He woke up with a vague sense of uncertainty. Much earlier, in 2016, he had defended in New York that Acemoglu and Robinson in their “A Theory of Political Transitions” of 2001 had chosen on dubious grounds to package their message in dense mathematical language. This fits in with the suspicion that political power positions can be acquired through economic-discipline based knowledge asymmetries. A suspicion that was reinforced for him by documentaries such as Charles Ferguson’s “Inside Job” from 2010.

The piece is about a model that describes political dynamics (the ability to alternate between democracy and dictatorship) on the basis of political-economic cycles in which individual variables (rich-poor, extent of democratic control, willingness to revolution yes-no, income , expenditures, taxes), collective variables (tax level, democracy dictatorship, willingness to give democratic control) and memory variables (was there a revolution in the previous cycle yes-no) play a role.

This set-up is perfectly suited to show with the help of agent-based simulation where heuristics can be made visible in a way that is clear to alpha minds. But a mathematical approach has been chosen that allows to define an equilibrium: roughly with “a pure-strategy Markov perfect equilibrium is the combination of strategies…

such that the equations (1) and (2) are solved.”

His problem with this approach is that the mathematical representation more or less obscures that, in order to make it operational at all, it requires an unacceptably extensive restriction of the parameter space that the model should support. A quote as an example: “We treat all poor agents as identical, and all members of the elite are also identical.” (p. 941) Since the model only knows these two types this reduces the mathematical complexity of a workforce and a networking elite to two straw or front men.

His uncertainty was based on the rock-solid trust that his brother, who is an economist, has in the qualities of Acemoglu and his people, and this has been increased by the last two books by Acemoglu and Robinson that contain qualitative analyzes of state forms and prosperity development (2012) and of the developments that relate the emergence, affirmation and loss of civil liberties (2019) to the dynamics in network configurations of economic, legal and political players (he characterizes them as different types of Leviathan).

These two books are a fantastic source for further research using simulations. Acemoglu’s Leviathan is then a complex adaptive system (something Hobbes has already foreseen) and his descriptions and analyzes lead to heuristics / hypotheses that can be modeled and whose initial values can be parameterized and whose effects can be analyzed and used for calibration. As a slow but deliberate form of deep learning.

And, oh yes: eco stands for household, and household encompasses more than just the wallet. The work of Acemoglu cs shows this quite clearly by becoming more ecological than economic over time.