Öberg: Sweden – a low inflation economy
I shall begin by thanking you for the invitation. It is always a pleasure to come back to Stockholm University, where I myself was a student.
I have worked with economics for most of my professional life. Prior to taking up my post at the Riksbank I had, for instance, been under-secretary of state at the Ministry of Finance, Director General of the National Institute of Economic Research and adviser at the International Monetary Fund (IMF) in Washington. I intend to use the experience this has given me to provide a long-term perspective to the current economic situation. My theory for the day is that "Things are better now".
I shall cover the following areas in my speech: The first part will deal with the Swedish economy in a slightly longer term perspective going back to the 1960s. There I shall discuss some episodes during the past 30 years that have coloured my perception of the Swedish economy and of what is important for monetary policy. The second part will deal with the current economic situation and current monetary policy.
Swedish economy 1960-2005
I shall begin with a description of the Swedish economy during the period 1960-2005 and start with a picture of inflation in Sweden and Germany during this period (see Figure 1). Inflation has been higher in Sweden than in Germany for long periods of time. During the 1960s we more or less kept pace with Germany. However, during the second half of the 1970s and in the 1980s, inflation in Sweden was several percentage points higher than that in Germany each year. Since the beginning of the 1990s, inflation in Sweden has once again been roughly the same as that in Germany. Sweden has become a low inflation economy.
What, on the whole, has happened since the 1960s? To illustrate this, I shall talk about some of the episodes of the past 30 years that have had great significance for economic developments and for my own view of the Swedish economy. I have been able to follow the economic policy decision-making process connected with these episodes at close quarters from my various posts in public administration. I would like to emphasise that this is my personal view of developments and not the Riksbank’s official view.
The episode I will talk about include the cost crisis in the 1970s and overheating in the 1980s, the change in exchange rate and monetary policy, the wage formation reform and budget consolidation of the 1990s.
The cost crisis of the 1970s
During the 1960s, economic growth in Sweden wad good. GDP growth was at a high level, inflation was moderate, unemployment was low, foreign trade was largely in balance and public finances were strong. I usually look at these five economic variables when trying to gain an impression of a country’s overall macroeconomic development and I will return to them often during my speech.
During the 1970s a number of events occurred that had serious consequences for the Swedish economy. Oil prices rose substantially in 1973, which led to much higher inflation than could be considered justified by the actual oil price. Internationally, this became the start of a long period of low growth.
In Sweden the government invested in economic stimulation to bridge over the coming economic recession. The central wage agreements signed for 1975-76 were therefore at a very high level, which together with wage drift and increased employer contributions meant that the wage costs per hour rose by a total of 38 per cent. Sweden’s competitiveness thus deteriorated substantially. Our labour costs per unit produced in industry in relation to the corresponding costs in other countries rose by around 25 per cent between 1974 and 1977 (see Figure 2).
At that time, Sweden had an exchange rate regime with a fixed exchange rate. After World War Two, almost all western countries’ currencies were pegged to the United States dollar, which in turn had a fixed price against gold. The system came into force in 1944 and was called the Bretton Woods System. When this system broke down in 1973, a currency collaboration was started up between a number of European countries, known as the currency snake, by which the krona was pegged to a currency basket dominated by the German D-mark. In 1977 Sweden left the currency snake and the krona was pegged to a currency basket where the dollar had a double weight in relation to the trade exchange. In 1991 the krona was instead pegged to the European currency unit, the ecu.
During the 1970s, Sweden was still applying credit and foreign exchange regulations. The credit regulation limited the banks’ lending and foreign exchange regulation limited the opportunities to move currency, securities and capital across borders. Both types of regulation entailed limiting the opportunities for efficient allocation of capital and they became increasingly difficult to maintain as time went by. The credit market was deregulated in 1985 and the final parts of the foreign exchange regulation were abolished in 1989.
The idea behind an exchange rate regime with a fixed exchange rate is that inflation in Sweden will adapt to inflation in other countries. Economists usually call this international inflation functioning as a nominal anchor. However, it did not succeed, as Sweden did not manage to stay even with other countries regarding inflation and cost developments. The economic policy conducted and the way the economy functioned during the 1970s were not compatible with a fixed exchange rate. This resulted in a series of devaluations, a smaller one of 3 per cent in 1976, two of them, 6 per cent and 10 per cent, in 1977, one of 10 per cent in 1981 and a large devaluation of 16 per cent in 1982.
