Heikensten: Eight Questions on Monetary Policy
Introduction
During the 1990s monetary policy has been the subject of intensive discussion. This is not surprising, given the serious crisis in the Swedish economy at the beginning of the decade and the role that monetary policy played both then and as the crisis was building up. Since that time, monetary policy has had an explicit role to safeguard independently the value of money—a role that is new in Sweden in the post-war period. With a floating exchange rate, the decisions currently made by the Riksbank on its instrumental rate are very real and important to the behaviour of households and businesses and to economic growth.
The monetary policy discussion has evolved over time. Questions and statements about monetary policy’s basic objective of price stability that were very frequent a couple of years ago are less so today. Understanding of the Riksbank’s actions seems to have increased. We are presently trying to evaluate this by asking the public how it views the Riksbank’s work and by following the media and newsletters written by the market’s players. It is also possible to form an opinion based on the market’s reactions to changes in the instrumental rates as to whether the policy is understandable.
In spite of the fact that knowledge and understanding of the Riksbank’s policy have gradually increased, I believe that from time to time there is a need to explain the basis for the policy and some of the principles that guide it today. The purpose of today’s speech is just that. I will endeavour to answer seven of the questions that I have encountered most often in recent years at meetings and in editorials and commentaries and that seem to permeate the discussion of monetary policy continually. By way of conclusion I will comment on an eight question.
Question 1: Why is low inflation good?
The experiences of previous decades in Sweden and a number of international examples show that high inflation is associated with economic costs. Moreover significant costs ensue when one is compelled to gear down from high to low inflation—something that sooner or later always happens.
It has been shown that high inflation is often associated with sharp fluctuations in the relative prices of goods. Rapid and abrupt price changes create uncertainty about the future for all the actors in the economy. It simply becomes more difficult to make long-range decisions. Households and businesses are compelled in practice to take greater risks and both then demand higher interest and profits. This reasoning is supported by research, which indicates that the growth rate of the economy may be lower, because the increased uncertainty reduces investment activity. In addition the risk of making erroneous investments increases. (Diagram 1. CPI and Growth)
That Sweden in the 1970s and 1980s lost so much ground in terms of prosperity in relation to many other European countries is certainly due in part to the stabilisation policy then being conducted. Prices and wages were allowed to drive each other up, while budget policy was far too weak. Devaluation followed devaluation, with low investment and weak productivity growth as the result.
High inflation also leads to arbitrary changes in the distribution of income and wealth in society. Wages and profits normally rise at the same rate as inflation. The value of real estate and some financial investments rises even faster. At the same time nominal fixed interest rates and transfers lag behind. The high profits made by the owners of houses and co-operatives in the 1970s and 1980s occurred in practice at the expense of tenants and those with bank savings. Between 1971 and 1974, for example, house prices rose by 50 per cent and agricultural property by 80 per cent. In contrast those who had their savings in bonds or in the bank saw their assets shrink by about a third in real terms.
The problems are amplified by a regime of taxes and subsidies that are wholly or partly indexed. Groups that have full indexing are favoured at the expense of others who do not receive subsidies or do not get full compensation. Also the economic costs may rise when the behaviour of businesses and households is negatively affected by taxes and subsidies in combination with a high rate of inflation. Many good examples of this can be found in the Sweden of the 1970s and 1980s. The american economist Martin Feldstein has estimated that if inflation increases from 0 to 2 per cent, then the economic costs in the United States will amount to the equivalent of 1 per cent of GDP, principally as a consequence of the disruption caused by the interaction between the tax regime and inflation.
Let me conclude the answer to this first question by emphasising that my reasoning is anything but new, even if some of the examples are, and there is new empirical research to rely on. The Swedish Trade Union economist Gösta Rehn argued in the 1950s and 1960s that full-employment policies must be combined with a tight demand policy. Wage formation could function well only if demand was restrained and the rules of the game were obvious. He also waged a battle for many decades against inflation for reasons of distribution under the slogan, ”Hate inflation”. Substantial, arbitrary changes in income and wealth pose a threat to ”the cement” in our society and thereby to the democratic system.
Question 2: Why has inflation been allowed?
