抄録

市場競争度を計る指標として用いられているH統計量について再考する.H統計量は,その市場が独占市場であることを否定する指標としては有用であるが,この数値の大小と競争度の間にはシステマチックな関係がないことを明らかにし,この指標を用いた実証分析は再検討の必要があることを明らかにする.

abstract

This paper reexamines the implications of H statistics, which are widely used as measures of competitiveness. Some researchers used H as a wrong way. They insisted that H<0 if and only if the market is monopoly and that H = 1 if and only if the market is perfectly competitive. I show that both are wrong. H can be negative or can be positive but strictly smaller one even in perfectly competitive markets. I also show simple examples illustrating that H statistics are related to the competitiveness.