, and how they can improve the world.

You can hear him talking about it at the start of his episode of this show: #79 — A.J. Jacobs on radical honesty, following the whole Bible, and reframing global problems as puzzles.

Keiran read an advanced copy, and is willing to put his reputation on the line to say that it’s a really wonderful book! You can find a link in the blog post for this episode.

One other update from another previous guest and friend of the show — in this case, the academic Philip Tetlock. Phil is now launching a new forecasting tournament, where they’re pairing together experts on specific existential risks — like pandemics, risks from AI, nuclear war, climate change, and so on — so they’re pairing people who are very knowledgeable about those existential risks are paired with people who are very knowledgeable about forecasting specifically, to see whether a combination of those two different forms of expertise can produce better predictions about how the world is going to play out in future than either one of them individually.

That research project is called the Hybrid Forecasting-Persuasion Tournament, and they’re currently recruiting a whole lot of people to participate, both as experts in forecasting and as experts in existential risks specifically. Philip says that some key features of the programme are: having really carefully assembled collections of early warning indicators for risks; looking at the interplay between smart generalists and subject matter experts; and the measurement focused not only on forecasting accuracy, but also on the power of the explanations for forecasts to move other people’s views hopefully in the right direction.

I think a lot of listeners to the show could potentially really be perfect to get involved in that. So we’ll stick up a link to more information about the nature of the tournament and how you can apply to potentially be involved in the blog post associated with this episode.

All right, The 80,000 Hours Podcast is produced and edited by Keiran Harris.

Audio mastering and technical editing by Ben Cordell.

Full transcripts and an extensive collection of links to learn more are available on our site and put together by Katy Moore.

Thanks for joining, talk to you again soon.