Heikensten: The Riksbank and stabilisation policy
Let me begin by thanking you for the invitation to attend this seminar. I look forward to an interesting and rewarding discussion.
The invitation to the seminar contained a list of questions for discussion. I intend to structure my contribution in terms of working my way down the list and giving my views on these questions.
Before I do so, however, I would like to take this opportunity to say that I appreciated the section on stabilisation policy’s possibilities and limitations in the LO economists latest “Economic Prospects” report. Although I do not agree with everything – which will become clear later – I perceived the article to be serious and searching in a positive way.
Is stabilisation policy important?
My answer to this question is (of course): Yes, stabilisation policy is important. One only needs to look at our experiences in the 1970s and 1980s to realise this. What happened then was that the policy conducted could not sustain an adequate balance in the economy. The policy conducted was often too expansionary, which created an environment where it was difficult to keep price and wage increases at a reasonable level. The rapid increases in wages and prices were repeatedly in collision with the fixed exchange rate. The result was that the krona was devalued several times from the mid-1970s onwards. The economy underwent "boom-bust-cycles” with alternate sudden braking and strong stimulus for acceleration. This policy had devastating effects on many aspects of economic developments. Sweden lost ground in terms of prosperity, while growth in real wages was largely non-existent. This situation finally become untenable and after the fixed exchange rate system was abandoned in November 1992 there was a clear shift in stabilisation policy regime.
When one looks back over the period since then, I believe that most people will agree with me that the new regime has been successful. Sweden is no longer afflicted with excessively high inflationary trends, recurring cost crises and high interest rates. The more stable environment has contributed to better growth in prosperity. The new regulations have also created other, better frameworks for wage formation, a point often emphasised by LO representatives. Trade union organisation no longer need to play along in a destructive game where high nominal wage increases are eaten up by inflation and real wage increases prove to be a disappointment year after year.
The fact that inflation has been lower and more stable during the period with inflation-targeting policy is a good result, although perhaps not surprising. What may have received less attention and been less evident, given the theories in this field, is that there has been less fluctuation in the real economy. It thus appears as though the new regime has also had a beneficial effect on the stability of the real economy.
Now that it appears as though the more fundamental stabilisation policy problems have been managed, it may seem natural to gradually become more ambitious and wonder if things could function even better.
However, it is important to have realistic expectations as to what the stabilisation policy can achieve. It has been easy to gain an impression in the recent debate that there are expectations that the fluctuations in both inflation and the real economy – perhaps especially the latter – can be eliminated entirely. When the heads of the National Labour Market Board make comparisons with a situation where unemployment is always around 4 per cent and then claim that the Riksbank bears the responsibility for the difference, they appear to be expressing this view.
It is unavoidable that the economy and inflation undergo rapid changes from time to time that cannot be predicted or fully counterbalanced in the short term. No central bank – or any other economic policy decision-maker – can thus steer inflation and economic activity completely in the long term. In actual fact, experiences show there is a substantial risk that actively fine tuning stabilisation policy can reinforce fluctuations in the economy. It may be worth noting in this context that the rise in unemployment we have observed in Sweden during the most recent economic downswing does not in any way appear remarkable from an international perspective.
It is also extremely difficult to know what comprises a “normal” rate of unemployment, particularly as this is also something that changes over time. It is important that the Riksbank does not commit itself too strongly to a particular figure. We must be willing to test our way forward and if there is sufficient confidence in our policy, we must be prepared to question old truths. However there is reason to be sceptical of those who think they have infallible answers in this field.
So, if I may modify my response to the question: Yes, stabilisation policy is important, but one should not overestimate what it can achieve.
The surprisingly weak development in the labour market is of course an important reason why the debate on stabilisation policy has recently intensified. I agree that we must take the problem of unemployment seriously. However, this is not achieved by primarily concentrating the discussion on stabilisation policy and repeatedly blaming weak employment growth on the monetary policy conducted in recent years. It is particularly grating when those pursuing this line of debate – the National Labour Market Board and others – have made essentially the same assessments of the economy and employment as the Riksbank at the points when the crucial decisions were made. We were all surprised by the relatively high growth rate combined with low inflation and almost unchanged employment.
In times such as these, with extensive structural changes and "shocks” succeeding one another, it is especially important to consider the economy’s capacity to adapt. The conditions for achieving lasting success in the form of higher employment and lower unemployment are greater if the measures are aimed at making the economy less sensitive to shock rather than at using stabilisation policy to try to fend off shocks. Moreover, it is much simpler to find useful measures that can be taken if one adopts the former policy rather than the latter.
The focus on stabilisation policy is also surprising given the fact that the nominal interest rate is the lowest it has been since the 1960s and 1970s and that even when measured in real terms, i.e. nominal interest rate minus expected inflation rate, it is among the lowest we have seen in 20 years. Monetary policy is thus clearly expansionary. At the same time, fiscal policy is also expansionary and has stretched the limits of what is compatible with the long-term regulations.
