We use experimental methods to examine the impact of costly information on firms' pricing strategies in a differentiated goods environment. The basic underlying feature of this differentiated goods setup is that consumers care about more than just finding the lowest possible price; they also consider how well the product attributes match their tastes and therefore search for a product with the best price-match combination. The experimental design is structured around Anderson and Renault's (2000) theoretical framework. Their model yields different price and welfare predictions depending on whether the information imparted to the consumer relates to price or product attributes. For instance, providing price information decreases the equilibrium price, while providing information about product characteristics curtails consumer search and raises the price. In fact, the price when consumers are match informed is greater than when they receive no information at all. This is in sharp contrast to a widely accepted proposition of economic theory that improved information leads to lower prices, thereby creating a positive externality for the consumers. Additionally, the timing of information also matters. The price is greatest when consumers are informed about the match values from both firms prior to engaging in costly search. This occurs because informed consumers first visit the firm from which their match value is greater and consequently decrease the demand inelasticity. This dissertation research project will use economic experiments to better understand how information about retail prices affects outcomes in markets where there are only a small number of sellers. The researcher will conduct duopoly experiments employing the posted offer trading institution, which closely resembles the conventional retail trading environments. Four different information environments will be compared. The No-Information treatment (NI) forms a baseline to which two other sets of treatments are compared. In the Price Information (PI) and the Match Information (MI) treatments, after all consumers have searched one of the firms, half of them become costlessly informed about the price or their match values from the second firm. Comparing these two treatments provides insight on how the content of information affects firms' pricing strategies. The impact of the timing of information is measured by comparing the Match information (MI) treatment with the pre-search Match information (PMI) treatment, in which half of the consumers know their match values from both firms before searching. This experimental design will improve our understanding of the mechanisms at work in retail market transactions, which are often plagued by informational asymmetries, and thereby provide a basis for sounder policy choices. For instance, various consumer protection regulations stipulated by the Federal Trade Commission that are aimed towards curbing the harms of "imperfect information" should take into account the nature and market impact of the informational imperfection, and not just its mere existence.
