Heikensten: Introduction on monetary policy
Let me begin by thanking you for your invitation to the Committee on Finance.
We have now been working with inflation targeting for just over ten years in Sweden. We have had open hearings in their current form since 1994. The hearings are one of the cornerstones of the Riksbank’s strategy regarding openness and are an important instrument with which to win democratic legitimacy for monetary policy.
Today is the last time I will have the opportunity to appear before the Committee in my capacity as Riksbank Governor. Against that background I thought I would begin with a look back at the last ten years. What have we achieved during these years? What has gone well and less well? I also intend to discuss a couple of issues concerning the formulation of the inflation targeting policy; what expectations are reasonable to have? How is policy being conducted and how can it be developed? These are relevant questions, not least given the discussion that we have seen recently. Finally, I intend to speak about the outlook for the economy and inflation that is presented in the Inflation Report and my own view on future monetary policy.
Developments in the past ten years
As I’m sure you all remember, our country was hit hard by a deep economic crisis at the start of the 1990s. This crisis could generally be described as a dramatic end to almost twenty years of stabilisation policy problems. The Swedish economy had been characterised by an unsustainable rise in costs, which had originated in price and wage increases that constantly came on a collision course with the fixed exchange rate. The result was repeated devaluations, modest economic growth, poor productivity growth and more or less stagnant growth in real wages. The performance was markedly weak both compared with earlier periods and in relation to other countries. During the crisis years the situation was aggravated further. Unemployment increased fourfold in the course of a few years and the central government finances deteriorated dramatically. Interest rates rose unchecked and the interest rate differential vis-à-vis Germany occasionally came to several percentage points.
Favourable results
Today, a little more than ten years after the crisis, the situation is completely different. The new focus of stabilisation policy – which involved crucial changes in the framework for both fiscal and monetary policy – has meant that inflation now remains relatively stable around the target of two per cent (Figure 1). The lower inflation rate has not been associated with weaker growth in output, rather the opposite. GDP growth has averaged roughly half a percentage point higher a year in the past decade than in the two previous decades, and the fluctuations in the real economy also appear to have lessened (Figure 2). Productivity growth has been surprisingly robust – stronger than in the rest of the EU – and there now seems to be broad consensus that the long-term rate of productivity growth, and thus the economy’s potential growth rate, has been raised. Real wages have also improved considerably; since 1995 they have risen by roughly 2.5 per cent a year on average, compared with less than 1 per cent on average during the 1970s and 1980s (Figure 3). Growth in employment has not been quite as good, but it nonetheless is worth pointing out that the situation today is far better than it was in the mid-1990s. At that time unemployment was commonly judged to have become entrenched around 8 per cent.
Those of us who experienced the crisis years know that this performance was by no means a certainty. That there was a shift in stabilisation policy is in itself perhaps not that surprising given both the seriousness of the situation at the time and the pressure imposed by an environment with free movement of capital. But the decisiveness and speed with which the changes were implemented is something that is still brought up when I speak with colleagues abroad. Compared with other countries it is perhaps above all the changes in fiscal policy that have made Sweden stand out, in spite of the fact that the limits in the fiscal policy framework have been stretched somewhat in recent years – inflation targeting is, after all, now conducted in some twenty countries around the world and there also are other reasons than inflation targeting that have contributed to lower price increases globally. The comparatively sound central government finances have been a great strength for our country over the past ten years. In the years ahead demands will rise even further if we are to meet the challenges posed by an ageing population.
More surprising, in a positive sense, has perhaps been wage formation. It is not that long ago since Swedish wage formation used to be highlighted in economic circles as a warning to other countries. But with the odd exception wage formation in Sweden since the mid-1990s has worked well from a macroeconomic perspective. I think the fact that the Swedish Trade Union Confederation has said that “the Riksbank is a wage-earner’s best friend” clearly shows that something has happened. Real wages have grown positively in negotiations that have set out from stable inflation expectations (Figure 4). That is not to say that wage formation and the way the labour market functions are perfect. The future is likely to involve even greater requirements for a more productivity-based local or individual wage formation.
The vigorous productivity growth has been another positive surprise in the performance (Figure 5). At the start of the 1990s there were no indications that this would be the case. Why then would the Government have appointed a productivity delegation? The firm productivity gains have, of course, been crucial for the increases in real wages. But they also have meant that companies have not needed to hire new staff on the same scale as before, which has contributed to the weak employment growth in recent years. We really do not know why productivity has improved so strongly even if there are a number of hypotheses according to which the IT explosion, deregulation and increased global competition are significant explanations. For the Riksbank, it is important to understand this better if we are to be able to produce accurate forecasts of inflation. Of even greater importance, of course, is to do what is possible through economic policy to ensure that the positive performance can continue. In the long run higher real wages presuppose increased productivity.
