Bergström: The Swedish economy-new or old?

The Swedish upswing has now begun in earnest and GDP growth is estimated at around 3.5 per cent per year during the next few years. In the course of 1999, it became increasingly clear that we were on the verge of an upswing. The negative atmosphere after the Asian crisis, the Russian suspension of payments and the problems in the hedge fund, Long Term Capital Management, had subsided and been replaced by optimism in many countries. The fact that the financial crises did not have such negative effects on demand as many had feared can probably be partly accounted for by the relaxation of monetary policy introduced by many central banks. The Riksbank was among these, reducing the repo rate from 4.4 to 2.9 per cent between spring 1998 and spring 1999. When the scenario of the Swedish upswing was confirmed and as resource use was accordingly expected to increase, it was natural to conduct a slightly more cautious monetary policy to avoid inflation accelerating.

Since November, the Riksbank has therefore increased the repo rate by 0.85 percentage points to 3.75. However, this is still an expansive monetary policy in the sense that the short real interest rates are so low that they should stimulate demand. And this is how it should be when 375,000 people are still unemployed and there is other unused capacity. As long as demand is lower than the potential production level, it can be allowed to increase more quickly than the potential growth rate without threatening the inflation target. According to the Riksbank's assessment, resource use is now increasing successively and will pass the potential level during the next few years.

It is easy to describe monetary policy in theory. Successful application of monetary policy in practice is difficult, however. As the interest rate changes do not reach full effect until 1½ to 2 years, monetary policy decisions have to be based on assessments of inflation quite a long way in the future. It is particularly difficult to make forecasts if the economy is undergoing major structural changes. Most forecast methods make use of experiences and correlation from earlier periods. During the nineties, the Swedish economy has been in a strong process of adaptation from a situation with major problems with central government finances, high unemployment and high inflationary expectations to an economy in which large parts can be described as being in balance. Central government debt has fallen from 80 per cent of GDP in 1995 to around 70 per cent in 1999, foreign debt has fallen from a maximum of 40 per cent of GDP in 1993 to only just over 20 per cent of GDP in 1999. The long-term inflation expectations have also been stable around the 2 per cent target for inflation for some time. It is true that unemployment continues to be high but it is falling. This is the most promising development for 30 years: We have strong growth at the same time as we have balance or large surpluses in central government finances and foreign trade.

The challenge for monetary policy is not to dampen demand unnecessarily but neither should demand be allowed to increase more than is compatible with the inflation target. In the long-term, monetary policy has, of course, no direct impact on the rate of growth although during a shorter period, demand and growth are influenced by the interest rate set by the Riksbank. If, for instance, monetary policy is made more stringent too late, there is a risk that inflation expectations will rise and gain a hold. To restore inflation to the target, a longer period of stringent monetary policy will then be required with weaker demand as a result. Then investment will also be throttled and capacity growth braked.

The inflation target has not only a value per se but is also first and foremost a means for creating the prerequisites for stable, sustainable growth. The best contribution the Riksbank can make is to maintain the inflation target but there is no simple rule of action as to how monetary policy is to be conducted. According to the Riksbank Act, the overall objective for the activities of the Riksbank is to maintain stable prices. However, according to the law, the Riksbank shall also support the objectives of general economic policy without setting aside the inflation target, in order to maintain sustainable growth and high employment.

As the Swedish economy is approaching balance in many respects and the long-term inflation expectations accord with the target, the conditions for carrying out monetary policy are more favourable in many ways than before.

This does not prevent there being many components of uncertainty now as well. Major structural changes are taking place in the world economy to some extent driven on by information technology. Deregulation, EU membership, and stiffer competition are other factors that can be assumed to have an effect on the economy's mode of functioning, although it is difficult to say how and over what period.

