Creating Institutions to Address Externalities TMS

Move from inefficient NE to SO Sometimes society can be stuck at a NE that is not efficient (not SO) The role of institutions is then the equilibrium to the SO The problem of IEAs is to devise institutions that are capable of doing so when monitoring and compliance are imperfect

Consider first – domestic context “Chicken” Game e.g. racing cars at one another He who “chickens out” loses big He who does not, is big winner If both do not, then both big losers

Strategic Form of Game Players (row, column) Strategies (pure, mixed) Pay-offs (row, column) Information (common knowledge) Play (simultaneous)

Chicken Game

NE in Strategic Form Games Recall the definition of NE (Nash conjecture) Each player chooses its optimal strategy given the environment they face Each player ignores the possible effects of its choice on other player’s choices  NE exists at any point where strategy choices are optimal given all other strategy choices (i.e. there are no incentives to re-optimise by any player given the environment in which they find themselves)

NE in Chicken Game Consider “Pure Strategies” (played with probability of 1.00) If Row plays Abate, and Col plays Pollute  No incentive for either to move unilaterally Likewise, if Col plays A and Row plays P  NE of A,P in Pure Strategies

NE in Chicken Game Note that the SO in Chicken Game is A, A  Joint payoff is higher for A,A than A,P But A,P is NE and A,A is not (because R would elect P if starting in A,A) ( and C would elect P if starting from A,A)

Move from NE to SO: What Institution? In this game we are looking for some sort of institution that will perform this “asymmetric matching” of strategies  When R goes, C stops (and vice versa) What sorts of institutions perform “asymmetric matching” like this? - when walking through doorways - when deciding who can put up a house on land

Role of Institutions In an uncoordinated environment, it can be the case that it is not possible to negotiate over every resolution (e.g. in chicken the car drivers are hurtling towards one another, and at intersections we have the same problem)  Institutions can move the uncoordinated players from the NE to the SO by the simple adoption or recognition of a particular rule or standard

Consider another Game – Pure Coordination

Pure Coordination Game What are the NE in this game (pure strats)? How do we determine them? Consider each element of the strategic game matrix and ask: does either player have an incentive to move the game out of this (poss) equilibrium?

Pure Coordination Game NE in Pure Coordination Game are: P,P and A,A SO in PC Game is P,P

NE and SO An institution may be necessary to move this society from NE to SO, if it finds itself located at the inferior NE Ex. Driving on left versus, driving on the right - best thing is for both cars to drive on opp sides, but it is prob best if all countries drove on the right (why?)

Required Institution: Coordination Game Again, if time and communication are possible, then a negotiated solution is probably possible (agree to both move to A,A since it is both NE and SO) -if time is not available, then it might be necessary to adopt some sort of custom or rule that indicates which strategy both will adopt (classic example is: who should phone back if cut off during call? Does not matter so long as both agree on the custom or std that prevails.)

Institutions for Solutions We now have 3 different forms of interaction and 3 different forms of institutional solutions: 1) NE involving excessive use (externalities)  negotiated contract (Coase) 2) NE requiring asymmetric matching (chicken)  Property rights institutions 3) NE requiring symmetric matching (coordination)  Coordinating Standards

Basic Lesson Interaction in the absence of correctly constructed institutions may lead society to rest at a Nash equilibrium (where all individuals or agents have optimised and have no incentive to re-optimise relative to perceived payoffs) NOT EQUAL to SO  Construction of institution enables new equilibrium at the SO

Fundamental Problem of International Interaction Due to the doctrine of national sovereignty, it can be very difficult to monitor or enforce agreed contracts between agents  Agents may choose to re-optimise (if payoffs reward re-optimisation) even if a contract or institution has been agreed that would indicate that the agents will choose otherwise

Consider Prisoners Dilemma Agents unable to communicate or enforce contracts Payoffs constructed to reward asymmetric matching but P,P is a NE (this time, unlike the chicken game)  Incentive to re-optimise to P from every starting point, and no capacity to enforce any other agreement

Prisoners Dilemma