A former British Treasury Minister, Harold Lever, once remarked that bankers should be satisfied that they are doing an important job; and they are not badly paid for it either. They should not expect to be loved as well.

The best that central banks can do to put themselves in a good light is to implement successful policies. The main thing I want to say about presentation is to echo the words of a popular British press lord who said: 'Publish and be damned' (Hugh Cudlipp). There is now criticism of the European Central Bank for not publishing its inflation forecast. If these forecasts were taken out of the cupboard, we would see that they were not very different from the average of other forecasts and interest in them would die down. The real misfortune is not that such forecasts are as yet undisclosed, but that they play such a large part in policy.

The chief economist of Goldman Sachs, Gavyn Davies, has recently reminded us that the Federal Reserve, the European Central Bank and the Bank of Japan now cover about 80 per cent of the developed world economy. "Governments have largely ceded to these three institutions the responsibility of controlling world inflation; and to do this they must necessarily influence the near-term path for GDP and unemployment. Rarely, if ever, can so much power have been wielded by such a small number of institutions outside the direct democratic process." But, like all forms of power, it will not last for ever.

The fashion at present is for using short-term interest rates as the main policy weapon. But central banks cannot control, although they can try to influence, long-term rates, even nominal ones. They cannot control real interest rates,even short term ones, except for highly temporary periods.

Ultimately, the main variable that central banks influence is the money supply. I am not suggesting that we should go back to a regime based entirely on monetary targets. But we should recognise that interest rate objectives are simply a poor proxy which have to be used because we do not know enough about movements of velocity or the appropriate monetary aggregates to target.

Please notice that I said 'influence', not 'control'. The relationship between the monetary base and the broader definitions, which are what most people understand by money, are neither rigid nor easy to predict. In any case the present powers of central banks will not last indefinitely. As Mervyn King, the deputy governor of the Bank of England, argued in a speech in Jackson Hole in 1999:"At present, central banks are the monopoly supplier of base money - cash and bank reserves. Because base money is the ultimate medium of exchange and final settlement, central banks have enormous leverage over the value of transactions in the economy, even though the size of their balance sheet is very small in relation to those of the private sector." But electronic transactions in real time hold out the possibility that "final settlements could be carried out by the private sector without the need for clearing through the central bank... Without such a role in settlement, central banks in their present form would no longer exist; nor would money." (Bank of England Bulletin, November 1999, pages 410-411).

In the meanwhile we should be a little careful about putting all the emphasis on the one objective of inflation targeting. It was not long ago that many central banks denied that they could control inflation, which they thought was a matter for incomes policy or fiscal policy. It is of course quite right to say that in the long term monetary policy mainly influences nominal variables and that growth and employment depend on the underlying behaviour of the real economy. But Keynes was also right to remark that in the long run we are all dead. Not only does the short run matter in its own right. Long periods of high unemployment or low growth can themselves have a deleterious effect on the long run or equilibrium rate of these variables.('Hysteresis').

Similary the recent habit of complete neglect of the exchange rate is just as questionable as the earlier habit of claiming that exchange rates were immovable, and then facing a political and confidence crisis when these untouchable rates were after all seen to move. Monetary policy is still an area where fashion reigns supreme.

Finally a brief word on asset prices. No-one has a perfect idea of how to take them into account. But simply giving them some weight - almost any weight - in the inflation target, in addition to the consumer price index, would surely be better than nothing. Even according to present fashions, central banks are responsible for the value of their currency. This value depends not only on domestic consumer prices but on what can be purchased with that currency abroad when it is exchanged for other currencies. It also depends on the movement of property prices and land prices.

The moral of all these examples, which I have given in such a desperate hurry, is that it is a mistake to subordinate everything to one objective, or to one definition, and that wise policy must weigh up many different goals. This was what I used to be told when as a junior journalist I went to see wise old central bankers. There is a danger however that today's central bankers try too hard to be technocrats and forget altogether the wisdom of their forefathers.