Heikensten: Monetary policy - some current issues
The whole speech
Let me begin by thanking you for the invitation to come here to the economic association. It has become something of a tradition that the Riksbank Governor comes here once a year to talk about current issues regarding monetary policy. I think this is excellent, particularly as it contributes to a lively, in-depth discussion of monetary policy in society. Those who have followed the Riksbank and monetary policy in recent years know that transparency and clarity have been guiding principles in our work.
Today I intend to provide some perspective on monetary policy by means of a backward glance over the past ten years. This is a natural foundation for the current discussion on interest rates which I will address a little later. I have no intention of giving any new monetary policy signals today, but to give an account of the basis for our repo rate decision last week. One particularly interesting question at present is productivity growth, which appears to have been better than expected last year. The conclusions that can be drawn from this development are my third theme. Finally, I shall give my views on the recent monetary policy debate, something I have been asked about on numerous occasions in the past couple of months.
The results of inflation targeting
My retrospective begins in autumn 1992, when the krona moved from a fixed exchange rate to a floating rate. It was then important to find a new anchor for monetary policy as rapidly as possible. The alternative chosen was inflation targeting. In January 1993 the Riksbank stated in a press release that the change in CPI inflation would in future be limited to 2 per cent a year with a tolerance interval of ± 1 percentage point.
The floating exchange rate and the inflation target constituted a drastic changeover of monetary policy. There were also a number of problems in the initial stages. The decision to introduce inflation targeting was controversial and much criticised in the public debate. Naturally, the Riksbank had no experience of this type of monetary policy regime and even in countries where it was already established it was a relatively new phenomenon. Moreover, the experiences of Swedish stabilisation policy in both the monetary policy and fiscal policy fields during the 1970s and 1980s were less than positive. This meant that there was considerable opposition at first and many believed the new regime would be short-lived.
The Riksbank had to balance the need to create confidence in the new, price stability-based regime against the need to conduct less restrictive monetary policy to soften the effects of the economic crisis. It was a difficult starting point for many reasons, including the fact that public finances were in very poor shape. There were also problems with inflation. When the crisis began, the Swedish inflation rate fell significantly, but up to 1995-96 it was still far above the target level. The problems were emphasised by developments in inflation expectations. It was not until 1997 that expectations stabilised around 2 per cent. (Figure 1).
However, around the mid-1990s confidence in economic policy increased. By then, the economy had grown for a couple of years, stimulated primarily by the weak krona rate. When drastic measures were taken to improve public finances, interest rates fell. Monetary policy contributed to this development by consistently fending off all tendencies towards excessively high inflation. It was now possible to have more stable economic growth. During the period 1995-2001, economic growth was an average of 3 per cent, which can be compared with an average of 2.4 per cent during the two previous decades. (Figure 2).
Developments in the labour market also indicate that the measures taken were successful. I recall that some economists wanted to interpret the high level of unemployment following the crisis years as an indication that the changeover of monetary policy regime had raised the economy's equilibrium unemployment level and caused a permanent deterioration in employment. However, once the crisis had been overcome and the new policy was established, the labour market situation soon improved. Unemployment is currently around 4-5 per cent, which can be compared with figures of above 8 per cent in the mid-1990s. However, we still have some way to go to return to what was considered full employment prior to the crisis years 1991-93.
All of this has been achieved parallel with a low inflation rate and during the period as a whole inflation has been in line with the Riksbank's target. From 1993 to the present day, the increase in CPI inflation has averaged 1.9 per cent, while the measure of underlying inflation on which we usually choose to base our monetary policy decisions - UND1X - has risen by an average of 2.1 per cent. (Figure 3).
It is not easy to determine what effect the various changes in economic policy have had individually. However, when taken as a whole they have led to an overall changeover in stabilisation policy. A clearer allocation of roles has been created in economic policy. The task of monetary policy is to attain price stability, while fiscal policy is to secure a long-term sustainable development in public finances. The government and the Riksdag (parliament) also have the task of increasing the number of those in regular employment to 80 per cent. This clear allocation of roles is an important reason behind the more successful economic policy over the past ten years.
The current monetary policy situation
The international recovery is increasingly evident
During 2001 there was a downturn in global economic activity, which was partly due to the financial setback following on from the stock market bubble in the second half of the 1990s. Resource utilisation fell and growth in the OECD area as a whole was 1 per cent that year. The downturn was remarkable symmetrical and counteracted by a series of interest rate cuts throughout the western world. Sweden was also affected by this downturn, of course, although to a lesser extent than many other countries, particularly because fiscal policy was more expansionary, but also due to the krona being undervalued. Resource utilisation was therefore assessed to be higher in Sweden than in many other areas, which affected the relative interest rate situation.
