The current situation for monetary policy
My thanks for the invitation to this Committee's public discussion of tendencies in the Swedish economy and the current situation for monetary policy. My opening remarks today broadly follow the earlier arrangement. Having presented them, I shall be prepared to answer questions.
Economic improvement
During the autumn there have been more substantial signs that activity in the Swedish economy is improving. Conditions for a continued improvement in economic growth also appear to be favourable. Nevertheless, in the coming two years it looks as though inflationary pressure will be moderate. Moreover, inflation expectations have remained low, which indicates confidence in Sweden's economic policy. Furthermore, there are many signs that Sweden's credibility as a country with low inflation is on the way to being established. All this adds up to a good possibility of inflation being in line with the price stability target. In any assessment of the future, however, allowance has to be made for risks and probabilities. I shall be returning to this shortly but first I want to consider some real economic tendencies.
The latest GDP statistics indicate that Sweden's economic growth in recent years has been higher than estimated earlier. Upward revisions have been made as regards both 1994 and 1995, as well as the first half of 1996. Between the summer of 1993, when the curved turned upwards, and the summer of 1996, GDP growth in annual terms is now put at 3 per cent. To date, therefore, in the 1990s the economic recovery has been more rapid than in the early '80s. At the same time, inflation has been considerably more subdued and even below the real growth rate. Not since the 1960s has Sweden had such a long period in which growth has been higher than inflation. A number of factors have probably contributed to this.
In the first place there is the obvious explanation that when the upturn began in the 1990s, total resource utilisation was lower than in the early '80s. This tended to curb inflation at the same time as the economy could expand more rapidly for a time as unutilised resources were activated.
But there are also underlying factors which may act more permanently. While it is still too early for any definite conclusions, it is conceivable that the Swedish economy's growth potential today is higher than in the 1970s and 80s, while its inflation propensity in a wide sense may have decreased. Support for these hypotheses is to be found in a number of interacting factors.
The higher overall growth rate in the 1990s stems from upward tendencies in productivity, capital formation and average working time. Regardless of the period chosen for comparison, productivity growth in the 1990s appears to be some tenths of a percentage point higher than in the '80s. The difference may seem slight but if its lasts, the cumulative effect over time becomes substantial.
Improvements in the workings of the Swedish economy may have been aided by the tax reform, various deregulations and altered replacement levels in the social security system. The low inflation is no doubt also helping to make the economy function better. In addition, competitive pressure has been strengthened by EU membership and other components of the globalisation of trade.
Another indication of economic tendencies with more permanent effects is the rapid rate of capital formation in recent years. Excluding the residential sector, gross capital formation has climbed to relatively high levels. While it is normal for industrial investment to pick up in response to rising capacity utilisation, compared with the 1980s the upturn in the '90s has been steeper and more pronounced.
New tendencies can also be discerned in the composition of foreign trade. The combined share for medical, pharmaceutical and telecommunication products has been growing rapidly in the 1990s. From 1990 to 1995 the pharmaceutical industry doubled its exports in volume terms, while telecommunication exports almost quadrupled. These developments have continued in 1996. Another new tendency is the notable expansion of exports to Asia. These are, of course, important reasons why our export performance has remained good in 1996 despite a weak tendency in some EU countries and the fact that the krona's appreciation has eliminated some of the improvement in relative costs that stemmed from its depreciation in the autumn of 1992. This can be taken as a sign that the Swedish economy has become more amenable to structural adjustment.
These are new and encouraging patterns in the Swedish economy which may herald a potentially higher long-term output. While the hypotheses are still tentative, they are pertinent to monetary policy because of their consequences for the assessment of inflation.
Although growth during the upswing has been strong, in the past year total production has tended to slacken. As I mentioned at the Committee's hearing in May, this is partly attributable to the stocks cycle.
During 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 a situation where steeply rising demand leaves them unable to meet customers' requirements. In time, firms find they have built up surplus stocks and initiate a downward adjustment. This has the opposite effect of curtailing production in relation to final demand.
It was this type of stock adjustment that occurred in Sweden towards the end of 1995 and continued into the spring. It had also occurred somewhat earlier in certain countries of importance for Swedish exports. Such stock adjustments are not unusual during this phase of an economic recovery. They happened at much the same juncture in the 1970s and '80s.
