Automobile Replacement: A DynamicStructural Approach

This paper specifies and estimates a structural dynamic model of consumer demand for newand used durable goods. Its primary contribution is to provide an explicit estimationprocedure for transaction costs, which are crucial to capturing the dynamic nature ofconsumer decisions. In particular, transaction costs play a key role in determining consumerreplacement behavior in both primary and secondary markets for durable goods. The uniquedata set used in this paper has been collected by the Italian Motor Registry and covers theperiod from 1994 to 2004. It includes information about sales dates for individual cars overtime as well as the initial stock of cars in the sample period. Identification of transactioncosts is achieved from the variation in the share of consumers choosing to hold a given cartype each period, and from the share of consumers choosing to purchase the same car typethat period. Specifically, I estimate a random coefficients discrete choice model thatincorporates a dynamic optimal stopping problem in the spirit of Rust (1987). I apply thismodel to evaluate the impact of scrappage subsidies on the Italian automobile market in 1997and 1998.