“Fragile, impermanent things”: Joseph Tainter on what makes civilizations fall
By Jessica McKenzie | March 12, 2025
In the introduction to his seminal 1988 book, The Collapse of Complex Societies, anthropologist and historian Joseph Tainter explained that lost civilizations have a vise-like hold on the human imagination because of the implications their histories hold for our own, modern civilization. Untangling how and why civilizations fall could, in theory, help humanity avoid a future calamitous collapse. “The reason why complex societies disintegrate is of vital importance to every member of one, and today that includes nearly the entire world population,” Tainter wrote. “Whether or not collapse was the most outstanding event of ancient history, few would care for it to become the most significant event of the present era.”
This issue of the Bulletin of the Atomic Scientists is dedicated to tipping points, primarily tipping points within the Earth’s climate system—when elements of the Earth system are pushed past a threshold and move from one stable state, to another, very different, stable state.
Civilizations are also complex systems, and ones that are not guaranteed to be indefinitely stable and secure. “Civilizations are fragile, impermanent things,” Tainter wrote, and collapse is actually quite common—“a recurrent feature of human societies.” What’s more, the kind of global civilization humanity currently enjoys is an aberration; for most of human history, people have lived in much simpler societies (which were also relatively stable systems). Consequently, Tainter wrote: “[W]e today are familiar mainly with political forms that are an oddity of history, we think of these as normal, and we view as alien the majority of the human experience. It is little surprise that collapse is viewed so fearfully.”
Do civilizations, then, have tipping points that determine their rise and fall?
Tainter’s theory of collapse is deceptively simple—especially when paraphrased. Collapse occurs, he argues, when the costs of complexity are greater than its returns to society. Complex societies are problem-solving organizations, and when the costs of coping with crises are too great, they fail.
One sentence in particular stood out: “Once a complex society enters the stage of declining marginal returns, collapse becomes a mathematical likelihood, requiring little more than sufficient passage of time to make probable an insurmountable calamity.” Although the phrase “tipping point” had not yet been popularized by Malcolm Gladwell or adopted by climate scientists at the time of the book’s publication, this certainly sounded like a kind of tipping point.
The book challenged some of my preconceived notions, and underscored things I don’t necessarily like or want to accept—that inequality is practically a requisite feature of complex societies, for example. Similarly, I chafed at some of the descriptions of class stratification, like the observation that: “Peasants are frequently disaffected, but they rarely revolt. They are usually passive spectators of political struggles.”
I also took note of the signs and portents of a civilization in decline. For example, when “tax rates rise with less and less return to the local level,” resulting in increasing dissatisfaction; or when “stress begins to be increasingly perceived, and…ideological strife” becomes noticeable; when the “system as a whole engages in ‘scanning’ behavior, seeking alternatives that might provide a preferable adaptation”; when “inflation becomes noticeable”; and “the hierarchy imposes rigid behavior controls…in an attempt to increase efficiency.”
Finally, I reached out to Tainter himself to discuss his theory, whether or not the tipping point metaphor is apt for civilizational collapse, and his more recent work on sustainability.
Editor’s note: This interview has been condensed and edited for brevity and clarity.
Jessica McKenzie: Could you briefly summarize your theory of collapse?
Joseph Tainter: Collapse is one of those major topics of history that are large and far-reaching, and had largely not been satisfactorily addressed. There’s substantial literature on the topic, but I was dissatisfied with all of it. There’s a chapter [in the book] where I go through all of that, describing what’s missing, what’s incomplete about existing theories.
I define a collapse in terms of the complexity of a society. Collapse is a rapid simplification. It’s a rapid loss of complexity. To address collapse, one also has to ask: “Why does complexity increase in human societies?” We’ve gone from small hunting, gathering, foraging bands of a handful of people, up to the complex societies of today.
Not every society on Earth did that by itself. Many did it because they were in contact with other complex societies, but basically, that’s been the course of human cultural evolution.
Now, complexity, we have to keep in mind, is not free. Complexity always has a cost. In the animal world, complexity has a metabolic cost. A deer is a more complex animal than a nematode. It also needs more calories per capita per unit of time. This is simply part of the nature of complexity. Complexity always has a metabolic cost, and this includes complexity in human societies. More complex societies require more energy as the basic unit of accounting—energy per person, per unit of time—than simpler societies do. So as societies have grown complex, they’ve also grown more costly per capita. If we go back in time, before the era of fossil fuels, when societies subsisted on what individuals could produce, either by hunting and gathering or by agriculture, increasing the complexity of a society meant that people worked harder. And so we would ask, “Why? Why would societies become more complex if it means that people have to work harder?”
