Ingves: Introduction on monetary policy
Thank you for the invitation to come here! I came here for the first time one year ago, as newly- appointed Governor of the Riksbank. A year passes quickly, and now it is once again time for me to summarise the past year and at the same time report on how the Executive Board of the Riksbank views the prospects for the economy and inflation.
Before I begin, I would like to welcome the fact that the Committee has decided to hold this hearing at a later date in relation to the publication of the Riksbank's report. Of course, this means I risk having to answer even sharper questions than usual, but it’s all in the line of duty. It is also important that these hearings are held under the best possible conditions.
My most important message today is:
- We have now begun to publish our own forecast for the repo rate. This means that we have taken a further step towards increased clarity and openness in monetary policy.
- Inflation rose last year, but was still below the target. One of the main reasons is the substantial productivity growth. Understanding which mechanisms lie behind the strong productivity has been, and is, an important task for the Riksbank.
- The economic prospects are bright. Economic activity is strong and resource utilisation is increasing. Given this, we raised the repo rate by 0.25 percentage points last week.
- Our assessment is that the repo rate needs to be raised by a further 0.25 percentage points over the coming six months. After that we can probably pause a while before it is time for another increase.
Changes in the Riksbank’s forecasting methods
Let me begin by saying something about a couple of new elements we have introduced in connection with the latest forecasting work.
As you have probably already noticed, one change is that we are now for the first time basing our forecasts for inflation and the real economy on our own assessment of what constitutes a reasonable future development of the repo rate. Previously, we based our forecasts on the assumption that the repo rate would develop in line with the expectations that could be read from the interest rates in the financial markets. Prior to this, our forecasts were based on the assumption that the repo rate would remain unchanged throughout the forecast period.
The biggest advantage in using our own forecasts for the development of the repo rate is that it will then be clear what monetary policy the Riksbank considers reasonable in the future. This change is above all a further step towards increased clarity in monetary policy. I don’t believe it is possible to be clearer than this with regard to our intentions for the repo rate.
I also believe that the monetary policy discussion gains by the Riksbank publishing its own forecast for the interest rate. It will enrich the debate, which will concern more the future development of the interest rate and not merely its current level. This has been clear over the past week. Our current decision to raise the interest rate by a further 0.25 percentage points came under the shadow of a discussion on the level of the repo rate in the longer term. Since the long-term development of the repo rate is much more important to most households and companies than an individual interest rate adjustment, this is a positive sign.
One thing it is important to emphasise is that we are talking about a forecast for the repo rate. This is the repo rate development that currently appears most likely given the information available. We are not making any promises. The fact that the Riksbank is presenting its own interest rate path does not mean that we are laying down a policy that we will commit ourselves to following. Previously, we did not commit ourselves to following the interest rate path based on market expectations. The reason is that the conditions are constantly changing. Various shocks occur that mean we have to regularly reconsider what would be an appropriate monetary policy. Adhering to one and the same interest rate path, despite conditions changing, is about as bad as always driving a car at the same speed, regardless of whether the road is dry or treacherously icy.
It is not particularly strange that the appropriate path for the repo rate should vary over time, but it is still worth emphasising this. It is sometimes raised as a potential problem that when a central bank presents its own policy rate forecast, it can be seen as a failure if the central bank does not later follow the path that it had forecast. However, I am quite convinced that the Swedish monetary policy discussion has now reached such a stage that this will not be a problem. There is currently widespread understanding that the assessments are uncertain and that new information may change the conditions for the monetary policy decisions quite substantially.
Others have pointed to a risk that the information value of setting interest rates in the market may deteriorate if the central bank publishes its own interest rate forecast. The reason is said to be that market agents would then no longer seek other information that has significance for long-term rates.
But I believe that this is also a rather exaggerated fear. The financial market agents know that we live in an uncertain world and will continue to make their own assessments of interest rate developments. I am therefore not particularly surprised that interest rates with a somewhat longer duration have only partially adjusted to the path we published last week. It is natural that from time to time some agents will have a different view of what would be an appropriate repo rate in the somewhat longer term than the Riksbank has, and this is also reflected in the setting of the interest rates.