The pattern is evident in the figures from Statistics Sweden. The growth of total output between the first halves of 1995 and 1996 is estimated to have been only 1.6 per cent. Production was held back by a downward adjustment of stocks that corresponded to 0.9 per cent of GDP. This implies that total demand actually rose 2.5 per cent. Firms achieve their desired adjustment of stocks by modifying production in relation to total demand.
Temporary production cuts may lead firms to lay off and even discharge labour. This has been the case in 1996 and is one reason for the weak tendency in the labour market. But if the underlying conditions for expansion remain good, the growth of production will pick up again when the adjustment of stocks is over. But there could also be a situation where underlying demand suffers a setback, for instance because a weaker labour market undermines households' expectations and thereby tends to curb private consumption.
It was this risk that prompted the Riksbank to supplement its main analysis last spring with a weaker alternative that pointed to a further easing of inflationary pressure. To date, however, the statistics suggest that the probability of such a scenario has decreased. The stock adjustments seem to be coming to an end and, despite the weak labour market, private consumption still appears to be on a stable upward path. That is one of the reasons for our assessment that a recovery in production is on the way. The signs of an improvement in activity abroad are another indication of this, as is the fairly sustained export growth to date this year.
A low inflation regime is being established
Another central feature of the picture is that inflation expectations, for both the short and the somewhat longer term, have moved down and are increasingly in line with the target. This implies that households, firms and financial agents now believe in Swedish economic policy's determination and ability to keep inflation down. In other words, during the past year the credibility of Sweden's economic policy has been appreciably enhanced. Confidence in the objective of price stability has to do not just with words but also with actions. The Riksbank's tighter interest rate policy between the summers of 1994 and 1995 demonstrated that we take the price stability target in earnest, while the consolidation of government finance is evidence that the political system is prepared to create long-term conditions for the maintenance of this objective.
As I have pointed out on several occasions to this Committee, if current inflation is kept in line with the target and a move towards sound government finance strengthens long-term confidence in economic policy in general, then sooner or later this will have an appreciable positive effect on bond rates and the exchange rate. During the past year we have experienced this type of confidence shock. Since the summer of 1995 the krona has appreciated about 13 per cent and the five-year bond rate has moved down almost 4 percentage points. This has had sizeable one-off effects on consumer prices. Import prices and house mortgage costs have fallen steeply. Here we have a dominant reason why the latest 12-month increase in consumer prices amounts to only 0.2 per cent, which by itself is below the downward tolerance interval for the price stability target.
However, given that the underlying rate of inflation is around 2 per cent, various types of temporary price shocks should normally be absorbed inside the tolerance interval. Moreover, the target is to be interpreted not in terms of isolated monthly figures but in an annual perspective. But it is conceivable that even for 1996 as a whole, the price trend may be somewhat below the lower tolerance limit.
An analysis of the present low rate of inflation in terms of the price stability target calls for an assessment of the underlying or trend rate of inflation today and its likely future course. With an approximate correction for the transitory effects from import prices and house mortgage costs, the underlying rate appears to be around 2 per cent; with all else equal, this means that registered inflation may gradually move up to that level as the transitory effects drop out of the statistics. I should immediately add that, for reasons to which I shall be returning, the various indicators of underlying inflation are very approximate.
Are wage increases and rising consumption inflationary?
Assessments of future inflation have to incorporate risks and probabilities. One risk that we have highlighted is wage formation. In the first half of 1996 it seems that the overall wage level, corrected for some transitory and statistical factors, moved up from the first half of 1995 by around 5 per cent. This represents an acceleration from the annual rate of just over 3.5 per cent for the second half of 1995.
In relation to the 2 per cent inflation target, a wage rise of just over 5 per cent gives a real wage increase of 3 per cent. Considering the current trend for productivity, this is a good bit above the economic margin. In the longer run, a development of real wages that deviates from the productivity trend is not compatible with the inflation target.
Today the Riksbank does not see any immediate risk of the existing wage agreements posing an appreciable threat as regards inflation in the coming two years. It is important, on the other hand, that an adjustment is achieved in the coming round of wage negotiations. This leaves the question of how the high wage increases should be interpreted.
