Svensson: Monetary policy and the interest rate path

Date 22/08/2007

“The Riksbank conducts what is known as ‘flexible’ inflation targeting. By this we mean that we attempt both to stabilise inflation around the target and to stabilise resource utilisation. ’Strict’ inflation targeting would mean that we only stabilised inflation around the target regardless of the consequences for resource utilisation. We would then try to attain the inflation target as quickly as possible. This would require many and considerable changes in the interest rate as soon as new information led to a forecast indicating that inflation would over- or undershoot the target. This would in turn lead to substantial fluctuations in resource utilisation and thereby output and employment. It is sometimes assumed in economic research that the central bank conducts strict inflation targeting, but this is never more than a pedagogical simplification. In practice, all central banks with inflation targets conduct a flexible policy and also attempt to stabilise resource utilisation,” began Mr Svensson.

”But the extent to which the central bank chooses to stabilise resource utilisation may vary over time. When the inflation target is new and it is important to establish credibility for the target, it is natural that the central bank should devote itself mostly to keeping inflation close to target and give less consideration to fluctuations in resource utilisation. When credibility has been built up and the central bank has begun to find its feet it is natural that it should give more consideration to also stabilising resource utilisation. But the importance given to stabilising resource utilisation may never be so great as to threaten the credibility of the inflation target," continued Mr Svensson.

”It is also important to remember that the only thing monetary policy can do is to stabilise resource utilisation around its average level. Monetary policy can not influence the average level. Therefore, it makes no sense to have a monetary policy target for average resource utilisation. In contrast, it makes a lot of sense to have a monetary policy target for inflation,” continued Mr Svensson.

”It normally takes time before inflation and resource utilisation react to changes in the interest rate. Milton Friedman used the expression ’long and variable lags’. It can take a year for output to react and another year for inflation. This is why effective inflation targeting must be based on forecasts, despite the fact that the forecasts are always uncertain (and become more uncertain the longer the forecast horizon). The Riksbank’s interest rate decision is therefore based on forecasts for inflation and resource utilisation at a 1-3 year horizon,” said Mr Svensson.

”Moreover, the current repo rate does in itself play an insignificant role for future output and inflation. The important thing is what firms, households and market agents believe will happen to the repo rate further ahead, in one or more years’ time. These expectations affect long-term interest rates, which have greater significance for decisions on output, consumption, investment, prices, wages and so on. In fact, central banks’ decisions on the current interest rate matter almost exclusively only to the extent that they affect expectations. Monetary policy actually consists mainly of ‘the management of expectations’, as economist Michael Woodford has expressed it,” continued Mr Svensson.

“Flexible inflation targeting means that the central bank must choose an interest rate path such that the resulting forecast for inflation and resource utilisation ‘looks good’. ‘Looking good’ means that inflation is on target and that resource utilisation is normal 2-3 years ahead, or that inflation is approaching the target and resource utilisation is approaching a normal level at an appropriate pace. The Riksbank also uses the expression ’well-balanced’ monetary policy for this. In my research I have called this method of using forecasts ‘forecast targeting’, as it involves choosing an interest rate path such that the inflation and resource utilisation forecast looks good,” said Mr Svensson.

“The forecasts for the interest rate, inflation and resource utilisation form a natural trinity. It is not possible to determine a forecast for inflation and resource utilisation without having a forecast or assumption for the interest rate path. This is why all central banks who aim to stabilise inflation, including the ECB and the Fed, somewhere in their documents preparing the decision also have a forecast of or assumption about the interest rate, even when they do not publish it,” continued Mr Svensson.

”The Riksbank’s decision to publish an interest rate path should be seen in this context. Monetary policy works by affecting expectations about the future interest rate. It is the entire interest rate path that is important for future inflation and resource utilisation, not merely the interest rate over the coming weeks. The Riksbank has therefore come to the conclusion that the only right thing is to explicitly discuss the interest rate path and to choose a particular path as the main forecast, as well as publishing the interest rate path and justifying its selection. This is in my opinion the most effective way of conducting monetary policy. Not to discuss and select a particular interest rate path as a main forecast would be an incomplete decision-making process. Not to publish the interest rate forecast would be to hide the most important information,” said Mr Svensson.

