Big pay, better work? You must be joking

Executive pay is among the most extraordinary social and economic phenomenons of our time.

Conventionally, chief executives' pay is determined by markets and performance. An influential branch of academic inquiry, agency theory, has legitimised the idea that pay should be used to incentivise managers to boost performance to the benefit of shareholders.

Corporate governance codes have striven to put this into operation - the Association of British Insurers has been congratulated in the press for updating its remuneration guidelines and announcing that 70 per cent of the FTSE-100 index of top companies listed on the London Stock Exchange met standards designed to ensure that executives are generously rewarded only for outstanding performance.

They must be joking. The best efforts of US and UK researchers have failed to unearth a link between pay and corporate performance in the short term. But some explosive research at the UK's Manchester and Royal Holloway business schools blows away the idea of any correlation between CEO pay and performance in the long term as well.

The results, says Manchester's Professor Karel Williams, one of the authors, are "pretty devastating". They show that from 1983 to 2002:

* Real CEO pay in the UK and US rose by 25 per cent a year, come rain or shine, compared with sales and profits growth of less than 3 per cent.

* Aggregated, while real sales and pre-tax profits of FTSE firms increased just over 50 per cent during the period, CEO pay shot up more than 500 per cent over his (and we do mean his) tenure of seven years - a chief executive could expect to see his salary double twice.

* While the real market value of the firms increased substantially (350 per cent), this indicator has least to do with management, being largely the product of the sharemarket boom and steady flows of investment funds.

* Although meaningful markets for top managers do not exist, a "going rate" for FTSE-100 CEOs sits at around £1 million ($2.7 million), regardless of performance.

* While CEO pay has towed other senior managers in its wake, there has been no spillover to average employee pay. UK CEOs now pocket about 50 times as much as ordinary employees compared with nine times 20 years ago. In the US, disparities have risen from 50 to 281 times.

In terms of individual companies, the value creation versus pay story is even less "outstanding". Of FTSE survivors, only GlaxoSmithKline showed long-run real sales and profits growth in line with extravagant increases in CEO pay. Of the three other companies where CEO pay had gone up most - more than 1000 per cent - Aviva had seen pre-tax profits plunge by 152 per cent, GUS by 22.3 per cent and Marks & Spencer 25 per cent.

Of course, such calculations are greatly affected by the chosen beginning and end points. But, in sales terms, a surprising number of the companies were actually smaller than 20 years before.

Williams says that giant firms are "GDP companies" - they grow more or less at the speed of the economy, and although managers do little to create long-term value, they are "uniquely positioned to enrich themselves without obvious victims as neither shareholders nor labour lose directly en masse".

The paper says "pay for performance", and corporate governance in general, can be seen more as ideological incantations for selling "market capitalism with responsibility" and high pay as an element of that, rather than a realisable programme.

Paradoxically, the academic theory that sets out to explain performance pay, and the regulatory framework to keep it in check, have served to free it from any other sort of control. As the paper notes, time after time outcry over "excessive" pay has preceded ineffectual attempts to regulate it that have only succeeded in setting a higher baseline.

But, as Williams says, breaking the link between pay and performance allows for very different thinking about executive salaries - and a more promising agenda than bankrupt agency theory debates.

On the first, the paper suggests that today's corporate managers are somewhat like landed aristocracy in the 19th century or political elites of the Third World: the benefits they receive, and any value they create, are the result of the prevailing form of development rather than any real contribution. That leads to important questions, some of which will be studied at a new Economic and Social Research Council centre for studying socio-cultural change, and in a forthcoming book.

Why is strategy in giant firms growing no faster than GDP? Shouldn't inclusion embrace top earners as well as the bottom?

Why should one set of managers be paid on a completely different scale from others having tasks of at least equal complexity and responsibility - prime ministers, permanent secretaries and generals? And what should be the going rate for a large company CEO? How many deserve any increase at all? i Observer

* "Pay for Corporate Performance or Pay as Social Division: rethinking the problem of top management pay in giant corporations," by Ismail Erturk, Julie Froud, Sukhdev Johal and Karel Williams.