Heikensten: Monetary policy and the new Executive Board

The autumn conference, arranged by the Centre for Business and Policy Studies, is a tradition I have come to enjoy and an opportunity of looking back on the course of monetary policy and considering future challenges.

My review last autumn began with the observation that a good deal had happened as regards economic developments and the conditions for monetary policy. Much the same applies on this occasion.

Two years ago we were in the throes of what looked like being a strong cyclical upswing with attendant price increases. My remarks at the autumn conference caused short interest rates to move up about 0.10 percentage points in expectation of a tighter policy. Little more than a month later the Riksbank did in fact raise the repo rate. Last year the situation was very different, with an international financial crisis and great uncertainty about its effects on the real economy. The krona had become markedly weaker and there was some discussion at the conference about the Riksbank’s recent currency market intervention. Later that autumn the Riksbank reduced the repo rate in two steps. Today the prospects for growth and employment are bright again and the krona is substantially stronger. The debate is focussed once more on the risks of overheating.

The setting for our activities is evidently liable to change at short notice. I shall begin, as usual, by talking about what has happened since we last met and commenting on the public debate. One of the changes this time has to do with the set-up for the Riksbank. New legislation from the beginning of this year has given the Riksbank a more independent status and a new managerial structure. So I shall also be saying something about the new conditions for the Riksbank’s communications. Finally I shall be looking ahead and commenting on our appraisal of the coming years. In doing so I shall consider a matter that has attracted a lively discussion recently: Has the trade-off between growth and inflation in Sweden changed? This is an issue we have had to consider carefully in recent years.

The international financial markets were very turbulent last autumn after Russia’s suspension of debt payments and the problems with the hedge fund Long-Term Capital Management. Analyses at that time suggested that the financial turmoil would have considerable global consequences for the real economy. The Riksbank judged that even with good domestic demand and a weaker exchange rate, the slacker international trend would cause inflation in Sweden to be below the target in twelve to twenty-four months’ time. In November we therefore reduced the repo rate in two steps from 4.10 to 3.60 per cent.

In December it was considered that both CPI and underlying inflation would be below the target in the coming one to two years. The downside risks were judged to predominate, mainly due to the uncertainty about the future course of international activity. The repo rate was reduced again, from 3.60 to 2.40 per cent, accompanied by signals that no further interest rate adjustments would be made before the changeover to the new Executive Board.

New statistics during the winter continued to show a weak international tendency. The Riksbank perceived a clear risk of this having negative effects on growth in Sweden, so that inflation prospects would also be affected. In February and March the new Executive Board therefore made two more repo rate cuts, from 3.60 to 2.90 per cent.

The mood swung as winter gave way to spring. The financial market unrest subsided, accompanied by the first signs that the development of activity was better than had been expected. At the monetary policy meeting on 22 April the Executive Board decided to keep the repo rate unchanged. A majority of the Board wanted to wait for a more comprehensive assessment of inflation prospects in connection with the June forecast before making any further interest rate adjustments.

In the late spring and the summer it became clear that economic trends were more favourable than had been predicted earlier. Economic growth in the United States had been higher than expected and the first signs of a recovery in Europe were now visible. Meanwhile, a distinct upturn in the Swedish economy was increasingly apparent. The forecasts were revised upwards and the Executive Board now agreed that further interest rate reductions were no longer on the agenda.

By the middle of August, at the monetary policy meeting on the 12th, the picture of a stronger economic development in Sweden was even clearer. The Board therefore concurred on a high future growth rate that would call for a tighter monetary stance. But price increases were low and the available statistics, on capacity utilisation and the labour market, for example, did not point to a rapid future increase in inflation. The Board’s majority therefore considered that an immediate tightening was not needed. With inflation expectations firmly anchored around the 2 per cent target, we judged there was scope for waiting with an interest rate increase.

Some of the earlier autumn conferences were preceded by monetary policy discussions that were rather intensive. In that respect the past year has been relatively calm, even though the situation has changed so rapidly.

Last autumn there was, of course, considerable anxiety but the Riksbank’s quick succession of interest rate cuts was well received. In Sweden, moreover, we were spared most of the European discussion about political pressure on monetary policy. As I see it, this was partly because our policy framework provided such clear arguments for our actions—inflation would turn out to be below the target if the interest rate was not reduced.

During the past year the Riksbank’s appraisal of economic activity and inflation prospects has been revised about as much as and more or less concurrently with the opinions of other observers. The sizeable changes in economic assessments are thus something we have in common with others. Individual economists and analysts have naturally voiced views or criticism from time to time but I at least have not encountered more concerted criticism of our economic and inflation assessments. (Figure 1)

The path of the krona last autumn generated more heated discussions. The Riksbank saw the depreciation as essentially connected with the international financial unrest. This made investors more averse to risk, with effects that hit small markets in particular. Other observers attached greater weight both to what they perceived as a lack of clarity in monetary policy and to structural problems in the Swedish economy and Sweden’s position outside the euro area. The risks in the latter respect were presented in a rather lurid light.

