Spatial competition in automobile retailing

ABSTRACT

Despite a rich literature on the determinants of automobile pricing, there is relatively little formal evidence regarding the effects of spatial intrabrand competition in auto retailing. Using an extremely large sample of tax registrations in Texas (USA) for 19 of the most popular sedans, we examine how the closeness of a dealer selling the same make and model of vehicle affected the final price. We find very strong evidence that the proximity of a same-brand dealer reduces prices significantly, though we find smaller effects for the lowest priced vehicles, a result consistent with the view that such vehicles are sold essentially at cost in order to satisfy Corporate Average Fuel Economy (CAFE) standards and related fleet characteristic regulations.

KEYWORDS:

• Automobile retailing
• price competition
• spatial competition
• car retailing

Acknowledgment

We are grateful to the National Automobile Dealers Association for making the IHS data available for our research

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Zettelmeyer, Scott Morton, and Silva-Risso (2006) include distance in an analysis of prices but do not report any results on the variable.

² Until now, available evidence was limited to the San Diego market.

³ Buyers are, by assumption, always located at the centroids of census areas, so that all buyers in the same area who purchase from the same dealer are assumed to have travelled the same distance.

⁴ We limit our attention to Texas because it is the state registration data is complete in the IHS Markit data, whereas data for others states is incomplete due to voluntarily submission.

⁵ Walden (2005) does find upgrades like airbags affect prices, but our attention to sub-models of cars and the near ubiquity of airbags in 2012 models suggests that these upgrades do not impact our analysis.

⁶ We also exclude a single observation of a Toyota Camry priced at about $200,000, which we presume is a coding error.

⁷ Transaction prices are for the first six months of 2012 and distance to nearest dealer ranges from 5 to 150 miles. Lowess calculated using a bandwidth of 1.0.

⁸ The coefficient on the six-cylinder engine for the Accord EX is 0.118 with a robust t-statistic of 57.64, implying a price differential of $3,181 (an increase of 12.6%). While we could have excluded the six-cylinder model, we were interested in the magnitude of the effect.

⁹ Full results are available from the authors upon request.

¹⁰ Detailed results are available upon request.