This dissertation encompasses three papers that empirically examine ongoing competitive interactions in retail gas markets. Each paper takes a different empirical approach to examine how organizational and cognitive factors influence pricing decisions in this market.
The first essay examines how multi-unit franchisee ownership and corporate ownership influences competitive behavior reflected in pricing. I use a panel dataset of pricing decisions of multi-unit franchisees and company-owned gas stations to compare two competing mechanisms by which ownership form influences pricing, double marginalization and strategic delegation. I find that franchisees charge higher average prices, supporting the greater influence of double marginalization on price. Contrary to agency theoretic predictions, firm size and geographic dispersion have a negative influence on the price of multi-unit franchisee stations.
The second essay explains how spatial distance and competitor similarity influence firm identification of a relevant competitors. In contrast to prior studies that have used surveys to identify competitors managers saw as most important, I identify a firm’s competitors by examining the competitive actions and responses of units using data that isolates the timing of price changes in the Los Angeles retail gas market. Consistent with predictions, I find that retail gas stations monitor a small number of rival stations. The results demonstrate that distance to a rival and similarity between competitors on price and the number of pumps at a station interact to influence the weights assigned to competitors. The findings suggest that managers categorize competitors based on a smaller number of key dimensions than previously theorized.
The third essay takes a behavioral approach to examining competitive market factors that lead to systematic pricing errors using non-experimental data. While management researchers have studied the causes of suboptimal pricing decisions, previous research has emphasized experimental or aggregate corporate data rather than pricing and performance data from actual competitive interactions. I utilize a hand-collected, longitudinal dataset of prices and performance outcomes for 26 retail gas stations to determine a daily, station specific profit-maximizing price. These prices are then compared to the actual prices charged to assess the accuracy of station pricing decisions. I find that the number of competitors in a market have a positive influence on the accuracy of pricing decisions at low numbers of competitors but a negative influence at high numbers of competitors. Stations with a visible competitor that compete head-to-head set more accurate prices than stations without a competitor visible competitor.