I am planning to take up some factors here in the risk scenario that must be taken into consideration in the ongoing evaluation of monetary policy. One issue concerns the extent to which we, inter alia as a consequence of the development of information technology, will see a higher productivity growth than we had expected and which, all other things being equal, would create scope for higher demand without threatening the inflation target. Another way of putting this question is whether we can expect to be on our way into something that is called "the new economy" in debate. Another issue which I have discussed in more detail in earlier speeches is that there is reason to fear that wage formation will not be able to cope with limitations set by the inflation target and productivity growth. How great, for instance, is the risk for the old pattern of wage spirals and increasing inflation being repeated? In this case, the Riksbank may be obliged to carry out a more stringent monetary policy to counteract growing inflationary pressure by subdued demand.

My ambition is not to give any answers but rather to point out what we have to take notice of in future.

An important component of the analytical framework of monetary policy is accordingly to make forecasts of potential growth. The supply of labour and productivity growth are important determining factors for how much we can produce in the long term in our country. A higher potential rate of growth creates, as I said earlier, scope for greater demand without inflationary pressure arising.

During the 50s and 60s and until the beginning of the 70s, the Swedish economy was characterised by very strong growth and very high productivity growth. Labour productivity grew from 1960-74 on average by as much as 4.3 per cent. At the same time, inflation was low during these years. The gross investment ratio was high and a large part of Sweden's incomes were invested during this period in expanding the productive capacity of the economy. Demand for goods could be satisfied at the same time, without the balance of payments showing large structural deficits.

In the mid-70s, productivity growth declined and it has since then been considerably lower than during the good years earlier and from 1974 to 1994, the growth in labour productivity remained at an average of 1.4 per cent. At the same time, the economy was characterised by high inflation. The low productivity growth was reflected by a drastic decline in the gross investment in relation to GDP in the mid-1970s. A smaller part of the national disposable income was invested to build up the capital stock (see diagram 1 and 2). Net investment measures the addition to the capital stock. It fell on average from an average of 16 per cent of disposable national income during the 1950s and 1960s to just over 5 per cent at the beginning of the 1980s. During this period, we also continually borrowed abroad to meet demand. Since 1995, productivity and the investment ratio in Sweden have increased again, although the improvement is modest and start from a low level.

In the USA, the picture of growth in labour productivity has in rough outline displayed the same pattern over time as the Swedish with one important difference. The improvement of labour productivity since 1995 has been larger in the USA than in Sweden in contrast to the earlier twenty-year period when labour productivity increased by about as much in both countries. In the USA, productivity growth between these periods has increased from 1.4 to 2.9 per cent while productivity growth between the same periods has increased from 1.4 to 1.9 per cent in the Swedish economy (see diagram 3) .

In the USA, the improvement in productivity has been preceded by strong increases in investment throughout the 1990s and investment during this upswing has contributed to GDP-growth to considerably higher extent than during previous upswings. The great productivity increase is one of the reasons why demand has been able to grow so strongly and for so many years without inflation accelerating. The questions raised are what underlies the U.S. development, and whether there is anything to indicate that we will see the same development in Sweden. The new economy does not seem to have come as far in Sweden as in the USA. We see no equivalent to the investment boom in the USA or even for that matter to that in Sweden in the 1960s. The information technology will have its main impact first when the new technology has penetrated the whole economy. To get that, huge investment activity is needed which we still are waiting for.

There are many explanations for the successes of the American economy. In common to most of these is reference to the economic policy carried out since the end of the 1980s. Price stability and a fiscal policy focused on budget consolidation have been highlighted as important components of the successful policy. The development of information technology and growth of IT related investments have probably been promoted by the stable economic environment and been an important driving force behind the unparalleled upswing.

It is not then possible to see any clear signs in the Swedish economy that we are on our way into a development similar to the American. At the same time, there are similarities, both with regard to the reorientation of the macroeconomic policy and to the penetration of computers and information technology in Swedish society. An important difference between USA and Sweden is that use of resources in the American economy has been high for many years while the Swedish economy was in a downturn for practically the whole of the 1990s. Investment usually increases only after a long period of high capacity use. The focus on stable prices and budget consolidation also started later in Sweden than in the USA. A very benevolent interpretation could be that we are moving in the same direction as the USA but with some time lag. The USA also waited for a long time for the productive effects of many computers introduced everywhere in the economy. The conclusion that can be drawn now is that it is anyway still not possible to base monetary policy on a real acceleration of potential growth.

