Are Consumers Strategic? Structural Estimation from the Air-Travel Industry
University of Michigan, Stephen M. Ross School of Business
Nelson F. Granados
July 17, 2013
INSEAD Working Paper No. 2013/85//TOM/ACGRE
Many consumers have learned to delay purchases, anticipating that prices might decrease. Such strategic or forward-looking behavior has attracted increasing attention from various disciplines, including operations management, information systems, marketing, and economics. However, there is currently limited empirical evidence of the extent to which this strategic purchasing actually takes place. Combining two unique data sources from the air-travel industry (posted fare data and booking data), we use a structural model to estimate the fraction of strategic consumers in the population, assuming di erent levels of sophistication in consumers' perception of future prices: perfect foresight and weak- and strong-form rational expectations. We nd that 5.2% to 19.2% of the population is strategic across markets, measured by the rst and third quartiles. Our inter-market analysis indicates that shorter trips with alternate modes of transportation have a higher proportion of strategic consumers. Using a non-parametric approach, we further nd that most strategic consumers arrive either at the beginning of the booking horizon or close to departure. Finally, our counterfactual analysis shows that, contrary to conventional wisdom, the presence of strategic consumers does not necessarily hurt revenues. Rather, the impact varies by market. Commitment to a non-decreasing pricing strategy is more likely to bene t business markets than leisure markets and it could even hurt leisure markets. Inter-market analysis shows that city-pairs with lower internet penetration, higher average price and shorter distances tend to bene t more from such commitment as well.
Number of Pages in PDF File: 42
Date posted: October 5, 2011 ; Last revised: July 17, 2013
© 2015 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo5 in 0.297 seconds