Bergström: Is the present low inflation rate a problem?
As one of my colleagues said recently, it appears as though we are now undergoing the gloomiest economic boom of our time. It is in many ways an unusual economic situation; growth is healthy, productivity is very strong, export demand is stable, wage trends look good, households are borrowing away and house prices are developing well. However, employment is weak and inflation is low.
The annual CPI inflation rate was 0.3 per cent in April and the UND1X rate was 0.4 per cent. This is below the interval of 2 per cent +/- one percentage point we have as our inflation target.
Of course, this is a failure, albeit one that is largely beyond our control. Our efforts, analyses and assessments are all aimed at keeping inflation as close to the target level as possible and within the interval we have specified ourselves. The question is what conclusions we should rightfully draw from this miscalculation, particularly with regard to employment.
Monetary policy’s effect one to two years ahead – a challenge for communication
To begin with the low inflation rate makes greater demands on the Riksbank’s ability to explain its policy. When the Riksbank is portrayed it is often the case, consciously or unconsciously, that the present inflation figures are linked to the current monetary policy conducted. If the most recent inflation figures are above the target level, the conclusion is often drawn that the Riksbank should raise the repo rate. If inflation is below the target level, as now, it is often said that the repo rate should be cut. The problem is that today’s monetary policy has very little effect on today’s inflation rate. The way monetary policy functions in practice, the impulse from monetary policy decisions spreads through the economy very slowly. By affecting the cost of loans, incentives to save, exchange rate trends and thereby exports and imports, etc. the interest rate has an effect on total demand in the economy and gradually also on inflation, but this process takes time. It is possible that the reaction time has been reduced somewhat as households have reduced the average period for fixing their loans, from around one and a half years to around one year. However, the accepted estimate is that an interest rate decision makes its greatest impact one to two years after the decision was taken.
The inflation figures registered today are thus largely due to interest rate decisions made one to two years ago, and the decisions we make now will not have their real effect until one or two years from now. The somewhat thankless situation for the Riksbank is therefore that we are always reviewed with hindsight.
I intend to briefly discuss both perspectives in turn and first explain why we made the decisions we made one to two years ago and how they have affected economic developments today – and then I will discuss how we use the monetary policy we conduct now to try to achieve the inflation target over the coming years.
Reasons for the current low inflation rate
The reasons why we misjudged inflation trends earlier are fairly easy to pinpoint now. Although some forecasters were more successful than others in their assessments, almost all made the same two mistakes; we underestimated productivity growth (and thereby overestimated resource utilisation in the economy) and we overestimated import prices. We, like many other analysts, assumed that the rate of productivity growth had increased in a number of sectors, partly due to IT investments combined with restructuring. Similarly, we knew that globalisation was pushing down import prices for many products. What we and many others misjudged was the magnitude of the changes. Productivity growth was even stronger than we had anticipated, with strong growth in the service sector, for instance, where IT use has increased. Import prices were slightly lower than we had expected, partly due to an unexpectedly high increase in the percentage of total imports from low-wage countries such as China and Poland. Consumer prices of imported goods have fallen since mid-2003. It is perhaps all too common that forecasters are reluctant to stick their necks out with regard to large underlying changes.
Moreover, there is another new trend, which was less evident one to two years ago, namely the speed at which competition in the food retail sector would intensify. The restructuring of the retail trade that took place earlier in the United States has now come to Sweden, albeit in a different form. New foreign companies and new business models reduce the labour-intensity and price levels in the retail food sector, where prices for staple products can differ as much as 30 per cent from low-price chains to normal food stores. There were few who could predict the speed of this change, with new food store chains that all together plan to establish hundreds of new low-price stores in one year.
The very nature of these forecasting errors is important to the way we assess the consequences. As employment has been weak during this economic upswing, and total unemployment is at a high level, many debaters have emphasised the link between low inflation and the labour market. They wish to imply that demand in the economy is insufficient because of an excessively tight monetary policy. But this is hardly the case; all of the effects the forecasters have underestimated are supply effects, which reduce inflation because of better and more efficient conditions for production, and not effects that reduce demand. What we are seeing is the reverse; that the expansionary monetary policy has a fairly tangible impact. Low interest rates are leading to unusually rapid growth in lending and to a rise in asset prices. Companies have ample liquidity and thereby good scope for investment. Nevertheless, employment is responding very slowly – instead of employing more staff, employers are using their existing labour more intensively and efficiently.
There are parallels here to what we have seen in the labour market during previous years in the similarly IT-intensive United States, where employment has increased with a lag relative to the cyclical upswing. This is because employers have initially pushed their organisations to the limit and got the most out of their earlier heavy IT investment before resorting to new recruitment. Given the intensive global competition and considerable scope for outsourcing operations, the labour market is constantly undergoing change with new opportunities for rationalising organisations. Many people are moving from an increasingly labour-scarce manufacturing sector to a service sector that is also rationalising. Data show that few of the sectors cutting staff in an economic downswing begin re-employing when economic activity improves again. The redundant employees are instead forced to move on to other occupations. The Swedish labour market appears to find it difficult to meet these structural changes. The major differences in unemployment between regions and between professions indicate that there is less mobility in the labour market than there could be and that few people are finding new job opportunities in new sectors or new locations.
