LESSONS FROM THE BRITISH EXPERIENCE OF MONETARY POLICY (Vinod Anand) The subject of this Paper is inflation targets as a technique for conducting monetary policy. I would like, however, to start by discussing the more general proposition that price stability is a desirable target for monetary policy, and then talk about inflation targeting, focusing in particular on the need for an inflation targeting strategy to be credible, and on the issue of inflation forecasting. Finally, I will touch upon two important issues that need to be settled when designing an inflation target strategy - what price index should you select for the target, and whether the target should be a single point or a band. My comments will be based largely on the experience of the United Kingdom with inflation targeting 1 do not wish to make any recommendations about how monetary policy in South Africa should be conducted. Inflation targeting is based on the belief that price stability is the proper objective of monetary policy. That belief can be justified from a number of viewpoints. In the language of economic theory, money is a social contrivance that adds to human welfare by conveying information about relative values in a form that is simple and easy to understand, Information about relative values is crucially important to a market economy because it guides decisions about resource allocation. Inflation makes the information about relative values which money prices convey harder to interpret because individual prices do no all change in step in the inflationary process. One of the consequences of inflation is continual temporary relative price changes which do not contain useful information and can therefore only be confusing. Inflation increases the noise-to-signal ratio of the information provided by money and in that way reduces welfare. Inflation is an unproductive process. A rising price level does nothing in itself to increase the total of goods and services available for consumption and investment. It does, however, have implications for the distribution of income and wealth: it benefits those who are well protected against it, perhaps because they are particularly financially sophisticated or politically powerful, at the expense of the ill protected, unsophisticated and politically weak. Inflation can thus be, and very often in history has been, socially divisive Why, in that case, does inflation happen? One reason is as follows. It is widely understood that a shift to a more expansionary monetary policy can cause a temporary acceleration in demand and output. It may well be the way to maximize short-term economic growth. If the more expansionary monetary policy causes demand to exceed the productive capacity of the economy, however, then at some point inflation can be expected to accelerate, Accelerating inflation is an indication that current developments in the economy arc unsustainable and that something needs to adjust in order to bring the economy back onto a sustainable path. The adjustments will normally be unexpected. They are therefore likely to disrupt the earlier plans of enterprises and consumers and thereby to reduce the productive capacity of the economy. In other words a policy designed to maximize short-term economic growth will often fail to maximize long-term economic growth. In particular, instability in monetary policy - by which I mean unexpected and therefore disruptive changes - is likely to damage the growth potential of the economy. That reasoning suggests that one of the features of a monetary policy that maximizes long-term economic growth is likely to be stability - in other words not many unexpected changes. Arguments about how inflation distorts price information apply with particular force to the information that guides inter-temporal decisions. Inflation erodes the value of money, and interest rates have to be higher to provide compensation. If interest rates are higher, the effective duration of loans is shorter, because the interest payments made during the course of the loan period have to make good the erosion of the real value of the principal induced by inflation. This effect artificially shortens the time horizons of economic decision makers, with some likely cost to the efficiency of investment and to economic growth If real interest rates are 4%, inflation is 10% and nominal interest rates are 14%, then someone who borrows money at 14% interest for say, 5 years, to finance a project pays back at the end of the first year not only 4% real interest but also 10% of the principal In make good the effect of inflation on the real value of the principal. In effect, 10% of the principal is repaid after one year, not five. In practice a positive correlation has been observed between the level of inflation and its variability. Uncertainty about inflation is likely to cause lenders of money to demand higher interest rates not only to compensate them for expected inflation but also to compensate them for the risk associated with the variability of inflation. That means that high inflation is likely to be associated with high real interest rates, which in turn are likely to inhibit investment and economic growth. Moreover uncertain inflation is likely to be accompanied by uncertainty about monetary policy, and as I said earlier, uncertainty about monetary policy is likely to be inconsistent with maximum long run economic growth. These are therefore reasons for expecting that a policy directed at maintaining low inflation is likely also to help maximize the productive capacity of the economy and to help create good conditions for economic growth. And indeed there is some evidence in the history of the last 40 years or so of a statistical correlation between low inflation and high economic growth. The estimated effects of inflation on growth look very small in any single year, but they accumulate over time, so that lower inflation can have a large effect of living standards if maintained for periods of decades All this suggests that low inflation or price stability should not be regarded as a macro-economic oh1ccnc having the same status as other macroeconomic objectives like short-term output growth. Rather, it should 6e regarded as a necessary pee-condition for maximizing sustainable output growth.