Understanding the price dynamics is always at the heart of studies on monetary policy. If prices are flexible, there is no need for monetary policy as a stabilization tool. If not, central banks should offer some nominal anchor so that welfare costs stemming from relative price dispersion should be reduced. Watanabe and Watanabe (2018) conduct careful analyses using micro data. Two main conclusions are: a simple menu cost model can explain the price developments in Japan well in that “there is a negative correlation between trend inflation and the share of items whose rate of price change is near zero,”1 and the “inflation norm” is too low in Japan. Watanabe and Watanabe provide comprehensive empirical facts to support these two conclusions. It also merits attention that the observed empirical regularities are replicated with simulations using a simple menu cost model. Structural interpretations of the empirical findings are offered albeit indirectly via model simulation. Not only do Watanabe and Watanabe offer convincing facts in understanding the Japanese economy from the view point of positive analysis, but also the findings in their paper have an immediate policy implication: It is not easy for Japan to attain an inflation target of 2%. Papers with important facts and significant policy implications are always good and highly appreciated. Watanabe and Watanabe's paper is one of these good papers. Thus, I do not have any particular comments to oppose their main conclusions. Let me comment on what we do not know from the data, and the policy implication. We do not know the markups in individual goods. In their menu cost model, nominal rigidities are favored as the way to explain price developments in Japan. This conclusion is mainly supported by the analyses in Sections 2.2 and 2.3 about the flattening of the reduced-form Phillips curve. However, we still cannot conclude that this less clearer relationship between inflation and unemployment rates is as a result of higher nominal rigidities possibly due to menu costs, since we cannot observe marginal costs. The relationship between inflation and unemployment rates can be just a reduced-form one. In the language of the New Keynesian model, unemployment or the output gap must be structurally related to markups. Otherwise, a causal argument is not possible. I am not fully sure whether such a relationship is stemming from menu costs or higher real rigidities around zero inflation rates. Real rigidities arise when marginal costs are sticky or marginal revenues are flexible. The former can be caused by habit formation in consumption, endogenous capacity utilization or investment growth adjustment costs as included in the canonical dynamic stochastic general equilibrium (DSGE) model a la Christiano et al. (2005), while the latter can emerge with a kinked demand curve as examined in the first-generation New Keynesian Model a la Kimball (1995). In order to identify the causes of the less clearer relationship between inflation and unemployment rates, it would be ideal if we can use the marginal cost data as examined in Eichenbaum et al. (2011). The inability to observe markups means that a careful interpretation of the analyses in Watanabe and Watanabe's section 4 is required. If the relationship between unemployment and markup is not stable nor structural, increasing real rigidities around zero inflation rates can explain the fact that “price stickiness as measured by the share of unchanged items increases as the inflation rate declines from a positive value to zero.” More stickiness around lower inflation rates should imply higher welfare costs stemming from relative price dispersions. On the other hand, menu costs are small near zero inflation rates. Besides the problems due to the zero lower bound on nominal interest rates, should the central bank really care about inflation developments near zero inflation rates? This point is related to the conclusion made by Nakamura et al. (2016) that “the main costs of inflation in the New Keynesian model are completely elusive in the data. This implies that the strong conclusions about optimality of low inflation rates reached by researchers using models of this kind need to be reassessed.”