Two problem areas
However, not everything has been as positive. Labour supply has not developed as well as one might have hoped. There are currently large groups outside of the labour market and employment has increased at a slower rate than is desirable over the past ten years. Continued weak growth in the labour supply will mean in concrete terms that the Swedish economy will reach its capacity ceiling at an earlier stage and that the Riksbank may be forced to raise the repo rate at a lower growth rate than would otherwise have been the case. Another dimension of the problem is that the ageing population gives rise to greater expenditure on medical treatment, care of the elderly and pensions, which burden government budgets. At the risk of sounding repetitive, as this is something pointed out by many in the debate, I would nevertheless like to emphasise that the only realistic means of alleviating these effects is for more people to work and pay taxes and for fewer to be dependent on the social welfare system. This area will offer one of the largest challenges to economic policy in the future.
Another factor that has proved surprisingly negative is the development of the krona. If someone had said to me ten years ago – when I first came to the Riksbank – that we would have as good macroeconomic growth as we have had, I would never have imagined that the krona would develop as weakly as it has done. The question is why this has happened? Traditional explanations based on competitiveness, relative inflation rates or growth rates are not sufficient. The Riksbank has therefore begun a large-scale study with the aid of external academics to try to understand this development. Questions relating to the taxation system, corporate location and investment trends will be examined in this context. At the same time, it is clear that the weak exchange rate has periodically been of assistance during, for instance, the latter part of the 1990s, when strong export growth was able to compensate for weak domestic demand. However, there is also another side to this issue. An excessively weak krona is not good for the Swedish people’s standard of living.
A few words about the Riksbank
During the hearings at the Committee on Finance I do not usually speak of developments within the Riksbank. Today I intend to make an exception and report some developments to you, our owners. In the mid-1990s we who formed the management of the Riksbank, my predecessor Urban Bäckström, my successor Stefan Ingves and myself, set a course for the organisation based on a few fairly simple principles. We would endeavour to concentrate on the Bank’s core tasks; monetary policy and financial stability. The emphasis would be on building up increased analytical capacity and competence within these areas. In addition, we would aim for greater openness, both internally and externally. These have been the guiding principles for our work over the past ten years. The changes that have been made are quite substantial. The number of employees at the Riksbank, including the Bank’s subsidiaries, has been reduced from over 1,200 to just over 400, which makes us one of the smallest central banks in Europe. Despite this, we have been able to significantly increase resources in our main fields. We now have a lot more people working on analysing issues related to monetary policy and financial stability than we had ten years ago. The academic level has also been raised substantially, which can be noted, for instance, in the fact that we have a much higher percentage of employees with postgraduate research degrees than we had ten years ago.
The emphasis on openness has also had clear consequences. We publish a comprehensive report showing the reasoning behind our interest rate decisions, the Riksbank Governor attends hearings at the Committee on Finance and members of the Executive Board travel around the country taking part in discussions and giving information on the Riksbank's work. In studies of central bank transparency, we are regularly ranked among the top few (Figure 6). In my opinion, this openness has been a central issue in gradually building up confidence in the Riksbank and its inflation target. It is also important to make it possible to regularly examine the Riksbank’s work in a serious manner.
One indication that we are on the right path to achieving a solid monetary policy regime is what happened when Urban Bäckström announced his resignation in May 2002 and when I myself did the same thing a couple of weeks ago. In neither case were there any significant effects on either the exchange rate or interest rates. Of course, I was a little uncertain as to whether I should just be pleased about this. But I think there is reason to be; it shows namely that we have a monetary policy regime which is stable and not dependent on single individuals.
Inflation targeting – some thoughts on the future
As I have mentioned, the macroeconomic outcome for Sweden has been relatively good over the past decade. Of course, this does not prevent there being questions worth discussing and much that can be improved. This also applies in the field of monetary policy. An important question in this context is what one can expect from monetary policy. If the expectations are excessive, there is a risk they will lead to the wrong issues being discussed and, if the worst comes to the worst, to political discontent that will undermine the current monetary policy system.
