Alfred Sloan's Memoir X: Summing UpIn Chapter 23 of My Years with General Motors, Alfred Sloan summarizes one of his central convictions concerning management of a multi-division corporation:
It has been a thesis of this book that good management rests on a reconciliation of centralization and decentralization, or "decentralization with co-ordinated control."
Eash of the conflicting elements brought together in this concept has its unique results in the operation of a business. From decentralization we get initiative, responsibility, development of personnel, decisions close to the facts, flexibility in short, all the qualities necessary for an organization to adapt to new conditions. From co-ordination we get efficiencies and economies. It must be apparent that co-ordinated decentralization is not an easy concept to apply. There is no hard and fast rule for sorting out the various responsibilities and the best way to assign them. The balance which is struck between corporate and divisional responsibility varies according to what is being decided, the circumstances of the time, past experience, and the temperaments and skills of the executives involved.
The concept of co-ordinated decentralization evolved gradually at General Motors as we responded to tangible problems of management. As I have shown, at the time its development began, some four decades ago, it was clearly advisable to give each division a strong management which would be primarily responsible for the conduct of its business. But our experience in 1920-21 also demonstrated the need for a greater measure of control over the divisions than we had attained. Without adequate control from the central office, the divisions got out of hand and failed to follow the policies set by corporation management, to the great detriment of the corporation. Meanwhile, the corporation management was in no position to set the best policies, since it was without appropriate and timely data from the divisions. A steady flow of operating data, for which procedures were later set up, finally made real co-ordination possible.
[. . .]
Much of my life in General Motors was devoted to the development, organization, and periodic reorganization of these governing groups [governing committees and policy groups] in central management. This was required because of the paramount importance, in an organization like General Motors, of providing the right framework for decisions. There is a natural tendency to erode that framework unless it is consciously maintained. Group decisions do not always come easily. There is a strong temptation for the leading officers to make decisions themselves without the sometimes onerous process of discussion, which involves selling your ideas to others. The group will not always make a better decision than any particular member would make; there is even the possibility of some averaging down. But in General Motors I think the record shows that we have averaged up. Essentially this means that, through our form of organizatipon, we have been able to adapt to the great changes that have taken place in the automobile market in each of the decades since 1920.
[pp. 429-430, 435, 1990 edition]