How can the Riksdag’s Finance Committee and others assess whether or not the Riksbank fulfills its price-stability objective? Especially since this is made difficult since the Riksbank does not have perfect control over the inflation rate and the inflation rate cannot always be exactly on the target. But it can indeed be done, by comparing average inflation over a longer period with the inflation target, and by also comparing with target fulfilment in other countries with inflation targets to judge what is possible.
The new Sveriges Riksbank Act took effect on January 1, 1999. The Riksbank, which is an authority under the Riksdag (the Swedish parliament), got a unique independent position in relation to the Riksdag and the Government. As discussed in the Government Bill 1997/98:40 (in Swedish only) that proposed the Riksbank Act to the Riksdag, a delegation of monetary policy to an independent Riksbank is consistent with democracy only under three conditions:
- The Riksbank must have an objective that has democratic legitimacy.
- The objective must be clear. If the objective is not clear, monetary policy is left at the discretion of the Riksbank and may be arbitrary and capricious.
- Such an extensive delegation of monetary policy requires that the Riksbank can be held accountable for its monetary policy. As the principal of the Riksbank, the Riksdag must be able to evaluate the Riksbank’s policy with regard to its stated objective.
Since the new Riksbank act took effect, the Finance Committee of the Riksdag has every year done an evaluation of the Riksbank’s monetary policy. The Committee also holds open hearings with representatives of the Riksbank.
The Riksbank Act states that “the objective of the Riksbank’s activities shall be to maintain prices stability.” The Riksbank has specified the price-stability objective as an inflation target, according to which the annual increase in the Consumer Price Index (CPI) shall be 2 percent.
How can then the Finance Committee and other observers assess whether the Riksbank fulfills or neglects the price-stability objective.
Assessing the target fulfillment is made more difficult since the Riksbank does not have complete control over the inflation rate. Inflation responds to monetary-policy actions with a lag of several quarters. It is also affected by different disturbances. Therefore, it is not possible to avoid inflation fluctuations at a 1- or 2-year horizon. It is not possible to always keep the inflation rate at 2 percent. The best the Riksbank can achieve is to have inflation fluctuating around 2 percent and be 2 percent on average.
- This means that average inflation over a longer period, say 5 years or more, can be used as a criterion for whether the Riksbank has fulfilled the price-stability objective or not. If average inflation over a longer period clearly over- or undershoots the inflation target, one may conclude that the Riksbank is not fulfilling the inflation target and guilty of prejudice to the price-stability objective.
A more technical advantage of looking at average CPI inflation over a longer period arises since the CPI includes a housing-cost component that varies with mortgage rates. These are in turn influenced by the Riksbank’s policy rate (the repo rate). Through the housing-cost component, changes in the Riksbank’s policy rate then cause short-term fluctuations in the CPI inflation rate that make a comparison with the inflation target less informative. Using average inflation over a longer period cancel these short-term fluctuations. What is left is only any trend due to a possible trends in the general level of interest rates, trends that mostly depend on a trends in real interest rates because of structural factors such as global imbalances, productivity, time preferences, and so on. Such trends do not prevent the Riksbank from keeping average inflation on target.
In figure 1, the grey line shows the CPI inflation rate in Sweden from 1995, when the inflation target took effect, and onwards (as always, click on the figure to enlarge it). (The inflation rate during 1995-2004 is real-time data according to the method Statistics Sweden then used to calculate the inflation rate. If the inflation rate is calculated according to the current method, as the percentage increase in the CPI, the inflation rate before 2005 would on average be 0.2 percentage point lower.) The blue line shows 5-year moving averages, that is, averages over the last 5 years. The red line shows the average inflation rate from 1995 onwards.
We see that the moving averages are clearly below 2 percent and that they never reach up to 2 percent. We also see that from 1997 and onwards for each year the average from 1995 to this year clearly falls below 2 percent. We also see that there is no trend in either the moving averages or the averages from 1995. They lie around 1.4 percent.
We may also note that from 2012 both the moving average and the average from 1995 fall. The moving average has not been as low as 1 percent since 2002.
Update, April 28, 2014: Updated figure 1 available here. Target achievement is now worse.
As a comparison to what is possible in other countries, we may look at CPI inflation in Canada. The Bank of Canada also has an inflation target of 2 percent (and a CPI with a similar housing-cost component as in Sweden). (For comparisons with more countries, see table 1, page 5, in this paper.) Figure 2 shows CPI inflation for Canada from 1995 onwards. We see that the moving averages are close to 2 percent and the averages from 1995 onwards are almost exactly on 2 percent after 2002.
- It is clear from figure 2 that Bank of Canada has fulfilled its price-stability objective with great success.
- It is also reasonable to conclude from figure 1 that the Riksbank has not fulfilled its price-stability objective but has systematically neglected it by letting the average inflation rate be too low. There is also no trend towards achieving the price-stability objective better over time.
When representatives for the Riksbank are confronted with the CPI inflation rate having deviated from the target, they usually emphasize that the inflation rate according to other price indices, such as the CPIF or the CPIX, are closer to the target. But if the target shall be clear and have any meaning, it is not acceptable to freely pick among indices to find one that in retrospect is closer to the target. For accountability it is not acceptable to shift between indices. And a clear target and accountability is required for the Riksbank’s independence to be consistent with democracy.
If the target is specified in terms of CPI inflation, it is CPI inflation that counts. And by looking at averages over a longer period one cancel , as was noted above, the direct effects on CPI inflation from short-term variations in mortgage rates and the Riksbank’s poliy rate – the effects that make comparisons of CPI inflation in the shorter term less informative.
The conclusion is that the Riksbank has neglected the price stability objective by allowing average inflation to fall clearly below the target. The consequences of this have been briefly discussed in my Vox columns here and here (and in previous Ekonomistas posts, English translations here and here). There is a great deal more to say about the consequences, and I will get back to that issue.