Human Research Relation at sears

UNIT -1

Case: Human relation research at Sears (Group assignment)

Source: Management, 3e, Thompson/Myths, pg 28

Focus:  To review the early methodology used to understand employee attitude and the roles of leadership and organizational structure in affecting satisfaction.

Abstract

During the 1940s, Roebuck conducted extensive research on organization behavior. This paper discusses two important features of that work: employee attitude surveys, and the Sears “X-Y Study “dealing with the impact of organization structure on employee behavior and the influence of executive attitudes on organization structure.

1.       Background

A significant amount of human relations research was conducted by Sears, Roebuck and Co. during the decade of the 1940s, with some carryover into the 1950s. Much of this was done in cooperation with the Committee on Human relations in Industry of the University of Chicago and Social Research, Inc., a private consulting firm headed by the two principals of Committee on Human Relation, W. Lloyd Warner and Burleigh B. Gardner, and largely staffed with personnel trained by the committee. The origins and accomplishments of the committee are recorded in the companion paper to this one by David G. Moore, a member of the Sears personnel staff during the period under review and the chief liaison between the company and the committee. The purpose of the present paper is to record Sears’ participation in these collaborative efforts.

Sears, Roebuck Management

The Sears organization built by General Robert E. Wood was remarkable in many ways (Worthy, 1984). Wood was an Army man, trained at West Point and with a distinguished record as head  of supply and manpower service in the building of the Panama Canal and as Chief of the Army Quartermaster Crops during the First World War. Because of his experience in large-scale procurement, he joined Montgomery Ward in 1919 as merchandising vice president, a post from which he was summarily field in the fall of 1924, chiefly for his insistent urging that the company add retail stores to its mail order operations, a course adamantly opposed by Ward’s president.

                     The head of Sears, Roebuck, Julius Rosenwald, recognizing his good fortune in thus having a man of Wood’s experience becoming unexpectedly available, hired him immediately, with a commitment to support his radical retail notions. Early in 1925, Wood opened his first Sears store, and others followed. The venture was eminently successful, and three years later Wood was named president and chief executive. Under his leadership, Sears not only survived the Great Depression but learned from it many lessons that enabled Wood to build in a few years’ time the largest and most profitable merchandising business the world had ever known.

Strong Personnel Function

For an Army man, Wood had a curios distrust of authoritarian direction and control from above. His store managers, and to a considerable extent the department managers within the local stores as well, had wide discretion in the selection of merchandising to be carried, the quantities to be ordered, and the prices to be charged. To Wood, it was fundamental that managers at all levels worked best if they were held accountable for the results but left free, within  broad limits, in the ways they accomplished those results (Worthy, 1984, chapt.7 “Organizational Policies”). But Wood did not believe in blind delegation. For this kind of organization to work, special care had to be taken in the selection and placement of people, and this called for a strong personnel department.

Wood was also well aware that a widely-dispersed system of mail order plants and retail stores, each functioning with considerable autonomy, required that positive measures to be taken to avoid friction and divisiveness between employees and management, to maintain high levels of morale, and to provide motivation for superior individual and group performance (Worthy, 1984, chapt.9, “Employee Policies”). This too, called for a strong personnel department.

In Clarence B. Caldwell, one of his best store managers, Wood found the man he needed to head that critical function.  Appointed director of retail personnel in 1934, his responsibilities broadened in due course to include the entire company and he became one of the most influential and respected of all Wood’s lieutenants. Encouraged by Wood to develop rational and soundly-based policies for all aspects of Sears’ personnel operations, Caldwell in 1938 established a research and planning function within his department, and it was my great good fortune to be hired for that responsibility.

With college majors in English and economics, I was not a trained management or research person. However, I had been called an assistant deputy administrator of the National Recovery Administration, deeply committed to the humane values of the early New Deal, and after aiding in writing the history of NRA following its demise had served as personnel manager for a Milwaukee department store. I was attracted to Sears because it was obliviously a fast-growing company with good opportunities for advancement, and because it was already gaining a reputation for excellence in management-employee relationship. I was also intrigued by the fact that General Wood was one of the few big businessmen who had been a strong supporter of President Roosevelt before breaking with him over his war policies. It seemed to me that Sears were the kind of company I wanted to work for. I was not disappointed.

II. Employee Attitude Surveys

One of the first undertakings of my new unit I the Sears personnel department was the design and implementation of a standard compensation plan for all retail employees. Other activities included studies of employee turnover and the demographic characteristics of company personnel, projections of future executive and employee requirements, designing better methods of selection and training, analyzing the operations of employee benefits plans, and exploring possible means for improving them.

