Bäckström: The Swedish economy and monetary policy

The repo rate increases

In order to understand why the Riksbank has raised the repo rate in two steps from 2.90 per cent last November to 3.75 per cent less than a fortnight ago, it is necessary to consider the situation in 1998. The mood at that time, in the global economy as well as in Sweden, was coloured by the Asian crisis, the suspension of Russian payments, the problems with the hedge fund Long Term Capital Management and falling stock-market prices.

It was considered that in addition to direct problems, the financial crisis in the autumn of 1998 would have negative effects on real economic activity. Forecasters altered their assessments and the prospects for global output and inflation were revised downwards. Central banks around the world did likewise and the Riksbank was no exception. Instrumental rates were reduced in many countries. Between the early summer of 1998 and the spring of 1999 the Riksbank cut the repo rate by a total of 150 basis points, bringing it down from 4.40 to 2.90 per cent.

The actions of central banks probably mitigated effects of the various shocks. They helped to check the signs of panicky behaviour in the market and then to restore a more optimistic assessment of the future. At the same time, the process displayed some ability to heal itself. All this has contributed to a rapid recovery in the global economy as well as in Sweden.

The current picture of many economies is thus very different from the uneasy situation that prevailed during 1998 and the early part of 1999. Global growth prospects are appreciably better and the risks connected with inflation have changed. That is also why instrumental rates were adjusted upwards by many central banks last autumn and early this year.

In principle it can be said that inflationary pressure follows a cyclical pattern. When economic activity is strong, demand growth tends to exceed the expansion of capacity and this normally leads in time to an accelerating rate of price and wage increases. Conversely, weak demand may lead to the prospect of inflation being below the target one to two years ahead. As a rule, the repo rate should be raised in the former case and lowered in the latter. In this way, the repo rate varies with the business cycle so that the rate of inflation remains more or less continuously in line with the target.

So, as rising economic activity successively reduces the degree of unutilised resources in the economy, the repo rate has to be increased. That is the process the Riksbank has initiated with the increases since last autumn.

From what I have said it follows that there should be a repo rate that is commensurate with stable inflation and a demand growth that matches the expansion of capacity while the economy is also balanced in other respects. It is around this level of interest rates that the actual repo rate ought to be varied as activity moves up and down. So what is the Swedish economy’s sustainable growth rate and what is the commensurate repo rate?

The Swedish economy’s sustainable growth rate

A simple way of estimating the potential growth rate is to add the figure for the trend rate of increase in labour supply to the growth rate for labour productivity.

The latest figures, based among other things on demographic changes and labour force participation, point to an annual increase in labour supply up to 2008 of 0.5 per cent. As regards labour productivity, most of the evidence suggests that the trend in the 1990s was higher than in the 70s and 80s. In the period 1974–93, for example, the annual increase in labour productivity averaged approximately 1.4 per cent, while the average since 1994 is about 1.9 per cent. It is admittedly true that as the economy has been in an upward phase since the summer of 1993, it is hardly surprising that productivity growth is above the trend; but a comparison with the corresponding upward phase in the 80s leads to the same conclusion: the rate of increase in labour productivity appears to be approximately half of a percentage point higher than before. This accordingly means that the Swedish economy’s potential annual growth rate is between 2 and 2.5 per cent. It may be asked, however, whether one can simply project productivity’s historical path.

This is where the discussion of what is known as the New Economy comes in. Productivity growth in the United States has risen appreciably in recent years. The investment boom in the 1990s has been one factor here; corporate investment has almost doubled since 1990 and a vary large share has gone into new information technology. This has led to what the economic jargon refers to as capital deepening, which amounts to a higher capital input per employee. A notable feature of the picture is thus that the new information technology seems to be having knock-on effects in the form of enhanced productivity in other segments of the U.S. economy. Many observers therefore consider that the American economy is now capable of expanding at a rate of 3 to 3.5 per cent without encountering shortages and an acceleration of inflation.

A similar change could well occur in other countries that reap the new information technology’s potential benefits. But it cannot yet be said that productivity growth in the Swedish economy has improved to the same extent. Neither have we registered a similar growth of investment, which could then be a leading indicator that the new technology is being introduced on a wide front. In the 1990s the stock of corporate capital, for example, grew twice as fast in the United States as it did in Sweden.

