The survey suggests that, while pushing more authority down to the operating level, companies are simultaneously strengthening central power - ready to take the rapid, big decisions on, for example, acquisitions and disposals. There is no contradiction here, because acquisition policy is frequently competition carried on by other means. That is, it becomes an extension of the operating manager's fight for market share, and that, for companies with multinational pretensions, increasingly means global share. The local manager needs to be closely involved in central decisions, and carries a wider, corporate responsibility than for his immediate business.
A spectacular example of a company trying to shake itself free of the head office burden occurred last year when Robert Horton took over as chairman of BP. Within a week he announced that more than 1,000 jobs would be cut from the massive headquarters in the City. With them went an elaborate hierarchical structure and some 70 committees, all supposedly needed to manage or co-ordinate the global business, but in fact stifling initiative and action in the operating units. A regional HQ in Brussels is being established to discourage head office interference in - or indifference to - European issues.
But Horton's plan goes much, much further. As he explained at the time: "What I am trying to do is to simplify, to refocus, to make it clear that we don't need any longer to have hierarchies. We don't need any longer to have baronial departments. This is a fundamentally different way of looking at the way you run the centre of the corporation."
Whether Horton's dream will be realised, and how effectively the company will work without those hierarchies, are still sub judice. What cannot be denied is that until the change BP's results, for all its head office brainpower, were no better (in many experts' opinions they were rather worse) than those of other oil companies. Nor could it claim a better record in foreseeing the great oil shocks, the collapse of communism or other world events affecting its business.
Horton's plan now is to develop a system of networking: an idea that is rapidly gaining ground among theorists, although, as he has admitted, general experience of it is limited. The term also describes information technology systems, in which BP is investing heavily. The two are, of course, linked. In the management sense, however, networking implies less reliance on hierarchical relationships, more on informal groupings of peers across departmental and national boundaries for specific tasks.
Many organisations - including the big accountancy partnerships - have been working at least partly in that way for years. The Japanese, in spite of their hierarchical instincts, also come close to it. The claimed advantages are better and quicker communication, better use of ideas and expertise, better motivation of staff. The disadvantages may be diffusion - if not confusion - of responsibility and commitment, and more difficult management relationships in times of crisis. The system relies upon maturity and mutual confidence among executives, which may be lacking or take some time to emerge. In any event, it means that the old centralised/decentralised axis loses even more value as a means of describing the power structure inside the company.
Few companies display BP's courage in attempting such far-reaching changes, in attitudes and methods of working as well as structure. Most prefer to modify what is there already, with, perhaps, different and rather more limited objectives in view. Cutting the weight and rigidities imposed by a large head office is only one of them.