My Years With General Motors

by Alfred P. Sloan Jr.

  Of course the Executive Committee had always been legally, and in fact, the top operating committee, but because it met jointly with the Operations Committee, and because decisions were made with the participation of both policy and operating people, the line between policy and administration was not sharply enough drawn. The first thing was to limit the Executive Committee to policy decisions made independently of the administrative, operating people.

  It was particularly important that independent policy-making be re-established because of the new conditions in the car market and the management problem that would arise if we restored the traditional five divisions, a move which I proposed.

  The situation was this: In the automobile market of 1933 the low price group grew relatively to encompass 73 per cent of the industry's unit car sales, as compared with 52 per cent in 1926. This meant, for the old car line, that we would have four lines in 27 per cent of the market and one line in 73 per cent of the market. Mr. Brown, for reasons of economy, preferred three divisions. I preferred five despite the extra cost, which I thought we could recover through increased volume. I stated, and in part restated, my longstanding views on commercial policy in a report to the Finance Committee on January 4, 1934, and, since they became the corporation's policy, I quote them here:

  THE FUNDAMENTAL CONCEPTION OF GENERAL MOTORS CAR PRODUCT PROGRAM

  Certain members of the Committee may recall that at the time Mr. P. S. du Pont became President of the Corporation, one of his first acts was to appoint a group to study the very important problem of car products. Up to that time there had been no fundamental conception or plan —there was no definite relationship between the products of one car division and another, or, in other words, no coordination. It was generally recognized that there should be a definite relationship and a certain amount of coordination, and the purpose of the study was to establish what that should be. This study was authorized by a resolution of the then Executive Committee, under date of April 6, 1921—now approaching thirteen years ago . . . [The study referred to here was that concerning product policy discussed earlier.]

  I took note of the general evolution of the automobile in those thirteen years under the pressure of intensified competition, and observed that the value of a car had come to be centered around certain considerations—appearance or style, technical qualities, price, and reputation. It was my impression at that time that differences in these matters were narrower than they had been in earlier years, and that because the state of the art was available to all, automobile technology would not in the future result in effective differences from a selling point of view. I was mistaken about that, although my general point, I believe, was valid, namely that selling had begun to focus on personal preferences of the consumer, especially in matters of style. I said:

  People like different things. Many people do not want to have exactly the same thing that the neighborhood has. The design of any car is an artistic and engineering compromise. No car can embrace all the desirable features. Relatively inconsequential features will often influence a sale[,] adversely to the customer's interests with respect to other far more consequential features. No prospect is intelligent enough to definitely determine the weighted value of all the elements that enter into any particular car. The consumer is also greatly influenced by the personal relationship with the dealer and some times rightly, other times, wrongly, becomes antagonistic to one particular dealer. General Motors in selling 45% of all the units in the industry, which is practically one out of every two, assumes a tremendous responsibility with respect to all these questions. Under those circumstances, new customers are hard to get and an old customer lost is hard to replace. It is quite different selling 45% than in selling 5%.

  It is perfectly possible, from the engineering and manufacturing standpoints, to make two cars at not a great difference in price and weight, but considerably different in appearance and, to some extent, different in technical features, both, in degree, built with the same fundamental tool equipment.

  In view of the concentration of volume within a narrow price range, is it desirable for the Corporation to place all its eggs in one basket in view of the above and many other considerations, or is it better to capitalize the recognized fact that different things appeal to different people; that not all the sound engineering ideas can be incorporated in one unit; that the influence of the dealer is an important consideration? . . .

  I answered these questions with a statement of my commercial policy as follows:

  ... I believe that in the low price field where from 80% to 90% plus of the volume is concentrating, we should have more than one appeal, and that our representation, whatever it may be, should contemplate the incorporation of sufficient differences in essential elements of design, as to make the broadest possible appeal to the largest number of people. I accept, as a corollary of this principle, the fact that manufacture and distribution is complicated. I regret that we can not make one thing that everybody will buy, but I do not think, under present circumstances, that is possible.

  In communities of high potential where there must necessarily be many dealers competing against each other, with the same product in the same market, I believe that it is better policy to limit the number of dealers competing for one line and obtain the increased number of contacts through a different appeal than it is to go the other route.

  To illustrate, let us take a community in which we might maintain "x" dealers. Rather than have those "x" dealers competing with each other on the basis of exactly the same merchandise, which is very demoralizing competition, I believe it is better to have a part of same, and a major part of course, competing against each other with the Chevrolet line and additional dealers competing with lines diversified from Chevrolet.