The answer I’ve suggested is that most of the time complexity increases because it’s useful to solve problems. Think about our society today—how are we addressing major problems like climate change? We have things like national legislation, we have state legislation, we have technological changes. These are all increases in complexity, but they also impose a cost on individuals. We think of the cost in terms of money, but the ultimate cost is energy, and this tells us why this is largely unrecognized today. The cost of complexity is largely unrecognized, because to us, complexity appears to be free. We pay for it with fossil fuels. That’s all it is. We have this subsidy of fossil fuels, meaning solar energy, eons ago, that mostly fuels our societies today.
In the ancient societies that I was studying as an archaeologist, the cost of complexity was more immediate. In a case such as the Roman Empire, which is a case I’ve worked on a lot, it meant that people—peasant farmers who were 90 percent of the population—had to pay higher taxes, and no one likes doing that. And what we find is that the Roman Empire, through time, encountered more and more challenges, particularly in the third century AD. This called for expanding the size of the government, expanding the size of the army, and making the army more complex. These are all elements of complexity, and this all had a metabolic cost that had to be paid with peasants’ agricultural production. And so we see over time in the Roman Empire that the cost of being the Roman Empire goes up and up in order simply to maintain the status quo. And I’ve expanded this to other ancient societies that were well documented.
What I’ve argued is that collapse results from diminishing returns to complexity. The cost of being a complex society goes up and up until finally a point is reached where the system simply can’t be maintained any longer. In the Roman Empire, the empire was being invaded by Germanic peoples from Central Europe, and eventually reached the point where they couldn’t defend themselves, and so the empire disintegrated. It collapsed, which is to say that it simplified.
destruction of a European city in the early 1600sDuring the Thirty Years’ War, much of Europe (particularly what was to become Germany) was devastated, in what some historians have labeled the General Crisis of the Seventeenth Century—a time of crop failures, economic hardship, extreme violence, and high mortality, all of which some trace to features such as the Little Ice Age. Some regions recorded population drops of as much as 70 percent. This painting by Matthäus Merian the Elder shows the siege of the city of Bautzen, Germany, in 1620. Public domain image.
McKenzie: The theme of this magazine issue is tipping points. Tipping points in the climate system have recently gotten a lot of attention, but they’ve also been discussed in the social sciences and other areas. Is the tipping point metaphor useful when thinking about societal collapse?
Tainter: It’s useful depending on how you define a point, because in the ancient societies that I study, the point might be a period of several decades. It’s not necessarily a single moment in time or a single incident, and maybe that’s the case today also.
McKenzie: While reading your book, I noticed that the people who were studying collapse before you, were very into tipping points. They were very into theories like, this plague, this environmental catastrophe, this single event caused X civilization to collapse.
Tainter: I hadn’t thought of the earlier literature in that way. But you’re right. That is their approach.
McKenzie: But then looking at your theory, you have this graph showing the marginal return on complexity, where there’s these key points at B1, C1, B2, C2, B1, C3, where marginal returns on complexity either slow, taper off, or even start to fall. And at each point, the risk of collapse becomes greater. Those all looked like potential tipping points to me, too. (See figure 1.)
chart on complexity Figure 1. The marginal product of increasing complexity. According to Tainter, at point B1, C1 on the curve, additional investments in complexity yields increased returns, but at a declining marginal rate. This makes a complex society more vulnerable to collapse. After B1, C2, the costs of complexity increase, but the benefits actually begin to decline. A society at B1, C3 is at very high risk of collapse. Graph courtesy of Joseph Tainter.
Tainter: You could say they were tipping points in that they were pinpointing phases of transition, but you wouldn’t point to a single year as the point. The third century AD was a very long period. It could be that that entire period was a tipping point, although the empire did recover from the crisis, it was a 50-year period of crisis, but it did so by becoming more complex and more costly.
