Pricing of Central Bank Payment Services

The economic history of technologies is largely a history of low-fixed-cost, high-marginal-cost technologies being replaced by newly invented, decreasing-average-cost (approximately, highfixed-cost and low-marginal-cost) technologies that come to be cost minimizing as demand for the product or service produced by the technology increases. The history of payment technology in the past century, and particularly in the last two decades, fits this description to a T. The adoption of decreasing-average-cost technologies in other industries (notably transportation and telecommunications) prompted the development of a body of economics regarding how use of these technologies ought to be priced. The highlights of this theory of non-linear pricing are that often, when average cost is decreasing, (1) multi-part pricing must be used to support a total-surplus-maximizing allocation and (2) price discrimination can maximize total surplus, if cost must not exceed revenue and single-part pricing must be adopted. My impression is that, although most central banks express concern about current or impending challenges in financing the operation of their payment systems, they tend to be more restrained than the operators of proprietary payment systems have been in taking full advantage of the opportunities that nonlinear pricing afford for generating revenue. My goal in this paper is to present—mostly in informal terms—the main ideas of the pricing theory relevant to payment systems, to discuss the relationship of those ideas to some specific issues