Airfare, competition, and spatial structure: New evidence in the United States airline deregulation

The dynamics of airline deregulation has caused dramatic changes in airfare and competition structure. This dissertation studies the variation of airfare, its relationship with the prevailing hub-and-spoke delivery network, and the complex competition structure in the deregulation environment. Results from this research show that airfare on average has decreased considerably since deregulation, but at the same time has also shown increasing variation across different market sectors. The spatial distribution of airfare has not been even. Pricing dynamics have resulted in geographic patterns of lower airfare for cities in the west and southwestern United States while higher airfare was found in the South, New England, and Midwest. The results from regression analysis show that: (1) route level airfare is significantly influenced by the hub status of the origin and destination airports; (2) the level of concentration and dominance at airport and route markets have important but variable impact on airfare; (3) the existence of low fare carriers has considerable impact on pricing in the market. Competition is effective in restraining the dominant carrier's market power; (4) there exists significant network autocorrelation among airfare at the level of route markets. Using a spatial autoregressive model with spatial dependence structure explicitly embedded can partially solve this problem and improve empirical results previously obtained from linear regression analysis.; The major contributions of this dissertation are: (1) this is the first empirical work using detailed disaggregated passenger data at the national level. It updates and in some way corrects the previous research with new results; (2) from a geographic perspective, this research focuses on analyzing the spatial structure of airfare and competition. Previously attention has rarely been paid to spatial influence in existing airfare analysis literature; (3) this research brings the concept of spatial network autocorrelation into air transportation research, and for the first time, applies geographic information systems and spatial regression techniques in airfare analysis.