Dictators and Their Viziers, Absent and Present
A preview of a paper in progress, which I will present next week at Yale. Zhaotian Luo, Spencer Pantoja, and I have been working on "A Relational Theory of Autocracy" that borrows from organizational economics to understand the nature of the relationship between dictators and their right-hand men. It's a theory paper, a dynamic agency model with the peculiar feature that the agent can try to become the principal—in our context, through a palace or military coup.
We build on earlier work by Konstantin Sonin and Georgy Egorov on "Dictators and Their Viziers." I remember Konstantin and Georgy working on their paper nearly 20 years ago, at the ISNIE meetings in Tucson. (Gary Libecap recruited the local stargazers club to place telescopes on the hotel roof for the gala dinner. We spent the evening looking at the moons of Jupiter and passing satellites.) Theirs was the first model of the loyalty-competence tradeoff in autocracies—a topic that Alexei Zakharov also explored.
Our model departs from this earlier work in three ways. First, rather than assume the presence of a vizier, we ask whether the ruler would recruit an agent of randomly drawn competence. Second, in our formulation, the threat to the regime comes from the agent himself, not from some outside actor. Finally, all of this happens in a dynamic setting, so that today's agent may be tomorrow's principal, with his own recruitment dilemma and agency concerns.
Three ideal types of the dictator-vizier relationship emerge from our analysis of this model, each represented by a particular historical regime.
In the first case, there is no agent whom the ruler would recruit to a stable relationship. We call this the Idi Amin Principle, as during his eight-year rule, Amin had no regular right-hand man.1 This arrangement is likely when the dictator is himself too incompetent, too bumbling, to trust that he can win a contest for power; the rents he would need to share to prevent a coup outweigh any benefit of having a right-hand man. Personalist autocracies with little power sharing also emerge when the rents from holding power are large, as the ruler must constantly worry that a recruited agent will try to seize those rents for himself.
Of course, many autocrats do have right-hand men, but that doesn't mean they would recruit just anybody. Rather, they balance the advantages of competence (more help in extracting rents, exerting control over the economy, or whatever the dictator hopes to accomplish) against the costs of ensuring that a competent agent stays in his lane. In general, dictators recruit agents who are neither too incompetent nor too competent. We call this the Molotov Principle, after Joseph Stalin's right-hand man—not quite "mediocrity personified," as Trotsky called Vyacheslav Molotov, but also not the brilliant strategist and rhetorician that was Trotsky.
Finally, more competent rulers are willing to recruit more competent agents. This is the Atatürk Principle. For most of Kemal Atatürk's fifteen years in power, he comfortably relied on the able assistance of İsmet İnönü, his successor as president of Turkey. Atatürk could afford this relationship, as he was himself supremely competent. Such alliances of talent rely not only on the competence of the ruler, but also on the right agent coming along at the right time. Atatürk was lucky to have available a brilliant diplomat and military strategist, an autodidact with political savvy, when he set out to modernize Turkey. Such partnerships are the exception, not the rule.
1 Amin frequently served, for example, as his own minister of defense—a typical arrangement in "big man" autocracies with little power sharing, as documented by Killian Clarke, Anne Meng, and Jack Paine. ↩