Average inflation in Sweden has been low over a very long-term perspective. That changed after the Second World War, both in Sweden and in the majority of other industrial countries. During the 1960s, 1970s and 1980s, inflation was on a rising trend and at the same time fluctuated considerably. Not until the 1980s was the growth in prices stabilised in many parts of the Western World. In Sweden, this stabilisation has come about in the 1990s. (Diagram 2. Price Developements in Sweden Since 1880)
Given what I have just said about the importance of low inflation, we may ask ourselves how inflation could have been allowed to reach such heights in recent decades. Why was there not a tighter policy all along the line to restrain inflation? Developments in other countries of the Western World appear about the same. This indicates that there is a common explanation.
It is clear that economic thought has shifted in recent decades and the shift in the price level trends that coincided with the Keynesian approach (which was not infrequently a grossly exaggerated interpretation of what Keynes himself had written or said) had a dominant influence on economic policy. The ambition to stabilise the economy and create full employment using economic policy tools grew stronger and stronger. For a long time it was believed that by allowing inflation to rise a little higher, less unemployment could be ’bought’. Both actual developments up until the 1970s and the theory then prevalent supported this approach.
This picture of how the economy functioned was modified in the 1970s. The economies in many Western countries at that time evinced both high inflation and high unemployment. In fairness it should also be mentioned however that the oil price shocks also contributed to the high inflation that decade; not all problems were a consequence of the failures of stabilisation policy.
Now most economists and political decision-makers accept that only in the short term is there a trade-off between inflation and employment. Real wages can be reduced owing to unexpected inflation and employment will thereby be stimulated. But when this pattern is recognised—as will very likely happen after a while—inflation expectations adjust to the new, higher inflation rate. Then the positive effect on growth disappears and employment reverts to its previous level. All that remains is a higher rate of inflation. In other words, in the long term there is normally no positive relationship between employment and inflation. Instead employment in the long term is decided by such conditions in the real economy as technological and population developments or labour market efficiency.
However the temptation to use monetary policy to stimulate employment still exists. Even for a government that knows that the long-term effects of increasing demand a little too much are undesirable, it can be tempting in the short run. The very fact that this possibility exists has made it more difficult to instil confidence in a policy of low inflation. Even in countries in which the governments and the central banks have clearly switched to a policy directed at achieving price stability, concern about a recurrence may persist. Such a concern is not without costs to the economy. Households and businesses will reckon that there is a risk of an increase in inflation. Consequently the interest rate level may be higher and economic development in general may be worse than what would have been possible if all the actors in the economy had had complete confidence in the anti-inflation policy.
Question 3: Why should central banks be independent?
This problem of building confidence in the anti-inflation policy may explain why in recent years many Western countries have changed their legislation or in other ways have made their central banks more independent. There are basically two problems: the temptation to resort to stimulating the economy in the short term may be too strong and inflation expectations become built into the system and make the policy less effective. By giving the central bank an independent status, the political decision-makers can demonstrate that they are serious about their ambition to let monetary policy be conducted with a view to long-term goals. In such a way price stability can be combined with better economic development.
There have been a good many studies showing that an independent position for the central bank tends to have a favourable impact on inflation. In countries with relatively independent central banks—that is, in which the political regime has only limited possibilities to influence monetary policy immediately—the fight against inflation has been more successful. However this should not be over-interpreted; the problem of inflation neither can nor should be solved solely by a few simple changes in the law.
The central bank’s position certainly depends on what the attitude to inflation is among the general public. It may very well be that in countries with a strong aversion to inflation, they have elected to give the central bank a comparatively independent position. (Diagram 3. The Relation Between a Central Bank´s Degree of Independence and Inflation)
In Sweden the Riksbank has, in practice, had a very independent standing in recent years, which compares quite favourably with that in other countries. This policy has also had a successful impact on price stability. This practice that has been developed has now received support in the form of a broad multi-party agreement to strengthen the position of the Riksbank. The agreement has led to a bill, which the Government introduced last November. This bill provides, inter alia, for the price stability objective to be confirmed by law and for the Riksbank to have a management that can formulate monetary policy independently.
It should be stressed that an independent position for the Riksbank is not the same thing as a monetary policy conducted without democratic control. The monetary policy objective has been made law by those who are popularly elected. It is making the objective operational and the ongoing work to secure the goal decided by the Riksdag that has been delegated to the Riksbank.
