The current situation for monetary policy
My thanks for the invitation to attend the Committee to discuss developments in the Swedish economy and the current situation for monetary policy.
Better conditions for sustained growth
The Swedish economy had another good year in 1995 (Diagram 1). GDP growth reached 3 per cent and industrial production rose as much as 9 per cent. Other positive factors were a continued expansion of investment by more than 10 per cent and a current account surplus of over 2 per cent of GDP. There was also a further reduction of the general government deficit.
The number in employment rose by 60,000 persons but unemployment remains high. From the low in 1994, the number in employment has increased by 100,000, which means that of the earlier loss of over 500,000 jobs, only 20 per cent have been recovered. In terms of hours worked, however, the picture is somewhat different; here, as shown in Diagram 1, the increase makes up for almost half of the earlier fall. The increase accordingly comes from longer average working time.
A continued economic improvement, with increased employment, presupposes sustained growth, which calls in turn for low inflation and confidence in Sweden's general economic policy. The earlier situation, with uncertainty about economic policy, resulted in a weak exchange rate and high bond rates. That made it difficult for the economy to get onto a path where growth would be balanced and sustained.
Since the upturn in the second quarter of 1993, GDP growth in annual terms has admittedly averaged 2.7 per cent. But three-quarters of this expansion has come from the industrial sector, which represents only about 20 per cent of the total economy.
This has altered the structure of the Swedish economy and generated a relatively large surplus on the current account. However, as the unutilised industrial capacity was soon activated, inflationary tensions showed up at an early stage in the upswing. Producer price increases accelerated, cost consciousness waned and difficulties arose in wage formation. The idea was spread that inflation would accelerate again, which meant that the higher inflation expectations were on the way to being fulfilled. In order to safeguard the inflation target, the Riksbank therefore raised the instrumental rates in the period from August 1994 to July 1995. The alternative would have been higher inflation, which would have had negative effects on future growth and employment.
The inflationary tendencies in the industrial sector have now subsided. The rate of inflation has been shepherded onto a path around 2 percent. This spring, the consumer prise rise has actually fallen back to 1.7 per cent.
The fact that inflation is now at its lowest level in more than 35 years is an achievement for economic policy. The downward movement of interest rates and the strengthening of the krona since last summer are indications that Swedish economic policy has earned increased confidence for a more permanent focus on price stability. We must not forget, however, that the improvement has benefited from the stability in international financial markets. Diagram 2 shows how the differential between German and Swedish bond rates and the value of the krona have developed since 1994. The fundamental conditions for a future with sustained, balanced and non-inflationary growth have accordingly improved.
The benefits of keeping inflation in check have become clear in the past six months. Previously, international slowdowns often coincided with a phase of high and rising inflation in Sweden. In the 1970s and '80s this tended to accentuate activity's downward path. Today the low inflation has given us a better starting point. During the spring the Riksbank has been in a position to lower the repo rate relatively quickly. The nominal instrumental rate is now at its lowest level since the early 1970s. The development of the repo rate since 1994 is shown, together with the money and bond rates, in Diagram 3.
During the spring it has been claimed that in lowering the repo rate, the Riksbank acted too late. Late last autumn, however, none of the more prominent observers of the Swedish economy envisaged that inflation would be in line with the target in 1996 and 1997.
For the Riksbank it has been vital, both now and earlier, to keep the inflation target firmly in sight. The easing of interest rate policy was initiated only when we saw that conditions had changed. The repo rate does not control the spectrum of market interest rates in an automatic manner. The possibility of influencing the structure of interest rates - above all the rates for securities with a long maturity - is based on the existence of confidence in economic policy. A premature cut in the repo rate would have generated confusion about the long-term direction of monetary policy, not least in the light of Sweden's unfavourable history of inflation. Such action by the Riksbank could have led to higher bond rates and a weaker exchange rate this spring or later on.
Neither should we be tempted to believe that by fine tuning economic policy it is possible to eliminate either cyclical or quarterly fluctuations in growth. Nor can we escape the fact that Sweden's economy is small and open, making it heavily dependent on developments in the rest of the world.
But a policy that focuses on keeping inflation in check can help to modify the fluctuations in activity. The greater the confidence that budget and monetary policy inspires, the more assured can households and firms feel about the future and the greater will be the Riksbank's room to manoeuvre in the short term.
