The decisions of dictators

Why do dictators so often make terrible decisions that lead to their downfall?

Dictators have a terrible habit of making truly bad decisions. Sometimes just how bad they are only becomes apparent in hindsight; sometimes it’s perfectly clear to everyone else at the time. Think of Vladimir Putin invading Ukraine, or Saddam Hussein invading Kuwait, or Mao Zedong’s Great Leap Forward. Why, when dictators are seemingly so powerful, do they mess up so badly so often? Why don’t they seek out sensible advice and listen to it?

Economists Georgy Egorov and Konstantin Sonin came up with an elegant theory to explain why this might be.

An autocrat who comes to power makes decisions based on the advice of those close to them. Seeking to preserve that power, they’re likely to pick the best advisers they can.

For the advisers, though, there are different incentives. It might not always be in the advisers’ interests for the autocrat to remain in power, and so savvy advisers will constantly think about what would happen if the autocrat lost power. Would they be better off in the new world, the world run by the opposition? If they would, they might choose to hasten that new world – by dispensing poor advice to the current leader:

“Deciding what advice to provide to the incumbent leader, the agent weighs two factors: first, the vulnerability of the incumbent, and, second, the agent’s own prospects under the new regime if the incumbent fails. Since the leader himself is not as well-informed as the advisor, the advisor has leverage: the advisor might tell the leader that he is not vulnerable, when he actually is. The wrong policy choice by a vulnerable leader will result in a change at the top.”

As a leader becomes more vulnerable, then, the quality of the advice they receive decreases.

But there’s a further vicious cycle involved. As a leader fears the threat of opposition, they might demand loyalty from their advisers. But a vulnerable leader who demands loyalty will receive even worse advice. Those advisers who are best informed will know which way the wind is blowing, and won’t be loyal to a leader with a limited shelf life. Those advisers who lack awareness, and fail to see that the leader is vulnerable, will offer their loyalty – and their poor advice. So, as the leader becomes more vulnerable, the quality of the advisers advising them will decrease; they’ll be surrounded by good advisers deliberately dispensing bad advice, and bad advisers doing so accidentally.

The problems for dictators don’t stop there, either. As well as demanding loyalty, they might seek to shore up their position by cracking down on the opposition. But a leader who represses the opposition harshly is more likely to be killed in a transition of power. After all, if you as the opposition think that you’re going to be killed should the person you’re overthrowing return to power, it becomes very tempting to kill them first.

This final dynamic completes the dilemma. An autocrat who is vulnerable receives less effective advice. An autocrat who is vulnerable tends to crack down on the opposition. An autocrat who cracks down on the opposition particularly fears losing power, because they’re likely to be eliminated if they lose it, and so hangs onto it for as long as possible. So they remain in power, receiving poor advice from loyal but ineffective advisers. At that point, it’s only a matter of time before they mess things up catastrophically.

This elegant model explains the behaviour not just of the leaders of repressive countries, but of all sorts of organisations. How many leaders of companies are subject to the same dynamics? Can Mark Zuckerberg’s disastrous pivot to the metaverse be explained in similar terms? What about Elon Musk’s acquisition of Twitter? How many companies become compromised in their decision-making as autocratic leaders, demanding loyalty, find the quality of the advice they’re receiving diminishes?

The point of Egorov and Sonin’s analysis is not that these dictators are somehow stupid, or irrational. At every stage they’re acting rationally to preserve power. They’re just trapped by something fundamental to the business of being a dictator, some fundamental dynamic that is always likely to emerge at some point. You’re better off, it seems, not being dictator in the first place. Now who’s going to tell Zuck?