3- Collusion Flashcards
What is Collusion?
Coordination of firm behaviour to restrict competition, exert market power and increase profits
What are the 2 main types of Collusion?
Explicit & Tacit
What is Explicit Collusion?
Firms coordinate and agree explicitly on their actions
What is Tacit Collusion?
Firms cooperate without communication, there is no explicit agreement
What are the 2 main components of a deterrence mechanism?
-Detection: colluding firms need to observe cheating
-Punishment: deviator must face credible punishment
Briefly explain the Incentive Compatibility Constraint (IC) for collusion?
The benefits of collusion must exceed those of deviating (profit + discounted firm value)
What is the Incentive Compatibility constraint when firms punish by setting p=mc indefinitely?
Firms reap whole monopoly profits for 1 period
πᴹ/n(1-δ) > πᴹ
What are the 4 steps to show whether collusion is sustainable in a single period?
1.Find equilibrium q, p, π* in absence of collusion as in standard Bertrand/Cournot
2.Find monopoly qᵐ, pᵐ, πᵐ
3.Sub qᵐ into initial FOC, from this find qᵈ, pᵈ, πᵈ
4.Show whether πᵐ or πᵈ is greater
How do you find the monopoly output a single colluding firm would produce?
1.Setup aggregate profit function πᵐ = (p-c)Q
2.Differentiate wrt Q, then solve for Q
3.Divide Qᵐ by n for qᵐ
How do you show whether collusion is sustainable in an infinite time horizon?
1.Sub equilibrium profits from one-shot game into Incentive constraint
2.Solve inequality for δ
How does growing demand change the Incentive Constraint?
A growth factor θ is applied in exact same way as discount factor δ:
πᴹ/n(1-δθ) > πᴹ
How does the multimarket setting change the Incentive Constraint?
Constraints in 2 markets with shares λ & λ-1:
λπᴹ/n(1-δ) + (1-λ)πᴹ/n(1-δ) > 2πᴹ
How does punishing for T periods change the Incentive constraint?
Post-punishment profits discounted at δᵀ⁺¹:
πᴹ/n(1-δ) > πᴹ + (δᵀ⁺¹πᴹ/n(1-δ))
How does punishing by producing q* change the Incentive constraint?
πᴹ/n(1-δ) > πᵈ + (δπ*/n(1-δ))
How does lagging punishment for t+n periods change the Incentive constraint?
Deviator reaps monopoly profits for additional periods:
πᴹ/n(1-δ) > πᴹ(1+δ+…+δⁿ)
