Sweden is now entering its third year with growth of over 3 per cent. This year, GDP is expected to increase by over 4 per cent and our assessments indicate continued annual growth rates averaging over 3 per cent for the next two years. Despite the strength of the economy, inflation is low. Why doesn’t inflation rise more? What balances do we have to strike as monetary policy decision makers? These are the issues that I want to give my view on in today’s speech.
Monetary policy strategy
I would like to start by drawing attention to what guides us when we make monetary policy decisions, i.e. how we have formulated the monetary policy strategy.
The new Riksbank Act, that came into force on 1 January 1999, confirmed by law that the objective of monetary policy is price stability. This objective was formulated in 1993 as an inflation target, where CPI is to total 2 per cent with a tolerance interval of plus/minus one percentage point. One reason for adopting CPI was because it is well-known and measures the prices consumers encounter for a large part of their consumption. Two per cent was chosen, among other reasons, because it was considered to be at a good European level at that time. When the new Executive Board of the Riksbank took office at the beginning of 1999, we found no reason to change the target formulation.
To make it possible to evaluate how we carry out our task, we endeavour to be as clear as possible about the grounds on which monetary policy decisions are made. Inflation assessments and interest rate decisions are explained in the Riksbank’s inflation reports.
The Riksbank has also stated a rule of conduct for what should normally lead to decisions on interest rate rises, and the justifications for departures from this rule of conduct. Finally, the published minutes of the Executive Board’s monetary policy meetings contain an account of how we voted and the reasoning underlying the decisions.
Resource use is important for inflation
Since the interest rate affects inflation with a time lag, monetary policy decisions are based on forward-looking assessment of inflation.
Highly simplified, the conceptual framework, used by the Riksbank in the appraisals of price tendencies, can be described as follows: When there is unused capacity in the economy, price pressure is low and inflation is under the target. When demand is sufficiently high to lead to output exceeding the level that coincides with the long-term trend level of production, inflation instead tends to exceed the target. Inflation will be on target when output is at the level of the long-term trend level of production – the output gap is closed. This reasoning assumes that inflation expectations are close to the target for monetary policy.
Inflation expectations are based therefore on forecasts for the development of the real demand, how many unused resources there are in the economy initially, assessments of the productivity trend and the long-term trend level of production.
However, consumer prices can be affected by factors that are not related to resource use in the Swedish economy. Inflation measured by CPI is also affected, for instance, by changes in supply and other factors that temporarily affect inflation. When CPI is affected by temporary factors of this kind that are unrelated to resource use, inflation diverges, in the simplified model, from the target even when the output gap is closed. UND1X is a measure that makes adjustments for some of these effects although far from all of them.
The rule of conduct
The Executive Board of the Riksbank has decided to formulate a decision rule that states that if inflation, measured as the annual change in CPI, is expected to rise about the target of 2 per cent during the nearest one to two years, then the interest rate should normally be increased and vice versa. The choice of time horizon is justified by our assessment that an interest rate change has the greatest effect on price developments during this period. However, interest rate changes also affect inflation in the shorter and longer term, so that there are reasons for giving some consideration to the inflation forecast for the coming year and for the time beyond two years.
However, a monetary policy that aims in all situations to keep inflation at 2 per cent can lead to large fluctuations in output. If monetary policy was to counteract price increases in CPI which are deemed to be temporary and not a consequence of excess demand, interest rates must be kept sufficiently high to hold down demand in the economy. Output would then be under the long-term trend level. Correspondingly, with the target of achieving 2 per cent inflation, monetary policy must stimulate demand and press output up above the long-term trend level if inflation was held down by factors that temporarily dampen prices.
There are therefore reasons for modifying the monetary policy rule of conduct to make it possible to achieve balance between variations in inflation and in growth. Excessively large fluctuations in output can entail high economic costs. Fundamentally, after all, the reason for monetary policy being given the task of maintaining stable prices is this is the way in which it is considered to make the best contribution to creating the prerequisites for high, stable growth.
If inflation is affected by temporary factors or if the economy is affected by such extensive disturbances that it can be difficult to bring inflation back on target within one to two years, there may therefore be reasons to bring inflation back to target more slowly than usually. However, we should give reasons for doing this.
Scope for major differences, despite a common strategy
Even if there is agreement on the strategy for monetary policy, there is, of course, considerable scope for differences in assessments of where the interest rate should be in each situation.
In the first place,
the Executive Board can make different assessments of the most probable price tendency. As we know, economics is not an exact science and forecasts of prices and growth potential always rest on many assessments of an enormous mass of information and correlations. Models can never capture the complexity of the economy and how people’s economic decisions interact and must therefore always be regarded as one of many aids in assessments.
