The Effects of Collusion in Laboratory Experiments

Introduction At least since the time of Adam Smith (1776), economists have believed that sellers are likely to discuss common price strategies, but that incentives to defect may be strong enough to break down collusive agreements. Collusion is difficult to study in naturally occurring markets, since sellers will try to hide illegal activities from buyers and from antitrust authorities. Even when a conspiracy can be established, the effective dates, the exact identities of all participants, and the real effects on prices are typically difficult to prove. In particular, price-fixing may result in stable common prices, but it is usually impossible to say what prices would have been without collusion, at least in the absence of precise cost and demand conditions. Even when markets seem to alternate between collusive and noncollusive phases, the price differences are difficult to interpret since a breakdown in collusion may be caused by a demand decrease that would have reduced prices in any case. This makes the laboratory an ideal setting to study factors that facilitate or hinder illegal price fixing. The main result from the experimental economics literature is that the market trading institutions are crucial in determining whether or not collusion will be successful in raising prices above competitive levels. The initial paper on this topic was Isaac and Plott (1981), who examined the effects of seller discussions between rounds of a continuous double auction. In particular, the sellers were allowed to come together to a corner of one of the rooms and confer after the close of one trading period and before the start of another. Attempts to collude were as ineffective as they were inevitable. The problem is that each seller has a strong private incentive to defect and lower the asking price during the course of the double auction trading (Clauser and Plott, 1993). In some sense, the Isaac and Plott (1981) result is consistent with Smith's (1981) finding that even a single-seller monopolist could not always find and enforce near-monopoly prices when trading with a number of buyers in a continuous double auction. Smith, however, did observe consistent supra-