Favourable economic development in Sweden
The economic situation at present in Sweden is better than it has been for a very long time. Last week, Statistics Sweden reported that GDP growth in 1999 averaged 3.8 per cent. This was accompanied by the creation of almost 90,000 new jobs, the largest increase for thirty years.
This expansion is occurring in an environment where, at least so far, the level of price and wage increases has remained comparatively low even though the economic upswing has been running for more than six years at an annual GDP growth rate that has averaged 3 per cent.
The unutilised resources that were available after the crisis in the early 1990s have naturally played an important role in this development but it is also significant that inflation expectations have been anchored around the target. The latter indicates that, at least to date, we have had a balanced development. The budget and current-account surpluses are further confirmation of this. Who would have thought a decade ago that the Swedish economy could achieve such a favourable performance?
The national accounts, which are now available from 1980 onwards with the new base year, demonstrate an interesting fact. In the upward phase that Sweden has been experiencing since the summer of 1993, a prominent part has been played by investment. But it has not amounted to an investment boom such as that in the United States.
Rapidly rising investment in new production facilities leads in time to higher potential output. This is clearly one reason why productivity growth in the 1990s has been somewhat stronger than in the ’80s. By contributing to a higher rate of potential growth, it points to some increase in the level of sustainable output. By historical standards, however, the improved rate of productivity growth is still fairly moderate.
Figure: Labour productivity between 1870 and 1999
Nevertheless, growth driven by investment can lead to a good circle. The supply side of the economy generates increased demand and a favourable composition of demand tends to strengthen the supply side even more.
The higher productivity growth that results, as I mentioned, from increased investment leads in turn to more optimistic expectations of future corporate profits, which often push share prices up. Higher corporate values stimulate these firms to invest even more and also make it possible for new enterprises to gain access to venture capital to finance their growth. At the same time, households with shareholdings find that their net assets have gained in value, at least on paper. Larger stocks of wealth and optimistic expectations lead to a stronger increase in consumption.
Figure: Investment relative to GDP
Since 1950, Sweden has experienced three distinct periods of different duration when investment rose rapidly.
The first period
lasted from the early 1950s up to 1967. The capital stock that was built up in those years provided the base for a comparatively major potential for business rationalisation. And as we all know, productivity growth was historically high and the investment boom continued for a long time. One factor—but by no means the only one—behind the strong growth of investment was massive residential construction.
The GDP share for investment peaked when a recession started around 1967. Corporate profit ratios fell, the first and second oil crises occurred in the 1970s, structural problems made themselves felt in the Swedish economy and from 1967 to 1982 the investment ratio declined sharply. This decline was partly a result of falling investment in residential construction but even when this item is excluded, the GDP share for investment shows clear signs of a standstill; if anything, it actually tended to fall. This was accompanied by a deterioration in potential output. Sweden lagged behind other countries in the growth league. Real wages were sluggish.
The second period
of strong investment growth began after the devaluation in 1982 and continued until the end of the 1980s. But it does not seem to have raised potential growth. The Swedish economy became overheated at fairly moderate growth rates in the late 1980s and productivity weakened markedly.
As a matter of fact, productivity growth in the 1980s was weaker than in any decade since the industrial era reached Sweden around 1870. This suggests that much of the investment in this period was not of the productive kind that helps to improve potential output. A look at the statistics shows that, at least in the latter part of the decade, a large proportion of the investment went into commercial property in connection with a speculative phase.
The third period
of investment began when the Swedish economy turned upwards in the summer of 1993. The krona had depreciated, corporate profits rose, global economic activity picked up and Swedish exports and manufacturing took off. Gross capital formation played a major part in the growth process. Moreover, the tendency is even clearer when housing investment is excluded.
Investment-driven growth can take different courses
Judging the sustainability of an ongoing period of growth is not an easy task. Some alternative scenarios can be constructed in terms of the interaction over time between supply and demand.
- One possibility, with an unduly optimistic investment propensity, is that the supply side expands too rapidly and capital stocks become excessively large. This can be likened to speculation in real capital formation. Such a course leads sooner or later to setbacks that are not necessarily preceded by an accelerating rate of price and wage increases. The real economy may then be forced into a troublesome process of downward adjustment. It is conceivable that elements of this type have been involved in the Japanese economic downturn in the 1990s. Excessive investment in real estate probably contributed to the problems in Sweden in the early 1990s.
