Heikensten: Monetary policy and the Swedish ecomony 1997
First a word of thanks for the invitation to talk with you about monetary policy.
Today I shall be considering this field in terms of three topical issues:
- First there is the path of inflation in recent years. In this context I shall dwell on the significance of transitory effects, not just for recent price tendencies but also more in principle.
- Next comes the issue of employment. What can monetary policy do to help create new jobs in Sweden?
- The third matter is the rule system for monetary policy. What is the point of delegating the management of monetary policy and the task of establishing low inflation to a central bank? In this connection I will also look at various proposals that are being discussed at present for amending the rules for the Riksbank.
1 Inflation and monetary policy
Price stability is the Riksbank's paramount objective, specified in the target of keeping the increase in the consumer price index to 2 per cent, with a tolerance interval of *1 percentage point. There is a time lag of one to two years before a monetary policy measure has its full effect on inflation. Our discussion should therefore focus on future inflation. But first I want to say something about inflation in the past year. Inflation and the transitory effects The CPI is used by the Riksbank to target inflation mainly because it is familiar, published monthly with a short time lag and seldom has to be revised.
However, using the CPI as a target variable is liable to face monetary policy with explanatory problems. The past year is a case in point. The low rate of inflation stemmed mainly from transitory effects; the consolidation of government finance and increased confidence in economic policy helped to strengthen the exchange rate. During 1996 it proved possible to lower the instrumental rates almost 5 percentage points.
Altogether, in 1996 decreased interest costs and the stronger exchange rate had a downward effect on the CPI of around 1 percentage point. In some countries the inflation target is assessed in terms of the CPI with various components excluded, for example indirect taxes and subsidies or mortgage interest costs. It may then be asked whether the CPI is the most appropriate target variable and whether some measurement of underlying inflation might be preferable. An advantage of focusing on a single underlying indicator is that it might help the Riksbank in its contacts with the general public and make monetary policy more transparent. Analysis and debate would be concentrated on factors that monetary policy influences. This, however, presupposes the existence of an underlying yardstick that can take every conceivable transitory effect into account. Such measurements are difficult to construct. Specifying all the circumstances that have to be considered so as to obtain the best picture of underlying inflation and doing this in advance is not a simple matter. Using underlying inflation also has the drawback that essential information about the inflation process may be overlooked. For example, changes in import prices are a part of the domestic inflation process because they affect price setting in trade. In the present state of our art, the drawbacks of changing to a measure of underlying inflation for the Riksbank's monetary policy are greater than the advantages. As I have indicated, however, the CPI may be markedly affected by transitory effects. It is therefore important to monitor measurements of underlying inflation that have been stable over time. The path of the underlying inflation process is an important factor in the Riksbank's monetary policy decisions.Gradual recoveryIn growth terms, 1996 was weaker than 1994 and 1995.
The slowdown mainly came from marked stock reductions and a contraction of public sector operations. But even the growth of final demand in the private sector tended to slacken in 1996. There are now clear signs of an economic recovery; the volume growth of exports remains high, along with indications that a stronger increase in private consumption is on the way. What, then, will be inflation's path in the years ahead? The Riksbank considers that at present the Swedish economy has unutilised resources and even if growth does pick up during 1997 and 1998, it is unlikely that these resources will be activated to such an extent that the inflation target is in danger. The scenario in the December inflation report involves a gradual increase in the rate of inflation in the coming years. The main explanation for this upward movement is that the transitory effects - from lower interest rates and a stronger exchange rate - which kept inflation down in 1996 will be fading. The registered rate of CPI inflation is then expected to approach inflation's underlying trend, which in the years ahead is judged to be more or less in line with the inflation target.Lower inflation propensity The new price and other economic information in recent months reinforces the impression that price pressure is low. To this extent, the conditions for fulfilling the inflation target have improved. On the other hand, a weaker exchange rate has already tended to stimulate activity. The outlook for inflation and the rapid progress in consolidating government finance suggest that there is room for a fall in long bond rates. Underlying economic factors indicate, moreover, that the recent depreciation has left the krona undervalued, so that it should appreciate. How should the exchange rate be assessed in connection with monetary policy decisions? According to experience in recent years, the impact of a depreciation on domestic prices is limited. But domestic demand has been weak.
