Debate on Deflation in Japan #4

Comment on Richard Katz

Tsutomu WATANABE (Hitotsubashi University, Japan)

Richard Katz's argument is clear and persuasive. I basically support his argument. The following is my short comment based on recent empirical research related with this issue.

Deflationary spiral
Mr. Katz points out that (1) the US economy experienced "deflationary spiral" in the 1930s, (2) but the current situation in Japan is different from it. I agree that the current situation in Japan is not characterized by "deflationary spiral." I go further than that: I am very skeptical about the existence of "deflationary spiral." I have never heard the definition (I should say, the careful definition) of "deflationary spiral" in the policy debate in Japan. As far as I know, there is no academic papers that successfully reproduce the phenomenon like "deflationary spiral" in economic models. This is not a coincidence, but a reflection that economists base their argument on the concept of "equilibrium," which is probably inconsistent with the idea of "spiral." I believe that some forces (for example, wealth effects) will start to work to prevent the Japanese economy from falling into the spiral, and take it back to the equilibrium.

Phillips curve
Mr. Katz presents a figure to show a positive relationship between the inflation rate and the demand-supply gap. This is nothing but the Phillips curve. The Phillips curve is still alive in Japan! The fact that the Phillips curve is still alive during the periods of deflation is very interesting. Let me make three comments on his finding. First, according to my empirical research with Kenji Nishizaki ("Output-inflation tradeoff at near-zero inflation rates" Journal of the Japanese and International Economies, 2000), the slope of the Phillips curve becomes flatter as the rate of inflation goes down. This means that a unit of policy stimulus (monetary or fiscal stimulus) creates a smaller increase in the inflation rate when the level of inflation is low. In other words, raising the inflation rate is more costly than before. Given this fact, I am against his idea of stimulating aggregate demand to stop deflation. Second, the existence of the Phillips curve does not rule out the possibility of supply shocks. In fact, our research mentioned above finds that supply shocks played an important role in lowering the inflation rate in the 1990s. Unfortunately, the sample period of our research ends in the 1990s, so our research does not say much about the role of supply shocks in deflation ("good deflation" in Mr. Katz's word) in the last two or three years. But, based on a casual "eyeball econometrics," I have an impression that supply shocks continue to play an important role. In this sense, I am against the view that weak demand is a dominant cause of deflation. Third, the Phillips curve looks stable in the 1990s but not so before that. For example, we did not observe high inflation (more precisely, CPI inflation) during the bubble period in the 1980s. As usual, we have to be very careful when we make an argument based on a statistical relationship.

Fisher equation
When Mr. Katz writes "The BOJ's power to create inflation is most impotent just when the "inflationists" say it is most needed," it gives me an impression that the combination of deflation and zero interest rates is just an unlucky coincidence. But this is not true: this combination is an inevitable one. The Fisher equation tells us that nominal interest rates tend to be low when the inflation rate is low. In this sense, the phenomenon of zero interest rates is a consequence of low inflation (or deflation). Aiming at a very low level of inflation, as the BOJ did in the 1990s, increases the probability of falling into a liquidity trap. This is an important lesson about how to avoid a liquidity trap, although this lesson says almost nothing about how to escape from the trap the Japanese economy has already fallen into.