Initiation of Attitude Survey Program

A particularly significant phase of the work undertaken by the research and planning unit was the development of systematic means of assessing employee attitudes and morale. By 1993, General Wood and other senior Sears officer were growing increasingly uncomfortable with the fact that they no longer had the close touch with day-to-day affairs of the rapidly expanding retail system they once had had, and Caldwell and I were directed to come up with a means for providing “a reliable and continuing flow of information about………… how we are getting along with our people “(Worthy, 1984,p.152; see also Jones, 1952).

It seemed clear from the outset that the best way to find out what employees thought about their jobs and the company was to ask them, and for this purpose an outside firm, Houser Associates, was retained to design and administer questionnaires which employees answered anonymously (Houser, 1938). By the end of 1939, a number of experimental surveys had been made of representative stores, and the results were encouraging. Wood and his fellow officers felt that a good start had been made and that the surveys were giving them at some of the information they wanted. The project continued on a gradually expanding basis until Pearl Harbor, when it was suspended for the duration of the war.

Houser’s approach was essentially that of traditional market research: count the number of people who respond in different way to a series a questions. The results were interesting and useful as far as they went, but enigmatic in many respects. Clearly, something more than counting was needed. As my staff and I studied the numbers, we became more and more convinced that social influences of some form were at work and that a sociological approach to the data might prove useful.

Interpretation of Survey Results

By fortune coincidence, early in 1941 a young man with a master’s degree in sociology from the University of Illinois applied to me for a job. He was David G. Moore, at that time a member of the personnel research staff of Western Electric Company. I was impressed by his ability to describe work relations phenomena in terms that seemed to make sense. I hired him at once and put him to work analyzing the data our employee attitude surveys were beginning to produce.

Moore organized the Houser data in new ways, with interesting results. While within all groups there were relatively normal statistical distributions, there were significant variations by age, sex, length of service, type of work, level of authority, size of store, size and economic base of city, geographical region, and so on. The data began to be more meaningful, but raised even more intriguing questions as to why there should be such large differences in the attitudes of different groups toward precisely the same thing.

As the survey data were studied further, it became increasingly clear that store organizations, and properly commercial and industrial organizations in general, were “social systems,” and that parts of the systems were closely interrelated. Compensation was not an independent variable, any more than hours of work, working conditions, or quality of supervision, and changes in any of these or in other significant elements in the working situation could not be made without affecting other elements and without modifying the whole. The importance of relationships and the analysis of attitudes and behavior in terms of group identify and organizational status loomed increasingly large in our thinking (Worthy, 1943).

During this period, Moore often spoke to me about Dr. Burleigh B. Gardner whom he had come to know at Western Electric. Gardner was a social anthropologist on the staff of William J. Dickson doing research on the Hawthorne counseling program, and Moore had been one of his assistants. From Moore’s own training and from what he had learned from Gardner, he was convinced that some of the concepts of social anthropology, particularly related to status and social class, would be useful in what we were trying to understand at Sear. Sometime in late 1942 or early 1943, Moore arranged for the three of us to have lunch at the Palmer House. That was the beginning of a long period of fruitful collaboration.

Shortly thereafter I met with Gardner and W.Lloyd Warner, who were then in process of organizing the Committee on Human Relations in Industry at the University of Chicago. I was as well impressed by Warner as by Gardner, and the plans they had in mind held intriguing possibilities. Sears became one of first corporate supporters of their new venture.

Post-War-Planning

During the wars years, General Wood directed his corporate staff departments to utilize whatever time they could spare from otherwise pressing duties to prepare for the period of peace that would eventually come. Among other things, the personnel department addressed the subject of improving the employee survey program, and the services of the Committee on Human Relations were engaged for that purpose.

The new survey program drew heavily on Warner’s and Gardner’s comprehensive grasp of the concepts of social anthropology and their experience with the methods of field research. Gardner’s involvement in the non-directive interviewing program of Western Electric was also useful. Because Sears planned to use its program company-wide rather than at a single plant as at Western Electric, reliance could not be placed on non-directive interviewing alone; there were simply too many people and too much ground to cover.

The program as it finally emerged ready for use as soon as the war was over combined questionnaires and interviews. A simplified questionnaire was administered to all employees in a store or other company unit, and the statistical results analyzed to make tentative identification of areas which appeared to represent problems. A team trained in non-directive techniques then interviewed executives and employees in the unit itself, and others likely to have knowledge of the problems, to learn more about what caused them. The questionnaire were used somewhat as “Geiger counters” to locate areas meriting closer investigation, which might be certain departments, or particular classes of employees, or some  aspect of company policy or local practice, or perhaps some recent event. In addition to locating areas of difficulty, the questionnaires also give some indication of what the difficulties might be. Armed with information such as this, the interviewers knew where to look and what to look for (Worthy,1947,1950).