McKenzie: There are plenty of people who believe we’re already on the path to collapse, and a lot of them point to climate change as a reason for that, or to environmental deterioration as a result of climate change. But based on your theory, that alone wouldn’t explain if and when current society collapses. Could you talk about how we can think about the risk of collapse now in the context of a changing climate?
Tainter: I would answer by asking: “Can we cope with climate change?” “Do we have adaptive mechanisms to deal with climate change?” I would be pretty sure that we do, but it means the society becoming more complex and more costly and so the question becomes: “Can we bear the cost?” —which, ultimately comes in terms of energy. Are we simply going to have to burn more fossil fuels? Can we make an energy transition in time?
McKenzie: Because we have a globalized society unlike any of the previous civilizations that have collapsed, are we in a unique civilizational moment?
Tainter: In ancient societies, it was isolated societies that collapsed. And there are no isolated societies today. Look at recent events: Syria is a great example. If it looks like it’s disintegrating, the outside powers have a tendency to meddle to various degrees—even to step in to occupy the place and try to establish a functioning government, a functioning economy.
McKenzie: The point you make in your book is that if civilization collapsed now, it would be a total collapse. There’s no one country or one region collapsing without the whole thing falling.
Tainter: I still tend to think that way. Yes, I cannot see a scenario where, let’s say, Western Europe all collapses and we don’t do anything about it. I just don’t see that.
McKenzie: Is it possible to get to a point where civilization is too big to fail? In your book, you make the point that collapse is not an enviable state to any individual unless they’re capable of feeding themselves and taking care of themselves. I’m not prepared to do that. And I think most people are not.
Tainter: We passed that point a long time ago. A collapse today would mean that billions of people would die in a short period of time.
McKenzie: That doesn’t sound great at all.
Tainter: It’d be awful. Yes.
McKenzie: Back to the scale of declining marginal returns. Where do you think we currently are as society on that curve? There are some symptoms that you identified in various civilizations that were on their way to collapse, things like inflation, or the feeling of being taxed too much, where people think that they’re getting less than they’re putting in, or they’re getting less than people who came before when they put in the same amount. And I think to some extent, there is the feeling that that is already happening in the United States, and maybe globally as well.
Tainter: I can remember when I was in my 20s and 30s, my age mates and I would complain that we were worse off than our parents were at that age. Today, I hear the 20- and 30- year old young adults saying the same thing about their relationship in respect to us old-timers. But there’s always some level of discontent. There are people who aren’t satisfied in their careers, people who have trouble with their children, people who don’t like individual politicians and so forth and so on. There’s always reasons for discontent. I don’t see any of that as leading to a collapse.
Now, there are people who are talking about civil war in the United States and so forth. I don’t see that happening, but there are people who think that way.
McKenzie: It does seem to me that we may be somewhere along that risk curve, if only because so many people are already saying we’re on the way to collapse. One parallel I was thinking of is the “vibecession.” We’re not in a recession, but people feel like we’re in a recession, and they’re responding accordingly. It’s all going on vibes. And I feel like there’s some parallel with collapse, in the sense that this impending feeling of collapse may itself indicate something not so good. Does that make sense?
Tainter: Oh, yes. And I think that is how people—excuse me for being blunt—in their ignorance, don’t comprehend what’s going on in the broader world. We evolved biologically in a context of small hunting and gathering bands. There was never natural selection for the ability to think broadly in time and space, because people never had an opportunity to and never had a need to. As a species, we do not think broadly in terms of time or space—or today, current events.
For example, people don’t pay attention to what’s in the news. I remember, going back to the days of the Watergate trial, I remember that in Washington, D.C., they would ask individual jurors for the Watergate defendants, have you heard of Watergate? And people would answer, “No.” My God, we were deluged with news about it for a year-and-a-half or more. It’s just astounding how people don’t know what’s going on in the world and have no inclination to find out. That is one of the species’ weaknesses.
McKenzie: I’m also interested in what you call “elite mismanagement,” and the role of greed in explaining collapse.
stone heads of Easter IslandStone heads known as “moai” on Rapa Nui (Easter Island). Scientists such as Jared Diamond think that human society on the island destroyed its natural environment, leading to a downward vicious cycle of warfare and population decline. Image courtesy of Yves Picq under Creative Commons Attribution-Share Alike 3.0 Unported license
Tainter: When I talk to audiences about, say, the collapse of the Roman Empire, there’s this popular mythology of Roman emperors as corrupt, deceitful, and inept. But, in fact, when I look at the historical record, most of the things most Roman emperors did were rational. It’s just that there was nothing they could do. The problems were beyond their capacity to solve—although they did, in fact, take very often rational steps to do so. Elite mismanagement is not really an explanation. It may be a short-term response, but it’s not the underlying long-term problem. If you want to understand a collapse, you have to go back several generations—at least several decades—and look at trends over time.
