地域間物価指数の理論と都道府県別物価指数の構築

阿部修人 (一橋大学経済研究所), 稲倉典子 (四国大学経営情報学部)

Regional Price Index Number: Theory and its Application to the Prefectural Price Index in Japan

Naohito Abe (Institute of Economic Research, Hitotsubashi University), Noriko Inakura (Faculty of Management and Information Science, Shikoku University)

要旨Abstract

　公開されている公式統計に基づき，日本の都道府県別物価指数を試算した．主要結果は下記の通りである．（1）帰属家賃を含む都道府県間物価指数は，全ての指数において大きな違いを示しており，物価の高い県と低い県の間には20％ポイント以上の差がある．（2）帰属家賃を除く場合，総務省による価格差指数同様，都道府県間の物価の違いは小さく，その差は10％ポイント以下程度となる．（3）食料品に関しては，物価指数算式間の差は少ない．（4）非食料品の指数算式間の差は大きく，特にJevons指数およびCUPI（CES Unified Price Index）と他の差が大きい．また，比較的同様の挙動を示すGeary-Khamis やFisher 指数，Gini やAbe-Rao 指数等の間でも1％から3％ポイント程度という無視できない違いが生じている．これらは，都道府県間の物価の違いを見る際，帰属家賃の影響が極めて大きいこと，また，それほどではないにしろ，指数算式間の違いも存在しており，分析・利用目的に応じた指数算式の選択が重要であることを示している．

　We construct the regional price indexes using the retail price survey and the family income and expenditure survey in Japan. We compare various index numbers, giving special attention to recent developments in index number theory. The main findings are as follows: 1) The prefectural price index, including imputed rents, exhibits heterogeneity across prefectures to a great extent for all the index number formulas. The largest difference across prefectures is greater than 20% points. 2) Excluding the imputed rents, we find the tendency for heterogeneity to be small, that is, less than 10% points at the most, which is consistent with the official regional price index number by the Japanese Statistical Bureau. 3) Prefecture level index numbers for foods are generally remarkably similar. 4) For non-food items, selecting the index number formula becomes important. In particular, the Jevons Index and the Constant Elasticity of Substitution (CES) Unified Price Index (CUPI) exhibit significant differences form other index number formulas, such as the Geary-Khamis, Fisher, and Abe-Rao indexes. The estimation results show that the effects of the imputed rents are substantial, whereas the choice of the index number formula is also important, suggesting the decisive role played by the selection of the “appropriate” index number formula in applied research.

Full Text

書誌情報Bibliographic information

Vol. 73, No. 2, 2022 , pp. 160-180
HERMES-IR（一橋大学機関リポジトリ）: https://hdl.handle.net/10086/74156
JEL Classification Codes: C13, C83, E01, E31