Question 4: Why 2 per cent?
The question of how the monetary policy objective should be made operational has several dimensions. By selecting a clear objective for monetary policy, it is possible to monitor and evaluate it in an effective way. This is especially important when the Riksbank is being given an independent status. A clearly defined inflation objective also has the advantage of being able to function as an ”anchor” for households’ and businesses’ expectations. This function alone can be an important feature, not least when a regime of low inflation is being established.
The objective for the Riksbank’s monetary policy is inflation of 2 per cent annually, as measured by the consumer price index. We have a stated tolerance interval of plus or minus 1 percentage point that will catch deviations from the target.
Alan Greenspan, the chairman of the US Federal Reserve, has said words to this effect: Inflation should be so low that the actors in the economy do not need to concern themselves with it when they make economic decisions. This seems reasonable, but what does it mean in more concrete terms?
Most likely it is not advisable to aim at price stability defined as zero-inflation. Some studies indicate that inflation as measured by the consumer price index is overestimated. There are many arguments for that being the case, one being that goods may increase in price because the quality has been improved and another that the items in the valuation basket whose prices are measured may be obsolete. This is a problem of methodology and measurement. Estimates for the American economy show that owing to errors in measurement, inflation tends to be overestimated by up to 2 per cent. Estimates by Statistics Sweden for the Swedish economy indicate a smaller error in measurement.
A very low inflation may also cause problems if the nominal wages are inflexible downwards, which is often the case. Some degree of inflation can then function like oil in the economic machinery. However price and wage rigidity do not follow any natural law, independently of how, for example, economic policy is formulated. After a period of low and stable inflation, it normally becomes more common and more generally accepted that prices and wages not only can be raised, but also can be lowered.
In addition striving after absolute price stability with margins of error that are in all likelihood valid will sometimes lead to price increase rates that may be negative. Such a situation is not entirely free of risk, since it may be difficult to get out of a deflationary situation.
Should inflation then not be allowed to be higher than 2 per cent? The basic argument against this assertion is that the disruption in the economy will be greater the higher inflation is allowed to rise. In practice the fact that other countries in Europe, especially Germany, have a criterion of 2 per cent for price stability has understandably had a decisive influence in Sweden. Moreover corresponding assessments have also been made by the majority of the countries outside Europe that work with explicit inflation objectives.
Question 5: Can´t the Riksbank do something about unemployment?
Sometimes the view is expressed that low inflation is indeed good, but that it leads to a very one-sided direction for monetary policy; the Riksbank should also have an objective for employment.
From what I have said thus far, I hope that it has been clear that there is no long-term conflict of interests between inflation and employment. On the contrary, there are many indications that low inflation is a prerequisite for stable and high growth and for favourable employment developments. Therefore, low inflation is the best contribution that monetary policy can make to the fight against unemployment.
It is only in the short term that high inflation may lead to lower unemployment. However to utilise such a trade-off has its risks, as the experiences from the 1970s and 1980s that I spoke about earlier indicated. The mere concern that policy will be too expansionary and that inflation may sprint ahead can push up a country’s interest rates and results not in higher but in weaker growth and employment developments.
Therefore it is crucial for monetary policy that confidence in the low inflation regime be preserved.
The very construction of the inflation objective implies that economic growth is taken into account in a certain way. A weaker growth, which creates excess capacity in the economy, normally means that inflationary pressures from the demand side will decrease. This in turn produces room for a more expansive monetary policy—interest rates will be lowered and demand will be stimulated. Conversely a very high demand signifies that production capacity risks hitting the ceiling. Prices and wages will then be pushed up. In order to prevent such occurrences, monetary policy must be tightened again and interest rates raised.
The time horizon over which the Riksbank strives to achieve the inflation target also has an effect on growth and employment in the short term. A quite short time horizon means that substantial changes in the interest rate are required in order to achieve the objective. Such substantial changes may in turn result both in monetary policy appearing abrupt and in undesirable and significant fluctuations in growth and employment. In contrast a very long time horizon will bring greater short-term stability to growth and employment, but at the same time lead to considerable fluctuations in inflation. These fluctuations may weaken confidence in the price stability policy and lead to higher instrumental rates over time. Here monetary policy is obviously faced with a balancing problem. The more credible the policy is, the more room there is to manoeuvre in practice.
