Abstract
Japanese data from 1971 and 1995 are used to estimate the marginal rate of substitution between private consumption and government consumption, and to empirically analyse the validity of the Ricardian equivalence theorem. The marginal rate of substitution between government expenditure and private consumption is found to be higher in Japan than in studies using US data. In other words, these two types of consumption are closer substitutes in Japan than in the US. The data negate the Ricardian equivalence theorem. However, it should be noted that the perceived wealth effect is small and has a value between 0.0087 and 0.019.