E-Finance Lab Team

In recent years nonlinear pricing has attracted a lot of attention from academics and practitioners due to increased possibilities of its application in the growing area of service industries. In case of wireless communication, internet access or TV pay-channels providers usually offer a set of optional tariffs that consists of a usage-independent fixed fee and a marginal price. Analysis of nonlinear pricing is, however, very complex and poses many difficulties because of the interdependency between the consumption level and marginal price. Willingness-to-pay functions account for this interdependency and can be used to predict consumer behavior, analyze the effect of price changes on the market as well as to design of the optimal nonlinear pricing scheme. So far, several studies made proposals on how to use survey data to estimate willingness-to-pay functions (Iyengar et al. 2006, Wolk and Skiera 2006). While those studies laid down the foundations for using survey data to estimate willingness-to-pay functions, they suffer from a limited comparison of data gathering and estimation procedures. In this study we review, enhance and compare various methods that use survey data for the estimation of willingness-to-pay functions. First, we focus on the data gathering procedure (i.e., ranking-based conjoint, choice-based conjoint, and contingent valuation) while in the second step we analyze various estimation procedures (i.e., one-step versus two-step procedures). In an empirical study we measure the perceived task difficulty, the time required for accomplishing the task, face, internal and predictive validity. Based on the results we provide recommendations concerning the most reliable and valid procedure for using survey data to estimate willingness-to-pay functions.