Buyers in many markets face multiple-discrete choices: they can purchase multiple-units as well as multiple-brands at the same time. This paper presents a multiple-discrete choice model for the analysis of differentiated products demand. Users maximize profits by choosing the number of units of each brand they purchase. The author estimates the model using micro-level data on the demand for personal computers. He uses the estimated demand structure to assess the welfare gains from computerization and technological innovation in peripherals. The estimated return on investment in personal computers is 92 percent. Moreover, a 10 percent increase in the performance-to-price ratio of microprocessors leads to a 22 percent gain in the estimated user surplus.

We present a dynamic model of adverse selection to examine the interactions between new and used goods markets. We find that the used market never shuts down, the volume of trade can be large, and distortions are lower than previously thought. New cars prices can be higher under adverse selection than in its absence. An extension to several brands that differ in reliability leads to testable predictions of the effects of adverse selection. Unreliable brands have steeper price declines and lower volumes of trade. We contrast these predictions with those of a model where brands physically depreciate at different rates.