Using Panel Data to Easily Estimate Hedonic Demand Functions

The hedonics literature has often asserted that if one were able to observe the same individual make multiple purchase decisions, one could recover rich estimates of preference heterogeneity for a given amenity. In particular, in the face of a changing price schedule, observing each individual twice is sufficient to recover a linear demand function separately for each individual, with no additional restrictions. Constructing a rich panel data set of buyers, we recover the full distribution of demand functions for clean air in the Bay Area of California. First, we find that estimating the full demand function, rather than simply recovering a local estimate of marginal willingness to pay, is important. Second, we find evidence of considerable heterogeneity, which is important from a policy perspective; our data-driven estimates of the welfare effects associated with a nonmarginal change in air quality differ substantially from those recovered using the existing approaches to welfare estimation.