Roger Myerson, “The Autocrat’s Credibility Problem”

The Autocrat’s Credibility Problem and the Foundations of the Constitutional State

Roger B Myerson

APSR 2008

Hardin’s “Why a Constitution” was pretty theoretical, but Myerson goes even a step deeper into theory in the Autocrat’s Credibility Problem. Actually, at the end of the paper, Myerson cites Hardin’s Why a Constitution paper. (I bet throughout this project, all of the papers I review will probably be linkable through citations, like a 6 degrees of separation thing…)

Anyways, this paper looks at trust. Trust is incredibly important to game theory and politics as well as our daily interactions with others. In many ways, trust in this article and “trust” as we normally think of it are similar. They both look to future actions of someone else, and our expectations as to what we think they will do.

Credibility/trustworthiness seems to be the ability to make promises which you can reliably assume other people will reliably assume you will follow.  Both parts are important: first that other people will expect you to follow through and that you expect them to expect you to follow through.

But, can’t we all just get along? Shouldn’t we just trust one another because we should give people the benefit of the doubt? Why is our default distrust? Or is it? Regardless of the normative or empirical answers to these questions, Myerson creates many models of why autocrats need a check on themselves if they break promises in order to maintain their status as leader.

What’s the main dilemma facing the autocrat and his/her captains? From page 134:

(1)   “an individual captain should not give costly support to a leader who is not expected to reward the support”


(2)   “a leader has no incentive to give costly rewards to past supporters if withholding these rewards would not reduce his expectations of future support.”

So, in the mind of the captain (or someone like a precinct leader, or someone who helps the leader get and maintain power):

“Ok, I have no guarantee that this guy is even going to keep me in the circle of people that get rewarded. He could just blow me off as soon as he gets into power. It certainly wouldn’t be the first time someone played “defect” in a prisoners’ dilemma game. Individually, I am also too weak to take him over unless I get a critical mass of other captains to overthrow him with me. And, sometimes we are bad communicators and even when enough of us don’t like him, we don’t do anything about it. Then we are all suckers. I am not going to be a sucker.”

In the mind of the autocrat:

“I need these guys to get into power and keep power. Whether or not I plan to try and make them into suckers, they are probably going to assume I will try to make suckers out of them. I’ll need to use my awesome charisma or something to initially get them to choose me as the leader, but in order to sustain it, they will need something more than my word. My word is just cheap talk. So, I will create this kind of institution like a court, but not the kind with robes and justices and whatnot. This court will be made of the captains, and they will sort of all get together and talk about me. If they are all unhappy with me, I guess they can kick me out. (But that is pretty unlikely- most people will probably only support a coup if other people support it first) I? They? We? will create these rules for my conduct that basically amount to “the autocrat shall not make suckers out of the captains”.

And those rules become his “personal constitution”. And they constrain him but also allow him to maintain power.

I am unclear on the transition, but we next look at the chicken game.

Chicken Game

Player 2Claims Player 2Defers
Player 1Claims -k, -k(Bad!) R,0(Good! For P1)
Player 1Defers 0, r(Good! For P2) 0,0 (Neither good nor bad? Inefficient?)

In this game, either player would like to lay claim to a piece of property, but only if the other person is not also going to lay claim to it. If they have reason to reliably believe the other person will play claim, they will probably want to defer. Likewise, if they reliably think the other person will not claim the property, they will go for it. But, if they can’t communicate about it, or they don’t trust the other person, it’s a tricky game.

There are three equilibria to this game as Myerson explains: 2 pure strategy equilibria where one claims and the other defers, and a third equilibrium where each player lays claim with a probability of r/(r+k) and the expected payoffs are 0.

What’s this have to do with Autocrats?

If there is one person that everyone recognizes as a source of authority, like the autocrat, then that autocrat has the power to use “focal point law” to essentially pick the winner of the chicken game (and the two cars don’t even have to crash into each other! Reference to PS 260 / AMCS 2601 “Game Theory in Science and Culture”). What that means is that it can become a “self fulfilling prophecy” as Myerson calls it as to who will actually choose to lay claim or defer.

If the mighty autocrat says that Player 1 should lay claim to the land, then that statement alone will likely have the effect of making P2 think that P1 will likely lay claim. As such, even if P2 is unhappy with that outcome, he will still likely defer. Then P1 will know that P2 knows that and P1 will feel confident in laying the claim.

So, this is just one of the perks of being the autocrat. You get to say who gets to lay claim to land. Well, not necessarily literally, but the chicken game is just an example of a class of games. But, if you get to say who can get certain benefits just by the “suggestion” of it, you can make that into a profit for yourself.


In a perfect segue to what I want to read next, Myerson ends on democracy. “A democratic constitution would be imperiled if its most powerful office were held by a leader who could be confident that his active supporters would still trust him after he openly violated the constraints of the constitution.” So, does the modern written constitution do the same thing, require the same things? To be continued!

What’s Next?

I think Barry Wiengast’s Political Foundations of Democracy will be next. Based on the abstract, he seems to apply these ideas to democracy and democratic functions: “It must be in the interests of political officials to respect democracy’s limits on their behavior”.

Parting thoughts:

(1)   Isn’t it possible to get credibility without having explicit constraints on you? What if you are “honest Abe”  and you just build a reputation of telling the truth? If you are playing an iterated prisoners’ dilemma, if you say you are going to play cooperate and then you do time and time again, even when no one can make you tell the true or punish lying, can’t you develop a reputation for being honest and then forgo the constraints?

(2)   What are more specifics on the collective action problems related to organizing a coup to overthrow a bad guy? Paper suggestions?

(3)   Myerson writes, “But laws are effective only to the extent that people are expected to punish their violations” pg. 127. This sounds like someone who hasn’t thought about the focal point/expressive nature of law. That sounds like law as only being a vehicle for cooperation enforcing. But, Myerson is definitely aware of expressive law and focal points and he brings it up starting on 133. So, what does that quote mean?


Posted on June 14, 2011, in Uncategorized and tagged , , , , , , . Bookmark the permalink. 3 Comments.

  1. Michael Wiebe

    “collective action problems related to organizing a coup to overthrow a bad guy?”

    Tullock, Paradox of Revolution

  1. Pingback: Wanted: Dead or Alive « The Law Rules Blog

  2. Pingback: Cheap Spam and Cheap Talk: Scam Artists Don’t Know Their Game Theory « The Law Rules Blog

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: