Heikensten: The Swedish economy and monetary policy
Let me begin by thanking you for inviting me to speak at the seminar.
As we all know, the Swedish people have voted to remain outside the single European currency. As regards monetary policy this means that we will carry on as before; price stability in Sweden will continue to be the focus of our efforts. We will be continuing to make forecasts of economic activity and inflation as a basis for our monetary policy decisions, and to publish inflation reports and minutes of monetary policy meetings explaining the reasons for these decisions.
To some extent the outcome of the referendum must be interpreted as a vote of confidence for the stabilisation policy regime we have had over the past ten years. Important components of this have been a budgetary policy with an expenditure ceiling and surplus target, and a monetary policy with an inflation target. The changes we have made in stabilisation policy can only be described as a success. The budget balance has been turned around from a deficit of 12 per cent of GDP in 1993 to a substantial surplus at the beginning of 2000. Inflation has been stable and has largely been close to 2 per cent. The inflation expectations of all economic agents today are firmly entrenched around 2 per cent. The outlook is similar for the future. We expect growth to exceed that in many other countries. Meanwhile, inflation is expected to be in line with our target. A surplus is forecast in both the public finances and the current account.
Today I will begin by discussing financial market developments since the referendum. Subsequently I will discuss how economic activity and inflation may develop over the coming years. After giving my view on the salient features of this, I will discuss a number of factors that are particularly difficult to assess. Finally, I intend to touch upon potential problems in the longer term.
The financial markets
In our inflation assessments during the summer, we forecast that the krona exchange rate against the euro would strengthen irrespective of the outcome of the referendum. We also forecast that this movement would mainly be smooth, despite the possibility of some unease in connection with the referendum itself, particularly in the event of an unexpected result. In the run up to the referendum, the krona weakened slightly more than forecast. Since then, it has appreciated substantially against both the dollar and the euro. It is likely that these recent developments have been due to market assessments of the relative strength of the Swedish economy compared with the euro area in the near future, but also to a revaluation of the dollar against the euro following comments from the G7 meeting in Dubai. All in all, developments in the krona have largely been in line with our forecast in the June Inflation Report.
So the appreciation of the krona so far is in line with previous forecasts. Should the krona continue to strengthen - and particularly if there is also reason to believe that the appreciation will last - it will of course have implications for our forecast. However, it is currently too soon to come to any real conclusions for inflation in the near future from the perspective that is significant for monetary policy. What these developments have mainly illustrated so far is perhaps how greatly exchange rates can fluctuate and how unexpectedly quick changes in sentiment can be.
Developments in the fixed income market have been less dramatic. However, the long-term yield spread on Germany widened slightly in connection with the referendum. In all probability, the current yield spread is partly due to recently better growth prospects in Sweden than in the euro area. It is also reasonable to expect that we will have to pay a certain premium in the form of a differential between Swedish and German long yields for remaining outside the single currency. We can illustrate this by comparing the yield spread in, for instance, Sweden, Finland and Italy (see Chart 1). Swedish long yields are higher than Finland's, whose situation in terms of growth and government finances has been roughly in line with our own. Our yields are also higher compared with Italy, despite their markedly weaker public finances.
The salient features of the economic outlook
World economic activity has been broadly in line with our forecast in the June Inflation Report. If anything, current tendencies are somewhat better, as concluded by the Executive Board at its monetary policy meeting in August. Signs of a recovery in the US economy have become increasingly evident, which is pleasing of course. The outlook is different, however, for the euro area. Perhaps there is reason to expect a slightly longer delay in the European revival, even if we have recently seen a number of positive signals of a forward-looking nature. On the other hand, it appears that economic activity in Japan will be stronger than expected. On balance, it currently seems likely that international activity will be roughly in line with the forecasts we made in August.
For Sweden, the majority of indicators point to an upturn in economic activity in the autumn. In accordance with our forecast during the summer, there are currently some unutilised resources (see Chart 2). This has mainly been due to a weakening in demand over the past two years. In addition to the relatively favourable capacity level, productivity growth has also been unexpectedly fast given the current economic climate. In the near future, demand is expected to mount due to both solid consumption growth and a small rise in investment. While GDP growth this year is forecast to be moderate, around 1.5 per cent, growth in both 2004 and 2005 is expected to be appreciably higher, approximately 2.5 per cent. So growth in coming years is anticipated to be somewhat higher than the potential growth we had forecast for Sweden in recent years.