As regards interventions, the Riksbank has been very restrictive in recent years but early last October we found they were called for. The marked depreciation of the krona did not reflect the fundamental situation in the Swedish economy. It is seldom possible to demonstrate clear effects of interventions and that could not be done on this occasion, though the depreciation was broken during the day in question. Later in the autumn the krona appreciated again.

It seems to me that two main factors lie behind the krona’s appreciation in the past twelve months. One is a normal rebound from last autumn’s excessive weakening. This happened during the winter and was particularly pronounced because the euro’s introduction at the beginning of the year had made some investors concerned about the krona. The other factor is the increasingly clear picture of a strong economy with a budget surplus, low inflation and high growth. Moreover, the higher growth in Sweden compared with the euro area has influenced appraisals of the respective levels of interest rates. Specific events and statements may have played a part in the timing of the appreciation but it seems to me that the crucial influence has come from more fundamental economic considerations.

The discussion of monetary policy has also acquired a new dimension. The procedure with the new Executive Board from the beginning of this year enables Board members to differ from the current majority view of inflation prospects and monetary policy measures. Board members have in fact differed on a number of occasions, as is clear from the published minutes. On two occasions, divergent opinions have also been registered regarding the level of the repo rate. This has naturally affected analyses of the Riksbank’s behaviour and resulted in a somewhat new type of discussion about monetary policy and, not least, the Riksbank’s communications. That brings me to the next point.

The new Riksbank legislation has altered the conditions for how the Riksbank operates and communicates externally. The strong statutory independence makes it extra important that openness is practised as much as possible and that what we do can be examined and evaluated. These are basic arguments in the interests of democracy. The law makes the Executive Board collectively accountable for running the Riksbank as well as for its monetary policy decisions. The Board concluded that the important discussion ought to be held in session, not in corridors, and ought also to be minuted in such a way that it can be assessed.

Having established this, the way in which we should communicate externally more or less presented itself. Each Board member should be answerable for her/his opinion. The members’ individual responsibility is thereby clarified. So when one of us makes a speech, it is not normally a joint monetary policy signal but the expression of a personal opinion. I willingly admit that this was accompanied by purely practical considerations to do with the difficulty of coordinating six strong wills before every speech or interview.

There probably are not many other collective decision-making assemblies that publish their members’ positions so openly or so soon after decisions have been made. That is probably not happenstance. It may jeopardise internal cohesion and besides making monetary policy decisions, we have to work together in running an organisation. Then there is the risk that a policy of openness inhibits the internal discussion. (That is the background to the published minutes not reporting exactly what each member said at the meeting.) A third problem and the one that has attracted most attention has to do with our external communication.

Before commenting more on the third point, it is perhaps advisable to point out that these problems had already been recognised when the new system was launched. That I mention them here should not be taken to mean that they have caused concern since then.

The changes from the beginning of this year were prepared at the Riksbank last autumn. After contacts with the Bundesbank, the Bank of England and others, a structure was drafted for the communication of monetary policy. Eight to ten monetary policy meetings a year would be held at set times. Press notices and inflation reports in connection with these meetings would form the corpus of our joint communications. Divergent opinions would not be revealed before the minutes have been published some time after each meeting.

This arrangement does impose some restrictions on the outspokenness of Board members. In the interval between a meeting and the publication of the minutes, members have to refrain from commenting on what happened at the meeting and may not express an opinion that differs from the majority view of the forward-looking monetary policy. That is why we made it clear from the start that we aimed to shorten this interval, as we have now begun to do.

My impression is that this system has worked quite well. There are no indications that the market as a whole misinterprets the Riksbank more than before. What we do see is a greater personification of the monetary policy debate; but that was predictable. It tallies with the media tendency, particularly in broadcasting, to dramatise events by presenting conflicting opinions. Today, different points of view can be clearly attached to particular individuals.

One feature of this discussion is the classification of Board members as doves or hawks. It so happens that when I joined the Riksbank, one of the largest business journals actually described me as a dove-hawk (the literal translation of the Swedish for goshawk). I wonder whether the writer was aware that the goshawk is the most murderous of all the hawks and preys mostly on doves. With a central bank that clearly targets inflation, however, this way of classifying board members is not particularly apt. A member who consistently strives for an unduly low interest rate will find, if he or she gets their way, that sooner or later the inflation target is exceeded. The opposite holds, of course, for those who always prefer a higher interest rate.

I expect we shall have to learn to live with this type of monetary policy personification. But I hope that in time the discussion that follows our decisions and published minutes will still lead to a wider understanding, among people in general as well as in markets, of the deliberations and choices the Executive Board has to make.