Favourable outlook for inflation
In January this year the consumer prise rise fell back to 2 per cent and then in February and March to 1.7 per cent. A breakdown of the CPI shows that the slowing of inflation in the early part of 1996 came mainly from prices of imported goods and mortgage interest costs. According to our measurements, the price rise generated in Sweden has also tended to decline during 1996; but development in this respect has been largely stable since the retardation of inflation began in 1992 (Diagram 4). The difficulty in isolating the full impact of the krona's devaluation on import prices probably accounts for the hump during 1993 in the diagram. The domestic price component also includes temporary effects of increased indirect taxes and decreased subsidies in connection with budget consolidation. It follows that for several years, domestic inflation, measured in this way, has shown a trend of around 2 per cent.
A shift in the rate of inflation, upwards or downwards, from this 2 per cent level would require either a shift, upwards or downwards, in inflation expectations or a change in resource utilisation so that this exceeds or falls short of the long-term potential, or else a combination of the two.
Turning first to the role of inflation expectations, it is generally the case that the longer the period during which inflation remains in line with the target, the more probable it becomes that a growing proportion of the economic agents will be confident that this state of affairs will continue. The expected rate of inflation is ultimately a reflection of the confidence in economic policy and its consistency. The fulfilment of the convergence programme is therefore of the utmost importance. Wage formation also plays a part.
Lower wage increases also improve the conditions for combining low inflation with a reduction of unemployment. Of course the construction of monetary policy is likewise important for the pattern of expectations. After all, inflation is ultimately a monetary matter.
As regards the development of demand, it seems that activity in the first quarter of this year was weak. Many economic observers have also made downward revisions to their forecasts for 1996. The assessments of this year have changed comparatively quickly, which is partly explained by a slowdown in international demand. As recently as last December the National Institute of Economic Research, for example, counted on growth in the region of 2 per cent both this year and next, accompanied by a successive acceleration of inflation to the upper part of the target's tolerance interval in the course of 1997.
Broadly speaking, the future outlook points to two alternatives: a main alternative in which activity in the Swedish economy gradually picks up again in the course of 1996, and a secondary alternative in which the recovery is more protracted and perhaps weaker. Permit me to discuss these two alternatives.
In the main alternative the persistently favourable development in the United States and an upturn in Japan contribute to successively stronger activity in Europe. The lower inflation in Europe has led to an easier monetary stance; together with lower bond rates, this is tending to stimulate Sweden's neighbouring economies. Strong growth in Europe and in the global economy as a whole favours Swedish exports.
An appreciable part of the cyclical fluctuations is a consequence of stockbuilding. From the picture of GDP and final demand in Diagram 5 it will be seen that in upward phases production tends to rise faster than demand; this was also the case in the latest recovery. The reason is that firms generally want to avoid being unable to supply customers when demand is rising steeply. Conversely, surplus stocks accumulate when demand slackens; during the subsequent stock adjustments, production is curtailed more sharply than final demand. The advances in information technology have enabled firms to handle materials more efficiently. A common example is Just-in-Time, whereby inputs are supplied exactly when they are needed in the production process. Even so, stockbuilding is quite often a sizeable factor, positive or negative, in the development of total production.
Better international demand can mitigate the contractive effect of the stock cycle that would be a normal feature of activity's current phase. Similar periods of stock adjustments occurred at much the same cyclical stage in both the 1970s and the '80s. Otherwise the underlying conditions for a continued expansion of the business sector in Sweden are favourable, with persistently good profitability and widespread investment activity.
Conditions for a cautious recovery may also be at hand for private consumption. Even with the budget reinforcements, household real income can be expected to rise in 1996 and 1997. Household balance sheets are also stronger than for many years. Moreover, there should be an accumulated need to replace durable goods.
When allowance is made for the methodological changes which EU harmonisation has entailed for the surveys by Statistics Sweden, it seems that households' assessments of their own economic future have become more positive. The earlier method shows that the net figure, which fluctuated around zero last autumn, has become more positive since the turn of the year. The new sample also includes people over 64 and this group turns out to be more pessimistic than those aged 16 - 64 years, who made up the earlier sample.
Retail turnover in January and February was weak compared with the same period a year earlier but this was some improvement from the closing months of 1995. But car sales appear to be weak.
Thus there are certain signs of some improvement in households' assessments of the future as well as in private consumption. It is too early, however, to draw more definite conclusions from this information.
