Publication Details


In my Governor's Address at the Ordinary General Meeting of Shareholders of the Reserve Bank on 24 August 1999, I stated that it is advisable to move away from the "eclectic" or informal inflation-targeting monetary policy framework to formal inflation targeting. This change is necessary because the eclectic framework has at times created uncertainties about Reserve Bank decisions and actions which were perceived as being in conflict with the stated guidelines for the growth in money supply and bank credit extension.

The question may be asked whether the adoption of inflation targeting will lead to a more effective monetary policy? I believe that this will be the case and that inflation targeting should minimise the social and economic cost of achieving price stability. In my address this evening I want to explain this position that has been taken by the Reserve Bank by elaborating on the following aspects of inflation targeting, namely:


What are the main characteristics of inflation targeting?

What are the advantages of inflation targeting?

Are there any disadvantages in shifting to an inflation-targeting framework?

What are the preconditions that have to be met before implementing inflation targeting?

What factors should be taken into consideration with the implementation of inflation targeting?




A number of industrial countries, such as New Zealand, Canada, the United Kingdom, Sweden, Finland, Australia and Spain, have adopted an inflation-targeting framework during the 1990s. More recently some of the developing countries have also opted for this monetary policy framework. The Czech Republic has operated a fully fledged inflation targeting regime since December 1997. Israel, Chile and Mexico are countries that have one-year inflation targets, while Poland, Hungary and recently Brazil have multi-year inflation targets.

The motives for the adoption of inflation targeting have varied considerably from country to country. In some countries, such as the United Kingdom and Sweden, the collapse of their exchange rates led to inflation targeting in order to assure the public that monetary policy would remain disciplined. Other countries, such as Canada, introduced inflation targeting because of problems experienced with the targeting of monetary aggregates.

Whatever the motives of the countries were, the adoption of an inflation-targeting strategy in all cases reduced the role of formal intermediate targets or guidelines, such as the exchange rate or the growth rate in money supply. Commitment to an intermediate target would be inconsistent with inflation targeting except if it is the only determining factor of inflation. Obviously this is unlikely. Although growth in money supply is a precondition for a general rise in prices, it is not the only factor that causes inflation.

In an inflation-targeting framework the central bank has to adopt a strategy of determining directly what the likely path of inflation will be. In inflation targeting close attention is typically given to changes in indicators which in the past have affected inflation. This has led to sophisticated models for the prediction of inflation and detailed assessments of factors that could affect it. The prediction of inflation is of the utmost importance in the implementation of monetary policy, because changes in policy measures are based on likely future price developments.

The fact that inflation targeting has to rely on forecasting has led to the criticism that this is a weakness of the technique because forecasts are inherently unreliable. Although it is true that an inflation-targeting framework is based on forecasts, it is also true that any other monetary policy framework has to take account of the fact that policy changes will only affect inflation some time in the future. If the objective of the central bank is the attainment of financial stability, it will always have to take a view on how its current policy stance will effect future price developments, whatever monetary policy framework it decides to pursue. The difference between inflation targeting and other monetary policy frameworks, is that inflation targeting makes forecasting explicit and transparent.

The adoption of inflation targeting by a central bank does not mean that the central bank must apply definite rules and is not left with any discretion. Exclusive emphasis on inflation goals over the short-term could lead to a highly unstable real economy in the case of serious supply shocks. If a severe supply shock hits the economy, keeping inflation close to the long-run target could be very costly in terms of lost output. In such cases, some discretion must be applied. Many of the central banks opt for escape clauses in these circumstances. Others have defined the inflation target in such a way that changes in food prices, the price of oil or the effects of value added tax are excluded from the index that they target.

The degree to which the central bank is held formally accountable for inflation outcomes varies considerably. In New Zealand the tenure of the Governor of the central bank is linked directly to the achievement of the inflation targets. In most other countries there are no explicit sanctions on central banks if they miss the target. However, missing the target by a very wide margin and consistently would, of course, lead to lost reputation or prestige and eventually perhaps to a request for the resignation of the officials involved in targeting. In most countries applying this framework it is still too early to determine what prospective penalties will be applied.

