Quis custodiet ipsos custodes? Juvenal's question would today surely apply not to the guards but to our new guardians: who will regulate the regulators?

The issue was given a twist this week when Northern Electric, one of the regional electricity suppliers, announced that it would give back £450m to its shareholders as part of its defence against a takeover bid from the construction and engineering group Trafalgar House. When companies find themselves subjected to a takeover bid they usually discover ways to boost their profits or reorganise their operations to the benefit of shareholders. But this sort of largesse is extremely unusual, and testimony to the extraordinary profitability of the regional electricity companies. The "recs", as they are called, are in the process of paying back money to shareholders by devices such as special dividends - payouts over and above the normal flow of funds the shareholders might expect.

Why are they so profitable? In a word, monopoly. Electricity is generated by several different power companies, so there is competition in that part of the chain. But the business of moving the electricity into our factories, offices and homes, through the national grid (which is what the regional companies own between them, and are all planning to sell) and the local distribution networks, is a natural monopoly. We cannot choose our supplier of electricity, for that depends on where we live.

It was because the market could not provide the usual commercial discipline that, when the public utilities were privatised, each was twinned with a regulator: Oftel for the phones, Ofgas for gas, Offer (the Office of Electricity Regulation) for electricity and so on. What no one knew was how these regulators should work, because we had no experience to go on. Nationalised monopolies had been formally regulated by parliament, through a minister who was notionally responsible for their activities. In practice such regulation was either non-existent or capricious - indeed one of the reasons why nationalised industries have failed as a form of corporate organisation is the difficulty of providing appropriate control.

So there was no model, no precedent, no template. Nor was their any foreign experience which could be drawn upon, for this country pioneered denationalisation. As a result our regulators had to learn on the job. They had to devise systems of controlling pricing which would allow the industries to be profitable enough to carry on their investment programmes, and provide adequate (or improved) quality of service, but not so profitable as to constitute an abuse of their monopoly power. More than this, they had to invent a whole new culture of regulation, working out what was expected of them and what would be going beyond their proper powers. Though they had duties under statute, much of the interpretation was left to one individual appointee, who inevitably sometimes made mistakes.

In the case of the "recs" the regulator, Professor Stephen Littlechild, made a mistake. His latest price formula, revealed last August and setting prices for the next five years, has clearly been too lax. There are two possible reactions to this. One is to call for the head of the regulator, and predictably there have been suggestions that Professor Littlechild should resign. The other is to ask whether we cannot refine our regulatory system so that it works demonstrably better. To do that we need to regulate the regulators.

There are two main regulatory failures. The first is the technical basis on which prices are controlled; the second the corporate governance of the regulatory body.

Typically prices are set by some formula. This may be a simple one: the retail price index plus or minus a figure. Or it may be complicated - Offer has drawn up an extraordinarily complex equation for the recs. But in either case it is rigid: once the formula is fixed, it is fixed, and in the case of electricity, for five years.

Yet surely there is a powerful case for allowing an annual adjustment to the pricing formula. Unfair on the companies? Not really. Many commercial firms have little idea of what prices the market will set for their products or their raw materials even a few weeks ahead. All that the regulator would be doing is to mimic a little more accurately the sort of conditions which the harsh competitive world would normally apply.

But a capricious regulator must not be able to impose frequent changes in an arbitrary fashion. The normal way western societies guard against individual arbitrariness is by splitting responsibility. To share responsibility among a board of regulators would reduce the danger of corruption - which mercifully we have avoided so far, but which is potentially enormous in any system where billions of pounds turn on the decision of a single individual. And it would increase the evident legitimacy of any decision.

Just how many people should be on the board, how they are appointed, how frequently they meet, how long they serve and so on are important details, but details none the less. The problem is the concept of the single individual regulator. It was not a bad first shot, more than a decade ago, when we had to invent a new type of body; and some (such as Oftel) have been a clear success. But we can now see a better way of ordering our affairs. The real lesson of the past week is not that Professor Littlechild should go; rather it is that all the regulators should go.

The wider point here is that quality of regulation is becoming one of the key determinants of competitive advantage between countries. As governments withdraw from direct involvement in producing goods and services they turn to regulation to ensure that the quality of service is maintained. But bad (or rather inefficient) regulation can be just as damaging to an economy as lack of regulation. The more successful we are in learning from the regulators that we have created, the more we can use them as a model to improve other services, public and private. This is a new game for government and will become an even more important one. We are not doing badly thus far, but we need to learn from our mistakes.