Although Gardner was deeply involved in designing the survey program and he and his committee staff in training survey teams and interpreting survey results, most of the actual survey work was conducted by members of the Sears employees in the course of developing the program that their level of confidence in the integrity to the company was sufficiently high that it would not be necessary for the questionnaires to be administered and the interviews conducted by staff of the Committee on Human Relations or by an outside firm, as had been done before the war with Houser Associates. This was a wise move because work as a member of survey teams proved to be valuable training for future personnel executives.

The New Survey Program

By this time Moore, under my general supervision, headed the Sears personnel research and planning staff and was in immediate charge of the survey program. His training as a sociologist, his experience with the personnel interviewing program at Western Electric, and his continuing collaboration with Gardner and Warner enabled him to provide the creative direction the program required, including training the program staff and interpreting survey results. Under Moore’s leadership and with Gardner’s aid, the staff developed high orders of analytical skill, and learned to use the questionnaire results much as a doctor uses tests as part of his procedure in diagnosing patient ills.

At one point, thought was given to the possibility of developing “a typology of the malfunctioning of organizations” which might be useful in diagnosing organization problems. This possibility was suggested by the frequency with which profiles of questionnaire and interview data tended to form patterns which began to grow familiar and the tendency for certain kinds of problems to occur in identifiable syndromes (Worthy, April 1950). Unfortunately, because of other research demands this possibility was not systematically pursed, although recognition of regularities definitely enhanced the skills of the interviewing teams.

The survey program was undertaken initially for the straight-forward purpose of finding how well employees liked their jobs and whether there were factors in the work situation causing dissatisfaction. The assumption was that once any such factors were discovered, necessary corrective action could be taken and all would be well. This assumption proved only partially valid, because in the new survey program we soon learned that many employee relations problems could not be understood, much less influenced, expect in terms of the broader context of the organization in which they had developed.

Gradually, the scope of the surveys was expanded to include the functioning of the unit (store, factory, etc.) as a whole and the entire plan pattern of technical processes and formal and informal relationships which compromise the unit. In recognition of this broader scope, the name of the program was changed from “morale surveys” to “organizations surveys,” and rather than simply discovering causes of employee dissatisfaction the purpose became to analyze strains or cleavages without operating units which might impede their functioning. In this context, determining levels of morale served chiefly as an aid in locating problem areas for further study by a field survey team (Worthy, January 1950).

A major early conclusion of the survey program was stated in a 1949 memorandum to Clarence Caldwell: “High morale is a by-product of sound organization. It is not a result that can be achieved by and for itself; above all, it is not a result of ‘being nice to people’ or plying them with favors. High morale is not something to be achieved at the expense of good operating results. The same policies, attitudes, and practices which are best calculated to produce good operating results are precisely the policies, attitudes, and practices which produce high level employee morale” (Worthy, 1949).

Findings of the Survey Program

Both internal and external factors were found to influence morale. Factors internal to Sears included job status, work pressures, tensions arising from differences in job goals and interests, tensions arising from hierarchal relationships, and disruptions caused by changes in management, company policies, job methods, and systems of rewards.

Significant influences on the quality of management-employee relations external to the company itself were also found, and while these were beyond the reach of management control it was important that management understand them. For example, the higher levels of morale generally prevalent among women employees as compared with men appeared to be explainable by differences in the social roles, expectations, and demands of the two sexes. (This was in the t 1940s; the comparisons are probably different today.) Other differences could be associated with social roles related to race and ethnicity. Levels of morale in retail stores were higher than those of warehouse and factories, apparently because of greater ideological agreement between salespeople and management. Closer ideological affinity also helped explain the fact that higher morale was found in the rural south than in the industrial north, and among employees with higher-rather than lower-ranked jobs. Morale tended to vary inversely by size of unit and size of city in which the unit was located. (For a typology of social factors, see Moore, 1950.)

At a broader level, we learned that human organizations tend to hold together to the extent that they have meaning to people in them, and ceremonials such as service anniversaries, retirement banquets, award presentations and the like provide opportunities for shared experience which help people identify with each other and with the organization. A great deal has been written about Sears’ famous profit-sharing plan, but its importance as a positive factor in employee relationships became better understood when its role was recognized as expressing the unity of interest between employees, management, and shareholders, in effect a unifying symbol about which the entire organization revolved.

Another valuable insight gained from the surveys was the fact that around any technology –store, mail order plant, factory, whatever –there develops a set of sub functions, departments, and jobs which determine to a great extent the kinds of relationships that will develop among employees in various activities. The importance of the job and job relationships in determining employee attitudes was particularly apparent to the survey team since jobs in the stores, while not uniform, were generally similar from unit to unit. Work on the selling floor, or the receiving room, or the auditing department varied little from one store to the next, and the survey team reached the point where it could anticipate how appliance salesmen, repair serviceman, office employees, and others would express themselves simply because of the nature of their jobs. This insight proved valuable in diagnosing organizational problems.