The pessimistic interpretation is that wage increases will remain high in the future. In that case, in the absence of extensive changes in the structure of the labour market so that this functions appreciably more efficiently, unemployment would continue at a high level. One ground for the pessimistic alternative is that wage increases on top of negotiated rates have accelerated in 1996 even though the labour market has been weak.
The main hypothesis adopted by the Riksbank is somewhat more optimistic, though it does not obviate the need to improve the workings of the labour market. The outcome for wages may have been conditioned by the settlements having been concluded before the credibility of the low inflation policy had been established; the 1995 wage negotiations were coloured by the contemporary inflation expectations, which were close on 4 per cent in annual terms. Moreover, in the spring of 1995 export demand was rising strongly and profits in the internationally oriented sector were very favourable.
If this hypothesis is reasonable, the rate of wage increases can be expected to slacken by degrees as the present agreements are replaced. However, the coming round of wage negotiations will be less concentrated in time. Two of the current settlements expire before the end of 1996. There have admittedly been statements by employee representatives that, if quoted correctly, are very remarkable. Allow me to cite (from Finanstidningen, 14th October 1996): ‘That inflation is low today is of no major importance for our wage claims’ and ‘There is no direct link between our coming wage claims and the number who are unemployed’. It is possible that the statements should be seen in the light of tactical considerations in the run-up to the negotiations, in which case they can be treated accordingly. But if they reflect an approach to wage formation that is representative of the employee organisations, it is clear that the labour market is suffering from larger problems to do with game rules, in which case something will have to be done about this in the general context of economic policy.
For the time being the Riksbank is adhering to its main hypothesis that the future will see a successive adjustment of wage increases. But there is every reason closely to follow the negotiations that will soon be beginning.
Another risk might be posed by an upswing for private consumption that is appreciably stronger than the Riksbank and most other observers expect. Considering how the financial situation of households has improved in recent years, accompanied by a weak development of consumption, a faster increase in consumption cannot be ruled out.
There may be an accumulated need to renew stocks of durable goods. Compared with other countries, Sweden's stock of cars, for example, has a high average age. However, experience from other countries also shows that the inflationary effect of rapidly rising consumption depends on a variety of factors, not least the level of unutilised resources. At present, therefore, a more pronounced upswing in consumption need not lead of itself to rapid price increases.
A decreased inflation propensity?
In contrast to these possible signs of inflation risks there are primarily the indications I mentioned earlier that the Swedish economy's inflation propensity in a wide sense seems to have decreased. Additional information that points in this direction has become available during the autumn. Although the current, very low inflation is mainly a consequence of transitory factors, it cannot be attributed to these alone. Some, albeit limited part of the fall in inflation seems to be of a more permanent nature.
I said earlier that the extent to which inflation is affected by various transitory factors cannot be determined exactly. The Riksbank's overall assessment is that inflation's underlying rate is now somewhere between 1.5 and 2 per cent.
If the inflation propensity turns out to be lower than the Riksbank has allowed for in the assessment of future inflation, this could have a positive effect on our inflation forecast. However, more information and time are needed in order to draw conclusions in this respect.
Monetary policy
In the course of 1996 there has been a sizeable adjustment of the repo rate. Since the series of cuts began, the level of the repo rate has been virtually halved, from 8.91 per cent to the current rate of 4.6 per cent. This, however, should be seen in the light of the favourable outlook for inflation as well as the krona's appreciation.
The repo rate has been lowered in small but continuous steps. The cuts have acquired a pattern. For the Riksbank it has been important to obtain continuous confirmation of our analysis from incoming statistics and also to sense that our assessment is shared by market agents. Ultimately it is the degree of confidence in economic policy that determines the Riksbank's room to manœuvre.
When the interest rate corridor and the repo rate were lowered on October 24th, the Riksbank announced an end to the pattern of regular cuts. To make this clear, a fixed repo rate of 4.60 per cent was set to apply for four weeks. At the same time the comments underscored that we did not want attention to focus strongly on the repo rate that is due to be set immediately afterwards.
The background to this is that while we perceive some limited room for a further lowering of the repo rate, the picture is not entirely unambiguous. We need time in which to follow and analyse incoming statistics. The conclusion to be drawn from the assessments that are made may also be an unchanged repo rate.