Reasonable expectations
According to the Sveriges Riksbank Act, the Riksbank is to safeguard price stability. We have chosen to implement this by keeping inflation at a target level of 2 per cent. The fact that the legislator has been clear that low inflation should be the target for monetary policy and nothing else, such as high growth or low unemployment, is because inflation is the only factor the Riksbank can govern in the long term. It is therefore unwise, and perhaps even counterproductive, to give a different impression. For instance, it is worth repeating that in order to reduce unemployment permanently, an attempt must be made to influence the structural component of unemployment, i.e. the component that is not due to fluctuations in the demand for goods and services in the economy. And this requires completely different measures than those of monetary policy. These measures must influence the functioning of the labour market and create a beneficial environment for new businesses and corporate sector growth. Of course, this does not prevent the Riksbank from contributing as far as possible to other economic policy targets being attained; this is quite in line with the preparatory works to the Sveriges Riksbank Act. The condition for this is that our own target of a low and stable inflation rate is not jeopardised.
The fact is that developments in recent years show that the task of ensuring inflation remains in line with our target is hard enough. There have been repeated structural changes that have been difficult to parry completely with interest rate adjustments. It is rarely possible to forecast structural changes well in advance, much less to estimate their quantitative effects on inflation. They can therefore often lead to inflation deviating from the target level. As I mentioned earlier, over the past ten years there have been a number of changes of a structural nature that have almost all contributed to holding back inflation.
When the Riksbank uses changes in its key rate to affect inflation, this is mainly by way of influencing aggregate demand. To put it simply, we usually put the brakes on when economic activity is rising and we accelerate when it is falling. Inflation targeting should thereby often be able to contribute to smoothing the business cycle. It is also possible that it can reduce fluctuations in unemployment. However, it is not reasonable to believe that the fluctuations in economic activity can be eliminated entirely. It is therefore seriously misleading to make the Riksbank responsible for the difference between actual unemployment and what could potentially have been attained if it had been possible to abolish fluctuations in economic activity. What we can hope for is a slightly more stable development, which we appear to have attained over the past ten years, although it is too early to draw any completely certain conclusions.
The policy is flexible
Another question which arises in the discussion now and then is how mechanical or how flexible the policy should be and how it should then be assessed. The Riksbank currently conducts what is known as flexible inflation targeting. We are definitely not what my British colleague Mervyn King would call “inflation nutters”. The focus lies not merely on stabilising inflation under any circumstances; as long as there is fundamental confidence in the policy, that we will have low inflation in line with our target, we also take into account other factors.
In January 1999 we published a clarification of our policy. There we made it clear that some deviations from the target level could be accepted, but that the Riksbank at the same time normally strives to bring inflation back to target level within two years. One reason for allowing deviations is that monetary policy does not make such a rapid impact and does not have the precision to closely steer inflation in the short term. Another reason is the need to be able to take account of other factors than inflation trends – for instance, growth in the real economy.
The principle of normally bringing inflation back in line with the target within two years can be interpreted as a restriction the Riksbank has imposed upon itself to make its policy more clear. This restriction helps clarify the consideration normally given to factors other than the price stability target. However, sometimes situations arise where flexibility with regard to the target needs to be greater to avoid the effects on the stability of the real economy – variations in production and employment – reaching unacceptable levels. In extreme cases the deviations from the inflation target may be so large that it is neither possible nor desirable to return inflation to target within the usual two-year perspective.
There are a number of examples of situations where considerations other than the assessment of CPI growth over the coming two years have affected our policy. During the mid and late 1990s we did not cut the repo rate as much as our normal principles might have recommended, as our assessment was that some of the fall in prices was due to our own interest rate cuts, which subdued the housing cost component of CPI, and to deregulation that was partly of a temporary nature. In the same way, we did not react as much as we could have done when inflation rose sharply for transient reasons in 2001 and 2003. A more recent example is the decision in March 2005 not to cut the key rate. Our reasoning then was that it might have been possible with the inflation forecasts we had, but that the expected high growth, combined with elevated house prices, called for some caution.
Our flexibility can also be said to include a preparedness to react to events where the final consequences are particularly difficult to survey. One such example is the acts of terrorism in the United States on 11 September 2001. In the wake of these attacks, many central banks around the world were quick to cut interest rates. This powerful response can be interpreted as a kind of "insurance" against a really poor outcome, rather than as the result of carefully-considered assessments of the most probable outcome.
It would be desirable, for a better understanding of monetary policy, to develop the principles governing how our reasoning in practice weighs future inflation against real or financial stability. However, this is for obvious reasons no simple task and there are no good examples of this from other central banks. Last spring, the Riksbank began work aimed at achieving this objective. One potential result of this work could be a development of the clarification from 1999.
A clearer picture of which deviations are allowed would also make it easier to communicate our policy and to assess it in a purposeful manner. At present, policy is often assessed solely on the basis of historical averages for inflation. While this can act as a natural starting point, the discussion should be taken further and include how the outcome of the policy relates to the considerations taken to various more transient price fluctuations. Real and financial developments should also be taken into account in contexts where they have affected the policy.