Provided there is stable confidence in the policy for price stability, price effects from a weakening of the exchange rate are generally transitory. A depreciation is a problem for monetary policy only when it has a lasting expansionary effect or if inflation expectations rise.
The economy's inflation propensity, which seems to have decreased in recent years, is an important factor to consider when assessing the threat from future inflation. The term inflation propensity stands here for the structure of the inflation process, which is liable to be changed by various factors. If stronger competition, for example, in certain segments of trade results in lower price increases than before when consumption is rising, then the economy has a lower inflation propensity. Examples of factors that may have helped to reduce the inflation propensity are the fall in market expectations of inflation and the higher acceptance of the Riksbank's inflation target. Another cause of a decreased inflation propensity can be lower inflation as such. Price comparisons are easier for households and firms when inflation is low. Price increases are more visible and producers have more difficulty in raising prices without a relative price shift. The situation in this respect is one of the main issues that the Riksbank will have to evaluate. We shall be returning to this later in the year.
All in all, the picture that emerges of the Swedish economy suggests that the monetary stance is well balanced. Major changes have occurred in the past year, for instance in the repo rate. Time is needed to assess their effects. This also motivates a cautious line in monetary policy.
2. Employment and what monetary policy can do
The low inflation in the past few years has been accompanied by respectable growth.
Since 1994 economic growth has averaged 2.7 per cent a year, which is higher than the normal rate in recent decades. Given the low capacity utilisation initially, the high growth as such is not remarkable. What is more surprising - and alarming - is that a recovery of employment seems to be considerably more sluggish than the upswing in total production.
The problem of generating new jobs is by no means peculiar to Sweden but something we have in common with much of Europe. The situation in this respect contrasts markedly with the United States. There are also exceptions in Europe, one being the Netherlands in the past decade. This suggests that cures for the high unemployment exist, though naturally the solutions in other countries may not apply as a whole to Swedish conditions.
The high unemployment in Sweden is now a subject of intensive discussion. Today I shall try to describe in principle what monetary policy is capable of doing and what it is not in a position to achieve.
Restricted potential in the longer run
In an analysis of monetary policy's potential for influencing employment it is natural to make a distinction between the short and the longer run. It is generally agreed that monetary policy can be used to affect production and employment in the short run. An important reason why monetary policy is capable, at least in principle, of influencing employment in this time perspective is the existence of rigidities in price and wage formation. Suppose, for example, that low wage settlements are concluded with a certain rate of inflation in mind. If monetary policy becomes more expansionary and inflation takes off, wage costs will be lower than expected in real terms. It will pay firms better to take on additional labour and employment will rise. In the short run, then, one can expect a trade-off between inflation (or the extent to which monetary policy is expansionary) and employment (or unemployment).
The existence of such a short-term trade-off does not mean, however, that it will serve to raise employment permanently. This is because when actual inflation changes, economic agents (firms, employees, etc.) normally adjust their expectations of future inflation. So in the next round of wage agreements, for example, employees may demand compensation for a higher expected rate of inflation. Changing expectations mean that the trade-off between inflation and unemployment ceases to be stable.
This is intuitively intelligible. If the normal rate of inflation is, say, 5 per cent, it seems reasonable that employment will rise if inflation moves up to 10 per cent because this reduces wage costs in real terms. In the figure this amounts to moving from A to B. In time, however, 10 per cent inflation will be regarded as normal and wage demands will be based on this level. With higher nominal wage increases, real wage costs will move up again and unemployment will tend to return to its initial level. In the figure this means moving from B to C. In order to boost employment in this situation, inflation would have to rise to say 15 per cent and the procedure will be repeated. All that is ultimately achieved is higher average inflation.
