Inflation and the interest rate
First I want to express my thanks for the invitation and opportunity to discuss Sweden's economy and the Riksbank's assessment of future inflation.
In last week's inflation report the Riksbank concluded that in the present situation the monetary stance is well balanced. As usual, the report starts from the technical assumption that the monetary stance is unchanged - that the repo rate continues to be 4.10 per cent - for the period into 1999 and then analyses what this implies for effects on demand and price movements. Of course this assumption should not be seen as a promise to leave the repo rate unchanged in the years ahead. The assessment that the monetary stance is well balanced means that we see no reason to adjust the repo rate in the near future in order to fulfil the 2 per cent inflation target up to the time horizon with which monetary policy operates, that is, usually between 1 and 2 years. But our analysis of the economy and risks of inflation in the years ahead does carry some reservations.
The report notes that in the current cyclical phase it is particularly important that incoming information is closely monitored for signs that the picture of inflation is changing. In view of economic activity's rising trend and our determination to fulfil the price stability target, this is only natural. There are uncertain factors and today I shall be discussing some of them. But that discussion should not overshadow the general picture of an economic situation in Sweden that is much better than it has been for a long time. The conditions are good for achieving a period of high, sustainable growth and persistently low inflation. This situation must be managed properly.
One source of uncertainty with crucial implications for inflation is cyclical activity and this raises the related question of how resource utilisation is developing. Another factor is inflation expectations and a third is wage formation. I shall be considering each of these factors in turn but first let me rehearse the conditions for the price stability target.
The price stability target
Disregarding various factors that have occasioned transitory shifts, upward or downwards, in the price level and therefore had just a temporary effect on the rate of inflation, it can be said that during the 1990s both actual and expected inflation have been moving towards the price stability target. This means that we are leaving a situation in which economic policy had to focus on bringing inflation down and approaching one where long-term price movements have to be kept in line with the price stability target.
A central parameter for such a policy is the economy's capacity to produce goods and services. This capacity is not unlimited. History has taught us that when demand outstrips the economy's sustainable level of production, there is a growing risk of capacity shortages and bottlenecks that lead to rising prices and wages. If there is a risk of such a situation arising, it is better to try and check it with a small increase in the interest rate at an early stage so that demand remains in line with the increase in the economy's potential output. Growth can then be sustainable. Predictable, timely action can hopefully make it possible to avoid sharp changes in the level of interest rates and still enable economic agents to go on relying on stably low inflation.
It should be underscored that long-term growth is determined by factors such as the accumulation and utilisation of physical capital, the promotion and utilisation of people's competence and the evolution of technology. None of these fundamental growth factors can be influenced directly by adjusting the Riksbank's repo rate. But what monetary policy can do is create certainty about inflation's future path, which probably contributes indirectly to a higher growth potential. Similarly, potential output may be weakened by a monetary policy that is erratic and undisciplined.
A monetary policy focused on maintaining low inflation does not automatically rule out economic growth rates of, say, 4 or 5 per cent. But growth as strong as that does require a supply side that functions properly, without generating inflationary impulses. If the economic structure is such, on the other hand, that potential growth amounts to only 1 or 2 per cent, then demand and thereby monetary policy must be adapted accordingly.
Once the price stability target has been established as a credible benchmark, the objective of preventing inflation from either rising above or falling below the target in the longer run can be likened in the normal case to ensuring that output neither exceeds nor falls short of the economy's sustainable capacity. Monetary policy's commitment to contributing to stable, sustainable growth is aided by the construction of the price stabiity target. A fall-off in economic activity normally implies lower inflationary pressure and this can provide room for interest rate cuts. Similarly, a tightening of the monetary stance is usually called for after a period of demand growth that exceeds the increase in potential output. It is therefore wrong to claim that monetary policy makes no allowance for employment and growth. Monetary policy is so constructed that production and employment are considered as long as they do not endanger low inflation. The symmetry inherent in a price stability target means that this type of consideration is taken automatically.
As a rule, the tolerance interval should suffice to accommodate shocks whereby one-off shifts in the price level lead to temporarily higher or lower inflation. From time to time, however, shocks may be so great that inflation moves outside the tolerance interval so that the inflation target, which refers to a longer, annual perspective, is not fulfilled. As you know, that is what happened in 1996, when the annual rate of inflation was some tenths of a percentage point below the lower tolerance limit. A major explanation for this was the steep fall in interest rates, which contributed to decreased house mortgage interest costs.
In situations where inflation moves outside the tolerance interval it is up to the Riksbank both to explain what has happened and to make it credible that monetary policy is such that inflation will return by degrees to the 2 per cent target. This applies regardless of whether it is the tolerance interval's upper or lower limit that is being transgressed.