- Another possibility is that rapidly rising asset prices prompt a strong expansion of credit. In this case, rising wealth generates demand for financial investment rather than for goods and services. To begin with, the yield from these financial investments is much higher than the borrowing costs. Financial bubbles of this type render the balance sheets of households and firms vulnerable and also jeopardise the financial system as a whole. Sooner or later the undue optimism evaporates and gives way to a difficult process of financial adjustment, perhaps with problems in the banking sector. This, too, has been a characteristic feature of the Japanese economy. Similar courses have also been observed in Sweden and other countries.
- A third possibility is that demand growth is too strong even for a faster expansion of capital stocks and supply. Expectations of future income become too optimistic. This results in a classic acceleration of price and wage increases.
- A fourth possibility—clearly the best alternative—is that supply interacts harmoniously with demand instead of getting out of step. This situation generates balanced growth and output, at least for a time, tends to be higher than before. The period of investment from 1950 to the mid 1960s is an example of such a long period.
The phenomenon that is often referred to as the new economy is driven by vigorous investment, whereby new technology spreads by degrees throughout the economy and renders production more efficient. As we have seen, however, there are alternative courses. Neither is everything for the best in that world, even though it is easy to get that impression from the general discussion. But it definitely does provide some interesting challenges and opportunities for consumers, entrepreneurs, financial players and, not least, those who decide economic policy. At the same time, it is important that agents in some part of the economy do not become unduly optimistic and make ill-considered decisions. That could transform the good circle into euphoria and history has taught us that the story then ends in a more or less heavy setback.
For a central bank it is thus a matter of ensuring that the rate of inflation is firmly anchored around the target but it is also important to promote the efficiency and security of the financial system in general. In this perspective, low, stable inflation is an indication that supply and demand are moving in harmony over time. But as I just said, we must also keep an eye on the financial system as a whole to ensure that no imbalances in other respects are built up that may threaten a favourable economic trend. What happened in the late 1980s and early ’90s is still casting a shadow over us.
Balanced economic development in Sweden
As far as Sweden is concerned, developments to date seem to be characterised by stability. There have been several occasions when the figures for inflation have, in fact, proved to be lower than most observers, including the Riksbank, had expected. At present, moreover, the expansion of credit appears to be reasonable. But asset prices have shot up to an extent that implies considerable expectations that economic growth in Sweden will be appreciably higher than before. So in this respect there are grounds for sending a warning signal.
Riksbank representatives have stated that our attitude to the new economy is tentative but not dismissive. We have not yet seen any appreciable evidence that productivity growth in the Swedish economy has moved up as it has in the United States. As I mentioned earlier, productivity growth does seem to have risen in the 1990s from the ’80s but it has not yet accelerated markedly, even according to the latest national accounts. We are, however, keeping a very close eye on this and searching for signs. After all, Sweden is often reckoned to be one of the leading countries when it comes to information technology in a wide sense and not just as regards Internet use among people in general, where we are actually top of the class.
The upward shift in productivity growth in the second half of last year is, however, an interesting observation. We also note that even with the strong growth of demand, domestic underlying inflation is low.
Swedish economy approaching full resource utilisation
An indication that demand in the Swedish economy is nevertheless rising faster than the supply side can cope with in the longer run is that unemployment is falling. This is fundamentally welcome, particularly after the period in the 1990s when unemployment in Sweden has been historically very high. But somewhere there is a point at which firms have difficulty in recruiting labour with the appropriate education and occupational experience. Such tendencies are already discernible in certain industries and localities. Moreover, unemployment among the highly educated has decreased rapidly to around 3 per cent, while among those with a lower education it has remained above 6 per cent.
That a labour market where unemployment is falling becomes successively more strained is only to be expected. The problem for a central bank is to know when this is liable to generate an accelerating increase in wage costs that threatens the inflation target. In that wages make up by far the largest share of corporate costs, rapid wage increases are a major driving force in an inflation process.
In this respect a central bank has to consider two problems. One is that wage increases accelerate as a rule some time after the onset of more widespread labour shortages. The other is the time lag before an interest rate increase has its full effect. The Riksbank has talked of a lag of twelve to twenty-four months, though of course we cannot be sure of the exact figure. But there is no evidence at present that the lag has become appreciably shorter; there are arguments both ways. Against this background, a central bank needs to avoid two conceivable alternatives.
One is that the central bank acts too early in an upward phase, so that unemployment remains unnecessarily high and the economy in general is prevented from utilising its full potential.
The other is that the central bank acts too late to prevent a process of accelerating price and wage increases from getting under way. Monetary policy then has to be restrictive and check demand harshly in order to bring inflation back to the targeted rate.