The strong increases in prices in the Swedish stock market indicate that there are expectations among those who handle risk capital that profits will increase sharply particularly in Swedish IT-related companies. In order for such a development to be realised, it is reasonable to assume that investment must increase on a broad front. A risk with a strong increase in share prices before any visible signs of potential growth accelerating is if consumption is strongly stimulated by households who own shares increasing their wealth. An increase in demand that is not corresponded to by an increase in potential growth increases inflationary pressure that the Riksbank must counteract by increasing interest rates. However, increased interest rates also subdue increased investment, which is unfortunate, as it would brake the growth of capacity.

It has been difficult during earlier periods of large increases in share prices to prove that Swedish households have really increased their consumption as a result of increased wealth. During recent years, however, the proportion of households owning shares has increased. When property prices increased sharply at the end of the 1980s, many households raised loans on their homes and used the money for consumption. A low housing production combined with growth that is strongly concentrated to the big city areas has contributed to housing prices again rapidly increasing in recent years. This development can also stimulate consumption with the risks that I have just mentioned.

A further factor of uncertainty for assessment of inflationary pressure in the ongoing evaluation of monetary policy is wage formation. Here it concerns the risk that the wage formation processes have not yet been tested in such a strong cyclical upswing as we now seem to be in. The mode of functioning of the labour market is also important for how quickly the Swedish economy can make use of the opportunities created by technical innovations.

The nominal wage increases must in the long term be kept within the scope permitted by productivity growth and the 2 per cent inflation target. Higher wage cost increases will initially either lead to reduced profit shares for businesses or higher inflation. However, sooner or later, excessively high wage costs will lead to lower growth either because the Riksbank is obliged to counteract an increased inflationary pressure with higher interest rates or to a reduction in the willingness to invest in Sweden confronted with the prospect of reduced profit margins.

This year and next year pay negotiations are to take place for 2.3 million employees and the mode of functioning of the labour market will then be seriously put to the test. The Swedish labour market has previously had a tendency to generate wage-wage spirals. It has, for instance, been difficult to change relative wages without this generating compensatory wage increases. The fact that the coming pay negotiations are taking place in a situation where it is expected that there will be a high demand for labour increases the risk of the historical pattern being repeated. In some industries and regions there are indications that a shortage of labour is developing. However, it is at the same time important to bear in mind that we are a long way from the scarcity levels experienced at the end of the 1980s.

Another challenge for wage formation is that the first pay negotiations taking place are for public sector employees. In this sector, there are large groups that have obviously fallen behind in their pay development, leading to shortages. Among these are certain categories of teachers and many health care employees. The Swedish labour market has not only had problems in dealing with pay balances between different occupations. It has also found it difficult to tackle regional imbalances in the same occupational category through adaptation of relative wages. Shortage of labour has, for instance, particularly affected big city areas, although it has hardly affected relative wages between areas where there is shortage and where there is unemployment, despite the cost of living and not least housing costs being lower outside the big cities. With a more flexible wage formation, the imbalances in the labour market would probably be reduced.

The starting situation for the pay negotiations is thus not unproblematic. However, although there are elements that cause concern, there is also a lot to indicate that the process of price formation has changed.

The most important change since the 1970s and 1980s is probably the change in the target for monetary policy and the support and credibility attained by the inflation target. I am convinced that the Riksbank's inflation target is understood and accepted by trade union members, officials and leaders. Knowledge about the inflation target and the role of monetary policy has become widespread in a quite different way to the fixed exchange rate norm. The fixed exchange rate regime basically required that wage cost increases should be adapted to the surrounding world in order not to affect employment negatively. When, despite this, wages rose more than in our competitor countries, employment was saved by recurrent devaluations. The high level of activity in the Swedish economy during the latter half of the 1980s also played an important role. No wage formation system would have coped with the overheating that existed at the end of the 1980s.