One of the most important things gained from working with Gardner and Warner and with the staff of the Committee on Human Relations, including William Foote Whyte who succeeded Gardner as executive director of the Committee, was the ability to explain complex human behavior to hard-nosed executives in terms they could readily understood. People who had learned to recognize the difference between a corner office and one down the hall readily grasped the concept of status symbols. The idea of social class and differences in class behavior made sense to them, and made it easier to explain some of the problems disclosed by the surveys.

In a field unrelated to employee behavior, I will never forget a meeting of Lloyd Warner with a group of senior merchandising executives in which he explained something that had bothered them for a long time: the fact that while Sear appliances enjoyed wide popularity among the rich and famous, even the wives of company executives were reluctant to buy their dresses from Sears. Warner pointed out that whereas a refrigerator or a power saw is relatively neutral in terms of social class identification, the clothes a woman wears are far more personal and an important way of expressing her idea of the place she considers herself to hold in society. The look of enlightenment that lit up faces in the audience was amusing to see.

Ancillary Values

In addition to its other values, because of the manner in which it was organized and administered the survey program developed into one of the company’s most effective general executives training tools. Care was taken from the outset to keep the program from being perceived as threatening, and in point of the fact the surveys were conducted primarily for the benefit of the manager or other executives in direct charge of the unit involved. Surveys were concerned not merely with identifying the existence and origin of difficulties; their primary purpose was problem-solving, and to this end great reliance was placed on the local executives themselves rather than on the survey team. The central objective was to give the responsible executives a better understanding of the way his organization was functioning, the various influences at work in his particular situation, and the ways these were impacting the attitudes and behavior of his people. At the same time, an effort was made to give the executive a more effective way of looking at his organization and its problem, with the expectation that with a better way of thinking about both he would be in a better position to take constructive action directed at the root of difficulties rather than at their superficial symptoms (Worthy, Gardner, and Whyte, 1946-47).

After a survey was completed and its results analyzed, the member of the personnel staff under whose direction it had been conducted would meet with the executive responsible for the unit and his staff. After reviewing the survey findings, the local executives would be questioned as to what interpretation might be put upon them, and a discussion would follow as to the best means for dealing with whatever problems had been uncovered. The survey “report” generally consisted of a memorandum to the manager saying, in effect, that “in our meeting we agreed that such and such things required attention and that such and such action would be taken.” The local executives were thus intimately example of the “case method” of teaching, with the considerable pedagogical advantage that the “case” was not some abstract problem but the company unit for which the executive himself was personally responsible. The learning results were impressive.

The program proved valuable in many other ways as well. It provided management with insight and understanding it could have gained in no other way of the actual state of affairs within the very large and widely-dispersed organization into which Sears by the late 1940s had grown. It produced a comprehensive body of factual material by which management was able to refine and improve personnel policies under the rapidly-changing conditions of the post-war period. In some ways, its most valuable contribution was the message it carried to employees of management’s genuine concern for running the business in ways that would serve their needs as well as the company’s.

The survey program developed into an invaluable administrative tool, and with modifications is still being used by Sears as a significant feature of that company’s widely-admired personnel system.

III. Social Research, Inc.

Sears’ work with the Committee on Human Relations in Industry went far beyond the survey program, and heavy demands were placed on the committee’s resources. We at Sears became increasingly impatient with the difficulties and delays of working through university channels. Our primary interest was in solving problems, and sometimes fast action was needed. But to start a project, it was necessary first to prepare a detailed written plan; which then had to be reviewed by university authorities and finally converted into a contract to be signed by appropriate officers of the university and the company. Weeks and sometimes longer were required to get projects under way, and while this might have been fast by university standard it was often exasperatingly slow by our needs. Early in 1947, therefore, I urged Gardner and Warner to establish an independent consulting firm with which we could deal. With Clarence Caldwell’s and General Wood’s ready approval, I offered them a contract that would underwrite the basic expenses of their new firm for its first two years of operation.

Thus was born Social Research, Inc., with Warner as chairman and Gardner as president. Warner continued for several years longer as professor at the University of Chicago, until moving to Michigan State in 1959 to introduce the study of social sciences into that university’s School of Business. Gardner served his formal connection with the university to devote himself full time to SRI, as the new firm quickly became known. He was succeeded as executive director of the committee by William Foote Whyte who served in that post until 1948 when he accepted a professorship at Cornell University.

With Whyte’s leaving, the Committee on Human Relations in Industry was disbanded. It had operated for a total of only five years, but those years had been remarkably productive, particularly in demonstrating the utility of applying the concepts of social anthropology to the practical problems of business and industry. And those concepts themselves were greatly enriched by exposure to modern forms of economic organization rather than to those of primitive societies which had previously been the discipline’s primary focus of attention.