For market players I continue to believe that fundamentally it is most important to arrive at their own sound assessments of inflation. Those who do so are likely to be fairly close to the mark in their predictions of the repo rate, at least in the somewhat longer perspective.

Let me now consider the present situation, starting from the Inflation Report the Riksbank presented last week.

The signs of a global recovery have been confirmed and also become stronger since the June Report. The Riksbank has accordingly made an upward revision to the forecast for GDP growth in the OECD area, to an annual rate of not quite 2.5 per cent up to the end of 2001. The growth is occurring on a wide front. The Asian and European economies are recovering, from different initial situations, and the slowdown in the United States is expected to be milder than we counted on earlier.

Even with the prospect of higher global economic growth, international price pressure is assumed to remain relatively low. One reason for this is that after the Asian crisis and its repercussions, there is still unutilised capacity in the world economy. Another important factor is increased price competition. In addition, the pass-through from external price increases to prices in Sweden will be dampened by an appreciating krona. In our main scenario, moreover, the barrel price of crude oil is assumed to fall back from the present level around USD23 to just over USD17.

The stronger international growth has also contributed to an upward revision of the Riksbank’s forecast for the domestic economy. Growth is now expected to be 3.6 per cent this year, followed by 3.8 and 3.0 per cent in 2000 and 2001. The upward revision also mirrors higher domestic demand growth. Firms as well as households are optimistic about the future. Real wages and employment are now rising rapidly and so is household income, particularly as fiscal policy is less restrictive. This is accompanied by rising corporate investment in the coming years. (Figure 2)

One component of the present picture that warrants the Riksbank’s attention is the development of asset prices. Share prices are high and a nation-wide increase in house prices has started in the past year. House prices in Greater Stockholm, for example, had already risen substantially before that. This increases the consumption and investment propensities of households and firms. To date, however, the levels reached by asset prices in Sweden are hardly a risk for economic development and stability in the way that some observers perceive to be the case in the United States.

All in all, the rate and breadth of the economic upswing in Sweden suggest that unutilised resources will be activated for production relatively quickly. Various indicators of resource utilisation suggest that certain capacity shortages may arise at the end of the forecast period. Unemployment falls to 4.6 per cent next year and to 4 per cent the year after, which means that we count on a higher rate of wage increases than before. In the coming years wages are now expected to rise 4–4.5 per cent. With the rapid increase in employment, moreover, registered productivity growth will be lower, particularly as most of the upswing is occurring in the services sector, where productivity is relatively lower. The forecast for this year has therefore been revised downwards and for the coming years the Riksbank counts on labour productivity rising in annual terms by just over 1.5 per cent. (Figure 3)

Against this background it has to be said that inflation expectations look good. There has naturally been some increase in the shorter run in the light of stronger economic statistics but the level remains low. Expectations for the medium and longer run are anchored around 2 per cent. (Figure 4)

The combination of all these factors led to the conclusion in the Inflation Report that in the forecast period inflationary pressure will be somewhat stronger. UND1X inflation, which excludes indirect taxes, subsidies and interest expenditure, is judged to be 1.8 per cent one year ahead and 2.1 per cent after two years. (Figure 5)

The Riksbank also attempts to identify any risks of developments that differ from the path in our main scenario. This time round we saw somewhat greater risks of inflation being above the target than below. The main reason for this is the uncertainty about how the stronger growth will affect the labour market and wages. To this I want to add another evident risk—that the price of oil does not fall back as we have envisaged. An oil price around USD20 in the coming years would lift inflation 0.1–0.2 percentage points above the main scenario. On the downside the main risk is a more marked slowdown in the US economy.

Now I want to touch on a matter that has been prominent in the debate after the publication of the Inflation Report. How will inflation be affected by the higher growth we foresee? The attention accorded to this question is only natural, for several reasons. The Riksbank now counts on growth that is clearly higher than we expected in June but has not made much of an upward revision to its assessment of inflation. Moreover, good growth is foreseen by virtually all observers and the consequences of this for inflation are therefore a central concern for everyone. But no one has the answer and that is what usually fuels a lively discussion.

The relationship between growth and inflation is a question that has occupied us at the Riksbank in recent years. Our inflation forecasts, as well as those of other observers, have almost consistently been above the outcome. The difficulty lies, however, in determining to what extent the lower outcome is attributable to transitory factors rather than to more fundamental behavioural changes in the Swedish economy.

In the simplest type of model, inflation is explained in terms of resource utilisation (however measured) and expected inflation.

