Cannibalization vs. competition: An empirical study of the impact of product durability on automobile demand

Abstract Many durable product categories have well developed and organized secondary markets that make it easy for consumers to purchase used versions in lieu of new ones. Manufacturers in these categories therefore face a dilemma in deciding on how much durability to build into their new products. High durability levels increase the risk that future new product sales are lost to used versions (the cannibalization effect ), but can also help the firm take sales from new and used versions produced by its rivals (the competition effect ). However, there has been no empirical research to determine the relative sizes of these two effects. Our research seeks to fill this gap in the literature via an empirical study of the demand for new and used vehicles in the US automobile market. We analyze two datasets; the first from the mid-sized sedan segment in the Indianapolis DMA during 2004–2006 and the second from the Entry Sports Utility segment in the Los Angeles DMA during the period 2003–2005. We propose and implement a durability metric that is based on the over-time trajectory of used car prices and estimate a structural model of differentiated product demand with heterogeneous consumers who choose among new and used vehicles and consider durability in their choice decisions. Both datasets reveal that durability significantly impacts the demand for new and used vehicles with mean durability elasticities estimated to be about 1.7. Both datasets also show that the competition effect (85%) is significantly greater the cannibalization effect (15%). While the prior theoretical literature generally recommends that firms avoid cannibalization by reducing product durability, our findings imply that manufacturers who increase the durability of their products may see net sales gains as a result.