Work with Social Research, Inc.

The work and influence of the committee did not die with it. Relieved of the burden of university red tape, n work with Gardner and Warner through Social Research, Inc. grew apace.

Some of the more significant of that work during the years 1947-52 included studies of special problem groups such as low status employees, commission salesmen, warehouse workers, and veterans returning from military service. A study of relationships between the company’s sources of supply and its buying organization was particularly useful, as was an analysis of internal relationships within the buying organization itself. A study of the social class background of upwardly mobile employees proved useful in refining the company’s recruitment and executive development policies.

In the course of several studies on a variety of subjects, attention was given to factors shaping the size and structure of administrative units within larger components of organization, and the influence of such factors on decision making processes (Worthy, 1952). All told, an extensive number and variety of projects were undertaken in cooperation with Gardner and Social Research, Inc. for the purpose of improving the effectiveness of the company’s organizational and managerial practices and for enhancing the quality of management-employee relationships.

While some of the results of this work were published, most of it was not. Unlike studies conducted under university auspices, which typically have publication of some kind in view, studies conducted in collaboration with SRI were basically done for administrative purposes, that is, to provide officers and executives with information and proposals that would be helpful in the discharge of their managerial responsibilities; many “reports” were simply in the form of memoranda. When a study had served the particular purpose for which it was made, it was usually filed away and attention turned to other more immediately pressing projects. Sponsored work with the Committee on Human Relations was essentially “scholarly” while that done with SRI tend to be much more along “consulting” lines. Fortunately, some of the more important results of this work found its way into Moore’s doctoral dissertation (Moore, 1950) and to that extent has a measure of permanence it would not otherwise enjoy.

IV. The Sears “X-Y Study”

In retrospect, one aspect of the collaborative research conducted by Sears and SRI during the late 1940s is of particular interest and significance. This was the study of the reciprocal effects of organizational structure on manage and worker behavior and on the dynamics and quality of working relationships within organizations. Because this study has been alluded to from time to time in various connections but has never been fully reported before, it merits presentation here in salient part (Worthy, 1959,1984).

In the spring of 1949 a sharp recession hit the retail industry, and General Wood with his customary decisiveness moved quickly to deal with the resulting problems. As one phase of this effort, the corporate personnel department was directed to investigate how representative store managers were coping with the problems of falling sales and profits and what could be learned from their experience. Other appropriate action, of course, was taken along lines of merchandising and expense control, but the personnel department was asked to make its own study, paying particular attention to staffing and organizational matters. The machinery of the employee survey program, by this time operating smoothly and the staff well-skilled in field research techniques, proved a great value for purposes of fact-finding and analysis. The resources of Social Research, Inc., and particularly the counsel of Gardner, were also called upon.

This study is described in detail in a manuscript written by me in 1952 and expanded in 1953, but because of the pressures of other work never completed for publication. I have a copy in my personal files, and another is on file in Sears Archives. An early draft was loaned to William Foote Whyte and summarized by him in his Modern Methods in Social Research (Whyte, 1952), and again in Man and Organization (Whyte,1959).

“Tall” vs. “Flat” Organization Structures

Studies previous to this had given indication of the superiority in many respects of relatively simple as compared with more complex forms of organization. It was Gardner who first called attention to the fact that Seas often violated the dictum that spans of management control should be kept as short as possible. For example, rather that the five or six subordinates per superior generally considered in the literature as the desirable limit, Sears had at this time nearly fifty senior merchandising executives reporting to a single merchandising vice president, with no intervening level of supervision; contrary to standard theory, the seemed to work very well.  Similar broad spans of control existed in other parts of the Sear organization, including some of the retail stores. Gardner coined the term “flat” and “tall” organizations to characterize the two types of structure.

Under General Wood’s highly decentralized plan of management, executives at each level were given wide latitude as to the manner in which they set up their organizations. For many years, Wood forbade the use of organization charts for the stores, saying to us in the personnel department that “If you publish a chart, managers will think what’s the way you want them to organize their stores. But I want them to organize their stores in the way they feel they can get the best results, and they can tell that better than you can.” After repeated request, Wood finally gave me permission to prepare organization charts to give to professors and other outside the company who asked for them, but with strict admonition that they were to have no circulation within the company.

Because of this latitude given store managers, many otherwise similar stores were organized differently: some, for example, were “tall,” while others were “flat.” The study on which the personnel department and SRI was now embarked undertook as its first objective an examination to determine whether either of the two types of organization was superior to the other, and if so in what ways.