At the same time, the labour market has been affected by changes both within the EU and in the world economy as a whole. Asia is rapidly expanding production and exports and reduced trade barriers are increasing the degree of competition even in markets that were not previously exposed to competition. The emancipation of the countries in eastern and central Europe has dramatically increased access to labour for manufacturing in these countries and also for work in Sweden. So far the temporary labour migration has been only a small, although growing, element in the Swedish labour market. However, the pressure from the labour supply from the new EU countries is bound to increase. The amount of foreign labour entering the various sectors will depend on the initial wage differences. The new member states’ wage levels are in the process of converging with Swedish wages in the long term, but it is dangerous to assume that neither wages nor employment in various sectors in Sweden will be affected. If we assume this, there is a risk that we will not be sufficiently willing to make changes and then unemployment will rise.
The Swedish labour force is thus facing competition both directly from foreign labour coming here and indirectly through households and companies choosing to buy goods and now even services that are produced abroad. Phenomena such as offshoring and merchanting can also be linked to the relatively high labour costs in Sweden. They indicate that Swedish companies can compete in the global marketplace, but their operations are such that they do not require very much labour, at least not in Sweden.
Different assessments have been made of how large a percentage of total unemployment is cyclical and in turn how much of this cyclical unemployment can be affected by monetary policy. The serious assessment attempts that I have seen usually end up with only a small fraction, according to some estimates only one or a few tenths of a percentage points, of total unemployment being affected by monetary policy. Those are calculations that are highly uncertain and should be taken with a large pinch of salt. However my point is that it is a gross exaggeration when, in the present debate, the Riksbank’s decision to lower or not lower the interest rate is presented as having momentous consequences for unemployment in Sweden. This kind of claim is not compatible with any serious assessments or forecasts.
On the whole, the reason for the problems in the labour market is not poor demand – on the contrary, lending is growing at a rapid pace and companies have ample liquidity – but these problems are rather related to structural factors. Even if the Riksbank had conducted an even more expansionary monetary policy one to two years ago, it is not certain that this extra stimulation and the possibly even stronger lending that would have ensued would have affected the labour market in the way some critics believe. The situation today is quite simply a different one from the classical shortage of domestic demand, where lending and liquidity are weak and an interest rate stimulus would have a considerable effect on demand and thereby employment.
Inflation outlook and future monetary policy
This brings us to the other question – namely what monetary policy can do today to ensure inflation reaches the target level one to two years from now.
What we have seen in the United States, which is considered to be slightly ahead of Sweden in the economic cycle, is that productivity growth is gradually subsiding. When the stimulus from monetary policy and fiscal policy has been maintained sufficiently long, companies are unable to respond with more efficient organisation and outsourcing alone, they have to begin recruiting. Productivity growth is dampened, employment rises and in the United States we have recently seen that inflation also rises.
This is also the kind of development we can expect to see in Sweden. The most likely scenario is that over the coming years we will see a more subdued growth in productivity, rising employment and an increase in resource utilisation, despite the pressure for change in the economy. Other factors that have temporarily held down inflation, such as the earlier krona appreciation, the rapidly-intensifying competition in the food retail sector and price pressure on clothing resulting from the EU abolishing textile quotas, are effects that are likely to decrease over time.
The fundamental factors are all in place; expansionary monetary and fiscal policy, strong growth in real wages, good underlying productivity growth and stable export demand, particularly from the United States and rapidly-growing emerging markets such as China and the countries in central and eastern Europe.
Therefore, we should not become exclusively preoccupied by the current inflation rate. Instead we should keep the right perspective. When developments gradually approach the usual recovery pattern, the indications are that resource utilisation will increase and inflation will approach the target level.
There has been concern that an interest rate cut would have a slight effect on consumption and investment, as interest rates are already low and lending is growing rapidly in a situation where housing loans can be obtained at just over 2.5 per cent interest. Companies also have good profits and ample liquidity. Meanwhile there is believed to be a risk that even lower interest rates would further push up house prices through cheaper housing loans. An interest rate cut could thus have a misdirected effect with the risk of future setbacks; little or no effect on achieving the inflation target but a strong impact on the demand for assets such as houses.
This line of reasoning applies when demand grows rapidly, as now, and when capacity utilisation rises. In my opinion, the interest rate effects would be different if the growth in demand slowed down, the labour market weakened and capacity utilisation ceased rising. Then there would be less upward pressure on house prices and I believe that an interest rate cut would be less risky in these terms.
International developments could also be so weak that insufficient export market demand would curb the recovery. The weak developments in our most important trading partner countries, the euro area, are particularly worrying. Like Sweden and the United States, Germany, France and Italy are affected by globalisation, but in these countries the labour market seems to be finding it particularly difficult to cope with change. Germany is at the same time undergoing extensive structural reforms, such as the reduction of compensation levels for unemployment benefits and other transfers, and also longer working hours without full compensation. In the long term, these measures are aimed at increasing the labour supply and competitiveness of the German economy. However, in the short term they lead to uncertainty for many individuals and thereby curb consumer confidence and demand. At the same time, fiscal policy is tied by weak government finances due to a lack of fiscal consolidation during the good times at the end of the 1990s which has made the obligations under the Stability Pact more or less binding, while monetary policy is common to the entire euro area. Germany thus has a domestic demand problem, which it is trying to solve through supply side reforms, while fiscal and monetary policy are tied by the EU and ECB. All this appears to have contributed to further prolonging the economic downswing in Germany and if the recovery in the euro area were to be delayed further, it would have an effect on Swedish export demand.
If these negative risks are realised in full, it could be more difficult for domestic demand to relieve the export-driven growth of recent years. This could break the present recovery impetus, which would lead to a need for further stimulus of the economy from monetary policy.
Conclusion
All in all, we should continue to be vigilant of risks and possible changes in the direction of business tendencies. However, it is important to hold on to a stable analytical framework and a consistent forecast assessment. The underlying factors still indicate a rise in resource utilisation and thereby a rise in inflation.