The insight that excessively high wage increases rather quickly lead to growth being dampened and an increase in unemployment with a monetary policy focused on an inflation target, probably serves to have a restraining effect on wage formation.

High credibility for the goals of monetary policy has led to a reduction in inflation expectations. And stable inflation expectations per se reduce the risk for pay negotiations leading to higher wage increases than there is room for. Stable inflation expectations reduce the uncertainty of the parties as to what a nominal wage demand means for the real wage. This can in turn reduce the risk for wage spirals.

Increased competition in globalised markets could also contribute to increased flexibility in wage formation. Globalised highly competitive commodity markets probably reduce the ability of businesses to compensate high wages by price increases. At the same time, it is probable that the strongly globalised capital market makes it more difficult for businesses to press down profit margins even in the short run. A conclusion is that businesses' opposition to wage costs increases that cannot be sustained by productivity increases may have increased.

Another change in the Swedish labour market is the tendency towards increased short-term employment. This could eventually increase the flexibility of wage formation and increase productivity. This development is particularly in evidence in the expanding services sector. According to the National Institute of Economic Research, the proportion of time-limited employment has increased from 10 to approximately 16 per cent of the total number of employees between 1990 and 1998. This development may have been reinforced by the difficult situation in the labour market during these years but many believe that the trend will continue. New methods of compensation are also spreading. Bonus wages, profit sharing and a 13^th monthly wage are examples of components in wage formation that increase flexibility in the labour market. Increased cost consciousness due to stiffer competition can be a factor driving on this development. Although short-term employment in various forms may reduce the inflationary pressure and make the labour market more flexible, this does not mean that this development is always positive for wage-earners and households. Uncertainty about future income makes long-term planning more difficult such as having children and buying a house.

In order to reduce the risk of various groups in the labour market competing against one another to drive up pay increases in a way that in the end creates a poorer outcome for all, various changes in the forms of wage formation have been carried out and are planned. The most important is the so-called Industry Agreement between some 20 organisations in the labour market. This agreement thereby affects some 20 per cent of the labour force and includes rules that reduce the risk of periods without agreements and rules that reduce the risk of conflicts breaking out.

In the state sector, adaptation of pay bargaining has been facilitated by the frame budget process, which entails that wage development in the export industry shall set the level for state wage costs.

With the intention of facilitating wage formation compatible with economic balance, the Government introduced a bill in December last year that included proposals to create a new agency, the Mediation Institute. The bill includes measures that aim to reduce the risk of conflict breaking out and are to come into effect if contractual solutions are not reached in some other way.

The state of the Swedish economy is better than for a very long time.

Inflation expectations have stabilised around the target and economic policy is focused on continued consolidation of public finances. The prerequisites for the Riksbank to conduct monetary policy so that it contributes to stabilising growth around its long-term path are thereby improved.

However, this does not mean that challenges are lacking with regard to striking the right balances in monetary policy. I have discussed some of these. One is the assessment of productivity growth and the potential growth in the next few years. There are still no signs that we will see a productivity growth such as in the USA. However, we must carefully follow and analyse the development in the area where monetary policy is formulated. The risk of incorrect assessments having more serious consequences should, however, be small as changes in this area do not take place overnight.

Another component of the risk scenario is how wage formation and the labour market will function in coming years. Negotiations are taking place in an economy undergoing strong growth at the same time, as there are occupations with a shortage of labour where pay has fallen behind. However, changes have taken place in the Swedish labour market that permit some optimism about developments. Above all, this concerns the support for and understanding of the inflation target and the low inflation expectations that have been established. Changes in the forms of wage formation, which have been carried out or are planned work in a positive direction. Finally, globalisation and competition can have increased businesses' cost consciousness. This may lead to a greater opposition to wage increases and also that businesses in their drive to keep costs down endeavour to introduce forms of employment that increase flexibility in the labour market.

In conclusion, it is important for the Riksbank as always to carefully monitor what is taking place in the economy when conducting monetary policy.

You will find diagram 1 and 2 in the enclosed Word file and diagram 3 in the enclosed PowerPoint file.