One reason why the analysis of the relation between growth and inflation is complex is that the actual level of total resource utilisation is so difficult to assess. One approach involves estimating the long-term level of potential output and use the difference between this and actual production (the output gap) for conclusions about the relationship with trend inflation. In recent years the Riksbank, along with other observers, has revised its assessment of potential growth. We believe an improvement in the productivity trend has helped to raise the potential growth rate so that this is now somewhat above 2 per cent. We have also altered our assessment of unutilised resources; the largest revision was made during 1996, partly in the light of new information and better estimation methods.

The Riksbank’s assessment of resource utilisation is appraised continuously as new information is obtained about, for example, shortages in various sectors, prices and wage formation. An improvement in the workings of the economy can show up just as a larger supply of resources, be they capital or labour. To date in 1999, for instance, employment has been provided for more than 100,000 persons without any signs of appreciable shortages. To me at least this is a welcome surprise. But one should not make too much of it; if we have in fact seriously misjudged how the labour market is functioning, one would expect wage increases to be lower than has been the case. It is also worth noting that the rate of wage increases in Sweden continues to be higher than elsewhere.

Another difficult matter is the effect on inflation from a given level of resource utilisation. This was examined in a box in the latest Inflation Report. The results suggest that the Swedish economy can cope with somewhat higher resource utilisation than before without an acceleration of inflation. There is also a good a priori case for this. Raising prices, for example, should be more difficult in an economy where prices fluctuate less. It follows that firms should then put up more resistance in wage negotiations. It is also conceivable that increased international competition, for instance, has left firms with less room for price increases, at least for a time.

Finally, then, what about the role of inflation expectations? They have indeed come down substantially in the 1990s, from 6–8 per cent to 2 per cent. The Riksbank’s picture of what this could mean for inflation is partly based on empirical studies that have been presented in earlier Inflation Reports. These results likewise suggest that the economy is somewhat less inflation prone, though the picture should not be over-interpreted. In a world with low and stable inflation, it is conceivable that wage demands may be formulated differently. A given level of real wages can be attained with considerably smaller nominal wage increases and perhaps with greater precision.

Another factor that is often mentioned in the discussion is the concept of a new economy, that is to say, how the trade-off between growth and inflation may have been modified by increased global competition and information technology. This theme has been stimulated not least by the example of the United States, where high growth has been achieved year after year without boosting inflation; but that is a topic for a separate address. Today I shall simply make two points. One is that effects of this type can enable the economy to cope with higher resource utilisation without an acceleration of inflation, that they can raise potential growth or they can lower the inflation propensity. Thus, the so-called new economy can conceivably affect the inflation assessment in the ways I discussed earlier. The other point is that demonstrating effects of this type on inflation empirically is presumably even more difficult in Sweden than it is in the United States. This is because insofar as such effects have been at work, this has happened in a period when inflation expectations shifted downwards, making it difficult to separate the one from the other.

The inflation assessment in the latest Inflation Report incorporates a marginal downward revision of the trade-off between growth and inflation compared with the June Report. This could stand for some change in the appraisal of resource utilisation as well as a somewhat lower inflation propensity, though perhaps we have placed more emphasis on the latter. At the end of two-year period we discuss, the effect on inflation amounts to some tenths of a percentage point. The Riksbank considered the revision was called for in the light of the available information about the labour market situation and price tendencies, for instance. The IMF has made similar downward revisions for the advanced economies and they are mirrored in OECD area forecasts compiled by Consensus.

In connection with the publication of the Inflation Report, the Riksbank decided to keep the repo rate unchanged. The reasons were much the same as in August. Economic growth is admittedly higher, which reduces the unutilised capacity and thereby leads to an increase in underlying inflationary pressure. But inflation is still below the target and is judged to rise relatively slowly in the future, so that twelve to twenty-four months ahead it will be in line with the target. It is only towards the end of the period that the 2 per cent target is marginally exceeded if the repo rate is not increased.

A notable problem in the assessment of inflation concerns the future development of wages: it is difficult to predict, yet its significance for future inflation is crucial. Our wage forecast is in line with those of other observers; at the same time, we see a higher wage outcome as the biggest risk for inflation. A lower wage outcome would provide room for quicker growth and higher employment.

The decision to defer an increase in the interest rate is naturally based on our assessment of inflation in the coming years. It also enables us to collect additional information. At the same time, the conditions will be better for getting people in general to accept the step we are taking, which is ultimately essential for the Riksbank’s legitimacy and credibility. Neither did we see a need at present to make our mark by raising interest rates. There appears to be confidence in the Riksbank’s commitment to maintaining low inflation and the results of surveys and other work we do to see how various economic agents are assessing future inflation give no cause for concern.

In connection with the publication of the Inflation Report we made it clear that the repo rate will need to be increased in the future if there is no new information that clearly alters the perspective. Even if inflation expectations continue to be low and there are certain signs that the economy is functioning more efficiently, the growth rate will have to be brought into line with the long-term potential. That we shall do in good time. We can thereby contribute to good, stable growth in the Swedish economy for a long time to come.