Collusion and price dispersion

Abstract

While there are suggestions in applied cartel studies that price dispersion changes when cartelization of a market occurs, there are few theoretical or empirical analyses of this effect. This article surveys the thin economic literature on the link between overt collusion and price dispersion. Formal theories and observation of cartel behaviour suggest that during successfully collusive periods prices become less variable and more negatively skewed compared to relatively competitive periods. Four empirical studies of cartels verify these predictions.

Notes

Mund (1960) notes that identical bids on project tenders is viewed as strong evidence of collusion. Elzinga (1984: 22) surveys cartel studies and mentions almost in passing ‘… the presumption that prices under cartels should be higher, less variable over time, and more uniform at a point in time, all other things equal’. There are no sources given for this notion.

A third factor, the density of demand, ‘plays a minor role’ (p. 490) and disappears when the cost curves become flat.

There are studies of cartel stability that examine the variation in prices as a measure of the degree of cartel discipline across multiple episodes (Suslow, 2001). Classic examples of reversion to more competitive pricing conduct are price wars. However, these studies are motivated by an interest in why cartel consensus breaks down from time to time or whether price-stabilization schemes are successful. They do not examine intra-episode price variation.

The product must be delivered in rail cars or truck trailers that cannot be used for any other product.