Overheating in the 1980s
The 1980s thus began with two devaluations in 1981 and 1982. The background to these was that the end of the 1970s and beginning of the 1980s had been very problematic. GDP growth had been poor, and production actually fell a couple of years. Unemployment had risen to 3-4 per cent, which was then regarded as a very high level. Inflation had soared to levels of 15 per cent. The current account and public finances showed a substantial deficit. Sweden found itself in a situation with weak real growth and twin financial deficits (see Figure 3).
I took part in the preparations for the devaluation in 1982, in connection with the change of government that year. The preparations were led by Kjell-Olof Feldt, Erik Åsbrink and Michael Sohlman. I had the task of making calculations of economic developments over the coming years, based on different economic policy alternatives. Two of the alternatives concerned expansionary fiscal policy to bring down unemployment and a strongly contractionary fiscal and monetary policy to eliminate the twin deficits. In both cases, substantial problems remained in the Swedish economy.
The third alternative concerned a devaluation of 20 per cent in autumn 1982, combined with a price stoppage and pegging the krona to the D-mark. The calculations indicated that industrial production and investment would increase, public finances would improve, the current account deficit would disappear, employment would increase and unemployment would be kept down. After Kjell-Olof Feldt had made contact with the Nordic finance ministers and central bank governors, as well as the German central bank governor, the size of the devaluation was altered to 16 per cent and the planned pegging of the krona to the German D-mark was postponed indefinitely.
I had reason to return to these calculations later on during the 1980s and I could observe that they were pretty accurate on most points, except with regard to price and wage developments. I had assumed that inflation and wage increases would adjust downwards to the international level. For devaluation to be successful, price and wage developments must be kept down after the devaluation. However, this was not the case. Sweden had an inflation rate much higher than that in Germany every year of the 1980s.
Economic policy was not sufficiently strict during the 1980s to keep down price and wage increases after the devaluations in 1981 and 1982. Although public sector finances improved substantially, this was primarily due to automatic improvements resulting from a number of years of good growth. While monetary policy was strict in the sense that both nominal and real interest rates were kept at a high level to protect the krona, at the same time credit regulations were abolished in 1985, which led to a huge credit boom that increased demand and pushed up property prices.
The excessive price and wage increases led to Sweden's cost level once again deteriorating substantially in relation to other countries. At the end of the 1980s, relative costs per produced unit had increased to the level prevailing prior to the devaluations at the beginning of the 1980s and we once again had problems with the twin deficits in the current account and public finances.
The problems were aggravated by an international economic downswing at the beginning of the 1990s and by the fact that household saving increased considerably after the tax reform implemented at the beginning of the 1990s. The property bubble that had arisen in the wake of the deregulation of the credit market burst, which resulted in the whole of the Swedish banking system being on the verge of bankruptcy in 1992. Large currency flows in autumn 1992 made fiscal policy spending cuts necessary and led to a powerful defence of the Swedish krona by the Riksbank with interest rates creeping up in stages. In the end, it was necessary to give up. The krona was allowed to float, leaving the peg against the ecu in November 1992, and it depreciated by 20-25 per cent (see Figure 4).
The conclusions I have drawn from developments during the 1970s and 1980s are that the economic policy conducted together with the functioning of the economy during this period could not lay a foundation for macroeconomic stability and good economic growth. Price and wage formation that got out of hand in the 1970s, a series of devaluations that created considerable pressure in the economy during the 1980s, deregulation of the credit market that triggered a large credit boom, which could not be counteracted by monetary policy as this was detailed to defend the krona’s fixed exchange rate, and a fiscal policy that could not withstand inflationary pressures. All in all, this led to repeated crises and finally to economic collapse. This leads me in to the next three episodes.
The reform of wage formation
Before I get into the changeover in exchange rate and monetary policy at the beginning of the 1990s, I want to mention two episodes that have contributed to making the changeover in exchange rate and monetary policy successful with regard to bringing down inflation to a level comparable with other countries. These are the reform in wage formation and the budget consolidation in the 1990s.
At the end of the 1980s it had become increasingly evident that Swedish wage formation did not function in a manner compatible with macroeconomic balance. It was common for central wage agreements signed during the 1980s to contain price trend guarantees, wage development guarantees, earnings guarantees, checkpoints etc., which led to prices and wages pushing one another upwards in an increasing spiral.