So, if I can modify my answer to the question even further: Yes, stabilisation policy is important, but it is even more important to undertake measures that enable the economy to adapt smoothly to new conditions. These are the type of issues we should discuss now that we actually have a system where stabilisation policy no longer contributes to creating severe economic fluctuations – as it did in the 1970s and 1980s – but instead seems to actually have a stabilising function.
How should roles be allocated between fiscal policy and monetary policy?
The change in regime in the 1990s involved the introduction of a distinct allocation of roles in Swedish economic policy. The exchange rate remained variable and monetary policy was given the main task of trying to keep inflation at a low, stable level. Fiscal policy, which had often resulted in large deficits under the previous regime of a fixed exchange rate, would now instead be targeted at achieving long-term stability and sustainability in public finances. Corresponding changes in economic policy had been made in many other parts of the world.
It is often said in the economic debate that one consequence of the change in system was that “the Riksbank was given responsibility for stabilisation policy”. As it is not always clear what is meant by this exactly – the interpretations appear to vary depending on whom is asked – it may be worth providing a brief summary of the inflation-targeting policy we conduct.
The Riksbank’s primary objective is to maintain price stability. This task often goes hand in hand with a stabilisation of the real economy; if demand in the economy develops in line with the long-term potential, it will often contribute to a stable development in inflation. However, inflation does not solely depend on our domestic resource utilisation; it is also influenced by supply shocks, international inflation, exchange rate trends, and so on. In certain situations it is necessary to find a balance between stabilising inflation and stabilising the real economy, for instance, when a severe cost shock pushes up inflation while the real economy is weakened. The typical result of weighing up these considerations is that – unless it would threaten credibility – the central bank will allow much longer time for inflation to return to target level than if the rate of price increase had been pushed up by increased demand. The negative effects on the real economy will therefore be much less significant than if the central bank had merely reacted to the higher inflation and used an aggressive policy to rapidly bring inflation back in line.
Considerations of this nature have been made by the Riksbank on several occasions during its decade of inflation targeting. There are examples from the end of the 1990s when deregulation contributed to very low inflation and the Riksbank chose not to fully utilise the scope for interest rate cuts that might been offered by a stricter interpretation of the inflation target. There are also examples of the reverse from 2001 and 2003, when temporary supply factors pushed up inflation and the Riksbank did not react by raising the interest rate. To be even more explicit; when we formulate our policy we have on several occasions deliberately chosen to allow deviations from the inflation target for short or long periods of time. When the consequences of the deviations have been lower interest rates than would have been the case with a stricter interpretation, there has been practically no criticism. However, it has been rather different with regard to deviations in the other direction.
Neither the Riksbank nor any other central bank using inflation targeting can be described as an “inflation nutter”, to borrow a famous phrase from Mervyn King, now Governor of the Bank of England. That is, we do not focus solely on stabilising inflation; we also take into consideration the real economy – conducting what is usually termed flexible rather than strict inflation targeting policy. This is also entirely in line with our task as laid down in law.
There are also other types of consideration that need to be made in monetary policy. One that is immediately apparent is how financial stability will be affected by the interest rate decisions. Over the past year, and also on a few earlier occasions, there has been extensive discussion of developments in the credit market at our monetary policy meetings. We have relatively short experience of how credit cycles affect the economy; up until the 1980s the financial markets were strictly regulated. Our knowledge of the risks entailed in a rapid increase in indebtedness and asset prices is thus limited. Although it can probably be said that the credit expansion and house price trends have not yet had decisive significance in our interest rate decisions, they are factors that must be included in the equation.
It is also important for the Riksbank to try to find a suitable balance between stabilising inflation and the real economy and also taking into account the consequences for financial stability. In the same way, fiscal policy must meet the requirements for long-term stability and sustainability in public finances, with the aid of budgetary targets and expenditure ceilings, which does not exclude consideration of the real economy. This allocation of roles has proved successful and we should therefore proceed with caution as regards making any drastic changes.
My reply to the question of how roles should be allocated between fiscal policy and monetary policy is therefore; as they are now. The same conclusion has been reached in many other countries as well as in academic research. What is important in this context is that one realises that the Riksbank takes real consideration when formulating its monetary policy, to the extent it deems this to be wise. It is possible, however, that this is something we need to be clearer about.
Can and should monetary policy and fiscal policy develop?
In my opinion, one can of course say that both monetary policy and fiscal policy are constantly developing. At the Riksbank we are working continuously on, for instance, improving our analyses and forecasting methods to provide the best possible basis for making decisions. Our decision to publish inflation forecasts based on market expectations in the March Inflation Report was an example of this. One can assume that the Ministry of Finance conducts equivalent development work.
