When Power Eats the Judgment Needed to Leave

In December 1989, Nicolae Ceausescu stepped onto a Bucharest balcony and heard something he had never, in thirty years of absolute rule, encountered: booing. He stopped mid-sentence. The expression is worth finding — it is not anger, and it is not fear. It is the face of a man who has genuinely never considered that such a sound might be directed at him. He was executed four days later, which in the circumstances was probably merciful.

The expression is the curiosity. Not the execution.

---

The standard explanation for why powerful men refuse to leave runs roughly as follows: they fear prosecution, they cannot bear irrelevance, they love power for its own sake. All of this is true. It is also beside the point, because it locates the problem in individual psychology rather than in the machinery that produces it.

The actual mechanism is less dramatic and considerably more depressing.

An advisor whose position depends on your continued tenure does not tell you that your last decision was wrong. He contextualises it. An inner circle assembled over decades on the basis of loyalty — since candour tends to be career-limiting in these environments — becomes, over time, a very expensive mirror. The information that reaches the top has been edited by everyone who handled it on the way up, each of them individually rational, each of them collectively catastrophic. Protocol fills in the rest: rooms go quiet, disagreement acquires the flavour of betrayal, and deference becomes so ambient that it ceases to register as deference at all.

By the time any of this becomes visible from the outside — the confused expression at the booing, the wrong speech read without noticing, the inner circle too frightened to enter the room — the capacity for self-correction is already gone. This is the structural irony: the thing that would allow a person to recognise their decline is precisely what power, given sufficient time, destroys.

---

The historical record is consistent to the point of tedium.

Mugabe ran Zimbabwe for thirty-seven years, from liberation hero to the man photographed reading his wife's Agriculture Ministry speech at a state event without apparently noticing, while his country issued hundred-trillion-dollar notes. Mao launched the Cultural Revolution in 1966 not from ideological conviction but because the Great Leap Forward had damaged his standing internally. His solution was to destroy the institutional structures that might constrain him. He had lived long enough inside a system that confirmed his judgment that the judgment being wrong did not present itself as a serious option. Stalin's inner circle, in 1953, were too frightened to enter his room after he collapsed — they had watched what happened to people who delivered unwelcome news — and he lay on the floor for some hours before anyone summoned a doctor. The system had produced a loop in which danger made truth unspeakable, which made the situation more dangerous, which made truth harder to say.

Churchill's second term (1951–1955) is the sympathetic case, and for that reason the most instructive. His private secretary, his physician, senior ministers — all documented the same decline in diaries not intended for publication. His colleagues could not bring themselves to say it directly. This was not cowardice, exactly. It was fifty years of accumulated deference, so habitual that overriding it felt like a kind of violence. He was eventually forced out, lived another decade in diminishing relevance, and died having predicted several of the disasters that followed his departure. This is not the comfortable ending.

Washington and Mandela are the exceptions, and they are clarifying for the same reason: both arrived at the moment of departure with a belief in the institution as something independent of themselves, formed before the institution had had sufficient time to remake them in its image. Most people do not get out in time. Most people do not notice that they have not got out in time. The machinery sees to it.

---

Which raises the obvious question: are we simply doomed to repeat this?

Not entirely. But the honest answer requires distinguishing between what works and what people prefer to believe works.

Hoping for better people does not work. The record is clear enough that exceptional character provides no reliable protection. The machine does not care. It processes whoever sits inside it.

What does work is institutional constraints designed before they are needed. Term limits are blunt instruments — they remove effective leaders alongside ineffective ones — but they work better than relying on voluntary departure. The American 22nd Amendment came after Roosevelt, not before him. The lesson is that you cannot build the exit mechanism after the person who would block it is already in place. Succession clarity made routine and discussable is similarly protective: the cultures where "what comes next?" is treated as a philosophical challenge to the incumbent's legitimacy are precisely the cultures most vulnerable to indefinite tenure. And independent institutions — a free press, an independent judiciary, a civil service with some protection from political loyalty tests — do not stop the process, but they shorten it. Their value is most visible when they are gone.

---

For bystanders specifically, the calculus is uncomfortable.

Say what you see, early, before the consensus forms. The sycophantic machine runs on the assumption that everyone else in the room also sees what you see and has decided not to mention it. The first person who says something out loud breaks that assumption. It is professionally costly and historically important, which is why so few people do it and why the ones who do tend to be remembered.

Distinguish between the institution and the person. The confusion of the two — treating criticism of the leader as an attack on the country, or the party, or the movement — is precisely what the machinery requires to function.

Support the unglamorous load-bearing structures. Term limits, inspector generals, freedom of information laws, judicial independence — these are not interesting and do not make for compelling political movements. They are, nonetheless, the actual counter-mechanisms. The bystander who votes reliably for their preservation is doing more than the one who writes the better editorial.

The pessimistic residue is this: bystanders historically do very little in time. They do considerable work in retrospect. The documentation of decline — the published diaries, the historians, the biographers — informs the next generation's thinking without saving the current one. The function of honest witness is real but it is mostly retrospective.

So: not doomed, no. The counter-mechanisms exist and some of them work. But they require maintenance before the crisis, which is precisely when they feel least urgent, and they require bystanders willing to be unpopular in rooms where unpopularity has a price.

History suggests this combination is available but not common.