医療費の都道府県格差分析

抄録

Using the OLS method, we investigate the difference of medical treatment expenditure between 47 prefectures, the largest administrative divisions of Japan. Paying attention to peculiarities of a medical treatment service, we examine the factors which determine medical treatment expenses both on demand and supply sides. Then, with a brief summary of institutional changes at that period, characteristics of changing figures of national medical treatment expenditure and medical treatment expenses throughout these years are pointed out.
The main part of our paper is a quantitative analysis of difference of medical treatment expenses between 47 prefectures. Using the OLS method, the prefectural difference of several medical treatment expenses are explained by demand, supply and accessibility factors. In order to avoid the multi co-linearity problem between independent variables, variables with high correlation coefficients are put into several groups, and no more than one variable within each group is used in each equation. The data used here is national medical treatment expenditure and medical treatment expenses of elderly persons at 47 prefectures in Fys 1990, 1993 and 1996.
Findings are as follows: 1. As long as the difference of medical treatment expenses by prefectures is influenced both by demand and supply factors, we cannot deny the ‘supply induced demand’ in medical treatment service. 2. PER CAPITA SICKBEDS is the best explanation variable for the most medical treatment expenditure. 3. Medical treatment expenses for elderly persons are explained especially by AVERAGE HOUSEHOLD PERSONNEL, which represents the difficulty/easiness of nursing care by family members, and by DID POPULATION RATIO, which represents the accessibility of medical treatment services. But an explanation power of the former variable is dropped in FY1996 reflecting the increase of nursing homes and the development of the visiting nursing system. 4. Each local government sets up the medical treatment plan to control expenditure. Although the most crucial index of it is the required sickbeds, an explanation power of EXCESS SICKBEDS RATIO is inferior to that of PER CAPITA SICKBEDS. 5. According to the analysis given here, the prefectural difference of medical treatment expenditure is influenced both by demand and supply factors. And since the medical treatment expenditure for outpatients has positive income elasticity, demand side constraint measures are needed to control the rapid increase of medical treatment expenditure.