There are ways to reconcile the tension between creating employment and forestalling price increases

This month's rise in base rates by one quarter of a per cent to 6 per cent led to the first real cries of protest from industry for many years. The remarks of Eddie George, governor of the Bank of England, this week about the "ultimately unsustainable growth of domestic demand" suggest that they are likely to rise further. In that case business protests will get louder, especially if the pound resumes its rise.

Today's Bank of England inflation report will undoubtedly enlarge on the governor's thinking. But I would be astonished if it resolved the fundamental dilemmas of monetary policy based on inflation targets.

Let me first put them in fairly popular form. The present British government, like all governments, is highly sensitive to job creation and destruction. It is happy to boast of how many jobs have been created since it came to office. It is also acutely aware of corporate decisions involving likely job losses, although it does not intervene as crudely as previous Labour governments.

Yet at the same time the Bank of England, like other central banks, employs a model in which inflationary pressure is ultimately closely tied to the jobs market.

When unemployment drops or vacancies rise, outside commentators expect that interest rates are more likely to rise. And they are right.

There is a theoretical way to resolve this contradiction between the government's employment objectives and the Bank's economic model. This is to say that, at any one time, there is a rate of unemployment below which inflation tends to accelerate. The same notion can be expressed, with less political risk, in terms of the closely related "capacity gap". But the labour market constraint is the more fundamental.

On this view, the aim of the more sensible job creation policies is to reduce the underlying level of unemployment at which inflation begins to take off. Meanwhile, central banks, working with a shorter time horizon, have to accept the structure of the economy as their equations suggest that it is, and wait until the evidence of better labour market performance is undeniable.

This reconciliation may be satisfactory at undergraduate level - at least in some universities. But it does not really work as a basis for monetary policy. For, in practice, central banks have only a vague idea of the underlying growth rate of the economy. Nor do they know the unemployment level and market conditions at which wage pressures build up - as they appear to have done from tentative end-1999 estimates - and employers are able to pass them on.

There thus seem to be two alternatives. One is to proceed as the US Federal Reserve does: allow pressures to build up in the labour markets and take decisive action only when there are unmistakeable signs of inflation gaining momentum. The danger here is of waiting too long and failing to forestall dangers in time. This can aggravate the boom-bust cycle and thus destabilise the economy.

The other alternative, which is nearer to that used by the Bank of England's monetary policy committee, is for the bank to make its best stab at analysing the structure of the economy and take whatever action it thinks necessary to stop inflation exceeding the 21/2 per cent target over a horizon of two years.

The danger here is embodied by the cliché, "not giving growth a chance". This is especially great when the structure of the economy is believed to be changing rapidly, due to more competitive labour markets, globalisation, information technology and all the rest.

If the only effect of errors is to aggravate the business cycle slightly, or risk inflation falling below target, we can live with it. But if short-term restrictive action slows growth and hits longer-term employment, we are paying an excessively high price for the forecasting exercise.

There is no magic solution. But I dislike ending articles by saying that "the choice will not be easy" or "it will be interesting to see what happens". I prefer to make some tentative policy suggestions.

I would move nearer the Fed practice and go by actual evidence of inflation. But this evidence should be treated in a commonsense way and not crudely restricted to what appears in the statistics on the day that the MPC meets. Over a horizon of, say, a year, inflation is largely determined by a mixture of past inflation and by pressures already apparent in the pipeline from exchange rate movements, import and raw material price changes and the like.

To these can be added corporate and anecdotal evidence about the state of competition and whether it is more or less easy to pass on cost increases in prices. Account should also be taken of suppressed inflation. This is often shown not in a trade or current account deficit on its own - which may well be supportable - but in a rapid increase of this deficit.

More fundamental forces, arising from either monetary policy changes or their impact on the labour market, become much more evident towards the end of the second year of the MPC's self-chosen target horizon. The assessment of these depends rather more on models of the economy.

But this is not quite all. One element of suppressed inflation, which central banks realise they do not take sufficiently into account, is when inflationary forces are diverted into asset prices, such as equities or houses. John Flemming, a former economic director of the Bank, has suggested giving asset prices a weight of a fifth to a third in an adjusted inflation index. He also suggests measuring the equity element by "the reciprocal of the dividend or earnings yield on a portfolio of pension fund assets". (Economic Outlook, October 1999, Blackwell, Oxford).

The above suggestions will not solve all problems, but they could be an improvement on the present excessive reliance on more remote crystal-ball gazing.