At the beginning of 2002, a wave of optimism began to spread, but it proved short-lived. The main reason for this was the increasing geopolitical unease concerning Iraq and later the war in Iraq. A recovery in the global economy was still expected, but at a later date. The weak economic activity led to further easing of economic policy during 2003. The United States Federal Reserve, the Riksbank and other central banks continued to cut their instrumental rates to all-time low levels.
Now most indications are that both the Riksbank and other analysts were slightly too pessimistic in connection with the Iraq war, particularly with regard to the view of economic activity in the US economy, but also developments in Asia. During the second half of 2003 we have had reason to make a further upward revision to our assessments of global economy growth and the picture of the global economy as a whole looks good at present.
The US economy is expected to grow at a slightly more rapid rate than we estimated in the December Inflation Report. In the short term, the most significant cause for concern in the US economy is that employment is still weak. Of course, the combination of rapid growth in GDP and a weak labour market mean that productivity growth is good, but the dismal labour market situation could nevertheless create uncertainty among households and contribute to subduing consumption.
However, it is difficult to believe that it would put an end to the upturn we are seeing now.
There are also signs of an upturn in the euro area now, although the picture is not as clear-cut as in the United States. Business tendency surveys for industry show an increased optimism, while industrial production is slowly beginning to rise. Developments in Asia, central and eastern Europe also appear to be slightly stronger than was previously assumed. However, the stronger euro may lead to a slightly slower upturn in the euro area.
All in all, the picture of international economic activity now looks somewhat brighter than we anticipated in December. The upturn in economic activity appears to be proceeding roughly as we had anticipated it would.
The Swedish economy has also turned
We assumed in the December Inflation Report that growth in Sweden would gradually increase during the course of 2003 and would amount to 1.5 per cent for the whole year. During 2004 and 2005 growth was expected to be around 2.5 per cent, which is in total a slightly more rapid rate than the Riksbank considers sustainable in the long-term without inflation rising.
The new statistics support this scenario, although the picture is not clear-cut. For instance, the National Accounts for Q3 indicate an increase in growth compared with earlier quarters, in line with the forecast in the Inflation Report. Other indicators of a recovery are the purchasing managers index and the National Institute of Economic Research's business tendency surveys, as well as the fact that retail trade continues to show good growth. Counteracting this is the fact that manufacturing output was weak during October and November and the number of new orders declined. In addition, new estimates for public finances indicate that public consumption this year may be somewhat weaker than was assumed in December. This is mainly connected with the financial situation in local government. The local government sector as a whole shows a deficit, which must be redressed over the coming years.
Although the dollar has weakened against the Swedish krona, in terms of the TCW index the krona has not appreciated as much. This is mainly because the krona, contrary to many expectations, has weakened against the euro. The most recent National Accounts figures and foreign trade statistics also indicate that we have so far been slightly too pessimistic with regard to Swedish exports.
The most evident cause for concern is the labour market, which has developed more weakly than expected, particularly towards the end of 2003. The strained financial situation in the public sector would seem to indicate weak employment growth for at least the coming year, while the economic upturn will probably only lead to a relatively small increase in employment in the private sector. There is, of course, a risk that the weak labour market will affect households' expectations and consumption propensity; a risk that we pointed out already in December and which may have been reinforced. The higher the rate of unemployment, the more caution households will probably show when purchasing and consuming. This could in turn lead to a risk of lower growth during 2004.
All in all, the view of the demand situation in Sweden remains essentially the same as in December; demand will increase and approach the economy's long-term sustainable level.
Lower domestic cost pressure
The relationship between unit labour costs and inflation is usually fairly clear. Thus, when we made a downward revision to the figures for unit labour costs in December, it was not entirely insignificant for the inflation forecast. The background was primarily the signs we saw of an improvement in productivity growth, combined with a weaker labour market. I shall return to the subject of productivity in more detail in a moment.
