The current situation for monetary policy
I should like to thank the Committee for the invitation to discuss the Swedish economy and the current situation for monetary policy. In my opening remarks today I shall begin by briefly summarising the recent years' economic tendencies and then consider the present construction of monetary policy. A look into the future will conclude. What I have to say is based on the latest inflation report and events since it was published.
From a dual situation to broader expansion
Sweden's economy is now entering the fifth year of expansion since the profound crisis at the beginning of the decade. Since GDP turned upwards in the summer of 1993, annual growth has averaged over 2.5 per cent. Annual inflation in this period has averaged only about 2 per cent. This is a marked contrast with the 1970s and '80s, when annual growth averaged approximately 2 per cent and inflation about 8 per cent. In one respect, however, the situation to date is still dismal, namely as regards the labour market and employment. If an economy is growing well but employment remains weak and unemployment high, there is a strong probability that the labour market is affected by structural problems.
Although growth in recent years has been satisfactory, the generation of a broader upswing in the economy has proved difficult. Initially, about three-quarters of GDP's increase came from manufacturing, which represents only about 20 per cent of Sweden's economy. This meant that bottlenecks occurred fairly quickly, accompanied by various types of inflationary impulses. Our relatively small industrial sector cannot be expected to sustain a growth rate that is higher than the total economy achieved in the 1970s and '80s.
The problem with the dual economy - a flourishing export-related industrial sector and a rather weak recovery in sectors with a more domestic orientation - stemmed in part from low confidence in economic policy's determination and ability to keep inflation low in the longer run. This in turn was a consequence of Sweden's poor record in this respect, coupled with serious problems with government finances. Long bond rates fluctuated widely and in 1994 and 1995 they actually climbed to the levels from the years of high inflation. The exchange rate continued along its depreciating path from the 1970s and '80s. For the Riksbank, the lack of confidence in the direction of economic policy left little room for manoeuvre. Although the instrumental rate was lowered during the spring of 1994, virtually all the market rates and particularly those for longer maturities rose. Sweden was approaching a point where economic policy had to demonstrate that the focus on stability was a firm commitment.
The monetary tightening and, perhaps to an even greater extent, the fiscal consolidation gradually generated increased confidence in Sweden's readiness and ability to achieve low long-term inflation. Against this background, during 1996 it was possible to cut the repo rate to its lowest level since the early 1960s. Low inflation and decreased inflation expectations also contributed to falling bond rates, accompanied by an appreciation of the krona from the earlier lows. Since December 1996 the repo rate has been unchanged at 4.10 per cent.
All in all, conditions were gradually created for a broader economic upswing. With the time lag that generally applies in economic policy, during 1997 we have observed a mood swing among households, leading for instance to increased private consumption, while export growth has been maintained. The Swedish economy is entering a phase of high activity with a combination of strong domestic demand and strong export growth.
The impact of the Riksbank's current monetary stance will take about two years to materialise in full. We therefore have to assess the outlook for inflation as far ahead as in 1999. Such an analysis was presented some weeks ago in the inflation report. I shall be returning shortly to that report and subsequent developments but first I want to discuss the approach in principle to the conduct of monetary policy over the complete economic cycle.
Monetary policy in a cyclical perspective
A matter that savers and investors are currently discussing is not whether but when the Riksbank will raise the instrumental rate. An expected increase in the repo rate is also indicated by the structure of market rates. The general impression among financial agents seems to be that a repo rate increase is probable some time in the coming winter.
The timing of a change in the monetary stance is determined by the Riksbank's assessment of inflation. The simple rule of thumb for the Riksbank can be formulated as follows: if the assessment of inflation, given an unchanged instrumental rate, points to a rate of inflation one to two years ahead that is in line with the target, then the monetary stance is well balanced. If the assessment indicates a risk of inflation rising above (falling below) the target, then the monetary stance is too expansive (contractive) and the repo rate should therefore be raised (lowered).
The time has not yet come for an increase in the repo rate but when it does, our action will hopefully be seen as a natural ingredient of a monetary policy for price stability. Our ambition is to take predictable and timely measures so that sizeable interest rate movements can be avoided.
Still, it is hardly surprising that the direction of monetary policy is being discussed. At present Swedish monetary policy's overall impact can be described as expansionary at a time when activity is becoming stronger.
The short-term real interest rate is between 2 and 2.5 per cent, depending on the exact maturity and how inflation expectations are measured. How a neutral monetary stance translates into the short real rate is difficult to determine precisely but it can be said that the current rate is comparatively low, with an impact on total demand that should be stimulatory. In addition, the level of the real exchange rate is weak in relation to a reasonable long-term equilibrium. This means that its effect is stimulatory, too. The combined effect of the short interest rate and the exchange rate is therefore tending to support the economic upswing.