I mentioned that we have made two changes in connection with the latest forecasts. The second change is that we have designed our report in a slightly different way. Our hope is that the new design will make the report more readable. The report has also changed its name and is now called the "Monetary Policy Report”. The reason for this is that the report will now contain monetary policy conclusions, as we are presenting our own forecast for the repo rate. There will be a clear link between the forecasts and the monetary policy considerations directly in the report. The idea is that the report will not only show how we think the repo rate should develop in future, but also as clearly as possible describe why we think the repo rate path we have chosen is the most appropriate. I will return shortly to our reasoning with regard to an appropriate interest rate path on this occasion. However, before doing so, I intend to look back at the past year. This is what we usually do at the first hearing of the year.
High productivity growth contributed to low inflation in 2006
Inflation rose last year, but was nevertheless below the 2 per cent target. CPI inflation rose by an average of 1.4 per cent. Underlying inflation, UND1X, was 1.2 per cent.
A comparison with our forecasts shows that inflation was lower in 2006 than we had anticipated in 2004. On the other hand we succeeded in predicting last year’s inflation in the forecasts we made in 2005.
One reason why inflation has been low in recent years is that prices of imported goods have developed weakly. This is probably linked to stiff international price competitive pressure. Another important reason is that productivity has risen faster than the Riksbank and other forecasters had expected. The unexpectedly high productivity growth has meant that companies have been able to increase production at a rapid rate without costs per unit produced rising to any great extent.
Understanding which mechanisms and changes in the economy that lie behind the strong productivity has been, and is, an important task for the Riksbank. Other forecasters have also assumed that companies’ production conditions should have been affected by the various structural changes we have seen acting in the economy over many years now. I am thinking here of phenomena such as globalisation and the rapid development of information technology. Most analysts probably agree that these changes have contributed to increased competitive pressure and that they have enabled production to be made more efficient in several ways. We have also been able to note that productivity growth has shown an increase trend over the past decade. This has confirmed our opinion that structural factors in the economy lie behind the strong productivity growth.
It is one thing to observe structural changes in the economy and have an opinion on the direction in which they will have an effect. It is another thing entirely to be able to predict with any accuracy how large and how lasting the impact will be. One of the conditions that must be taken into account is that productivity growth usually varies over the business cycle. At the beginning of the business cycle, for instance, productivity growth is usually quite high. Part of the strong productivity in recent years has probably therefore been cyclical. But exactly how much is difficult to say.
At the same time, the assessment of productivity growth is an important building block in the Riksbank’s inflation forecasts. Gaining a deeper knowledge of the driving forces behind productivity growth will therefore remain a high priority area for the Riksbank. The Bank already has a project aimed at improving understanding of the correlation between, among other things, IT investments and productivity growth. We also intend to invite various researchers to an in-depth discussion on the driving forces behind productivity in Sweden and abroad. Hopefully, this will give us some new insights into productivity development.
As a complement to our statistical studies of productivity and other subjects, we will also create channels for direct contacts with companies. This is to improve our knowledge of the current status and current events in the economy.
At the same time, I would like to point out that one should not have overly high expectations that the work will lead to significantly more accurate forecasts. We will probably have to continue to live with some surprises in the economy. This applies in particular when structural changes in the economy change the historical pattern.
But improving understanding of what has happened is still important. The cause behind fluctuations in inflation and growth is important for how monetary policy should be designed.
The high productivity growth has meant that the Riksbank has been able to conduct an expansionary monetary policy. In other words, the economy has managed a high growth rate without prices rising much. Under these circumstances a natural question might be whether we should have cut the rate more in order to push up inflation towards the target more quickly?
With hindsight it is fairly easy to note that we could possibly have made the interest rate cuts at a slightly faster pace when we began making cuts a few years ago. We have already noted this in various contexts and it is hardly controversial.
On the other hand, I do not believe that it would have been wise to cut the rate much more in recent years. In 2005 the interest rate had fallen to a historically low level of 1.5 per cent. Although the Riksbank predicted in 2005 that inflation would remain low in 2006, albeit rising slightly, we were counting on strong growth in demand, partly due to the expansionary monetary policy. The low interest rates had also led to a rapid rise in household borrowing and house prices. If we had tried in that situation to push up inflation more quickly by means of further interest rat cuts, I believe there was a risk that the economy would have overheated later on, with excessively high price and wage increases and other imbalances. In that case we would have been forced to tighten monetary policy much more than we now envisage. Ultimately, we would have risked an unstable economic development.