McKenzie: Wasn’t there an example from one of your more recent papers of a complex society that was able to systematically simplify and avoid collapse?
Tainter: The Byzantines. When the Arabs erupted out of Saudi Arabia and conquered the most productive lands of the Byzantine Empire, the empire’s revenue was cut in half, and it could very well have been completely overrun by the Arab forces. But it survived—very largely by simplifying. I mean, it lost half its revenue. It disbanded its professional paid army and reverted to peasant militias for at least a period of a couple hundred years. Literature virtually disappeared, except for lives of states. The whole of what remained of the empire, which was largely the Balkans and Anatolia—what’s now Turkey—all of the cities disappeared and reverted to fortified hilltops. There were only two cities left in the empire: Constantinople, and Thessalonica in Greece. I mean, it survived by simplifying, but that was a collapse. When I give talks, I’m asked: “Can a society simplify and survive?” There is this one example, but it’s the only one, and they didn’t do it voluntarily. They did it because their backs were to the wall.
McKenzie: So you still think a systematic simplification of civilization won’t work, that it’s not really possible?
Tainter: No, because problems arise and they’re dealt with by increasing complexity. And I have argued that we can’t reduce consumption of resources voluntarily over the long term, because even if we do so, something’s going to come up requiring us to use those resources again to increase complexity to solve problems.
McKenzie: What about regime change or large-scale shifts in how civilization is run—so not de-complexifying, but changing for the better?
Tainter: It takes awareness. It takes consensus. People need to understand the resource basis of how we live today. As I said, to us, complexity appears to be free because we pay for it with fossil fuels. There is an effort underway to pay for complexity with other forms of energy. I hope it’s successful. It seems that it’s necessary. In our own part of the world, it’s not going rapidly. But I am not entirely without hope, not entirely without optimism—although I am a realist.
McKenzie: Do you have thoughts on the popular perception of collapse today? There is, for example, an entire subreddit dedicated to collapse.
Tainter: I don’t follow anything like that. But COVID made me aware that people perceive a threat to their way of life, even if they can’t articulate it. I do a few interviews a year. When COVID hit, I suddenly noticed I was getting more requests for interviews. Journalists were picking up on the fact that people were worried, that they saw a threat to their way of life. People perceive threats to their way of life. They may not think in terms of collapse, but they perceive threats to their way of life.
McKenzie: How is that different from the existential fear at the height of the Cold War, at the height of fear of nuclear weapons?
Tainter: I’m a child of the Cold War. I remember hiding under our desks in grammar school, as if that would do any good. I grew up in San Francisco. I just assumed that it would happen. As a seven, eight, nine-year-old, I thought, it’s going to happen, and I used to have daydreams about living out in the countryside somewhere camping out with my family. Somehow, we all survived, but the city was ruined, and I have these memories as a child.
I think it’s less explicit now. During the days of the Cold War, it was explicit. People knew it could all go away. And this persisted in my reckoning, into the Reagan administration, when nuclear war was seriously discussed as a possibility. And then, since then, those fears have largely died down. There are occasional fears of recession. Back in ‘08-’09, there was a fear of major economic collapse, the worst depression since the ‘30s. People would talk about it—and people are vaguely aware of those things without understanding underlying causes. So they think, you know, the President’s responsible for a recession, for example. It goes back to the days of kings and emperors, where a bad harvest was blamed on the king; today, a bad economy is blamed on the president. It’s just how people think about their world—again, excuse me for being blunt—but in ignorance.
McKenzie: If collapse isn’t guaranteed, what are the prescriptive things that you would have people in power—or people who may soon be in power—do?
Tainter: The way I like to answer that is that we are a species that muddles through. It’s all we’ve ever done, all we ever will do. The Cold War is one instance where there was continuity of national policy, the need to confront the Soviet Union and the global spread of communism. But other than that, we muddle through.
