China Built the World's First Meritocracy. It Worked So Well It Broke Everything.
Somewhere around 605 AD, a Tang dynasty emperor made a decision that would echo for thirteen centuries: he decided that the people running his government should get their jobs by being smart, not by being born into the right family. The tool he used was a standardized test. Sound familiar?
The keju — China's imperial examination system — is one of the most under-discussed experiments in human organization. It ran, more or less continuously, from the Sui dynasty through the Qing, finally abolished in 1905. That's a longer run than the Roman Empire, longer than the United States has existed, longer than most institutions humans have ever built. And for most of that stretch, it genuinely did what it promised: gave talented men from ordinary backgrounds a path to real power.
The part nobody tells you is what happened when it worked too well.
The Idea Was Radical. The Execution Was Remarkable.
Before the keju, Chinese government ran on aristocratic patronage. You got a position because your family had positions. The emperor's court was a closed loop of inherited influence — which meant the quality of governance depended heavily on the accident of birth.
The examination system blew that open. Any male subject could, in theory, sit for the exams. The tests covered classical literature, philosophy, poetry, and statecraft. They were grueling, multi-day affairs conducted in individual cells, and the content was standardized enough that a farmer's son from a rural province was answering the same questions as a nobleman's heir from the capital.
The results, over time, were genuinely impressive. Historical records show significant social mobility in dynasties that enforced the system rigorously. Families rose from obscurity to the highest levels of government within a generation or two. The bureaucracy, by the standards of the ancient world, was remarkably competent. When European observers encountered it in the 17th and 18th centuries, they were stunned — and immediately began advocating for something similar back home. (The British civil service exam, introduced in the 1850s, was explicitly modeled on the keju.)
For centuries, the system was genuinely meritocratic in ways that most of human history simply wasn't.
Then the Arms Race Started
Here's where it gets uncomfortable.
As the examination system matured and the rewards for passing became clearer — government salary, social status, family honor, exemption from certain taxes — the investment required to compete escalated with it. Preparation stopped being something a bright kid could do with a good teacher. It became a years-long, family-consuming project.
Wealthy families hired private tutors. They built study pavilions. They kept sons in preparation for a decade or more. The content of the exams, meanwhile, narrowed over time into an increasingly rigid format — the famous eight-legged essay (baguwen) — which rewarded not original thinking but perfect mastery of a prescribed structure. Examiners knew what they were looking for. Candidates learned to produce exactly that.
The result: the exam still sorted people, but increasingly it sorted for the ability to prepare for the exam. A candidate from a wealthy family who'd spent fifteen years drilling classical forms had a structural advantage over a brilliant kid from a poor village who'd spent five. The credential still meant something. But what it measured had quietly shifted.
Passage rates stayed low — often under 5% at higher levels — so the credential remained prestigious. But the number of people spending their lives preparing for it kept growing. By the later Qing dynasty, there were hundreds of thousands of men who had passed lower-level exams and were technically credentialed but had no government position waiting for them. They'd done everything right. The system just didn't have room for them.
The Credential Without the Job
This is the part that should make American readers sit up straight.
By the 19th century, China had produced a massive class of highly educated men — credentialed, trained in a specific set of skills the system had told them were valuable — for whom the promised reward simply wasn't available. The pipeline had been optimized for producing candidates. Nobody had optimized it to produce enough positions.
The social consequences were severe. Educated men with no economic outlet became a destabilizing force. Some historians link the surplus of frustrated credentialed elites to the Taiping Rebellion (1850–1864), one of the deadliest civil conflicts in human history. Hong Xiuquan, the rebellion's leader, had failed the imperial examinations multiple times.
The system wasn't abolished because it stopped working. It was abolished in 1905 because the Qing government, modernizing under pressure, recognized that it was producing the wrong credentials for a changing world — and that the credential class it had built was too large, too frustrated, and too poorly matched to available opportunity to sustain.
They'd built a perfect machine for a world that no longer existed.
The Uncomfortable Parallel
America's college admissions system isn't thirteen centuries old. But the psychological pattern running underneath it is identical, and the keju gives us the clearest possible view of where credential arms races go.
The original promise — that educational attainment would be the meritocratic sorter, rewarding talent over birth — was real. For much of the 20th century, a college degree functioned the way the keju did in its early centuries: a genuine signal of capability that opened genuine doors.
What happened next is keju dynamics in fast-forward. As the reward became clear, investment escalated. Preparation started earlier. The format narrowed — test prep, extracurricular optimization, application coaching — in ways that increasingly rewarded families with resources to invest. The credential retained its prestige partly because admission rates at selective schools stayed low, but the population spending enormous resources preparing for it kept expanding.
And now, like late Qing dynasty China, America has a substantial population of people who did everything the system asked — accumulated the credential — and are finding the promised return isn't reliably there. The graduate degree arms race is the eight-legged essay. The credential inflation at every level of the job market is the surplus of lower-level exam passers with no government post.
What History Actually Says About This
The keju ran for 1,300 years, so it would be wrong to call it a failure. It was one of the most durable human institutions ever built, and it produced genuinely competent governance across many dynasties.
But the record also shows a consistent pattern: meritocratic systems tend to be most functional in their early centuries, before the optimization pressure fully kicks in. Once families know what the credential requires, they begin producing the credential rather than the underlying competence the credential was meant to signal. The system doesn't collapse — it drifts. The sorting continues; it just sorts for different things than advertised.
Human psychology hasn't changed. The incentive to optimize for measurable signals rather than underlying quality is as old as measurement itself. Every generation of exam-takers, from Tang dynasty scholars to today's SAT-prep industry, is running the same calculation.
The keju is a five-thousand-word warning written in a language we keep refusing to read.
