Heikensten: Monetary policy and wage formation
Firstly, I would like to thank you for the invitation to come here to the Riksdag and discuss wage formation together with Anders Lindström from the National Mediation Office and Ingemar Hansson from the National Institute of Economic Research.
I shall begin by describing how we take wage formation into account when making decisions on interest rate changes. Then I shall comment on the assessment made in our most recent Inflation Report (published in December), which we did not find reason to alter significantly at our last monetary policy meeting on 6 February. I shall continue by discussing how wage formation may develop in future. In conclusion, I shall summarise and mention some of the challenges facing wage formation.
Just over a decade ago, there was a comprehensive realignment in economic policy. The budgetary process was tightened up. At around the same time, monetary policy changed direction, the exchange rate was allowed to float and the new target was to limit inflation to 2 per cent, with a tolerated deviation of plus/minus 1 percentage point. The economic policy realignment can only be described as a success. On average, the inflation rate amounted to just under 2 per cent a year during the period 1993-2002, and the budget balance was turned around from a deficit of 12 per cent of GDP in 1993 to a surplus of 4 per cent in 2001. Real wages increased on average by 2 per cent a year, which can be compared with the stagnant or even declining real wages in the 1970s and 1980s. At the same time, unemployment was halved from 8 per cent in 1993 to the current figure of 4 per cent.
Thus, the Swedish economy is currently in a much better condition than it was 10 years ago. However, there are many indications that it may become more difficult to combine a good growth rate with a low inflation rate in future. At present there is no unutilised capacity in the economy as there was directly following the crisis at the beginning of the 1990s. It also appears as though the labour supply will grow at a slow rate in the near future as a result both of demographic reasons and of a high rate of absence due to sickness. This impairs the opportunities for growth in the Swedish economy. It could also mean that, when the economy shows an upswing, the Riksbank will be forced to conduct a tighter monetary policy to prevent tension arising between supply and demand on the labour market.
Wage costs have great significance
As total wage costs comprise almost 70 per cent of the total value added in the economy, developments in wage costs have decisive significance for the development of costs and prices. Making an assessment of future developments in wage costs is therefore a natural and central element in the Riksbank's analysis of inflation prospects. If there is an overly rapid increase in wage costs, companies may be forced to raise their prices significantly and this would result in inflation exceeding the target level. This would require tightening monetary policy, i.e. raising the repo rate. Although this would bring inflation back to the target level, it would be at the cost of an increase in unemployment. In the long term, however, developments in unemployment are essentially dependent on how well the labour market and wage formation function.
Monetary policy has affected the wage formation process; the uncertainty over future price developments has declined thanks to inflation targeting. Wage-earners no longer need worry that their pay rises will be eaten up by overly rapid price increases. At the same time, employers find it easier to determine the size of the scope for wage increases. It also appears as though what is usually termed "monetary illusion" may have disappeared. The labour market organisations appear today to focus to a greater extent on real wage cost increases. The strong confidence in the inflation target has meant that these organisations' inflation expectations have been at a level of around 2 per cent since 1996.
The changed conditions for wage formation have probably also contributed to the two most recent bargaining rounds being characterised by a high degree of cooperation between the labour market organisations. At the same time, there appears to be support among wage-earner organisations for the manufacturing industry to lead the way in collective wage agreement through what is known as the Industrial Agreement. Altogether these changes have entailed an increase in nominal wages of an average of 4 per cent a year over the past decade, which can be compared with an average figure of 8 per cent a year during the 1980s.
Developments in wage costs over the coming years
When we make our wage forecasts at the Riksbank, we base them on a total assessment of collective wage agreements, demand pressure in the labour market, productivity, profit margins, inflation expectations and taxes, to mention just a few factors. This autumn will see the beginning of wage bargaining rounds for more than 2 million employees and the outcome will have crucial importance for the development of inflation and thus monetary policy. The assessment in the December Inflation Report was that the Swedish economy would show a recovery towards the end of the year, following a relatively mild downturn in economic activity. We envisioned a still relatively strong labour market. However, employment was expected to fall slightly to begin with, before slowly showing an upturn once again. The strained economic situation among a number of municipalities and county councils meant that we assessed it to be difficult for the Swedish Municipal Workers' Union to achieve any further far-reaching wage increases for its members. It was estimated that, all in all, the demand situation would have a subduing effect on wage demands and we thus did not expect to see any major contagion effects in the 2004 wage bargaining rounds from the early termination of the Municipal Workers' Union's wage agreement. We did not find reason to make any fundamental changes to these assessments at the beginning of February, when we had our most recent monetary policy meeting.
Wages comprise only part of the total costs for labour among companies; on top of these come costs for negotiated and legislated employer contributions, various types of fringe benefits and shorter working hours. As it is the total wage cost that is important to inflationary pressure, it is also necessary to make an estimate of all of these factors. The assessment in the Inflation Report was that the wage costs for the economy as a whole would increase by just under 5 per cent this year and just over 4 per cent next year. It was also assumed in the forecast that productivity growth would recover after a decline in 2001 and approach the more normal, trend level.