With regard to wage developments, the wage bargaining rounds are currently a central issue. Their outcome will have great significance for the rate of wage increase and thereby also to a large extent for inflation in coming years. It is clear that the labour market situation calls for moderate wage increases. Unlike the situation prevailing during the two previous bargaining rounds, unemployment has increased and employment has declined recently. Nor are there any labour shortages, other than in a few areas. It is also a positive sign that the rate of wage increase has been declining over the past year and that the same applied to expected inflation. Given the present low inflation rate, even limited nominal wage increases will lead to a fairly good development in real wages, which trade union representatives usually emphasise. The nominal wage increases put forward now are lower than those of recent years.
Meanwhile, there is of course always a risk of shocks from the actual negotiation process. It is not unthinkable that some of the increase in productivity is used as justification for higher wage-paying capacity. This could be particularly worrying if the wage negotiations assume that the improvement in productivity is permanent and it later shows a rapid decline.
At our meeting last week we considered that there was reason to make a further slight downward revision to unit labour cost figures, which in turn had effects on domestic inflationary pressure. The labour market has been weak and most indications are that there is reason to be slightly more pessimistic with regard to employment in future. The rate of wage increase could therefore be slightly lower than we assumed in December. Meanwhile, productivity growth has continued to be higher than expected and we believe that it may rise even further in the near future. Inflation has also been a little lower than anticipated. We are expecting a low underlying inflationary pressure in the short term, but this will probably increase as economic activity picks up. However, import prices are very important to our assessment of future inflation. These have had a significantly subduing effect in recent times, but an upward turn appears probable over the coming year. All in all, our assessment is that inflation will begin to rise later this spring and gradually approach our target level two years ahead. However, we do not believe that this upturn will be particularly dramatic.
Let me emphasise that the changes that have occurred since our December meeting are not particularly great. They mainly entail some signs that domestic cost pressure could be lower than anticipated. As early as December we made some adjustments to our estimates in this respect, but the signs were fairly new then and a majority of the Executive Board preferred to await further information that would make the picture clearer.
Given this, the assessment was that inflation would be slightly lower throughout the forecast period than was assumed in December; a conclusion that was not particularly sensitive to the choice of inflation measure. The Executive Board of the Riksbank therefore decided to cut the repo rate by 0.25 percentage points.
One question that has been raised in the debate is whether it is too late to cut the rate now, given the economic recovery. This is not the decisive issue for us. Our method of working means that we base our decisions on inflation forecasts. These show that there is scope to cut the rate. In the short term, inflation is expected to be much lower than 2 per cent and to cautiously approach the target level only 1-2 years from now. It would therefore appear that there is plenty of time for an upward adjustment of the interest rate in the future if our forecasts indicate this is necessary. The situation would be different if there was reason to expect a rapid economic upturn now, which could threaten fulfilment of the inflation target in the shorter term. A cut in the interest rate can perhaps also be regarded as an insurance against weaker growth in demand; although we have assumed a clear improvement in economic activity, this is not guaranteed.
Finally, I should like to comment on an element in the present discussions on monetary policy. Many debators have shown a tendency to argue as though it were possible to conduct a perfect, fine-tuning policy. I believe it is almost impossible. I think that it is always possible to claim with hindsight that a little bit more or a little bit less could have been achieved either earlier or later. However, the Riksbank's most important contribution is not this type of "fine-tuning", but rather maintaining the inflation target and upholding confidence in price stability. When we achieve this, we can contribute to stabilising real growth in production and employment. But this is not our main objective.
Productivity growth - an expansion
Last year, employment in the business sector showed an unexpectedly weak increase, while growth was roughly as expected. This was reflected, as I have already mentioned, in the fact that productivity growth was unusually good. The most recent National Accounts figures also indicate a stronger productivity growth last year than we assumed in our December assessment. The question is what this implies about the future; is what we are seeing a sign of higher productivity growth which will continue over a number of years or is it a temporary rise as a result of an improvement in economic activity? Or is the upturn perhaps a chimera and the result of statistical problems connected with, for instance, difficulties in measuring productivity in certain sectors, such as the telecom industry?
In the United States, the pattern of a weak growth in employment, combined with a recovery in economic activity has been established for a number of years. According to the National Bureau of Economic Research, NBER, the US economy bottomed out in November 2001. Although the size of the decline in employment measured varies from one source of statistics to another, the overall picture is still clear; employment has been weaker than anticipated, while economic growth has been stronger. This phenomenon is known as jobless growth, or jobless recovery.
