Heikensten: Monetary Policy and Wage Formation

In my introductory speech today I will discuss four issues. First, I will present the argument for adopting price stability as the Riksbank’s chief objective and say some words about the Swedish debate on this matter. Then I will comment on developments during the last decade, when we abandoned the high-inflation policy of the 1970s and 1980s. My third subject will be the current monetary policy situation and the choices that we now have to make. Lastly, I will comment briefly on wage formation and the labour market in the light of this situation.

Discussion of the price stability objective

In the early 1990s, there was a shift in Sweden’s economic policy towards establishment of a low-inflation regime. A few years later, price stability was established as the objective of monetary policy. This change was the result of a realization among policy-makers that the policy of high inflation and recurring devaluations that had been pursued in recent decades had failed. After 20 years of stagnating prosperity and major imbalances, it was time for a change of policy. A contributory factor was the changing view of the potential of monetary stabilization policy since the late 1960s, as a result of which almost all industrial countries adopted a price stability objective during the 1980s. To continue to pursue an independent monetary stabilization policy was hardly an option for Sweden, especially not if it was thinking of joining the EU.

An important task for the Riksbank was to gain acceptance for the change of policy and create understanding for our work and our independent status. This has not been all that easy. I think there are several explanations for this. One is quite obvious. The changeover to low inflation was accompanied by a sharp downturn in the Swedish economy. Production declined for a period of about two years and there was a dramatic increase in unemployment. The blame for this was laid at the door of the policy shift. That was of course unfair, although I must admit that the decline was aggravated by some policy mistakes. The basic problem was that the economic policy that was pursued in the second half of the 1980s resulted in imbalances, in particular, bubbles in the financial and property markets and an overvalued exchange rate. It was necessary to correct these imbalances, and this process would have been costly in any circumstances.

However, these were not the only reasons why it proved difficult to gain acceptance for a low-inflation policy in Sweden and why the thought of entrusting monetary policy to an independent institution raised doubts. The Keynesian tradition has been extremely strong among Swedish economists and across party lines ever since the 1930s. Therefore, the idea that the state can continuously control or influence the economic situation has probably been more persistent here than elsewhere. Our strong parliamentary tradition, which is based on the idea that all power derives from Parliament, has also played a part. Countries where power is shared more evenly among the political institutions have found it easier to accept the delegation of monetary policy to an independent authority.

Having said this, it is worthwhile recalling that there is in fact a completely different Swedish tradition. The policy shift of the early 1990s is in fact a return to a long tradition of price stability that was temporarily interrupted for a few decades. Between 1870 and 1970, the average rate of annual inflation was about 2 per cent.

Without wishing to labour this comparison, it is perhaps nevertheless worthwhile emphasizing that the twentieth century was a time of unprecedented economic growth; only Japan’s economy grew faster than Sweden’s. But this historical pattern came to an abrupt end in the 1970s and 1980s.

Price stability does not in itself create growth and economic development. These are facilitated mainly by other factors: the legal and institutional framework, the level of education, the will to invest and the availability of capital, technological development etc. But price stability focuses attention on the rules governing the economy. Those who set prices and negotiate wages do not have to worry about an erosion of future purchasing power but can expect price levels to remain stable. Similarly, there is less uncertainty when it comes to savings and investments, and indeed economic decisions of any kind. Furthermore, the distribution of wealth and income is less arbitrary. This was a key argument for Gösta Rehn, one of LO’s most prominent economists of all time, when he campaigned against inflation under the motto "Hate inflation".

Although gaining a hearing for the policy of low inflation was an uphill struggle to start with, we can see that the idea enjoys growing support today. Studies show that more and more people back the Riksbank’s inflation target. The fact that the Riksbank now has an independent legal status is perhaps a sign that Swedes have come to accept the price stability regime.

Results in recent years

It may be interesting against this background to examine developments in the last few years. Are there any indications that the policy shift around 1990 has already had some effect? Obviously, the return to price stability is not the only change that has taken place. Economic policy has changed in other respects too. For example, budgetary policy has been tightened up, public expenditure has been cut and deregulation and other structural reforms have been implemented.

