## Theories of oligopoly - non-collusive

The various models of oligopoly can be classified under two main headings: non-collusive or competitive oligopoly and collusive oligopoly. We shall consider each in turn:

### Non-collusive or competitive oligopoly

In this case, each firm will embark upon a particular strategy without colluding with its rivals, although there will of course still exist a state of interdependence, as possible reactions of rivals will have to be considered.

There are three broad approaches that might be adopted by firms in a situation of competitive oligopoly:

- Observe the behaviour of rival firms but make no attempt to predict their possible strategies on the basis that they will not develop counter strategies. This was the essence of the earliest model of oligopoly developed by Cournot as far back as 1838: each firm acts independently on the assumption that its decision will not provoke any response from rivals; this is not generally accepted nowadays as providing a useful framework in which to analyse contemporary oligopoly behaviour.
- Make the assumption that a given strategy will provoke a response from competitor firms, and assess the nature of the response using past experience. This is the basis of the kinked demand curve model, described below, in which it is assumed that any price cut by one oligopolist will induce all others to do likewise, whilst a similar price increase would not be matched.
- Formulate a strategy and try to anticipate how rivals are most likely to react, and be prepared with suitable counter measures.

This is the basis of **game theory** in which competition under oligopoly is seen as being similar to a game of chess in which every potential move must be regarded as a strategy, and possible reactive moves by opponents and subsequent counter-moves must all be carefully considered. The application of the theory of games to economics was first introduced in 1944 by J. von Neuman and O. Morgenstern. Game theory involves the study of optimal strategies to maximise payoffs, taking into account the risks involved in estimating reactions of opponents, and also the conditions under which there is a unique solution, such that an optimum strategy for two opponents is feasible and not inconsistent. A zero-sum game is one in which one player's gain is another's loss, and a non-zero-sum game is one in which a decision adopted by one player may be to the benefit of all. Look up 'The Prisoner's Dilemma'!

For some further resources on the Prisoner's dilemma, why not have a look at:

- Serendip - a prisoner's dilemma game/simulation
- Prisoner's dilemma - an online simulation
- Prisoner's dilemma - Wikipedia