These superstars get everywhere, and there is nothing to stop them

Diane Coyle On the force behind rising inequality

AFTER the space shuttle Challenger exploded soon after take-off in January 1986, the cause of the tragedy turned out to be a fault in a simple doughnut-shaped rubber seal, known as an O-ring, linking two connecting pipes. One of the simplest components in the multi-billion- dollar shuttle lay behind the loss of life and the destruction of the work of thousands of people.

The report the following June of the presidential commission appointed to investigate the accident concluded: "The failure was due to a faulty design." More broadly, it said, the number of shuttle flights scheduled had put too much strain on the capacity of the workforce.

The lesson is not just the obvious one that a chain is only as strong as its weakest link. For economists there is an additional moral: the smallest error in one component or by one team member can put at risk the quality of an entire finished product or project. The higher the quality required, the more serious an issue this becomes - and in the advanced economies, quality is rising all the time as manufacturing and services progress steadily towards higher and higher value added activities.

This observation forms the basis of one explanation for otherwise inexplicable patterns of change in earnings, namely the fact that increasing income inequality is to be observed not just between high-skill and low-skill employees, but also within different professions where skill levels must be similar. The rise in inequality has what has been described as a "fractal" nature, to be found at any level you look in the figures. It is not just that top lawyers earn more than their secretaries, but that earnings among lawyers and amongst secretaries are also more unequal.

One widely accepted explanation for this is based on the "superstar" hypothesis. This says that, just as the existence of a mass market for films allows some individuals, perhaps only slightly more talented than others, to capture a huge market share because cinema-goers would rather see a known quantity like John Travolta or Gwyneth Paltrow than an unknown. It is virtually costless for the stars to reach a bigger audience - they only act the once during filming - and the audience overwhelmingly demands the market leader. The superstar phenomenon is being extended by new technologies through ever wider swaths of the economy because these supply and demand conditions are replicated in all sorts of weightless industries. Now we have superstar lawyers, designers, doctors and even intellectuals.

What, then, does the O-ring example add as an explanation for increasing income inequality? The answer, set out in a new book by Daniel Cohen, a French economist*, is that, as Nasa discovered, all the workers in a team need to have a similar level of competence. No matter how much it would cut the wage bill, it is not worth hiring less than the very best lab technicians you can find. Similarly, the top law firms need the top secretaries, whose pay will reflect their value to the company, whereas secretaries working for neighbourhood solicitors will earn considerably less. It means that very slight differences in skill and performance are magnified into big disparities in income. There is no longer a single market for any particular type of skill - an individual's prospects depend on which company they are employed by, and career paths are much more volatile.

This phenonemon is technology-fuelled in two ways. One is that advances in technology are driving the advanced economies down the path of adding more and more value as low-value added production can so easily be carried out in emerging economies. The other is that it is increasingly easy for goods and services to be produced more flexibly with various components sub-contracted out. Less mass production means smaller units of production which can more easily become homogeneous in terms of the skill levels of the individuals within them. An outfit which is very high skill across the board can buy in any low-skill goods or services it needs. Thus the City law firm will subcontract its cleaning as cheaply as possible while paying its secretaries a relatively high wage.

The really bad news about this process, which Professor Cohen calls "assortative matching", is that it is disastrous for the low-skill minority. It makes their exclusion from higher incomes look pretty much insuperable. Indeed, the usual prescription of improving the education and skill level of the workforce might help in the long run but in the short run can end up excluding the people at the bottom of the skills heap even more thoroughly. It all looks rather gloomy for politicians - or unions - who want to halt the trend towards obscene inequality observed in the US and UK.

The New Labour Government's "Fairness at Work" White Paper this week is unlikely to offer any correctives to the fundamental economics driving the trend. Nor are issues like the minimum vote needed for union recognition, the TUC's chosen terrain for battle, anything more than a sideshow compared to addressing the huge underlying shifts in the situation of the underdogs and the top dogs in our society.

One possibility for shaping future policies, however, lies in recognising that the O-ring theory implies that the productivity level of individual workers does not depend on their own efforts alone. It also depends on the productivity of their colleagues at all levels. It is not brute strength in numbers that might form a basis for collective organisation in the workplace, as in old-style unionism, but rather genuine economic interdependence. For all that management gurus spout about the importance of teamwork, few managers seem to take it seriously, or at least see it as applying only to select categories of employees. That will change.

Another question is the openness of high-value, high-income jobs to all comers. It will be essential, for reasons of fairness and social peace, to make sure that the jobs market is as mobile as possible, with no closed shops. This is less of a worry in an entrepreneurial economy like the US where anybody can start up their own business. Even so - as with so many other aspects of economic policy - regulation rather than intervention could turn out to be the key issue.

However, once an economy has started down the path of inequality, when the organisational, legal and cultural barriers to some categories of people receiving higher and higher incomes have vanished, it is impossible to turn back. Tax away high salaries? Many of us might cheerfully see the top income tax rate go up to 50 per cent, say, but there is no appetite among voters and politicians anywhere for a return to punitive taxation. Introduce a pay policy that sets upper earnings limits? It would be unpopular and unenforceable.

In the very long run increasing the productivity of the entire population through education and training is probably the only answer, as it was after the Industrial Revolution. Meanwhile, fairness will be best served by making sure the opportunities for superstardom and access to high-value jobs are as widely available as possible. Where there is no possibility of more equality of outcome, equality of opportunity is the only alternative.

d.coyle@independent.co.uk

*'The Wealth of the World and the Poverty of Nations', Daniel Cohen, MIT Press pounds 19.95.