If Swedish economic activity is to rise as forecast, there will need to be a pick-up in investment. This began to fall as early as 2001 in manufacturing, and the following year it declined in the whole economy. Like many other forecasters we viewed the decline in investment as more transitory than it later proved to be. A year ago, our forecasts pointed to a turnaround at the beginning of 2003 (see Chart 3). In the June Inflation Report, it was forecast that investment would be delayed and begin to rise at the beginning of 2004. During the second quarter this year, gross capital formation fell by approximately 2 per cent according to the National Accounts. And in its latest forecast from August, the National Institute for Economic Research expects an increase in investment to be delayed further. The situation is clearly uncertain, even if the fundamental conditions for a rise in investment are good, with low interest rates, stable equity market growth and expectations of a recovery in the international economy.
Another source of concern is connected with Swedish household consumption. This has increased at the same time as we have witnessed a relatively substantial decline in the value of household net wealth. In recent years, household indebtedness has risen sharply and is now approaching the levels seen shortly before the banking crisis at the beginning of the 1990s. The interest ratio, i.e. post-tax interest expenses in relation to disposable income, is however at a relatively low level historically. So far the Riksbank has not judged these developments to be a problem for financial stability, at least not as long as any future rise in interest rates is accompanied by an increase in growth and income. But it is clear that households' scope for consumption in the near future will be limited by their levels of debt and that this could subdue the recovery.
As capacity utilisation in the economy today is relatively low, stronger growth is not expected to constitute a more acute problem with regard to inflation. But when growth exceeds its potential, this normally leads to a gradual intensification of inflationary pressures. Since our last assessment of the situation, some new data has shown, for example, that inflation in August was higher than we previously thought, owing mainly to electricity prices remaining higher than expected. Nonetheless, it is likely that inflation will continue to fall rather sharply over the coming six months. Monetary policy is focused on inflation a couple of years ahead, however. In August, inflation, measured in terms of CPI and UND1X, was forecast from this perspective to rise slowly to around our 2-per-cent target. There is currently no reason to alter this forecast markedly.
Some factors that are difficult to assess
Of course there are factors that could thwart the main scenario I've just presented.
One such factor is a delay in the European economic recovery. Several countries in the euro area are contending with growth problems. Germany's GDP, which comprises one-third of that of the euro area, declined during the second quarter. Investment in Germany has fallen since 2000 while consumption has been weaker than the euro area average. There was also a contraction in second-quarter growth in France, whose GDP comprises roughly one-fifth of that of the euro area. This came after the country had enjoyed relatively robust growth in consumption in recent years, which had compensated for weak investment growth. There are also more fundamental problems in the euro area which could have a negative impact on economic activity. For instance, Germany is still suffering from the after-effects of reunification. Moreover, the euro has appreciated sharply. In light of this, there is no doubt that the world economy is strongly dependent on a US revival.
The outcome of the forthcoming wage bargaining round will also be of great importance, both for economic activity and inflation prospects. It is clear that if real wage growth in the coming years is in line with productivity growth, this will facilitate the Riksbank's efforts. The Riksbank's forecast during the summer rested on the assumption that a weaker labour market would restrain the rate of wage increases to just over 3.5 per cent. But there are examples of wage bargaining rounds in which wages increased excessively despite weak demand pressures, e.g. 1995. The conflict over the Municipal Workers' Union's agreement is also indicative of tensions in the labour market. However, the two most recent wage bargaining rounds have been characterised by greater cooperation between labour market organisations. In addition to current economic conditions, this suggests that the forthcoming wage negotiations should result in economically reasonable wage rises of around 3.5 per cent. It is worth underscoring, however, that no one factor is as important for inflation - and consequently for the repo rate - as wages. The Riksbank will therefore be monitoring the autumn wage bargaining round closely.
One factor that has proved difficult to assess in recent years is energy prices, in particular those for electricity. Energy prices have not fallen back as much as we expected during the summer. In August, UND1X inflation excluding interest expenditure, among other things, rose by 2.2 per cent, of which 0.7 percentage points were due to the increase in energy prices. Current forward prices suggest that electricity prices will be approximately 10 per cent higher during the autumn and winter than was indicated before the summer. Also, the world market price of oil has moved higher than expected.
Thus far, the stance of monetary policy has been based on the assumption that the rise in energy prices will be temporary. We have refrained from raising interest rates despite inflation being considerably higher than 2 per cent at times. An alternative way of explaining our approach is to say that we have focused on underlying, demand-driven inflation. Should energy prices remain elevated for a long period or continue to increase, inflation will of course exceed our forecasts thus far. This could have other implications for monetary policy if it were to have a marked impact on cost levels and begin to affect inflation expectations.