One way to understand the impact of the policy shift is to compare developments in recent years with the preceding twenty-year period with a high-inflation regime. Naturally, all such comparisons must be made with great caution, since the outcome depends very much on which periods you take as your starting-point.

The argument for low, stable inflation is that it ensures a better environment for investment, improved productivity, economic growth and increases in real wages and, in the long run, higher employment too.

Let us make a comparison between the period starting with the recession of the early 1970s up to 1989, when annual inflation averaged over 9.5 per cent, and the period from 1993, i.e. the end of the last recession, up to now, with an average inflation rate of just over 1.5 per cent.

Broadly speaking, the latter period presents a brighter picture. The total rate of investment, measured in terms of year-on-year changes, is higher on average during this period, especially if investments in housing are excluded.

The average productivity of the economy as a whole is broadly the same during the two periods. However, behind the average during the first period lay some years of very strong growth in productivity in the early 1970s, offset by poor productivity growth during the 1980s.

The Swedish economy has grown by 2.3 per cent on average since the start of 1993. This is only slightly higher than growth during the period 1971-1989. However, the composition of growth is different in the two periods. During the second period, growth in industry has been just under 3.5 per cent, which is about 1 per cent higher than during the previous period. The difference in growth would also be more marked if the latter period were to start only six months later. Average growth from mid-1993, when the recession levelled out, to the end of 1999 was 2.9 per cent.

The employment rate has increased substantially during the last few years. The difference between the two periods is even greater if we confine ourselves to employment in the private sector. In fact, employment in the sector fell substantially during the 1970s and 1980s, at the same time as public sector employment increased rapidly, a development which would not have been sustainable in the long run.

The return to low inflation has also yielded higher real wage increases for employees. Although nominal wage increases have been moderate, low inflation has meant that real wage increases have been considerably higher than previously. Nominal wage increases of over 9 per cent resulted in real wage increases of only 0.7 per cent between 1971 and 1989, while nominal wage increases of less than 4 per cent yielded real wage increases of over 2 per cent between 1993 and 1999.

The general picture is thus encouraging. All in all, the policy has yielded satisfactory results. Naturally, it is too early to draw any definite conclusions from this. Many external factors have also had a favourable effect on developments in recent years. The later period also started with very low capacity utilization. But despite these reservations the situation certainly looks promising.

The current situation

The question today is: will this favourable development continue? This depends not least on the development of wage formation. Before I go into detail on this subject, however, I would like to say a few words about the Riksbank’s appraisal of the situation at the moment.

The Swedish economy is currently making a strong recovery. In the last Inflation Report the Riksbank viewed that growth would be around 4 per cent this year and average almost 3.5 per cent in the next three years. This seemed to be a higher rate of growth than the economy could sustain in the long run. As available resources are utilized, price pressure will grow. The task of monetary policy is to ensure that growth does not deviate too far from a rate that is sustainable, thus causing shortages and price pressure.

The Riksbank was of the view that, in the next few years, the inflation rate would be lower than what might be expected with such strong growth. One reason for this was that the modest increase in underlying domestic inflation anticipated in our forecasts was offset by a stronger krona and lower oil prices. Our assessments also took into account the fact that in recent years domestic price pressure had been lower than expected. We do not know exactly why domestic inflation has been so low in the last few years. Deregulation of the electricity and telecom markets is one important factor. But competition has increased in other product sectors too, due to globalization in general and to membership of the EU in particular.

Our forecasts are also strongly influenced by our estimates of capacity utilization in the economy and of potential growth. There is some uncertainty in both cases. In the last few years we have changed our minds and become more optimistic when it comes to capacity utilization. The potential rate of economic growth may also have increased somewhat. The slightly higher rates of investment and productivity suggest that a somewhat higher rate of growth may be possible. Nowadays, the Riksbank assumes a potential rate of growth of between 2 and 2.5 per cent, which is a slight upwards adjustment compared with the 1980s. However, uncertainty about these factors makes it necessary for the Riksbank to keep a close watch on developments and revise conventional wisdom where necessary.

