If the Bank’s forecasts are wrong, maybe its decisions are too?
If the Bank of England admits it got its old forecasts wrong, why should we think its new predictions on the economy are right? Perhaps the Bank just isn’t very good at guessing the future?
The Bank has got it wrong on the recession, house prices, jobs, inflation and growth. Just three months after predicting 3.4 per cent growth for 2011 it now says the economy will expand by 2.7 per cent. The 2012 figure comes down from 3.5 to 3.1 per cent too and with 2010 more than half gone, the short-term forecast for this year has been trimmed too.
But why should we think the new figures are any more accurate? They are still well ahead of what other forecasters are looking for. The suspicion must be that the Bank is using the wrong forecasting model.
Indeed, the Bank admits as much: its annual accounts show it has spent £1.1m reviewing its model and is spending another £2.4m on a new one.
Given that the current model dates only from 2004, that’s worrying. If that computer programme got it right for a few years, that might simply be luck – or more likely, be because it was designed to forecast the recent past rather than the future.
It is easy for any model trying to forecast a slowly moving trend to get it right in the short, even medium, term. If it is a bit below the curve at first, a bit over later and perhaps then a bit below again, the discrepancies are small and average down. Imagine drawing a straight line through the edge of a large circle – a chord close to the tangent, to put it mathematically. Just before it cuts the circumference, while it passes through the circle, and for a time after it leaves again, it gives a close approximation to the curve – but either side of that it is increasingly wrong.
Over the past four years, the Bank’s forecast of inflation has been wrong in almost every month. Given that the Bank’s sole role is to keep inflation on target, that is worrying. On seven occasions it has been so wrong – more than 1 percentage point away from the 2 per cent target – that the governor has had to write to the chancellor explaining why. He has written so often that the letters have become junk-mail.
Macro economics is the governor’s specialist subject, yet having failed on that, he is being given supervision of the banking system too when the FSA regulator is broken up. And the Bank’s track record there is not perfect either: as well as BCCI, Barings and Johnson Matthey failing when the Bank was last the supervisor, it was slow to act to keep Northern Rock afloat.
Maybe we should simply accept that the country’s central bank is not perfect. Forecasting is a means to its end, not its purpose, and the new Office of Budget Responsibility, as well as the Treasury and a myriad of private forecasters produce predictions that the Bank can read as well as its own. But if it has got its assumptions wrong on its forecasts, maybe it has also got its decisions wrong on setting interest rates?