However, it was also observed in the report that there was a risk that wage cost increases in coming years could be higher than this figure of 4-5 per cent. At our meeting in February we saw no reason to make any appreciable change to this assessment. The newly-received statistics on wages and prices have not reinforced the upside risk. The National Accounts for the third quarter of 2002 and the revisions for earlier quarters made by Statistics Sweden indicate that productivity growth last year may have been slightly higher than we anticipated. In addition, the outcome in the short-term wage statistics showed lower increases than expected in the central government sector. However, the local bargaining rounds are not yet complete and retrospective revisions will be made, so the statistics should still be interpreted with caution. Other factors here are price rises on electricity and oil, which could contribute to higher wage demands if they influence the view of future inflation.
The long-term scope for wage increases
The Riksbank has the task of ensuring that the inflation target is achieved, broadly speaking regardless of how wage costs develop. Developments in wage costs play an important role, as they influence inflation, but we do not really need to decide whether one particular development in wage costs is more favourable than another. At the same time, it is evidently good to have a wage development that makes it easier for the Riksbank to conduct its monetary policy and safeguard low inflation without overly high real costs. It is not sustainable in the long term to have a wage development that entails the cost of labour increasing more than the value of the goods and services produced. This means that real wage costs can only increase at the same rate as productivity growth. The stronger the growth in productivity, the higher the wage costs can be without triggering inflation, as long as they take into account any reduction in working hours and other secondary wage costs.
The basis for this reasoning is that the economy is in balance, which is characterised by, for instance, stable wage shares and profit shares. Profit shares - profits as a percentage of value added - have fluctuated around 30 per cent for the entire economy. Since 1995, real wage costs have risen more quickly than productivity growth and wage shares. This is not sustainable in the long run. It is necessary, in the long term, for wage shares and profit shares to be at a level where profitability in Sweden is on a par with our competitors to safeguard a level of capital formation and employment that will best promote the development of our welfare.
A couple of different approaches/rules of thumb are used in the wage formation debate to evaluate the scope for wage increases in the long term, or to estimate a "wage cost norm". These are all based on a reliable monetary policy with an average inflation rate of 2 per cent, but the assessments of productivity growth can differ. One problem in this context is that it is not self-evident which sectors should be used for measuring productivity growth. The public sector is particularly difficult to measure and productivity growth here is therefore often assumed to be zero. This means that a rule of thumb based on productivity growth in the economy as a whole is "conservative" in the sense that it gives a slightly lower scope for wage costs than other approaches. Despite the problems, the different approaches tend to give rather similar results in an interval of 3.5-4 per cent.
Of course, there is reason to be cautious when making this type of assessment. Past experiences of wage formation in Sweden are that a wage cost norm can easily be perceived as a floor for negotiation. Another factor advocating caution is that Sweden could be a member of the Eurosystem in the not too distant future. In recent years, wage costs in trade and industry have risen by just over one percentage point more in Sweden than in the euro area. This is not sustainable unless productivity also rises at a more rapid rate in Sweden. A further factor calling for restraint is that the favourable productivity growth prevailing during the greater part of the 1990s was partly explained by growth in the telecom sector. If the telecom industry is excluded from the figures for trade and industry, productivity increased by only 1.9 per cent during the period 1995-2000, compared with a growth of 2.8 per cent if it is included. This sector has now greatly declined in importance and there is apparent confusion over its future development. Given this, the average rate of wage increase should not exceed 3.5 per cent over the coming years.
In conclusion, it may be worth pointing out that if Sweden joins the Eurosystem, there will be other factors affecting the nominal wage cost scope. This is because the ECB's inflation target may be even more ambitious than the Riksbank's. In this case, and if productivity growth remains unchanged, a lower rate of increase will be needed for nominal wage costs.
Summary
I shall conclude with a summary. The realignment of stability policy has contributed to wage formation functioning better than previously in the two most recent wage bargaining rounds. The rate of increase in nominal wages has approached reasonable levels, while wage earners have experienced large increases in real wages. This development is underpinned by a good development in productivity.
When we look ahead, the picture is more worrying. Profits are under greater pressure. It is not certain that productivity growth can remain as high as it was during the latter part of the 1990s. Given these factors, and taking into account international developments, there is a need to bring the average rate of wage increase down to, or even below, 3.5 per cent. This is no easy task in the present situation. The labour market is quite different today than it was 5-10 years ago. Unemployment is relatively low and there is a labour shortage in some parts of the economy, despite the slowdown in economic activity. Moreover, there are tendencies towards demands for compensation, which, if they were to spread, could provide further complications for wage formation.
The trends with regard to labour supply are also worrying. Demographical developments, such as an increasing number of elderly people over the coming decade, will create a demand for more employees to work in geriatric care, for instance. At the same time, the number of persons available on the labour market is declining. An addition problem is the high frequency of absence due to sickness.
It is evident that a central target for economic policy should be to get more people in work. This would facilitate wage formation and create greater scope for economic growth in future. This is perhaps just as important a theme for a hearing in the Committee on Finance as wage formation in itself. It concerns measures that are largely in the hands of the Riksdag.
Thank you!
