Inflation measures and the ECB’s mandate

Regarding the not-so-subtle innuendo in one of my last blogs, I will write a little bit about how the ECB measures inflation and why this may be insufficient.

Inflation is the term economists use to describe an overall increase of prices. Of course, some prices rise more than other, while others fall. Statisticians came up with a trick and define a “basket” of goods that are frequently used. Then inflation is the change of the weighted average price of said basket. Normally, these baskets are updated every five years with new products to replace obsolete ones (think Blu-Rays instead of DVDs). There are other measures of inflation, that have certain advantages and disadvantages over a goods basket. But since the ECB measures inflation by means of a goods basket, let us stick to that measure.

The relevant basket for the ECB is the Harmonised Index of Consumer Prices (HICP). Its mandate is price stability (Article 127(1) of the Treaty of the Functioning of the EU). This is defined as:

Price stability is defined as a year-on-year increase in the Harmonised Index of Consumer Prices (HICP) for the euro area of below 2%. Price stability is to be maintained over the medium term.” Today, the Governing Council confirmed this definition (which it announced in 1998). At the same time, the Governing Council agreed that in the pursuit of price stability it will aim to maintain inflation rates close to 2% over the medium term. This clarification underlines the ECB’s commitment to provide a sufficient safety margin to guard against the risks of deflation. It also addresses the issue of the possible presence of a measurement bias in the HICP and the implications of inflation differentials within the euro area.

(, own emphasis)

Good. Low inflation, well anchored and over the medium term, i.e. no panic if the inflation rate goes a bit off path in the short run. However, in the wake of the crisis it becomes clear, that this is just not good enough.

First of all there are problems with the HICP, not measuring housing adequately. But that is not the main point. Kantoos had a very good article some years ago. (Damnit Kantoos, why did you close your blog?)

The ECB has an average inflation target for the whole, heterogeneous Euro area. As I argued before, it is hard (impossible) to conduct sensible monetary policy for all countries of the Euro area.

The foremost reason for the ECB’s ultra loose monetary policy is the attempt to keep the Euro area together – while noble, I’m not perfectly convinced that a) this is covered/demanded by its mandate and b) this does not violate the notion of price stability demanded by the Treaty. Remember that is the Treaty that is important; the ECB’s interpretation of what exactly price stability is, is just a decision by the governing council which can be changed anytime.

My point is rather statistical: The ECB’s inflation target is a triple average: over the goods in the HICP, over time and over the members of the Euro area. Every mathematically inclined person knows that an average has little information without knowing its variance. (Read something about it here. Sorry, German only.)

Since the inflation rate is a triple average there are also three variances. The most obvious swing can be seen in a time series of the inflation rate. There is a huge increase in 2008 followed by a huge decrease in 2009. Both anomalies basically cancel each other out so that there is little left of them in the medium-term average. However the swings are severe and increase the variance.

Red (left-hand axis) is the inflation rate. Blue (right-hand axis) YoY change in per cent. Source: ECB
Red (left-hand axis) is the inflation rate.
Blue (right-hand axis) period-on-period change in per cent.
Source: ECB

Then there is the average over the goods in the HICP basket. Here swings are not that bad, since it reflects healthy markets when prices react in both directions to changing circumstances. Moreover, it is not the central bank’s job to micro-manage goods prices but rather to make sure prices can adjust.

Finally, and this is the most severe criticism is the measuring the inflation rate as an average over the whole Euro area. Yes, the ECB has rightly the European point of view because it conducts the monetary policy not for a bunch of countries but for all people using the Euro, regardless of where they live. This is a chart of the HICP inflation rates in Germany (red) and Spain (blue).

HICP inflation rates (monthly, annualised) for Germany (red) and Spain (blue). Source: ECB
HICP inflation rates (monthly, annualised) for Germany (red) and Spain (blue).
Source: ECB

As you can see, average inflation in Spain was almost always way above the magical 2 per cent (inflationary bubble, anyone?). Meanwhile in Germany, inflation was almost always below the magical 2 per cent (sub par growth, anyone?).

The problem is now that the ECB claims to have achieved their target. Well, 2.06 per cent of average inflation since 1999 is pretty close to the target. But still above 2 per cent. Anyway, details.
More troubling is that despite the “accomplished goal” of the ECB there has been an inflationary bubble in Spain. Hardly accomplished, if you ask me.

The proposal would now be to change the ECB’s notion of price stability. Abandoning the union-wide average and replacing this by member-wise averages would be the first step. Price stability would then only be achieved if all the members fulfill the goal.
A second step would be to implement bands within the inflation rate may vary. Another approach would be to limit the rolling variance to a certain threshold. After all, the variance itself is only an average – that of quadratic deviations from the average. So limiting the variance of the, say, last 5 years to a certain threshold would be considered price stability.

The important and far too often neglected aspect of price stability is that also the variance has to be stable.

What do you think? Is this feasible? Is there need for change in the first place? I’m looking forward to your comments (Also in German.)

Published by


I write about economics and politics. I take Ordnungspolitik seriously. While not blogging, I study monetary unions for my doctorate.

2 thoughts on “Inflation measures and the ECB’s mandate”

  1. “The proposal would now be to change the ECB’s notion of price stability. Abandoning the union-wide average and replacing this by member-wise averages would be the first step. Price stability would then only be achieved if all the members fulfill the goal.”

    The question is in how far it’s possible to achieve price stability under these circumstances. If the notion of price stability is made very strict, it may prove to be impossible to achive the goal with the current set of Euro members.
    Furthermore the question remains, how to adjust the monetary policy, if some members face deflationary pressure, while others are above the 2 % inflation target – a completely fictitious example of course. Obviously this already is a current problem of the ECB. I certainly hope, that the ECB already looks at the inflation figures for single countries, even if they are not doing it in public (at least they should start doing it now, after what has happened in some member countries).

    1. Tobi,

      you are perfectly right to point out, that this would make price stability as we know it now highly unlikely or impossible.
      Not only because of the heterogeneity of the EMU but also because of non-aligned business cycles (granted, this is one source of heterogeneity).
      I would make the case for a certain band fo both inflation and inflation variance. Within these bands any fluctuation is possible. This would on the one hand allow for monetary flexibility without worrying about price stability that much, but on the other hand also takes the heterogeneity into account. Basically, we need the EMU to adopt a notion of price stability that does not depend on some average but on all the parts. This is necessary because we don’t have labour mobility within the EMU to counter the effects of changes in the real exchange rate.

      Your example is not so far from the actual EMU problems. The core has inflationary tendencies, especially in housing, while the periphery suffers from deflation to recover from the inflationary boom of the 2000s. There is no common monetary policy to cover for this, because it cannot be expansionary (periphery) and contractionary (core) at the same time. Besides, the problem of the periphery is a real one. Either productivity must come up or prices/wages must come down to realign the real exchange rate with the core. Monetary policy can only bring you so far.

      On a wider perspective, it was precisely this average notion of price stability that made the mess possible in the first place. The core had below average inflation and the periphery above average inflation. The ECB should have stepped in by providing less liquidity to periphery banks and more to core banks. Hence, we also have to think about the mechanisms of liquidity provision by the ECB.

      By the way, the numbers are there. They are just not used short of computing the average.

Comments are closed.