Last week, Ben Bernanke, former chairman of the US Federal Reserve Bank and Nobel Laureate in economics, published his review into the process of forecasting and policymaking at the Bank of England.
You really need to be a linguist to have a full appreciation of the report. It is a matter of translating the formal, polite language of official bureaucracy into the more colourful form of English in everyday use.
At one level, the report is careful, hedged and qualified in what it says. But in the vernacular, Bernanke puts the boot into the Bank in no uncertain terms. The kicking is a severe one.
A great deal of criticism has already been levelled at the Bank for its failures to forecast inflation over the past few years.
Bernanke points out that this is in fact a feature, not a bug in the main model used by the Bank to predict inflation, a model which he states has “serious shortcomings”.
He explains that the embedded expectations of inflation are “well anchored”. A seemingly innocuous jargon phrase.
But it actually means that the Bank assumes that in the longer run inflation forecasts are fixed at two per cent. And if expectations of inflation are fixed at a low level, any deviations of actual inflation above this, no matter how large, must be only temporary.
In the type of model used by the Bank, inflation in the current period depends upon the current level of output and the expectation of inflation in the future.
The expectation of inflation can be above two per cent and if so, inflation will rise in the next period. But the “anchor” level which is assumed by the Bank exerts constant pressure to pull it back down.
The technical model which underpins the Bank’s forecasting errors leads into a second major criticism of the Bank in the Bernanke report.
He states that “the most serious problems we found are the deficiencies of the Bank’s forecasting infrastructure – the tools the staff uses to produce the quarterly forecast and supporting analyses”.
Key software is out of date and lacks functionality. This in turn leads to the fact that the Bank’s inflation models are “not adequately maintained”.
By this, Bernanke means that the models are not updated sufficiently. When new economic data appears – which it does all the time – the model should incorporate it on a regular basis and be revised if necessary. This is particularly the case when large forecasts errors have been made.
This is not rocket science. It does not even involve wrestling with the intricacies of the latest economic theory or statistical technique. It is a matter of basic competence.
A third main area of criticism is that the forecasting approach should include “rich and institutionally realistic representations of the monetary transmission mechanism, allowing for alternative channels of transmission”.
Translated, this simply means that the Bank pays little or no attention to the money supply. The sharp rises and falls in inflation over the past few years have been predicted much more accurately by monetarist economists, a success ignored by the Bank.
Bernanke goes on to note other areas where the Bank’s models lack sufficient coverage. Supply-side factors are largely absent and, rather surprisingly for a central bank, there does not appear to be a detailed model of the financial sector.
Earlier in my career I spent a good few years as an economic forecaster. So I have every sympathy with those that are involved in this challenging task. But even so, it is hard to describe the Bernanke report as anything other than a devastating criticism of the Bank and its forecasting procedures.