“X” Stores and “Y” Stores

Two groups of stores, carefully matched for comparability in all respects but one, were selected for study. Sears retail stores were classified as “A,” “B,” and “C,” depending on size and lines of merchandise carried. All stores included in the study were medium-size “B” stores located in the same general geographic area and in communities of approximately similar size and economic base. They carried the same lines of merchandise, and were roughly comparable in number of employees and volume of business. They differed in only one important respect: structure of organization. Because all were “B” stores, for convenience of the study those with “flat” structures were designated “X” "and those with “tall” structures “Y.”

Stores in the “X” group had a minimum number of staff and supervisory employee and simple structures. There was a store manager at the top, a single assistant manager, and approximately 30 managers of merchandise departments –“divisions,” in Sears’ terminology. There were, in addition, the usual non-selling or “service” activities such as shipping and receiving, customer service, maintenance, unit control, etc. in theory, all heads of the merchandise departments reported directly to the store manager, with the assistant manager usually responsible for the service functions. In practice, the store and the assistant managers usually worked together as a team, dividing responsibilities on a more or less informal, ad hoc basis. There was no level of supervision between the two at the top and approximately thirty-five heads of selling non-selling activities.

Stores in the “Y” group were set up on a more hierarchical basis. Instead of all functions reporting directly to a two-management team, the organization was broken down into a series of sections. Instead of a single assistant manager, there was likely to be hard lines merchandise manager supervising the hard lines departments, a soft lines merchandise manager supervising soft lines, and an operating assistant supervising service activities. In some cases there was also a personnel manager, who typically reported to the operating assistant.

In separating out these two groups of stores for purposes of study, the investigators had been preparing to find differences between them. They did, in fact, find differences in payroll costs and profits as percentage of sales –in favor, as expected, of the more simply organized “X” group with fewer higher-paid supervisors. But they found other important differences as well.

Differences between “X” and “Y” Stores

For one thing, the department managers in the “X” stores tended to be more skillful merchants than those in the “Y” stores, a trait apparent not only from personal interview but from the merchandising results of their departments. Morale as measured by attitude surveys was higher in the “X” than in the “Y” stores, and this was especially true at the department manager level. Finally, a review of the records of the two groups going back over a period of several years disclosed that larger numbers of people had been promoted out of the “X” than the “Y” stores to higher levels of responsibility. Differences such as these between the two types of structures had not been expected, and a search began for explanations.

To the team of investigators made up of men and women experienced in the methods of field research, the answers were not difficult to find. For one thing, in the “X” stores, the manager and his assistant –the “management team” –were spread very thin. The number of department heads reporting to them was such that the amount of time they could devote to any one of them, on average, was small. Under these circumstances, the only possible way for them to accomplish their won jobs ways to have high quality people heading their departments. For example, the only way a store manager could be sure of having a good hardware business was to find –either from within the organization or elsewhere –a good hardware merchant; and so on through all the other departments. Precisely because they themselves were spread so thin, the manager and his assistant were compelled to seek out people of initiative and ability, people who could take responsibility, who could move ahead on their own without having to clear everything in advance. By and large, the department managers in the “X” type stores measured up to these specifications and their results showed it.

The situation tended to be rather different in the “Y” stores. Here, people at the merchandise manager and operating assistant level tended to be of fairly high caliber, but those at the department manager level tended to compare rather less well with their opposite numbers in the other group of stores. The investigators found that managers with specialized assistants tended to look for results and were not under the same compulsion to build high levels of competence at the individual department level. The differences of caliber of department manager personnel were thus the direct consequence of differences in organization structure and of the styles of management the structures tended to call for.

But the explanation for the differences in capabilities did not stop here. In addition to pressures on the managers of the “X” stores to staff their departments with superior talent, those placed in department manager positions in such stores were exposed to a kind of experience that tended to develop their capacities in a particularly favorable manner. Because the store manager and his assistant could give only a limited amount of time, on average, to each department, the heads of that department were forced to assume responsibility. To a rather remarkable extent, they had to make their own decisions on running their departments and stand or fall by the results.

The task of their counterparts in the “Y” stores was considerably less demanding. The head of the hardware department submitted his merchandising plans to the hard lines merchandise manager, and because this individual would be held responsible by his store manager he typically would review the plans in great detail, make numerous changes as his judgment dictated, and the final plan would be his and not the department head’s. The learning environment in these stores was much less favorable than in the other.

Close observation of the way the two types of stores functioned indicated that one of the definite disadvantages of the more complex organization was that it safeguarded people too closely against making mistakes. It was difficult for department managers to get far off base or to stay off base very long. But precisely for this reason, they were deprived of one of their most valuable means of learning. Much of the superiority of department manager personnel found in the stores with the flat organization structures was clearly attributable to the fact that the nature of the structure within which they worked tended to force them to accept responsibility, to exercise initiative, to develop self-confidence, and to learn from experience.

