As Walter Heller once said, ceremonial occasions-presidential, memorial, or inaugural addresses-in particular seem to evoke musings on the troubled or even dismal state of our science. Suffice it to mention Wassily W. Leontief who criticized the mathematical superstructure of pure or speculative economic theory for its weak and all too slowly growing empirical foundations; John Kenneth Galbraith who condemned neoclassical and neo-Keynesian economics for eliding power and thus destroying its relation with the real world; Joan Robinson who blamed orthodox neo-neoclassical economics for its disregard of the most pressing problems involving war, poverty, environment and the quality of life of industrialism.
Another samples of criticism are easily found in many other presidential addresses; E. H. Phelps Brown viewed the usefulness of current works in economics as not equal to its distinction because it is built upon assumptions about human behavior that are plucked from the thin air; G. D. N. Worswick observed that, notwithstanding the appearance of formidable progress in techniques of all kinds, the performance of economics seems curiously disappointing and that there could be what looks like specutacular progress in economic theory and yet no progress at all in economic science in the broader sense. Nicholas Kaldor went so far as to say that the powerful attraction of the habits of thought engendered by equilibrium economics has become a major obstacle to the development of economics as a science, so that, in terms of gradually converting an intellectual experiment into a scientific theory, the development of theoretical economics is one of continual degress, not progress.
But today, stemming this current vogue of economic masochism, I should like to serve as defense counsel for equilibrium economics. In my belief, it would be at least as reasonable to judge a discipline by its success as by its failures. Having paid my devoirs to the critics mentioned above, I do not intend any point-by-point appraisal or refutation. Rather my object is to gain a more balanced perspective by focussing on the nature, significance, and contributions of equilibrium economics.
After discussing the economic relevance of equilibrium theory and the useful role of modern mathematical techniques, I review, with the purpose of illustration, certain recent investigations on the two failures of the neoclassical equilibrating mechanism, the first, the non-existence of futures goods markets, and the second, the persistence of underemployed resources. These developments bring money, uncertainty, transaction costs and trading at false prices into the framework of general equilibrium, and their main features are sequence economies, stochastic equilibria, equilibria relative to information, and non-Walrasian underemployment equilibria. Though all of these works are still in their infancy, the whole subject is plainly on a promising track. No doubt the range of phenomena the theory can deal with is being extended, which rebuts a demonstration that the theory at present cannot deal with anything at all.

View full abstract

The purpose of this paper is to analyze the causes of the high rate of return of capital of the major drug companies in Japan.
First, we examine empirically the industry spokesmen's explanation that it is the rewards for their intensive R & D activities, and we conclude that it is hardly convincing.
The majority of the important ethical drugs introduced in Japan after the World War II are discovered abroad and introduced by the major Japanese drug companies. Special attention is paid to this fact and the form of their patent licensing. With a simple economic analysis we reach the conclusion that, when patented drugs are licensed in such a way as is prevailing in this industry, the licensees always receive a part of the monopoly profits.
Using the cross-sectional data of the major drug companies for 10 years, we obtained the following conclusions.
(i) Throughout the period we studied, the amount each company received as the licensee's share of the monopoly profits explains significantly the differences of the rate of return of capital among the major companies.
(ii) Throughout this period, the licensees' share of the monopoly profits has been one of the major factors that explain the high rate of return of the major drug companies.

View full abstract

The purpose of this paper is to examine the optimal behavior of the competitive firm when output price is seen by it as a random variable. The paper distinguishes itself from previous work in that it provides a general model of the firm which permits the presence of any finite number of inputs and a less restrictive treatment of the firm's attitude toward risk than before. In particular, the concept of the "normal" utility-production structure is introduced to show that, as with the certainty case, inputs are likely to be gross complements, which previous work failed to show even with the two-input case.

View full abstract

The conventional method of testing any structural change of a model is to apply the "Chow Test" at a particular time-point specified a priori by the analyst. On the other hand, we are also able to apply the Chow Test to every possible timepoint without introducing any a priori information and to find turning points from the results. In practice, when the data sample is hypothesized to have been generated by more than a single regression model, what is at first necessary for best description of the data sample is to find any boundary time-points, that is turning points, of the data sample. M. Brown's method in "On the Theory and Measurement of Technological Change" and "Piecewise Regression" by McGee V. E. and W. T. Carleton belong to the latter approach.
Both approaches are not exclusive but mutually supplementary. In this paper we attempt to show a modified and extended version of the latter approach, which we call "Stepwise Chow Test", in comparison with Piecewise Regression. (We think Stepwise Chow Test is an alternative method to Piecewise Regression.)
The first step of our procedure is (1) to apply the Chow Test to every time-point of the sample period, (2) to plot all calculated F-values to find their peak if any F-values are significant and (3) to separate the sample period into two subperiods at the time-point of the peak F-value. The second step is to apply the Chow Test to every time-point within each subperiod and to find the peak of significant Fvalues. As a general rule, we repeat this procedure stepwise until we cannot find any significant F-value within every subperiod. Thus we are able to find the number and time of structural turning points.
Judging from our Monte Carlo experiment, it is shown that our procedure is effective in finding structural turning points. And it is also shown that the effectiveness of Stepwise Chow Test is never inferior to that of Piecewise Regression, and moreover, our work load is about as half as theirs.

View full abstract