The achievement of the target is, however, also important because it affects the credibility of the central bank. An inflation-targeting framework can only be successful if the public is convinced that the central bank is serious about containing inflation. This means that the government and the central bank cannot simply change to inflation targeting without reference to anyone else. It must ensure that the ordinary people in the street are in favour of price stability. Any misconception that the central bank is not concerned about economic growth and employment creation must be removed. It is important that the public does not get the impression that the central bank is dogmatic about the containment of inflation and does not care about other critical issues of importance to the economy.

It is also important that the public is ensured that there are no hidden agendas, in other words that the objectives stated will be the ones actually pursued. Bill Allen, an official of the Bank of England, at the beginning of this year in a speech at the Reserve Bank indicated that this was a major issue in the United Kingdom when they adopted inflation targeting in 1992. This was partly due to their mixed historical record at that time in achieving monetary policy objectives and partly because monetary policy had traditionally been conducted in an environment of considerable mystery and secrecy. Fortunately, we have progressed somewhat on both these fronts in South Africa and there should not be any difficulties in switching to inflation targeting.



What are the advantages of inflation targeting for South Africa? Firstly, inflation targeting can improve the co-ordination between monetary policy and other macroeconomic policies depending on the way the target is set and whether the target is consistent with other policy objectives. The setting of inflation targets should preferably (but not compulsorily) be a joint effort between the government and the central bank. Inflation targeting will be most effective when economic policies are well co-ordinated. This co-ordination must be clearly spelled out and all stakeholders should be consulted, including business and the trade union movement.

A formalised co-ordinated approach will probably lead to the best results. The Reserve Bank is therefore of the opinion that with the adoption of inflation targets a policy target agreement should be drawn up that would be signed by the Minister of Finance on behalf of Cabinet and the Governor of the Reserve Bank to define precisely the co-ordinated effort needed to contain inflation in pursuit of the broader economic objectives of sustainable high economic growth and employment creation. While this agreement should define the practical implication of the Reserve Bank's goal of price stability, it should also guarantee the instrumental independence of the Reserve Bank.

Secondly, the announcement of inflation targets clarifies the central bank's intentions and reduces uncertainty about the future course of monetary policy. Inflation targets make policy transparent. They make the central bank's intentions explicit in a way that should improve the planning of the private sector.

Thirdly, inflation targeting helps to discipline monetary policy and strengthens the central bank's accountability. If targets are not met, the central bank has to explain what went wrong. This leads to a better understanding on the part of the public on what basis monetary policy decisions were made.



Although inflation targeting has certain definite advantages when compared with other monetary policy frameworks, it could also have certain disadvantages. One of the limitations of inflation targeting is that it is a complicated approach to implement. Although all monetary policy frameworks should be forward looking, inflation targeting relies heavily on forecasts. Where inaccurate forecasts are made public it can obscure the central bank's objectives and reduce its credibility.

Compared with other monetary policy frameworks there is also the risk that inflation targeting could lead to inefficient output stabilisation. This can occur particularly in the event of significant supply shocks, such as sharp changes in the price of oil.



In view of these potential disadvantages of inflation targeting, it is important that certain preconditions are met before a decision is taken to implement this framework. In this regard it is important that a central bank is free to pursue financial stability. The inflation target may be jointly set by the government and the central bank. However, once the target has been determined the central bank should be free to use any instrument to achieve the ultimate objective.

In the application of inflation targeting it is further important that there is a commitment by all authorities to the objective of price stability. Preferably there should be close co-operation between monetary policy and other policies, and in this co-ordinated effort the inflation objective should be an inherent part of overall policy.

To implement inflation targeting it is important that well-developed financial markets exist in a country. The policy instruments generally used by monetary authorities require effective money, capital and foreign exchange markets. If financial markets do not react quickly to the instruments applied, it obviously reduces the effectiveness of monetary policy and leads to a delay in impacting on inflation.