One thing I’ve observed in my lifetime is that things that seem like an urgent crisis and have Congress and politicians tied up in knots, a little while later, they just sort of fade into the background, and the next crisis comes along. I tell my students, “Look, this too shall pass.” It’s just how I see things, that the things we live through for the most part are minor crises. A recession is a minor crisis. COVID is a minor crisis, compared to some crises that societies in the past went through and that we might want to go through again. So muddle through. That’s the only thing I can say.
McKenzie: That’s a pretty healthy perspective, not to get too bogged down in pessimism.
Tainter: I have, at times. I did in the ‘08-‘09 crisis, but not because of the economy. I got depressed because oil was up to $140 a barrel, which was a lot in those days, and I was thinking, oh boy, the end may be coming. What I didn’t foresee at the time was fracking, which largely alleviated that crisis. There are things about it we don’t like, but fracking, in my mind, averted a major crisis for our way of life.
McKenzie: How does renewable energy fit in?
Tainter: Well, it’s hopeful. I have colleagues who think that if we can transition to a renewable energy economy, then all our problems are solved; there will be no future crises. I point out that crises are inevitable. Problems are inevitable. Many problems and crises call for solutions which can be costly, can increase complexity, and have a metabolic cost.
One thing that worries me about renewables is the question of can we rapidly increase production of energy in a renewable energy society? There are things about a liquid fuel, advantages that are simply insurmountable. And in the future, if we don’t have the ability to rapidly produce a liquid fuel, because we’re relying on renewables and probably then electricity, what does that mean for coping with crises? I just don’t know the answer. Technology is not my field.
I’m a social scientist, and maybe there would be ways of doing it. Maybe we devote a lot of renewable energy to producing liquid energy in one form or another. But then the problem of pollution comes up. So I don’t know the answer to that. But as a historian, I think in terms of crisis, and I wonder, what happens in a crisis if we’re dependent on renewable energy?
McKenzie: Since your work on collapse, you’ve switched your focus to sustainability. I was wondering if you could tell me what sustainability means to you?
Tainter: Sustainability to me, means continuity. It means maintaining a complex society and not having it collapse. Other people think about it in different terms. Some people think in terms of sustainable development, which just strikes me as primarily development. It’s continuity, continuing our way of life.
McKenzie: Some of your more recent research has looked at whether we can innovate our way to indefinite sustainability.
Tainter: Yes. There are colleagues of mine—some in economics, some in other fields—who argue that we never need to worry about resources, as long as there is the price mechanism and free markets and the government doesn’t get in the way, that there will always be incentives to innovate and innovate our way out of problems. We’ll either discover new resources or new ways of using the old ones, or we’ll develop new technologies. Whatever it is, that as long as we have incentives to innovate, we always will, and therefore resources can be just dropped out of the equation.
McKenzie: But you don’t necessarily agree.
Tainter: No. It seems to me that there is an implicit assumption in that line of reasoning—that my colleagues who reasoned this way aren’t aware that they’re making—that the productivity of innovation remains constant, or maybe even grows.
Innovation is like any human activity; it has benefits and costs. The system of innovation that we have today, whether it’s commercial innovation or scientific innovation, it’s like any human activity. It undergoes increasing complexity over time.
If you think back to the 19th century, even the 18th century, the age of what are sometimes called “lone-wolf naturalists,” people like Charles Darwin, Gregor Mendel. Science in those days was the province of individual scholars pursuing it, often as a hobby or because it just intrigued them. Compare that to today, where science is a complex, interdisciplinary enterprise.
Joseph Tainter. Image courtesy of Utah State University.
In the 1970s when I was working on my PhD, you could pick up an issue of Science or Nature, the two primary journals, and flip through them, and you would usually find a single author per paper, sometimes a couple of authors per paper. You open most journals today and there are multiple authors per paper. It’s a symptom of the fact that research—whether it’s in the commercial arena or for pure science, curiosity-driven research—has become a highly complex, interdisciplinary enterprise. It takes fairly large, costly institutions like universities, think tanks, commercial firms with R&D departments. Science is growing more complex, but more costly at the same time.