Monetary policy’s ability to stabilise the economy
Let me also briefly mention in this context the question of monetary policy’s ability to stabilise the economy in general.
By considering how quickly we should aim to bring inflation back on target, we try to contribute to the most stable development of production and employment possible. But the fact that we give consideration to the real economy when formulating monetary policy does not mean that we have any targets for employment or production. There are several reasons for this.
Firstly, it could probably easily lead to unrealistic expectations of what we can achieve by raising and cutting the interest rate. Monetary policy cannot in the long run affect either production or employment. Well-judged adjustments in the interest rate can ensure a stable development of inflation in the long term. In addition, monetary policy can, at best, contribute to alleviating cyclical fluctuations in production and unemployment. However, the interest rate is much too blunt an instrument to be able to fine-tune the economy to any great extent.
Secondly, a production target or employment target would only be purposeful if we were fairly sure how large the growth potential of the economy was and what level of employment represented some form of equilibrium. In practice, it is not possible to determine this with any great precision. Both potential growth and employment vary over time with structural changes in the economy in ways that are difficult to quantify.
Thirdly, an extra difficulty in this context is that the picture of how the economy develops changes over time as statistics are revised. This applies in particular to GDP statistics. It becomes particularly clear how difficult it would be for the Riksbank to have a target for productivity growth if one looks at how the statistics were revised at the end of 2004 and beginning of 2005.
The information available at the beginning of 2005 indicated a fairly clear slowdown in production growth. Later on in 2005 the picture of a slowdown in the growth of the economy was confirmed, but it then appeared to have a more prolonged course of events. Since then the picture has been revised again, and now the economic downturn around the end of 2004/beginning of 2005 appears much milder than was indicated by the statistics at that time. Today we have a new description of history to base our own assessments on.
This type of statistical measuring problem was one of the things that made Francesco Giavazzi and Frederic Mishkin advice against having targets for production or employment in their evaluation of Swedish monetary policy. I agree with them. The best we can do is to balance monetary policy so that the economy follows what is roughly considered to be a long-term favourable path. This balance should of course be within the scope of our inflation target.
The current situation and monetary policy considerations
This seems like a good point to move on to talk about the Riksbank’s current views on inflation prospects.
The economy has grown rapidly in recent years. Over the coming years we are expecting growth in the economy to slow down and approach a growth rate that is more sustainable in the long term. One reason is that our assessment is that international economic activity will slow down; another is that the productivity increases will slow down.
Over the coming year, GDP growth nevertheless looks to be relatively high. Employment will rise and households will increase their consumption at a rapid rate.
We are assuming that employment will continue to rise during the forecast period, albeit at a slower rate. This will lead to unemployment falling. When we make forecasts for the labour market, we also have to take into account the fact that the labour supply will increase when economic activity strengthens. This is a normal pattern. The Government’s policy will probably also mean that the number of persons in the labour force increases slightly more than has been usual in previous economic upturns. The fact that the labour supply increases will mean that unemployment will fall at a slower rate than employment rises.
With regard to inflation, we expect it will fall over the coming six months. It is the large fluctuations in energy prices that make an impact on inflation in the short term. We cannot – and should not – do much about this type of short-term fluctuation in inflation. In the longer term we are expecting inflation to rise and be close to 2 per cent two years ahead.
One might wonder, if inflation is now expected to fall in the short term, what reason do we have to believe that it will rise to 2 per cent in the longer term?
One important reason for this forecast is that productivity growth should be more modest in future. Over the past two years we have been able to note that productivity growth has slowed down compared with the years before that. This indicates that productivity growth is about to slacken to a more normal level, seen from a long-term perspective.
As employment rises and the lack of labour increases, it is also reasonable to assume that the rate of wage increase will be gradually raised. The Riksbank’s forecast assumes that wage increases will grow as resource utilisation in the labour market rises.