One of the most important variables when it comes to assessing future inflationary pressures is wage trends. This is perfectly natural, since wage costs represent almost 70 per cent of the total value added in the economy and are therefore crucial to the development of costs and prices. Historically, rising wage costs have covaried with inflation, even though substantial deviations can occur in some years since inflation is also influenced by external factors, such as changes in oil prices.

At this stage of economic recovery, the labour market has functioned better in some respects than the Riksbank, and many other observers too, expected. Nominal wage increases have, as I said before, been moderate and vacancies have been filled without this leading to any substantial shortages. At the same time, it would be premature to breathe a sigh of relief. The labour market and the wage formation system have not been yet tested to the full. The real test will come in the next few years when wage negotiations take place in the context of high capacity utilization and high demand for labour.

What should we expect of the labour market and the wage formation system?

Basically, there is a general, long-term macro framework for how high wage increases can go before triggering rates of inflation that exceed the target. This framework only allows wages to increase in parallel with productivity and inflation. If we assume that productivity growth in Sweden is about 1 to 2 per cent, this would allow wage increases of between 3 and 4 per cent over time, given an inflation target of 2 per cent.

I know that many of those who are engaged in wage negotiations would not agree with this description of the scope for wage increases. Let me therefore mention some important reservations. First, this analysis is only valid in a situation where the economy is in balance as regards employment. In a situation with high unemployment, lower wage increases are desirable, since they stimulate growth in employment. The analysis also assumes that profits are balanced in the sense that sufficient investments are made to ensure continued desirable growth. This was a problem when the EFO model was applied in the late 1960s and early 1970s(1). Around 1970, wage increases became too rapid, and gradually investment in Swedish industry declined.

The most important reservation, however, is that the analysis relates to the general, average scope for wage increases in the country as a whole. This both can and must conceal considerable differences between sectors and regions. The wage formation system must achieve something better than a result for the economy as a whole which is commensurate with average inflation and productivity. If we want low unemployment and price stability, wage outcomes must be adjusted to productivity in different sectors of the economy, indeed to the situation in individual enterprises and to different individuals. The design of pay systems and pay determination is also crucial to the distribution of resources and thus productivity and economic growth. In other words, higher productivity can create more scope for wage increases, but a functioning wage formation system can also help to improve productivity.

Sweden’s success in this area in the next few years will depend both on the measures taken by Parliament and the Government and on management and labour. Parliament and the Government establish the framework for the functioning of the labour market and wage formation. A more expansive financial policy would make the task of management and labour and of monetary policy more difficult in the next few years. The legal framework, for example the rules on industrial disputes, is another factor. Tax and benefit systems are important and affect both wage formation and the behaviour of private individuals, i.e. their willingness to find work, move etc. The Spring Budget Bill noted that in many households it is still not profitable for unemployed persons to seek a job or for those in employment to work longer hours because of the way the tax and benefit systems are designed. These marginal effects have, moreover, increased since 1991.

The system of wage agreements also has a bearing on the possibility of combining price stability and satisfactory employment growth. The fact that several agreements concluded at the national level do not include any individual pay adjustment guarantees, or very low ones, should increase the possibility of determining wages locally and gearing them to productivity development in various sectors. The cross-sector agreements that have been concluded, including the Industry Agreement between about 20 labour market organisations, include rules that reduce the risk of periods without valid pay agreements and the risk of conflicts. This is also a change which should help to improve wage formation.

Let me conclude with a classic quotation from a previous president of the U.S. Federal Reserve, William McChesney Martin. He is supposed to have said that the role of the central bank was to "take away the punch bowl just when the party gets going." That is one way of putting it. The central bank is always on its guard, always ready to pounce! But look at the story from another angle. It is in the guests’ own interests not to let things get out of hand. Behave yourselves well at the party and the punch bowl will still be on the table when you go home! How long and enjoyable the present party will be will depend on many of you here, and on how we in Sweden succeed in establishing a functioning wage formation process and labour market.

Thank you.

1)According to this model, the scope for wage increases in a country with a fixed exchange rate is determined by price trends in other countries and domestic productivity growth in the competitive sector.