In late 2005, I spent a couple of years at Arizona State University. I met an economist there named José Lobo who understood what I was arguing. He had a colleague, a geographer named Deborah Strumsky, who was putting together a database of patents that I saw would allow us to test the proposition that science is growing more complex and producing at higher costs, and perhaps, producing diminishing returns. We took this database of about 3 million patents, beginning in 1974 through 2012, and we estimated the cost of developing an individual patent, and how many individuals did it take to produce a patent? The converse of that is patents per author. And patents per author has been declining that entire time, it’s down by something like 22 percent in the data set that we developed, and shows there’s no reason to think that pattern is going to change. It’s hard for people to understand this, because you can get online or go in an electronic store, and there’s always new electronic widgets that you can buy. When I bring this up, people often refer to Moore’s Law, the idea that the number of transistors on a chip doubles with every 18 months at half the price. And I say, “Yeah, great, but it now takes five times as many engineers to keep Moore’s Law going as it did in the 1990s.”
Another example I like to use: cell phones. How many elements from the periodic table did it take to develop cell phones in the 1990s versus today? I don’t have the exact figures off hand, but in the 1990s to develop a cell phone it took a few elements from the periodic table. Today it’s big chunks of the periodic table, and it almost looks like we’re running out of elements. It’s all an example of how science, innovation, commercial development all goes according to the normal human behavior of first plucking the low-lying fruit. In science, the low-lying fruit was basic theories like the theory of evolution, basic knowledge of electrics, so forth and so on. And those discoveries are no longer out there waiting for us to stumble across them. We’ve plucked the low-lying fruit in research, and so science has had to move on to ever more complex fields of inquiry.
McKenzie: The example that stood out to me, of marginal returns on research and health, specifically, was your point that we now we have to spend the same amount or more to fight evolving pathogens that we previously conquered with vaccines, but we still have to keep iterating to ensure that we have the next updated vaccine, just to keep a baseline level of health.
And in terms of climate, two of the big ideas that technological optimists have put forward for fixing the climate crisis are solar geoengineering—spraying stuff into the atmosphere to block the sun’s rays and lessen global heating—and direct air capture. And both of those things are expensive or difficult or risky or complex. Ultimately they just get us marginally closer to where we want to be in terms of global temperature or atmospheric carbon dioxide levels. It just staves off a little bit of damage.
Tainter: Yes. And actually, if I can go back to tipping points for a moment—some thing that got me worried about what could be a tipping point was COVID, which brought to my attention the risks of globalization. There’s this famous photograph that I showed in one of my classes of a vast parking lot full of General Motors trucks all complete, that can’t be sent to market because they’re missing one computer chip. Markets were breaking down because of the failures of globalization during the COVID crisis, not sufficiently to cause what I would consider a societal collapse. But it did cause some rather substantial economic retrenchments, and it made me think that one of the risks today is globalization—that a total breakdown of globalization could cause a collapse.
McKenzie: One quote stuck with me: “Once a complex society enters the stage of declining marginal returns, collapse becomes a mathematical likelihood, requiring little more than sufficient passage of time to make probable an insurmountable calamity.” If further complexity is inevitable, and will eventually result in declining marginal returns, does that mean that the collapse of global civilization is inevitable?
Tainter: I don’t want to say it’s inevitable, but we have an enormous job to do, educating the population about the risks. If I was half-a-century younger, I might want to spend my time talking to K-12 educators so that they could try to teach young children to think more broadly in time and space, in a way that our educational system mostly doesn’t do today.
Now, I’m not a K-12 educator. I don’t know how you educate children at that age. Is it possible to teach them?
But if we could teach people, it would have to start at an early age, to be curious about what’s going on in the world, in areas over the horizon, things they can’t see, things that don’t affect them immediately, but could in the future. I hope it would make a difference, and perhaps an ultimate crisis could be averted that way. It might require people to accept different way of living, to accept less consumption than we are used to today. So I don’t want to say it’s inevitable.
McKenzie: I have to admit I had to really grapple with the systems-level thinking in your book. In some ways it made me feel pretty powerless, to think about civilization itself as an independent working machine, and the limited power of individuals to influence civilization’s progress in either direction. Does that make sense?
Tainter: Yes, well, one of the definitions of culture in anthropology is that it’s extra somatic. It’s something external to us that operates by its own rules as a system unto itself, including complex cultures such as ours. I mean, that is correct to a degree, but it doesn’t mean that it couldn’t be out of human control. It comes down, as I say, to knowledge and curiosity and awareness.