The Collapse of Complex Societies

"This was hard. What I was trying to do is find a book that gets at the idea of rise and fall, which is intrinsic to economies. You could easily read some economic histories about Japan in the nineties, maybe even of China today and some of the struggles around the real estate industry there, to get a sense of a boom and bust. That comes with industrial policy. I prefer Joseph Tainter’s work in The Collapse of Complex Societies , because it’s anthropological, a field we haven’t had on this list so far. (Alice Amsden is economics, The Politics of Large Numbers is from a sociology of statistics perspective, Lowen is coming from a historical point of view, and Kicking Away the Ladder is development and institutional economics). Now we’ve got anthropologists and we’re actually looking at individual humans and how they respond to things. Tainter has a very simple theory, which is that most social systems start very simple, and we benefit a lot when they become complicated. We live in a modern economy where everyone hones a different skill. You are doing this interview, I am reading books. We were talking about a soufflé—a chef is making incredible food. I don’t know how to make a soufflé, but we both benefit from the fact that I make money somehow, you make money somehow, and we’re able to pay someone else to do something we can’t do. There are benefits to complexity. We actually want to create more complexity, more subdivisions of labor, more expertise, because then we’re able to trade it and do something magical. But Tainter argues that there’s a certain point in almost any of these economic systems, or any society, where we go over the peak of benefits and start getting more and more uselessly complex over time. There’s a popular argument in progress studies and technology circles today that we can’t build anything in the United States. We want to build a subway line in New York and we just opened East Side Access, a 20-year project that was $12 billion in the making—in order to get a couple of suburban trains into Grand Central. It’s the most expensive subway station in the world. What went so wrong? There are too many planners, too many designers, too many consultants, too many construction engineers. No one knows what they’re doing. No one has experienced building a subway in a generation. So it costs astronomical sums as everyone gropes for an answer. I think that’s really important, because when we’re setting an industrial policy, a lot of the argument is that you have to be quick. You’ve got to allocate effectively, it’s got to go to the right people. If any of that goes wrong, it doesn’t really work. Tainter is saying: this is what it looks like when it fails. Things get really bureaucratic. They get extremely slow. Money’s not getting allocated well. Support Five Books Five Books interviews are expensive to produce. If you're enjoying this interview, please support us by donating a small amount . Look at the European Union . Some industrial policy obviously happens there, but a lot of folks would say the European Union has not done well in building up its industry. It’s not a lead in fabs and semiconductors, it doesn’t have a lead in a lot of other industries. East Asia has hollowed out the core of industry in Europe, particularly in Britain and France, less so in Germany. What went wrong? Well, if you have multiple levels of government, they’re not interacting with each other, there’s all this complexity, and you actually spend more time on the complexity itself than the actual building. That’s a huge problem. Tainter has a lot of examples. He’s looking at everything from Mesoamerica and the Mayans to Mesopotamia and a lot of the civilizations there, as well as China and some of the dynasties that have collapsed. He has more than a dozen civilizations he’s analyzing. He’s trying to develop a theory of why these civilizations or societies rise and then fall. And what he tries to show is that too much complexity leads to the decline of a lot of these civilizations. The bureaucracy becomes too intense. In China, you see this with a lot of the dynasties. They start taxing fairly low, and then it reaches this pinnacle where you’re taxing too much and no one can actually make any money, people underinvest, and then the regime or the dynasty falls. He’s interconnecting a lot of different examples. It was well received. It’s been popular more recently, for various reasons. Everyone thinks we’re at the end of times, particularly during and after COVID. I’m not sure it’s a perfect book, but I do think it has a lot of good insights. What I particularly loved was a section where he describes how you really have to do this using statistics. He’s more on the quantitative side. You can always feel, ‘Oh, it’s the worst time, we’re about to collapse,’ and the reality is that, for example, the Romans said that and then it took 400 years. Generations went by and everything was fine. So you just never know unless you’re really looking at the math, and thus he’s arguing for the benefit of statistical work. Quantification rears its head again! Tainter’s book is probably the best argument against industrial policy. It’s one that’s maybe most relevant to the United States in particular because entrenched interests become really hard to un-entrench when a policy or approach needs to move on. So if the last decade in technology was the rise of social media, and that was where the action was, now it’s AI. Yes, there are overlaps, and there are people who might be borrowed from one to the next, but every field starts from scratch. In the artificial intelligence world today, you need chips, you need fabs. You need to program these large language models. It’s a totally different industry than the one that we’re just coming from. That means you need an entirely different set of incentives, of approaches, of thinking about it. What often happens, particularly in the US where lobbying is very effective, is that you can’t beat the fact that the last industry was successful and has money. We keep wanting to print money with the old money machine that we used to make it with in the past. Industrial policy is about being done and saying, ‘Okay, that’s over, you’re on your own now. We need to move on to the next thing.’ If you look at South Korea in particular, the government is really good at moving on. There’s definitely a survival-of-the-fittest mentality. They’re like, ‘Look, you are in place, you are competitive. It’s up to you to keep going. We’re not going to keep subsidizing this. We’re not going to keep protecting you. You must be market driven. At some point, you must compete and win. We’ve gotten you to that point and our job here is done.’ Then they walk away and go on to the next industry. The challenge in the United States—and less so but also in Europe—is entrenchment. It’s lobbying of Congress. We see this particularly in defense. Defense is a huge part of industrial policy for the United States. We build a fighter jet in 50 states because we want to cover every congressional district. We want to make sure everyone has a little bit of their hand in the till, so to speak, so the program doesn’t get canceled. This leads to these suboptimal outcomes where the Navy will recommend—this was reported a couple of weeks ago— that the littoral combat ship be decommissioned because it doesn’t work and we don’t need it anymore. But we’re going to maintain it anyway for another 25 years because maintenance jobs are affected in all these different congressional districts. There’s no way to get rid of it. That’s the challenge for a lot of folks who do industrial policy today in the United States. On the one hand, there are clearly a lot of industries that would benefit from more industrial policy. On the other hand, we have never proven ourselves capable of letting go when we need to let go. So you have a little bit of pragmatism from some folks who say, ‘We should never do it to begin with because we’re always going to end up in some terrible situation at the end’ versus those who are a little bit more idealistic who say, ‘No, we can get it right. There’s a lot of benefit from this.’ I’m a pragmatist, but I wish the idealists would win every now and then—imagine what we could build!"