Publications and Research

1997 conference

Price Stability, Inflation Targets and Monetary Policy

Proceedings of a conference - May 1997

Table of Contents

Session 7 - General Discussion

W. Paul Jenkins (PDF, 30 kb)
Norman Cameron (PDF, 28 kb)
William Scarth (PDF, 25 kb)

Prepared by Irene Ip and Maral Kichian

Pierre Fortin said that the conference had made important distinctions along two dimensions: first, whether an inflation target is believed and, if it is, whether it is acted upon by the public; and second, whether the target is credible and also the right target.

He noted that, based on our own experience and that of the panel of central bankers, there was universal agreement among the participants that it is better to have targets than not, and that the United States was now the odd one out among central banks. But on the question of whether the target is acted upon by the public once it is believed, he said that there remains considerable uncertainty.

He also argued that, while there was universal agreement that there are real and important benefits from central bank transparency and credibility with financial markets, there was less certainty about the impact of credibility on the real sector of the economy. One explanation he gave for this is that, even if the many players believed absolutely in the central bank's objective and that it was feasible, they would still take notice of other wage setters in order not to be left behind, leading them into a "prisoner's dilemma."

With regard to whether the target is credible and is also the right target, he suggested that much more work is needed, because our knowledge of the analysis of the trade-off between inflation and unemployment is piecemeal and incomplete. He said that there is still great uncertainty about the benefits and costs of bringing inflation down from 4 per cent to 1 per cent, as the papers by Ambler and Cardia and by Black, Coletti, and Monnier underlined. The Ambler and Cardia paper, he added, also stressed the difficult nature of the econometric estimation of the right relationship between inflation and growth.

In response to Scarth's wrap-up remarks, Charles Freedman stressed that the monetary conditions index was not an intermediate target, but more an operational target, in the same class as the short-term interest rate.

Freedman also responded to Cameron's proposal to use as the target the current indicators of the private sector's expectations; Freedman argued that this approach was not in fact workable. He argued that private sector expectations take two things into account: what the central bank might do, and what would happen to the rate of inflation if the central bank did not do it. Thus, if it were believed that the central bank would act appropriately, everybody would expect the midpoint of the central bank's target to be realized, regardless of what shocks might hit the economy. Since the Bank must distinguish between the effects of credibility and those of particular shocks, the private sector forecast would not be very helpful to the Bank.

Cliff Halliwell suggested that one method of achieving some aspects of the benefits of price-level targeting while avoiding some of the costs would be to put the inflation target¯say, of 2 per cent a year¯in terms of a price-level target that is 2 percentage points higher than the current price level a year from now. Abstracting from the fact that there are many other strategic considerations when actually implementing monetary policy because of the lags, the Bank might start rolling forward inflation rates for 12, 24, and 36 months in terms of a targeted price level one, two, or three years from now, and still have some chance to let some of the bygones be bygones as they rolled this thing along.

Daniel Racette noted that three major conclusions could be drawn from this conference. First, that there was a high variance among the findings of the various papers presented. That is, while it seems clear that there are both benefits and costs to lowering inflation, the magnitudes of these are not precisely known. Second, he stressed that the Bank should not give the impression to the public that it has good control over certain economic variables when it actually does not. He explained that other conferences have indicated that the instruments of monetary policy are also imprecise, so that strategies based on these, such as making the MCI a target, could be potentially dangerous to pursue. Finally, he urged the Bank never to engage in an inflationist strategy (for instance, by widening the bands) and to remember the costs of high inflation.

Pierre Duguay followed up on Cameron's comment on the incredibility of a price-level target by discussing price-level versus inflation targeting. He suggested that to talk of a decline in the price level as deflation, as Cameron had, was incorrect. Deflation, Duguay argued, is bad essentially because it tends to be unexpected so that it generates uncertainty. In other words, without an anchor, one does not know where it will end, so its only true cost is that, in contrast to disinflation, there is no automatic corrective mechanism of falling interest rates. When there is a price-level target, if there is a negative demand shock, so that the price level goes down, it is possible to return to the targeted level using the exchange rate, asset prices, or other types of instruments. In this case, expectations are that the price will go back to its targeted level, so that the unexpected decline in prices is no longer a big problem. In the case of a shock that pushes the price level above its target¯which is the shock Cameron focussed on¯Duguay argued that it is possible to have a policy-induced decline in the price level at little cost, provided that the central bank's reaction function is well understood so that the decline is anticipated.

Duguay also commented on the indexation issue of the tax system, saying that the great difficulty in indexing the tax system is that the tax system is based on standard accounting practices that do not index for inflation. He added that the cost of inflation coming from the lack of indexation of the tax system is just the tip of the iceberg, and that similar costs exist in every aspect of private planning, investment planning, and so on, since these activities are all based on a non-indexed accounting system. He concluded by suggesting that these costs of inflation are best examined in a general-equilibrium model.