Some notes on the return of inflation

 

Last week Jim O'Neill, head of global economics at Goldman Sachs, made the point in a note to clients (and in a short summary article in the Financial Times) that the talk about the return of inflation was essentially nonsense. He particularly criticised commentators who focused on changes in relative prices. He also made the point that the rise in commodity price was a demand shock, not a supply shock as it used to be in the 1970s, and therefore irrelevant. Furthermore, inflationary expectations are well anchored, given that most central banks in the world pursue some form of direct inflation targeting.

 

We disagree with each one of his points. As far as the price level vs relative price debate is concerned, we believe that variances matter, not just for individuals but for the economy as a whole. To see this, look at the following simplified model of cost-push inflation. Society consists of various groups each facing a different inflation rate. There is, for example, research evidence from the UK, showing that both poorer people and older people face structurally higher inflation rates than the official rates. (And there is much anecdotal evidence for other groups, including bankers). Now assume that each group negotiates wages on the formula inflation plus some percentage of sectoral productivity. There is a high degree of asymmetry in the system in that above-average inflation sectors use their sectoral inflation rate as the basis, while below-average inflation sector use the central's banks rate as the basis. In other words, the central bank's inflation rate acts as a lower bound. This is quite a realistic setting for how wage bargaining occurs in the euro area, for example.

 

Under such a setting, it is quite easy to construct an aggregate problem for monetary policy. If a central bank cares only about the average inflation rate, it is possible for the economy to develop higher inflation expectations than the central bank's target. (This is what might have happened in the UK, where wage bargainers focus on a different inflation index than the central bank). If the variance of inflation rates between the different groups is small, this effect is small. But if it is not, the effect could be large. In other words, relative price differences matter, not just the price level.

 

O'Neill says that the phenomenon of differential inflation rates was important only for well-off consumers of modern services - such as financial services professionals who consume private education and prime-location housing. That is not true. Apart from old and poor people, young single people also have higher average rates (if they spend their money on restaurants and hotels), as do families (who also consume service, such as nannies, schools). In general, the higher the service content of consumption, the higher the rate of inflation. Inflation indices, by contrast, account for services only in an inadequate way.

 

O'Niell says the rise in commodity prices is not a fundamental problem, since it is due to higher global demand, rather than due to supply shocks as was the case in the 1970s. While demand shocks are more benign - global prices go up, but so does global capacity - they can still cause inflation if monetary policy reacts wrongly. We believe that the present, slightly elevated rates of inflation in the US and the UK are primarily due to this factor - central banks have persistently underestimated those shocks, and have run policies that were persistently too accommodating.

 

This leads to his final argument that inflation expectations are firmly anchored. It is true that they are more anchored than they were during the 1970s, but as we have seen in the UK and the US, it does not take a lot to lose the anchor. We also think that central banks will eventually do everything to contain inflation - which is why we do not see any long-run inflation problem either - we may well experience periods of price instability in the medium-term. One of the trends we have observed during this decade, in the US, in the UK and in the euro area, was that the central bank's inflation forecast proved persistently too optimistic. The ECB has used the information from its monetary analysis as a corrective to its own internal forecasts, while the Fed and the BoE have not. It is therefore not that difficult to conceive of a situation where inflationary expectations, once firmly anchored, may get unstuck. An example has been the recent comment by Fed governor Frederick Mishkin, who said it might too costly, in terms of economic growth, to get inflationary expectations back into the 1-2% comfort zone, the Fed's unofficial target range for core inflation. US headline inflation has been outside this range for the last three years, and this example shows clearly how easy it is to get off-track, how costly it may be to get back into the range, and how much resistance one would encounter if one tried. If the Fed had been serious about its inflation target, it would have raised interest rates to 6%, and higher. In the end, we trust that the Bank of England to get inflation back into its range, but we also believe that this comes at a cost. In other words: if O'Neill is right about inflation, then only because central banks do not care about short-term growth. That is true of some central banks, but not of others. On average, we would therefore expect global inflation to rise.

Eurointelligence ECB Watch

---