Why Civilizations Break Down (And What It Means for Us)

Have you ever noticed how fixing one problem sometimes creates two more? It turns out that pattern doesn't just apply to your to-do list — it may be a fundamental law of how civilizations work.

Anthropologist Joseph Tainter made a simple but powerful observation: societies solve problems by adding complexity. A village needs food? Appoint someone to manage irrigation. Need protection? Build an army. Need fairness? Write laws.

Early on, each new layer of organization pays for itself many times over. But eventually, you hit a wall. Each new fix costs more than the problem it solves, and society spends all its energy just keeping the lights on — with nothing left in reserve for the next crisis.

This Has Happened Before

The Roman Empire started with a clean, simple legal code. Centuries later, it had ballooned into a massive, contradictory bureaucratic mess. To pay for it all, Rome debased its currency from pure silver to base metal with a silver wash. The Ottoman Empire followed a similar playbook in the 1800s, adopting Western-style bureaucracy that ultimately failed to halt its decline.

The pattern is always the same: more rules, more administrators, more cost, less resilience.

It's Happening Now

The numbers in modern America tell a striking story. The US Federal Register of regulations went from about 2,500 pages in 1936 to over 90,000 today. The original income tax law was 400 words; the current tax code runs past 4 million. And between 1975 and 2010, the number of doctors grew by 150% while healthcare administrators grew by 3,000%.

The numbers in Canada tell a similar story. Federal regulations contained roughly 400,000 restrictive words in 1970; today, they exceed a million. The Income Tax Act was six pages in 1917; it now runs to 1.1 million words across more than 1,400 pages. And in British Columbia alone, health system administrative spending has nearly tripled since 2017 — outpacing both frontline care spending and population growth. And Ontario physicians now spend up to 40% of their time on paperwork rather than patient care.

We're spending more and more energy managing the system itself, and less actually doing the thing the system was built for.

Corporations Do It Too

General Electric is a textbook case. Its famous Six Sigma quality system initially drove huge gains, but over time, it hardened into a rigid bureaucracy that crushed innovation. GE's sprawling financial arm nearly killed the company in 2008. Eventually, the whole conglomerate had to be broken up. When something gets too complex to understand, simplification isn't optional — it's survival.

Roughly two-thirds of software projects end in partial or total failure. For large-scale projects, the success rate drops below 10%. The 2013 launch of Healthcare.gov — built by 55 different contractors with no central coordination — is the poster child. Six people managed to enroll on day one.

Can We See It Coming?

Maybe. Physicist Yaneer Bar-Yam built a model showing that when global food prices cross a specific threshold, social unrest spikes dramatically. It predicted the Arab Spring. Historian Peter Turchin used a massive database of past civilizations to predict a peak in US social instability around 2020, years before it arrived.

A 2023 study in PNAS found that the median lifespan of a state is roughly 200 years, and that the older a state gets, the more brittle it becomes. A young society can absorb a shock; an old, complex one shatters.

Tainter's core argument is simple. Societies are problem-solving machines, and their primary tool is complexity — more roles, more rules, more hierarchy, more specialization. Early investments in complexity yield huge returns: a little bureaucracy produces a lot of order. But each successive layer of complexity delivers less benefit at greater cost, following a curve of diminishing marginal returns.

Once a society has eaten through all the easy gains, it finds itself on the flat or declining part of that curve — spending enormous resources just to maintain the status quo. At that point, it has no surplus left to absorb shocks (drought, invasion, economic disruption), and what looks like a mighty civilization is actually brittle. Collapse, in Tainter's view, isn't a mysterious catastrophe — it's a rational, even economical, response. Populations shed the costly complexity because a simpler organization actually delivers better returns per unit of investment. Roman peasants in some provinces welcomed barbarian rule because it meant lower taxes.

He backs this up with three detailed case studies — the Western Roman Empire, the Classic Maya, and the Chacoan society of the American Southwest — and argues that each followed the same essential trajectory: rising complexity, diminishing returns, loss of legitimacy, and simplification.

Collapse isn't a failure of will or morality but a structural inevitability built into the economics of complexity. That's what makes it feel so uncomfortably relevant to modern debates about regulatory bloat, institutional sclerosis, and the sustainability of ever-growing administrative states.

The Bottom Line

Complexity is a tool with a built-in expiration date. Every civilization in history has eventually hit the point where its own machinery became the problem. Some researchers argue the collapse, when it comes, looks less like a slow decline and more like a sudden phase change — water turning to ice.

The question for the 21st century is simple: can we learn to voluntarily simplify our systems before they simplify themselves? No civilization has pulled it off yet. Whether we'll be the first is still an open question.

• Joseph Tainter — The Collapse of Complex Societies (Cambridge University Press, 1988)

• Mancur Olson — The Rise and Decline of Nations (Yale University Press, 1982)

• Jonathan Rauch — Demosclerosis: The Silent Killer of American Government (Times Books, 1994)

• McKinsey & Company — Healthcare.gov pre-launch assessment (2013)

• Yaneer Bar-Yam et al. — Food price / social unrest model (New England Complex Systems Institute, ~2011)

• Peter Turchin — Work on cliodynamics, elite overproduction, and the Seshat Global History Databank

• Rolf Landauer — "Irreversibility and Heat Generation in the Computing Process," IBM Journal of Research and Development (1961)

• Ilya Prigogine — Dissipative structures framework ( Nobel 1977)