Constructing willingness to pay for product attributes in the probability space.

This paper outlines several approaches used to construct measures of willingness to pay (WTP) for product attributes. We argue that measures based on consumer nomination require an unrealistic task and/or inconsistent with most revealed preference data. Compensating Variation, based on MNL model output, uses estimates of parameters constructed from a more familiar task, product choice. Compensating Variation (CV), however, is dependent on the base market from which comparisons are made. It also presents anomalies in relation to expected values of consumer WTP for changes in price.In this paper, a WTP measure is constructed by equating changes in probability in product attributes with changes in price. As a result, WTP for each product attribute can be constructed from choice model estimates, but address the independence of base problem and allow for many alternatives and attributes. We demonstrate its practicality in expressing changes in products on various features in monetary terms. In turn, it provides an easily communicated and managerially informative measure for determining price, especially in markets where consumer evaluation of features is not well understood (e.g., new products).