Last year's downward tendency in the CPI figures, whereby inflation dipped below the tolerance interval for the price stability target, was a temporary phenomenon. Registered inflation is now returning to the underlying rate of around 2 per cent. Thus, the registered increase was expected and, like last year's transient fall, it does not have any fundamental consequences for monetary policy. In this respect our actions have been symmetric.
Cyclical activity and resource utilisation
The inflation report we presented last week noted that the economic recovery in Sweden is continuing. As previously, the Riksbank considers that GDP growth this year can be around 2 per cent, followed by around 3 per cent in 1998 and 1999. Although this means that demand will probably be rising at a stronger rate than the supply side can match in the longer run, in the coming years there should not be any general capacity shortages. Most indicators suggest that the economy has available resources which can be activated by such an expansion of demand. It looks, moreover, as though the upswing will be divided relatively equally between export and domestic markets, which will also help to counter any inflationary bottlenecks.
In the absence of serious tensions between supply and demand, the underlying rate of inflation ought to be more or less constant for the time horizon that is relevant for monetary policy, that is, one to two years ahead. The expected upward tendency in the consumer price rise is mainly a consequence of transitory factors. This process is already at work. Registered 12-month inflation has moved up successively this year, from a low of 0 per cent in the spring to the most recent figure of 1.5 per cent. Average CPI inflation is expected to rise from around 1 per cent in 1997 to around 2 per cent in 1998 and 1999.
Because of the long time lag before monetary policy acts on demand, we have to make do with somewhat uncertain economic forecasts for the coming two years. We are not alone in working with forecasts. Many economic agents do the same. We publish inflation reports four times a year in which our assessments are checked against current tendencies and revised accordingly.
We do not know just how much the economy can produce without an increase in inflation. Output gap estimates - attempts to measure the difference between actual production and the level at which inflation is liable to accelerate - are basically too uncertain to be used as an exact pathfinder for monetary policy. One therefore also needs to look at how resource utilisation is developing in particular segments of the economy.
For manufacturing, various indicators show some increase in capacity utilisation between this year's first two quarters. In the business tendency surveys from the National Institute of Economic Research there has been some increase in the proportion of firms reporting the supply of production factors as a restriction on output. In the second quarter there was also an increase in the shortages of skilled workers and salaried technicians. Capacity utilisation has also moved up in the services sector. Moreover, delivery times became somewhat longer in the second quarter. Another sign of higher capacity utilisation is that firms planning price increases in the third quarter make up a larger proportion that those planning price cuts.
But while most things accordingly point to some increase in capacity utilisation, the main hypothesis continues to be that the supply of resources is still sufficiently good to match rising demand.
In a rapidly expanding economy there may well be temporary bottlenecks in that new facilities do not always come on stream just when demand is growing. Experience from the upswing in 1994 and 1995, which mainly came from manufacturing, admittedly suggests greater flexibility than before as regards both the utilisation of the existing capital stock and prompt investment in new machinery and plant when demand picks up. But demand for manufactural products was then so strong that capacity problems still arose and led to accelerating producer price increases. Moreover, matching problems in the labour market may cause some delays before firms manage to recruit the labour they require. The current demand will not necessarily be directed at just those industries and regions that were hit by falling demand in the early 1990s. The unchanged level of unfilled job vacancies is indicative of certain matching problems in the labour market.
The economy's complex structure makes it necessary to monitor a large variety of factors in order to form a picture of resource utilisation and judge whether there is a risk of bottleneck problems that might generate rising inflation.
The tendencies to date do not suggest that the future risks of inflation are unduly great. But we are now in an upward phase and need to be on the watch for changes in the outlook.
Inflation expectations
The importance of bringing inflation expectations down into line with the price stability target is evident from a series of economic policy decisions in the 1990s. Examples of such decisions are the formulation of the target for the Riksbank's monetary policy, the tightening of the monetary stance from the summer of 1994 and during 1995, and the broad political agreement on the status of the Riksbank. The price stability target has also influenced the process of consolidating the government budget, though this process has been directed by other considerations. The emphasis on price stability has also yielded results.
Inflation expectations, as measured by survey data, have fallen markedly in recent years. This is mirrored in the lower bond rate as well as in the stronger exchange rate. The present situation is in marked contrast to the 1970s and ¨80s, when annual inflation averaged 8 per cent.
We have made good progress along the road to establishing confidence in the low inflation policy. It is essential, however, to keep a close eye on how inflation expectations develop as activity goes on rising. There must be no doubts about the important function of the price stability target in general economic policy or about the Riksbank's commitment to this target.
In a setting where inflation expectations are firmly anchored to the inflation target, monetary policy has some room in which to manoeuvre and can support an economic recovery. This is evident not least in countries were low inflation has been leaving its mark for some time. It may seem paradoxical that the greater a central bank's credibility in combating inflation, the more room it will have for supporting general economic policy. But it has to happen in that order: credibility first and then room to manoeuvre - not the other way round.