It is certainly conceivable that the high level of the Stockholm stock exchange represents expectations that productivity growth will also move up in Sweden. The sum of the values of individual companies can serve as a yardstick of what is expected of the stock market as a whole and this presumably reflects expectations of how the total economy will develop. And if current share prices are regarded as reasonable, there will clearly have to be an appreciable increase in future profits. This can then be interpreted as expectations of an overall improvement in productivity growth and thus of a higher potential output. An alternative interpretation could be that the stock market is at least partially over-valued. A third possibility is, of course, that the development of share prices is explained by a combination of these two factors.

It still seems too hazardous, however, to assume that potential growth in Sweden has moved up towards the level in the United States. It is possible—perhaps even probable—that similar supply factors will emerge in the future and the Riksbank must be alert to this. But it seems safest at present to count on a potential annual GDP growth rate of between 2 and 2.5 per cent.

Growth can be temporarily higher

When the labour force is not being fully utilised, the increase in employment can exceed the trend rate of about 0.5 per cent. In a transitional period the economy can then grow faster than its potential rate in the longer run.

Unemployment in Sweden is still historically high. That might suggest that the labour market has plenty of unutilised resources at its disposal. Some bottleneck problems are already showing up, however, and are an indication that there is only a limited number of persons with the appropriate training and occupational experience for the jobs that are available.

Moreover, statistics from the National Labour Market Board show that the number of unfilled vacancies at employment offices has risen to the highest level since 1990 and business tendency data from the National Institute of Economic Research reveal that certain sectors are now finding it difficult to recruit labour with the required training and occupational experience. The problems are accentuated by the fact that a large part of the increased production and hence the expansion of employment is concentrated to the metropolitan regions and other university cities. This means that bottleneck problems are starting to impede the growth of production and employment. The infrastructure in a wide sense in the metropolitan areas, perhaps in the Stockholm area in particular, may affect the possibilities of continuing to benefit from the existing demand. The rapid prise rise for houses and tenant-owned flats in the Stockholm area is a clear illustration of certain supply-side problems. Another sign is the shortage of rented dwellings, or rather the lack of mobility in this market.

Thus it is not only labour demand that will determine the future development of unemployment. A growing part will be played by supply-side factors in the form of occupational and geographical mobility.

So while it can be concluded that in the near future the Swedish economy’s growth rate may be capable of topping the potential level of—as I just indicated—between 2 and 2.5 per cent, this applies only for as long as there is a supply of labour with the appropriate training and occupational experience and the possibility of being geographically available.

According to the most recent national accounts, published just before Christmas, GDP growth in the first three quarters of 1999 was over 3.5 per cent. The fourth quarter outcome is still difficult to estimate precisely but demand growth seems, if anything, to have accelerated. An unusually good Christmas season appears to have resulted in a strong trend for private consumption, for example. A continuation of the strong development is suggested, moreover, by such leading indicators as household confidence, a high increase in the money supply and rapidly rising asset prices.

This implies that sooner or later, if nothing were done, the rapid growth at present would lead to overheating and an acceleration of inflation. One factor behind the current trend has been monetary policy’s stimulatory direction ever since the repo rate cuts that were initiated in 1998. This illustrates that over a complete business cycle, the repo rate of 2.90 per cent that applied up to last November is too low. GDP growth at the current rate of 3.5 to 4 per cent implies an expansion of demand that clearly exceeds the rate of 2 to 2.5 per cent that the economy can cope with in the longer run. It is only feasible in a limited, transitional period, that is, for as long as there is a supply of labour for the available jobs.

The repo rate hike may not suffice; inflation prospects set the pace

As monetary policy works rather like the helm of an oil tanker—it takes rather a long time to shift the economy onto a new course—it was necessary to make the adjustment to a more neutral monetary stance. Thus, GDP growth will continue at a higher rate for some time to come and then, after perhaps one to two years, successively slacken towards the economy’s long-term potential.

