How Empires Learned to Promote the Wrong People (And Never Stopped)

Every organization that has ever existed has had to solve the same problem: who gets to be in charge next, and how do you make sure it's someone competent rather than someone well-connected? Thousands of years of evidence suggest we've never really cracked this one.

The failure mode is so consistent across time and culture that it starts to look less like a policy problem and more like a psychology problem. Which, it turns out, it is.

The Exam That Was Supposed to Fix Everything

Around 165 BCE, the Han Dynasty did something genuinely radical. Instead of staffing the imperial bureaucracy with the relatives and cronies of whoever happened to be powerful, they started testing candidates. Written exams. Standardized questions. In theory, the smartest person who could master the Confucian classics got the job, regardless of family background.

This was a real innovation. For stretches of Han history, it worked reasonably well. Talented people from modest backgrounds rose to positions of genuine influence. The bureaucracy functioned. The empire was administered by people who had at least demonstrated the capacity to learn complicated things and write coherently about them.

Then the system started doing what systems do.

Preparing for the imperial exams required years of intensive study — tutors, texts, time. Those things cost money. Families with money could afford better preparation than families without. The exam that was designed to route around inherited privilege gradually became a mechanism that laundered inherited privilege into something that looked like merit. By the later dynasties, the examination system had become so entrenched and so gameable that it was producing officials who were excellent at passing exams and increasingly bad at actually governing.

The Han wanted meritocracy. They built a system. The system got captured. The empire's administrative quality declined. This cycle repeated across Chinese dynastic history so reliably that historians have a name for the pattern.

What "Meritocracy" Actually Selects For

Here's the uncomfortable mechanism. When an organization announces that it promotes on merit, it still has to define merit somehow. And whoever defines merit tends to define it in terms of things they can measure — which tends to be things that look like what the current leadership values, which tends to be things that the current leadership is good at.

This isn't conspiracy. It's psychology. People recognize competence most easily in people who remind them of themselves. They trust people who communicate in familiar ways, who have navigated similar paths, who signal belonging through shared references and shared networks. This was true in a Han Dynasty examination hall, and it is true in a modern performance review.

The result is that "meritocracy" in practice often selects for cultural fit with the existing leadership cohort, which in most organizations means it selects for people who share the background, communication style, and social network of whoever is already at the top. The exam score becomes a credential. The credential becomes a signal. The signal gets gamed. The system produces the same people it was designed to replace.

The Tang Tried Harder. It Didn't Help.

The Tang Dynasty (618–907 CE) looked at what had happened to Han meritocracy and tried to tighten the system. They expanded the examination curriculum, added more subjects, and introduced measures meant to prevent the children of officials from having an unfair advantage. For a while, Tang China had one of the most sophisticated civil service systems in the world.

And yet. By the mid-Tang period, the bureaucracy had stratified into competing factions organized around — wait for it — family networks and patron-client relationships. Officials who had passed the same exams in the same year formed cohorts that functioned like alumni networks, advancing each other's careers in ways that had nothing to do with administrative competence. Powerful families learned to ensure their sons passed the exams by controlling access to the best preparation.

The system wasn't corrupt in the sense of being obviously broken. It was operating exactly as designed. The problem was that the design couldn't survive contact with human social instincts, which are very old and very strong and very oriented toward kin and reciprocity rather than abstract institutional fairness.

Your Company Is Running This Experiment Right Now

American corporations spend a significant amount of money on talent management systems, leadership development programs, and performance review frameworks — all nominally designed to identify and advance the best people. Most of them are running a version of the Tang exam system, with the same structural vulnerabilities.

The research on this is pretty consistent. Studies on promotion decisions repeatedly find that factors like physical resemblance to current leadership, shared alma mater, shared social networks, and communication style similarity predict advancement at rates that shouldn't be possible in a genuine meritocracy. The performance review, like the imperial exam, measures something real — but the something real it measures is often "how well does this person fit the template of people who are already here."

This isn't an argument that performance reviews are useless or that credentials don't matter. It's an argument that the gap between what these systems claim to do and what they actually do is not a modern management failure. It's a documented feature of how humans evaluate other humans, and it has been producing predictable organizational outcomes for millennia.

What Happens When the Wrong People Accumulate at the Top

The dynasties that promoted badly tended to fail in characteristic ways. The bureaucracy filled up with people whose primary skill was navigating the bureaucracy. Innovation slowed because the people rewarded for advancement were the people who didn't challenge the existing framework. When external shocks arrived — drought, invasion, economic disruption — the administrative apparatus was staffed with people who were very good at performing competence and not particularly good at exercising it.

Modern organizations tend to fail more slowly and less dramatically than empires, but the pattern is recognizable. Companies that have spent a decade or two optimizing for internal political navigation end up with leadership cohorts that are skilled at internal political navigation. When the market shifts, or a competitor does something unexpected, or a crisis requires genuine problem-solving rather than credentialed performance, those cohorts are often poorly equipped for what's actually required.

The people who could have told them this were usually passed over for promotion.

The Part Nobody Wants to Hear

There's no clean solution here. The Han tried standardized testing. The Tang tried expanding the system. Modern HR departments have tried 360-degree reviews, blind resume screening, structured interviews, and algorithmic assessment tools. All of these work somewhat, and none of them work enough to fully override the social instincts that have been shaping human group hierarchies since long before any of us showed up.

What history suggests is not that meritocracy is impossible, but that it requires active, continuous, uncomfortable effort to maintain — and that the moment an organization decides it has solved the problem, the problem has almost certainly already started reasserting itself.

The Han knew this. They wrote about it. The Tang knew this. They wrote about it too. The warnings are in the historical record, which is one of the reasons The Past Market exists.

Five thousand years of data. Use it.