Review of Industrial Organization
12: 413–416, 1997.
1997 Kluwer Academic Publishers. Printed in the Netherlands.
Ramsey Pricing and Regulator’s Social Welfare
Weights: An Empirical Application
Department of Economics, Universidade Federal do Rio de Janeiro, Av. Pasteur 250, 22290, Rio de
Janeiro-RJ, Brazil, and Queen’s College, University of Oxford, Oxford, OX1 4AW, U.K.
Abstract. The paper aims at uncovering the implicit regulator’s social welfare weights within a
generic Ramsey pricing structure allowing for distinct weights. The methodology is applied to U.S.
water utilities for residential and non-residential water services, implying that the weights were
virtually identical and thus no cross-subsidies occurred across those goods.
Key words: Ramsey pricing, welfare weights, water industry.
The study of second-best pricing in regulated industries has brought a large, pre-
dominantly theoretical, body of literature (see e.g., Baumol and Bradford, 1970).
The empirical counterpart of such effort has evolved at a slower pace however,
what provides a motivation for additional research.
Rate setting in regulated industries, constitutes one of the good examples,where
academic research has had some impact in terms of policy making, specially in
terms of peak-load pricing (see Faulhaberand Baumol, 1988). Examples of empiri-
cal studies related to Ramsey pricing appear in Scott (1986) and Kim (1995). In the
present paper, we intend to explore another aspect, namely to uncover the implicit
regulator’s social welfare weights that can be obtained within a generic Ramsey
pricing structure, taking as reference detailed information on U.S. water utilities.
The paper is organized as follows: the second section describes the methodology;
the third section discusses the data and presents the empirical results. The fourth
II. Regulator’s Social Welfare Weights
The most generic Ramsey pricing scheme wouldinvolve potentially distinct weights
according to the good or consumer group; it should be noted however, that the two
weighting alternatives are typically not independent, as the “merit” goods may
be deﬁned conditional to their weight in the expenditure of a particular consumer