As always in a slowdown, however, there are risks of the outcome being weaker. One can therefore consider a secondary alternative where weaker activity in Europe leads to a poorer development of exports and larger stock adjustments. That in turn could have repercussions on the domestic economy if it has negative effects on the labour market and weakens household confidence. It is also conceivable that wage increases elicit rationalisation in domestically-oriented sectors and affect unemployment.
Against this background, what is the outlook for inflation in 1996 and 1997?
The inflationary risks from the demand side would seem to be limited. This applies even in the main alternative. Behind this conclusion lie indications that the potential growth rate has risen somewhat compared with the 1970s and '80s, as well as some signs that the Swedish economy has become less inflation prone. But a further adjustment of wage formation is needed in the medium term. The demand-side risks of inflation are no doubt smaller in the secondary alternative than in the main alternative. Thus, given low inflation expectations and confidence in economic policy, the conditions remain good for inflation in line with the target in the coming years.
Future monetary policy
Generally speaking, the monetary stance can be neutral, expansionary or contractive. A neutral monetary policy is appropriate when expected inflation is low, capacity utilisation is normal and consequently there is a good probability of continuing to fulfil the inflation target. A contractive policy is suitable when there is a risk of inflation accelerating, while an expansionary tendency is admissible if expected inflation is low and resource utilisation is either low or falling, because there is then an increased probability of being below the inflation target.
Determining what constitutes a neutral, contractive or expansionary monetary policy is not a simple matter. For a small, open economy such as Sweden's, it is important that both the exchange rate and interest rates are included in the assessment of the overall monetary conditions. Assuming that the krona is in the process of adjusting to its equilibrium level, at present the monetary conditions in Sweden can reasonably be described as broadly neutral.
The short treasury bill rate is currently in the interval from 6 to 6.5 per cent. The rate varies slightly, depending on the time to maturity and on whether it is expressed in simple or effective annual terms. The latest 12-month figure for inflation is just under 2 per cent. The real money market rate can then be said to be approximately in the region of 4.5 per cent.
Experience from the past fifteen years suggests that such a level of the real market rate for money is somewhat higher than the historical international average. In Germany and the United States, for instance, the average since the early 1980s up to the present is around 4 per cent. For Sweden, however, the corresponding average is somewhat higher, just over 4.5 per cent. This probably reflects difficulties with confidence in the low-inflation regime on which the fixed exchange rate policy rested.
A G-10 study has shown that, since the 1960s, the real bond rate in the industrial world has moved up approximately 1 percentage point, from 3 to around 4 per cent. One important explanation is that problems with public sector deficits in many industrial countries have involved decreased global saving. Another is that the removal of administrative regulations has been followed by an international adjustment of the artificially low levels which the earlier regulations had entailed. In the 1970s, for example, it was not uncommon for real interest rates to be negative. Thus it is difficult to calculate historical real interest rates for lengthy periods.
The future course of the average real interest rate will be partly contingent on confidence in national economic policies and on the global balance between saving and investment. This means that the historical average will not necessarily hold in the future.
In general, as a consequence of the monetary policies of central banks, real money rates have fluctuated round the average level (Diagram 6). In the first half of the 1980s the US real rate was between 6 and 7 per cent; at times it was even higher. The German real rate was also high in the early 1980s but then fell back. In the late 1980s there was a tightening of monetary policies and the real rates moved up to 5 per cent in the United States and 6 per cent in Germany. As inflationary pressure was pushed down in the early 1990s, monetary policy was eased again in these countries. In the United States the real money rate bottomed at around 0 per cent, which is a very low level, and was then raised from the spring of 1994 onwards to around 3 per cent, which is its current level. In Germany the short real rate has been lowered from the end of 1992 to about 2 per cent.
As I pointed out earlier, an analysis of the overall monetary conditions must include exchange rate movements. It has been found that real money rates are often lower in countries where the exchange rate has developed more strongly than in other countries. This is one of several reasons why continental European countries now have lower real money rates than other countries. In Germany, for instance, the real effective exchange rate has appreciated relatively strongly in recent years.
In the light of this discussion and of the Riksbank's appraisal of the outlook for inflation, what conclusions can be drawn about future monetary policy in Sweden? Regardless of which of the two alternatives for the real economy that proves most probable, the Riksbank considers that there may continue to be some room for repo rate reductions. But the room for manoeuvre is contingent on the alternative that materializes, on the attendant inflation outlook, and on the confidence in economic policy. The Riksbank will therefore be keeping a close watch on developments.