The parallels to the recent developments in Sweden are obvious (Figure 4). Despite the increasingly clear signals of stronger economic activity, growth in the labour market is very slow, particularly in the business sector. Productivity growth is tangible in both the United States and Sweden, while many countries in Europe are still struggling with a relatively weak growth. However, one important difference in the comparison between the United States and Sweden is that both the rate of growth and the increase have been much more dramatic in the United States. The same applies to the size of the decline in employment, which has been much more significant in the United States. This means that comparisons between Sweden and the United States should be interpreted with caution.
There is reason to reflect over the explanations that might lie behind a productivity increase of a more fundamental nature, similar to that we have seen in the United States in recent years. Personally, I have found it helpful to use the events of the late 1990s as a starting point; a period when productivity was also surprisingly strong. The driving force behind this development is usually said to be technological progress in information and communications technology, which led to productivity increases in the IT-producing industry and price falls on IT products. In addition, there was extensive investment in IT capital in other industries, also stimulated by favourable price developments, which resulted in higher productivity outside of the IT sector. Expressed in a different way, productivity growth at this stage was said to be driven by increases in total factor productivity within the IT-producing industry and by an increased IT capital intensity in other sectors of the economy.
Some analysts have pointed to the possibility that the recent productivity increase could be at least partly due to IT capital now making an efficiency impact on production even outside the sectors producing IT capital, where it has enabled rationalisation of labour and a reduction of the investment requirement. This theory thus explains the combination of rapid productivity growth, weak employment growth and a slow increase in investment, which has recently characterised both the US and Swedish economies.
So what - apart from the actual developments in productivity, employment and investment - indicates that this is the explanation?
One piece of the puzzle that supports this theory is that the improvements in productivity can scarcely be explained merely by business cycle effects. This becomes clear if we study productivity growth in the corresponding phase of the business cycle during earlier periods (Figure 5). The measure of economic activity we have chosen is, however, more or less arbitrary and does not entirely correspond to the assessment of resource utilisation in the Riksbank's Inflation Report. Nevertheless, this does not appear to affect the general conclusion that productivity growth is unusually rapid at present, bearing in mind economic activity and could thus be a sign of a sustained change.
Another piece of the puzzle is the development of factor productivity (Figure 6). As I mentioned earlier, a more rapid growth in factor productivity at aggregate level could be a sign that IT capital is being used more efficiently even outside of the sectors producing IT capital. The best method would be to study factor productivity excluding the sectors producing IT capital, but this is not an easy task, due to various statistical problems. The calculations of factor productivity we have made at the Riksbank indicate a relatively clear upturn in recent years.
It is also interesting in this context to compare how developments in Sweden compare with other countries, particularly the United States. These comparisons are made more difficult by the fact that comparable statistics are hard to find. However, the statistics that are available indicate the factor productivity growth is relatively good in Sweden; much better than in the euro area, for instance.
To summarise, we can conclude that there has been an upturn in productivity development in recent years. Earlier economic patterns are not being repeated entirely. However, it is difficult at present to establish what lies behind the upturn. The situation is reminiscent to some extent of that we saw for a few years at the end of the 1990s. Then, as now, there was reason for a wait-and-see policy, but not to reject outright the idea that what we see is a clear change. In any case, some importance should be attached to the new figures when taking a policy stance, although we must try even harder to understand what is happening.
The monetary policy debate
Allow me to say a few words in conclusion regarding the recent, lively debate on monetary policy. This is because I have received many questions on this subject in recent weeks. However, it is not always easy to get a subtle message across when the media debate is so intense. I therefore decided at an early stage of the discussion to take up this question in greater depth here at the Economics association, a suitable forum for a penetrating and relatively non-media discussion.
Before I try to explain my views on some elements of the discussion, I cannot resist pointing out that I found the tone of the discussion somewhat surprising. Sweden has come through the recent economic slowdown better than many other countries. Personally, I also felt much greater concern over economic developments and monetary policy at the end of spring and beginning of summer 2003 than I have done in recent months. The reason is simple. At the end of last spring, the inflation rate was still high. Although we believed it would fall, there was a clear risk that it could become entrenched. Meanwhile, there was considerable uncertainty over international economic activity; would there be a turnaround and if so, when? This could mean we are forced to make difficult judgements. For instance, we might be forced to raise the interest rate even if economic activity weakened. Now the situation is fundamentally much better, with declining inflationary pressure combined with increasing growth. Nevertheless, the debate is much tougher. Why, one might ask.