I believe that an even deeper discussion of what is entailed in the considerations made between inflation targeting and the real economy can be a good thing. During the first years of inflation targeting, the Riksbank sometimes sounded like an “inflation nutter”. However, this stance was abandoned after a few years and since then we have communicated that we also take into consideration financial and real factors. A decision by the Executive Board of the Riksbank in 1999 clarifies our views on these issues. However, we can probably become even better at analysing and explaining the considerations we have to make when conducting a flexible inflation-targeting policy. The Executive Board has held several speeches on this theme over the past year and the aim is to publish some form of joint document on this subject in the near future.
With regard to fiscal policy, it has a stimulating effect on the economy primarily through what are known as the automatic stabilisers, i.e. through systems for taxes and expenditure, which are very extensive in Sweden, being designed to allow some “automatic” adjustment to cyclical developments. In addition, as I mentioned earlier, fiscal policy can play a decisive role in the economy’s adjustment to shock, through its taxation and expenditure systems. The government’s economic policy arsenal also contains labour market policy, which can play an important role in both facilitating adjustment and maintaining demand for labour in the short term within individual sectors or regions. There is reason to continue looking for ways to develop policy in all of these fields and to make it more effective.
However, in practice there is a risk that the potential for allowing automatic stabilisers to take effect or for putting in more extensive labour market policy measures is limited unless public finances are sufficiently strong. This is a conclusion that I share with the LO economists, if I have understood correctly. However, in practice the LO economists have rarely manned the barricades and argued in favour of stricter policy even during good times. Although a change may be under way here.
But what the question of whether monetary and fiscal policy should be developed probably refers to is whether there is reason to modify the allocation of roles between monetary and fiscal policy that has prevailed since the introduction of the new stabilisation policy regime. This is based of course on LO’s suggestion that the Riksbank should “receive assistance” from discretionary fiscal policy in stabilising the real economy. The idea is, if I have understood correctly, that the Riksdag Committee on Finance or some part of it – a special stabilisation policy committee – would be given a greater role, with an analysis function and the authority to implement discretionary fiscal policy decisions when necessary.
I have no faith in this idea, for several reasons. One fairly obvious reason is that a stabilisation policy committee would of course face the same difficult problems as the Riksbank and the government with regard to predicting the future. The fact that inflation is currently low is largely due to surprisingly positive surprise shocks that were not predicted by any analysts, rather than to the Riksbank making assessments that deviated from others or holding back demand. Difficulties in predicting and fine tuning the economy are something that everyone entering this field will face and scarcely a good reason for establishing a new public body.
Another important reason for my doubts towards this proposal is connected with the leading principle for reforming monetary policy, namely that it should be formulated "at arm's length" from the current party politics. From what I understand of the proposal, the idea is that these problems will be managed by setting rules in advance that will indicate the situations in which discretionary fiscal policy measures shall be implemented and phased out. Of course, one can have different opinions as to how successful this might be. But if one compares it with the rigorous institutional framework that encompasses monetary policy, it is difficult to rid oneself of the thought that the “arm" which would mark the distance from present party politics would be on the short side.
I believe that there are some worrying signs here. It appears as though there is a tendency, from a political point of view, to see stabilisation policy, or more precisely the repo rate, as the main solution to the problems we are currently seeing in the Swedish labour market. As I see it, there is an evident risk that this type of view would also colour the discretionary fiscal policy with an increased stabilisation policy responsibility.
However, before I round off, allow me to say that it is possible that situations might arise that are so serious from a stabilisation policy point of view that discretionary fiscal policy measures could be warranted. Here I am mainly thinking of situations with a severe fall in demand in the nature of a depression, where both fiscal policy and monetary policy would be forced to make substantial departures from the normal principles. These would be extremely unusual circumstances, however, and do not provide any argument for setting up a permanent stabilisation policy committee.
So, my answer to the question on whether monetary and fiscal policy should be developed is yes, but not in the direction LO appears to have in mind. With regard to monetary policy, this means continued improvement of analysis and forecasting methods and to find ways to clarify the balance between the inflation target and the real economy. For fiscal policy it could mean finding ways to make the automatic stabilisers even more effective, but primarily it entails designing preventive systems to enable smoother structural adjustments in the economy when various shocks unavoidably occur. It does not entail any changes in the fundamental role allocation as regards stabilisation policy.
Summary
Let me briefly summarise my responses to the three questions. Stabilisation policy is important, but one should not overestimate what it can achieve. Other economic policy measures are much more important to long-term growth in prosperity. A smoother adjustment process when the economy suffers shocks could benefit employment more than an active stabilisation policy. The allocation of roles between fiscal policy and monetary policy with regard to stabilising the macro economy has functioned well in the new regime and I see no good reason to change it. Both fiscal policy and monetary policy can and should develop, but this should occur within the framework of the present allocation of roles regarding stabilisation policy. For monetary policy this is a question of analysing and describing the considerations made between the inflation target and developments in the real economy. The most important issue regarding fiscal policy is not measures aimed at stabilisation, but providing a good general foundation for prosperity growth.
Thank you!
