Just what has Cedric Brown started?

The princes of the corporation are uncomfortable. They feel they're being watched. And they are right. Their pay packets are being opened in public. Shareholders are calling senior management to account. Most famously, at last week's AGM, small shareholders in British Gas used Mrs Thatcher's idea of shareholder democracy as a stick with which to beat institutional investors. It was as if "one share, one vote" had subtly metamorphosed into "one shareholder, one vote". The notion of people's capitalism made popular in the Eighties has clearly generated a new set of expectations about the governance and accountability of companies.

In response to the row over executive pay at British Gas and elsewhere, the Greenbury Committee is preparing guidance on boardroom remuneration. And Labour is talking of requiring shareholders to approve directors' pay in advance, and insisting that pension funds publish investment plans before AGMs.

But the debate about British Gas is not just over pay: it's also about "just who are companies for?" This is the issue that yesterday's Royal Society of Arts (RSA) report on "Tomorrow's Company" has put centre-stage. The RSA urges firms to let more "stakeholders" into their citadels of power. The US Department of Labour published research this week supporting their view that companies perform better when they involve workers in decisions.

There are three main propositions in response to the question about who companies are for. For the sake of convenience, let's call them the "shareholder value", the "multiple stakeholder power", and the "managerial power" models.

In mainstream British and US circles, the orthodox view is that companies are there to deliver returns to their owners, the shareholders - no less, no more. The free market economist Milton Friedman once described charitable giving by companies as "theft" from shareholders. The for-profit firm, it is argued, is the most efficient organisational form because shareholders have a direct financial incentive in the size of their dividend cheque and the value of their shares, to keep down costs, innovate, and to respond quickly to changing market conditions. On this view, forcefully advocated by Samuel Brittan of the Financial Times, global competition will weed out firms that try to satisfy other interests, because their costs and their sluggishness will mean they're left behind. Even worried German and Japanese commentators are starting to pick up these ideas.

The second view, put forward by the RSA and by Will Hutton in his book The State We're In, is that companies work best in the long run, and secure public legitimacy, if they take account of the interests of many other stakeholders. These include their own employees, the communities living near their offices and plants, their customers and the environment. As the RSA says, their governance must be "inclusive". They argue that high morale among employees makes for greater productivity. Mr Hutton's supporters on the left use these arguments to support the Social Chapter and the minimum wage. They go on to claim that greening the firm makes for long- run sustainable profits, that corporate philanthropy is part of long-run recruitment and staff development, and can help to identify market opportunities. If efficiency is the watchword of the "shareholder value" view, then "legitimacy" is the hallmark of the multiple stakeholder model.

The third view is that salaried managers have run firms for decades, largely in their own interests and looking first to their own power and pay, and they always will. Made famous in the Thirties and Forties by James Burnham and repeated this week by Anthony Sampson in his book Company Man, the idea is that firms are driven by managerial power and dominance rather than efficiency or legitimacy.

This all looks rather like an old-fashioned left-right battle, with the left claiming to speak for the "long term" and the right pointing to the "bottom line", with a bunch of "cynics" on the sidelines pointing to megalomania as the driving force.

But in their simple forms, all three views are misleading. A policy of pure self-interest isn't even in the long-term interests of shareholders. Indeed, sensible shareholders recognise this and aren't obsessed with the quarterly figures. Many already take along-term view, and recognise a core of good sense in the "multiple stakeholder" view. They don't insist that all the slack be taken out of the firm because they know that companies need some spare capacity to respond when conditions change.

The danger, on the other hand, of involving all stakeholders in governance is that decision-making slows down or locks up altogether. Buying off special interests is expensive, and leaner firms may drive the kindly humanitarians out of the market, exactly as some German and Japanese business people fear. Alternatively, the RSA and Will Hutton model can be a mask for managerial power. Many managers rather like a cluster of relatively weak interests to play off, because behind the legitimacy of these groups they can consolidate their own power - and become even more complacent.

Many employee-owned companies aren't doing well these days. The John Lewis Partnership is finding the high street a tough place, and United Airlines has been failing for some time. While no firm has yet collapsed from too much charitable giving or environmental correctness, these budget lines are risky. Managerial power isn't everything either. Management- dominated companies can fail spectacularly: many people argue that this is why IBM recorded enormous losses by the early Nineties.

The first function of the company will always be to make money - in the long term - for shareholders. Criticism of excessive pay for chief executives is motivated by a feeling that they enjoy too much power and exercise that power in their own narrow interests. Boardroom pay matters when directors and senior managers aren't delivering long-term shareholder value. It also matters when the rest of the organisation hears a message that everyone should take what they can get out of the company as quickly as possible, rather than concentrate on the firm's long-term health. There is evidence that firms that pay managers too much perform badly.

However, it doesn't follow - as Labour is suggesting - that institutional investors are the problem and that new legal duties on them are a solution. In most firms, pension funds and insurance companies are usually more vigilant towards management than individual shareholders have the time or inclination to be. Erecting new legal hurdles in the process of decision- making within the firm simply makes for red tape, without offering a viable way to reinforce enlightened shareholder self-interest.

Getting ownership and governance right is this week's obsession, and there are practical steps that many companies can take to improve it. Many have been suggested by the Cadbury Committee which was established in 1991 and which in a survey last week showed that ideas for remuneration committees, clearer roles for chairs and non-executive directors and new audit arrangements are being widely adopted.

Creating roles or rights for ever more stakeholders, complaining about institutional shareholders and chanting slogans about shareholder democracy won't solve the problems of companies like British Gas. There is no single right model of ownership and governance and a healthy economy needs a variety. How companies are managed and how the economy is being managed matters much more for all stakeholders than how their firms are governed.

THE SHAREHOLDER CONGLOMERATE

Hanson is an example of a UK company that says it puts shareholders' interests first. Lords Hanson and White have pursued a ruthless strategy of buying and selling businesses to make money for its shareholders. There is little that it is not prepared to do if it thinks the result will be to increase the market value of the company. Its loyalty is ultimately to the bottom line, not to the people it employs. The weaknesses of the group are twofold. It lives or dies by acquisitions: if they dry up, so too does its growth. And it has a wholly autocratic management style. It is unlikely that the company will survive when Hanson and White retire.

Norwich Union is a mutual society, owned by its policyholders and run by a board that in effect decides its own successors. Recent results have been disappointing.

Poor ownership? Critics of mutual ownership blame the absence of shareholder pressure on managers to improve performance. There are so many owners of a mutual society, each with a tiny stake, that their influence is extremely diffuse and rarely exercised.

The end of the mutual? In 10 years' time, mutual insurance societies will look increasingly isolated. Building societies will mostly have converted to banks.

The John Lewis Partnership has been owned by its workers since 1929.

Worker partnership: Worker-partners share in profits through an annual bonus. The chairman could be voted out of office by a council dominated by partners.

Performance: Critics say John Lewis is slow moving, over-staffed and under-invested. But a 1992 study found overall performance matched publicly quoted rivals.

Continuity: In 10 years' time, John Lewis is likely to be recognisably the same as today, just rather bigger.

THE CUSTOMER-OWNED SOCIETY

THE WORKER-OWNED COMPANY