Abstract
This study examines how used-vehicle markets responded to the automobile hydrocarbon emissions by linking used-vehicle price to the large scale emission test data that contain 74 vehicle models manufactured over 18 years. An additive semiparametric hedonic model is estimated to analyse the relationship between vehicle price and hydrocarbon emissions. The estimation procedure is novel and involves a local polynomial estimator nested in a backfitting algorithm with the bandwidths chosen by a data-driven plug-in method. The results indicate that hydrocarbon emissions have a significant negative impact on vehicle price, but the negative association is evident only at low emission levels. The price discount appears to be unrelated to the increased costs from recent emission regulations that mainly target high-polluting vehicles.
Acknowledgements
The authors thank Steve Polasky and Vic Tremblay for useful comments. The authors also thank Jerry Copper of Oregon Department of Environmental Quality for providing automobile emission data from the Vehicle Inspection Program and Li Liu for helpful research assistance.