  For all the above reasons, I personally feel that the policy as outlined many years ago [1921] by the Executive Committee, should be modified substantially as follows:—

  THAT giving recognition to the concentration of volume in the low price field, it shall be the policy of the Corporation to increase its representation within that field, but in doing so the utmost consideration must be given to the importance of the greatest possible diversity in all elements of consumer appeal in such additional offerings so as to build the strongest foundation of acceptability to the consumer.

  This proposal to maintain diversity in cars and separate divisional selling efforts in a crowded price class required new forms of co-ordination. And the more you co-ordinate, the more questions you draw up into the policy area, and therefore the finer must be the distinctions between policy and administration. For example, when two or more divisions use common components, the independence of each division is limited to the extent that there must be a common program between them. Someone therefore must coordinate such a program. As this process is elaborated, more questions come into the policy area which were formerly in the administrative area. I have always believed it is imperative to maintain the distinction between policy and administration. Without that distinction a decentralized organization would be in constant confusion as to what was decentralized and what was not. So the big question was the policy one. And it demanded a general solution. The solution we worked out at that time remains today the basic decision-making process in General Motors. It was contained in the following recommendation which I made to the Executive Committee in October 1934:

  It has already been developed that the creation of a policy might emanate either from the Central Authority or from within the Operating Division or Subsidiary. The finalization or authorization of a policy is exclusively within the province of the Central Authority as presented by a Governing Committee. From whatever source the policy may originate, it is essential that those involved in authorization should be qualified to pass on the proposal with a full knowledge of its implications as affecting the present and future position of the business. Where a policy involves important consequences, as in the case of the operations of General Motors Corporation
, it is essential that the broadest possible foundation of thought and fact and the approach of the proposal from as many different angles as possible, be established. To the degree that authorization becomes superficial, to that degree does the position of the business become jeopardized, or its progress adversely affected.

  The above more or less philosophical discussion is for the purpose of establishing the reasons for and the desirability of formulating a broader conception of the policy creation phase of administration than has heretofore existed in the General Motors scheme of things.

  This Procedure establishes, for the first time, the following principle:—

  l—that the development, or creation of constructive and advanced policies, as defined, is of vital influence in the progress and stability of the business.

  2—that in the organization of General Motors Corporation, the fact just recited, should be recognized through a specialization of policy creation, independent of policy execution, so far as that is reasonably practical.

  The concept of policy creation described above was embodied in a set of new organs in General Motors, called policy groups. These groups, generally speaking, had functional titles, such as Engineering Policy Group, Distribution Policy Group, and the like, with later some line groups such as the Overseas Policy Group. They combined the top executive officers, including the president, with functional staff men, and each group was charged with making policy recommendations in its functional area to the top operating policy committee of the corporation. The divisional managers, being charged with administration, were specifically excluded from these groups. The groups themselves, as groups, were not given authority over the divisions, or any authority to finalize policies, but since the groups contained the principal officers of the corporation, policy group recommendations were normally adopted by the governing committee in the area of operations. We tried out policy groups in the engineering and distribution areas between 1934 and 1937, and in the latter year expanded their use to other functions and some operations and adopted them officially for the corporation. (Note 10-2.) They express in more sophisticated form the management policy I first formulated in the "Organization Study" of 1919-20, that is, of decentralized operations with co-ordinated control.

  There are now nine policy groups in the corporation, divided into two categories. First, there are those that deal with functions —namely, Engineering, Distribution, Research, Personnel, and Public Relations—all largely related, though not exclusively, to the car operations. And then there are those that deal with certain group operations—namely, Overseas, Canadian, General Engine, and Household Appliance. In the case of each functional activity, these policy groups work under the auspices of the section of the general staff to which they are related; the Engineering Policy Group, for example, is related to the Engineering Staff through the Engineering Staff vice president. In the case of operations, each group is under the auspices of the group executive for that set of operations. Members of these various groups carry great influence at the highest level in the corporation. The chairman of the board and chief executive officer, for example, is a member of all but three of the policy groups, and the president a member of all but two. In the case of the Distribution, Engineering, Research, Personnel, and Public Relations policy groups, the members of the Executive Committee as well as others of the operating management are members. Their full membership comprises a cross-section of the whole of the corporation's executive talent. Hence their great weight in uniting staff and line operations, preparing policy recommendations, and developing the basis for executive decisions.

  The activity of the policy groups varies with the changing needs for policy determination. The Engineering Policy Group, for example, meets regularly to work on new product programs. (Note 10-3.) In this activity the general managers individually and collectively, in person and through their functional departments, are in intimate contact with the work of the policy groups. They are not members of the policy groups, as I have said, because those groups are charged with policy development and the general managers are charged with administration.