Both alternatives are unnecessarily costly for the real economy and therefore undesirable. The way to avoid landing in situations such as these is to forecast inflation so that the central bank is in a position to adjust the repo rate with its sights on roughly the same time horizon as for monetary policy’s impact on inflation. This gives a course that is smoother both for the path of the repo rate and for the real economy, that is, for production and employment.
Forecasts can err but that is manageable
In other words, monetary policy must be forward-looking. Otherwise we would generate instability instead of stability in the real economy.
But this does require a great deal of forecasters as well as of those who construct macro models. The forecasts concern inflation one to two years ahead, which calls, for instance, for reasonable precision in the relationship between, for example, resource utilisation and inflation. Knowledge is also required about the strength and timing of the impact on inflation that a repo rate adjustment would have.
It is evident that in the 1990s forecasters and macro economists, including central banks, faced major challenges. Inflation forecasts, for example, were frequently on the high side. There are many explanations for this. A rapid economic transformation is in progress in Sweden and many other countries. This may alter relationships between demand and inflation. For one thing, low inflation may lead to pricing behaviour that is different from when inflation is high. For another, the economy may be exposed to increasingly strong competition, due to either internal deregulation or the general process of globalisation or a combination of the two. Thirdly, a shift in productivity growth may alter potential output.
The Riksbank, like other forecasters, is making intensive and continuous efforts, however, to improve its predictions. The more experience we have of the new monetary policy regime and its macro-economic consequences, the better should our forecasts and macro-economic assessments become.
Moreover, in the changeable world in which we live, new, relevant information about economic developments may have to be accorded greater weight than more formal results from models, which of necessity are based on historical relationships.
So it ought to be possible to deal more successfully with the fact that forecasts of economic developments some time ahead can never be completely accurate and still allow for the comparatively long time lag before monetary measures have their full effect.
Interest rate level an important adjunct to inflation assessment
As I mentioned to begin with, the Swedish economy seems to be robust and the Riksbank has the task of promoting a continuation of this trend. This involves, as I just said, keeping inflation close to the targeted rate.
The Riksbank can influence inflation by aiming to adjust total demand to the long-term path for economic growth and/or by working on expectations of future inflation.
Our simple rule of thumb for monetary policy involves raising the repo rate when forecast inflation is above the target and vice versa. For this to affect demand in practice, however, it is important that the adjustment occurs in the real repo rate, so that this then acts to a varying extent on real interest rates for loans with longer maturities.
Besides determining the adjustments that are called for, however, monetary policy has to consider the level that is pertinent. This is where arguments about the equilibrium real interest rate come into the monetary policy discussion. This is not a separate issue from the discussion of inflation forecasts but an important adjunct in the assessments.
When the economy is in equilibrium, in the sense that expected inflation is on the target and output is on the long-term trend, then the real interest rate is also at its equilibrium level. This implies that there is a level of the real repo rate that corresponds to that level and can be described as neutral. Consequently, when the real repo rate is above this level, monetary policy can be described as restrictive and when it is below, policy can be said to be relaxed.
However, this neutral level is elusive because the equilibrium real interest rate is not directly observable and has to be estimated. The real interest rate is dependent both on the rate whereby domestic saving and investment are in equilibrium and on the international real interest rate, which depends in turn on the global relationship between saving and investment.
Experience shows that in practice the real interest rate also varies over fairly long periods, which should indicate that even the equilibrium real interest rate shifts over time. Historical series on the real interest rate show that it was negative in the 1970s, whereas in the ’80s and ’90s it was around 4 per cent. In the period 1960–99 as a whole, the short-term real interest rate was around 2 per cent. From this it can be concluded that the historical picture poses difficulties in assessing the level that should serve as a starting point for monetary policy.
As a rough approximation, however, the Executive Board of the Riksbank considers that monetary policy is relaxed at present, which implies that the real repo rate is somewhat below the equilibrium level. As economic activity becomes increasingly strong and we move further into the future, resource utilisation can be expected to become more and more strained. Monetary policy will therefore also need to adopt a neutral stance; the question is what that level of interest should be and how quickly we should approach it. It is clear from various analyses and market letters that these issues are occupying many financial market players at present.
From what I have said, however, it should be evident that there is nothing mechanical about this, no predetermined path or fixed steps for future interest rates. As always, the interest rate movements are guided by the process of inflation. A good deal has happened recently. The economy appears to be somewhat stronger than we envisaged in December. At the same time, productivity seems to be developing more favourable. Perhaps there are grounds for somewhat greater optimism about wage formation. After the latest repo rate increase in February, against this background I consider that the inflation risks still seem to be fairly balanced.