Structure and Behavior

Note has already been taken of the larger number of promotable people originating in the “X” stores. This was due in part to the greater necessity in such stores for picking high quality personnel and the more stimulating environment for personal growth in which they worked. But beyond these factors were marked differences in the styles of leadership typically found in the stores.

Because manager and assistant in the more simply-organized stores did not have time to supervise in detail, their leadership tended to be chiefly in terms of setting goals, leaving much of the task of determining the means for achieving them to the department mangers themselves. In stores with the taller, more hierarchical organizations, the mangers and their intermediate staffs generally not only set the goals for the various departments, but worked out in considerable detail the manner in which the goals were to be accomplished. Much more than those in the “X” stores, department managers in d “Y” stores had the task of carrying out plans others had made and striving for goals they had little part in setting. Leadership in the “X” stores tended to be “goal oriented,” whereas that in the “Y” stores was primarily “means oriented.”

The two styles could be characterized with equal validity as “people oriented” and “system oriented.” In the “X” stores, the very nature of the organization structure tended to force managers to concern themselves primarlily with people. Store problems had to be looked upon essentially as problems of people rather than problems of things. If the performance of a department was unsatisfactory, the way to improve it was to strengthen the management of the department, for only in this way could the improvement be lasting. This kind of organization structure not only place a premium on selecting well-qualified personnel in the first place; it also placed pressures on the manager and his assistant to work with people as their individual needs required to bring them to levels of competence where they could be entrusted with larger measures of personal responsibility.

Most emphatically, this type of structure did not involve blind delegation. Managers had to have good judgment as to which individual could move ahead largely on their own and which required help. A considerable part of the time of the manager and his assistant was devoted to the newer department managers or to those whose performance was not up to par or who otherwise needed assistance. Throughout, the emphasis was on building individual competence, bringing people along to the point where they could run their departments with a minimum of direction and supervision.

In the “Y” stores, on the other hand, the emphasis was quite clearly on the manager and his immediate staff doing the thinking and planning and depending on close supervision and essentially mechanical controls to ensure proper execution at the department manager level. The intervening layer of supervision, moreover, tended to deprive department managers of direct contact with the store manager, which in the “X” stores was an important factor in their development.

A Second Look

The findings of the study to this point were fairly clear; equally clear, apparently was the course of action the company ought to take. If the differences in the performance of the two types of stores were related so directly to differences in organization structure, the logical step seemed to be converting the “Y” stores to the “X” structure. But in part because of General Wood’s well-known reluctance to interfere with the way the managers ran their stores, those conducting the study thought it wise to check further.

The stores in the two groups were revisited, and this part time particular attention was paid to the personalities and ways of behaving of the store managers themselves. Although each store manager was very much an individual with personal characteristics of all his own, managers of the two groups of stores tended to group themselves roughly into two fairly distinct personality types. The differences between them were particularly apparent in their genera attitude toward employee as revealed in the way they talked about their people and in their behavior in face-to-face relationships.

Managerial Personalities

Managers of the “X” stores generally took a great deal of pride in their employees. Walking through the store with managers in this group, they were likely to pause frequently to introduce the visitor from the national office to department managers, salespeople, and others, making complimentary remarks about their performance or relating some recent incident of merit. They tended to take special pride in their promotable younger people, and would often recount their history, their special qualifications, what was being done to further their training, how soon they would be ready for promotion, etc. these managers were not indiscriminate in their praise. They held their people to high standards and were critical of those who failed to measure up. They were usually good judges of people, able to evaluate the strengths and weakness of subordinates and to deal with them accordingly. In part because of their skill on this store, they usually did have people in whom they could take pride.

Managers of the “Y” stores tended to be rather different. They had much less of the warm, outgoing orientation of the “X” men. They were often distrustful of people, and at some time during the store visit they were likely to make remarks to some such general effect as: “ It’s hard to get good people these days,” or “They’ve been spoiled by the government and the unions,” or “There’s something wrong with the education system –people just don’t believe in doing a fair day’s work anymore, they’re out for all they can get and give as little as they can get away with, “ or “Young people have lost their ambition; you don’t see many of them these days who are willing to work to get ahead the way we had to.” These mangers often seemed to expect the worst of their people, and not infrequently their fears were justified. They felt that people had to be watched, that their work had to be checked closely –else no telling what might happen.

This sketch of the two types of managers has been deliberately overdrawn to emphasize their differing orientations. But there was a definite difference in their general outlook on life and in their basic attitudes toward people. This difference was reflected in the kinds of organizations they set up and the ways they related themselves to the people in them.