All else equal, an expansionary monetary stance leads to a successive acceleration of demand. In time, this may generate a growth rate that exceeds the economy's long-term potential. The ability of the economy to cope with a high demand trend is not unlimited. As unutilised resources are activated and the output gap closes, bottlenecks of various kinds arise in the labour market or in real capital. It is in such a situation that inflation may accelerate.
In order to avoid the occurrence of shortages and accelerating inflation, the monetary stance must not remain too expansive for too long in an upward phase. Instead, a timely adjustment - bearing in mind the lag associated with monetary policy - must be made to a more neutral position before the output gap closes. This promotes a continued growth of demand at a rate the economy can absorb - for instance with the help of new investment, improved efficiency and labour supply - without generating inflationary bottlenecks. The purpose of such a monetary adjustment is not to break the expansionary trend but to achieve growth that is more sustainable. In other words, a shift from an expansionary stance to one that is more neutral is designed to bring the economy onto the growth path that matches its structure - not to bring growth to a halt. Neither should this be described as a contractive turn; it is rather a move that yields a less expansionary stance.
In the public debate the Riksbank's monetary policy is sometimes described as though repo rate adjustments were a form of "punishment" or "reward", aimed at measures that have been taken either by the political system or by labour market organisations in connection with wage negotiations. My point here has been that interest rate adjustments are a natural feature of monetary policy's path over the economic cycle. Of course this does not mean that the construction of monetary policy is not affected by fiscal expansion or contraction, or by low or high wage settlements. What I want to make clear is that adjustments to the instrumental rate in the course of the economic cycle are a natural phenomenon and their purpose is to create conditions for sustained growth and low inflation.
If a central bank waits too long before shifting from an expansive to a more neutral monetary stance, problems with bottlenecks are liable to become more widespread and the economy may become overheated. Besides entailing rising inflation, this causes firms and households to start accustoming themselves to a high level of inflation. Monetary policy may then be forced to adopt a contractive stance in order to bring inflation back to the target rate. So if the central bank waits too long, the interest rate increases may be all the larger. In the meantime, moreover, economic imbalances and tensions may become so great that they lead to economic setbacks. We are familiar with such a course of events from the 1970s and '80s, and perhaps not least from the early 1990s
Against this background, in situations when capacity is coming under pressure, it is clearly better for a central bank to act at an early stage.
Future inflation
The assessment in the inflation report, published just over a month ago, is that growth in the Swedish economy will be about 2 per cent in 1997 and around 3 per cent in 1998. In 1999 the rate is expected to go on accelerating to just over 3 per cent. This assessment envisages that, in an expanding world market, export growth remains relatively high. This is accompanied by an accelerating increase in private consumption, while public consumption at least stops falling.
Economic growth in Sweden in 1998 and 1999 would then most probably exceed the potential rate, which is estimated to be just over 2 per cent in annual terms. The broad rise in all the major demand components will therefore gradually exert pressure on the available resources. This successive increase in resource utilisation is already evident in some indicators. A growing proportion of firms are reporting, for example, that output is being restricted by productive factors rather than by demand.
The gap between total demand and total resources for production is an important but elusive indicator of inflationary pressure. The Riksbank considers that the output gap, which is an attempt to measure the degree of unutilised resources in the total economy, is 2 per cent of GDP. With the GDP forecast from the inflation report, this means that the gap will narrow by degrees and close some time during 1999, in any event towards the end of that year.
In the light of the assessments of resource utilisation and other factors, in the inflation report the Riksbank forecast that the annual increase in the CPI will be around 1 per cent in 1997 and around 2 per cent in both 1998 and 1999, with some upward tendency in the latter two years. This is our main scenario.
Like most other forecasts, however, inflation assessments are subject to uncertainty. Our forecasts as a foundation for monetary policy therefore do not consist of single point estimates of future inflation. In practice we construct a number of alternative outcomes and assess their probabilities. Normally there is one main scenario and two alternatives.
Economic activity could be stronger than we have envisaged, so that inflation is somewhat higher; but it could also be weaker, giving lower inflation. Most of the evidence suggests, however, that since the inflation report, an outcome which is appreciably weaker than the main scenario has become less likely.
The forecasts for all the demand components are naturally uncertain. It is, however, particularly worth discussing the forecast for private consumption, since this component represents about one-half of total demand. Some factors suggest that the growth of private consumption could become stronger than most observers count on today.
