Deriving nonlinear pricing schemes using modified least deviations spline regression

Nonlinear pricing is a form of price differentiation in which the average price per unit is a nonlinear decreasing function of demand. It is widely used in many industries such as telecommunications or utilities. Rationales for using nonlinear pricing are manifold, ranging from taking account of decreasing marginal utilities of customers to building up switching costs. As with other forms of price differentiation, the potential for capturing consumer surplus and increasing profits is considerable. Nonlinear pricing schemes come in various forms, such as discounts, buy-one-get-one, power shopping, two-part tariffs, and block-tariffs. All these forms are well known. However, little is said about how exactly to systematically derive an optimal nonlinear pricing scheme. In this paper, I propose a new systematic empirical approach using a modified least deviations Spline regression-estimator.