When monetary balances are introduced into the consumer's utility function and budget constraint, the resulting demand functions for money and for the consumption goods in general do not exhibit the standard Slutsky-Hicks properties. Many researchers have studied this issue and have attempted to remedy the outcome; for example, see M. Morishima (1952), C. Lloyd (1964, 1971), J. Hadar (1965), and Dusansky-Kalman (1972, 1973). Their paramount goal has been to provide a neoclassical integration of money demand and consumer choice theory that will yield the traditional demand predictions that are the hallmark of neoclassical microeconomic theory. The most recent work on this subject is by Samuelson-Sato (1984). They convincingly demonstrate that with a weakly separable utility function it is possible to reinstate symmetry and negative semidefiniteness and to generate testable predictions. In this paper, however, I show that these predictions are not those of traditional theory but a modification of them, having alternative implications (for example, the assumption of a non-inferior good is no longer sufficient to ensure a downward-sloping own demand curve). The problem is that the weak separability assumption is not sufficient to recover the traditional properties. I present an alternative class of utility functions for which the commodity demand functions do exhibit the usual Slutsky-Hicks properties. This result also holds for money demand. I. Implications of the Sanuelson-Sato Results

One of the most startling tax policy prescriptions to appear in recent years is the tax of Yew-Kwang Ng (1987). Diamond goods are goods, such as diamonds, precious stones and metals, expensive fur coats, and luxurious cars, which are valued for their values more than for their intrinsic consumption effects. Consumers of diamond goods are viewed as deriving utility from exhibiting their wealth and/or by using these goods as store of value and/or giving them as gifts of value. In these cases it is the value of the good that matters most. After analyzing the taxation of diamond goods, Ng formally asserts three propositions and corollary. His principal findings are that a change in the price of diamond good leaves its value and the amounts of all other goods consumed, and hence the utility level of the consumer unaffected (Proposition 1) and that the demand curve for diamond good is rectangular hyberbola with unit elasticity throughout the whole range where it remains pure diamond (Corollary 1). These findings form the basis for his conclusion that a pure diamond good has an infinite tax in an optimal tax system (Proposition 3). The taxation of diamond goods thus represents burdenless tax policy panacea. The purpose of this paper is to reconsider the microeconomic analysis of the taxation of diamond goods. We show that the striking results of Ng depend on the particular way in which the diamond good is specified. We demonstrate that an alternative-and eminently reasonable -specification yields drastically different results, in that Ng's principal findings do not hold. Ng considers representative consumer choosing among single pure diamond good and n -1 ordinary consumption goods. The utility function is assumed to be well behaved and is written