The Riksbank uses a lot of different approaches and indicators to make assessments. The focus of the Riksbank’s analysis has been until now on the real economy and how it affects price trends. In many areas, we have come further than other central banks, while we have not come so far as regards the analysis of different financial aggregates and their linkage with inflation and real demand. An endeavour has certainly been made to take into account information provided by the development of the quantity of money and growth of credit. However, we have not for instance developed methods that can give guidance as to the level of indebtedness and the quantity of money that is compatible with the long-term sustainable growth of the economy on the basis of the inflation target. While the financial economy is in principle only a reflection of the real economy and vice versa, changes can take place as to which component is in the driving seat. A well-developed analysis of both sides would probably improve the basis for decision-making in monetary policy. Structural changes in the financial markets make it difficult to interpret developments and to find stable correlations between, for instance, the quantity of money and inflation, although this is true for many parts of the economy. It applies, for instance, to methods for estimating such central variables as how many unused resources there are in the economy and how great the potential growth can be. Different methods then produce different results.
The ECB is an example of a central bank that builds up its analysis from the real side and from the financial side – monetary policy rests on two pillars. There are partly historical reasons for monetary policy strategy being conducted in this way. Their strategy is perhaps not the best way to take financial trends into consideration. The risk with such a clear division is that the fact that it is actually about mirror images is disregarded and that the consistency between the real and financial analysis is lost. The advantage is that financial trends can clearly be taken into consideration when monetary policy decisions are made.
In the second place,
we can make different appraisals of the risk picture for the price trend during the forecast period. As there is normally great uncertainty in the inflation assessment of the main scenario, the Riksbank also produces risk scenarios we also publish estimates of the probability distribution of different inflation outcomes. The distribution around the probable outcome is taken into consideration when decisions on the interest rate are taken.
In the third place,
we can make different assessments as to how monetary policy is to be conducted to strike a reasonable balance between fluctuations in growth and inflation. As I mentioned earlier, the Executive Board has agreed that we can decide not to bring inflation back to target within one to two years if CPI is affected by temporary factors or if the economy is affected by extensive disturbances. Making assessments and evaluating when – within these frameworks – there are sufficient reasons for avoiding output variations by extending the control horizon is far from self-evident, however.
To maintain the credibility of monetary policy aimed at price stability, it is important therefore to justify such departures.
In the fourth place
– and I am now starting to sound as if it not at all possible to conduct monetary policy with a demand for clarity –we do not know exactly the stimulation or tightening produced by a particular interest rate. This depends on our, like everybody else’s, limited knowledge of the "transmission mechanism". It is also due to it being difficult to exactly define at what real interest rate, monetary policy has a neutral effect on demand. The difficulty is increased by the "neutral" real interest rate changing, for instance when the productivity growth in the economy changes. On the basis of experience, simple rules of thumb and rough calculations, everyone can per se obtain an idea of how monetary policy operates. However, this is another area that needs to be developed.
Altogether, this review shows that monetary policy decisions can never be made mechanically with the aid of simple rules of thumb. In order to firm up the analysis and make it easier to describe the basis for our decisions, it is notwithstanding this important to describe and keep to a conceptual framework, around which we construct our analysis. As we gain increased knowledge, our analysis and conceptual framework are complemented or modified. In order not to lose clarity, we endeavour in our inflation reports, speeches and publications to describe our analytical apparatus and its development.
Challenges in the present situation
The growth prospects for the Swedish economy have improved considerably in recent years. The Riksbank has been successful in creating support for and making the inflation target credible, central government finances have improved and many structural reforms have been carried out, although much remains to be done.
However, although for the most part, it looks good, there is a risk that the Swedish economy is now growing too rapidly due to over-optimistic assessments of growth prospects. This must in my view be taken into consideration in the conduct of monetary policy. Excessively optimistic expectations of the future could lead to financial imbalances with excessive indebtedness building up in the Swedish economy. High stock exchange prices and increasing property prices can intensify this tendency. In this situation, an expansive monetary policy, based on the same over-optimism, would contribute to imbalances being built up.
However, an important issue for the Riksbank is whether such a scenario is possible without different indicators for inflation signalling in time that overheating is on the way so that monetary policy has time to react. Experiences from other countries show that low consumer price inflation and healthy central government finances do not by themselves constitute a guarantee for future good development. Although one should always be cautious about drawing conclusions from comparisons between countries, it has been seen, for instance, that experiences from Japan at the end of the 1980s and from Korea in the mid 1990s that overheating can be expressed in exaggerated stock exchange and property prices, high lending rates and low yield investments rather than in rising consumer prices.
I can conceive of at least two factors in the Swedish economy that lead to an increasing overheating not being expressed in rising consumer prices and therefore risk leading the Riksbank and the general public to draw the wrong conclusions about the state of the economy:
In the first place
I believe that both wage and pricing behaviour are affected by inflation expectations being low and stable. It takes time before an increased use of resources has an effect in price and pay increases! In the same way as it takes time to adjust inflation expectations downwards, there is inertia in shifting them upwards. This would mean that even if demand and production are higher than what is sustainable in the long-term, that this is not marked in price development before the use of resources has come under great strain.
In the second place deregulations and increased competition can lead to consumer prices being dampened as a result of excess profits being forced down in previously sheltered industries. Initially, increased competition will perhaps primarily lead to redistribution of real incomes. In the somewhat longer term, productivity will also be improved. However, it is uncertain how quickly and to what extent potential growth will increase as a result of increased competition.
