Modeling Leadership
Tuesday, December 24, 2013
Daron Acemoglu and James Robinson

So leadership does seem to matter. But then, how do we incorporate into our political economy models?

The answer is not clear, mostly constituting an area for future research, though there are a few attempts.

One approach to this has been pioneered in organizational economics, for example by Benjamin Hermalin’s work, focuses on how leaders, such as top managers, can take actions to “lead by example” and to motivate other workers in the organization (e.g., work visibly hard on a project to signal the value of putting more effort for a specific project).

Roger Myerson’s work has pursued a different line, modeling leadership as a form of reputational equilibria (the leader builds a reputation for sticking to his word, for example).

Perhaps more important in political economy is the role of leaders in solving collective action problems and coordinating beliefs and behavior. There is much less on this.

One approach is developed by Daron and Matt Jackson. Leaders are those individuals whose actions are “prominent” meaning that they are seen by more agents and more precisely. In a dynamic coordination game, this creates an advantage. If an individual who is not prominent wishes to switch from the Pareto dominated equilibrium of the coordination game to the Pareto dominant one, she will rightly worry that others will not see her action or its signal or that the signal will not be sufficiently informative. The prominence of leaders enables them to circumvent this problem, enabling them to change the equilibrium if they want to.

There are several directions for extending this. First, rather than taking prominence as exogenous, one can try to endogenize prominence by the past actions of an individual. For example, Nelson Mandela became prominent because of his key role in the African National Congress’ struggle, his organizational efforts and then unwillingness to make a deal with the Apartheid regime.

Second, one can introduce similar ideas in the context of a game of collective action (rather than simply a game of coordination between two players). Then the role of leaders would become closer to that of “solving the collective action problem,” a hallmark of political leadership.

This work takes a first, and specific, step in thinking about the role of leadership as coordinating beliefs, but there are other possibilities. For example, leaders solve the collective action problem not because of their prominence, but by introducing new ideas (something we will discuss in later posts) or by more effectively communicating their actions or information (for example, by forming or exploiting their social networks). But these ideas have not yet been, at least to the best of our knowledge, systematically investigated.

Another important role of leaders, which has been investigated even less, is to form new coalitions. They may be able to solve again affecting beliefs or by exploiting their network of friends, followers and associates.

Article originally appeared on Why Nations Fail by Daron Acemoglu and James Robinson (
See website for complete article licensing information.