What can be said about the discussion itself? Let me begin by pointing out the obvious; it is good to have a vivid, lively discussion of monetary policy. A central bank that has been given clear, independent responsibility for monetary policy has good reason to be open about the bases for its decisions. This facilitates a more exact discussion of monetary policy as well as assessment afterwards. An open debate forces and encourages us to improve our arguments and to continuously develop methods and assessments, in other words to become better. This is all for the good and in actual fact one of the most important motives behind the openness we have striven to achieve over the past ten years.
Various assessments of the situation have been made recently, both by external analysts and within the Riksbank. For instance, in December two of my colleagues believed we should cut the repo rate. Essentially, I think that the debate has been factual in that the demands for rate cuts have usually been based on assessments of future inflation differing from those made by the Riksbank. It is only in exceptional cases that these exhortations appear to have been based on inflation targeting as such being questioned.
However, this does not mean that all elements of the discussion have been good and clear-cut. One issue that has led to questions and been discussed considerably is that some leading politicians have actively involved themselves in the debate and on repeated occasions expressed very specific requirements for immediate interest rate cuts and in one case even a change in our meeting schedule. The question is whether this is appropriate or even wise.
This is not a new issue, and it is one I have faced on numerous occasions since I began work at the Riksbank. My point of view has always been that we have a statutory task to perform. We shall perform this independently and to the best of our abilities. The law is crystal clear; we are not allowed to accept political instructions or even advice. Moreover, even before the new legislation was in place we had chosen to design a clear framework for our work. Our target of maintaining price stability has been specified as an inflation rate of 2 per cent. We have also "tied" our actions in other ways, such as publishing forecasts and systematically following particular working methods. This enables those outside of the bank to assess our actions and motives and thereby to be sure that we are not taking account of inappropriate considerations. Given this, over the years I have had a fairly relaxed attitude to politicians' interference in the general debate.
However, this issue has further implications than such. Over the Christmas holidays I read a book by the former US Treasury Secretary Robert Rubin(1) . He takes up this issue and motivates why he strongly advised President Clinton against holding public opinions on monetary policy, advice followed by the President. According to Rubin, there is always a risk that the bond market will be affected if there is the slightest suspicion that the central banks' actions are governed by political pressure(2). I think that the fact that he - with his decades of experience on Wall Street - brings up this argument and that he does so despite the widespread respect enjoyed by Greenspan's Federal Reserve, shows that the argument cannot be waved aside. One could object that the Riksbank's legal independence is greater than that of the Federal Reserve. On the other hand, the low-inflation regime is relatively new in Sweden and our market is also small and influenced to a large degree by external agents, who often lack in-depth knowledge of how our system works.
At present, the economic and political developments in Sweden are so stable that political manoeuvring of the type we have seen hardly has any effect on the markets, although some concern has been expressed by foreign analysts. However, politicians who enter the discussion with strong opinions in a stable situation should probably reflect over how they might act in future in more problematic situations.
Rubin also points out that the president was able, by showing a clear respect for the Federal Reserve's independence, to increase his own credibility and build-up confidence in economic policy in the United States and in the functioning of the financial markets. This is a view I can sympathise with. The question is then how we in Sweden should act to ensure that confidence in our total economic policy is as great as possible. Essentially, this of course means doing the right things. In our case, ensuring price stability in accordance with our own definition. However, I believe that the ways in which the authorities responsible for fiscal policy and monetary policy cooperate can also play a role here. In this aspect I can see clear advantages in the interplay between government and central bank in the United States and the United Kingdom. Maintaining mutual respect for one another's tasks, they have a close cooperation to conduct an overall policy that will benefit their country as much as possible. This is a desirable form of cooperation, particularly for small countries and I believe that we in Sweden have come a long way towards this since the mid-1990s. It would be a negative development if the pattern were broken and we instead slipped into a form of interplay similar to that in continental Europe, where national finance ministers blame their problems on the ECB and where sharp words are regularly aimed in the opposite direction.
Thank you.
Notes:
1) See Rubin, R.E and J. Weisberg, In an Uncertain World: Tough Choices from Wall Street to Washington, New York: Random House, 2003.
2) Lawrence Summers, who was Treasury Secretary towards the end of Clinton's period in office, has even designated criticism of the Federal Reserve by politicians as "a fool's game", which is at best ineffective and at worst counterproductive. "[I]t does not change short-term interest rates because the Fed does not respond, but it does increase long-term interest rates because the bond market does". (http://www.ustreas.gov/press/releases/rr2401.htm)
