POLS1002 Lecture Notes - Lecture 5: Albert O. Hirschman, Andrew Leigh, Repeated Game
23 May 2018
School
Department
Course
Professor
L5 COOPERATION AND CONFLICT
1. Why are cooperation and conflict central concepts in political science?
2. What is game theory?
3. Is game theory a useful tool for explaining 'real world' collective action problems?
Game Theory
- Pisoes Dilea → represents the basic structure of collective action problems for the provision
of public goods, each actor has a dominant strategy not to cooperate, leading to an inefficient
outcome in which all the participants are worse off than if all cooperate. No zero-sum game.
o If change to iterated game with finite repetitions, what do you think will happen?
o Prisoners will cooperate until the last round, at which point it will become a one off.
o Solutions: Repeated interaction → Aelods Tit fo Tat, Putas soial capital (bowling alone),
Ostos o-production e.g. community health, neighbourhood watch.
- Tit for Tat → strategy of starting to cooperate and doing unto others as they do unto you can lead
to sustained cooperation. Mutual cooperation is more likely the greater the uncertainty as to the
length of the collective relationship and the higher the no. of interactions.
- Backward Induction → reasoning leading to mutual defection when the end of the interaction is
known.
- Conflict or zero-sum game: an interaction in which the gains for some people imply losses for
others.
*Colomer only refers to exit and voice. Loyalty is implied.
- Collective Action Problem
o Whether or not to contribute to collective action depends on the collective action function
▪ R = B*P – C + D
e.g. Andrew Leigh MP BBQ flyer.
R: <> 0, B: Clean park space, P: 100%, C: Opportunity cost i.e. Saturday afternoon, D: BBQ
▪ Interested in a clean park, but cost of contributing to clean up may be too high time energy,
materials, risk, expectation of success
▪ Voice, complain/suggest alternative for clean up
▪ Exit, to hell ith the pak, I oig to a e eighouhood ith leae paks.
- Exist, Voice and Loyalty
o Influenced book by Albert Hirschman (1970)
o Outline how to deal with undesirable changes to oes pesoal eioet.
o E.g. Changes to status quo
▪ ANU increases student fees by $2000
• ANUs deisios of that to do depeds o hat the studets do. Cat aise tuitio if all the
studets ae goig to uit. Doest at to eate situatio hee it looks ad i the edia,
or is at the receiving end of vandalism.
▪ Govt. decriminalises soft drugs
▪ Govt. welcomes 25, 000 Syrian refugees
▪ Drop voting age to 16
find more resources at oneclass.com
find more resources at oneclass.com
▪ What can you do?
o Hirschman says that there are three options
▪ Exit: Accept the unpleasant change to your environment and alter your behaviour to achieve
the best outcome possible given the new environment
▪ Voice: Complain, protest, lobby, take direct action to try and return the environment back to its
original condition
• If you participate in a mass demonstration (collective action, therefore subject to collective
action function)
▪ Loyalty: Accept the e app situatio ad ake o hage to ou ehaiou
▪ How you respond depends on several factors
• What do you expect to happen when you make your exit/voice/loyalty choice?
• Is eitig a edile theat?
• Your strategy might be contingent on what another actor chooses
- Game Theory
o Can use game theory to analyse strategic situations where my decision of what to do in order to
achieve my goals depends on what you are going to do in order to achieve yours goals and vice
versa.
o Not really a theory, more of a tool predicated on a set of assumptions
o Gae: a situatio i hih a idiidual’s aility to ahiee her goals depeds o the hoies of
others
▪ Games have players >=2, e.g. Student v ANU
▪ Student has to choose: exit, voice, loyalty
▪ If student chooses voice, ANU has to choose to react: back down on free increase, or ignore
▪ If ANU ignores, then student has to choose exit or loyalty
o Games have strategies: a complete plan of action that outlines what players would do in any/every
situation that can arise in the game
o To solve the game, we have to identify strategies that a rational decision maker would make is she
is trying to do as well as possible e.g. chess
o Nash equilibrium: a set of strategies in a game such that no player has an incentive to unilaterally
change her mind given what other players are doing → prisoners dilemma. Equilibrium = no
incentive to change → max min strategy. E.g. having to meet some at ANU but not telling there
where or when.
o Focal point is similar to a social norm – a behaviour that is considered appropriate without having
to resort to the use of laws to coordinate behaviour e.g. what side of the sidewalk to walk on. In
one way an adhered to social norm is a kind of Nash equilibrium. However, you can rig the game to
get what you want.
o Payoffs: the rewards that are associated with each outcome of the game. An important rule of the
game is that players prefer higher payoffs for lower payoffs.
o Extensive form games: players make their moves sequentially, like chess
o Normal/strategic form games, players make moves simultaneously, like paper/rock/scissors.
No zero sum
Zero sum
Mutual cooperation
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Tit for tat strategy of starting to cooperate and doing unto others as they do unto you can lead to sustained cooperation. Mutual cooperation is more likely the greater the uncertainty as to the length of the collective relationship and the higher the no. of interactions. Backward induction reasoning leading to mutual defection when the end of the interaction is known. Conflict or zero-sum game: an interaction in which the gains for some people imply losses for others. Collective action problem: whether or not to contribute to collective action depends on the collective action function, r = b*p c + d e. g. andrew leigh mp bbq flyer. R: <> 0, b: clean park space, p: 100%, c: opportunity cost i. e. saturday afternoon, d: bbq. Influenced book by albert hirschman (1970: outline how to deal with undesirable changes to o(cid:374)e(cid:859)s pe(cid:396)so(cid:374)al e(cid:374)(cid:448)i(cid:396)o(cid:374)(cid:373)e(cid:374)t, e. g.