I imagine that these are questions to which the Committee will wish to return in future, both as a result of the work conducted within the Riksbank and perhaps also as a result of the external assessment of monetary policy the Committee intends to make. The Riksbank is also holding discussions with external parties on a regular assessment, where a preliminary starting point would be comparisons between the assessments made by the Riksbank and other forecasters in order to provide a better basis for a correct discussion.
The economic situation
Let me now comment on our current assessment of inflation prospects and our interest rate decision yesterday.
International cyclical upswing continuing
World economic growth overall has been relatively firm for a couple of years. Last year in particular, growth figures were very high, driven by the marked upswing in the United States and by stable firm growth in large parts of Asia, but also in other emerging market economies. However, during the spring new data releases prompted us and other forecasters to question the strength of the international economic upswing that we had envisaged. Growth statistics from the euro area were weaker than expected and the information from the United States was providing a somewhat mixed picture.
Judging by the data that have been released over the summer and at the start of the autumn, most indications are that the slowdown in economic activity that we saw at the beginning of the year was temporary. In the United States, growth continued to be high in the second quarter. In China and the rest of Asia the economic expansion has continued and it is pleasing that the Japanese economy now looks stronger than it has for a long time. On the other hand the performance in the euro area has remained weak, especially in the large economies, but there are nonetheless some indicators there as well pointing to a turnaround.
Overall, it appears that GDP growth abroad will continue to be firm this year and that there are prospects that the global cyclical upswing will persist in the coming years, with the United States, China, and the rest of Asia as the main engines.
But nothing is certain when it comes to economic activity. One obvious source of uncertainty is the oil price, which has turned out much higher than we anticipated as late as in June. Forward prices, which reflect expectations of future oil prices, have also risen and judging by these, expectations in general seem to be set on a slowly falling, but nevertheless still high, oil price. Even though the oil price has dropped back recently we have revised up our forecast for the coming years.
It is hard to judge what effects the high oil price will have on growth and inflation. Among other things, this depends on whether the high price reflects mounting demand or whether it has been caused by supply shocks. Pointing to the first explanation is the fact that the price rise in recent years has gone hand in hand with an increase in output. A dampening factor is that most industrialised countries are considerably less dependent on oil than they were a couple of decades ago. Against that background, we estimate in our assessment that the high oil prices will have limited effects on economic activity. But, of course, it is not possible to rule out either higher inflation or a more negative impact on demand. Signs of the former can be seen in the United States. Another source of uncertainty is the large savings imbalances around the world and how a correction might come about.
Strong domestic demand in Sweden
In Sweden, too, economic growth was strong last year. During the spring this year, however, there were signals indicating that a slowdown was on the way. Moreover, in June growth data were released for the first quarter, which were weaker than expected by all economic forecasters. That prompted the Riksbank to revise down its growth forecast for 2005, as well as our outlook for future cost pressures. But we nevertheless stuck to our assessment that most factors – not least the expansionary economic policy – pointed to comparatively firm growth in the period ahead.
That now seems to have been a correct assessment. The statistics for the first quarter were revised up over the summer and thus were not quite as weak as they first appeared. During the second quarter GDP growth also rose, according to the data we now have. The performance is positive as regards both domestic demand – consumption and investment – and exports. Looking ahead, we expect that domestic demand will be a more important driving force than it has been so far in the cyclical upswing.
However, demand is likely to begin to slacken by degrees as a result of ordinary mechanisms that dampen economic activity. One such mechanism is that interest rates rise as resource utilisation and cost pressures pick up.
Previously, the forecasts in our main scenario have been produced on the basis of the assumption that the repo rate is constant during the forecast period. In this Report the forecast in the main scenario is based on the assumption that the repo rate evolves in line with implied forward interest rates. According to this assumption the repo rate will remain at 1.5 per cent up to about the second quarter next year, when it gradually begins to be raised.
Overall, the Swedish economy is forecast to grow at a comparatively firm rate during the forecast period. GDP growth is estimated to be 2.3 per cent this year and 3.0 per cent next year. Thereafter, it will gradually fall towards just over 2 per cent in 2008. Compared with the assessment in the June Report, this is an upward revision and the fundamental outlook is now more positive. For this year it is mainly the revision of the outcome for the first quarter that has affected the forecast. After that it is, among other things, the more expansionary economic policy compared with our June Report – both a slightly lower repo rate and more expansionary fiscal policy – that have raised the forecast.(1)
Resource utilisation to increase gradually
There are now signs that the situation in the labour market is improving. Among other things, the number of hours worked has begun to rise and indicators of the number of redundancy notices and new job vacancies also point to a more robust labour market.