On 1 March 1990, the Government appointed the Rehnberg Commission to follow negotiations of new central wage agreements in the labour market and to act to achieve new agreements that would contain price clauses to replace the agreements for 1990 and new agreements for 1991 and lead to a radical reduction in total wage increases, including wage drift, during 1991. Based on the Commission’s proposals, stabilisation agreements were signed for the period 1 January 1991 – 31 March 1993 for just over 75 per cent of the labour market, which contributed to the rate of wage increasing falling from around 10 per cent in 1990 to 3 – 4 per cent a year during 1992 – 1993. After that, however, the rate of wage increase rose again to around 6 per cent in 1996.
In April 1997 I was given the task by Deputy Labour Minister Ulrika Messing of making an inquiry into wage formation and mediation. When I was given this task, the social partners in the industrial sector had recently (in March 1997) signed a new cooperation agreement, known as the Industrial Agreement, whereby the parties took joint responsibility for wage formation in their part of the labour market and had agreed on a regulatory framework for negotiating wage agreements. The agreement was followed up by similar, but not as far-reaching, agreements in other sectors. At the same time, there were substantial parts of the labour market that had no such type of agreement.
The inquiry led to a new National Mediation Office being established, with the tasks of mediating in labour disputes and acting to promote an efficient wage formation process. The most important change was in my opinion that a third, independent, party was established in the labour market to promote efficient wage formation. The National Mediation Office has the task of keeping up-to-date with the situation in central wage negotiations and acting to ensure that macroeconomic considerations are taken into account in wage negotiations. The Office was also given increased powers of authority to act in connection with central agreement negotiations, for instance, the possibility to call the parties to talks, to provide mediation even if one party was opposed to mediation and to postpone industrial action. The wage statistics improved. The National Institute of Economic Research was given the task of annually producing reports on the economic conditions for wage formation.
Over the past 10 years wage formation has functioned much more efficiently. Nominal wage increases have been around 4 per cent a year on average, compared with 8 per cent a year in the 1980s. The agreements no longer contain clauses regarding inflation and so on and wage drift is much more limited. Real wage increases have been on average 3 per cent a year on average, compared with 0 per cent a year in the 1980s (see Figure 5).
Of course, one can discuss how much regulatory frameworks and institutional conditions surrounding wage negotiations have meant for this development in relation to the general conditions and the increased understanding of the inflation target and of the importance of low, stable inflation. My belief is that they improve the conditions for succeeding in monetary policy’s objective of maintaining price stability.
Following the change of government in October 1994, I returned to the Ministry of Finance after a couple of years as adviser at the International Monetary Fund in Washington to become Under-Secretary of State to Göran Persson. There was basically one task on the agenda; to eliminate the very large budget deficit. The deficit in public sector finances amounted to around 11 per cent of GDP in 1993.
After a majority of the Swedish population had expressed their support for Sweden joining the European Union at a referendum held in autumn 1994, this became an additional task, as well as financing Sweden's membership of the EU, which cost in the order of SEK 20 billion a year.
The timing for dealing with the deficit was ideal from both a political and an economic viewpoint. Studies made by the OECD indicate that powerful budget consolidation processes have only been carried out in connection with serious crises, and the 1990s began with the most serious recession in Sweden since the Depression of the 1930s. GDP fell three years in a row, 1991-1993, and employment fell by more than 500,000 persons during these three years. At the same time, letting the krona float free in 1992 had led to a depreciation of the krona of around 20-25 per cent and a large increase in exports. There was thus a strong growth in demand from abroad that made it easier to constrict domestic demand.
The budget consolidation succeeded beyond expectation. Public finances were improved from a deficit of around 11 per cent of GDP in 1993 to a surplus of around 2 per cent of GDP in 1998. The target had been aimed at achieving a balance in public finances in 1998. The major part of the budget consolidation was achieved by reducing the public expenditure as a percentage of GDP.
There are considerable similarities between the problems at the beginning of the 1980s and the 1990s. In both cases, Sweden’s international competitiveness had deteriorated and we had large deficits in both the current account and public finances. However, the major difference between developments in the 1980s and developments in the 1990s was that fiscal policy was much stricter in the 1990s and inflation was brought down to a level comparable with other countries.
