Why the Bank is probably right on rates again

Press reports have suggested that officials of both the Treasury and the Bank of England recommended that the Chancellor should have raised interest rates last Wednesday, but once again Ken Clarke turned down this advice. Despite the fact that Christmas retail sales were only so-so, and that recent inflation data have been encouraging, at least in the manufacturing sector, it seems to me that the balance of the argument once again favours the politically impartial judgement of Bank and Treasury officials, rather than that of the Chancellor. Nevertheless, the case is by no means open and shut. There are three main areas for debate.

The first is the extent of labour market tightening in the economy at present. The official unemployment statistics which were published last week for December would certainly seem to leave no room for debate.

The decline of 45,000 in the claimant total (a definition which includes only those unemployed people actually claiming benefit), following the record drop of 95,000 the previous month, means that the official figures have never before fallen as fast as they did in the final quarter of 1996.

Yet there are definite reasons for believing that these figures are distorted. According to the Department for Employment, the replacement of unemployment benefit by the jobseekers' allowance has reduced the claimant total by a maximum of 35,000 in the past two months, which explains about one quarter of the decline.

Furthermore, the alternative important source of information about the labour market - the Labour Force Survey (LFS), which is based on sample surveys of the population at large - indicates that a significant proportion of the rise in unemployment between last September and November might have been due to a toughening in benefit procedures.

Between these months, the LFS shows that claimant unemployment fell by 117,000, while non-claimant unemployment actually rose by 85,000, a combination which suggests that a large number of people have been pushed off benefit and into the "lump" of invisible unemployed. If this is indeed the case, then the labour market is not tightening by anything like as much as the decline in the claimant unemployment total suggests.

Even allowing for this factor, however, there is plenty to worry about. As the graph shows, the ratio of unfilled job vacancies to short-term unemployed, which is one good measure of labour market slack, has risen to record levels, both on the LFS and on the claimant measure of unemployment.

The vacancies total has risen by a remarkable 40 per cent during 1996, and in the past the current level of vacancies has always been consistent with GDP growth of well above 4 per cent, a figure which would almost certainly lead to a rise in inflation pressures if seen during 1997.

This brings us to the second area for debate, which is the accuracy of the GDP figures. This is clearly crucial for Mr Clarke, since every time the Chancellor takes a decision on base rates, he refers to recent GDP growth as one of the prime considerations in his mind.

On the latest published data it seems that GDP in 1996 has expanded by only 2.3 per cent. Taken at face value, this looks like a comfortable situation, in view of the fact that the underlying growth of capacity in the economy is probably around 2.5 per cent per annum, and that the level of output may still be about almost 3 per cent below trend.

But the conclusion would look very different if the government statisticians were under-recording the level and growth of GDP by a meaningful amount, as they have frequently done in the past.

Unfortunately, it seems quite likely that this will turn out to be the case, even though the Office for National Statistics (ONS) has been making strenuous efforts to correct the problem.

The graph (prepared by David Walton of Goldman Sachs) shows how important the mismeasurement of GDP could prove to be in assessing whether the economy is currently generating latent inflationary pressures.

After a recent study of their own past track record, ONS statisticians have reported that their initial estimates of the annual rate of GDP growth typically prove around 0.7 per cent too low during the upswing phase of the economic cycle. Using this result, Walton calculates that the correct level of output at present is 2 per cent higher than the official data indicate, which in turn is only fractionally below the normal capacity of the economy.

Furthermore, if we make the more pessimistic assumption that GDP is being under-recorded by the same amount as it was during the exact same phase in the 1980s economic upswing, then output is actually a little above capacity.

Although this may seem to be straining every sinew to see the dark side of life, it would certainly explain why the labour market is tightening so markedly at present. And since the growth of output looks certain to be above trend this year, the situation is likely to worsen before it improves. When Goldman Sachs has argued this before, some people have accused us of making up the GDP data to suit our purposes. We disagree.

If the ONS openly admits that it has a repeated tendency to underestimate GDP in the present phase of the cycle, then this is surely something which we should build into our central assessment of economic conditions today - or we should certainly do so if we have a risk-averse approach to the control of inflation. If policymakers had done this at similar junctures in the past, some of our most damaging macro-economic policy errors would have been avoided.

This leaves the third and most difficult area for debate, which is whether the rise in sterling obviates the need for higher interest rates at home. The Chancellor seems to think so, to judge from the vehement way he described the disinflationary effects of sterling's rise on Thursday.

In fact, listening to his emphasis on the exchange rate, it was hard to imagine another rise in base rates occurring before the election. He has a point here. If sterling stays where it is, then it will tighten overall monetary conditions enough to ensure that the inflation target is hit over the next 18 months, even with base rates no higher than the present 6 per cent.

In fact, when we weight together sterling and interest rates into a single monetary indicator (a procedure which the Bank of England hates, with some justification), we find that, thanks to sterling, monetary policy has already tightened by more in the current episode than it did in 1994/95, when base rates rose by a point and a half.

But the great question is why sterling should remain this high if base rates do not rise. Much of the exchange rate appreciation has been directly due to the anticipation of higher base rates, while a significant element is, quite frankly, difficult to explain in terms of standard or "fundamental" economic models.

Probably some of the rise will therefore prove "frothy" and unreliable, while much of the rest is dependent on the Chancellor eventually choosing to raise base rates.

If we arrive in the summer without a significant further base-rate rise, sterling will drop back and latent inflation pressures will be unleashed.