There may be reasons for completely abstaining from reacting to certain inflationary impulses of a one-time nature. An example of such “one-time inflation” is indirect tax hikes. If price disruptions are judged to have merely temporary effects on inflation and do not affect expectations of future inflation, it would imply unnecessary real economic instability to prevent the price increases from having an impact on the consumer price index.
Question 6: Doesn’t the Riksbank itself ”talk up” inflation?
Now and then the criticism is made that the Riksbank stresses the risks of inflation in the economy and is thereby guilty of ”talking up” inflation. When this is carefully done the Riksbank—say some of the critics—can use the expected higher inflation (that we ourselves have created) as an argument for making the interest rate increases we have wanted to make all along. This type of criticism was common last autumn, for example, when the Riksbank spoke in various connections about its inflation outlook and explained that interest rate increases might be necessary.
I have the following answer to these criticisms. To begin with, the Riksbank’s focus on inflation developments is entirely natural. It is undoubtedly our task to show clearly the reasons for our decisions. Without a candid accounting of monetary policy from the Riksbank’s perspective it would be more difficult for the political decision-makers and the general public to evaluate monetary policy and our fulfilment of our objective. Think what type of debate we would have and how monetary policy would be seen if the Riksbank only made its decisions without commenting at all on the outlook for inflation and without giving reasons for its policies.
Thus the need for the Riksbank to speak about inflation is obvious. Naturally it cannot be ruled out that in doing so it will influence households and businesses in their views on price developments, however our influence should not be exaggerated. The cause-and-effect relationship is rather the reverse: the Riksbank reacts to inflationary developments, which in turn are determined by a great number of decisions by businesses and households. However, of course the Riksbank does not react in a mechanical manner to a measured change in inflation expectations. The Riksbank currently monitors a number of indicators of future inflation developments and, responses to its surveys about inflation expectations are just one of many.
By way of conclusion there is reason to call attention to the fact that inflation expectations to an ever greater degree are being guided by the inflation objective and not by our or others’ discussions of the risks of inflation. The signs of this are clear. The expected inflation has crept down markedly and has now been stabilised in the vicinity of the objective for most of the groups currently monitored by the Riksbank. This is a profoundly positive development, which actually facilitates the monetary policy task. (Diagram 4. Inflation Expectations of the Players in the Money Market Agents)
Question 7: Is the policy asymmetrical?
It takes time before a change in the instrumental rate affects the economy and inflation. Therefore monetary policy always has to be formulated with the aid of forecasts of future inflation. When the Riksbank formulates monetary policy it is done with the aid of a forecast one or two years into the future. A simple rule of thumb should be able to describe monetary policy: if our overall assessment of inflation—assuming that the instrumental rate remains unchanged—indicates that inflation will deviate from the objective for two consecutive years, the instrumental rate must be changed in the corresponding direction.
Monetary policy is always adjusted with the aim of achieving the 2 per cent objective. Occasionally the argument has been presented that the Riksbank should accept that inflation has risen to the upper range of the interval around the objective. This argument is difficult to reconcile with the conduct of monetary policy in general. To begin with, it is clear that if the Riksbank did so, inflation would more often be outside the tolerance interval. In the second place the rate of inflation would on average be higher than 2 per cent unless the Riksbank had the same tolerance for inflation staying at the lower edge of the band in an economic downturn. However such a policy would instead reinforce the real swings in the economy, in contrast to a policy that is always focused on the objective. In upswings the Riksbank would allow more growth and inflation and in downturns we would do the opposite.
There is not any asymmetry in monetary policy and there should not be any either. One of the purposes of a clearly formulated objective is simply to inform what the policy aims at. In such a way the inflation objective may anchor the price expectations in the economy. The Riksbank should be evaluated in light of this objective. If the Riksbank systematically aimed below the objective or always overestimated the risks of inflation, it would be manifested over time by too low inflation on average.