The task the Riksbank is endeavouring to perform is to achieve an economic transition from a fairly normal upward phase to a growth trend that is more sustainable. This task, which is not simple, follows implicitly from our statutory function of ‘maintaining price stability’. As I mentioned earlier, a low and stable rate of inflation normally means that the economy is continuing to expand in accordance with its long-term potential.

One cannot be certain, however, that the repo rate hike since November will suffice to bring about such an adjustment of the path for economic growth. The impact on demand from monetary policy is probably still stimulatory. But the repo rate that represents a more neutral monetary policy impact cannot be specified exactly. We simply do not know enough about this. Besides which, the level probably varies over time.

Empirical guidance can be derived from countries that are ahead of Sweden in the business cycle. The action of the U.S. Federal Reserve, for example, illustrates the economic advantage of a forward-looking approach that raises the interest rate for a gradual reduction of monetary stimulation. The Fed began a series of increases in the instrumental rate in February 1994, from a level of 3 per cent, and then adjusted the rate within a more normal range of 4.5 to 6 per cent. This served to prevent an acceleration of inflation without plunging the economy into a recession. On the contrary, the U.S. economy is now experiencing the post-war era’s longest upward phase. While other factors besides monetary policy have, of course, contributed to this, the American example does show that a judicious upward adjustment of the interest rate definitely can contribute to good economic effects provided the sustainability of growth is also taken into account.

I do not cite the case of the United States to imply that the Riksbank has already decided to make further rapid interest rate increases in the near future. That is not how we work. Those who follow what we do know that the foundation for monetary policy measures is our inflation forecast.

At the time of the latest repo rate increase the Riksbank’s Executive Board judged that inflation is rising only moderately and will be approximately in line with the target in one to two years’ time. It must be borne in mind, of course, that new information may alter our assessment and that our time horizon in the upward phase is moving ahead. In the short perspective, however, things look a little different. The rate of inflation may actually tend to slacken this spring, partly due to effects of the deregulation of telecommunications. But as I said, the perspective for monetary policy lies further ahead.

Forecasting is difficult

The difficulty for the Riksbank lies in constructing forecasts for the period—usually one to years ahead—that is generally regarded as the time lag before an interest rate adjustment has its full effect. Anyone with some experience of forecasting knows how difficult this can be and is aware that forecasts not infrequently miss the mark. We could in fact attach more importance to the current as opposed to the forecast future rate of inflation but that would probably call for rather large repo rate adjustments in order to keep inflation stable around the target. Such an approach would result in sharp fluctuations in real economic activity, which would naturally not be desirable.

Are there any developments which suggest that the time lag for effects of interest rate adjustments may be becoming shorter? More and more home owners are switching from fixed to variable mortgage rates, which may indeed tend to speed up the impact of monetary measures. But there are other factors which may be working in the opposite direction, towards a longer time lag. One example is the costs involved in altering prices. One of the benefits of low and stable inflation is precisely that it facilitates contracts with a longer duration. Insofar as the establishment of a low-inflation regime has led to price movements being more restrained for this reason, it follows that it will take longer for monetary policy measures to impinge on prices. For monetary policy’s time horizon, there are therefore arguments in both directions.

The problem of faulty forecasts is limited, however, in that monetary policy is implemented continuously. Moreover, current information about economic tendencies can be accorded more weight than events that lie further ahead and the forecasts can be revised as time passes. And when the repo rate is comparatively far from a more neutral monetary stance, monetary policy can be adjusted in larger steps than when the rate is closer.

Thus, there are various ways of making up for the fact that economic forecasts beyond the short term not infrequently miss the mark and still allow for the circumstance that it takes a comparatively long time for the full effect of monetary measures to materialise.

Conclusion

I should like to conclude by saying that while we all know that the Riksbank’s main concern is the future path of inflation in the Swedish economy, our deliberations about how to hold inflation at a stable level around the 2 per cent target can just as well be seen as the Riksbank’s endeavour, by means of the repo rate, to contribute to a continuation of the Swedish economy’s favourable trend. In many situations our statutory duty—the maintenance of price stability—is fully commensurate with the requirement that the Riksbank helps to make the development of production and employment in the Swedish economy as stable as possible.

1) See LU 1999/2000 Bilaga 1 (Medium Term Survey 1999/2000, Annex 1, Population, labour force and employment), Ministry of Finance.