  The function of the Engineering Policy Group in the development of a new model in General Motors is a good example of the work of a policy group. The initiation of a product program in any division lies within the responsibility of the general manager of that division in collaboration with its engineering department, influenced of course by the market as reflected by its sales department and importantly co-ordinated with the needs of the other operating divisions. If we were to go back twenty-five or thirty years, we would find little co-ordination between a program proposed by one division and one that might be put forward by any of the other car divisions. But over the course of time a very great amount of co-ordination has become necessary. In other words, a divisional product program, instead of being integral in itself, is deeply involved with the product programs of many of the other divisions; therefore it must be developed from the corporate point of view. The time between the conception of any new program and its final execution is presently about two years. In the case of an advanced engineering concept, it is very likely to be more than two years. And during that time it is subject to an untold amount of change. Therefore, during the development period, detailed contact must be continually maintained between the engineering departments of all the car divisions, the Styling Staff, the Fisher Body Division, and more or less the accessory divisions—because they are all commonly concerned with and working together on a single problem. Here the corporation's Engineering Staff comes into the picture and collaborates with the other divisions in order to effect the necessary co-ordination. The co-ordination agency— if I may call it such—that passes on the problems involved in this process is the Engineering Policy Group. Its decisions are normally adopted by the Executive Committee, whose members are present through the whole process.

  It was the depression, and the economies it forced in co-ordination of the product, as I have shown, that led to this new type of co-ordination in management. With the establishment of policy groups in 1937, the scheme of management which originated in the Organization Study of 1919-20 was complete.

  Long reflection on the distinction to be made between policy and administration led me in 1937 also to consider the application of this concept more precisely to the organization of the governing committees of the corporation. Early that year I proposed that in place of the Finance and Executive committees we should have a single policy committee to concentrate on making over-all policy for the corporation and an administration committee to concentrate on the carrying out of policy. After considerable discussion we adopted this change in May of that year. The Finance Committee and the Executive Committee were discontinued. A new Policy Committee was formed. It was composed, of course, entirely of members of the board, and brought operating officers, financial officers, and outside directors into one senior group. The new Administration Committee was composed entirely of operating officers.

  The Policy Committee took over all the responsibilities of the former Finance Committee and additional responsibilities for operating policy. From 1937 through 1941 the Policy Committee made policy for operations in a number of important areas. For example, it established the broad outlines of labor policies and programs and also decided a number of distribution policies, particularly those concerning dealer relations. As international conditions became increasingly unsettled, more and more of its time was given to determining policies for the overseas subsidiaries. As war approached, the Policy Committee had to deal with increasing shortages of raw materials, relations with the government, and the impact on our civilian operations of government requests for the production of airplane engines, tanks, and other military items.

  The entrance of the United States into the war in December 1941 required a special change in our committee organization. To make the rapid shift to all-out war production, we set up on January 5, 1942, a War Emergency Committee composed of six top executives, mos
tly from the Policy Committee. This committee met once a week and sometimes more often. From January through April 1942 it ran General Motors. Temporarily the Policy and Administration committees did little more than ratify the War Emergency Committee's work. But in May 1942, as we settled into war production, we abolished the War Emergency Committee. We then transformed the Administration Committee, consisting of all the general executives and group vice presidents, into the War Administration Committee. For the next two to three years the War Administration Committee practically ran the organization. This was because our wartime policy was set and nearly all the corporation's work was war production. Aside from the technical problems of production, our policy decisions were concerned primarily with our relationships with various government departments.

  By 1945 the Policy Committee once more had come into its own as we were developing our broad postwar planning. Since the problems of reconversion and postwar business were so enormously important, nearly all major questions, even some concerned with the running of the business, came to the Policy Committee. As a result of this overburdening of the Policy Committee, we began to reconsider the structure and function of the committees of the board.

  A single policy committee was an ideal solution to the problem of separating policy from administration. But two factors promised to make it impractical in the new circumstances that were developing. First, it was apparent that the growth in the volume and complexity of the corporation's activities would require a greatly enlarged responsibility both in finance and operations at the board level. Second, it was going to be difficult to obtain the services of experienced outside directors with adequate time to serve on a committee dealing with operations as well as financial policy. Therefore, in 1946, we dissolved the Policy Committee and replaced it with the two traditional committees representing the primary functions of finance and operations, now renamed the Financial Policy Committee and the Operations Policy Committee. In 1958 we restored their old names, Finance Committee and Executive Committee, and in a further refinement, expanded their membership to obtain more overlapping of the individuals on the two committees.