“The Riksbank is not the first central bank to implement this monetary policy innovation. The Reserve Bank of New Zealand began to do this as early as 1997. A number of researchers, including myself, have argued for several years that more central banks should publish their own interest rate forecasts. I was therefore very pleased when the Norwegian central bank began to do so in 2005, and I was even more pleased when the Riksbank began in February this year. The publication of interest rate paths is now being widely discussed in both academic and central bank circles around the world. I would not be surprised if more central banks begin publishing their own interest rate forecasts in the future,” said Mr Svensson.

”There has been much discussion of the two interest rate forecasts the Riksbank has so far published, in February and June this year. Many market agents considered that the interest rate forecast in February was too low. However, in connection with the June forecast some agents were surprised at how much we had adjusted the forecast upwards. But I do not consider it strange that some analysts were surprised by our forecasts for the interest rate. It is natural that it will take some time for the new system to become established, including the interplay between the forecasts for the interest rate, inflation and resource utilisation. That the interest rate forecast can be revised when new information has been received is quite natural. The interest rate path we publish is a forecast based on the information available at that time and is certainly not a promise! It is in fact a probability distribution and is also published in the form of a mean value with a substantial uncertainty interval and with alternative scenarios. Even when the system has become established, there will be scope for different interpretations of new information and what change in the interest rate path this may justify,” continued Mr Svensson.

”Forecast targeting, the method of using forecasts I described earlier, also shows how monetary policy with an inflation target has to take into account new information regarding the economy, economic activity and international developments. This new information is only important for monetary policy if it affects the forecast for inflation and resource utilisation when the interest rate path remains unchanged. If the new information does not affect the forecast in any definite way it therefore has no impact on monetary policy. If the new information pushes up the forecast for inflation or resource utilisation (or both), there is reason to adjust the interest rate path upwards. If the forecast for inflation or resource utilisation is pushed down, there is instead reason to adjust the interest rate path downwards. If the new information pushes the forecasts for inflation and resource utilisation in different directions, there is a more complicated trade-off.  Market agents and analysts wishing to predict how the Riksbank will act on new information at the next monetary policy meeting therefore have to analyse how this new information will affect the forecast for inflation and resource utilisation,” continued Mr Svensson.

”The forecasts for the interest rate, inflation and resource utilisation in the main scenario of the February Monetary Policy Report described what the Riksbank then considered to be a well-balanced monetary policy given the information available at that time. The information received between the February and June reports implied that wage increases would be higher, productivity lower and fiscal policy more expansionary than the Riksbank had assumed in February. With an unchanged interest rate path this information pushed up the forecasts for inflation and output, and our assessment was that there was a risk that inflation and resource utilisation would be too high. We therefore needed to adjust the interest rate path upwards, which then pushed back the forecasts for inflation and output, so that they showed a better development for inflation and resource utilisation. The interest rate path that we Executive Board members agreed on prior to the June report was finally higher than the alternative high-wage scenario published in the February report. This is consistent with the new information also including lower productivity and more expansionary fiscal policy, both of which increase inflationary pressures,” said Mr Svensson.

”New data have been received since the June report was published. I am not particularly interested in interpreting week by week how this information may possibly change the forecasts for inflation and resource utilisation and possibly justify changing the interest rate path, and I do not consider this to be a particularly efficient way of working. Instead I prefer to gain an overall grasp of the new situation prior to the next monetary policy meeting in September. With regard to the unrest that has arisen in the financial markets, I can say that we are closely monitoring developments. Our continuous monitoring of the financial markets means that we are well-informed. The Swedish money market and the financial system are functioning well,” said Mr Svensson.

”Let me finally say a few more words about resource utilisation. As is obvious from what I have said today, I see resource utilisation as an important variable in monetary policy. It can be measured in several different ways.. Research on monetary policy indicates that the gap between actual output and ’potential’ output, the so-called output gap, is an appropriate measure of resource utilisation in monetary policy. But there is considerable uncertainty, both theoretically and empirically, regarding how potential output should be estimated and forecasted. Consequently, the measures of potential output used by central banks are fairly uncertain. There are thus strong reasons for the Riksbank and other central banks to do more research in this area,” concluded Mr Svensson.