Supra-Competitive Pricing in Laboratory Markets (for the New Palgrave Dictionary of Economics

Despite the robust tendency of laboratory markets to generate competitive outcomes, some market designs deviate persistently from competitive predictions. This essay discusses the primary drivers of supra-competitive prices that have been observed in market experiments. Trading Institutions and Market Performance. The robustness of competitive market predictions stands as one of the most impressive results in experimental economics. Laboratory markets regularly generate competitive outcomes in environments populated by just two or three sellers. However, as in natural contexts, competitive outcomes do not always emerge from laboratory markets. This essay reviews results of laboratory markets in which price increases are driven by factors such as the exercise of unilateral market power or by collusion. Prior to reviewing this literature I offer two observations. First, laboratory methods represent an important but limited complement to existing empirical tools for investigating supra-competitive pricing. Given the stark simplicity and limited duration of laboratory markets, experimentalists can aspire to say little about specific naturally occurring markets. Experiments can, however, provide important insights into the behavioral relevance of theories upon which antitrust policies are based. Second, the trading rules defining negotiations and contracting can exert firstorder effects on market competitiveness. For example, markets organized under the double auction trading rules used in many financial exchanges, are much more robustly competitive than markets organized under the posted-offer trading rules used in most retail exchanges. Indeed, one of the motivating factors in the emerging field of institutional design was an interest in developing institutional rules that promoted efficient market outcomes. For specificity I focus here on results from posted-offer markets, primarily because posted-offer markets allow a particularly intuitive illustration of the factors affecting market competitiveness. However, a host of other trading institutions exist, ranging from single and multi-unit auctions, to multi-sided computerized ‘smart’ markets, and again to institutions that exist primarily as theoretical constructs, such as quantitysetting Cournot mechanisms. The competitive implications of each of these institutions must be evaluated independently. Posted-Offer Markets and Unilateral Market Power. Unilateral market power is perhaps the most frequently observed reason why prices in laboratory markets deviate from competitive predictions. This market power exists when one or more sellers, acting on their own, find it profitable to raise prices above the competitive level. The supply and demand structures, shown in the two panels of Figure 1 illustrate how capacity restrictions can create market power. In each panel, the market consists of three sellers, S1, S2 and S3, each of whom offers four units for sale, under the conditions that two units cost $2 and two units cost $3. A buyer will purchase a fixed number of units (seven in the left panel or ten in the right panel) at prices less than or equal to $6.00.