Comment on “Why Has Japan Failed to Escape from Deflation?”

• DOI: http://doi.org/10.1111/aepr.12214
• Author: Kazuo Ueda
• Published date: 01 January 2018
• Date: 01 January 2018

Comment on “Why Has Japan Failed to

Escape from Deflation?”

Kazuo UEDA†

Kyoritsu Women’s University and CARF at the University of Tokyo

JEL codes: E31, E5

Accepted: 31 July 2017

The major findings of Watanabe–Watanabe (W-W) (2018) are as follows: (i) in the

mid to late 1990s, Japan’s consumer prices did not fall by as much as expected given

the deterioration of the economy. One way W-W show this is through the sharp rise

in the share of items with near zero price changes (S(0%), henceforth); (ii) the number

of S(0%) items has since remained high, that is to say the mode of the price change

distribution has been zero; and (iii) for many other countries, mode of the price

change distribution is 2–3%.

W-W offer essentially two explanations for these findings. One of them is the usual

Ball–Mankiw (B-M) type argument that in the presence of fixed costs of price adjust-

ment, prices change more frequently with higher trend inflation. According to this

view, the difference between Japan and other countries boils down to differences in

trend inflation. The other explanation is based on what W-W call a default inflation

rate or an inflation norm, as explained in W-W’s footnote 9. The default inflation rate

or an inflation norm is low (around 0%) in Japan, and much higher (around 2%) in

the rest of the world. As I will now explain, the presence of these two explanations

seems to be a source of some confusion,

The B-M type view explains W-W’s conclusions (i) and (ii) fairly well. According

to this view, there is a range of trend inflation where price nonadjusters numerically

dominate the outcomes. Thus, as inflation started to fall in Japan in the 1990s, initially

price adjusters dominated. But well before inflation reached zero, the nonadjusters

came to dominate. Since then, many firms have stayed in the nonadjuster status even

though there have been some ups and downs in trend inflation. Such a story is consis-

tent with S(0%) rising fairly sharply in the mid to late 1990s, and remaining high in

the Abenomics period despite some increases in actual inflation.

I hasten to add that the W-W might want to emphasize more the special, though

well known, feature of a near zero trend inflation rate. Small fluctuations in the aggre-

gate inflation rate around zero do not force individual price setters to change their

†Correspondence: Kazuo Ueda, Kyoritsu Women’s University and CARF at the University of

Tokyo, 2-2-1 Hitotsubashi, Chiyoda-ku, Tokyo 101-8437, Japan. Email: ueda@e.u-tokyo.ac.jp

44 © 2018 Japan Center for Economic Research

doi: 10.1111/aepr.12214 Asian Economic Policy Review (2018) 13, 44–45