Inflation targets as focal points: Revising the ECB’s monetary framework
The toughest job central banks face in the next five years is managing uncertainty. In the euro area, inflation is persistently low and the ammunition available to raise it is minimal. Meanwhile, structural changes, including the rise of the digital economy and its effects on productivity, and the threat to global open trade, imply that we do not know how the economy will work. At the very least, the European Central Bank will have to design policy not just for specific circumstances but for as many circumstances as possible.
The first guiding principle in revising the ECB monetary policy framework should be to reduce the degree of uncertainty the bank itself brings to the system. In other words, the ECB policy framework should be clarified to minimize self-generated policy uncertainty, and the starting point here will be to have a clear focal point to anchor expectations of inflation.
We propose two changes to ensure that the ECB’s inflation target acts as such a focal point. For any inflation target to be an effective focal point, it requires two things: The target needs to be clear for all to understand and it needs to be implemented in a framework in which it can be assessed.
The current numerical objective communicated by the ECB is “below but close to 2 percent.” This is not as precise as it could be. The word “below” suggests that there is a downward bias in the definition, which might make current conditions more difficult to escape from. By contrast, the word “close” suggests that the ECB might actually be targeting implicitly a lower level of inflation, say 1.8 percent. And there would then be no downward bias, because the ECB might be targeting inflation symmetrically around that smaller number.
But it is not clear which one is true and there is no benefit from adding disagreement to the ECB’s definition of price stability. The first step to achieving a focal point would be to change the mandate to read “the inflation objective is 2 percent.”
However, this is not enough. A point number alone is irrelevant because inflation will seldom be at exactly 2 percent. What does that mean? Is the Central Bank simply seldom successful? What is needed is a band around the point number that would provide meaningful information in terms of what is tolerated. Too broad a band, and the target becomes meaningless again, as it means that any number is really tolerated. But too narrow a band (and at the limit a point number) and the signaling value of the inflation target also disappears as the target will seldom land within these narrow limits.
An inflation target becomes meaningful if there are appropriate bands around it. The monetary policy “game” then allows for agents to observe the inflation outcome and evaluate whether the Central Bank has been successful or not. A few successes and the Central Bank would become credible, meaning expectations are anchored to the numerical target. A few failures and expectations could become de-anchored. Importantly, credibility is itself the outcome of performance, and constrains or releases the Central Bank’s ability to achieve its goal.
What qualifies then as “appropriate band width”? This depends on the level of uncertainty in which central banks operate. If uncertainty is high, then inflation is less likely to land in tight bands. The Central Bank ends up being seldom successful. If uncertainty is low, broad bands risk blurring the signaling quality of the target. The Central Bank risks unnecessarily having agents not anchoring their expectations on the target.
Most central banks that have an explicit inflation target of 2 percent typically tolerate inflation between 1 and 3 percent. But if the ECB has to operate in high uncertainty, what might have been an acceptable tolerance band up till now may no longer be enough. With this in mind the ECB should define what it is prepared to accept. If the ECB is confident to be able to control inflation between 0 and 4 percent, then a tolerance band between 0.5 and 3.5 percent may be a more effective width to provide a meaningful signal but also acknowledge high levels of uncertainty.
Source: kathimerini.gr