- Tamim Bayoumi, Guy Meredith, and Bijan Aghevli
- Published Date:
- June 1998
Masahiko Takeda opened his remarks by noting two underlying themes in the papers presented by the staff. The first theme was that, in the period after the bursting of the bubble until 1995, monetary policy had been too slow to support the economy. The second theme was that one of the reasons for this failure was that the Bank of Japan had focused on the wrong indicators and that, in particular, the use of a monetary conditions index (MCI), which weighted together the short-term interest rate and the exchange rate, would have provided superior signals about the appropriate stance of monetary policy, as would have an inflation target with a lower floor for inflation. On the first point, Takeda noted that one could plausibly argue that the Bank of Japan had remained too contractionary over much of the 1990s, although this assessment involved a good deal of hindsight, and had been much less clear at the time. His main concern with the staff’s analysis, however, was with the use of the MCI as a guide to policy. The MCI was a mixture of a policy variable under the direct control of the Bank of Japan (short-term interest rates as defined by the official discount rate) and another variable whose price was determined in asset markets (the exchange rate), but which was significantly affected by the behavior of short-term interest rates. Sorting out the truly exogenous impact of the exchange rate from the induced impact from monetary policy was a difficult econometric exercise. Furthermore, it remained unclear whether the MCI incorporated all of the information contained in its individual components, let alone other information relevant for the conduct of monetary policy. For this reason, he felt that the Bank of Japan was likely to retain its more eclectic approach on the conduct of monetary policy. Takeda was also skeptical about the value of an inflation target. He noted that inflation had remained low over the bubble period, and hence would not have provided a useful signal for monetary policy at that time. He also felt that there could be definite advantages to not being too open about the conduct of monetary policy, at least in an economy such as the Japanese, in which the central bank had considerable anti-inflation credibility.
Joseph Gagnon emphasized that Japanese monetary policy had been extremely successful in providing low and stable inflation. At the same time, he felt that the staff’s assessment that there had been too little easing of policy in the early 1990s, and that, in particular, too little emphasis had been put on movements in the yen, was accurate—in contrast to the 1980s, when too much emphasis may well have been put on the yen. He was sympathetic to the idea of setting an inflation target range (with a lower bound greater than zero), as it gave the market a criterion by which to judge the performance of the central bank, but had more mixed feelings about the use of an MCI. On the one hand, Gagnon acknowledged that weighting alternative monetary channels together was a useful exercise, which could provide insights as to the net effects of monetary policy actions; but on the other hand, the appropriate monetary response to an MCI shock depends on the behavior of the individual components of the MCI, thereby calling into question the usefulness of any composite MCI measure. He also mentioned that a potentially important factor in assessing recent Japanese monetary policy was the health of the banking system. More generally, he felt that the presentation of staff views could have been improved by providing a model of the operations of monetary policy. Finally, in regard to the research paper on monetary transmission, Gagnon found the results less surprising than did the authors, particularly the evidence of structural change. The difference in results between nominal and real interest rates must reflect the transition from high to low inflation in Japan. In the inflation equations, the dependent variable should be the change in the inflation rate, since a steady state is consistent with any constant level of inflation. This change should eliminate the perverse result that higher interest rates cause more inflation.
Fumio Hayashi opened the general discussion by asking why, in the research paper, only bivariate VARs had been used, and why no terms in the levels of the underlying variables had been included. William Alexander noted that the MCI was only of use if it was assumed that most exchange rate movements reflected “news” rather than shifts in interest rates. Ralph Bryant said that he remained skeptical about the value of an MCI, in part because it remained unclear to him whether it represented a measure of monetary policy or a target, and he saw little value in weighting together variables rather than looking at the components separately. Both Masaru Yoshitomi and Shigemitsu Sugisaki said that the behavior of the banking system was an important factor in determining monetary policy. Kazuo Ueda noted that the Bank of Japan’s operations were not transparent, compared to those of the Federal Reserve in the United States, and that the Bank of Japan had failed to avert economic volatility in the 1980s and 1990s, The increasing political oversight contained in the new Bank of Japan Law might well increase the need for a policy rule to make the conduct of policy easier to explain and justify. Several other speakers, however, were considerably more skeptical about the value of an inflation target, because of the high level of anti-inflation credibility enjoyed by the Bank of Japan. On the same theme, Jean-Claude Chouraqui questioned whether its pragmatic policy approach had served the Bank of Japan well over the 1990s, but felt that an explicit inflation target might put too much weight on the behavior of the exchange rate.