Markets with more than one entrant, more than one participant, but with a relatively small number of participants who have market power
Inbetween PC and monopolies
Firms don’t cooperate
what you’re going to do in a game, is based on what you think your opponent is going to do
Game theory - equilibrium concept
The point at which the game ends, the point at which all of the players are satisfied. In particular, when the market is actually in equilibrium
When no firm wants to change its actions, given what the other firms are doing; everything is basically stable
The Nash equilibrium consists of the strategies of each firm, so it is not a strategy by one player. A market with more than one participant is a non-monopolistic market. The Nash equilibrium often does not maximize social welfare.
The dominant strategy is the best strategy no matter what the other player does. The strategy does not necessarily maximize social welfare. There will be a dominant cooperative strategy and a dominant non-cooperative strategy.
In the prisoners’ dilemma - If the prisoners cannot communicate, what is the dominant strategy in the prisoner’s dilemma scenario presented in lecture?
The dominant strategy is to talk to the police. If the other prisoner talks, then you receive 2 years in jail if you talk and 5 years in jail if you stay silent. If the other prisoner stays silent, then you spend 0 years in jail if you talk and 1 year in jail if you stay silent. In either case, talking is the best strategy.
Inefficient Nash Equilibria
Non-cooperative equilibria will be worse than the cooperative equilibria, but still be the stable Nash equilibrium
How does a repeated game change the dominant strategy?
Essentially, you could use threats to enforce the cooperative equilibria and solve the problems of an inefficient Nash equilibria
(As long as you don’t advertise, I won’t. But, if you advertise once, then I’ll advertise forever)
What is required in repeated games, to ensure cooperative equilibria?
You have to commit to this decision forever. If you plan on ‘exiting’ the game, your incentives will change the last year…which will change the incentives the year before, etc. ultimately returning to the non-cooperative equilibria
Given that you know the repeated game runs out, you’ve destroyed the repeated game. So, repeated game solution is only good if it goes on forever
Cournot model of non-cooperative oligopoly
Set of quantities for each firm, holding other firms quantities constant, no firm can achieve a higher profit by choosing a different quantity (similar to Nash equilibrium concept)
The Cournot model is a model of non-cooperative oligopoly by definition. The difference between the Cournot model and the prisoner’s dilemma is that firms have a continuum of choices about what quantity to produce in the Cournot model.
Five steps of cournot model
- Solve for residual demand - what is the demand for your good, given what the other person is selling
- Develop MR function of the other guy’s quantity
- Set MR = MC for profit maximization
- Repeat this for all firms
- Solve “n” equations in “n” unknowns
Reaction/best response curve
Dominant strategy given what the other firm will do, for every possible decision the other firm could make (profit maximizing quantity given the residual demand for every single output that the other firm could produce, etc)
Where the best response curves of the firms intersect
The point at which they are both satisfied; can’t do any better given what the other firm is doing. No reason to undercut the other firm at this point; profit-maximizing point for both firms.
By definition, the Cournot equilibrium is the stable equilibrium where firms do not have an incentive to deviate. This will be when both firms have a dominant strategy, which will be when the best response curves intersect. This may not be when marginal cost equals the price, because in a Cournot model, firms have market power. The Cournot model is a model of non-cooperative oligopoly.
Oligopolistic actors get together and behave cooperatively (basically as a monopoly); makes every producer in the market better off
By cooperating, cartels can restrict the quantity supplied to the monopolistic solution and charge the monopolistic price. The monopolistic quantity and price will maximize overall producer profits, which oligopolists can then split amongst themselves.
Why don’t we see more cartels?
- Cartels are unstable - each participant has an incentive to cheat. By cheating, they get the benefits of additional sales at half the cost (shares the burden of the lower price with other participant)
If firm B increases production, then the price of the good will fall. Firm B’s profits will increase because they get all the benefit from the increased production and only half the loss from the poisoning effect (the fall in price). Firm A’s profits will decrease because they get no benefit from the increased production but are hurt by the poisoning effect. Combined profits will decrease, because combined profits were maximized at the monopolistic solution, but after firm B cheats, production is higher and price is lower than at the monopolistic solution.
- Cartels are illegal. Anti trust laws were created to break up cartels (aka trusts)
How does the quantity produced by non-cooperative oligopoly actors compare to monopoly and perfect competition? What about social welfare?
Non-cooperative oligopolistic actors will have higher quantity produced and lower profits than a monopolist, because the oligopolistic actors will not fully internalize the poisoning effect from increasing production, which will lead them to produce more than the monopolist and drive down overall profits. Non-cooperative oligopoly leads to lower social welfare and lower quantity produced than perfect competition because the oligopolistic actors can use their market power to restrict supply, which will prevent socially beneficial trades from occurring.
In general, what can we infer about social welfare as we study profits & quantity produced across various equilibria?
At the highest quantity, you’ll have zero profits (perfect competition), and you’ll have no DWL. You’ll maximize social welfare.
As the quantity goes down, you start to move towards oligopoly and monopoly equilibria - meaning the quantity produced overall is lower, profits are higher, and you’re moving away from the competitive equilibrium & creating DWL - therefore lowering social welfare.
As the number of firms increases in the market…
You get closer to perfect competition, increase social welfare, and lower profits.
As the number of firms increases in cartels…
it more quickly breaks down, much harder to enforce the cooperative equilibrium
Trade-off with mergers?
Market power vs. economies of scale/scope
Bertrand model of price competition
Model of price competition, compete over price and let the market tell them the quantity
Two firms may be enough to reduce market power if the firms are identical. Price is driven down to marginal cost.
In the Bertrand model, oligopolistic firms compete over price instead of quantity. Each firm can capture the entire market if they slightly reduce their price. This incentive will hold at any price above marginal cost (where capturing the entire market yields profit), so price will be driven down to roughly marginal cost.
Bertrand model vs. Cournot model - when is each more likely?
Cournot model, quantity competition - more likely when there are lags in the production (i.e. cars)
Bertrand model, price competition - more likely when there is instantaneous production (i.e. cereal)
How can a firm in a market with short production lags avoid price competition?
They differentiate their products.
Only in Bertrand competition when products are IDENTICAL.
Promotes innovation, improves welfare
product differentiation would allow the firm to patent their product and protect themselves from price competition.