I have since had reason to go through what happened when confidence was restored in the Swedish economy and economic policy. Despite the decision on budget consolidation essentially being taken by the Government and Parliament in autumn 1994, it took some time before various figures believed it would succeed.
The first international recognition of our policy came from the European Commission in spring 1995. I accompanied Göran Persson to Brussels and gave an account of Sweden’s convergence programme to Giovanni Ravasio, who was one of the leading economists at the Commission. Each member state of the EU was to draw up a convergence programme that would show what policy they intended to conduct to achieve stable public finances, low inflation and low interest rates. The Swedish convergence programme was aimed at achieving balance in public finances in 1998, and had the broad support of all of the parties in Parliament with the exception of the Left Party. Ravasio was convinced that the programme was sustainable.
It was possible to see in the financial markets that the interest rate differential between Swedish and German long-term bonds began to decline in spring 1995 when the Swedish convergence programme had been decided and even more so in autumn 1995, when it had been approved by the EU’s finance ministers. For the Riksbank it took until the beginning of 1996 before short-term interest rates began to be cut. The krona was still weak during 1995 and inflation was close to 3 per cent.
I have also had reason to go over how fiscal policy has been managed, since leaving the Ministry of Finance in 1997 to become head of the National Institute of Economic Research and later General Director of Statistics Sweden in 1999. In June 1999, Finance Minister Bo Ringholm appointed me to assess the reformed budget process. The budget process had been reformed in stages since the end of the 1980s. Following the assessment and with broad political agreement, a Budget Act came into force on 1 January 1997. It contained regulations on, for instance, an expenditure ceiling and guidelines for public expenditure.
The main conclusion of the assessment was that the new budget process had functioned well. It had contributed to public expenditure as a percentage of GDP being reduced significantly and to public sector financial saving being turned around from a large deficit to a surplus. Of course, there had been some creative accounting to bring public expenditure under the ceiling, but this was on a relatively small scale.
The changeover in exchange rate and monetary policy
In 1992, when the fixed exchange rate for the krona could no longer be defended, and we changed over to a floating exchange rate, Sweden found itself in a situation where it had not been since the 1930s. The krona was floating and the fixed exchange rate could no longer function as a nominal anchor for monetary policy.
In 1993, the Riksbank specified the objective of monetary policy as being that inflation as measured by the Consumer Price Index (CPI) with effect from 1995 should be 2 per cent a year, with a tolerated deviation of one percentage point above and below this. The exchange rate and monetary policy regime had thereby changed over from a fixed exchange rate regime with monetary policy aimed at maintaining the fixed exchange rate to a floating exchange rate where monetary policy was aimed at keeping inflation low. The inflation target had become the new nominal anchor. However, it was necessary to build up credibility around the new monetary policy and the inflation target.
Part of the work of strengthening confidence in monetary policy was to make the Riksbank independent. At the beginning of the 1990s, the Riksbank had begun to be more independent. In 1990 there was a change in the principle that the Under-Secretary at the Ministry of Finance would be chairman of the Riksbank’s General Council. During the changes in government in 1991 and 1994 the Riksbank Governor was not replaced, which had previously been the accepted rule. Membership of the EU in 1995 carried the requirement of adapting Swedish legislation to the EU’s regulatory framework. During the years 1996-97 there was a commission of inquiry that investigated granting the Riksbank formal independence and in 1998 there was a broad political agreement that this would be granted. In 1999 the Riksbank formally received independence, with an Executive Board that makes monetary policy decisions and a statutory inflation target.
The changeover in exchange rate and monetary policy and the Riksbank’s increased independence have been important conditions for bringing down inflation to a level that corresponds to other countries. This development has been supported by the budget consolidation and the wage formation reform.
Sweden is now once again a low inflation economy, after cost crises and overheating. Over the past 10 years, inflation has varied between just below zero and just over three per cent on an annual rate and inflation expectations have stabilised at around 2 per cent a year. International inflation is also lower now. Both in the EU and the OECD area, inflation is close to 2 per cent a year. The major distinction from before is that the difference in inflation between Sweden and other countries has declined. We once again have an inflation rate comparable with that in Germany.
