Bäckström: The inflation targeting approach
First a word of thanks for the invitation to attend this conference and discuss challenges for monetary policy together with such a distinguished panel.
In the last decade a growing number of countries have chosen to conduct monetary policy with an explicit target for inflation. One reason behind this choice has no doubt been these countries’ poor experience with a fixed, but adjustable, exchange rate regime.
The first country to formally adopt a policy of targeting inflation was New Zealand, in 1990. Canada did the same in 1991, followed by the United Kingdom, Sweden and Australia. Since the early 1990s other countries have also introduced some variant of inflation targeting.
It is interesting to note that before the 1990s the predominant view was that a floating exchange rate regime was not suitable for a small open economy. Today, I believe that experience among those countries that are targeting inflation has been exceptionally good. In fact, even some emerging market countries are now building the same kind of regime.
It is not without some pride that I recall that, in a sense, Sweden was actually something of a pioneer in explicitly focusing monetary policy, with a flexible exchange rate, on a specified target for prices.
Back in 1931, after Sweden had been obliged to abandon the gold standard, its government and the central bank declared that the overriding objective of monetary policy would be the stabilisation of prices. More specifically, the finance minister of the day described the task of the central bank as that of, "using all means available, preserving the domestic purchasing power of the Swedish krona". In this way, the Riksbank became the first central bank to declare price stabilisation as its policy norm with a floating exchange rate.
The norm was based on a proposal made much earlier, in 1898, by Knut Wicksell, often regarded as Sweden’s greatest economist at the time. Wicksell’s rule for monetary policy, which stemmed from his extension of the classical quantity theory of money, was simple: "The central bank is to raise the discount rate as long as prices are rising, lower it as long as prices are falling and keep it constant when prices are stable".
In other words, price stability was to be achieved through interest policy. Although the rule applied to the price level rather than to inflation and was not forward looking, it is a clear parallel to the rules of thumb that are used today by those central banks that target inflation.
Sweden’s price stability norm in the 1930s lasted only a few years and was followed by a return to a fixed exchange rate regime. During those years, however, it performed the central function of stabilising expectations as well as prices. This was important not least because to some extent it helped shield Sweden from the global economy’s depressive impulses and thereby contributed to the recovery being speedier than in most other countries.
Here I should perhaps add that although this Swedish price stability "experiment" attracted a good deal of attention at the time, it certainly did not leave any deep or lasting mark on the modern approach to inflation targeting. In the aftermath of the depression and the second world war, Keynesianism gained ground. The introduction of inflation targeting in Sweden in the 1930s was therefore not as pioneering as when this kind of regime was introduced in the early 1990s, first by New Zealand and later by others.
A great advantage with a regime that explicitly targets inflation, is that the central bank adjusts the interest rate continuously in the light of inflation prospects. Compared with a fixed exchange rate regime, imbalances can then be prevented from building up to the same extent. A slight upward adjustment of the interest rate at an early stage, to reduce the risk of inflation picking up, is something quite different from having to hike the short-term interest rate in order to defend a fixed exchange rate.
As a personal reflection I can say that explaining the importance of low inflation is much easier – not least for a central bank governor – than the task of selling the advantages of a stable currency. This is particularly true as the latter task becomes increasingly pressing if the economy has been hit by problems with costs and is facing speculative attacks. Some of you may remember that for a few days in the autumn of 1992 the instrumental rate in Sweden was raised to as much as 500 per cent. In my view, this is probably one important explanation why Sweden is not a member of the ERM system – at least not at this stage.
The challenge during the 1990s for the inflation targeting countries has been to build a rigorous framework around the new regime. Answers have been needed in particular to questions like the following: What rate of inflation should be targeted, what index should be used, are bands around the target necessary, the degree of transparency etc.
It has also been important to formulate a monetary policy reaction function, that is, to define the balance between inflation variability and output variability. The reason why central banks do not aim for maximum short-term stability in either inflation or output is that attaining only one of the targets would be costly.
On the one hand, the central bank’s legitimacy among the general public could be lost if short-term interest rates have to move sharply and cause output to fluctuate widely while inflation is being held stable and exactly on the target. Maximum stability in output could, on the other hand, result in sharp fluctuations in the rate of inflation and erode the target’s credibility. Instead it is a matter of finding a point somewhere midway between the two extremes.