The highly irregular developments in electricity prices are a considerable problem for many decision-makers in the economy, not just for forecasters of inflation. A comparison of electricity price developments in Sweden and the euro area speaks volumes (see Chart 4). These developments raise important questions about how the electricity market functions, questions that concern both competition and the conditions for the production of electricity.
Summary of the outlook
Even when taking into account the factors I have just discussed, the outlook for economic activity and inflation has not changed much since August. The upturn in the US appears to be slightly more evident, with less risks of a setback. In Europe, the outlook is instead somewhat weaker. For Sweden, it seems likely that the economy will experience a rather normal recovery next year. Consequently, it is also probable that inflation a couple of years ahead will be close to 2 per cent, roughly as forecast by the Riksbank in August, even if inflation in the forthcoming year may be lower.
Long-term perspective
Before rounding off, allow me also to touch upon the longer-term perspective.
Over the last decade, the economic policy debate has largely centred on stabilisation policy. The fact that stabilisation policy was in focus during the first half of the 1990s is easy to understand. Following a severe crisis caused largely by mistakes in stabilisation policy, we changed our exchange rate regime and altered the focus of both our monetary policy and fiscal policy. In the EMU debate, it has been natural to continue the discussion of monetary policy and exchange rate systems. I, at least, believe that this discussion went a bit too far. It was easy to get the impression that welfare in Sweden depended solely on the Riksbank's actions.
This is not the case of course. The objective of monetary policy over the past ten years has been to ensure price stability. This has proved rather successful - inflation is in line with the Riksbank's target - and it is important therefore that it lays a solid foundation for growth. But when we alter interest rates, it is mainly with a view to affecting demand around a long-term output trend. While this is important, it must be remembered that we do not affect the trend itself. And it is this trend that determines our welfare!
So we must now shift our focus in the public debate from monetary policy and stabilisation policy to issues that determine Sweden's ability to create growth and welfare. In fact, this has been important for a long time since Sweden suffered a considerable decline in welfare terms relative to almost all other countries, especially during the 1970s and 1980s. But it is particularly significant now in the light of the challenges facing us in the near future, particularly as a result of our ageing population.
Essentially, economic growth is determined by how much we work and how productive we are. Nothing can change this. The number of hours worked is affected by a large number of factors; e.g. the number of people of working age, the number of these that are employed, absenteeism, etc. Labour productivity, measured as GDP per hour worked, is partly dependent on technological advances (e.g. more modern machinery) or on real capital being supplied (a long-term expansion of the means of production). In addition to this are changes in productivity that result from the movement of labour from less productive to more productive sectors.
The number of hours worked grew relatively favourably for a number of years after the crisis at the beginning of the 1990s. But it was essentially demand-driven. There were many people of working age who were unemployed after the crisis years and then found employment. But also from the supply side developments were, for a period at least, positive, partly because there was an increase in the working-age population. In recent years, however, it has become evident that many people have ended up outside the labour market, partly as a result of a rise in sick leave.
As regards productivity growth, the trend in recent years has been more positive. At GDP level it has been approximately 2 per cent per year or slightly above. This differs from the situation in the 1970s and 1980s. It is likely that the changes in monetary and fiscal policy and increased transparency with stiffer competition have been important in this regard.
So how should we feel about developments in the near future? When it comes to productivity, there is reason for optimism, even if the high figures seen in the 1990s partly reflect both rapid growth in the private sector compared to the public sector and robust growth in the ICT sector. Our economy is more open and has greater competition and downward price-pressure in many sectors. The cost pressures in the public sector also have a similar impact.
The outlook is considerably gloomier for the labour supply. As I mentioned, developments are now on a downward trend. One cause of this is that economic policy in some respects has become less focused on raising labour force participation. In terms of both transfers and taxes, changes have been implemented that reduce the incentive to work. Further changes are also being discussed along the same line, including shorter working hours. Furthermore, the age structure of the population is becoming less advantageous from a labour perspective (see Chart 5).
In a country like Sweden, where the costs of an ageing population have such a strong impact on the public sector, the supply of labour is particularly important. It is only if more people work and pay tax, and fewer make use of transfers, that it will be possible to notably relieve the pressure on the public budgets. And this will be necessary if we are to be able to maintain our current commitments in terms of nursing, health care and pensions. So it is high time that we put the labour supply in particular in the focus of the public debate. We should have learned from the crisis of ten years ago the importance of dealing with problems in time. If this does not happen, they could become seven times worse in a matter of years, when the population trend has become more markedly negative. This could then require greater, more painful changes under less orderly circumstances.
Thank you!