For the implementation of inflation targeting the central bank must have the necessary resources, i.e. human, technological, etc. Inflation targeting requires sophisticated forecasting models which need advanced computer hardware and software. In developing economies in particular, this could be an important constraint on the implementation of inflation targeting.

Fortunately, the Reserve Bank's staff should be able to refine and strengthen the Bank's forecasting framework, especially with the help of international experts in this field. The Bank has enough experience in information technology for the implementation of inflation targeting and the other preconditions are generally met in the case of South Africa.



Before inflation targeting is implemented, it is important that:


a decision is taken on the inflation target that will be set;

a satisfactory forecasting framework is developed; and

a high degree of transparency is ensured.

In all inflation targeting countries the practice has been to specify the target in terms of the consumer price index, or some variant thereof. There is, however, no consensus on how inflation should be measured. Preferably the index should include a range of products whose prices fully describe changes in the cost of living and are generally accepted by the public. Including prices over which policy has no control may lead to misleading signals when these prices move out of line. Policy makers in many instances have accordingly opted for a target that excludes certain measurable components that are either unpredictable, volatile or unresponsive to policy.

In addition, a choice has to be made between setting a single point or a target range. This choice reflects a tradeoff - a narrower band may be interpreted as indicating a stronger commitment to the inflation target, but frequent breaches could undermine credibility. A fixed point target is much more difficult to hit than a band. A single point provides the best focus for inflation expectations and avoids the disadvantage of a band which tends to concentrate expectations towards its upper boundary. A band, however, leaves some discretion to the central bank and can provide more flexibility in the case of unforeseen price shocks.

It is also important to decide at what level the target should be set. Preferably it should rather be on the low rather than on the high side. High targets may give the impression that the central bank is not serious about combatting inflation. It is accordingly better to lengthen the period over which the target will be reached, than starting at too a high level.

The implementation of inflation targeting requires that the authorities must be able to develop a satisfactory forecasting framework. Although South Africa has experienced major structural changes in recent years, it seems that we are in a position to forecast inflation. The Reserve Bank already makes regular inflation forecasts using a large macroeconomic model which is based on a wide range of financial and economic equations. In the application of inflation targeting it will be prudent to supplement this large model with a number of smaller and single equation models.

Mervyn King, deputy governor of the Bank of England, recently stated that forecasting, more than any other aspect of economics, brings the discipline into disrepute. Forecasts should not be presented as a single number, but rather as a probability distribution. The Bank of England has accordingly adopted a so-called fan-chart forecasting technique which shows the relatively likelihood of possible inflation outcomes. Careful consideration will have to be given to these techniques of other central banks before inflation targeting is introduced in South Africa.

It is important that the Reserve Bank enhances transparency for the effective operation of an inflation-targeting framework. Transparency introduces predictability, and helps to ensure that expectations are consistent with the objective of price stability, thus lowering the cost of achieving the inflation target. As pointed by prominent economists out many years ago, monetary policy should avoid exacerbating fluctuations of output and employment by introducing unnecessary uncertainty.

A transparent monetary policy will mean that changes in short-term interest rates should not surprise the market. Markets should be able to anticipate decisions taken by the Reserve Bank. Transparency should promote the predictability of policy. In this regard Mervyn King has stated that a successful central bank should be boring, it is like a referee whose success is judged by how little his or her decisions intrude on the game.

Although the Reserve Bank has progressed considerably in recent years in providing more detailed information to the market and in explaining its policy stance, the Bank has decided that with the introduction of inflation targeting it will improve the transparency of monetary policy further by periodic testimony to Parliament, the regular publication of a monetary policy review report and focussed statements after the meetings of the Monetary Policy Committee. In both cases the Bank will provide an assessment of how it perceives the underlying economic and financial conditions and explain the monetary policy stance.



Although inflation targeting is a difficult monetary policy framework to apply in practice, most of the preconditions for its implementation can be met in South Africa. Inflation targeting should lead to closer economic co-ordination and greater transparency and accountability. It is, however, at the same time important that our house is in order before we implement this framework. In other words we must get our ducks in a row first.