Output gap lets in growth without inflation

When the Chancellor and Bank Governor met last week to discuss the possibility of a further base rate cut, it is a fair bet that the concept of the 'output gap' was a determining factor in their decision. Output 'gapology' is on the rampage. In fact, 'mind the gap' is a phrase heard almost as often in the Treasury nowadays as on the London Underground.

Unfortunately, it is a devilish task to derive an empirical estimate of the output gap in practice, and this leaves scope for the usual wide range of disagreement among economists about the appropriate setting for monetary policy in the year ahead. Despite this, the output gap - or one of its close cousins - is so fundamental for policy at present that it demands discussion.

The output gap is simply the difference between the actual output of the economy and the level of output at 'normal' capacity working. Normal capacity is reached roughly in the middle of the economic cycle, halfway between boom and slump. Most economists would agree that when the output gap is zero, inflation should be stable. Some, like Tim Congdon, believe that inflation will continue to decline for as long as the output gap is negative (which he believes it is at the moment), and that this is the end of the story.

Others, like myself, believe that the level of the output gap is indeed an important determinant of inflation, but that the change in the gap also contributes to the picture. Therefore, if the output gap is negative but declining (ie the level of output is below normal capacity, but the growth in output is above average), there will be two offsetting forces at work, and inflation might either rise or fall.

But obviously this disagreement is pretty irrelevant if the output gap cannot be measured in the first place. The estimation problem hinges on the definition of 'normal' capacity working in the economy. Some economists gauge capacity simply by reference to the availability of idle machinery in manufacturing, and on this basis the picture is not encouraging.

According to the latest CBI survey, the proportion of firms working below capacity fell to 57 per cent, which is slightly below the historical average for this series. At a similar stage in the last cycle, about 77 per cent of firms said they were still working below capacity.

WILLING TO INVEST

However, firms can of course add to their capacity by increasing fixed investment so that potential problems disappear. According to the CBI survey, firms are showing a definite willingness to invest in new capacity - 25 per cent of firms investing say they are doing so to raise capacity at present, against around 12 per cent at the same stage in the last cycle.

Furthermore, the CBI survey reports that about half of all firms take employment levels into account when assessing whether they are working at full capacity. One of the features of the recent recession was that firms laid off workers at an earlier stage in the downturn than normal, and they were therefore 'hoarding' less excess labour than usual as the economy began to recover. This may cause higher readings for 'capacity utilisation' by individual firms, but the economy as a whole would not face constraints until the supply of unemployed skilled workers ran dry.

These points on capacity are relevant to an assessment of the output gap. Those who believe the output gap is small tend to use methods for estimating trend GDP that incorporate the CBI measures of industrial capacity (or use the CSO coincident indicator, which in turn is partly based on the CBI capacity series). The dotted line ACDE in the graph is derived by one such method. This accepts the CSO assertion that the economy is working 'on trend' whenever the coincident indicator stands at 100, and simply joins up these points to produce a series for trend GDP.

Based on this method, the economy is estimated - amazingly enough - to have returned to trend working late last year, when the CSO coincident indicator returned to 100. If true, this would indicate the output gap is now zero. On this basis, the Chancellor should be considering an imminent rise in interest rates.

REASONS FOR SCEPTICISM

However, there are reasons for being very sceptical about this method. First, the CSO coincident indicator is likely to be heavily revised if the economy now embarks on a significant period of rapid growth. This is what happened in 1980-83. The CSO initially estimated that the economy had returned to trend working as early as the first quarter of 1982 - only 12 months after the deepest recession since the 1930s had ended. At the time, trend GDP was shown by the line AB on the graph. However, as activity began to expand, the CSO significantly increased its estimate of trend, and the latest estimate is shown by the line AC.

Furthermore - and this is the key point - even after the economy returned to trend in 1983, there followed about four years of rapid GDP growth before inflation pressures started to rise. During this period, the CSO's estimate of trend GDP also grew very rapidly, so that the output gap did not turn positive. A similar effect may occur from 1994 to 1998 if the economy grows at 3 per cent per annum.

In economic terms, the GDP trend shown by the dotted line ACDE can perhaps be interpreted as being determined by short-run plant capacity, which can be influenced by plant closures, scrapping and new investment. As the graph shows, this trend has actually fallen in the past few years, suggesting that the capacity of the economy has been shrinking. If this has really happened, it can only have been because of plant closures and scrapped machinery. But in the longer term, these losses can be made good by extra investment.

In fact, over very long-term horizons, stretching through much of the past century, the economy appears to attain an extremely steady trend growth rate of about 2-2.25 per cent a year. This trend does not seem to be much affected by short- term events, like recessions or booms, but seems to be determined by much more fundamental forces, presumably focusing on labour supply and technical progress.

The UK's long-term trend growth rate of around 2.25 per cent per annum has been used to construct the 'main trend' for GDP growth shown by dashed line AD in the graph. If we extrapolate this trend to the present (the point D at the end of 1990 is selected as a plausible quarter in which GDP was last at trend), we find the economy is nowhere near its ultimate capacity. In fact, there is still a huge output gap of the order of 5.7 per cent, and plenty of scope for inflation- free growth in years ahead.

If the experience of the 1980s is repeated, inflation will stop falling when the economy's short-term trend is reached (ie some time soon), but will not start to rise until the ultimate trend is exceeded, which may not be for a few years. In this period of steady inflation, monetary policy should have a 'steady as she goes' bias, avoiding upheavals wherever possible. This peaceful interlude could be quite a challenge for the UK authorities, which have become accustomed to dramatic lurches in both directions.

(Graph omitted)