The level of unemployment to which an economy returns when inflation expectations have adjusted is usually referred to as equilibrium unemployment. It is determined by structural factors such as demographic changes, technology and the institutional set-up in the labour market, factors that are immune to monetary policy. The important conclusion to draw from this is that permanently low unemployment cannot be achieved by accepting permanently high inflation. Under favourable circumstances, to which I will shortly return, monetary policy can promote an economic recovery; but an appreciable improvement in the unemployment situation calls for measures in many other fields, above all in the workings of the labour market and wage formation.
What should monetary policy aim for?
If a central bank is incapable of "creating jobs" on a permanent basis, what should be its primary task? By means of an appropriately balanced monetary policy a central bank can determine, or at least strongly influence, the rate of inflation that prevails in the economy and underlies economic decisions by firms and households. Price increases are, of course, conditioned by many other factors, such as the economic situation, the fiscal stance, wage formation and so on. But inflation is a monetary phenomenon and in the longer run the Riksbank, by controlling the supply of money in a wide sense, can also control its value in terms of inflation.
The perception that the primary task of a central bank should be to maintain price stability can be said to have won general acceptance. In that low, stable inflation helps to improve the functioning of markets, it seems reasonable that, indirectly, monetary policy also promotes economic growth and employment. In their economic decisions, households and firms have one less source of uncertainty - inflation - to worry about.
The inability of monetary policy to affect long-term employment and the fact that it should focus on maintaining price stability does not mean that its capacity to condition production and employment in the short run can not sometimes be used for stabilising activity. Monetary policy can contribute to an economic recovery in so far as this is feasible without jeopardising price stability. Problems may arise, however, if monetary policy is perceived as focusing primarily on the stabilisation of employment. As we have seen, people's expectations play a major part in the inflation process. If people expect monetary policy to concentrate in the first place on stabilising employment, then firms and employees will not regard unduly high wage increases as all that serious because they will count on measures being taken to stimulate demand if employment is in danger.
Sweden in the 1970s and '80s is a good illustration of the problems that may arise. Wage and price increases in that period were systematically higher than in other countries, which led to recurrent cost crises and problems with competitiveness. An important reason for the lack of adjustment to wage increases in the rest of the world was that employees and firms reckoned that the government and the Riksbank would ultimately bail them out with devaluations, which is what actually happened. In this way inflation expectations were adjusted in advance to an anticipated pattern of economic policy reactions, whereby high price and wage increases were accepted or accommodated by devaluating the krona. As a result, strong inflationary mechanisms were built into the economy. As we have seen in recent years, once such mechanisms have been established, they are difficult to change.
3. Monetary policy's rule system
Experience has shown that more is needed than proud declarations from political decision-makers that they will safeguard a stable value of money and are not prepared to adapt policy to excessively high wage increases. We heard plenty of that in the 1980s. One way of enhancing credibility that has been found to work well and is supported by economic theory is for the democratically elected body to delegate monetary policy to a central bank with a clear, specified obligation to safeguard the value of money. The central bank is given full control over the instruments, principally the short interest rate, for fulfilling the stipulated objective. This arrangement limits both the room for short-run attempts to boost employment and the risk of inflationary mechanisms taking root in the economy. That also lessens the risk of a repetition of what happened in the early 1990s when the inflation bubble burst and unemployment rose.
Credibility provides greater manoeuvrability
It may be worth underscoring that a system where price stability is credibly delegated to the central bank can confer greater freedom of action in economic policy. This may seem paradoxical but if people feel they can rely on the central bank, in an economic slowdown it may be possible to ease the monetary stance without triggering higher inflation expectations and bond rates, because everyone counts on policy being tightened before inflation moves up. We have experienced something of this sort in Sweden in the past year.
In a system with poor credibility, a monetary easing may be perceived as a sign of a lasting upward shift in inflation. This in turn may generate self-fulfilling expectations, with future problems when efforts are made to control inflation once more. It is understandable that the risk of this is seen as particularly great in economies where inflation has been high for a long time. A more independent central bank is thus in a better position not only to maintain low inflation but also to operate at times with a policy that includes an element of stabilisation (though the problems of fine tuning stabilisation policy should not be underestimated). The German central bank is one example; over a series of decades, economic slowdowns could be parried by lowering the short interest rate fairly rapidly and markedly.
