Stephen King: Bank should wait and see rather than take a risk

Strict adherence to the implications of surveys could lead to an increase in policy errors

Economists are fairly simple folk. They have a very straightforward understanding of cause and effect. They know, for example, that strong growth leads to higher inflation and, hence, higher interest rates. They also know that weak growth implies lower inflation and, therefore, should justify interest rate cuts. They also believe that, once an economy begins to strengthen, it's on its way towards an untamable boom. Likewise, they know that a weakening economy is, like it or not, probably heading straight into recession.

A lot of these arguments are based on the "straight lines" that can be drawn on the back of simple multiplier calculations. An initial boost to demand from, say, government spending, triggers a series of additional spending increases, ranging from higher consumption in the light of rising employment to firmer investment growth in the light of a cyclical expansion in profits. Inflation targeting is designed to limit some of the more extreme consequences of these "straight-line" effects. If the economy expands too quickly, the central bank can place its foot lightly on the monetary brake, in the hope that the economy can avoid the nasty effects of an eventual "hard landing".

If only life was this simple. Why, if this is all so straightforward, do policy makers have to scratch their heads all the time? Why is it that Sushil Wadhwani can leave the Monetary Policy Committee of the Bank of England apparently holding on to a rather different view of how the UK economy works compared with some of his more hawkish colleagues. Why is itthe European Central Bank is thinking of raising interest rates in a situation that would barely get Alan Greenspan out of his bath in the morning?

Some of these questions are, ultimately, rather philosophical in nature: I dealt with some of the answers in this column two weeks ago. On top of these philosophical difficulties, however, there are also practical challenges. As a policy maker, how can you be sure that your economy is sticking to one "straight line" and not veering off in all manner of unpredictable directions?

Ultimately, of course, this is a question about the predictability of economic performance. Financial markets tend to get carried away at the first sign of surprise either in an upward or downward direction. The reality, however, is often a lot more surprising. For example, supposedly one of the best "early warning signals" for any economy is the news coming from business surveys. Whether it be the CBI survey in the UK, the ISM survey in the US or the IFO survey in Germany, business surveys are regarded as useful and reliable lead indicators of future economic health. Indeed, there appears to be a fairly consistent relationship between the results of these surveys and overall economic activity, as measured by GDP (see charts).

Put another way, whenever the results of these surveys rise to certain levels, they give the impression that economies are well on the road to a sustained recovery, leading to the "straight line" argument in favour of faster growth, higher inflationary pressures and, eventually, higher interest rates. The evidence, however, suggests that this kind of knee-jerk response is a long way from the truth. Surveys may be good enough to suggest that things aren't going to get any worse – and, even here, there have been some spectacular failures – but they say very little about the pace of recovery and the dangers of inflationary excess.

The problem is one of sustainability. Business surveys give a good impression of what is happening now. They are not so helpful, however, in telling us what will happen a year or two down the road. Put another way, they are quite useful at telling us whether economic policy was about right a year or so ago but they give us few clues as to whether current interest rates are about right, too high or too low.

Consider, for example, the relationship between the various business surveys and GDP growth. In the US, for example, the ISM survey has particular resonance at certain key levels. When it rises above the 40 mark, it's typically consistent with US GDP growth shifting from negative into positive territory. When it rises above the 50 mark, it's typically consistent with a shift in the pace of economic growth from below to above the long-term sustainable rate of expansion.

Yet, over the past 30 years, these simplistic relationships have given only a very imprecise guide. In the following two years after a breach of one of these key levels, the economic outcomes have been rather random. Sometimes, there has been a very powerful expansion. Sometimes, the rate of expansion has been rather weak. On one occasion – at the beginning of the 1980s – the surveys marked the high watermark of economic success and the US economy promptly went back into recession.

Given this range of outcomes, it's hardly surprising that inflation can head off in all directions and that, on a number of occasions, the Federal Reserve has chosen to either leave interest rates unchanged for a prolonged period of time or, alternatively, has chosen to cut rates despite evidence of renewed expansion (the early 1990s is an interesting example of this apparently counter-intuitive approach).

The same observations apply for the UK and Germany. Like the US, the CBI and IFO surveys are quite good at telling us what's happening to the UK and German economies today. Like the US, however, these surveys give a rather ambiguous guide to what's likely to be happening in a few months' time. Of course, this creates a problem for policy makers who need to know the future effects of current policies, not the current effects of past policies.

Indeed, strict adherence to the implications stemming from surveys could lead to an increase in policy errors. An increase in interest rates today on the back of a recovery in current survey conditions could be a short-lived affair. If policy makers subsequently discover that the pick-up in growth was not sustainable, they could then be forced into a rather embarrassing policy reversal, clouding the whole process of monetary decision making.

On this basis, a "wait and see" approach might be more appropriate. This might not sit very easily with the "pre-emptive" requirements of a modern central bank but, with inflation relative low, the risk-reward ratio would appear to favour caution. Ultimately, survey data are only one piece of a very elaborate economic jigsaw puzzle. Surveys may be recovering but there are other reasons for concern: the ongoing weakness of equities, for example, suggests something about the risk to economic expansion that is, potentially rather different to the message coming from the survey data. In other words, even if interest rates do rise this year, there are no guarantees that rates will continue to rise during 2003. If history tells us anything, it's that economies do not travel in the straight lines so loved by the economics community.

Stephen King is managing director of economics at HSBC.