Personalities and Structures

As has already been noted, Sear under General Wood in the 1930s and 1940s was administered with a remarkably high degree of decentralization. Individual store managers were vested with wide latitude and discretion, and this extended, within limits, to the manner in which they set up their organizations. As the study of the two groups of stores proceeded, it became clear that the two fairly different types of managers had tended to create the kinds of organizations with which they could work comfortably. The “Y” manager, who tended to have little confidence in their people, felt that employees had to do the “real thinking,” and that close controls had to be imposed to make sure that those further down the line followed through as they were instructed.

Managers of the “X” stores, on the contrary, tended to have more confidence in the capacity of their people, and enough skill in evaluating them to be rather sure in their judgments as in whom and how far they could place their confidence. They relinquished none of their responsibility for guidance and direction or for final results, but they sought to capitalize on the initiative and good sense of their subordinates rather than trying to do the “real thinking” for them. Their primary method of solving problems was to work with the people involved, to the end of solving not only the immediate problem but to strengthen the ability of their subordinates to deal with other problems in the future. To this end, they liked to work directly with their department managers rather than through an intermediate staff which they were likely to feel would only get in the way. The “X” managers no less than the “Y” had tended to create the kinds of organizations with which they could work comfortably.

Viewed in this light, the task of changing organization structures loomed as by no means the simple matter it had first appeared. With all the apparent superiority of the “X” structure, there was a real question as to whether the “Y” type manager could function effectively within it. This doubt was reinforced by what often happened following store manager changes. It was not uncommon to fund that a manager transferred from a “Y” store to one organized on an “X” basis would soon begin reconsisting the organization along lines more nearly like that to which he was accustomed.

These changes were usually not begun immediately nor, once begun, accomplished rapidly, but the direction was often unmistakable. Always, of course, the changes were quite “logical” –inventory problems, the need for building up, say, the soft line departments, the scarcity of high caliber people for department manager jobs, and so forth. Whatever justification there might have been for the moves, the end result tended to be an organization structure of the manager’s new store not too different from that he had left (assuming, of course, that the two stores were approximately the same size and type).

The reverse process was equally pronounced. An “X” manager transferred to a “Y”-type store (again assuming reasonable comparability) was likely before long to report to his regional office that as he had become familiar with his new store he had found he had more merchandise managers than he needed. Sometime later, he might report that as soon as he could make certain moves to strengthen some of his department managers he would probably have another merchandise manager available for transfer. Here too, the end result was likely to be conversion of the organization of the manager’s new store to correspond more closely with that of his old.

The task of changing organization structures thus appeared to be considerably more complicated than issuing directives or re-drawing organization charts. In the light of the study it was clear that, given some degree of latitude, the way an organization is structured is likely to reflect the personality and temperament of the key people in it –above all, the personality and temperament of the person at the top.

For Sears’ purposes, the results of the study were inconclusive. It appeared unwise to try to change organization structures arbitrarily, without regard to the styles and personalities of incumbent managers. On the other hand, it was obvious that personality changes of any magnitude in managers are extremely difficult to bring about and likely to require the equivalent of psychoanalysis or religious conversion –both well beyond normal administrative capabilities. The chief lesson Sears drew from the study was recognition of the importance of selecting people for key positions who exhibited “X” rather than “Y” characteristics.

V. Conclusion

This study was in effect the climax of the collaboration between Sears, Roebuck and Social Research, Inc. as successor to the Committee on Human Relations in Industry. Social Research continued under Gardener’s leadership to operate as a successful consulting firm until 1985, but did little further work of importance for Sears in the areas of organization structure and employee behavior. To a very large extent, the close collaboration that had characterized the 1940s had depended on the personal relationships between Gardner, Moore, and me, and while the three of us to this day enjoy each other’s friendship we have had little opportunity to work together for many years.

Moore left Sears in 1950 to complete his doctorate at the University of Chicago, and went on from there to a distinguished academic career. At the beginning of 1953 I joined the Eisenhower Administration as Assistant Secretary o Commerce. By this time, Clarence Caldwell’s health was falling, General Wood was soon to retire, and there was a change in the leadership of the personnel function. In light of the results of the study just described, it came as no surprise that important changes in managerial style ensued. The new style was effective in many ways, and served Sears well for many years thereafter. But it left no room for the kind of objective research in organizational behavior which had characterized the preceding ten years. It had been a productive ten year, and perhaps had gone as far as it was practical to go.

Analysis

a.   Why would external factors influence employee morale as found in the Sears survey programs?

b.  What was the value of the Sears profit-sharing program to the employees? To the company?

c.   Explain the differences between “X” and “Y” stores. Relate them to the notion of tall and flat organization structures.

d.  Why were managers most likely to be promoted from “X” store than from “Y” stores?

e.  How could the organization structure in “X” and “Y” store? Could a “Y” manager be as effective in an “X” store? Why?