It may be in place to point out that credibility and, for that matter, independence can be achieved in various ways. Some countries have recently clarified their regulations so as to bring them into line with the arrangement I have just advocated. One example is New Zealand. In other countries, for instance the United States, policy has been conducted so that it has gradually earned increasing confidence within the time-honoured system. In my opinion the rule system is important; it reflects the general commitment to monetary policy and can provide a short cut to credibility. In a democratic society, however, monetary policy is ultimately founded on popular support.
That is one reason why it is important for the Riksbank to operate in such a way that confidence is created in the policy of price stability. Among other things, this calls for a clear intellectual framework that can be evaluated and for an open discussion about the direction of monetary policy.
Delegates accountable to a democratic body
An argument that is heard from time to time in the monetary policy debate is that the delegation of monetary policy's construction to an independent central bank would be undemocratic. In my view that is incorrect. The Riksbank is an agency of the Riksdag (Sweden's parliament), which clearly can stipulate the objective which the Riksbank is to aim for. It is worth emphasising, however, that it would naturally be inappropriate for the Riksdag to set up objectives that the Riksbank is not in a position to fulfil, like that - as we saw earlier - of "creating jobs" on a permanent basis. When, but only when, the Riksbank's objective has been established, will it be possible to monitor operations effectively. It should be realised that the delegation of monetary policy to an independent central bank is a way of generating confidence that the declared policy of the Riksdag and the government will in fact be implemented.
In the coming years the rule system for monetary policy will be changed. As an adherent to the Maastricht Treaty, Sweden has undertaken to make changes in, for example, the status of the Riksbank's governor. He or she shall not be dismissed except on explicit grounds of serious misconduct, etc. Neither is the Riksbank to be given instructions from third parties, e.g. government, except in a statutory form. It also looks as though the objective of the Riksbank is to be formulated in law and will be price stability. The question of eligibility to serve on the governing board will have to be reviewed. Active members of parliament are delegates in Sweden, which is entirely contrary to traditions in central Europe.
The Riksbank differs from most central banks in Europe in that we are formally responsible for exchange rate policy. The Government sees this as a problem in EU cooperation and has therefore instituted an enquiry into exchange rate competence. In the light of what I have just said, these conclusions are not sound. The proposal to transfer responsibility for exchange rate policy from the Riksbank to the government makes it more difficult for the Riksbank to safeguard the value of money effectively and thereby contribute to a favourable level of interest rates and also - when this does not endanger price stability - to stimulate the economy and employment.
In a world with free capital flows there is a direct link between monetary and exchange rate policies. The two policy fields may have conflicting goals. Moreover, if the government, via exchange rate decisions, can affect the ability of the Riksbank to fulfil the inflation target, there may be misunderstandings about where responsibility for the operation of monetary policy lies.
In my opinion it may be reasonable - though this, too, is questioned by some economists - for the government to decide the type of exchange rate system Sweden is to have in relation to other EU countries, whether the krona should join the European exchange rate mechanism or remain flexible. There is a close connection here with the rules and regulations that apply in the rest of the EU and the matter is very much a political issue. But within a given exchange rate system, it is the Riksbank that should decide an appropriate rate for the krona or the band within which it may move. Such decisions can be made after an exchange of information with the government but in the final analysis it should be up to the Riksbank to determine whether or not they are compatible with price stability. With a flexible exchange rate, policy should operate, as today, with an inflation target, and the government should have no possibility of influencing the construction of exchange rate policy.
Democratic control is important. But it should be exercised via the Riksbank's principal, the Riksdag, on whose behalf monetary policy is conducted. It should be noted that in exceptional situations, when the price stability objective may have to be suspended for the time being (e.g. a major oil price shock), both the government and the Riksbank may call on the Riksdag to amend the Riksbank's statutory objective. In that way the democratic control of monetary policy can be clearly ensured.