If monetary policy assessments are based on an overoptimistic view of growth prospects and developments in consumer prices do not give any signals about this, this can lead to considerable imbalances being building up.
Sooner or later, however, inflation will accelerate, however. When the scope for reduced profit margins as a result of deregulations and increased competition begins to be exhausted and demand pressure finally begins to be expressed in rising prices, there is a risk that price increases will be large and come quickly. Expectations of future growth prospects can then be abruptly interrupted. The consequences can be that demand drops if companies and households are then too highly indebted. If a lot of companies during the overheating period, have based their investment decisions on incorrect expectations of the demand trend, the fall can be great.
The example shows a development where monetary policy by acting too late creates high variability in both inflation and output. The effects need not either be visible until after the forecast horizon. The economic costs of waiting too long to change the direction of monetary policy can be high if imbalances have grown great and the fall in production is large.
The same course of events as in U.S.A.?
Although the prospects for the Swedish economy look very bright, we cannot therefore disregard the risk of overoptimism. History has shown many times that an optimism that was well-founded to start with has become self-reinforcing and has ended up by creating considerable imbalances in the economy. Developments in Japan in the 1980s and in Asia in the 1990s are clear examples of this.
The exceptionally long upswing in the USA has contributed to strengthening expectations of faster long-term growth in Sweden as well. It is therefore important to understand whether an American level of growth is sustainable in the long-term and whether anything similar can take place in Sweden. An important basis for the favourable development in the USA is probably the focus on budget consolidation and price stability that has guided economic policy there for a long time. We have had a similar development in these areas in Sweden, although the process started later here. At the same time, there are similarities with regard to the strong penetration of information technology in the economy, although we are also later in this development. As in the USA, high asset prices and strong optimism, have stimulated demand and increased the level of indebtedness in the Swedish economy. Although it is not yet possible to write off the risk of financial imbalances building up the USA – imbalances that can have negative effects for the American economy –it now seems as if the American economy may be on its way to a soft landing. However, there still remain the global financial imbalances that have been built up under many years with very high saving in Japan and very low saving and high indebtedness in the USA. We have still not seen how these imbalances are to be resolved and the consequences this will have for the world economy.
Although there are some indications that the Swedish economy is experiencing a similar development as the American, the structural differences cannot be disregarded, which probably create greater inertia in the Swedish economy than in the American. One such area is the labour market. Considerable barriers to competition remain in the Swedish economy as well. Another difference is that the cyclical upswing we are now experiencing is taking place at the same time as an increase in growth and resource use in the rest of the world. We cannot therefore benefit in the same way as the USA from there being unused resources in the rest of the world.
Conclusions
The Riksbank makes the assessment now that GDP growth will be over 4 per cent this year and average over 3 per cent during the next two years. This rise is taking place after a number of years with considerable growth. Despite this, inflation is low and is expected to remain so for the whole forecast period. We cannot with certainty determine the processes concealed behind the weak increase in prices. I drew the conclusion during the spring that we had probably underestimated the output gap, i.e. the amount of unused resources available in the economy initially and this is probably an explanation for the price increases being surprisingly low. Some productivity improvements and indications of shrinking profit margins in the economy as a whole also indicate that deregulations and increasing competition play an important role. I furthermore consider that the stabilisation-focused macropolicy and measures such as the tax reform have moreover increased the potential growth in the Swedish economy.
The strong growth that the Riksbank – assuming an unchanged repo rate – calculates with, is slightly too low in my view. At the same time, I have a somewhat less optimistic appraisal than the majority of the mode of functioning of the labour market and competition in a number of markets in Sweden. However, my assessment is that the low inflationary expectations are making it take a longer time before overheating tendencies are expressed in increasing consumer prices. This leads me to make the assessment that prices will not start to rise until the end of the forecast period but that price increases will then be stronger than the majority believes. My assessment is therefore that inflation at the end of the forecast period, in my main scenario, will end up marginally over the forecast that was reported in the last inflation report, which the Executive Board majority did not find reason to revise.
My assessments of developments during the next two years differ then only marginally from the majority view. I am more concerned about the price tendency and the risks I see in developments after the forecast horizon. The assessments of the use of resources and future production possibilities are uncertain. Structural changes of the Swedish as well as the international markets make it particularly difficult to draw conclusions about future growth prospects. There is a risk that we go to excess. High, increasing indebtedness calculated as a share of GDP for both households and companies make me concerned for financial imbalances that can be on their way to being created, although our knowledge of what can be assumed to be long-term sustainable development is small.
In my view, there is a risk that we overestimate potential output in the coming years and that we therefore accept an excessively high level of demand in the economy. Low inflation expectations and the effects of deregulations and increased competition can lead to the excessively strong demand not producing an effect on CPI until a late stage but primarily in other parts of the economy. The economic consequences that can ensue from an incorrect assessment can be high. Although the probability of a dramatic development is small, the high costs that can arise if this comes about must in my view be taken into account when making monetary policy decisions. Caution indicates that monetary policy should at least be neutral in the present situation. I regard monetary policy as being expansive at present. I therefore consider that the interest rate should be increased.
