Tim Knox: Bank betrays bias in its gambles on the future

For the most part, these forecasts add some jollity to everyday life. They are essentially trivial, in the sense that – unless you actually win the Lottery, or are a compulsive gambler – the outcome is relatively unimportant.

Even so, we all carefully check our lottery numbers or our betting slip to see whether we have won a few pounds.

So, logically, you might expect that the more important a forecast was, then the more we would try to work out how accurate the forecaster was. Logical, but wrong. For in one of the most important areas of economic forecasting – that of the inflation rate – that analysis has only just now been completed.

And the results are worrying. As Ryan Bourne shows in a recent Centre for Policy Studies paper, the Bank of England recently had a brilliant record at forecasting inflation. For example, between 2001 and 2004, the forecasts were almost spot on.

In the 12 quarterly inflation reports from August 2001 to May 2004, the Bank’s average forecast for inflation a year ahead was 2.2 per cent per annum while the eventual result was 2.3 per cent. Most impressive.

Things got a bit worse in the next three years. In the 12 quarterly inflation reports from August 2004 to May 2007, the Bank’s average forecast for inflation a year ahead was 1.9 per cent per annum, with actual inflation of 2.3 per cent. The average error was, therefore, +0.4 percentage points – less accurate but still well within normal forecasting margins.

Since then, however, things have tobogganed downhill. In the 12 quarterly inflation reports from August 2007 to May 2010, the Bank’s average forecast for inflation a year ahead was 1.9 per cent. Actual inflation was 3.2 per cent each year.

This represents a much larger average error of 1.3 percentage points. Considering that inflation was only meant to be 2.0 per cent, this is a huge error. Indeed, if you wanted to know the inflation rate, it turns out that you would probably have recently got a better picture from Mystic Meg than from the Bank of England.

And it is not just the fact that the forecasts have been wrong that matters. What is most striking is the consistency of the direction of bias. The Bank keeps telling us that underlying deflationary forces or excess capacity will curb inflation in the medium-term. But they haven’t.

Why are we to trust their forecasts now when they have been so wrong over the past three years?

And don’t forget that the ultra-low interest rates resulting from the Bank’s forecasts are not politically neutral.

Yes, the CBI and business people have spoken about the damaging effects that interest rate hikes would have on investment and consumer confidence.

But it is also strongly in the interest of the Government to see a modest level of inflation to allow the real depreciation of the very high debt levels of the Government and the banks. The Government wins. The deeply-indebted banks win. All those of us with mortgages win.

But there are losers. Savers in particular, are seeing the real value of their savings erode while also receiving a negative real rate of interest. And it is surely wrong to see the thrifty and the honest take the hit for bailing out the spendthrift.

And the Bank of England is not the only forecasting body that tends to be over-optimistic about our macro economic outcomes. Another recent CPS paper has shown that HM Treasury forecasts of GDP growth three years ahead for the period 2000-2010 were also over-optimistic on average by 1.3 per cent.

Our economic masters seem to be living in La-La land, a world in which there is now an in-built over-optimism of the UK’s inflation and growth prospects.

George Osborne’s decision to transfer responsibility for growth forecasts to the Office of Budget Responsibility is welcome.

For, at best, these biases and errors show the difficulty of macro-economic forecasting, especially in light of unanticipated deep recessions.

At worst, however, they suggest that economic policy makers may be willing the ends without actually implementing the tough and probably unpopular policies needed to achieve such ends.

Better models or more honest forecasting – either way, there are improvements to be made.

Tim Knox is Acting Director of the Centre for Policy Studies. From spot on to inaccurate: The recent history of MPC inflation forecasts by Ryan Bourne can be downloaded from www.cps.org.uk