The background here is that for a number years the growth of household disposable income has been held back by fiscal consolidation. After 1998, however, the direct restrictive fiscal impact on household finances is expected to diminish. But in spite of the weak or negative income trend, private consumption has already been rising for a number of years, albeit rather slowly. The reason is that, besides the effect of income on private consumption, spending by households, as well as their view of the future, are influenced by the increase in wealth.
Households have recently become appreciably more optimistic about the future. House prices and share prices have been rising markedly for some years. The realised and latent capital gains of households in 1997 add up to the equivalent of about 25 per cent of their income, a figure that is as high as in the late 1980s. It would be surprising if such an increase in wealth did not contribute sooner or later to higher consumer demand. The key question, however, is how rapidly private consumption will grow.
The above applies provided the increase in asset prices is maintained. Since share prices turned upwards in the autumn of 1992, the increase on the Stockholm Exchange has been one of the largest since the early 1920s, which is as far back as the statistics go. The overall valuation of corporate profits is also historically high. The present level of share prices calls for a further rapid increase in profits for several years to come. Minor changes in this picture of the future could affect current share prices.
Another element of uncertainty concerns the economy's supply side. We cannot be sure of just how much the economy can produce without an increase in inflation. Moreover, it is not only the total reserve of resources that may affect inflation. The rate at which resource utilisation rises may also be important. A rapid activation of the unutilised resources could generate stronger inflationary impulses even if surplus capacity still exists at total level.
CPI inflation has moved up sharply during 1997. The 12-month change was down around 0 per cent for a couple of months last spring but after that it moved up fairly quickly to 1.9 per cent for September. In September alone the CPI rose 0.9 per cent, which was more than most observers had expected. The figure also aroused some uncertainty among savers and investors. The question is whether the increase in inflation represents a new and unforeseen tendency.
The main features of this development are not unexpected. One factor behind the rising CPI is increased indirect taxes, for instance for tobacco, and higher administered prices. Another explanation is that the impact on the CPI from the falling interest rates in 1996-97 is diminishing. All in all, this means that the CPI is closing in on underlying inflation as measured in various ways by the Riksbank.
Still, the CPI outcome for September is somewhat higher than the Riksbank had counted on. Figures for a single month admittedly do not warrant definite conclusions but it is certainly important to pay heightened attention to price movements in the coming months.
Normally it is underlying or trend inflation that is primarily of interest for monetary policy. This holds unless transitory price effects influence inflation expectations. In a setting where inflation expectations are firmly anchored to the inflation target, monetary policy has some room for manoeuvre. But if expected inflation starts to rise, we must be on the alert, particularly in that the current monetary stance is stimulatory. The importance of expected inflation for price and wage formation was evident not least in connection with the 1995 round of wage negotiations. At that time, expectations of future inflation seem to have been a bit above 4 per cent, which probably contributed to the excessively high level of the ensuing wage agreements. Now that the Swedish economy is on the threshold of very important wage negotiations, there must be no doubt that the Riksbank is truly serious about the inflation target.
Future wage formation is yet another risk factor that the Riksbank has to consider. There are some encouraging signs that wage formation is moving in the right direction, even though the parties have gone in for a certain amount of bidding. This tends to be a process that conveys a varying degree of realism. It is important to bear in mind that wage formation is not just a matter of the level of wage agreements. The outcome for wages is also conditioned by the wage drift that mirrors labour supply and demand. The less efficiently the labour market works, the sooner bottlenecks arise that generate increased wage drift. Tendencies to labour shortages in certain sectors are already discernible, despite the high unemployment.
Conclusion: monetary stance well balanced; changes in the spectrum of risks
A monetary policy that aims for sustained price stability is not at odds with good economic growth. This presupposes that the supply side functions so well that growth does not give rise to inflationary impulses. But if the structure of the economy is such that only moderate growth is feasible, demand and accordingly the monetary stance must be adjusted accordingly. The alternative is demand growth that exceeds what the economy is capable of producing without inflationary bottlenecks. Accelerating inflation is a threat to sustainable growth and may even result in serious economic setbacks, as history has taught us.
As economic activity becomes stronger and monetary policy adjusts its sights on the future picture, the Riksbank's monetary stance will have to gradually move away from its current expansionary position. The timing of such a move has to be assessed in the light of new information and today one cannot say when it will happen.
Our present assessment is that the monetary stance is well balanced. There has been some shift in the spectrum of risks but our inflation forecast for the coming years is unchanged.
