分配の公正に関する経済理論:展望

• Published in: 季刊 理論経済学 Vol. 25; no. 2; pp. 1 - 14
• Main Author: 村上, 雅子
• Format: Journal Article
• Language: Japanese
• Published: 日本経済学会 1974
• ISSN: 0557-109X; 2185-4408
• DOI: 10.11398/economics1950.25.2_1

This paper attempts to survey and appraise various contributions to the theory of optimal income distribution focused attention being on the works useful to determine the optimal degree of inequality between income classes. First we shall examine the theory of justice presented by J. Rawls. If he considered the division of constant amount of income, the consequence of his "maxi-min" principle might be equal distribution. He didn't think so, because he assumed the existence of "chain-connection" effect between the rise of the least advantaged person's position and the improvement of the other citizen's. If this reciprocal postive chain holds, his "maxi-min" principle being consistent with efficiency, becomes more generally acceptable one. So, an appropriate formulation of this chain effect must be done if we want to determine optimal inequality between income classes based on Rawlsian principle. We shall next discuss how this requirement is met in part with the efforts of "optimal income tax" theorists; J. Mirrlees, A. Atkinson, E. Sheshinski and Y. Itsumi. In their general equilibrium models, all tax revenue is spended on minimum guranteed income given to the people with no earnings. So, "maxi-min" principle means tax-revenue maximization. On the one hand, the rise of tax rate may increase consumption of least advantaged, which in equilibrium of total production, can increase all earnings. But on the other hand, the rise of tax rate may suppress labour supply which shall reduce tax revenue. Therefore optimal tax rate may depend on the elasticity of labour supply, the shape of individual utility function and the type of ability dispersion. Finally we shall discuss the measurement of marginal utility of money. K. Tsujimura measured it by applying the theory of consumer's equilibrium. He assumed a linear marginal utility function of each commodity in simplified form, (1) ui=bi+aiiqti Consumer's equilibrium condition, (2) ui/pi=λt, i-1,…, n. He estimated the parameters in marginal utility function and ,It with the least-square method on formula (3). We think this least-square method is questionable, because he treated ,λt as a parameter, the least-square method must be applied to formula (4). (3) λt=bi1/pti+aiiqti/pti (4)qti/pti=λt/aii+bi/aii(-1/pti) Then, we must devise the method to separate a • from the first and the second parameters. If we can find the device this modified Tsujimura method will be a plausible one to measure marginal utility of money, on which we shall we shall give some welfare meaning as an indicator of rdlative necessity of money between income classes.