It was this that prompted the Riksbank not to lower the instrumental rate before January 1996 - we wanted to have a safety margin and underscore the commitment to low inflation. I hope that commentators who did not then see monetary policy from this angle do so today. It is by promoting price stability, low inflation expectations and thereby low market interest rates over the whole yield curve that monetary policy can contribute to growth that is sustainable - I should like to emphasise sustainable. Without the credibility represented by inflation expectations in line with the target, monetary stimuli are liable to cause the economy to revert to the misery of inflation, with negative consequences for production and employment. While a marginal deviation from 2 per cent inflation would not necessarily be disastrous for the workings of the economy, it would undermine the price stability target's function as a clear foundation both for economic policy and for economic agents. It suffices to recall the early 1970s as an illustration of all the economic consequences this would have. The collapse of the Bretton Woods system, of which Sweden was a member, marked the start of a long period of high and fluctuating inflation. That is why the inflation target must be carefully safeguarded. It serves as a clear reference point and gives decision-makers in the economy a distinct message about the direction of economic policy.
It was not until the early 1990s that the establishment of a low inflation regime began in Sweden. We must therefore be particularly on our guard lest the direction of policy contributes to rising inflation expectations. The risk of this has probably diminished in recent years, both because inflation has been low and in that the central government has demonstrated in deeds that price stability is a central pillar of economic policy. But watchfulness is the order of the day, particularly in that monetary policy is now exercising a stimulatory effect. Inflation expectations in line with the target contribute to stable conditions for all the contracts that are a daily feature of a dynamic economy. This applies not least in the labour market, as we shall see in the run up to a very important round of wage negotiations.
Wage formation
Wage formation is naturally affected to an essential degree by the expected level of future inflation. The comparatively large wage increases in recent years, despite high unemployment, are probably explained to a considerable extent by the circumstance that when the agreements were concluded, inflation was expected to be substantially higher than it proved to be. The level of the wage increases accordingly precluded a reduction of unemployment.
Besides inflation expectations, a major factor for wage development is the labour market's institutional arrangements or, speaking more generally, how the labour market functions. It is still difficult to form a clear opinion as to whether the talks that have been held on future wage formation and the cooperation agreements that have been concluded in segments of the labour market will cause wage formation to function better in practice from 1998. What counts in the final analysis is results; an indication of the problems that exist in wage formation is provided by the level of unemployment.
The room for future wage increases is highly contingent on productivity. In the 1990s labour productivity has been rising at a considerably higher average rate than in the 1980s. It must be asked, however, whether this is a lasting shift that can serve as a basis for future wage formation or whether it has explanations of a more temporary nature. While it does seem probable that economic changes in Sweden in recent years will yield better productivity growth in the future, it seems premature to draw far-reaching conclusions about a break in productivity's structural trend. Part of the higher productivity growth in the 1990s does in fact appear to stem from temporary factors. The cost crisis and the steep fall in production in the early part of this decade gave rise to widespread rationalisation in the business sector. As a result, the number of persons in employment and total working time both fell more than output in those years and this showed up as increased labour productivity. Another transitory factor that may have contributed to the high registered productivity in recent years is that when demand is rising, the initial increase can normally be met with a firm's existing work force. Historically, upswings have been associated with marked improvements in productivity but such variations usually average out over the cycle a a whole. It is also conceivable that the high wage increases in recent years have occasioned rationalisation with a view to countering increased costs. Neither can such effects on productivity be used as an argument for general wage increases, particularly if the aim is to bring unemployment down.
Future wage formation is thus another risk factor which the Riksbank has to take into account. There are encouraging signs which suggest that wage formation is moving in the right direction but a lasting reduction of unemployment is also required before one can say that wage formation is no longer a problem.
Conclusion
The Riksbank considers that at present the monetary stance is well balanced. With a monetary policy focused on a price stability target, the need for interest rate adjustments is determined primarily by domestic economic tendencies and the assessment of future inflation in Sweden. Thus we are not directly affected by whether interest rates are raised or lowered in other countries. This was apparent earlier when other countries lowered their interest rates and the Riksbank continued with a higher level; it is also clear at present, when the instrumental rate has been raised in Finland and we leave the rate unchanged.
Economic conditions in Sweden are good for establishing a phase of sustainably high growth, based on macroeconomic stability. For this to be feasible we must be alert to changes that may lead to rising inflation. This is particularly important now that growth in the coming years is expected to be above the trend. In this situation some of the risks that may lie ahead must be discussed even though the picture, including the present assessment of these risks, appears to be favourable.
If the arrangement is such that it is up to us at the Riksbank to identify potential risks, everyone else can hopefully concentrate on getting the best out of the stable economic development that price stability offers. It is several decades since the economic conditions for Sweden have been as favourable as they are at present. Provided the labour market functions properly, in years ahead this may also lead to a lasting reduction of unemployment. The coming upswing will show whether this is to be the case.