That the Riksbank always aims at the objective of 2 per cent unfortunately does not mean that the price increases will always be limited to 2 per cent. First knowledge of the path leading from interest rate changes to their impact on inflation is insufficient to enable us always to expect to be able to achieve the objective. Second price increases as measured by the consumer price index will continually be affected by various kinds of temporary factors. There may be, for example, changes in indirect taxes or external shocks in the form of changed oil prices. Often such disruptions should not be counteracted (as long as they do not risk affecting the long-term inflation process). In addition, there is a continuous flow of new information that can change the inflation outlook. These factors can make the consumer price index diverge from the target ex post, so to speak. This insight is the reason for the Riksbank’s decision to establish a tolerance interval around the inflation objective. (Diagram 5. CPI and Underlying Inflation)
How has it worked then? Is there any basis for asserting that the policy has been asymmetrical? Since the inflation objective came into effect in 1995, the consumer price index has, on average, increased by 1.5 per cent per year. During the same period the underlying inflation was on average between 1.5 and 2 per cent, depending on how it was measured. It is primarily the significant decline in interest rates, which lowered home ownership costs in the consumer price index, that explain the discrepancy.
Interest rate costs are not included in the underlying measurements. Accordingly, if one takes into account the impact of the one-time effects on the consumer price index, the Riksbank has been relatively successful in meeting the inflation objective. However in order to make a more meaningful evaluation, a longer time period is required.
Question 8: Why did the Riksbank raise the interest rate and what happens now?
To conclude, let me make this discussion of monetary policy principles somewhat more tangible by describing what was behind the Riksbank’s decision to raise the interest rate in December and say a little about how we look at the situation today.
To simplify, one can say that the Riksbank bases its assessment of future inflation on two factors. The one is supply and demand. The other is how inflation expectations develop.
As a rule, we assume that it will take one to two years until an interest rate change has its full impact on the economy. Thus the interest rate reductions that were made in 1996 can be assumed to be an important driving force behind the upswing that now seems to be on the way. The strengthening of economic growth in Europe also gives important support.
The monetary policy measures taken by the Riksbank affect price developments with a lag and therefore the measures have to be forward looking. In an economic upswing it is thus important that monetary policy is not expansionary for too long. The switch-over should come early and be made gradually before the economy reaches full capacity utilisation. Such a gradual change in monetary policy will slow down the expansion of the economy in the short run, but the result will be better conditions for a more stable increase in production and employment in the long run. If instead the Riksbank were to wait too long, there would be a risk that the interest rate changes would have to be bigger, resulting in negative short-term effects.
In the Inflation Report of December, the assessment was that economic growth would be around 3 per cent in 1998 and somewhat more in 1999. The rise in demand was expected to result in a gradual reduction of the output gap in the coming years and in increasing inflationary pressure. Given an unchanged repo rate, inflation was expected to rise gradually to 2.5 per cent during 1999. In order to avoid more drastic changes in monetary policy at a later stage, we chose to adjust monetary policy in a somewhat less expansionary direction. The repo rate was raised in December from 4.10 to 4.35 per cent. (Diagram 6. Inflation Forecast December 1997)
The current crisis in Asia was discussed chiefly in our risk scenarios. A small reduction in the growth of exports was made in the main forecast. But it was primarily a possible deepening of the crisis compared with the situation in December that, it was deemed, might have substantial effects on economic and price developments in Sweden. Thus the Asian crisis was taken into account in our risk assessment.
Since then the crisis in Asia has worsened, both in terms of its effects on the financial markets and its likely consequences for real economic growth. For Sweden’s economy it means that the negative effects of the Asian crisis probably will be somewhat more extensive and prolonged than we expected in December.
All in all, the prospects for inflation from the demand side look a bit better than they did when the repo rate was last raised. Another positive sign is that the wage agreements that have been concluded so far in 1998 indicate a rate of wage increases that is somewhat lower than in the main scenario described in December. However the recent weakening of the krona is a cause for concern. If it turns out to be permanent, it could affect inflation prospects.
In the Riksbank we are now finalising the next Inflation Report, in which we will present an inflation forecast reaching into the first quarter of the year 2000. There, factors such as the Asian crisis, the wage negotiations and the exchange rate developments will be weighed and put into a larger context. The view of the inflation prospects that will result from this process will form the basis for future monetary policy.