Things also look better in other regards than they did in the 1970s and 1980s. During the 1960s and up to the mid-1970s, Sweden enjoyed high growth, with the exception of a few individual years. This was followed by a period up to the early 1990s when growth was lower and varied considerably. Some years it was even negative. Over the past ten years we have once again enjoyed good and relatively stable growth. If we compare the period 1970-1993 with the past twelve years, we can observe that growth was much higher during the latter period of low inflation (almost 3 per cent a year) than during the earlier period of high inflation (almost 2 per cent a year) (see Figure 6).
Real wages have also developed better over the past ten years than during the interim period with higher inflation, despite the fact that nominal wage increases have been lower. During the 1960s and up to the mid-1970s, real wages rose every year, i.e. wage were rising more rapidly than prices. Then followed a period of 20 years when real wages did not rise at all on average. Since the mid-1990s, we have once again enjoyed a period of rising real wages. Higher productivity growth has contributed to this.
The current account is showing a very large surplus. Public finances are also showing a surplus, although this does not reach the target of 2 per cent over an economic cycle.
On the other hand, unemployment is now higher than before. During a very long period up to the beginning of the 1990s, unemployment was very low in Sweden. Open unemployment varied between 1 and 4 per cent. In recent years, it has varied between 4 and 6 per cent, following a peak of 9 per cent in 1997 (see Figure 7).
When I studied economics at university, we were taught that lower inflation was linked to higher unemployment. This relationship was illustrated by the Phillips curve. However, later research has shown that in the longer term there is no connection between inflation and unemployment. At the same time, I believe that the problems in the 1970s and 1980s could partly be explained by the Swedish economy being unable to manage the low unemployment we had then without excessive price and wage increases ensuing. For me, one of the most important tasks of economic policy over the coming years is to bring down unemployment. However, the Riksbank is unable to do very much about this. The Riksbank’s contribution is to safeguard the inflation target and in this way create favourable conditions for growth and employment.
All in all, my picture is still that the Swedish economy has functioned much better over the past ten years than in the 1970s and 1980s. Things are better now.
Current monetary policy
The other part of my presentation is concerned with economic developments and current monetary policy and the reasons why the Executive Board of the Riksbank decided at its most recent monetary policy meeting on 22 February 2006 to raise the repo rate by 0.25 percentage points to 2 per cent. The Riksbank’s view of economic developments over the coming years is presented in the Inflation Report, which was published at the same time as the interest rate decision. This part of my presentation does not contain any new information or new assessments in relation to the February Inflation Report and the separate minutes of the Executive Board meeting.
In the light of the picture of economic developments outlined in the February Inflation Report, the Riksbank concluded that there was reason for gradually moving towards a more neutral monetary policy. Although inflation is currently low, developments in the real economy are strong and as capacity utilisation increases, inflation is expected to approach the target of 2 per cent. Raising the repo rate by 0.25 percentage points therefore appeared reasonable. The rate for the continued adjustment of monetary policy will depend as usual on the information received.
To understand the aim of monetary policy, it is necessary to be aware that monetary policy normally has an impact with a relatively long time lag. It is not aimed at the current low inflation rate, but at the expected inflation rate a couple of years ahead. One must also take into account the fact that even after raising the repo rate to 2 per cent, monetary policy is still expansionary. During the entire forecast period, the repo rate is expected to be lower than what can be considered an average key rate in the long term, and lending is expected to continue to increase relatively strongly, even after an interest rate rise.
In my opinion, the information received recently does not significantly change the picture outlined in the Inflation Report. The National Accounts for the fourth quarter of last year were slightly weaker than expected, but at the same time, the outcome for previous quarters was revised up, so that GDP growth for the year 2005 as a whole was as expected. Inflation was slightly lower than expected in January, but in February inflation was slightly higher than expected. The moving three-month averages in the Labour Force Surveys in February showed largely the same picture of labour market developments as that on which our interest rate decision was based.
In conclusion, I would like to return to my theme for this talk, namely that things are better now. Of course there are problems now, and of course one can discuss whether monetary policy should provide a little more or less stimulation. However, with the long-term perspective I have tried to provide here today, I believe that Sweden has a much better economic situation now than during the problem years between the mid-1970s and the mid-1990s. We have a regulatory framework for exchange rate and monetary policy, as for fiscal policy, which promotes macroeconomic stability and good economic growth. It is important that we safeguard these regulations.