A rising rate of wage increase and a stabilisation of productivity increases at a more normal level will mean that cost pressures in the economy increase in the coming years. With more rapid cost increases, price increases will also be greater in the long term.
There is thus good reason to believe that inflation will increase during the forecast period. At the same time, what we are expecting is a modest upturn in inflation, despite a strong upturn in employment and relatively high GDP growth. This is due to continuing favourable supply conditions. For one thing, productivity growth is expected to remain at a relatively high level, and for another thing the labour supply will probably increase more than during previous economic upturns. A larger supply of labour will hold back the pressure for higher wage increases. Moreover, it appears that international price pressures will remain high. These hold back price increases on imported goods.
But I would also like to emphasise that the inflation forecast is uncertain. This is illustrated by the uncertainty bands around the inflation forecast that is included in our report. The uncertainty bands are based on the Riksbank’s historical forecasting errors. Their breadth indicates that it is not particularly meaningful to focus on tenths of percentage points in inflation.
But the uncertainty bands also show that there is a large probability that inflation will remain within the tolerance interval of 1-3 per cent over the coming years. It is also worth reminding that inflation seen over a longer period of time has not on average deviated significantly from the target.
In our current assessment the rate of wage increase is particularly uncertain. On the one hand, the fact that many agreements in the labour market will be renegotiated in a favourable labour market situation entails a risk that wages will increase more than we had assumed in our main scenario. On the other hand, it is also possible that we are underestimating how much the labour supply may increase. This means there is also a downside risk for wages. As before, we cannot ignore that there are risks linked to the rapid increase in house prices and loans. Imbalances could build up. When imbalances sooner or later have to be corrected, this can lead to undesirable fluctuations in inflation and the real economy.
At our most recent monetary policy meeting we weighed these risks and considerations against one another. This resulted in the assessment that the repo rate needed to be raised by a further 0.25 percentage points and that it would then need to be raised once more over the coming six months. After that we considered it would be probably possible to pause before making a further increase.
In our Monetary Policy Report we highlight two other possible interest rate paths to show the considerations that need to be made. One of these involves the interest rate being raised at a slower pace and the other a more rapid pace than in the main scenario. With our current knowledge of the state of the economy and inflationary pressures, our assessment is that faster interest rate increases would lead to excessively high inflation. Correspondingly, slower interest rate increases would lead to excessively low inflation.
Let me point out in conclusion that the inflation assessment we make now is not particularly different from the one we made in October, when we published our previous report. From the comments in the mass media, one might get the impression that the Riksbank had made a drastic change in its views of inflationary pressures in the economy. But this is not the case. On the other hand, we now believe that growth will be a little higher and employment growth will be stronger than we forecast in October. At the same time, it appears that the labour supply will increase a little more than we assumed then. Our inflation forecast therefore remains largely the same.
Our assessment of the interest rate has therefore also changed very little. In October we based our inflation forecast on the market’s expectations of the development of the repo rate. The forecast for the repo rate that we presented last week agrees fairly well with the market expectations in October.
But this does not mean that we will not need to revise our assessment in the future. I would once again like to emphasise that the path for the interest rate is a forecast and therefore uncertain. It is not a promise. But nevertheless it is our best current assessment of how the repo rate will need to be adjusted in the near future. It is reasonable that we report it in a similar manner to the way we talk about what we think will happen with production growth and the labour market and a number of other variables.
I have today emphasised that our decision to publish our own forecast for the interest rate should be regarded as a further step towards greater openness and clarity in monetary policy.
When looking back, I observed that inflation in 2006 was held down by, among other things, the strong development in productivity. This has contributed to a high GDP growth without any significant rise in prices. Gaining a deeper understanding of the driving forces behind productivity growth is a difficult task. It is also an important one, to which we devote considerable effort.
I have also described the background to our decision to raise the interest rate by 0.25 percentage points and to our assessment that a further increase is necessary over the coming six months before we probably pause before making another increase. Continued good economic activity and an increase in employment will lead to inflation rising in the coming years. But we are nevertheless assuming that the inflation upturn will be modest. This is due to continued favourable supply conditions in the economy. Finally, I have pointed out that our forecast, as always, is uncertain. This applies equally to our interest rate forecast.