Looking ahead, I believe there are at least two major questions that central banks have to address more thoroughly.
First, how should monetary policy react to asset prices, such as equity and/or real estate prices? We know from history that the development of asset prices can have a significant impact on both inflation and real economic activity. We need to establish whether or not there are actions that central banks can and should take to minimise the likelihood of macroeconomic instability arising from extreme fluctuations in asset prices.
Second, how should monetary policy react to structural changes in the economy? On this issue quite a lot of thinking has already been done, both among academics and policymakers. Nevertheless, I believe that more must be done in this field, especially since at least some parts of the world economy have seen something of a productivity shock in recent years. I am carefully trying to avoid using the buzzword "new economy" here, since history has been full of expressions like that ever since the turn of the 19th century.
Unfortunately, I do not have the time to elaborate on the first question. Instead I will say a few words about the second.
Although the discussion about monetary policy and a productivity shock is taking place mostly in the United States, in the light of the impressive economic performance there and the acceleration in the rate of productivity growth, one could certainly ask: What about Europe?
Europe is, on the one hand, now in a strong upward economic phase. In addition, new technology, especially the Internet and telecommunications, is gaining ground in many European countries. On the other hand, Europe also differs from the US. The rapid growth of US investment has not yet been seen in Europe. One could also argue that the existence of quite a few structural obstacles in Europe makes an acceleration of productivity growth less likely in the immediate future.
A handful of countries – including the Netherlands, Ireland, the UK and, more recently, Sweden and Finland – have achieved strong economic growth and in some cases also low inflation. But with the exception of Ireland, these countries have not yet enjoyed the same strong productivity growth as the US.
Nevertheless, the possibility of a productivity breakthrough also in Europe cannot be ruled out. It will probably depend on how well Europe succeeds in reforming microeconomic policy.
Higher – or accelerating – productivity growth does indeed pose new challenges for monetary policy. It could set in motion a complex of effects on aggregate supply and demand, on inflation, equity prices and interest rates. Just to give a few examples, let me briefly mention the following effects:
First, a positive productivity shock raises the economy’s potential growth rate. It also affects aggregate demand through new, profitable investment opportunities, higher equity prices and expectations among households of higher permanent income. One question for policymakers is whether the demand effects are so powerful that they have the potential to outpace the growth of aggregate supply in the short and medium perspective.
Second, there could be temporary disinflation effects at work if the increase in productivity is unexpected initially. If wages are slow to adjust, higher productivity growth lowers unit labour costs and thereby reduces price inflation. Here again there are important questions for policy makers. Should this effect be allowed to result in a temporary reduction of price inflation or unemployment?
Third, an increase in the trend rate of productivity growth could also result in a higher equilibrium real interest rate, when the output gap is closed, in order to balance saving and investment. Fiscal policy and the development of public financial balances could at least partly offset this effect, as could the economy’s degree of openness.
All this implies that monetary policymakers, also in Europe, could face quite a few challenges. One of them is to identify structural changes and distinguish these from cyclical variations. Another challenge is to detect the magnitude of the effects. Furthermore, looking ahead in monetary policy is hard when the economy is going through major changes that make it difficult to forecast the future path of the economy and inflation.
It is quite clear that this demands 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 measurement of the output gap and the relationship between the output gap and inflation.
It is already evident, for example, that in the 1990s inflation forecasts in many countries were frequently on the high side. I believe that the best we can do, at least at this stage, is to pay close attention to incoming data in the forecasting process, and not mechanically use the results from models that rely too heavily on historical relationships. We have to deal more successfully with the fact that forecasts of economic developments can never be completely accurate and still allow monetary policy to be forward looking.
When discussing monetary policy of today, we must remember that we have come a long way. The world economy shows strong growth coupled with low inflation and monetary policy has contributed to this development. However, there are interesting, though somewhat difficult, challenges ahead of us which must be met if we are to preserve this bright picture of the world economy. This conference will contribute to a better understanding of the issues and problems involved.
