Prolegomena to the Testing of Utility Theory

Abstract

The arguments of this paper are based on the assumption that, if economics is to advance as a science, economists must increasingly resort to empirical testing of the theories of economics. The reader who rejects this premise will be out of sympathy with the thoughts that follow, because the substance of these thoughts is a series of proposals concerning direct experimental testing of the theory of consumer's preferences. The theory to be subjected to the proposed testing is the ordinal theory of utility. This theory assumes that the consumer is capable of ranking various collections or budgets of goods according to their desirability to him. There is no assumption that the consumer is able to convey, either consciously or unconsciously, to an observer the number of units of utility that he receives from any budget of goods, i.e., there is no implication that utility is measurable in the neo-classical sense. A further assumption is made to the effect that the consumer will choose the most desirable or the highest ranking collection of goods that his income will allow him to choose.l Before testing is undertaken, it is desirable that the would-be experimeters think out, insofar as it can be done, the problems they are likely to encounter. It is toward this goal of pre-experimental planning that this paper is oriented. First, it may be observed that in some cases the decision as to the desirability of various goods will be a group decision. In the case of unattached individuals this, of course, is not true, but in the case of families the decisions may be group decisions and special problems may arise from this fact. Another problem that should be considered is the possibility that individuals behave differently toward different types or classes of goods. This raises certain interesting questions as to the structure of the preference function itself. The problems that might arise from the family group decision process and from the different regard that individuals have for different types of goods will be considered below. In addition, certain means of testing that have been suggested by various authors will be reviewed. Finally, a different type of attack on the problem of testing the theory will be suggested.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 1 citations exist upstream.