Relatively firm growth in domestic demand in the period ahead, coupled with the fact that it should become increasingly difficult for companies to boost output without also adding to their workforces, means that we foresee higher demand for labour in the coming years. The announced labour market policy programmes for next year entails that total employment will rise more than regular employment, but there should also be an increase in the number of regular jobs. Our forecast is that the number of employed will rise by 0.4 per cent this year and 1.1 per cent next year. Open unemployment is estimated to be 5.9 per cent this year, 5 per cent in 2006 and then to drop somewhat more.
Wages are forecast to increase faster as the labour market improves. During the latter part of the forecast period growth in productivity is also expected to diminish, which is natural when growth in the economy will be driven to a greater extent by domestic demand and higher output in the private services sector. Overall, that means that companies’ costs will begin to increase faster again, but our assessment is that the rise will be comparatively slow and that cost pressures will be moderate throughout the forecast period.
It is difficult to determine exactly how much the productivity upswing will be dampened. It is possible that the rate of increase in productivity has been raised on a more permanent basis compared with before. One cause may be the large-scale investment in IT undertaken by companies at the end of the 1990s. But it is also possible that the continued economic upturn will be followed by a rapid decline in productivity growth. Our assessment today is that the high growth in productivity will last somewhat longer than assumed in the previous Inflation Report and the forecast for average productivity growth therefore has been revised up slightly over the forecast period.
Inflation rising but still low
Inflation has risen somewhat recently but remains low. In September this year, the annual CPI and UND1X inflation rates stood at 0.6 per cent and 1.0 per cent, respectively. That is slightly higher than we forecast in June, owing mainly to the unexpectedly high oil price.
In the coming years, mounting domestic cost pressures are anticipated to contribute to a rise in domestic inflation. However, the rate of this increase is estimated to be modest due to both continued fairly strong productivity growth and stiff competition.
That imported inflation in recent years has been contained is due, among other things, to the strengthening of the krona and the fact that there has been a substitution in favour of cheaper imported goods. The strength of these effects is likely to diminish slightly during the forecast period, as a pick-up in global resource utilisation leads to higher international price pressures. Neither do we believe that the krona will appreciate as much in the period ahead.
The overall assessment is that inflation will rise in the coming years, but at a fairly moderate pace. For the larger part of the forecast period inflation is expected to be below 2 per cent. Our forecast is that the twelve-month change in the CPI will be 1.5 per cent in one year and 2.2 per cent in September 2007. The corresponding forecast for underlying inflation UND1X is 1.2 per cent and 1.8 per cent, respectively.
The question then is what this entails for monetary policy. What is clear is that the low inflation has meant that monetary policy has been highly expansionary for a rather long time. This, of course, has helped stimulate economic activity and contributed to the firm economic growth. It also has had consequences for, among other things, developments in house prices and households' debt situation.
As I observed earlier, the assessment in the Inflation Report’s main scenario is based on the technical assumption that the repo rate will develop in line with implied forward rates. It must be emphasised that the assumed interest rate path should not be interpreted as reflecting the path the Riksbank considers to be most desirable.
The inflation development I have described and which follows on from the current interest rate assumption is judged to be reasonable, but there are several alternative paths for the future repo rate, which would give roughly the same outcome for inflation. As before, there is also reason to note that household borrowing and house prices are continuing to rise rapidly. Given this, the decision we made yesterday was to leave the repo rate unchanged at 1.5 per cent. The precise future stance of monetary policy will depend as usual on new information regarding economic developments in Sweden and abroad and on the Riksbank's assessment of the inflation outlook.
-------------------------
By way of conclusion I would like to refer back to the first part of my speech, where I presented some information regarding the outcome of the policy that has been conducted in Sweden in the past ten years and of the Riksbank's efforts. Let me stress that the good results that have been achieved – with a stable and relatively strong economic performance – are very much due to decisions taken here in the Riksdag. I am thinking of those decisions you made during and after the crisis years with regard to budget policy, both its framework and its implementation, and the various decisions you have taken that have helped to make the Riksbank independent.
It has been a great privilege to have been involved in this process and to have had the opportunity to work for ten years in the management of the Riksbank. I would like to take this opportunity to thank you for the faith you have shown in me and the support and encouragement you have given me, as well as for your excellent cooperation, and that applies to you all irrespective of political affiliation.
Thank you!
Notes
1) The comparison refers to the forecast based on the assumption that the repo rate evolves in line with implied forward interest rates, which in the previous Report was presented in a box.
