Measuring Depreciation for Japan: Rejoinder to Dekle and Summers

Abstract

Recently, my claim that depreciation reported in the Japanese national accounts is underestimated by a substantial margin has been challenged by Dekle and Summers (NBER Working Paper No. 3690), on the ground that the implied depreciation rate (ratio of depreciation to the capital stock) is implausibly high. I argue in this rejoinder that Japan's high depreciation rate can be attributable to two factors. First, the depreciation rate for owner-occupied housing is much higher in Japan. Second, equipment capital (a component of the denominator in the depreciation rate) in the Japanese national accounts seems underestimated. Therefore, my estimate of the level of depreciation for Japan does not seem exaggerated.