Applying detailed consecutive daily micro data at the gasoline station level from Sweden we estimate a structural model to uncover the degree of competition in the gasoline retail market.We find that retailers do exercise market power, but despite the high upstream concentration, the market power is very limited on the downstream level. The degree of market power varies with both the distance to the nearest station and the local density of gasoline stations. A higher level of service tends to raise a seller's market power; self-service stations have close to no market power. Contractual form and brand identity also seem to matter. We find a clear result: local station characteristics significantly affect the degree of market power. Our results indicate that local differences in station characteristics can more than offset the average market power found for the whole market.

We estimate a structural model to uncover the degree of competition in retail gasoline markets using daily station-level data on quantity and price from the Swedish market. The structural model enables us to consider key features on both the demand and supply side that are important when evaluating retailers' ability to obtain market power. Endowed with station-level information on service level, contractual form and number of nearby stations, we take into account the main drivers of differentiation in the local market. Our findings suggest that retailers in general exercise significant intermediate levels of market power. Further, local station characteristics significantly affect to which extent stations are able to extract market power. Results are robust to different estimation methods.

No abstract

Firms may want to coordinate industry-wide price jumps that are predictable for rivals, however, unpredictable for consumers. We show how such coordination is carried out in Norwegian gasoline retailing. Overnight, the market leader initiated an equilibrium transition from regular to non-regular price jumps. Prior announcements of a non-transaction price variable, recommended prices, are used to coordinate the timing and the level of industry-wide price jumps.

First, we analyze how regular days off from competition and a time-dependent price pattern affect firm performance. Second, we examine the effects on firms' profitability from consumers' changing search-and timing behavior. We use microdata from gasoline retailing in Norway. Since 2004, firms have practiced an industry-wide day off from competition, starting on Mondays at noon, by increasing prices to a common level given by the recommended prices (decided and published in advance). Hence, firms know when and to what level to raise their price. In areas without local competition, retail prices are always equal to the recommended prices. Hinged on this, we regard recommended prices as the monopoly price level. In turn, a foreseeable low-price window is open before every restoration. During the data period, we observe an additional weekly restoration on Thursdays at noon. We show that an additional day off from competition increases firm performance. As expected, a conventional price search of where to buy reduces firms' profitability. In contrast, consumers who are aware of the cycle and spend effort on when to buy have a positive impact on firms' profitability.If consumers spend effort on when to buy rather than where to buy, price competition might be softened even in the low-price windows.* We are grateful for financial support from the Centre for Applied Research at NHH (project CenCES